Treatise, Book 1

David Hume

i: Ideas

Part i: Ideas, their origin, composition, connection, abstraction, etc.
1: The origin of our ideas
special name to mark the difference.1 [In this work, ‘name’ is

All the perceptions of the human mind fall into two distinct
kinds, which I shall call ‘impressions’ and ‘ideas’. These
differ in the degrees of force and liveliness with which they
strike upon the mind and make their way into our thought
or consciousness. The perceptions that enter with most force
and violence we may name ‘impressions’; and under this
name I bring all our sensations, passions, and emotions,
as they make their first appearance in the soul [= ‘mind’; no
religious implications]. By ‘ideas’ I mean the faint images of
the others in thinking and reasoning: for example, all the
perceptions aroused by your reading this book—apart from
perceptions arising from sight and touch, and apart from
the immediate pleasure or uneasiness your reading may
cause in you. I don’t think I need to say much to explain
this distinction: everyone will readily perceive for himself
the difference between feeling (·impressions·) and thinking
(·ideas·). The usual degrees ·of intensity· of these are easily distinguished, though there may be particular instances
where they come close to one another. Thus, in sleep, in a
fever, in madness, or in any very violent emotions of soul,
our ideas may become like our impressions; as on the other
hand it sometimes happens that our impressions are so faint
and low that we can’t distinguish them from our ideas. But
although ideas and impressions are fairly similar in a few
cases, they are in general so very different that no-one can
hesitate to classify them as different and to give to each a
1

often used to cover not only proper names but also general terms such
as ‘idea’.]

Another division of our perceptions should be noted; this
one cuts across the line between impressions and ideas. It is
the division into simple and complex. Simple perceptions—
that is, simple impressions and ideas—are ones that don’t
allow any distinction or separation ·among their parts·. Complex perceptions, on the contrary, can be distinguished into
parts. Though a particular colour, taste, and smell, are qualities all united together in this apple, it’s easy to perceive that
they aren’t the same as one another and can least be distinguished from each other—·and so one’s total perception of
the apple is complex ·.
Having through these divisions ordered and arranged our
subject-matter (·perceptions·), we can now set ourselves to
consider more accurately their qualities and relations. The
first fact that springs to my attention is that our impressions
greatly resemble our ideas in every respect except their degree of force and liveliness. Perceptions of one kind seem
to be, in a way, reflections of perceptions of the other kind;
so that all the perceptions of the mind do double duty, appearing both as impressions and as ideas. When I shut
my eyes and think of my study, the ideas I form are exact
representations of the impressions I felt ·when I was in my
study·; every detail in one is to be found in the other. And I

I hope you will allow me to use the words ‘impression’ and ‘idea’ in senses different from their usual ones. Perhaps, indeed, I am restoring ‘idea’ to
its original sense, from which Mr Locke has perverted it by making it stand for all our perceptions. By the term ‘impression’ I don’t mean anything
about how our lively perceptions are produced in the soul; I merely label the perceptions themselves; and for this I don’t know any particular name,
in English or any other language.

1

Treatise, Book 1

David Hume

find the same resemblance and representation when I survey
my other perceptions: ideas and impressions seem always
to correspond to each other. This remarkable fact holds my
attention for a moment.
Surveying the field more accurately, I find I have been
swept along by how things first appeared to me, and that I
must—with help from the simple/complex distinction—limit
this general thesis that all our ideas and impressions are
resembling. I observe that •many of our complex ideas never
had impressions that corresponded to them: I can imagine a
city such as the New Jerusalem, with golden pavements and
ruby walls, though I never saw such a thing. And I observe
that •many of our complex impressions are never exactly
copied by ideas: I have seen Paris, but I can’t form an idea of
that city that perfectly represents all its streets and houses
in all their detail.
So I perceive that although there is in general a great
resemblance between our •complex impressions and ideas,
it is not true across the board that they are exact copies of
each other. Now let us consider how the case stands with
our •simple perceptions. After the most accurate examination I am capable of, I venture to say that here the rule
holds without exception: that every simple idea has a simple
impression that resembles it, and every simple impression
has a corresponding idea. The idea of red that we form in
the dark differs only in •degree ·of intensity·, not in •nature,
from the impression ·of red· that strikes our eyes in sunshine. You can satisfy yourself that I am right about this by
going over as many of your simple impressions and ideas as
you like; it’s impossible to prove my point by going over all of
them! But if anyone should deny this universal resemblance
·between simple impressions and simple ideas·, I don’t know
how to convince him except by asking him to show •a simple
impression that doesn’t have a corresponding idea, or •a sim-

i: Ideas

ple idea that has no corresponding impression. If he doesn’t
answer this challenge—and it’s certain that he can’t—then
his silence and our own observation will suffice to establish
our conclusion.
Thus we find that all simple ideas and impressions resemble each other; and as the complex are formed from
simple ones we can say generally that these two sorts of perception exactly correspond. Having uncovered this relation,
which requires no further examination, I am curious to find
some of the other qualities ·of impressions and ideas·. Let
us consider what brings them into existence: as between
impressions and ideas, which are causes and which are
effects?
The full examination of this question is the subject of this
book; so I shall here content myself with establishing one
general proposition:
All our simple ideas, when they first appear,
are derived from simple impressions which correspond to them and which they exactly represent.
In looking for phenomena to support this proposition, I can
find only two kinds; but the phenomena of each kind are
obvious, numerous, and conclusive.
·As a preliminary to the first kind of phenomenon·, I first
go over again in my mind, and make myself certain, of the
proposition that I have already asserted, that every simple
impression is •attended with a corresponding idea, and every
simple idea is •attended with a corresponding impression.
From this •constant conjunction of resembling perceptions
I immediately conclude that there is a great connection between our corresponding impressions and ideas, and that
the existence of the one has a considerable influence on the
existence of the other. Such a constant conjunction in such
an infinite number of instances can’t arise from chance, but
clearly proves a dependence of the impressions on the ideas
2

Treatise, Book 1

David Hume

or of the ideas on the impressions. Wanting to know which
way the dependence runs, I consider the order in which
these ·simple impressions and ideas· first appear; and I find
by constant experience that the simple impressions always
come first—it is never the other way around. To give a child
an idea of scarlet or orange, of sweet or bitter, I present
objects ·that are that colour or taste·—that is, I give him
those impressions. I don’t do anything as absurd as trying
to give the child the impression by arousing in him the idea!
When our ideas occur they don’t produce the corresponding
impressions; we don’t see any colour or feel any sensation
merely by thinking of them. On the other hand we find that
every impression— whether of mind or body—is followed by
an idea that resembles it in every way except its degree of
force and liveliness. The •constant conjunction of our resembling perceptions is a convincing proof that the one are the
•causes of the other; and the fact that the impression always
comes first is an equal proof that impressions are the causes
of our ideas, not vice versa.
This is confirmed by another plain and convincing phenomenon, namely: whenever someone happens to lack the
faculty that gives rise to impressions of some kind—e.g. when
someone is born blind or deaf—he lacks not only impressions
of that kind but also the corresponding ideas; so that his
mind never shows the least traces of either of them. This
holds not only where the relevant organs of sensation are
entirely destroyed, but also when they haven’t yet been put
into action to produce a particular impression; we can’t form
an accurate idea of the taste of a pineapple without having
actually tasted it.
But there is one phenomenon that goes the other way,
and may prove that it is not absolutely impossible for ideas to
occur in advance of their corresponding impressions. I think
you’ll agree that the various ideas of colours that enter by the

i: Ideas

eyes are really different from each other, though there are
resemblances amongst them; similarly for ideas of sounds
that are conveyed by the ·sense of· hearing. If this is true
of •different colours, it must equally hold for the •different
shades of the same colour that each of them produces a
distinct idea that is independent of the others. (If not, then
it is possible by the continual gradation of shades to run a
colour imperceptibly into what is most remote from it. ·We
can create a sequence of colours, each barely perceptibly
different from its neighbours, with some colour at the start
of the sequence and a totally different one at the end·. If you
won’t allow any of the intervening pairs of neighbours to be
different, you can’t without absurdity say that the colours at
the ends of the sequence are different—·which they patently
are·.) Now take the case of someone who has had the use of
his eyesight for thirty years, and has become perfectly well
acquainted with colours of all kinds except for one particular
shade of blue, which he happens never to have encountered.
Let all the different shades of blue except that single one be
placed before him, descending gradually from the deepest
to the lightest. Obviously, he will perceive a blank ·in the
sequence· where that shade is missing, and will be aware
that the qualitative gap between neighbours is greater at
that place than anywhere else in the sequence. Now I ask:
Can he fill this gap from his own imagination, raising
up in his mind the idea of that particular shade, even
though ·an impression of· it had never been conveyed
to him by his senses?
I think most people will agree that he can; and this may
serve as a proof that simple ideas are not always derived
from corresponding impressions. But this instance is so
particular and singular [those are Hume’s adjectives] that it is
hardly worth noticing, and isn’t enough on its own to require
us to alter our general maxim.
3

Treatise, Book 1

David Hume

But I ought to mention that the principle that impressions
come before ideas is subject not only to the exception (·about
the missing shade of blue·) that I have just sketched but also
to another limitation, namely: just as our ideas are images [=
‘copies’] of our impressions, so we can form secondary ideas
that are images of primary ones; and my own theory allows
for this. This is not strictly speaking an exception to the rule
·that impressions come first·, but rather an explanation of it.
Ideas produce the images of themselves in new ·secondary·
ideas; but as the first ·or primary· ideas are derived from impressions, it still remains true that all our simple ideas come
from their corresponding impressions—either immediately or
·as secondary ideas· through the mediation of primary ideas.
This, then, is the first principle I establish in the science
of human nature. Don’t despise it because it looks simple. It
is a remarkable fact that the present question about •which

i: Ideas

comes first, impressions or ideas, is the very one that has
created so much noise when expressed as the question of
•whether there are any innate ideas, or whether all ideas
are derived from sensation and reflection. Notice that when
philosophers want to show the ideas of extension and colour
not to be innate, all they do is to show that those ideas are
conveyed by our senses. To show that the ideas of passion
and desire are not innate they observe that we have a prior
experience of these emotions in ourselves. Now, if we carefully examine these arguments we shall find that they prove
only that ideas are preceded by other more lively perceptions,
from which they are derived and which they represent. I
hope this clear statement of the question will remove all
disputes about it, and will render this principle of more use
in our reasonings than it seems to have been up to now.

2: Division of the subject
Since it appears that our simple impressions come before
their corresponding ideas, and that the exceptions to this
are very rare, it seems that the methodical procedure would
be to examine our impressions before turning to our ideas.
Impressions can be divided into two kinds, those of sensation and those of reflection. •Impressions of sensation arise
in the soul itself, from unknown causes. •Impressions of
reflection are largely derived from our ideas, in the following
way. An impression first strikes on the senses and makes
us perceive heat or cold, thirst or hunger, pleasure or pain,
of some kind or other. Of this impression the mind makes
a copy which remains after the impression ceases; and we
call this copy an ‘idea’. When this idea of pleasure or pain
recurs in the soul, it produces new impressions of desire

and aversion, hope and fear, which may properly be called
‘impressions of reflection’ because they are derived from reflection. These impressions are in turn copied by the memory
and imagination and become ·sources of· ideas, which in
their turn may give rise to yet other impressions and ideas.
Thus the impressions of reflection come before their corresponding ideas but come after impressions of sensation
and are derived from them. The study of our sensations
belongs more to anatomists and natural philosophers than
to moral philosophers [= ‘belongs more to anatomists and natural
scientists than to philosophers and scientists interested in the human
condition’];

so I shan’t go into it here. And as the impressions
of reflection—that is, the passions, desires, and emotions—
that mainly deserve our attention arise mostly from ideas,
4

Treatise, Book 1

David Hume

we must reverse the method that seems most natural at first
sight: in explaining the nature and principles of the human
mind, we must deal in detail with ideas before we proceed to
impressions. That is why I have chosen to begin with ideas.

i: Ideas

in his day, namely that of ‘source’, ‘cause’, ‘drive’, ‘mechanism’ or the
like. From now on, every occurrence of the word in that sense will be
written as ‘principlec , suggesting ‘principle = cause’. A ‘principle’ without the subscript is a proposition, usually a premise but sometimes a

[Important note: Most of Hume’s uses of ‘principle’ in Treatise I, in-

conclusion.]

cluding the one we have just met, give it a meaning that it often had

3: Memory and imagination
We find by experience that when an impression has been
present to the mind, it re-appears there later as an idea;
and it can do this in either of two ways: •when in its new
appearance it retains a good deal of its first liveliness and is
intermediate between an impression and an idea; or •when it
entirely loses that liveliness and is a perfect idea. The faculty
by which we repeat our impressions in the first manner is
called the ‘memory’, and the other the ‘imagination’. You
can see at a glance that the ideas of the memory are much
livelier and stronger than those of the imagination, and that
the memory paints its objects in sharper colours than the
imagination uses. When we remember a past event, the idea
of it flows in on the mind in a forcible manner; whereas in
the imagination the perception is faint and languid, and the
mind can’t easily keep it steady and uniform for any considerable time. Here, then, is a noticeable difference between
one species of ideas and another. But of this more fully
hereafter, in I.iii.5.

der and form as the original impressions had, whereas the
memory is in a way tied down in that respect, without any
power of variation.
It is evident that the memory preserves the form in which
its objects were originally presented, and that when we depart from that form in recollecting something, this comes
from some defect or imperfection in that faculty. An historian may find it more convenient to relate one event before
another which in fact occurred before it, but then, if he is
careful, he comments on this re-ordering, and thereby puts
the relevant ideas back in their right order. Similarly with
our recollection of places and persons with which we were
formerly acquainted: the chief exercise of the memory is not
to preserve the simple ideas, but to preserve their order and
·temporal· position. In short, this principle is supported by
so many common and everyday phenomena that I needn’t
trouble to insist on it any further.
Equally evident is the second principle, of the imagination’s liberty to transpose and change its ideas. The fables
we meet with in poems and romances put this quite beyond
doubt. Nature there is totally confounded, with stories full
of winged horses, fiery dragons, and monstrous giants. This
liberty of the fancy [= ‘imagination’] won’t appear strange when
we consider that all our ideas are copied from our impres-

Another, equally obvious, difference between these two
kinds of ideas is this: though neither the ideas of the memory nor those of imagination— neither the lively ideas nor
the faint ones—can appear in the mind unless their corresponding impressions have gone before to prepare the way
for them, the imagination isn’t bound to keep the same or5

Treatise, Book 1

David Hume

sions, and that no two impressions are perfectly inseparable;
·so that there are no constraints on how freely ideas may be
assembled and re-arranged·. Not to mention that this is an
evident consequence of the division of ideas into simple and

i: Ideas

complex. Wherever the imagination perceives a difference
among ideas it can easily produce a separation.

4: Association of ideas
place, and •cause and effect. [Hume calls the three relations ‘qualities’, but only in this paragraph.] I don’t think I have much need
to show that these ·three· relations produce an association
between ideas such that when one appears it is naturally
followed by another. It is plainly the case that in the course
of our thinking and in the constant turn-over of our ideas
our imagination runs easily from one idea to any other that
•resembles it, and that this quality alone is for the imagination a sufficient bond and association. It is likewise evident
that as the senses have to move from object to object in a
regular manner, taking them as they lie •contiguous to each
other, so the imagination also must become accustomed to
following the same pattern in its thinking, and run along
the parts of space and time in conceiving its objects. The
•relation of cause and effect will be thoroughly examined
later, so I shan’t say much now about its role in creating
associations of ideas. I merely say that there is no relation
that produces a stronger connection in the fancy, and makes
one idea more readily recall another, than the relation of
cause and effect between their objects.

The imagination, then, can separate ideas and then re-unite
them in whatever form it pleases; so its operations would
be perfectly inexplicable—·a meaningless jumble of random
events·—if it weren’t guided by some universal principlesc
that give some uniformity to its doings at different times. If
ideas were entirely loose and unconnected, they would be
joined ·in the imagination· purely by chance; and in that
case it couldn’t happen—as in fact it often does—that the
same simple ideas regularly came together into complex ones.
For that there needs to be some •bond of union among them,
some •associating quality by which one idea naturally introduces another. This •uniting principlec among ideas is not an
unbreakable connection, for that has been already excluded
from the imagination; nor should we conclude that the mind
cannot join two ideas without this uniting principlec , for
nothing is more free than the imagination, ·which can join
any two ideas it pleases·. We should regard the uniting
principlec only as •a gentle force that •usually dominates,
·not as an •irresistibly strong one that •always dominates·.
Among the things it explains is the fact that languages so
nearly correspond to one another: it is because Nature has
(in a way) pointed out to everyone the simple ideas that are
most suitable for being united into a complex one.

To understand the full extent of these ·three· relations,
we must grasp that two objects are connected together in the
imagination not only when •one is immediately related to the
other by resemblance or contiguity or cause-effect, but also
when •a third object comes between them and is related in
one of the ways to them both. This can be carried on to a

The relations that give rise to this association ·of ideas·,
in this way carrying the mind from one idea to another, are
these three: •resemblance, •contiguity [= ‘nextness’] in time or
6

Treatise, Book 1

David Hume

great length, though each lengthening of the chain considerably weakens ·the association of ideas that comes from·
the relation. Fourth-cousins are connected by causation (if I
may so express myself), but not as closely as brothers are,
let alone children and their parents. In general all the bloodrelationships depend on cause and effect, and are regarded
as close or distant according to how many connecting causes
are interposed between the persons.
Of the three relations above mentioned, causation is the
most extensive. Two objects can be considered as related
by it not only when one is the cause of •the existence of
the other but also when one is the cause of •some action
or motion of the other. . . . This line of thought goes further:
two objects are connected by the cause- effect relation not
only when one •does produce a motion or action in the other
but also when it •has the power to produce it. This, we can
see, is the source of all the relations of ·self·-interest and
duty by which men influence each other in society, leading
to some being governors and others subordinates. A master
is someone whose situation gives him—whether by force or
by prior agreement—a power of directing some of the actions
of someone else whom we call servant. A judge is one who
can by his opinion settle questions of ownership that are disputed between members of the society. When someone has a
power, all that is needed to turn it into action is the exercise
of his will; and his exercising it is in every case regarded as
possible, and in many cases as probable—especially in ones
where there is authority, where the subject’s obedience will
bring pleasure and advantage to the master.

i: Ideas

These, then, are the principlesc of union or cohesion
among our simple ideas, providing in our imagination a substitute for the tighter links that ideas have in our memory.
This ·association of ideas· is a kind of attraction, which in the
mental world will be found to have effects as extraordinary
as ·those of attraction or gravity· in the physical world, and
to show itself in as many and as various forms. Its effects
are everywhere conspicuous; but its causes are mostly unknown, and must be assigned to basic qualities of human
nature that I don’t claim to explain. What a true philosopher
needs most is to restrain the immoderate desire to search
into causes; and, having established a doctrine on the basis
of a sufficient number of experiments, to rest content with
that when he sees that a further enquiry ·into its causes·
would lead him into obscure and uncertain speculations.
In that case he would spend his time and energy better in
examining the •effects of his principle than in examining its
•causes.
Among the effects of this union or association of ideas,
none are more remarkable than the complex ideas that are
the common subjects of our thoughts and reasoning, and
that generally arise from some principlec of union among
our simple ideas. These complex ideas can be divided into
•relations, •modes, and •substances. I shall briefly examine
these in order; and shall ·in section 7· add some considerations about our general and particular ideas. That will bring
me to the end of ·Part i and of· the present subject, which
can be considered as the elements of this philosophy.

7

Treatise, Book 1

David Hume

i: Ideas

5: Relations
I. The first is •resemblance. This is a relation without
which no philosophical relation can exist, for no objects can
be compared unless they have some degree of resemblance.
But though resemblance is necessary for all philosophical relation, it doesn’t follow that it always produces a connection
or association of ideas. When a quality becomes very general,
and is common to a great many individuals, it doesn’t lead
the mind directly to any one of them, because there is too
great a choice for the imagination to fix on any single object.
2. •Identity can be counted as a second kind of relation.
This relation I here consider as applied in its strictest sense
to constant and unchanging objects, without examining the
nature and foundation of personal identity, which will be
discussed later (·in I.iv.6·). Identity is the most universal of
all relations, because it is common to every thing that stays
in existence for any period of time.
3. The next most universal and comprehensive relations,
after identity, are those of •space and time, which are the
sources of an infinite number of comparisons, such as distant, contiguous, above, below, before, after, etc.
4. All objects that admit of •quantity or number can be
compared in that respect, which is another very fertile source
of relation.
5. When two objects have a quality in common, the
•degrees to which they have it form a fifth species of relation. Thus, of two objects that both have weight, one may be
lighter or heavier than the other. Two colours of the same
kind may be of different shades, and in that respect admit of
comparison.
6. The relation of •contrariety might at first be regarded
as an exception to the rule that no relation of any kind

The word ‘relation’ is commonly used in two senses considerably different from each other: either for a •quality by
which two ideas are connected in the imagination so that
one naturally introduces the other, in the way I have explained; or for a •particular basis on which we may see fit to
compare two ideas which we choose to bring together in the
fancy (·without either of them naturally leading to the other·).
[In Hume’s time, ‘comparing’ two things could be simply bringing them
together in a single thought, not necessarily a thought about their being
alike. (We still have that usage in the expression ‘Let’s get together and
compare notes’.) That broader, weaker sense of ‘compare’ seems clearly
to be sometimes at work in the present section, but in the paragraph la-

What
is called a ‘relation’ in common speech is always the former; only in philosophy do we extend the word to cover any
particular basis of comparison when there is no connecting
principlec . For example, distance will be classified by philosophers as a true relation, because we acquire an idea of it by
comparing objects; but in everyday speech we say things like
‘Nothing can be more distant than the furthest star and the
earth; no two things can have less relation’, as if distance
and relation were incompatible. [For Hume, relations don’t divide
belled ‘1.’ our more usual narrower sense seems to be assumed.]

into •philosophical and •natural; and when he writes as though they do,
he is using ‘philosophical’ as short-hand for ‘only philosophical’ or ‘philosophical and not natural’.] It may be thought an endless task
to enumerate all the qualities that make objects admit of
comparison, and by which the ideas of philosophical relation
are produced. But if we look carefully we shall find that they
can easily be put into seven kinds, which can be considered
as the sources of all philosophical relation ·and thus of all
relation ·.

8

Treatise, Book 1

David Hume

can hold between two things without some degree of resemblance between them. But bear in mind that no two ideas are
in themselves ·flatly and absolutely· contrary except those
of existence and non-existence; and it is clear that ·even·
these—·contrary though they are·—are resembling, because
each of them conveys an idea of the object, though the latter excludes the object from times and places at which it is
supposed not to exist.
7. All other objects—such as fire and water, heat and
cold—are found to be contrary only by experience, and from

i: Ideas

the contrariety of their causes or effects; and this relation of
•cause and effect is a seventh philosophical relation, as well
as a natural one. The resemblance implied in this relation
will be explained later.
You might naturally have expected me to include difference among the relations; but I regard difference as a
negation of relation rather than as anything real or positive.
Difference is of two kinds, as opposed either to identity or
resemblance. The first is called difference of number, the
other difference of kind.

6: Modes and substances
I have a question for those philosophers who base so much
of their reasoning on the distinction between substance and
accident [= ‘quality’], and who imagine that we have clear ideas
of each. Is the idea of substance—I ask—derived from impressions of sensation or of reflection? If ·the former, that
is, if· it is conveyed to us by our senses, I ask: Which of our
senses, and how? If it is perceived by the eyes, it must be
a colour; if by the ears, a sound; if by the palate, a taste;
and so on with the other senses. But I don’t think anyone
will say that substance is a colour, a sound, or a taste! So
the idea of substance must be derived from an impression of
reflection, if it really exists. But the impressions of reflection
come down to our passions and emotions, and none of those
can possibly represent a substance. So we have no idea of
substance other than the idea of a collection of particular
qualities, and such collections are all we can meaningfully
refer to when we talk or think about ‘substance’.

can recall that collection to ourselves or to others. But the
difference between these two sorts of ideas comes from the
following facts about the ideas of •substances. The particular qualities that form a substance are commonly referred
to an unknown ·and fictional· something in which they are
supposed to inhere; or, when this fiction doesn’t occur, the
qualities in the collection are at least supposed to be closely
and inseparably connected by the relations of contiguity and
causation. The effect of this is that when we discover some
new simple quality to have the same connection with the rest
·as they have with one another·, we immediately include it
among them, even though it didn’t enter into our first conception of the substance. Thus our idea of gold may at first
be a yellow colour, weight, malleableness, and fusibility; but
when we learn that it is soluble in aqua regia we join that to
the other qualities and suppose it to belong to the substance
as much as if its idea had been part of the compound one
from the outset. The uniting force of the qualities is regarded
as the chief part of the complex idea, so it provides a way
into the complex idea for any quality that turns up later—

The idea of a •substance, as well as that of a •mode, is
nothing but a collection of simple ideas that are united by
the imagination and assigned a particular name by which we
9

Treatise, Book 1

David Hume

letting that quality in along with the ones that first presented
themselves.
To see that this can’t happen with •modes, consider their
nature. The simple ideas out of which modes are formed
either represent
qualities that are not united by contiguity and causation, but are scattered through different subjects;

i: Ideas

or, if they are all united together,
the uniting principlec is not regarded as the foundation of the complex idea.
The idea of a dance is an instance of the first kind of mode;
the idea of beauty an example of the second. It is obvious
why complex ideas of this kind can’t admit any new idea
without changing the name that has been given to the mode.

7: Abstract ideas
An important question has been raised about abstract or
general ideas, namely: Are they general or particular in
the mind’s conception of them? A great philosopher—Dr
Berkeley—has challenged the usual opinion about this, and
has asserted that a general idea is nothing but a particular
idea attached to a certain word that gives it a wider application and makes it recall (when needed) other individuals that
are similar to it. As I regard this as one of the greatest and
most valuable scholarly discoveries that has been made in
recent years, I shall try here to confirm it by some arguments
that I hope will put it beyond all doubt and controversy.

ticular sizes or qualities at all. The former of these is
absurd, because it implies an infinite capacity in the
mind. So we must opt for the latter, and suppose that
our abstract ideas represent no particular degree of
quantity or quality.
I shall try to show that this inference is erroneous, by arguing •that it is utterly impossible to conceive any quantity or
quality without forming a precise notion of its degrees; and
secondly •that though the capacity of the mind is not infinite
we can form—all at once—a notion of all possible degrees of
quantity and quality. However imperfect our way of doing
this may be, it may at least ·be good enough to· serve all
the purposes of thought and conversation. (·My first point
challenges the argument’s conclusion; my second undercuts
one of its premises·.)
To begin with the first proposition, that the mind can’t
form any notion of quantity or quality without forming a
precise notion of degrees of each, we can prove this by the
following three arguments. Firstly, I have observed that
whatever objects are different are distinguishable, and
that whatever objects are distinguishable are separable by the thought and imagination.
Now we should bring in the converse propositions:

It is evident that in forming most (if not all) of our general
ideas we abstract from every particular degree of quantity
and quality, and that objects aren’t prevented from belonging
to the same species by small differences in size, duration, or
other properties. It might be thought that we are here confronted by a plain dilemma, which will let us settle the nature
of those ‘abstract ideas’ that philosophers have speculated
about so much. ·Here is how the argument runs·:
The abstract idea of man represents men of all sizes
and all qualities, and there are only two ways it might
do that: by •representing all possible sizes and all
possible qualities at once, or by •representing no par10

Treatise, Book 1

David Hume

i: Ideas

liveliness. What I have said ·about impressions’ having to be
determinate in quantity and quality· wasn’t based on their
having a certain degree of liveliness, so it must hold equally
for perceptions that are less lively. So: an idea is a weaker
impression; and, as a strong impression must necessarily
have a determinate quantity and quality, the same must
hold for its copy or representative.
Thirdly, it is a principle generally accepted in philosophy
that every thing in Nature is individual, and that it is utterly
absurd to suppose (for instance) a really existent triangle
that has no precise proportion of sides and angles. If this is
absurd in fact and reality, therefore, it must also be absurd
in idea, since nothing of which we can form a clear and distinct idea is absurd and impossible. But forming the idea of
an object and forming an idea is the same thing; describing
the idea as ‘of an object’ merely relates it to something outside it and says nothing about its own character. Now, as it
is impossible to
form an idea of an object that has quantity and quality
but no precise degree of either,
it follows that it is equally impossible to
form an idea that itself has quantity and quality but
no precise degree of either.
So abstract ideas are •in themselves individual, even when
they become •general in their representation. The image
in the mind is only that of a particular object, though the
application of it in our reasoning may be the same as if it
were universal. ·Now I turn to the question of how we apply
such ideas in our reasoning·.
This application of ideas beyond their nature—·that is,
their being used universally although in their own nature
they are particular·—comes from our bundling together all
their possible degrees of quantity and quality, in a rough and
ready way that serves for everyday purposes. This is the sec-

whatever objects are separable are also distinguishable, and whatever objects are distinguishable are
also different.
For how could we separate what is not distinguishable, or
distinguish what is not different? With this in hand, let
us examine whether all the circumstances that we abstract
out of a general idea are distinguishable and different from
those that we retain as essential parts of the idea. It is clear
straight off that •the precise length of a line is not different
or distinguishable from •the line itself; ·and more generally
that· the precise degree of any quality is not different or
distinguishable from the quality. Since these don’t admit
of distinction and difference, they don’t admit of separation
either (·following the second of the propositions displayed
above·). So they are ·inseparably· conjoined with each other
in the conception: our general idea of a line, notwithstanding
all our abstractions and refinements, has in its appearance
in the mind a precise degree of quantity and quality. Even if
it is made to represent other lines that have different degrees
of both, ·it doesn’t do this by not having any degree of either·.
Secondly, it is admitted that no object can appear to
the senses— i.e. that no impression can become present to
the mind—without being determinate in its degrees both of
quantity and quality. Impressions are sometimes confused,
but only because they are faint or unsteady, not because
the mind can receive any impression that in itself has no
particular degree nor proportion! Such an impression would
be a contradiction in terms, and even implies the flattest of
all contradictions, namely that it is possible for something
both to be and not to be.
Now, since all ideas are •derived from impressions and
are nothing but •copies and •representations of them, whatever is true of the one must be admitted to hold also for the
other. Impressions and ideas differ only in their strength and
11

Treatise, Book 1

David Hume

ond proposition I proposed to explain ·in my initial criticism
of the ‘dilemma’ argument· . When we have found a resemblance among a number of objects that we often encounter,
we apply a single name to all of them, whatever differences
we may observe in the degrees of their quantity and quality,
and whatever other differences may appear among them.2

For this is one of the most extraordinary aspects of this
business, that after we have done some reasoning with an
individual idea in our mind, the associated custom—revived
by the general or abstract word ·that we use to name the
original idea·—readily brings to mind any other individual
·to which the word also applies·, if by chance we have gone
wrong in our reasoning. For example, if we used the word
‘triangle’ and formed the idea of a particular equilateral one
to correspond to it, and if we went on to assert that the three
angles of a triangle are equal to each other, all the individual
triangles that we have overlooked—the ones that are not
equilateral—would immediately crowd in on us and make
us see the falsehood of what we had just said, even though
it was true in relation to the idea we ·first· formed. If the
mind doesn’t always suggest these ideas when it would be
appropriate to do so, that comes from some imperfection in
its faculties—an imperfection that is often the source of false
reasoning and sophistry. But this is principally the case
with ideas that are abstruse and compounded. On other occasions the custom is more entire [= ‘holds together more firmly’],
and we don’t often run into such errors.

After we have become accustomed to using the word in
that way, the hearing of it revives ·in our mind· the idea of
one of these objects, and makes the imagination conceive it
in all its particular detail. But the same word is supposed
to have been frequently applied to other individuals that are
different in many respects from the idea that is immediately
present to the mind, and the word can’t revive the idea of all
these individuals; so all it does is to touch the soul (if I may
put it like that) and revive the custom that we have acquired
by surveying them. Those individuals are present to the mind
not •actually but •only potentially. We don’t portray them
all distinctly in the imagination, but keep ourselves ready to
survey any of them when we are so prompted by a present
plan or by necessity. The word raises up an •individual idea
along with a certain •custom, and that custom produces any
other individual idea that comes to be appropriate. But as
the production of all the ideas to which the name may be
applied is in most cases impossible, we shorten the work by
a more partial consideration, and we find few inconveniences
arising in our reasoning from that abridgment.
2

i: Ideas

Indeed, so entire is the custom that the very same idea
may be attached to several different words and used in different reasonings, with no danger of mistake. Thus the idea
of an equilateral triangle of an inch perpendicular may serve
us when we are talking of a

It is obvious that different ideas—even simple ones—can have a similarity or resemblance to each other; and the respect in which they are alike
need not be distinct or separable from respects in which they differ. Blue and green are different simple ideas, but they are more alike than are blue
and scarlet; though their perfect simplicity makes it impossible to separate or distinguish their respect of similarity. The same holds for particular
sounds, tastes, and smells. These can be alike in countless ways, taking them as wholes, without having any ·separable· feature in common. And we
can be sure of this general point by considering the very abstract phrase ‘simple idea’. This covers all simple ideas, and these •resemble each other
in that they are all simple. Yet precisely because they are simple, and thus have no complexity or compoundedness about them, this •respect in
which they are all alike is not distinguishable or separable from the rest. It is the same case with the different degrees of a quality: they are all alike,
yet the quality in any individual is not distinct from the degree—·we can’t, even in thought, separate a thing’s bright-blueness into two components
of which one is mere blueness·.

12

Treatise, Book 1

David Hume

‘figure’,
‘rectilinear figure’,
‘regular figure’,
‘triangle’, or
‘equilateral triangle’.
In this case all these terms are associated with the same
idea; but as the terms are standardly applied to smaller or
larger ranges of particulars, each arouses its special habit
·and no confusion need arise·. . . . Before those habits have
become entirely perfect, the mind may sometimes not be content with forming the idea of only one individual, and may
instead run over several ideas so as to make itself grasp its
own meaning and the range of the collection that it intends
to express by the general term. Wanting to fix the meaning
of the word ‘shape’, we may revolve in our mind the ideas of
circles, squares, parallelograms, triangles of different sizes
and proportions, not resting on one image or idea. Still, it is
certain that •when we use any general term we form the idea
of individuals, that •we can seldom if ever bring all these
individuals to mind, and that •the ones we don’t bring to
mind are represented only by means of the habit by which
we recall them when there is a need to. This then is the
nature of our abstract ideas and general terms; and this is
how I deal with the foregoing paradox, that some ideas are
particular in their nature but general in their representation. A particular idea becomes general by being attached to
a general term; i.e. to a term that is related by a customary
conjunction to many other particular ideas which it readily
recalls in the imagination.
If there remains any difficulty in this subject, it must
concern the custom that so readily recalls every particular
idea for which we may have need, and is triggered by any
word or sound to which we commonly attach it. The most
proper method of explaining this act of the mind, I think, is

i: Ideas

by producing other instances that are analogous to it, and
other forces that help it to operate. It is impossible to explain
the ultimate causes of our mental actions; it’s enough to give
a satisfactory account of them from experience and analogy.
First, then, I observe that when we mention any great
number, such as a thousand, the mind has generally no
adequate idea of it but only a power of producing such an
idea through its adequate idea of the decimals under which
the number is comprehended. This imperfection in our ideas,
however, never affects our reasonings; so this seems to be
an instance parallel to the one about universal ideas that I
have been discussing.
Secondly, we have several instances of habits that can be
revived by a single word; as when a person who has learned
by rote a speech or poem, and then can’t remember it, will
call the whole thing to mind once he is given the single word
or phrase with which it begins.
Thirdly, if you examine what goes on in your mind in
reasoning, I think will agree with me that we don’t attach
distinct and complete ideas to every term we use: in talking
of ‘government’, ‘church’, ‘negotiation’, ‘conquest’, we seldom spread out in our minds all the simple ideas of which
these complex ones are composed. Despite this imperfection,
however, we can avoid talking nonsense on these subjects
and can perceive any conflicts among the ideas as well as if
we had them fully in our minds. Thus, if instead of saying
that in war the weaker always have recourse to negotiation
we should say that they always have recourse to conquest,
the custom we have acquired of attributing certain relations
to ideas still follows the words and makes us immediately
perceive the absurdity of that proposition. This is like the
way in which one particular idea can serve us in reasoning
concerning other ideas, however much they differ from it in
some respects.
13

Treatise, Book 1

David Hume

Fourthly, when the individuals are collected together and
put under a general term on the basis of their resemblance
to one another, this relation must •make it easier for them
to enter the imagination, and •make them more likely to be
suggested when there is a need for them. And, indeed, if we
consider the usual way our thoughts move in private thought
and in conversation, we shall find good reason to be convinced of this. The imagination has an admirable readiness
to suggest its ideas and to present them at the very instant
when they are necessary or useful. In collecting the ideas
that belong to a subject, the fancy runs from one end of the
universe to the other: one might think that we could see the
whole intellectual world of ideas all at once, and that all we
did was to pick out the ideas that best suited our purpose.
But it may be that really the only ideas that we have ·at such
a moment· are the ·seemingly ‘picked out’ ones·—the very
ideas that are thus collected by a kind of magical faculty in
the soul. This faculty is always most perfect in the greatest
geniuses—and is properly what we call ‘genius’—but it can’t
be explained by the utmost efforts of human understanding.
Perhaps these four reflections will help to remove the
obstacles to accepting my hypothesis about abstract ideas,
contrary as it is to what has previously prevailed in philosophy. But to tell the truth, my chief source of confidence lies
in what I have already proved about the impossibility of general ideas according to the usual account of them. ·Because
of that proof·, we have to look for some other account of
general ideas, and clearly mine is the only candidate. If ideas
are •particular in their nature and •finite in their number,
the only way they can become•general in their representation and •contain infinitely many other ideas under them is
through custom.
Before I leave this subject, I shall employ the same principles to explain that ‘distinction of reason’ that is so much

i: Ideas

talked of, and so little understood, in the schools [= (roughly)
Examples
of this are the distinctions •between shape and the body
that has the shape, and •between motion and the body that
moves. It is hard to make sense of this ‘distinction’ in light
of the principle— explained above—that all ideas that are
different are separable. For it implies that if the shape is
different from the body ·that has it·, their ideas must be
separable as well as distinguishable, ·which plainly they are
not·. . . . What then is meant by a ‘distinction of reason’,
since it implies neither a difference nor separation? To remove this difficulty, we must rely on the account I have given
of abstract ideas. It is certain that nobody would ever have
dreamed of distinguishing a shape from the shaped body—
from which it is in reality not distinguishable or different
or separable—if it hadn’t been noticed that even this simple
shaped body (·which is ‘simple’ in the sense that it can’t
be divided into two elements, the shape and the body·) has
many different resemblances and relations ·to other things·.
For example, when we see a globe of white marble, we receive
only the impression of a white colour laid out in a certain
shape, and we can’t separate and distinguish the colour from
the shape. But when we later see a globe of black marble
and a cube of white, and compare them with our former
object, we find two separate resemblances in something that
formerly seemed, and really was, quite incapable of being
separated out into two components. After a little more practice of this kind, we begin to distinguish the shape from the
colour by a distinction of reason; that is, we consider the
shape and colour together, since they are in effect the same
and undistinguishable; but still we view them in different
aspects, according to the resemblances they can enter into.
Wanting to consider only the shape of the globe of white
marble, we actually form an idea of both the shape and the

‘philosophy departments dominated by Aristotelian ideas’].

14

Treatise, Book 1

David Hume

colour, but we have our eye on its resemblance to the globe
of black marble; and when we want to consider its colour
only, we look to its resemblance to the cube of white marble.
In this way we accompany our ideas with a kind of reflection
of which custom makes us largely unaware. Someone who
asks us to consider the shape of a globe of white marble

i: Ideas

without thinking of its colour is asking for an impossibility;
but what he means to ask is that we consider the colour and
shape together, but still keep our eye on the resemblance to
the globe of black marble or to any other globe of whatever
colour or substance.

15

Treatise, Book 1

David Hume

ii: Ideas of space and time

Part ii: The ideas of space and time
1: The infinite divisibility of our ideas of space and time
When a philosopher comes up with something that looks like
a paradox and is contrary to basic beliefs of ordinary folk,
·it often fares better than it deserves, for two reasons·. •It is
greedily embraced by philosophers, who think it shows the
superiority of their discipline that could discover opinions so
far from common beliefs. •When something surprising and
dazzling confronts us, it gives our minds a pleasurable sort of
satisfaction that we can’t think is absolutely baseless. These
dispositions in •philosophers and •their disciples give rise
to a relation of mutual comfort between them: •the former
furnish many strange and unaccountable opinions, and •the
latter readily believe them. I can’t give a plainer example of
this symbiosis than the doctrine of infinite divisibility. It will
be the first topic in my discussion of the ideas of space and
time.

from Latin meaning ‘no end’.] So it is certain that the imagination reaches a minimum, and can form in itself an idea of
which it can’t conceive any subdivision—one that can’t be
diminished without a total annihilation. When you tell me
of •the thousandth and •ten thousandth part of a grain of
sand, I have a distinct idea of these numbers and of their
different proportions; but the images I form in my mind to
represent the things themselves are not different from each
other and are not smaller than the that image by which I
represent •the grain of sand itself, which is supposed to be
so much bigger. What consists of parts is distinguishable
into them, and what is distinguishable is separable. But,
whatever we may imagine of the thing, the idea of a grain of
sand is not distinguishable or separable into twenty different ideas—much less into a thousand, ten thousand, or an
infinite number of them! The impressions of the senses are
the same in this respect as the ideas of the imagination. Put
a spot of ink on paper, fix your eye on that spot, and move
away just far enough so that you lose sight of it: it is obvious
that the moment before it vanished the image or impression
·of the spot· was perfectly indivisible. Why do small parts of
distant bodies not convey any sensible impression to us? It
is not for lack of rays of light ·from them· striking our eyes.
Rather, it is because they are further away than the distance
at which their impressions •were reduced to a minimum
and •couldn’t be diminished any further. A telescope that
makes them visible doesn’t produce any new rays of light,
but merely spreads out the rays that always flowed from
them: in that way the telescope •gives parts to impressions

Everyone agrees—and the plainest observation and experience makes it obvious—that the capacity of the mind is
limited, and can never attain a full and adequate conception
of infinity. It is also obvious that whatever is capable of being
divided in infinitum must consist of an infinite number of
parts: if you limit the number of parts, you thereby limit
the ·possible· division. It doesn’t take much work to conclude from this that the idea we form of any finite quality
is not infinitely divisible, and that by proper distinctions
and separations we can reduce it to lesser ideas that are
perfectly simple [= ‘without parts’] and indivisible. In denying
that the mind’s capacity is infinite we are supposing that
it will come to an end in the division of its ideas; and there
is no possible escape from this conclusion. [‘Infinite’ comes
16

Treatise, Book 1

David Hume

that had appeared simple and uncompounded to the naked
eye, and •advances to a minimum what was formerly imperceptible. The explanation of what a microscope does is
essentially the same.
From this we can discover the error of the common opinion that the capacity of the mind is limited on both sides,
and that the imagination can’t possibly form an adequate
idea of anything below a certain size or above a certain size.
Nothing can be more minute than some ideas that we form in
the imagination, and some images that appear to the senses,
for there are ideas and images that are perfectly simple and
indivisible, ·and nothing can be smaller than that·. The only
defect of our senses is that they give us wrongly proportioned
images of things, representing as tiny and uncompounded
what is really large and composed of a vast number of parts.
We aren’t aware of this mistake. ·Take the example of a
very tiny insect such as a mite·. When we see a mite we
take that impression to be equal or nearly equal in size to

ii: Ideas of space and time

the mite itself; then finding by reason that there are objects
much smaller than that—·for example, the small parts of the
mite·—we rashly conclude that these things are smaller than
any idea of our imagination or impression of our senses. But
it is certain that we can form ideas that are no bigger than
the smallest atom of the animal spirits of an insect a thousand times smaller than a mite. [‘Animal spirits’ were thought to
be extremely finely divided fluids in animal bodies —more fluid and finely

We ought rather to conclude that the
difficulty lies in enlarging our conceptions enough to form a
just notion of a mite, or even of an insect a thousand times
less than a mite. For in order to form a just notion of these
animals we must have a distinct idea representing each part
of them; and that, according to the system of infinite divisibility, is utterly impossible, and according to the system of
indivisible parts or atoms it is extremely difficult because of
the vast number and multiplicity of these parts.
divided than air or water.]

2: The infinite divisibility of space and time
When ideas adequately represent objects, the relations, contradictions, and agreements among the ideas all hold also
among the objects; and we can see this to be the general
foundation of all human knowledge. But our ideas are adequate representations of the tiniest parts of extended things,
so no parts of the things—through whatever divisions and
subdivisions we may suppose them to be arrived at—can
be smaller than some ideas that we form. The plain consequence, to be drawn with no shuffling or dodging, is that
whatever appears impossible and contradictory in relation to
these •ideas must be really impossible and contradictory ·in
relation to the •things·.

Everything that is capable of being infinitely divided contains an infinite number of parts; otherwise the division
would be stopped short by the indivisible parts that we would
arrive at. So
if •anything of finite size is infinitely divisible, then
•it can’t be a contradiction to suppose that an extended thing of finite size contains an infinite number
of parts;
and, putting the same thing the other way around,
if •it is a contradiction to suppose that a finite thing
contains an infinite number of parts, then •no finitely
extended thing can be infinitely divisible.
17

Treatise, Book 1

David Hume

·who believe that whatever is extended is divisible·, what is

The thesis that a finite thing can be infinitely divided is absurd, as I easily convince myself by considering my clear
ideas. I first take the smallest idea I can form of a part of
the extended world, and being certain that there is nothing
smaller than this idea, I conclude that whatever I discover
by means of it must be a real quality of extended things. I
then repeat this idea once, twice, thrice, and so on; this repetition brings it about that my compound idea of extension
grows larger and larger, becoming double, triple, quadruple,
etc. what it was before, until eventually it swells up to a
considerable size—larger or smaller depending on how often
I repeat the same idea. When I stop adding parts, the idea of
extension stops enlarging; and if I continued the addition in
infinitum, my idea of extension—this is clear—would have
to become infinite. From all this I infer that the idea of •an
infinite number of parts is just the idea of •an infinite extension; that no finite extension can contain an infinite number
of parts; and, consequently that no finite extended thing is
infinitely divisible.3

extended is always a number of items and never resolves
itself into a unit or indivisible quantity; from which it follows
that what is extended can never exist! It is no use replying
that a determinate quantity of extension is a unit, though
one that admits of an infinite number of fractions and can be
subdivided without limit. For by that standard these twenty
men can be considered as a unit. The whole planet earth,
indeed the whole universe, can be considered as a unit. That
kind of ‘unity’ involves a merely fictitious label that the mind
can apply to any quantity of objects that it collects together;
that sort of ‘unity’ can no more exist alone than number can,
because really it is a true number ·masquerading under a
false label· .The unity that can exist alone and whose existence is necessary to that of all number is of another kind; it
must be perfectly indivisible and incapable of being resolved
into any lesser unity.
All this reasoning applies also to ·the infinite divisibility
of· time, along with a further argument that we ought to take
notice of. A property of time that it cannot lose—it is in a
way time’s essence—is that its parts come in succession, and
that no two of them, however close, can exist together. Every
•moment must be distinct from—later or earlier than—each
other •moment, for the same reason that the •year 1737
cannot coexist with the present •year 1738. This makes it
certain that time, because it exists, must be composed of
indivisible moments. For if we could never arrive at an end
of the division of time, and if each moment as it succeeds
another were not perfectly single and indivisible, there would
be an infinite number of coexistent moments or parts of time,

Let me add another argument, proposed by a noted author (Monsieur Malezieu), which seems to me very strong and
beautiful. It is obvious that existence in itself belongs only
to unity, and is applicable to number only on the strength
of the units of which the number is composed. Twenty men
may be said to exist; but it is only because one, two, three,
four, etc. are existent; and if you deny the existence of the
individual men the existence of the twenty automatically
falls. So it is utterly absurd to suppose that a number ·of
items· exists and yet deny the existence of individual items.
Now, according to the common opinion of metaphysicians
3

ii: Ideas of space and time

It has been objected to me that infinite divisibility requires only an infinite number of proportional parts, . . . . and that an infinite number of proportional parts does not form an infinite extension. (·The objector is thinking of things like the division of a line into a half, followed by a quarter,
followed by an eighth, . . . and so on·.) But this is entirely frivolous. Whether or not the parts are proportional, they can’t be smaller than the minute
parts I have been talking about, and so the conjunction of them can’t generate a smaller extension.

18

Treatise, Book 1

David Hume

·namely the parts of the moment ·; and I think this will be

ii: Ideas of space and time

sible for them to have a sound basis. ·This will occupy the
remainder of this section; in section 3 I shall present some
further doctrine about the ideas of space and (especially)
time, and sections 4–5 will address objections to this further
doctrine as well as objections to my view about divisibility·.
It is an established maxim in metaphysics that

agreed to be an outright contradiction.
The infinite divisibility of space implies that of time, as
is evident from the nature of motion. So if time can’t be
infinitely divisible, space can’t be either.
Even the most obstinate defender of infinite divisibility
will surely concede that these arguments are ‘difficulties’,
and that no perfectly clear and satisfactory answer can be
given to them. Let me point out here the absurdity of this
custom of trying to evade the force and evidentness of something that claims to be a demonstration [= ‘a logically rigorous
proof’] by calling it a ‘difficulty’. It doesn’t happen with demonstrations, as it does with probabilities, that difficulties crop
up and one argument counterbalances another and lessens
its force. If a demonstration is sound, it can’t admit of an opposing difficulty; and if it is not sound it is nothing—a mere
trick—and can’t itself be a difficulty. It is either •irresistible
or •without any force at all. If in a topic like our present
one you talk of ‘objections’ and ‘replies’, and of ‘balancing’
arguments ·pro and con·, you are either accepting that human reasoning is nothing but word-play or showing that you
don’t have the intellectual capacity needed for such subjects.
A demonstration may be difficult to understand because of
the abstractedness of its subject; but it can’t have difficulties
that will weaken its authority once it has been understood.
It is true that mathematicians are given to saying that
there are equally strong arguments on the other side of our
present question, and that the doctrine of indivisible points
is also open to unanswerable objections. I shall examine
these arguments and objections in detail ·in sections 4 and
5·; but first I will take them all together and try to prove
through a short and decisive reason that it is utterly impos-

Whatever the mind clearly conceives includes the idea
of possible existence—that is, nothing that we imagine
is absolutely impossible.
We can form the idea of a golden mountain, from which we
conclude that such a mountain could actually exist. We can
form no idea of a mountain without a valley, and therefore
regard it as impossible.
Now, it is certain that we have an idea of extension, for
how otherwise could we talk and reason about it? It is
also certain that this idea as conceived by the imagination,
though divisible into parts or smaller ideas, is not infinitely
divisible and doesn’t consist of an infinite number of parts;
for that would exceed the grasp of our limited capacities. So
there we have it: an idea of extension consisting of parts
or lesser ideas that are perfectly indivisible; so this idea implies no contradiction: so it is •possible for extension reality
also to be like that; so all the •arguments that have been
brought against the •possibility of mathematical points are
mere scholastic quibbles that don’t deserve our attention.
We can carry this line of argument one step further, concluding that all the purported demonstrations of the infinite
divisibility of the extended are equally invalid; because it is
certain that these demonstrations cannot be sound without
proving the impossibility of mathematical points; which it is
an evident absurdity to claim to do.

19

Treatise, Book 1

David Hume

ii: Ideas of space and time

3: The other qualities of our ideas of space and time
For deciding all controversies regarding ideas, no discovery could have been more fortunate than the one I have
mentioned, that
impressions always precede ideas, and every ·simple·
idea that comes into the imagination first makes its
appearance in a corresponding impression.
These •impressions are all so clear and evident that they
there is no argument about them, though many of our •ideas
are so obscure that it is almost impossible even for the mind
in which they occur to say exactly what they are like and
how they are made up. Let us apply this principle with a
view to revealing more about the nature of our ideas of space
and time.
On opening my eyes and turning them to the surrounding
objects, I see many visible bodies; and when I shut my eyes
again and think about the distances between these bodies, I
acquire the idea of extension. As every idea is derived from
some impression that is exactly like it, this idea of extension must come from some impression, which can only be
either some sensation derived from sight or some internal
impression arising from these sensations.
Our internal impressions are our passions, emotions, desires, and aversions; and I don’t think you’ll say that they are
the model from which the idea of space is derived! So there
remain only the ·external· senses as sources for this original
impression. Well, what impression do our senses here convey to us? This is the main question, and it decisively settles
what the idea is like.
My view of the table in front of me is alone sufficient
to give me the idea of extension. So this idea is borrowed
from, and represents, some impression that appears to my

senses at this moment. But my senses convey to me only the
impressions of coloured points arrayed in a certain manner.
If you think the eye senses anything more than that, tell me
what! And if it is impossible to show ‘anything more’, we can
confidently conclude that •the idea of extension is nothing
but •a copy of these coloured points and of the manner of
their appearance.
Suppose that when we first received the idea of extension
it was from an extended object—or composition of coloured
points—in which all the points were of a purple colour. Then
in every repetition of that idea we would not only place the
points in the same order with respect to each other, but
would also bestow on them that precise colour which was
the only one we had encountered. But afterwards, having
experience of other colours —violet, green, red, white, black,
and all the different combinations of these—and finding a resemblance in the layout [Hume’s word is ‘disposition’] of coloured
points of which they are composed, we omit the peculiarities
of colour as far as possible, and establish an abstract idea
based merely on the layout of points—the manner of appearance that is common to them all. Indeed, even when the
resemblance is carried beyond the objects of one sense, ·and
the sense of touch comes into the story·, the impressions of
touch are found to be similar to those of sight in the layout
of their parts, and because of this resemblance the abstract
idea can represent both. All abstract ideas are really nothing
but particular ones considered in a certain light; but being
attached to general terms they can represent a vast variety,
and can apply to objects which are alike in some respects
and vastly unalike in others.
The idea of time is derived from the succession of our
20

Treatise, Book 1

David Hume

perceptions of every kind—ideas as well as impressions, and
impressions of reflection as well as of sensation. So it’s
an example of an abstract idea that covers a still greater
variety than does the idea of space, and yet is represented
in the imagination by some particular individual idea of a
determinate quantity and quality.
As we receive the idea of space from the layout of visible
and tangible objects, so we form the idea of time from the
succession of ideas and impressions ·in our minds·. Time
cannot all on its own make its appearance or be taken notice
of by the mind. A man in a sound sleep, or strongly occupied
with one thought, is unaware of time; the same duration
appears longer or shorter to his imagination depending on
how quickly or slowly his perceptions succeed each other. A
great philosopher (Mr Locke) has remarked that our perceptions have certain limits in this respect— limits that are fixed
by the basic nature and constitution of the mind— beyond
which no influence of external objects on the senses can ever
speed up our thought or slow it down. If you quickly whirl
around a burning coal, it will present to the senses an image
of a circle of fire, and there won’t seem to be any interval
of time between its revolutions. That is simply because our
perceptions can’t succeed each other as quickly as motion
can be communicated to external objects. When we have
no successive perceptions, we have no notion of time, even
though there is a real succession in the objects—·as when in
a single circling of the burning coal, the second quarter of the
journey follows the first quarter·. From these phenomena,
as well as from many others, we can conclude that time can’t
make its appearance to the mind •alone or •accompanied by
a steady unchanging object, but is always revealed by some
perceivable succession of changing objects.
To confirm this we can add the following argument, which
strikes me as perfectly decisive and convincing. It is evident

ii: Ideas of space and time

that time or duration consists of different parts; for otherwise
we couldn’t conceive a longer or shorter duration. It is also
evident that these parts are not coexistent: for the quality of
having parts that coexist belongs to •extension, and is what
distinguishes it from •duration. Now as time is composed
of parts that don’t coexist, an unchanging object, since it
produces only coexistent impressions, produces none that
can give us the idea of time; and consequently that idea must
be derived from a succession of changing objects, and time
in its first appearance can never be separated from such a
succession.
Having found that time in its first appearance to the mind
is always joined with a succession of changing objects, and
that otherwise we can never be aware of it, we now have
to ask whether •time can be conceived without our conceiving any succession of objects, and whether •there can be a
distinct stand-alone idea of time in the imagination.
To know whether items that are joined in an impression
are separable in ·the corresponding· idea, we need only to
know whether the items are different from one another. If
they are, it is obvious that they can be conceived apart:
things that are different are distinguishable, and things that
are distinguishable can be separated, according to the maxims I have explained. If on the contrary they are not different
they are not distinguishable, in which case they can’t be
separated. But this ·latter state of affairs· is precisely how
things stand regarding •time in relation to •succession in our
perceptions. The idea of time is not derived from a particular
impression mixed up with others and plainly distinguishable
from them; its whole source is the manner in which impressions appear to the mind—it isn’t one of them. Five notes
played on a flute give us the impression and idea of time,
but time is not a sixth impression that presents itself to
the hearing or to any other of the senses. Nor is it a sixth
21

Treatise, Book 1

David Hume

impression that the mind finds in itself by reflection, ·thus
yielding time as an idea of reflection·. To produce a new idea
of reflection the mind must have some new inner impression:
it can go over all its ideas of sensation a thousand times
without extracting from them any new original idea, unless
it feels some new original impression arise from this survey.
·And, returning now to our flute·, these five sounds making
their appearance in this particular manner don’t start up any
emotion or inner state of any kind from which the mind, observing it, might derive a new idea. All the mind does in this
case is to notice the manner in which the different sounds
make their appearance, and ·to have the thought· that it
could afterwards think of it as the manner in which other
things—·other than the five flute-notes·—might appear. ·For
the mind to have the idea of time·, it must certainly have
the ideas of some objects [here = ‘events’], for without these it
could never arrive at any conception of time. Time doesn’t
appear as a primary distinct impression, so it has to consist
in different ideas or impressions or objects disposed in a certain manner—the manner that consists in their succeeding
each other.
Some people, I know, claim that the idea of duration is
applicable in a proper sense to objects that are perfectly
unchanging; and I think this is the common opinion of
philosophers as well as of ordinary folk. To be convinced of
its falsehood, however, reflect on the above thesis that
the idea of duration is always derived from a succession of changing objects, and can never be conveyed
to the mind by anything steadfast and unchanging.
It inevitably follows from this that since the idea of duration
can’t be derived from such an object it can’t strictly and
accurately be applied to such an object either, so that no
unchanging thing can ever be said to have duration, ·i.e. to
last through time·. Ideas always represent the objects or

ii: Ideas of space and time

impressions from which they are derived, and it is only by a
fiction that they can represent or be applied to anything else.
We do engage in a certain fiction whereby we apply the idea
of time to unchanging things and suppose that duration is
a measure of rest as well as of motion. I shall discuss this
fiction in section 5.
There is another very decisive argument that establishes
the present doctrine about our ideas of space and time; it
relies merely on the simple principle that our ideas of space
and time are compounded of parts that are indivisible. This
argument may be worth examining.
Every idea that is distinguishable is also separable; so
let us take one of those simple indivisible ideas of which the
compound idea of extension is formed, separate it from all
others, and consider it on its own. What are we to think are
its nature and qualities? Clearly it isn’t the idea of extension;
for the idea of extension consists of parts, and we have stipulated that the idea we are considering is perfectly simple
and indivisible ·and therefore has no parts·. Is it nothing,
then? That is absolutely impossible. The compound idea
of extension is real, and is composed of ideas just like this
one we are considering; if they were all nonentities, there
would be an existing thing composed of nonentities, which
is absurd. So I have to ask: What is our idea of a simple and
indivisible point? If my answer seems somewhat new, that is
no wonder, because until now the question has hardly ever
been thought of. We are given to arguing about the nature
of mathematical points, but seldom about the nature of the
ideas of points.
The idea of space is conveyed to the mind by two senses,
sight and touch; nothing ever appears to us as extended
unless it is either visible or tangible. The compound impression that represents extension consists of several smaller
impressions that are indivisible to the eye or feeling, and
22

Treatise, Book 1

David Hume

may be called
impressions of atoms or corpuscles endowed with
colour and solidity.
But this is not all. For these atoms to reveal themselves to
our senses, it is not enough merely that they be coloured
or tangible; we have to preserve the idea of their colour or
tangibility, if we are to grasp them by our imagination. The
idea of their colour or tangibility is all there is that can make
them conceivable by our mind. Deprive the ideas of these
sensible qualities and you annihilate them so far as thought
or imagination is concerned Now, as the parts are, so is the
whole. If a point is not considered as coloured or tangible,

ii: Ideas of space and time

it can’t convey any idea to us, in which case there can’t be
an idea of extension that is composed of the ideas of these
points. If the idea of extension really can exist, as we are
aware it does, its parts must also exist, which requires them
to be considered as coloured or tangible. So we have no idea
of space or extension as anything except an object either of
our sight or feeling.
The same reasoning will prove that the indivisible moments of time must be filled with some real object, some
existing item, whose succession forms the duration and
makes it conceivable by the mind.

4: Objections answered
ideas of the manner or order in which objects exist ·or in
which events occur·. This means that it is impossible to conceive either •a ·spatial· vacuum, extension without matter, or
•·a temporal vacuum, so to speak·, a time when there is no
succession or change in any real existence. Because these
parts of my system are intimately connected, I shall examine
together the objections that have been brought against both
of them, beginning with those against the finite divisibility of
extension.

My system about space and time consists of two intimately
connected parts. •The first depends on this chain of reasoning.
•The capacity of the mind is not infinite. So
•any idea of extension or duration consists not of an

infinite number of parts or smaller ideas, but of a
finite number that are simple and indivisible. So
•it is possible for space and time to exist conformable
to this idea, i.e. as only finitely divisible. So
•space and time actually do exist in that form, since
their infinite divisibility is utterly impossible and contradictory.

1. The objection that I shall take first really has the effect
of showing that the two parts of my system depend on one
another, rather than of destroying either of them. In the
schools they have often argued like this:

•The other part of my system is a consequence of this. Dividing ideas of space and time into their parts, one eventually
reaches parts that are indivisible; and these indivisible parts,
being nothing in themselves, are inconceivable unless they
are filled with something real and existent. So the ideas of
space and time are not separate or distinct ideas, but merely

•A mathematical point is a nonentity; so
•no assemblage of such points can constitute a real

existence; so
•the whole system of mathematical points is absurd;

·so
23

Treatise, Book 1

David Hume

•there is no coherent account of where the division of

ii: Ideas of space and time

one of the bodies and the preservation of the other, without
our being able to tell which is which. Before the contact we
have the idea of two bodies; after it we have the idea only of
one. ·This is the only way we can make sense of ‘penetration’,
for· the mind can’t possibly preserve any notion of difference
between two bodies of the same nature existing in the same
place at the same time.

extended things would end if it did end; so
•such a division doesn’t end·; so
•anything extended must be infinitely divisible.
This would be perfectly decisive if there were no middle way
between •the infinite divisibility of matter and •the nonentity
of mathematical points. But there is such a way, namely
conferring colour or solidity on these points; and the absurdity of the two extremes is a demonstration of the truth and
reality of this middle way. (The system of physical points,
which is an alternative middle way, is too absurd to need
a refutation. A real extension such as a physical point is
supposed to be must have can’t exist without parts that are
different from each other; and when objects are different
they are distinguishable and separable by the imagination,
·which means that the supposed physical point isn’t a point
after all·.)
2. The second objection to the view that extension consists of mathematical points is that this would necessitate
penetration. A simple and indivisible atom that touches another (the argument goes) must penetrate it; for it can’t touch
the other only at its external parts because it, being simple,
doesn’t have parts. So one atom has to touch the other
intimately, in its whole essence, [then some Latin phrases],
which is the very definition of ‘penetration’. But penetration
is impossible; so mathematical points are impossible too.
I answer this objection by substituting a sounder idea of
penetration. What we must mean when we talk of penetration is this:
two bodies containing no empty space within them
come together and unite in such a way that the body
resulting from their union is no bigger than either of
them.
Clearly this penetration is nothing but the annihilation of

Taking ‘penetration’ in this sense, now, as meaning the
annihilation of one body on its contact with another, I ask:
Does anyone see a necessity that a coloured or tangible
point should be annihilated upon the approach of another
coloured or tangible point? On the contrary, doesn’t everyone see clearly that from the union of these points there
results an object that is compounded and divisible and can
be distinguished into two parts—each part preserving its
existence, distinct and separate, despite its being right next
to the other? ·If help is needed·, aid your imagination by
conceiving these points to be of different colours, to help you
keep them distinct. Surely a blue and a red point can lie
next to one another without any penetration or annihilation.
For if they can’t, what can possibly become of them? Shall
the red or the blue be annihilated? Or if these colours unite
into one, what new colour will they produce by their union?
What chiefly gives rise to these objections, and at the same
time makes it so hard to answer them satisfactorily, is the
natural infirmity and unsteadiness of our imagination and
our senses when employed on such tiny objects. Put a spot
of ink on paper and back away to a place from which the spot
is altogether invisible: you will find that as you move back
towards the spot it at first •becomes intermittently visible,
then •becomes continuously visible, and then •acquires a
new force only in ·the intensity of· its colouring, without
getting any bigger; and afterwards, when it has increased
enough to be really extended, it will still be hard for your
24

Treatise, Book 1

David Hume

imagination to break it into its component parts, because of
the uneasiness you will experience in the conception of such
a tiny object as a single point. This infirmity affects most of
our reasonings on the present subject, and makes it almost
impossible to answer intelligibly and accurately the many
questions that can arise about it.
3. Many objections to the thesis of the indivisibility of
the parts of extension have been drawn from mathematics,
though at first sight that science seems favourable to my doctrine. Anyway, although it is contrary in its demonstrations,
it perfectly agrees with me in its definitions. My present
task, then, is to defend the definitions and to refute the
demonstrations.
A surface is defined to be length and breadth without
depth; a line to be length without breadth or depth; a point
to be what has neither length, breadth, nor depth. It is
evident that all this is perfectly unintelligible on any other
supposition than that of the composition of extension by
indivisible points or atoms. How else could anything exist
without length, without breadth, or without depth? Two different answers, I find, have been made to this argument ·of
mine·, neither of them satisfactory in my opinion. •The first
answer is that the objects of geometry—those surfaces, lines,
and points whose proportions and positions it examines—are
mere ideas in the mind; they never did and indeed never can
exist in nature. They never did exist, because no-one will
claim to draw a line or make a surface that perfectly fits the
definition; and they never can exist, because we can produce
demonstrations from these very ideas to prove that they are
impossible.
But can anything be imagined more absurd and contradictory than this reasoning? Whatever can be conceived by
a clear and distinct idea necessarily implies the possibility of
existence; and someone who claims to prove the impossibility

ii: Ideas of space and time

of its existence by any argument derived from the clear idea
is really saying that we have no clear idea of it because we
have a clear idea! It is pointless to search for a contradiction
in something that is distinctly conceived by the mind. If it
implied a contradiction, it couldn’t possibly be conceived.
So there is no middle way between allowing at least the
possibility of indivisible points and denying that there is any
idea of them. And that principle is the basis for •the second
answer to the argument of mine that I have been defending.
It has been claimed that though it is impossible to conceive
a length without any breadth, we can consider one without
bringing in the other, doing this by means of an abstraction
without a separation. It is in this way (they say) that we can
think the length of the road between two towns while ignoring
its breadth. The length is inseparable from the breadth both
in Nature and in our minds; but that doesn’t rule out ·our
giving the length· a partial consideration, thereby making a
distinction of reason.
In refuting this answer I shan’t again press the argument
that I have already sufficiently explained, namely that if the
mind can’t reach a minimum in its ideas, its capacity must
be infinite in order to take in the infinite number of parts of
which its idea of any extension would be composed. Instead,
I’ll try to find some new absurdities in this reasoning.
A surface terminates a solid; a line terminates a surface;
a point terminates a line; but I contend that if the ideas
of a point, line, or surface were not indivisible we couldn’t
possibly conceive these terminations. ·Here is how I argue
for that·. Suppose that the ideas in question are infinitely
divisible, and then let your mind try to fix itself on the idea
of the last surface, line, or point; it will immediately find this
idea to break into parts; and when your mind seizes on the
last of these parts it will ·again· lose its hold because of a
new division—and so on ad infinitum, with no possibility of
25

Treatise, Book 1

David Hume

arriving at a terminating idea. The number of fractions bring
it no nearer the last division than the first idea it formed.
Every particle eludes the grasp by a new fraction, like quicksilver when we try to take hold of it. But as in fact •there
must be something that terminates the idea of any finite
quantity, and as •this terminating idea can’t itself consist of
parts or smaller ideas (otherwise the terminating would be
done not by this idea but by the last of its parts, and so on),
this is a clear proof that •the ideas of surfaces don’t admit
of any division in depth, those of lines can’t be divided in
breadth or depth, and those of points can’t be divided in any
dimension.
The schoolmen [= roughly ‘mediaeval Aristotelians’] were so
well aware of the force of this argument that some of them
maintained that, mixed in with •particles of matter that
are infinitely divisible, Nature has a number of ·indivisible·
•mathematical points, so as to provide terminations for bodies; and others dodged the force of this reasoning—·the reasoning of the preceding paragraph·—by a heap of unintelligible point-scorings and distinctions. Both these adversaries
equally yield the victory: a man who hides himself admits
the superiority of his enemy just as clearly as does one who
fairly hands over his weapons.
Thus it appears that the •definitions of mathematics destroy the purported •demonstrations: if we have ideas of
indivisible points, lines, and surfaces that fit their definitions, their existence is certainly possible; but if we have
no such ideas, it is impossible for us ever to conceive the
termination of any figure, and without that conception there
can be no geometrical demonstration.
But I go further, and maintain that none of these demonstrations can carry enough weight to establish such a principle as that of infinite divisibility. Why? Because when they
treat of such minute objects they are built on ideas that

ii: Ideas of space and time

are not exact and maxims that are not precisely true, so
that they are not properly demonstrations! When geometry
decides anything concerning the proportions of quantity, we
shouldn’t expect the utmost precision and exactness—none
of its proofs yield that. Geometry takes the dimensions and
proportions of figures accurately—but roughly, with some
give and take. Its errors are never considerable, and it
wouldn’t it err at all if it didn’t aim at such an absolute
perfection.
I first ask mathematicians what they mean when they
say that one line or surface is ‘equal to’, or ‘greater than’,
or ‘smaller than’ another. This question will embarrass any
mathematician, no matter which side of the divide he is on:
maintaining that what is extended is made up of •indivisible
points or of •quantities that are divisible in infinitum.
The few mathematicians who defend the hypothesis of
indivisible points (if indeed there are any) have the readiest
and soundest answer to my question. They need only reply
that lines or surfaces are equal when the numbers of points
in each are equal, and that as the proportion of the numbers
varies so does the proportion of the lines and surfaces. But
though this answer is sound, as well as obvious, I declare
that this standard of equality is entirely useless and that
it is never from this sort of comparison that we determine
objects to be equal or unequal with respect to each other.
The points that make up any line or surface, whether seen
or felt, are so tiny and so jumbled together that it is utterly
impossible for the mind to compute how many there are;
so such a computation can’t provide us with a standard by
which we may judge proportions. No-one will ever be able to
determine, by a precise count ·of constituent points·, that
an inch has fewer points than a foot, or a foot fewer than
a yard; which is why we seldom if ever consider this as the
standard of equality or inequality.
26

Treatise, Book 1

David Hume

As for those who imagine that extension is divisible in
infinitum, they can’t possibly give this answer to my question, or fix the equality of lines or surfaces by counting
their component parts. According to their hypothesis •every
figure—large or small—contains an infinite number of parts;
and •infinite numbers, strictly speaking, can’t be either equal
or unequal to one another; so •the equality or inequality of
any portions of space can’t depend on proportions in the
numbers of their parts. It can of course be said that the
inequality of a mile and a kilometre consists in the different
numbers of the feet of which they are composed, and that of
a foot and a yard in their different numbers of inches. But
the quantity we call ‘an inch’ in the one is supposed to be
equal to what we call ‘an inch’ in the other, this equality has
to be fixed somehow. ·Perhaps by sameness of numbers of
millimetres·! If we are not to embark on an infinite regress,
we must eventually fix some standard of equality that doesn’t
involve counting parts.

ii: Ideas of space and time

but useless. We must therefore look elsewhere for an answer
to my question.
Many philosophers refuse to assign any standard of
equality. To give us a sound notion of equality, they say, it
is sufficient to present two objects that are equal. They hold
that without the perception of such objects all definitions
are fruitless, and when we do perceive such objects we don’t
need any definition. I entirely agree with all this. I contend
that the only useful notion of equality or inequality is derived
from the whole united appearance and the comparison of
particular objects.
It is evident that the eye—or rather the mind—is often
able at one view to compare the size of bodies, and pronounce
them equal or unequal to each other without examining or
comparing the numbers of their minute parts. Such judgments are not only common but in many cases certain and
infallible. When the measure of a yard and that of a foot are
presented, the mind can no more question that the first is
longer than the second than it can doubt the most clear and
self-evident principles.
So there are three proportions that the mind distinguishes in the general appearance of its objects, and labels
as ‘larger’, ‘smaller’, and ‘equal’. But though its decisions
regarding proportions are sometimes infallible, they aren’t
always so; our judgments of this kind are as open to doubt
and error as those on any other subject. We frequently
correct our first opinion •by a review and reflection, and
judge objects to be equal that we at first thought unequal,
or regard an object as smaller than another though it had
formerly seemed to be larger. And that isn’t the only way in
which we correct these judgments of our senses: we often
discover our error •by putting the objects side by side; or,
where that is impracticable, •by applying some common and
invariable measure ·such as a yardstick· to each, learning

There are some who claim that equality is best defined
by congruence, and that two figures are equal if when they
are placed one on the other all their parts correspond to and
touch each other. To evaluate this definition I must first
make this preliminary point: equality is a relation; it isn’t a
property in the figures themselves, but arises merely from
the comparison the mind makes between them. So if equality
consists in this imaginary application and mutual contact
of parts, we must at least have a clear notion of these parts,
and must conceive their contact. In this conception, obviously, we would follow these parts down to the tiniest that
can possibly be conceived, because the contact of large parts
would never make the figures equal. But the tiniest parts we
can conceive are mathematical points! So this standard of
equality is the same as the one based on the equality of the
number of points, which we have already seen to be a sound
27

Treatise, Book 1

David Hume

in that way of their different proportions. And these corrections themselves are subject to further correction, and to
different degrees of exactness depending on the nature of
the measuring-instrument we use and the care with which
we use it.
So when the mind •has become accustomed to making
these judgments and to correcting them, and •has found that
when two figures appear to the eye to be equal they are also
equal by our other standards, •we form a mixed notion of
equality derived from both the looser and the stricter methods of comparison. But we are not content with this. Sound
reason convinces us that there are bodies vastly smaller than
those that appear to the senses (and false reason tries to
convince us that there are bodies infinitely smaller!); so we
clearly perceive that we have no instrument or technique of
measurement that can guarantee us against all error and uncertainty. We are aware that the addition or removal of one
of these tiny parts won’t show up either in the appearance or
in the measuring; and we imagine that two figures that were
equal before can’t be equal after this removal or addition; so
we suppose some imaginary standard of equality by which
the appearances and measuring are exactly corrected, and
the figures are related by that standard. This standard is
plainly imaginary. For as the idea of equality is the idea of
a specific appearance, corrected by placing the things
side by side or applying to each a common measure,
the notion of any correction that is finer than we have instruments and techniques to make is a mere fiction of the mind,
and is useless as well as incomprehensible. Although this
standard is merely imaginary, however, the fiction is very
natural: the mind often continues in this way with some procedure, even after the reason that started it off has ceased
to apply. This appears very conspicuously with regard to
time. Obviously we have no exact method of comparing pe-

ii: Ideas of space and time

riods of time—not even ones as good as we have for parts
of extension—yet the various corrections of our ·temporal·
measures, and their different degrees of exactness, have
given us an obscure unexpressed notion of perfect and entire
equality. The same thing happens in many other subjects
as well. A musician, finding that his ear becomes every day
more delicate, and correcting himself by reflection and attention, continues with the same act of the mind—·the same
thought of progressive refinement·—even when the subject
fails him ·because he is thinking of refinements that he can’t
actually make ·; and so he is led to entertain a notion of a
perfect major third or octave, without being able to tell where
his standard for that comes from. A painter creates the same
fiction with regard to colours; a mechanic with regard to
motion. To the former light and shade, to the latter swift
and slow, are imagined to be capable of exact comparison
and equality beyond the judgments of the senses.
We can apply the same reasoning to curves and straight
lines. Nothing is more apparent to the senses than the difference between a curved line and a straight one, and our
ideas of these are as easy to form as any ideas that we have.
But however easily we may form these ideas, it is impossible
to produce any definition of them that will fix the precise
boundary between them. When we draw a line on paper it
runs from point to point in a certain manner that determines
whether the line as a whole will look curved or straight;
but ·this ‘manner·, this order of the points, is perfectly unknown; all we see is the over-all appearance ·that results
from it·. Thus, even on the system of indivisible points we
can form only a distant notion of some unknown standard
to these objects. On the system of infinite divisibility we
can’t go even this far, and are left with merely the general
appearance as the basis on which to settle whether lines are
curved or straight. But though we can’t give a perfect defini28

Treatise, Book 1

David Hume

tion of ‘curved’ or ‘straight’, or come up with any very exact
method of distinguishing curved lines from straight ones,
this doesn’t prevent us from correcting ·our judgment based
on· the first appearance by •a more accurate consideration
and by •applying some standard of whose accuracy we are
more sure of because of its past successes. It is from these
corrections, and by carrying on the same ·correcting· action
of the mind past where there is any basis for it, that we form
the loose idea of a perfect standard for straight and curved,
without being able to explain it or grasp what it is.
Mathematicians, it is true, claim to give an exact definition of a straight line when they say that it is the shortest
distance between two points. ·I have two objections to this
supposed definition·. First: this is a statement of the properties of a straight line, not a sound definition of ‘straight’.
When you hear ‘a straight line’ mentioned, don’t you think
immediately of •a certain appearance, without necessarily
giving any thought to •this property? ‘Straight line’ can be
understood on its own, but this ‘definition’ is unintelligible
without a comparison with other lines that we conceive to
be longer. Also, in everyday life it is established as a maxim
that the straightest journey is always the shortest; but if
our idea of a straight line was just that of the shortest distance between two points, that maxim would be as absurd
as ‘The shortest journey is always the shortest’! Secondly, I
repeat what I showed earlier, that we have no precise idea of
equality and inequality, shorter and longer, any more than
we do of straight and curved; so the former can never yield a
perfect standard for the latter. An exact idea can’t be built
on ideas that are loose and indeterminate.
The idea of a plane surface is no more susceptible of
a precise standard than that of a straight line; we have no
means of distinguishing such a surface other than its general
appearance. It is useless for mathematicians to represent

ii: Ideas of space and time

a plane surface as produced by the flowing of a straight line.
This is immediately open to ·three· objections: (1) that •our
idea of a surface is as independent of •this way of forming
a surface as •our idea of an ellipse is of •the idea of a cone
(·though mathematicians ‘define’ an ellipse as something
made by cutting a cone in a certain way·); (2) that the idea of
a straight line is no more precise than that of a plane surface;
(3) that a straight line can flow irregularly and thus form a
figure quite different from a plane, so that ·for purposes of
the mathematicians’ definition· we must suppose the straight
line to flow along two straight lines parallel to each other and
on the same plane, which makes the definition circular.
So it seems that the ideas that are most essential to
geometry—namely the ideas of
equality and inequality,
straight line, and
plane surface
—are far from being exact and determinate, according to
our common method of conceiving them. We are not only
incapable of telling in difficult particular cases whether these
figures are equal, whether this line is straight, whether that
surface is plane; we can’t even have a firm and invariable
idea of equality or straightness or planeness. Our appeal is
still to the weak and fallible judgment that we make from
•the appearance of the objects and correct by •a compass or
·other· everyday device or technique; and if we bring in the
supposition of •some further correction, it will be either useless or imaginary. It is pointless to resort to the usual line of
thought that brings in God, supposing that his omnipotence
enables him to form a perfect geometrical figure, and draw a
straight line without any curve or inflection. As the ultimate
standard of these figures is derived from nothing but the
senses and imagination, it is absurd to talk of any perfection
29

Treatise, Book 1

David Hume

beyond what sense and imagination can determine, because
the true perfection of anything consists in its conformity to
its standard.
Since these ideas are so loose and uncertain, I want to
ask any mathematician:
What entitles you to be so utterly sure of plainest and
most obvious principles of your science (let alone of
the more intricate and obscure ones)? How can you
prove to me, for instance, that two straight lines can’t
have a segment in common? Or that it is impossible
to draw more than one straight line between any two
points?
If he replies that these opinions are obviously absurd, and
in conflict with our clear ideas, I answer:
I don’t deny that. When two straight lines approach
each other •with a perceptible angle between them, it
is absurd to imagine them to have a common segment.
But suppose two lines to approach at the rate of •one
inch in sixty miles, I see no absurdity in asserting
that when they meet they become one. Please tell
me what rule or standard you are going by when you
assert that the line in which I have supposed them
to come together can’t make the same straight line
as those two that form so small an angle between
them? Presumably you have some idea of a straight
line to which this line doesn’t conform. Well, then,
do you mean that the line in question doesn’t take
the points in the same order and by the rule that is
special and essential to a straight line? In judging in
this way you are allowing that extension is composed
of indivisible points, which may be more than you
intend; but let that pass. My present point is just
that •this is not the standard by which we form the
idea of a straight line; and that •even if it were, our

ii: Ideas of space and time

senses and imagination don’t provide anything firm
enough to determine when such an order is violated
or preserved. The original standard of a straight line
is in reality nothing but a certain general appearance;
and it is evident that straight lines can be made to
coincide and yet correspond to this standard, even if
it is corrected by all the means either practicable or
imaginable.
Whichever way they turn, mathematicians are still caught in
this dilemma. ·On one side of it·: If they judge of equality etc.
by the accurate and exact standard of the enumeration of
the minute indivisible parts, they •employ a standard that is
useless in practice, and •they rely on the truth of something
they have been trying to explode, namely the doctrine of indivisible parts of extension. ·On the other side of the dilemma·:
If they employ (as they usually do) the inaccurate standard
derived from the general appearance of objects when they
are considered together, corrected by measuring and putting
the objects side by side, their first principles are too coarse
to afford any such subtle inferences as they commonly draw
from them. The first principles are certain and infallible;
but they are based on imagination and the senses, so what
is ·soundly· inferred from them can never go beyond those
faculties, much less contradict them.
This may open our eyes a little, and let us see that no
geometrical ‘demonstration’ of the infinite divisibility of extension can have as much force as we naturally attribute to
every argument supported by such magnificent claims. At
the same time we may learn why it is that geometry fails to
convince us on this single point, while all its other reasonings
command our fullest assent and approval. And indeed there
seems to be more need to explain why this exception exists
than to show that it really is an exception and that all the
mathematical arguments for infinite divisibility are utterly
30

Treatise, Book 1

David Hume

sophistical. For it is obvious that as no idea of quantity is
infinitely divisible it is a glaring absurdity to try to prove that
quantity itself admits of such a division, arguing for this by
means of ideas that are directly opposite to that conclusion.
And as this absurdity is very glaring in itself, so every argument based on it is accompanied by a new absurdity and
involves an obvious contradiction.
I could cite as instances those arguments for infinite divisibility that are derived from the point of contact—·that is,
the point at which, supposedly, a circle is in contact with a
straight line that is tangential to it·. I know no mathematician will agree to be judged by the diagrams he draws on
paper, these being rough sketches (he will tell us) that serve
only to convey more easily certain ideas that are the true
basis of all our reasoning. I accept this, and am willing to
base the controversy merely on these ideas. So I ask our

ii: Ideas of space and time

mathematician to form as accurately as possible the ideas of
a circle and a straight line; and then I ask whether in his conception of their contact he can conceive them as touching at
a mathematical point, or whether instead he has to imagine
them to coincide for some space. Whichever side he chooses,
he runs himself into equal difficulties. •If he says that in
tracing these figures in his imagination he can imagine them
as touching only at a point, he allows the possibility of the
idea of a point, and thus the possibility of points. •If he says
that in his conception of the contact of those lines he must
make them coincide ·for some tiny distance·, he is implicitly
admitting the fallacy of geometrical demonstrations that are
carried beyond a certain degree of minuteness; for he certainly has such demonstrations against a circle’s coinciding
·for any distance· with a straight line. . . .

5: The same subject continued
·At the start of section 4, I pointed out that my account of

philosophers and scientists think they are free to join either
side in this controversy, as their fancy leads them. But
whatever basis there may be for a controversy about vacuum
and plenum themselves, it may be claimed—·and by Locke
it was claimed·—that the very ·existence of the· dispute is
decisive concerning the idea: men couldn’t possibly argue
for so long about a vacuum, and either oppose or defend it,
without having a notion of what they refuted or defended.

space and extension has two parts. I devoted that section to
the first part, namely the thesis that what is extended consists of indivisible parts. Now we come to· the second part
of my system, namely that the idea of space or extension is
nothing but the idea of visible or tangible points distributed
in a certain order. If that is true, it follows that we can
form no idea of a vacuum, or space where there is nothing
visible or tangible. This is met by three objections that I shall
examine together, because my answer to one of them is a
consequence of my answer to the other two.

Secondly, if this argument should be rejected, the reality
or at least the possibility of the idea of a vacuum can be
proved by the following reasoning. Every idea is possible
that is a necessary and infallible consequence of ones that
are possible. Now, even if we suppose the world to be at
present a plenum, we can easily conceive it to be deprived

First, it may be said that men have disputed for ages
about a vacuum and a plenum [= ‘space that is entirely full’]
without being able to reach a final decision, and even today
31

Treatise, Book 1

David Hume

of motion —this idea must be allowed as possible. It must
also be allowed as possible to conceive that God in his omnipotence annihilates some portion of matter while nothing
else moves. [For the rest of this paragraph Hume continues
to expound (in very Humean terms) this argument for the
possibility of vacuum; and to defend it against a certain
reply (that of Descartes), in order to set it up for his own
reply.] For as every idea that is distinguishable is separable
by the imagination, and as every idea that is separable by
the imagination may be conceived to be separately existent,
it is evident that •the existence of one particle of matter no
more implies •the existence of another than •one body’s having a square shape implies that •every body is square. This
being granted, I now ask what results from the concurrence
of these two possible ideas of rest and annihilation—what
must we conceive to follow from •the annihilation of all the
air and subtle matter [= ‘matter that is finer than air’] in a room,
supposing the walls to remain the same, without any motion
or alteration? There are some metaphysicians—·such as
Descartes·—who answer that since •matter and •extension
are the same, the annihilation of one necessarily implies
that of the other; so ·if there is now •no matter· between
the walls of the room there is now •no distance ·between
them either·; that is, they touch each other, just as my hand
touches the paper I am writing on. But though this answer
is very common, I defy these metaphysicians to conceive the
matter according to their hypothesis, or to imagine the floor
touching roof and the opposite walls touching each other
while nothing moves! . . . . If you change the position of the
roof, floor, and walls, you suppose •a motion; if you conceive
anything between them, you suppose •a new creation. But
keeping strictly to the two ideas of •rest and •annihilation, it
is obvious that the idea resulting from them is not that of a
contact of parts, but something else that is concluded to be

ii: Ideas of space and time

the idea of a vacuum.
The third objection carries the matter still further, and
contends not only that the idea of a vacuum is real and
possible but that it is necessary and unavoidable. This assertion is based on the motion we observe in bodies: this, it is
maintained, would be impossible and inconceivable without
a vacuum into which one body must move in order to make
way for another. I shan’t expound this objection at length,
because it principally belongs to physics, which lies outside
our present sphere.
In order to answer these objections I must dig pretty deep
and consider the nature and origin of various ideas, lest we
argue without perfectly understanding what we are arguing
about. The idea of darkness is obviously not a •positive one,
but merely the •negation of . . . . coloured and visible objects.
When a sighted man looks around him in complete darkness,
he receives no perceptions except ones he shares with someone born blind; and it is certain the latter has no idea either
of light or darkness. So the impression of extension without
matter couldn’t come from the mere removal of visible objects; the idea of utter darkness can never be the same as the
idea of vacuum. Now, suppose a man to be supported in the
air and to be—·without seeing or feeling anything·—gently
carried along by some invisible power; it is obvious that this
invariable motion doesn’t make him aware of anything, and
doesn’t give him the idea of extension or indeed any other
idea. Even if he moves his limbs to and fro, this can’t convey
that idea to him. He feels a certain sensation or impression,
the parts of which are successive to each other; they may
give him the idea of time, but certainly they are not laid out
in a way that could convey the idea of space or extension.
So it appears that darkness and motion, •in the absence
of everything visible and tangible, can’t give us the idea of
extension without matter, i.e. of a vacuum. So now we must
32

Treatise, Book 1

David Hume

ask: can they convey this idea •when mixed with something
visible and tangible? . . . . If we are to know whether sight
can convey the impression and idea of a vacuum, we must
suppose that in a complete darkness there are luminous
bodies presented to us, their light revealing only these bodies themselves and giving us no impression of surrounding
objects. And we have to form a parallel supposition about
touch. It won’t do to suppose a perfect absence of all tangible
objects: we must suppose that something is perceived by the
sense of touch. then after an interval and motion of the hand
or other sense-organ another tangible object is met with,
then another, and so on, as often as we please. The question
is: do these intervals give us the idea of extension without
body? To begin with the case of sight: it is obvious that
when only two luminous bodies appear to the eye we can
see whether they are conjoined or separate, and whether the
distance between them is large or small; and if that distance
changes, we can perceive it getting larger or smaller as the
bodies move. But in this case the distance is not anything
coloured or visible, so it may be thought that what we have
here is a vacuum or pure extension, not only intelligible to
the mind but obvious to the senses.

ii: Ideas of space and time

man gets from his eyes or what is conveyed to us in the
darkest night, it must have the same properties; and as
•blindness and darkness give us no ideas of extension, it is
impossible that the dark and undistinguishable •distance
between two bodies can ever produce that idea.
The sole difference between absolute darkness and the
appearance of two or more visible luminous objects consists,
as I said, in the objects themselves and how they affect our
senses. ·Don’t think that the distances are also perceived·.
Philosophers commonly agree that it is reason, more than the
senses, that . . . . tells us how far away from us a given body
is. The only perceptions from which we can (·by reasoned
inference·) judge the distances are
•the angles that the rays of light flowing from the objects form with each other,
•the motion the eye has to make when it goes from
looking at one object to looking at the next, and
•the different parts of the organs that are affected by
the light from each object.
But as each of these perceptions is simple and indivisible,
they can never give us the idea of extension.
We can illustrate this by considering the sense of touch,
and the imaginary distance or interval between tangible or
solid objects. I have supposed two cases:
•a man supported in the air and moving his limbs to
and fro without meeting anything tangible;
•a man who feels something tangible, leaves it, and
after a movement of which he is aware feels another
tangible object.
What is the difference between these two cases? No-one will
hesitate to reply that it consists merely in the perceiving
of those objects, and that the sensation arising from the
movement is the same in both cases. Well, that sensation
can’t give us an idea of extension when it isn’t accompanied

This is our natural and most familiar way of thinking, but
if we think a little we’ll learn to correct it. Notice that when
there is perfect darkness in which two ·luminous· bodies
present themselves, the only change that is revealed is the
appearance of these two objects; all the rest continues to be,
as before, a perfect negation of light and of every coloured or
visible object. This is true not only of what may be said to be
far away from these bodies but also of the very distance that
interposes between them; for all that consists of nothing but
darkness, or the negation of light—without parts, without
composition, unchanging and indivisible. Now, since this
distance causes no perception different from what a blind
33

Treatise, Book 1

David Hume

by some other perception, so it can’t give us that idea when
mixed with impressions of tangible objects, because that
mixture does not alter the sensation.
But although motion and darkness—alone or accompanied by tangible and visible objects—don’t convey •any idea
of vacuum or extension without matter, they are the causes
for •our falsely imagining we can form such an idea. For that
motion and darkness are closely related to a real extension,
a real complex of visible and tangible objects. ·There are
three components to this relation·.
First, we may observe that two visible objects appearing
in the midst of utter darkness •affect the senses in the same
way, •form the same angle by the rays that flow from them,
and •meet in the eye ·in the same way·, as if the distance
between them were filled with visible objects that would give
us a true idea of extension. Similarly, the sensation of motion when there is nothing tangible between two bodies is
the same as when we feel a complex body whose different
parts are outside one another.
Secondly, we find by experience that when •two bodies
so placed as to affect the senses in the same way as •two
others that have a certain extent of visible objects between
them, the former two can come to have the same extent of
visible objects between them without anything’s perceptibly
bumping into or penetrating anything else and without any
change in the angle they subtend at the eye. Similarly, when
there are •two objects of which we can’t feel both unless,
between the two feelings, time elapses and there is a sensation of movement in our hand, experience shows us that
•the two objects could be felt with the intervening time being
filled by that same sensation of hand-movement together
with impressions of solid and tangible objects. Summing up
these two points: an invisible and intangible distance can be
converted into a visible and tangible one without any change

ii: Ideas of space and time

in the distant objects.
Thirdly, these two kinds of distance have nearly the same
effects on every natural phenomenon. All qualities—heat,
cold, light, attraction, etc.— grow weaker as the distance increases; and we observe little difference ·in this effect· when
the distance is •marked out by compounded and perceptible
objects from what it is when the distance is •known only by
how the distant objects affect the senses.
So here are three relations between the distance that
conveys the idea of extension and that other distance that
isn’t filled with any coloured or solid object. •The distant
objects affect the senses in the same way, whether separated
by one distance or the other; •the second species of distance
is found to be capable of receiving the first; and •they both
equally lessen the strength of every quality.
These relations between the two kinds of distance easily
explain why one has so often been mistaken for the other,
and why we imagine we have an idea of extension without
the idea of any object either of sight or feeling. For we can
accept it as a general maxim in this science of human nature
that
whenever there is a close relation between two ideas,
the mind is very apt to mistake them, and to use one
in place of the other in all its discourses and reasonings.
This phenomenon occurs so often, and is so important, that
I can’t resist stopping for a moment to examine its causes.
Let me say in advance that the phenomenon mustn’t be confused with my account of its causes: if you have doubts about
my explanation of the phenomenon, don’t let them become
doubts about the phenomenon itself. It may be real even
if my explanation of it is chimerical. Though it is complete
wrong to do so, it is very natural for us to infer that something doesn’t exist from the falsity of a purported explanation
34

Treatise, Book 1

David Hume

ii: Ideas of space and time

ideas related together, but •the actions of the mind that we
employ in considering them are so alike that we can’t distinguish them. This fact is of great importance. Quite generally
we can say that whenever the actions of the mind in forming any two ideas are the same or very alike, we are apt to
confound these ideas and take the one for the other. We’ll
see many examples in the course of this book. But though
resemblance is the relation that most easily produces a mistake in ideas, the other two—causation and contiguity—can
also contribute to it. We could prove this with the examples
of poets and orators, if it were thought proper (it is certainly
reasonable) to draw arguments from that quarter in metaphysical subjects. But metaphysicians may think this to be
beneath their dignity, so I shall get a proof from an observation that can be made about most of the metaphysicians’ own
discourses—namely that it is usual for men to use •words
instead of •ideas, and to •talk instead of •thinking in their
reasonings. We use words in place of ideas because they
are commonly so closely connected that the mind easily mistakes them. This also explains why we substitute the idea
of a distance that is not taken to be visible or tangible for the
idea of extension, which is nothing but a complex of visible
or tangible points arrayed in a certain order. The relations of
causation and resemblance both contribute to this mistake.
As the first sort of distance is found to be convertible into the
second, it is in this respect a kind of cause; and the relation
of resemblance comes in through the similarity in how the
two sorts of distance affect the senses and diminish other
qualities.

of it; and the naturalness of that error is a clear instance of
the very principle that I am now about the explain! When ·in
section 4i · I accepted the relations of resemblance, contiguity, and causation as sources of union among ideas, doing
this without looking into their causes, I was busy pressing
my first maxim, that we must in the end rest contented with
experience; it wasn’t that I had nothing attractive and plausible to say on the subject of the causes. It would have been
easy to make an imaginary dissection of the brain, and to
show why on our conception of any idea the animal spirits
run into all the nearby channels and rouse up the other
ideas that are related to it. But though I passed up any
advantage that I might have gained from this line of thought
in explaining the relations of ideas, I’m afraid that I must
now have recourse to it so as to account for the mistakes
that arise from these relations.
The mind is endowed with a power of arousing any idea
it pleases: whenever it despatches the spirits into the region
of the brain containing a certain idea, they always arouse
the idea when they run precisely into the proper channels
and rummage the cell that belongs to it. But their motion
is seldom direct, and naturally turns a little to one side or
the other; and for this reason the animal spirits, falling into
nearby channels, present other related ideas instead of the
one the mind at first wanted to look at. Sometimes we aren’t
aware of this switch; we continue the same train of thought,
make use of the related idea that is presented to us, employing it in our reasoning as if it were the one we asked for. This
is the cause of many mistakes and sophisms in philosophy,
as you can imagine; and it would be easy to show this, if
there were any need to do so.
Of the three relations I have mentioned, resemblance is
the most fertile source of error; and indeed most mistakes in
reasoning owe a lot to that source. Not only are •resembling

After this chain of reasoning and explanation of my principles, I am now prepared to answer all the objections that
have been offered, whether derived from metaphysics or
physics. •The frequent disputes about vacuum, or extension
without matter, don’t prove the reality of the idea on which
35

Treatise, Book 1

David Hume

the dispute turns; for there is nothing more common than to
see men deceive themselves in this regard, especially when
some close relation presents them with another idea which
may be the occasion of their mistake.
We can make almost the same answer to •the second objection, derived from the conjunction of the ideas of rest and
annihilation. When everything in the room is annihilated,
and the walls don’t move, the chamber must be conceived
in much the same way as at present, when the air that fills
the room is not an object of the senses. This annihilation
leaves the eye with the fictitious distance that is revealed
by the different parts of the organ that are affected, and by
the degrees of light and shade; and it leaves to the sense of
touch the fictitious distance that consists in a sensation of
motion in the hand or other member of the body. It is no use
our looking further. On whichever side we turn this subject,
we shall find that these are the only impressions such an
object can produce after the supposed annihilation; and I
have already pointed out that impressions can give rise only
to ideas that resemble them.
Since we can suppose a body to be annihilated without
producing any change in its neighbours, we can easily conceive how a body might be created anew without affecting
anything else. Now, the motion of a body has much the
same effect as its creation: the distant bodies are no more
affected in one than in the other. This suffices to satisfy our
conceptual demands, and proves that there is no inconsistency in supposing such a motion. Afterwards experience
comes in play to persuade us that two bodies situated in
the manner described above really can receive ·a new· body
between them, and that there is no obstacle to converting
the invisible and intangible distance into one that is visible
and tangible. However natural that conversion may seem,
we can’t be sure that it is practically possible until we have

ii: Ideas of space and time

experience of it.
Thus I seem to have answered the three objections mentioned above [on pages 23– 25], though I realize that few people
will be satisfied with these answers, and most will immediately propose new objections and difficulties. It will probably
be said that my reasoning is irrelevant to the real question,
and that I explain only •how objects affect the senses, without trying to account for •their real nature and operations.
What I have said goes like this:
When there is nothing visible or tangible between two
bodies, we find by experience that the bodies can be
placed in the same manner, with regard to the eye and
hand-movement, as if they were divided by something
visible and tangible. This invisible and intangible distance is also found by experience to contain a capacity
of receiving body, i.e. of becoming visible and tangible.
That is the whole of my system; and nowhere in it (the complaint runs) have I tried to explain the cause that separates
bodies in this way, making them able to receive others between them, without any collision or penetration.
I answer this objection by pleading guilty, and by admitting that I never intended to penetrate into the nature
of bodies or explain the secret causes of their operations.
This is no part of my present purpose, and anyway I am
afraid that such an enterprise is beyond the reach of human
understanding, and that we shall never be able to claim to
know body otherwise than by the external properties that
reveal themselves to the senses. As for those who try to go
further: I can’t approve of their ambition until I see at least
one example of success in it. But at present I content myself
with knowing perfectly how objects affect my senses, and
knowing what experience tells me about their connections
with one another. This suffices for the conduct of life, and it
also suffices for my philosophy, which claims only to explain
36

Treatise, Book 1

David Hume

the nature and causes of our perceptions, i.e. impressions
and ideas.4
I shall conclude this subject of extension with a paradox
that the arguments I have given will easily explain. This
paradox is that •if you choose to give the name ‘vacuum’ to
distance of the invisible and intangible sort—in other words,
to the ability to become a visible and tangible distance—then
extension and matter are the same, and yet there is a vacuum! •If you choose not to give it that name, then motion
is possible in a plenum without collisions running on to
infinity or returning in a circle, and without penetration.
But however we express ourselves, we must always admit
that we have no idea of any real extension without filling it
with perceptible objects and conceiving them as visible or
tangible.
As for the doctrine that time is nothing but the manner
in which some real objects ·or events· exist: this is open to
the same objections as the similar doctrine regarding extension. If our disputing and reasoning about •·spatial· vacuum
is a sufficient proof that we have the idea of it, we must
for the same reason have the idea of time when nothing
happens—·that is, of •temporal vacuum·—because there is
no commoner subject of dispute. But it is certain that we
really don’t have any such idea. For where could it come
from? Does it arise from an impression of sensation or of
4

ii: Ideas of space and time

reflection? Point the source- impression out distinctly to us,
so that we can know its nature and qualities! But if you
can’t point out any such impression you may be certain that
you are mistaken in thinking you have any such idea.
But although it is impossible to show an impression from
which an idea of time without something that changes could
be derived, we can easily point out the appearances that
make us fancy we have that idea. We may observe that
there is a continual succession of perceptions in our mind,
so that the idea of time is always present to us; and when
we consider an unchanging object at five o’clock and then
again at six •we are apt to apply our idea of time to it in
the same way as if the object had been moving or altering
throughout. The first and second appearances of the object,
being compared with the succession of our perceptions, seem
as far apart ·in time· as if the object had really altered. To
this we may add, what experience shows us, •that between
these appearances the object was capable of such a number
of changes ·as we fictionally imagine it to have undergone·;
as •also that the unchanging or rather fictitious duration
has the same effect on every quality increasing or lessening
it—as does the succession that is ·real, because it is· obvious
to the senses. Because of these three relations we are apt to
confound our ideas, and imagine we can form the idea of a
time and duration without any change or succession.

As long as we confine our theorizing to the sensory appearances of objects, without getting into their real natures and operations, we are safe from all
difficulties and can never be embarrassed by any question. For example, if we are asked ‘Is the invisible and intangible distance between two objects
something or nothing?’ we can easily answer that it is something, namely a property of the objects that affect the senses in such and such a way. If we
are asked ‘When two objects have an invisible and intangible distance between them, do they touch or not?’, we can answer that this depends on the
definition of ‘touch’. If objects are said to touch when there is nothing perceptible placed between them, then these two objects touch. If objects are
said to touch when their images affect adjoining parts of the eye, and when the hand feels both objects successively without any interposed motion,
these objects do not touch. The appearances of objects to our senses are all consistent; and no difficulties can ever arise except from the obscurity
of the terms we employ.

37

Treatise, Book 1

David Hume

ii: Ideas of space and time

6: The ideas of existence and of external existence
It may be a good idea, before we leave this subject, to explain
the ideas of existence and of external existence, which have
their difficulties as well as the ideas of space and time. This
will help to prepare us for the examination of knowledge and
probability, when we understand perfectly all the particular
ideas that can enter into our reasoning.

Whatever we conceive, we conceive to be existent. Any idea
we please to form is the idea of a being; and the idea of a
being is any idea we please to form.
If you oppose this, you are obliged to point out the distinct
impression from which your idea of entity [= ‘existing thing’] is
derived, and to prove that this impression is inseparable
from every perception we believe to be existent. This, we can
say without hesitation, is impossible.

Every impression or idea of every kind, in consciousness
and in memory, is conceived as existent; and obviously the
most perfect idea . . . . of being is derived from this consciousness. This gives rise to a splendidly clear and conclusive
dilemma: that since we never remember any idea or impression without attributing existence to it, the idea of existence
must either be •derived from a distinct impression that is
conjoined with every perception or object of our thought or
be •the very same as the idea of the perception or object.

My reasoning ·in section 7i · about ·the so-called ‘distinction of reason’·—the distinction of ideas without any real
difference—won’t do anything for us here. That kind of distinction is based on the fact that a single simple idea may
resemble several different ideas ·in different respects·. But
no object can resemble a second object with respect to its
existence while differing from a third in that respect, since
every object that is presented ·as a candidate for comparison·
must necessarily be existent.

This dilemma is an obvious consequence of the principle
that every idea arises from a similar impression, so there is
no doubt about how we should choose between the horns of
the dilemma. So far from there being any distinct impression
attending every ·other· impression and every idea, I don’t
think that any two distinct impressions are inseparably conjoined. Though certain sensations may at one time be united,
we quickly find they can be separated and can appear apart.
And thus, though every impression and idea we remember
is considered as existent, the idea of existence is not derived
from any particular impression.

Similar reasoning will account for the idea of external
existence. It is a philosophical commonplace as well as a
pretty obvious truth that nothing is ever really present to the
mind except its perceptions—its impressions and ideas—and
that external objects become known to us only through the
perceptions they give rise to. To hate, to love, to think, to
feel, to see—all this is just to perceive.
Now, since nothing is ever present to the mind but perceptions, and since every idea is derived from something
that was previously present to the mind; it follows that we
can’t so much as conceive or form an idea of anything that is
specifically different [= ‘different in fundamental kind’] from ideas
and impressions. Look outside yourself as much as you can;
chase your imagination to the heavens or to the outer limits

The •idea of existence, then, is identical with •the idea
of whatever it is that we conceive to be existent. To reflect
on something •simply, and to reflect on it •as existent, are
exactly the same procedure. When the idea of existence is
conjoined with the idea of an object, it adds nothing to it.
38

Treatise, Book 1

David Hume

of the universe; you’ll never really advance a step beyond
yourself, and you can’t conceive any kind of existent other
than the perceptions that have appeared within the narrow
compass ·of your mind·. This is the universe of the imagination, and we have no ideas of anything that is not produced
there.
The furthest we can go towards a conception of external

ii: Ideas of space and time

objects, taking them to be specifically different from our perceptions, is to form a relative idea of them without claiming
to comprehend the objects themselves. Generally speaking,
we don’t suppose them to be specifically different; we take
them to differ from our perceptions only in respect of some
of their relations, connections, and durations. But of this
more fully hereafter—·in 2iv ·.

39

Treatise, Book 1

David Hume

iii: Knowledge and probability

Part iii: Knowledge and probability
1: Knowledge
There are (as I said in section 5i ) seven different kinds of
philosophical relation:

soning or reflection. Not even the simplest phenomenon can
be explained purely in terms of the qualities of the objects
as they appear to us, or be foreseen by us without the help
of our memory and experience.

resemblance
identity
relations of time and place
proportion in quantity or number
degrees in any quality
contrariety
causation.

It seems, then, that of these seven philosophical relations
there remain only four that can be the objects of knowledge and certainty because they depend solely on ideas,
. These four are resemblance, contrariety, degrees in quality, and proportions in quantity or number. Three of these
relations are discoverable at first sight, and belong in the
province of intuition rather than of demonstration. [In Hume’s

These relations can be divided into two classes. •In one
class are the ones that depend entirely on the ideas that we
compare together, ·so that the relation can change only if
the ideas change·. •In the other class are relations that can
be changed without any change in the ideas. •The idea of a
triangle shows us the relation of equality that its three angles
have to two right angles, and this relation is invariable as
long as our idea remains the same. On the other side, the relations of contiguity and distance between two objects can be
changed merely by moving the objects, without any change
in them or in their ideas; and how things move depends
on a hundred different events that can’t be foreseen by the
mind. Similarly with identity: two objects can be numerically
different from each other—·that is, can really be two ·—even
though they perfectly resemble each other, and even if they
appear at different times in the very same place. And with
causation: the power by which one object produces another
can never be discovered merely from the ideas of the objects;
so it is evident that cause and effect are relations that we
learn about from experience and not from any abstract rea-

day, ‘intuition’ stood for ‘seeing something, straight off, as self-evidently
true’; while ‘demonstration’ is the procedure of proving something by rigorously valid argument, each step in which is warranted by intuition.]

When two objects resemble each other, the resemblance will
immediately strike the eye, or rather the mind, and seldom
needs a second look. Similarly with contrariety: no-one can
doubt for a moment that existence and non- existence destroy each other and are perfectly incompatible and contrary.
And with the degrees of any quality: although it is impossible to judge exactly concerning degrees of a quality—such as
colour, taste, heat, cold—when the difference between them
is very small, it is easy to decide which is the more intense
when their difference is considerable. And we pronounce
this decision at first sight, without any enquiry or reasoning.
We can proceed in the same way in fixing the proportions of quantities or numbers: where the difference is very
great and remarkable, we can see at a glance which figure
or number is the larger of two. As to equality or any exact
40

Treatise, Book 1

David Hume

proportion—·that is, any judgment about exactly how much
larger one item is than another·—a single look will yield us
only a guess, except with very small numbers or very limited
portions of extension, which can be taken in all at once and
where we perceive that we can’t fall into any considerable
error. In all other cases we must settle for approximations,
or else proceed in a more artificial manner.
I have already observed, ·near the middle of 4ii ·, that
geometry, or the technique by which we fix the proportions
of figures, never achieves perfect precision and exactness
(though its results are much more general and exact than the
loose judgments of the senses and imagination). Its first principles are drawn from the general appearance of the objects,
and when we know something of the prodigious minuteness
of which Nature is susceptible we can’t feel secure about
general appearances! Our ideas seem to give us a perfect
assurance that no two straight lines can have a common
segment; but if you attend to the ideas that we have when
we think this you’ll find that they always suppose the two
lines to be inclining perceptibly towards one another, ·so that
the angle between them is fairly large·. When the angle they
form is extremely small we have no standard of straight line
precise enough to assure us of the truth of this proposition.
It is the same with most of the primary decisions [Hume’s
phrase] of mathematics.
There remain, therefore, algebra and arithmetic as the
only sciences in which we can carry on a chain of reasoning
to any degree of intricacy while preserving perfect exactness
and certainty. We have a precise standard by which to judge
concerning the equality and proportion of numbers; and on
the basis of that standard we can determine the relations
between numbers without any possibility of error. When
two numbers are brought together so that each always has
a unit answering to every unit of the other, we pronounce

iii: Knowledge and probability

them ‘equal’. The reason why geometry doesn’t quite qualify
as a perfect and infallible science is that it doesn’t have a
comparable standard of equality in size.
But it may be as well here to remove a difficulty that
could arise from my asserting that, though geometry falls
short of the perfect precision and certainty that arithmetic
and algebra have, it still excels the imperfect judgments of
our senses and imagination. The reason why I attribute any
defect to geometry is that its first basic principles are derived
merely from appearances; and you might think that this
defect must follow it all the way through, preventing it from
ever being able to compare objects or ideas more exactly
than we can by relying purely on our eye or imagination. I
accept that this defect follows it far enough to prevent it from
ever aspiring to full ·exactness or· certainty: but since its
basic principles depend on the easiest and least deceitful
appearances, they give to their consequences a degree of
exactness that the consequences couldn’t have if they were
taken singly. It is impossible to see by looking that the angles
of a thousand-sided figure are equal to 1996 right angles, or
to guess at anything remotely like this result; but when the
eye determines that straight lines cannot coincide, and that
we can’t draw more than one straight line between two given
points, its mistakes can never be of any consequence. And
this is the nature and use of geometry, to take us back to
appearances which, because of their simplicity, can’t lead us
into any considerable error.
I shall take this opportunity to offer a second observation
about our demonstrative reasonings. . . . It is usual with
mathematicians to claim that the ideas that are their objects
are so refined and spiritual that they can’t •be conceived in
the imagination but must •be comprehended by a pure and
intellectual view of which only the higher faculties of the soul
are capable. The same notion runs through most parts of
41

Treatise, Book 1

David Hume

philosophy, and is principally made use of to explain our
abstract ideas, and to show how we can form an idea of a
triangle, for instance, which is to be neither isosceles nor
scalar, nor confined to any particular length or proportion of
sides. It is easy to see why philosophers are so fond of this
notion of ‘spiritual and refined’ perceptions, since it helps
them to cover up many of their absurdities, and lets them
refuse to submit to the decisions of clear ideas by appealing
to ideas that are obscure and uncertain ·though ‘spiritual
and refined’·! To destroy this trick we need only to reflect on
the principle I have stressed so often, that all our ideas are

iii: Knowledge and probability

copied from our impressions. From that we can immediately
conclude that since all impressions are clear and precise,
the ideas copied from them must be clear and precise too, so
that it’s our own fault if they ever contain anything dark and
intricate. An idea is by its very nature weaker and fainter
than an impression; but being in every other respect the
same, it can’t bring with it any very great mystery. If its
weakness makes it obscure, it is our business to remedy
that defect as much as possible by keeping the idea steady
and precise; and till we have done that it’s pointless for us
to engage in reasoning and philosophy.

2: Probability, and the idea of cause and effect
I think that’s all I need to say about those four relations that
are the foundation of science; but there is more to be said
in detail about the other three—the ones that don’t depend
on the ideas, and can be absent or present even while the
ideas remain the same. These three relations are identity,
situations in time and place, and causation.

but only a passive allowing in of the impressions through
the organs of sensation. According to this way of thinking,
we ought not to classify as ‘reasoning’ any observations we
make about identity or relations of time and place; for in
none of those does the mind go beyond what is immediately
present to the senses, whether to discover the real existence
of ·other· objects or to discover the relations between them.
Only causation produces a connection that can assure us,
on the basis of the existence or action of one object, that
some other existence or action followed or preceded it. And
the other two relations—identity, and location in time and
space—can be used in reasoning only to the extent that they
affect or are affected by causation. •There is nothing in any
objects to persuade us that they are either always distant or
always close; and when from experience and observation we
discover that their spatial relation doesn’t change, we always
conclude that some secret cause is separating or uniting
them. The same reasoning extends to identity. •We readily
suppose that an object can continue individually the same—

All kinds of reasoning consist in nothing but a comparison, and a discovery of the relations—constant or changing—
that two or more objects have to one another. [In Hume’s time,
‘comparing’ two things could be simply bringing them together in a single
thought, not necessarily thinking about their being alike. The present
section seems to use the word sometimes in that broader, weaker sense
of ‘compare’ and sometimes in the narrower sense that is common to-

We can make such a comparison when both the objects
are present to the senses, or when neither is present, or
when only one is. When both the objects are present to the
senses along with the relation ·that holds between them·,
we call this ‘perception’ rather than ‘reasoning’: in this case
there is no exercise of thought, no action properly so-called,
day.]

42

Treatise, Book 1

David Hume

·that is, can continue to be that very same object·—even if

iii: Knowledge and probability

turn them on all sides in order to find the impression that
produces this prodigiously important idea. I see straight off
that I mustn’t search for it in any of the particular qualities
of the objects: whichever of these qualities I pick on, I find
some object that doesn’t have it and yet does fall under the
label of ‘cause’ or ‘effect’. And indeed everything that exists,
whether externally or internally, can be considered as either
a ‘cause’ or an ‘effect’, though it is plain that no one quality
universally belongs to all beings and gives them a title to
that label.
So the idea of causation, ·since it doesn’t come from any
•quality·, must be derived from some •relation among objects; and that relation is what we must now try to discover.
The first thing I find is that only contiguous pairs of objects [=
‘immediate neighbours’] are considered as cause- effect related,
and that nothing can •operate at a time or in a place other
than—even if extremely close to—the time and place that
it •exists in. It sometimes seems that one object acts on
another that is at a distance from it, but they are commonly
found on examination to be linked by a chain of causes,
with each link contiguous to the next, and the end links
contiguous to the distant objects; and in any particular case
where we can’t discover such a chain we still presume it
to exist. So we can take it that contiguity is essential to
causation; at least we can suppose it to be so, according to
the general opinion, until we can find a better occasion—in
section 5iv —to clear up this matter by examining what objects are and what are not capable of being brought together
and conjoined.
The second relation that I shall claim to be essential to
causes and effects is not so universally acknowledged ·as
contiguity·, being a subject of some controversy. It is the relation of the cause’s priority in time to the effect. Some claim
that it is not absolutely necessary for a cause to precede its

in our perception it comes and goes; we attribute to it an
identity, despite the interruption of the perception, as long as
we conclude that if we had kept our eye or hand constantly
on it would have given us an invariable and uninterrupted
perception. But this conclusion ·about what would have
happened· goes beyond the impressions of our senses and
has to be based on the connection of cause and effect; and
we need cause and effect if we are to be sure that the object
has not been switched on us, however much the new object
may resemble the one that formerly appeared to the senses.
Whenever we discover such a perfect resemblance, we consider whether it is common in that kind of object; whether
possibly or probably any cause could be at work producing
the switch and the resemblance; and our judgment about
the identity of the object is based on the answers to these
questions concerning causes and effects.
So we find that of the three relations that don’t depend
purely on the ideas, the only one that can be traced beyond
our senses, and that informs us of existences and objects
that we don’t see or feel, is causation. So I shall try to explain this relation fully before we leave the subject of the
understanding. ·This explanation will occupy most of the
remainder of Part iii of this work·.
To begin in an orderly fashion, we must consider the idea
of causation and see from what origin it is derived. It is impossible •to reason soundly without understanding perfectly
the idea about which we reason; and it is impossible •to
understand an idea perfectly without tracing it back to its
origin and examining the primary impression from which it
arises. •The examination of the impression gives clearness
to the idea, and •the examination of the idea gives a similar
clearness to all our reasoning.
Take any pair of objects that we call cause and effect, and
43

Treatise, Book 1

David Hume

effect, and that any object or action can in the very first moment of its existence exert its productive quality, giving rise
to another object or action that is absolutely simultaneous
with it. But experience in most instances seems to contradict
this opinion, and anyway we can may establish ·the essentialness of· the relation of priority by a kind of inference or
reasoning, ·as follows·. It is an established maxim, both in
physics and the human sciences, that
an object O1 that exists for some time in its complete
state without producing another object O2 is not the
sole cause of O2 ·when it does occur·, but is assisted
by some other factor that pushes O1 from its state
of inactivity and makes it exert the energy which it
secretly possessed.
Now if any cause could be absolutely simultaneous with its
effect, it is certain, according to this maxim, that all causes
must be simultaneous with their effects; for any one of them
that holds back its operation for a single moment doesn’t exert itself at the very time at which it might have operated, and
so it is not the whole cause of the effect. The consequence
of this would be nothing less than the destruction of the
succession of causes that we observe in the world—indeed,
the utter annihilation of time. For if one cause were simultaneous with its effect, and this effect with its effect, and so
on, there would plainly be no such thing as succession, and
all objects would be coexistent.
If you find this argument satisfactory, good! If not, I ask
you to allow me the same liberty that I took in the preceding
case, of supposing it to be satisfactory. You will find that the
affair is of no great importance.
Having thus discovered or supposed the two relations
of contiguity and succession to be essential to causes and
effects, I find myself stopped short: this is as far as I can
go if I attend only to single instances of cause and effect.

iii: Knowledge and probability

When bodies collide, we think that the motion in one causes
motion in the other; but when we consider these objects with
the utmost attention, we find only that one body comes up to
the other, and that the former’s motion precedes the latter’s,
though without any interval that we can perceive. It does
no good for us to rack ourselves with further thought and
reflection on this individual case: we have said all we can
about it.
You might want to stop looking at particular cases and
define ‘cause’ as ‘something that is productive of something
else’; but this doesn’t say anything. For what would you
mean by ‘production’? Could you define it except in terms of
causation? If you can, please produce the definition. If you
can’t, you are here going in a circle, producing merely one
synonymous term instead of a definition.
Shall we then rest contented with •contiguity and
•succession as providing a complete idea of causation? By
no means! One object can be contiguous and prior to another without being thought to be its cause. There is also
a •necessary connection to be taken into account, and that
relation is much more important than either of the others.
So I return to the particular case—·for example, the
collision·—and look at it from all angles trying to discover
the nature of this necessary connection by finding the impression(s) from which the idea of it could be derived. When I
cast my eye on the known •qualities of objects, I immediately
find that the relation of cause and effect doesn’t depend in
the least on them. When I consider the •relations between
them I can find only contiguity and succession, which I have
already regarded as imperfect and unsatisfactory. Should I
despair of success, and accept that what I have here is an
idea that is not preceded by any similar impression? That
would be strong evidence of light-mindedness and instability,
given that the contrary principle has already been so firmly
44

Treatise, Book 1

David Hume

established as to admit of no further doubt—at least until
we have more fully examined the present difficulty.
So we must proceed like someone who, having searched
for something and not found it where he expected, beats
about all the neighbouring fields with no definite view or
plan, hoping that sheer good luck will eventually guide him
to what he is looking for. We have to leave the direct survey of
this question about the nature of the necessary connection
that enters into our idea of cause and effect (·returning to
it at the start of section 14·), and try instead to find some
other questions the answering of which may afford a hint on
how to clear up the present difficulty. I shall examine two
such questions [the second question is here considerably expanded
from Hume’s formulation of it]:
What is our reason for holding it to be necessary that
everything whose existence has a beginning also has

iii: Knowledge and probability

a cause?
Why do we conclude that causes of kind K1 must necessarily have effects of kind K2 , and what is going on
when from the occurrence of a K1 we infer that a K2
will occur, and how does it happen that we believe the
predictions generated by such inferences?
Before going further, I should remark that although the ideas
of cause and effect are derived from impressions of reflection
as well as of sensation, for brevity’s sake I usually mention
only the latter as the origin of these ideas. Whenever I say
anything about impressions of sensation, please take it to
be said about impressions of reflection as well. Passions
are connected with their objects and with one another just
as much as external bodies are connected together. So the
same relation of cause and effect that belongs in the external
world belongs in the mind as well.

3: Why a cause is necessary
To begin with the first question, about the necessity of a
cause ·of coming into existence·: It is a general maxim in
philosophy that whatever begins to exist must be caused to
do so. This is commonly taken for granted in all reasonings,
without any proof being given or asked for. It is supposed
to be based on intuition, and to be one of those ·immediately
self-evident· maxims that men can’t really doubt in their
hearts, even if they deny them with their lips. But if we
examine this maxim in terms of the idea of knowledge that
I have explained, we shan’t discover in it any mark of any
such intuitive certainty. Quite the contrary: we’ll find that it
is of a nature quite foreign to what can be known intuitively.

as the ideas don’t change. These relations are resemblance,
proportions in quantity and number, degrees of any quality,
and contrariety, none of which is involved in the proposition
Whatever has a beginning has also a cause of existence. So
that proposition is not intuitively certain. At any rate, if you
want to maintain that it is intuitively certain you must deny
that these four are the only infallible relations, and must
find some other infallible relation to be involved in the proposition we are examining. When you do that, we can look at
it! Anyway, here is an argument that proves at one blow that
our proposition is not intuitively or demonstrably certain. To
demonstrate that (1) there must be a cause for every new
coming-into-existence and for every alteration of something
already in existence, we would have to show that (2) it is en-

All certainty arises from the comparison of ideas, and
from the discovery of such relations as don’t change so long
45

Treatise, Book 1

David Hume

tirely impossible for anything to begin to exist without some
productive force ·making it do so·; so if (2) can’t be proved,
we have no hope of ever being able to prove (1). And (2) is
utterly incapable of demonstrative proof, as we can assure
satisfy ourselves by considering that as •all distinct ideas
are separable from each other, and as •the ideas of ·a given·
cause and ·of its· effect are evidently distinct, we can easily
conceive an object coming into existence without bringing in
the distinct idea of a cause or productive principlec . So the
separation of the idea of a cause from that of a beginning of
existence is plainly possible for the imagination; and consequently the actual separation of these items is possible to the
extent that it doesn’t imply any contradiction or absurdity;
and so it can’t be refuted by any reasoning from mere ideas,
without which it is impossible to demonstrate the necessity
of a cause.

iii: Knowledge and probability

absurd in the one case, it must be so in the other; and if
the absurdity isn’t clear without a proof in the one case, it
will equally require a proof in the other. So there can be no
question of showing the absurdity of one supposition and
inferring from that the absurdity of the other; for they are
both on the same footing and must stand or fall by the same
reasoning.
The second argument that is used on this topic (by Dr
Clarke and others) runs into similar trouble. It goes like this:
Everything must have a cause; for if anything lacked
a cause it would produce itself, i.e. exist before it
existed, which is impossible.
But this reasoning is plainly invalid, because it assumes that
something’s lacking any cause involves it having a cause,
namely itself. No doubt the notion of a thing’s bringing itself
into existence is an evident contradiction. But to say that
something comes into existence without a cause is not to say
that it is itself its own cause! On the contrary, in excluding
all external causes the statement excludes the thing itself
that comes into existence. An object that exists absolutely
without any cause is certainly not its own cause; and when
you assert that the one follows from the other you are taking
for granted the very point that is in question . . . . Exactly the
same trouble infects the third argument that has been used
by Mr Locke to demonstrate the necessity of a cause:
Whatever is produced without any cause is produced
by nothing, i.e. has nothing for its cause. But nothing
can never be a cause, any more than it can be something, or be equal to two right angles. By the same
intuition that we perceive that nothing is not equal to
two right angles, and that nothing is not something,
we perceive that nothing can never be a cause; and
this forces us to see that every object has a real cause
of its existence.

Accordingly, when we look into the demonstrations that
have been adduced to show the necessity of a cause we shall
find them all to be fallacious and sophistical. ·I shall show
this with respect to the three main ones·. Some philosophers
(including Mr Hobbes) argue like this: all the points of time
and place in which we can suppose any object to come into
existence are in themselves equal; and unless there is some
cause that is special to one time and to one place, and by that
means determines and fixes the coming- into-existence, the
‘Where?’ question must remain eternally unanswered, and
the object can’t come into existence because there is nothing
to fix where and when it will do so. But I ask: Is it any
harder to suppose the time and place to be fixed without a
cause than to suppose the coming into existence of the object
to be determined without a cause? The first question that
comes up on this subject is always Will the object come into
existence or not?, and the second is When and where will it
come into existence? If the removal of a cause is intuitively
46

Treatise, Book 1

David Hume

I don’t think I need employ many words in showing the weakness of this argument, after what I have said of the other
two. All three are based on the same fallacy, and are derived
from the same turn of thought. I need only to point out that
when we exclude all causes we really do exclude them: we
don’t suppose that nothing or the object itself causes of the
object to come into existence; so we can’t argue from the absurdity of those suppositions to the absurdity excluding all
causes. . . . Even more frivolously, some say that every effect
must have a cause because having-a-cause it is implied in
the very idea of effect. It is true that every effect must have
a cause, because ‘effect’ is a relative term of which ‘cause’ is
the correlative. But this doesn’t prove that everything real
must be preceded by a cause, any more than it follows from
‘Every husband must have a wife’ that every man must be
married. The right question to be asking is: Must every item
that begins to exist owe its existence to a cause? I hope

iii: Knowledge and probability

that by the foregoing arguments I have shown well enough
that the answer Yes is neither intuitively nor demonstratively
certain.
So the opinion of the necessity of a cause for every new
production isn’t based on ·a priori· knowledge or scientific
reasoning, and must therefore arise from observation and
experience. The natural next question is: how does it arise
from experience? But I shall postpone that for a while, because I find it more convenient to sink this question in two
others:
•Why do we conclude that such-and-such particular

causes must necessarily have such-and-such particular effects?
•Why do we form an inference from cause to effect?
It may turn out eventually that a single answer will serve for
both questions.

4: The component parts of our reasonings about cause and effect
Although the mind in its reasonings from causes or effects
carries its view beyond the objects that it sees or remembers,
it must never lose sight of them entirely; it mustn’t reason
merely on its own ideas, without some mixture of impressions (or at least of ideas of the memory, which are equivalent
to impressions). When we infer effects from causes, we must
establish the existence of the causes; which we have only
two ways of doing. We can do it either by •an immediate
perception of our memory or senses, or by •an inference from
other causes; but then we must ascertain the existence of
these in the same way, either by a present impression ·or
memory· or by an inference from their causes, and so on
·backwards· until we arrive at some object that we see or

remember. We can’t carry on our inferences ad infinitum;
and the only thing that can stop them is an impression of
the memory or senses. Beyond that there is no room for
doubt or enquiry.
For an example, choose any point of history, and consider why we either believe or reject it. Thus, we believe that
Caesar was killed in the senate-house on the ides of March,
because this is established on the unanimous testimony
of historians, who agree in assigning this precise time and
place to that event. Here are certain words that we see or
remember, words that we remember to have been used as
the signs of certain ideas; and these ideas—·the ones in the
minds of writers of the history books·—were those of people
47

Treatise, Book 1

David Hume

who •were immediately present at assassination and received
their ideas directly from it, or who •got their ideas from the
testimony of others, who relied on yet earlier testimony, and
so on backwards until the slope stops at those who saw the
assassination. It is obvious that all this chain of argument or
connection of causes and effects is initially based on words
that are seen or remembered, and that without the authority
of either the memory or senses our whole reasoning would be
chimerical and without foundation: every link of the chain
would hang on another; but there would be nothing fixed
to one end of it that could support the whole chain, and so
there would be no belief. And this is actually the case with
all hypothetical arguments, or reasonings from a supposi-

iii: Knowledge and probability

tion, for in them there is no present impression and no belief
about a matter of fact.
You may want to object: ‘We can reason from our past
conclusions or principles without having recourse to the impressions from which they first arose.’ This is true, but not a
sound objection; for even if those impressions were entirely
wiped from the memory, the belief they produced may still remain. All reasonings about causes and effects are originally
derived from some impression; just as one’s confidence in
a demonstration always comes from a comparison of ideas,
though the confidence may continue after the comparison
has been forgotten.

5: The impressions of the senses and memory
In this kind of reasoning from causation, then, we use materials that are of a mixed and heterogeneous nature: however
inter- connected they are, they are still essentially different
from each other. All our arguments about causes and effects
consist of •an impression of the memory or senses, and of
•the idea of the real object or event that ·we think· caused
or was caused by the object of the impression. So we have
here three things to explain: •the original impression, •the
transition ·from that· to the idea of the connected cause or
effect, and •the nature and qualities of that idea.

whether they are true or false, whether they represent Nature
justly or are mere illusions of the senses.
When we search for the feature that distinguishes memory from imagination, we see straight off that it can’t lie in
the simple ideas they present to us; for both these faculties
borrow their simple ideas from •impressions, and can’t ever
get beyond •those original perceptions. Nor are memory and
imagination distinguished from one another by how their
complex ideas are arranged. It is indeed a special property
of the memory to preserve the original order and position of
its ideas—·or, more strictly speaking, to preserve its ideas
in the order of the original corresponding impressions·—
whereas the imagination transposes and changes its ideas
as it pleases. But this difference is not sufficient to tell us
whether in any given case we have memory or imagination;
for it is impossible to bring back the past impressions in order to compare them with our present ideas and see whether

As for the impressions that arise from the senses: in my
opinion their ultimate cause is utterly inexplicable by human
reason; we will never be able to decide with certainty whether
•they arise immediately from the object, or •are produced
by the creative power of the mind, or •are caused by God.
But this question doesn’t affect our present purpose. We
can draw inferences from the coherence of our perceptions,
48

Treatise, Book 1

David Hume

the arrangements are exactly alike. So the memory is not
known by •the nature of its simple ideas or •the order of its
complex ones; so the difference between it and imagination
must lie in •memory’s greater force and liveliness. You can
indulge your fancy by imagining a past scene of adventures;
and you couldn’t distinguish this from a memory of those
events if it weren’t that the ideas of the imagination are
fainter and more obscure.
It often happens that when two men have been involved
in a course of events, one remembers it much better than
the other and has great trouble getting his companion to
recollect it. He recites various details—the time, the place,
who was there, what they said, what they did—all with no
result, until finally he hits on some lucky circumstance that
revives the whole affair and gives his friend a perfect memory
of everything. Here the person who forgets receives all his
ideas ·of the event· at first from what his friend says; he has
the right ideas of the circumstances of time and place ·and
so on·, though to him they are mere fictions of the imagination. But as soon as the detail is mentioned that triggers his
memory, those very same ideas now appear in a new light,
and in a way feel different from how they did before. Without
altering in any way except in how they feel, they immediately
become ideas ·not of imagination but· of memory, and are
assented to.
Since the imagination can represent all the same objects
that the memory can offer to us, and since those ·two· faculties are distinguished only by how the ideas they present
feel, we ought to consider what the nature is of that feeling.
I think everyone will readily agree with me that the ideas
of the memory are stronger and livelier than those of the
imagination.
A painter wanting to represent a passion or emotion of

iii: Knowledge and probability

some kind would try to get a sight of a person in the grip
of that emotion, in order to enliven his ideas of it and give
them more force and liveliness than is found in ideas that
are mere fictions of the imagination. The more recent this
memory is, the clearer is the idea; and when after a long time
he wants to think again about that passion, he always finds
his idea of it to be much decayed if not wholly obliterated.
We are frequently in doubt about ideas of the memory when
they become very weak and feeble; and can’t decide whether
an image comes from the imagination or from the memory
when it is not drawn in colours that are lively enough to
point ·certainly· to the latter faculty. . . . .
As an idea of the memory can by losing its force and liveliness degenerate so far that it is taken to be an idea of the
imagination, so on the other hand an idea of the imagination
can acquire such force and liveliness that it passes for an
idea of the memory and has a counterfeit effect on belief and
judgment. We see this in liars who by frequently repeating
their lies eventually come to believe them, ‘remembering’
them as realities. In this case, as in many others, •custom
and habit have the same influence on the mind as •Nature
does, and implant the idea with the same force and vigour.
It appears, then, that the •belief or assent that always
accompanies the memory and senses is nothing but the
•liveliness of the perceptions they present, and that this is
all that distinguishes them from the imagination. In such
cases, believing is feeling an immediate impression of the
senses or a repetition of that impression in memory. It is
simply the force and liveliness of the perception that constitutes the basic act of judgment, laying the foundation for the
reasoning that we build on it when we track the relation of
cause and effect.

49

Treatise, Book 1

David Hume

iii: Knowledge and probability

6: The inference from the impression to the idea
are remembered; but whenever we •reason about them, only
one is perceived or remembered, and the other is supplied
on the basis of our past experience.

It is easy to see that when we think our way along this relation, the inference we make from cause to effect is not based
merely on probing these particular objects and learning
enough about their inner natures to see why one depends
on the other. If we consider these objects in themselves
and never look beyond the ideas we form of them, we shall
find that none of them implies the existence of anything
else. Such an inference—·based purely on the ideas·—would
amount to knowledge, and would imply the absolute contradiction and impossibility of conceiving anything different,
·that is, of conceiving the predicted effect not to follow·. But
clearly there can’t be any impossibility of that kind, because all distinct ideas are separable. Whenever we pass
·inferentially· from a present impression to the idea of some
other object, we could have separated the idea from the impression and have substituted any other idea in place of
it.

Thus, in moving on through our topic we have suddenly
come upon a new relation between cause and effect—finding
this when we least expected it and were entirely employed
on another subject. This relation is the constant conjunction of cause with effect. Contiguity and succession are not
sufficient to make us regard two objects as cause and effect unless we see that these two relations are preserved in
a number of instances. Now we can see the advantage of
leaving the direct survey of the cause-effect relation in order
to discover the nature of the necessary connection that is
such an essential part of it. Perhaps by this means we may
at last arrive at our goal! But, to tell the truth, this newly
discovered relation of constant conjunction doesn’t seem to
take us far along our way. ·Here is an expansion of that
pessimistic thought·:

So it is purely by experience that we can infer the existence of one object from that of another. The experience goes
like this. We remember having had frequent instances of the
existence of one sort of object, and also remember that individuals of another sort have always gone along with them,
regularly occurring just after them and very close by. Thus
we remember seeing the sort of object we call ‘flame’ and
feeling the sort of sensation that we call ‘heat’. We recall also
their constant conjunction in all past instances—·always
flame-then-heat·. Without more ado we call the one ‘cause’
and the other ‘effect’, and infer the existence of the heat from
that of the flame. In all the instances from which we •learn
the conjunction of particular causes and effects, both the
causes and effects have been perceived by the senses and

The fact of constant conjunction implies only that
similar objects have always been placed in similar
relations of contiguity and succession; and it seems
evident that this can’t reveal any new idea; it can
make our ideas more numerous, but can’t make them
richer. What we don’t learn from one object we can’t
learn from a hundred that are all of the same kind and
are perfectly alike in every detail. Our senses show us
in one instance two bodies (or motions or qualities) in
certain relations of succession and contiguity, and our
memory presents us with a multitude of cases where
we have found similar bodies (or motions or qualities)
related in the same ways. The mere repetition of a
50

Treatise, Book 1

David Hume

past impression—even to infinity—won’t give rise any
new original idea such as that of a necessary connection; and the sheer number of impressions has in this
case no more effect than if we confined ourselves to
one only.
But although this reasoning seems sound and obvious, it
would be folly for us to despair too soon. So I shall continue
the thread of my discourse: having found that after the discovery of the constant conjunction of any objects we always
draw an inference from one object to another, I shall now
examine the nature of that inference, and of the transition
from the impression to the idea. Perhaps we shall eventually
find that •the necessary connection depends on the inference
rather than •the inference’s depending on the necessary connection! It appears that the transition from an impression
that is present to the memory or senses (and said to be of
a ‘cause’) to the idea of an object (which is said to be an
‘effect’) is founded on past experience, and on our memory
of their constant conjunction. So the next question is: how
does experience produce the idea ·of the effect·? Is it done by
the •understanding or by the •imagination? Are we caused
to make the transition by •our reason or by •some ·nonreasoned· association and relation of perceptions? ·I shall
start with the former suggestion, giving it about a couple of
pages·.
If reason did the work, it would have to be relying on the
principle that
Instances of which we haven’t had experience must
resemble those of which we have; the course of Nature
continues always uniformly the same.
In order to clear this matter up, therefore, let us consider
all the arguments that might be given to support such a
proposition. They will have to be based either on •·absolutely
certain· knowledge or on •probability; so let us look into each

iii: Knowledge and probability

of these degrees of certainty, to see whether either provides
us with a sound conclusion along these lines.
My previous line of reasoning will easily convince us that
no demonstrative arguments could prove that instances of
which we have had no experience resemble those of which
we have had experience. We can at least conceive a change
in the course of Nature; which proves that such a change is
not absolutely impossible. To form a clear idea of anything
is an undeniable argument for its possibility, and can all on
its own refute any claimed demonstration against it.
Probability doesn’t concern the relations of ideas as such,
but rather the relations among objects; so it must be based
in some way on the impressions of our memory and senses,
and in some way on our ideas. If our probable reasonings
didn’t have any •impressions mixed into them, their conclusions would be entirely chimerical: and if there were there
no •ideas in mixture, the action of the mind in observing the
relation—·that is, in taking in that such-and-such makes
so-and-so probable ·—would strictly speaking be sensation,
not reasoning. In all probable reasonings, therefore, there is
•something present to the mind that is either seen or remembered, and from this we infer •something connected with it
that is not seen nor remembered.
The only connection or relation of objects that can lead
us beyond the immediate impressions of our memory and
senses is that of cause and effect, because it is the only one
on which we can base a sound inference from one object
to another. The idea of cause and effect is derived from
experience, which informs us that certain specific ·kinds of·
objects have always been constantly conjoined with each
other; and as an object of one of these kinds is supposed
to be immediately present through an impression of it, we
on that basis expect there to be an object of the other kind.
According to this account of things—which I think is entirely
51

Treatise, Book 1

David Hume

unquestionable—•probability is based on •the presumption
that the objects of which we have had experience resemble
those of which we have had none; so •this presumption can’t
possibly arise from •probability. One principle can’t be both
the cause and the effect of another. This may be the only
proposition about the causal relation that is either intuitively
or demonstratively certain!
You may think you can elude this argument. You may
want to claim that all conclusions from causes and effects are
built on solid reasoning, saying this without going into the
question of whether our reasoning on this subject is derived
from demonstration or from probability. Well, please produce
this reasoning so that we can examine it. You may say that
after experience of the constant conjunction of certain ·kinds
of· objects we reason as follows:
This kind of object is always found to produce an
object of that kind. It couldn’t have this effect if it
weren’t endowed with a power of production. The
power necessarily implies the effect; and therefore
there is a valid basis for drawing a conclusion from
the existence of one object to the existence of another.
The •past production implies a •power; the •power
implies a •new production; and the new production is
what we infer from the power and the past production.
It would be easy for me to show the weakness of this reasoning •if I were willing to appeal to the observations I have
already made, that the idea of production is the same as
the idea of causation, and that no existence certainly and
demonstratively implies a power in any other object; or •if
it were proper to bring in here things I shall have occasion
to say later about the idea we form of power and efficacy.
But these approaches might seem •to weaken my system by
resting one part of it on another, or •to create confusion in
my reasoning ·by taking things out of order·; so I shall try to

iii: Knowledge and probability

maintain my present thesis without either of those kinds of
help.
Let it be temporarily granted, then, that the production
of one object by another in any one instance implies a power,
and that this power is connected with its effect. But it has
already been proved that the power doesn’t lie in the perceptible qualities of the cause, yet all we have present to us are
its perceptible qualities. So I ask: why, in other instances
where those qualities have appeared, do you presume that
the same power is also there? Your appeal to past experience
gives you no help with this. The most it can prove is that that
very object which produced a certain other object was at that
very instant endowed with a power to do this; but it can’t
prove that the same power must continue in the same object
(collection of perceptible qualities) ·at other times·, much
less that a similar power is always conjoined with similar
perceptible qualities ·in other objects·. You might say: ‘We
have experience that the same power continues ·through
time· to be united with the same object, and that similar
objects are endowed with similar powers’; but then I renew
my question about why from this experience we form any
conclusion that goes beyond the past instances of which we
have had experience. If you answer this in the same way
that you did the previous question, your answer will raise
a new question of the same kind, and so on ad infinitum;
which clearly proves that this line of reasoning had no solid
foundation.
Thus, not only does •our reason fail to reveal to us the
ultimate connection of causes and effects, but even after
experience has informed us of their constant conjunction
we can’t through •our reason satisfy ourselves concerning
why we should extend that experience beyond the particular
instances that we have observed. We suppose, but can never
prove, that objects of which we have had experience must
52

Treatise, Book 1

David Hume

iii: Knowledge and probability

(·‘not the sole causes’·), I still contend that the only general
factors that associate ideas are •resemblance, •contiguity,
and •causation.
Ideas are indeed subject to a uniting force that may at
first sight seem different from any of these, but will be found
ultimately to depend on the same origin. When every individual of some kind of objects is found by experience to be
constantly united with an individual of another kind, the
appearance of any new individual of either kind naturally
conveys our thought to its usual attendant. Thus, because a
particular idea is commonly attached to a particular word,
nothing is required but the hearing of that word to produce
the corresponding idea; and this transition will be one that
the mind is hardly able to prevent, however hard it tries. In
this case it is not absolutely necessary that on hearing the
sound we should reflect on past experience and consider
what idea has usually been connected with the sound. The
imagination, unaided, takes the place of this reflection; it
is so accustomed to pass from the word to the idea that it
doesn’t delay for a moment between hearing the word and
conceiving the idea.
But though I acknowledge this to be a true principlec of
association among ideas, I contend that it is the very same
as that between the ideas of cause and effect, and is an
essential part of all our causal reasonings. The only notion
of cause and effect that we have is that of certain objects
that have been always conjoined together, and in all past
instances have been found inseparable. We can’t penetrate
into the reason for that conjunction. We only observe the fact
itself: from constant conjunction, objects acquire a union
in the imagination. When the impression of one becomes
present to us, we immediately form an idea of whatever usually accompanies it; and consequently we can lay this down
as one part of the definition of opinion or belief, that it is an

resemble the ones that lie beyond the reach of our discovery.
I have called attention to •certain relations that make
us pass from one object to another even when no reason
leads us to make that transition; and we can accept as a general rule that wherever the mind constantly and uniformly
makes a transition without any reason, it is influenced by
•these relations. That is exactly what we have in the present
case. Reason can never show us a connection of one object with another, even with the help of experience and the
observation of the objects’ constant conjunction in all past
instances. So when the mind passes from the idea or impression of one object to the idea of or belief in another, it isn’t
driven by reason but by certain forces that link the ideas of
these objects and unite them in our imagination. If among
•ideas in the •imagination there were no more unity than
the •understanding can find among •objects, we could never
draw any inference from causes to effects, or believe in any
matter of fact. The inference, therefore, depends solely on
the ·unreasoned· union of ideas.
The principlesc of union among ideas come down to three
general ones, I maintain; and I have said that the idea
or impression of any object naturally introduces the idea
of any other object that is •resembling, •contiguous to, or
•connected with it. These are neither the infallible nor the
sole causes of union among ideas. They are not infallible
causes, because someone may fix his attention for a while
on one object, without looking further. They are not the sole
causes, because ·some of our transitions from impressions
to ideas owe nothing to these three relations·: our thought
has a very irregular motion in running along its objects, and
can leap from the heavens to the earth, from one end of the
creation to the other, without any certain method or order.
But though I concede this weakness in these three relations
(·‘not infallible’·), and this irregularity in the imagination
53

Treatise, Book 1

David Hume

iii: Knowledge and probability

conjunction—it’s only in its role as a •natural relation that it
produces a union among our ideas and enables us to reason
on it and draw inferences from it. [See note on page 8.]

idea related to or associated with a present impression.
Thus, though causation is a •philosophical relation—
because it involves contiguity, succession, and constant

7: The nature of the idea or belief
The •idea of an object is an essential part of the •belief in it—
·of the belief that it exists·—but not the whole. We •conceive
many things that we don’t •believe. Let us now investigate
more fully the nature of belief, or the qualities of the ideas
that we assent to.
Obviously, all reasonings from causes or effects end in
conclusions about matters of fact—that is, about the existence of objects or of their qualities. It is also obvious that
the •idea of existence is not different from the •idea of any
object, and that when after •simply conceiving something
we want to •conceive it as existent, this actually doesn’t add
to or alter anything in the first idea. For example, when we
affirm that God is existent we simply form the idea of such
a being as he is represented to us, and the existence we
attribute to him is not conceived by a particular idea which
we join to the idea of his other qualities and could again
separate and distinguish from them. But I go further than
this. I say not only that •the conception of the existence of
an object adds nothing to •the simple conception of it, but
also that •the belief in its the existence doesn’t add any new
ideas either. When I
think of God, then
think of God as existent, then
believe God to be existent,
my idea of him neither grows nor shrinks. Still, a simple conception of the existence of an object certainly differs greatly
from a belief in it; and as this difference doesn’t consist in

the parts or structure of the relevant idea, it follows that it
must consist in how we conceive it.
Suppose that someone in conversation says things to
which I don’t assent— that Caesar died in his bed, that silver
is more fusible than lead, that mercury is heavier than gold.
It is obvious that despite my incredulity I clearly understand
his meaning, and form all the same ideas as he does. My
imagination has the same powers as his: he can’t conceive
any idea that I can’t conceive, or conjoin any ideas that I
can’t conjoin. So I ask: what makes the difference between
believing a proposition and disbelieving it? The answer is
easy with regard to •propositions that are proved by intuition
or demonstration. In that case, the person who assents not
only conceives the ideas according to the proposition but is
forced—either immediately or by the interposition of other
ideas—to conceive them in just that way. Whatever is absurd is unintelligible, and the imagination cannot conceive
anything contrary to a demonstration. But in •reasonings
from causation, and about matters of fact, this sort of necessity isn’t present and the imagination is free to conceive
both sides of the question; so I ask again, what makes the
difference between incredulity and belief?. . . . Here is a bad
answer:
A person who doesn’t assent to a proposition that you
advance first conceives the object in the same way as
you, and then immediately goes on to conceive it in
a different way and to have different ideas of it; ·and
54

Treatise, Book 1

David Hume

this different conception is his disbelief·.

in every detail. When you want somehow to vary your idea of
a particular object, all you can do is to make it more or less
strong and lively. If you change it in any other way it will
come to represent a different object or impression. (Similarly
with colours. A particular shade of a colour may acquire
a new degree of liveliness or brightness without any other
variation; but if you produce any other change it is no longer
the same shade or colour.) Therefore, as belief merely affects
how we conceive any object, all it can do—·the only kind of
variation that won’t change the subject, so to speak·—is to
make our ideas stronger and livelier. So an opinion or belief
can most accurately defined as: a lively idea related to or
associated with a present impression.5

This answer is unsatisfactory—not because it contains any
falsehood but because it doesn’t reveal the whole truth.
Whenever we dissent from what someone says, we do indeed conceive both sides of the question, ·and that is the
truth in the ‘bad answer’·; but we can believe only one side,
so it evidently follows that belief must make some difference between the conception to which we assent and the
one from which we dissent. We may mingle, unite, separate, run together, and vary our ideas in a hundred different
ways; but until there appears some principlec that fixes one
of these different combinations ·as the one we believe·, we
have in reality no opinion. And this principlec , as it plainly
adds nothing to our previous ideas, can only change how we
conceive them. All the perceptions of the mind are of two
kinds, impressions and ideas, which differ from each other
only in their different degrees of force and liveliness. Our
ideas are copied from our impressions and represent them
5

iii: Knowledge and probability

Here are the main points of the arguments that lead us
to this conclusion. When we infer the existence of one object
from the existence of others, some object must always be
present either to the memory or senses to serve as the foundation of our reasoning (the alternative being a regress ad

I take this opportunity to comment on a very remarkable error which, because it is frequently taught in the schools [= ‘Aristotelian philosophy departments’], has become a kind of established maxim and is accepted by all logicians. This error consists in the division of the acts of the understanding
into
conception, judgment, and reasoning,
and in the definitions given of them. •Conception is defined as the simple survey of one or more ideas, •judgment as the separating or uniting
of different ideas, and •reasoning as the separating or uniting of different ideas by the interposition of others which show how they are related to
one another. But these distinctions and definitions are seriously faulty. (1) It is far from being true that in every judgment that we form we unite
two different ideas. In the proposition God is—or indeed any other proposition about existence—the idea of existence is not a distinct idea that we
unite with that of the thing that is said to exist, forming a compound idea by the union. (2) Just as we can thus form a proposition containing only
one idea—·as the idea of God is the only idea in the proposition God exists·—so we can exercise our reason employing only two ideas, not bringing
in a third to serve as an intermediary between them. We infer a cause immediately from its effect; and this inference is not only a true example
of reasoning, but is the strongest of all, and is more convincing than when we interpose another idea to connect the two extremes. What we can
in general affirm regarding these three acts of the understanding is that properly understood they all come down to the first of the three, and are
nothing but particular ways of conceiving our objects. Whether we consider a single object or several, whether we dwell on these objects or run from
them to others, and in whatever form or order we survey them, the act of the mind doesn’t go beyond a simple conception, and the only remarkable
difference that sometimes occurs is when we join belief to the conception and are convinced of the truth of what we conceive. Belief is an act of the
mind that has never yet been explained by any philosopher; so I am at liberty to propose my hypothesis about it, which is that belief is only a strong
and steady conception of an idea—one that approaches in some degree to an immediate impression.

55

Treatise, Book 1

David Hume

infinitum). Reason can never satisfy us that the existence of
any one object ever implies the existence of another; so when
we pass from an impression of one to an idea of or belief in
another, we are driven not by reason but by custom, or an
associative force. But belief is something more than a simple
idea. It is a particular manner of forming an idea; and as
an idea can be varied—·without being turned into another
idea·—only by a variation of its degree of force and liveliness,
it follows from all this that belief is a lively idea produced by
a relation to a present impression, which is the definition I
gave.
This operation of the mind that forms the belief in any
matter of fact seems to have been until now one of the greatest mysteries of philosophy, though no-one has so much as
suspected that there was any difficulty in explaining it. For
my part, I have to admit that I find a considerable difficulty
in this, and that even when I think I understand the subject
perfectly I am at a loss for words in which to express my
meaning. A line of thought that seems to me to be very
cogent leads me to conclude that an •opinion or belief is
nothing but an idea that differs from a •fiction not in the
nature or the order of its parts but in how it is conceived.
But when I want to explain this ‘how’, I can hardly find any
word that fully serves the purpose, and am obliged to appeal
to your feeling in order to give you a perfect notion of this
operation of the mind. An •idea assented to feels different
from a •fictitious idea that the imagination alone presents to
us; and I try to explain this difference of feeling by calling it
‘a superior force’, or ‘liveliness’, or ‘solidity’, or ‘firmness’, or
‘steadiness’. This variety of terms, which may seem so unphilosophical, is intended only to express the act of the mind
that makes realities more present to us than fictions, causes
them to weigh more in thought, and gives them a superior

iii: Knowledge and probability

influence on the passions and the imagination. Provided we
agree about the thing, we needn’t argue about the labels. . . .
I admit that it is impossible to explain perfectly this feeling
or manner of conception ·that marks off belief·. We can
use words that express something near it. But its true and
proper name is ‘belief’, which is a term that everyone sufficiently understands •in common life. And •in philosophy
we can go no further than to say that it is something felt
by the mind which distinguishes the ideas of the judgment
from the fictions of the imagination. It gives them more force
and influence, makes them appear of greater importance,
anchors them in the mind, and makes them the governing
forces of all our actions.
This definition will also be found to fit perfectly with everyone’s feeling and experience. Nothing is more obvious
than that the ideas to which we assent are more strong,
firm, and vivid, than the loose dreams of a castle- builder.
If one person sits down to read a book as a romance, and
another ·reads the same book· as a true history, they plainly
receive the same ideas in the same order; and they attach
the very same sense to what their author writes, despite
the incredulity of one and the belief of the other. His words
produce the same ideas in both, but his testimony doesn’t
have the same influence on them. The believing reader has
a livelier conception of all the incidents. He enters deeper
into the concerns of the persons; he represents to himself
their actions and characters, their friendships and enmities;
he even goes so far as to form a notion of their features and
manners. While the disbelieving reader, who gives no credit
to the testimony of the author, has a more faint and languid
conception of all these particulars, and can’t be much entertained by it unless he is held by the style and ingenuity of
the composition.

56

Treatise, Book 1

David Hume

iii: Knowledge and probability

8: The causes of belief
Having thus explained the nature of belief, and shown that
it consists in a lively idea related to a present impression, I
now enquire into what forces produce belief—·that is·, what
gives the idea its liveliness.

right about this just by considering what it is for things to
be ·naturally· related, and the ease of transition that is essential to this. But I have to say that ·we can’t, and that· my
confidence in my account comes mainly from experience. As
the first experience that is relevant to our topic, I note that
when we see a picture of an absent friend our idea of him is
plainly enlivened by the resemblance ·of the picture to the
friend·, and that every passion that the idea of our friend
gives us—whether of joy or of sorrow—acquires new force
and vigour ·because we see the picture·. This effect comes
from the joint operation of •a relation and •a present impression. ·The relation·: if the picture isn’t at all like the friend,
or at least wasn’t intended to be a picture of him, it doesn’t
so much as carry our thought to him. And ·the present
impression·: if the picture is absent as well as the friend, the
mind may pass from the thought of the picture to that of the
friend, but in this case it feels its idea ·of the friend· to be
weakened rather than enlivened by that transition. We enjoy
seeing a picture of our friend when it is set before us; but
when the picture is removed, we prefer thinking about him
directly to thinking about him as reflected in a picture which
is as distant and dark to us as he is.

I would like to have it established as a general maxim
in the science of human nature that when an impression
becomes present to us it not only •carries the mind to such
ideas as are related to it but also •passes on to those ideas
a share of its force and liveliness. All the operations of the
mind depend to a large extent on its state at the time when it
performs them; and the action will always have more or less
vigour and liveliness according to whether the energy-level is
high or low and the attention more or less fixed. So when an
object is presented which •elevates and enlivens the thought,
every action the mind performs will be stronger and more
vivid as long as •that state continues. Now, it is obvious
that how long the state continues depends entirely on what
the mind is thinking about, and that •any new object ·of
thought· naturally draws the energies in a new direction and
changes the mind’s state; while on the other hand when
the mind fixes constantly on •the same object, or passes
easily along related objects without being aware that they
are different, the state lasts much longer. So it comes about
that when the mind is enlivened by a present impression it
proceeds to form a livelier idea of the related objects, by a
natural transition of the state—·the level of liveliness·—from
one to the other. The change of the objects is so easy that
the mind is hardly aware of it, and applies itself to the conception of the related idea with all the force and liveliness it
acquired from the present impression.

The ceremonies of the Roman Catholic religion may be
considered as events of this sort. The devotees of that strange
superstition usually plead, in excuse of the weird rituals they
are scolded for, that they feel the good effect of those external movements, postures and actions in enlivening their
devotion and their fervour, which would decay if they were
directed entirely to distant and immaterial objects. They say:
We represent the objects of our faith in perceptible
symbols and images, and make them more present to

It would be nice if we could satisfy ourselves that I am
57

Treatise, Book 1

David Hume

us by the immediate presence of these symbols than
we could make them merely by an intellectual view
and contemplation. Perceptible objects always have
a greater influence on the imagination than anything
else, and they readily pass this influence along to the
ideas to which the objects are related and which they
resemble.
I shall only infer from these practices and this defence of
them that the effect of resemblance in enlivening the idea
is very common; and as in every case a resemblance and a
present impression must work together, we are abundantly
supplied with phenomena to prove the reality of the ·ideaenlivening· force of which I have spoken.
We may reinforce these phenomena by ·bringing in· others of a different kind, noting the effects of contiguity as well
as of resemblance. Distance certainly lessens the intensity of
every idea; and when we are getting near to an object, even
though it isn’t yet present to our senses, it operates on our
mind with an influence that imitates ·that of· an immediate
impression. Thinking about an object readily carries the
mind to things that are contiguous to it; but only the object’s
actual presence carries the mind ·to an idea of contiguous
objects· with a superior liveliness. ·Here is an example of
what happens where there isn’t a relevant present impression·. When I am a few miles from home, whatever relates to
it touches me more nearly than when I am six hundred miles
away, though even at that distance reflecting on anything
in the neighbourhood of my friends and family naturally
produces an idea of them. But as in this latter case both the
relevant objects of the mind are ideas, the easy transition
between them can’t heighten the liveliness of either, because
there is no immediate impression at work.
No-one can doubt that •causation has the same influence
as do •resemblance and •contiguity. Superstitious people are

iii: Knowledge and probability

fond of the relics of saints and holy men, for the same reason that they want symbols and images, in order to enliven
their devotion and give them a stronger and more intimate
conception of the exemplary lives they want to imitate. It is
clear that one of the best relics a devotee could get would be
something made by the saint ·and thus causally related to
him·; and if his clothes and furniture are ever considered as
especially desirable in the same way, that is because they
were once at his disposal and were moved and affected by
him, which makes them partial effects of the saint, and connected with him by a shorter chain of consequences than any
of the ones from which we learn that he really existed. This
phenomenon clearly proves that a present impression with a
relation of causation can enliven any idea, and consequently
produce belief or assent; which fits my definition of ‘belief’.
But we needn’t search out other arguments to prove that
a present impression with a relation or transition of the imagination can enliven an idea, because this very example—our
reasonings from cause and effect— suffice for that purpose
all on its own! Here are three certainties:
•We must have an idea of every matter of fact that we
believe.
•This idea arises only from a relation to a present impression.
•The belief adds nothing to the idea, but only changes
how we conceive it, making it stronger and livelier.
The present conclusion about the influence of ·a natural·
relation follows immediately from these steps, and every step
appears to me sure and infallible. All that this operation of
the mind contains is: •a present impression, •a lively idea,
and •a relation or association in the imagination between the
two.
. . . . It is the present impression that is to be considered
as the true cause of the idea, and of the belief that comes
58

Treatise, Book 1

David Hume

with it. So we should consult our experience in order to learn
what special qualities the impression has that enable it to
produce such an extraordinary effect.
•First kind of experience: the present impression doesn’t
have this effect through its own power and efficacy, considered alone as a single perception and limited to the present
moment. I find that an impression from which I can draw
no conclusion when it first appears can later become a basis
for a belief, after I have had experience of its usual consequences. ·For such a transition to occur·, we must in every
case have observed the same ·sort of· impression in past
instances, and have found that there is some other ·sort of·
impression with which it is constantly conjoined. This is
confirmed by such a multitude of events that there can’t be
the slightest doubt about it.
From a •second kind of experience I conclude that the
belief that comes with the present impression, and is produced by a number of past impressions and pairs of events,
arises immediately, without any new operation of the reason
or imagination. I can be sure of this, because I never am
conscious of any such operation in myself and don’t find
anything in the situation to operate on. When something
comes from a past repetition without any new reasoning or
conclusion, our word for it is ‘custom’; so we can take it as
certainly established every belief that follows on a present
impression is derived solely from custom. When we are accustomed to see two impressions conjoined, the appearance
or idea of one immediately carries us to the idea of the other.
Being fully satisfied about this, I make a •third appeal to
experience in order to learn whether the production of this
phenomenon of belief needs anything more, in addition to
the customary transition. So I change the first impression
into an idea; and then I note that though the customary
transition to the correlative idea still remains, there isn’t any

iii: Knowledge and probability

real belief or conviction. So a present impression is absolutely required for this whole operation; and when I go on
to compare an impression with an idea, and find that they
differ only in their degrees of force and liveliness, I reach the
bottom-line conclusion that belief is a more vivid and intense
conception of an idea, coming from its relation to a present
impression.
Thus, all probable reasoning is nothing but a kind of
sensation. We must follow our taste and sentiment not only
in poetry and music but also in philosophy. When I am
convinced of some principle, it is only an idea that strikes
me more strongly. When I prefer one set of arguments to
another set, all I do is to decide on the basis of how they feel
which is the more powerful [Hume’s exact words: ‘I do nothing but decide from my feeling concerning the superiority of
their influence’]. Objects have no discoverable connection
with one another, and the only factor that lets us draw any
inference from the appearance of one object to the existence
of another is custom operating on the imagination.
It is worth noting that the past experience on which all
our judgments about cause and effect depend can operate
on our mind so imperceptibly that we don’t notice it, and it
may even be that we don’t fully know it. A person who stops
short in his journey when he comes to a river in his way foresees the consequences of going forward; and his knowledge
of these consequences comes from past experience which
informs him of certain linkages of causes and effects. But
does he reflect on any past experience, and call to mind instances that he has seen or heard of, in order to discover
how water effects animal bodies? Surely not! That isn’t how
he proceeds in his reasoning. ·In his mind· the idea of •water
is so closely connected with that of •sinking, and the idea of
•sinking is so closely linked with that of •drowning, that his
mind moves from one idea to the next to the next without
59

Treatise, Book 1

David Hume

help from his memory. . . . But as this transition comes from
experience and not from any primary connection between the
ideas, we have to acknowledge that experience can produce a
belief—a judgment regarding causes and effects—by a secret
operation in which it is not once thought of. This removes
any pretext that may remain for asserting that the mind
is convinced by reasoning of the principle that instances of
which we haven’t had experience must resemble those of
which we have. For we here find that the understanding or
imagination can draw inferences from past experience without so much as reflecting on it—let alone forming a principle
about it and reasoning on the basis of the principle!
In general we may observe that in all the most established
and uniform conjunctions of causes and effects—gravity, impact, solidity, etc.—the mind never consciously reflects on
any past experience; though in cause- effect linkages that
are more rare and unusual the mind may engage in such
reflections as an aid to the custom and transition of ideas.
Indeed, in some cases •the reflection produces the belief
without the custom; or— more accurately—•the reflection
produces ·the belief by producing· the custom in an oblique
and artificial manner. Let me explain. It is certain that
not only in philosophy and science, but even in common
life, we can come to know of a particular cause by a single
experiment, provided it is judiciously made with a careful
removal of all extraneous and irrelevant circumstances. . . .
A habit can’t be acquired from a single instance, so it may be
thought that belief in this case can’t be the effect of custom.
But this difficulty will vanish if we consider that, though
we are here supposed to have had only one experience of
a particular effect, we have millions to convince us of this
principle:
•Like objects placed in like circumstances will always
produce like effects.

iii: Knowledge and probability

And as this principle has established itself by a sufficient
custom, it makes convincing and firm any opinion to which
it can be applied. The connection of the ideas is not habitual after one experiment; but this connection is covered
by another principle that is habitual; which brings us back
to my hypothesis. In all cases we transfer our experience
to instances of which we have no experience, doing this
consciously or implicitly, directly or indirectly.
I mustn’t leave this subject without remarking that it is
very difficult to talk perfectly properly and accurately about
the operations of the mind, because common language has
seldom made any very fine distinctions amongst them, generally calling by the same word all that closely resemble each
other. And as this is almost inevitably a source of obscurity
and confusion in an author, so it may cause you to have
doubts and objections that you otherwise would never have
dreamed of. Thus, my general position that
•an opinion or belief is nothing but a strong and lively
idea derived from a present impression related to it
may be liable to the following objection, because of a little
ambiguity in the words ‘strong’ and ‘lively’:
It is not only an •impression that can give rise to
reasoning—an •idea can have the same influence, especially given your principle that all our ideas are derived from corresponding impressions. If I now form
an idea whose corresponding impression I have forgotten, I can still conclude from ·the existence of· this
idea that such an impression did once exist; and this
conclusion comes as a belief; so •what is the source
of the qualities of force and liveliness that constitute
this belief?
I am ready with an answer: •it comes from the present idea.
This idea is not here considered as •the representation of
an absent object but as •a real perception in the mind, of
60

Treatise, Book 1

David Hume

which we are intimately conscious; so it must be able to
bestow on whatever is related to it the same quality (call it
‘firmness’, or ‘solidity’, or ‘force’, or ‘liveliness’ [Hume throughout uses ‘vivacity’]) with which the mind reflects on it and is
assured of its present existence. The idea here takes the
place of an impression, and so far as our present purpose
goes it is entirely the same.
For the same reason, we needn’t be surprised to hear of
the memory of an idea—that is, the idea of an idea—and of
its having more force and liveliness than the loose concep-

iii: Knowledge and probability

tions of the imagination. In thinking of our past thoughts we
don’t just sketch out the objects of which we were thinking;
we also conceive the action of our mind in doing this— that
certain je-ne-sais-quoi of which it is impossible to give any
definition or description but which everyone understands
well enough. When the memory offers an idea of this, and
represents it as past, it is easy to see how that idea could
have more vigour and firmness than ·the idea that occurs·
when we think of a past thought without having any memory
of it. . . .

9: The effects of other relations and other habits
However convincing those arguments may appear, I mustn’t
rest content with them, but must turn the subject on every
side in order to find new points of view from which I can
illustrate and confirm these extraordinary and fundamental
principles. Philosophers are right when they conscientiously
hesitate to accept a new hypothesis; their attitude is necessary for progress towards the truth, and should be respected.
So I must produce every argument that may tend to their
satisfaction, and remove every objection that may stop them
in their reasoning.

all this out I was confirming by analogy my account of our
judgments about cause and effect. But this very argument
might be turned against me, becoming an objection to my
hypothesis rather than a confirmation of it. The objection
goes like this:
If all the parts of your hypothesis are true, namely:
•these three kinds of relation are derived from
the same principlesc ,
•their effects in giving force and liveliness to our
ideas are the same, and
•belief is nothing but a more forceful and vivacious conception of an idea,
it should follow that belief can come not only from
the relation of •cause and effect, but also from those
of •contiguity and •resemblance. But we find by experience that belief arises only from causation, and that
we can draw no inference from one object to another
unless they are connected by this relation. So we can
conclude that there is some error in the reasoning
that has led us into such difficulties.

I have often remarked that in addition to •cause and effect
the two relations of •resemblance and •contiguity are associating forces of thought, capable of conveying the imagination
from one idea to another. I have also noted that when two
objects are linked by either of these relations, and one of the
objects is immediately present to the memory or senses, the
mind is not only •carried to the linked object by means of
the associating force, but •conceives that object with an additional force and vigour through the combined operation of
the associating force and the present impression. In pointing
61

Treatise, Book 1

David Hume

That is the objection; now let us consider its solution. It
is obvious that whatever is present to the memory, striking
on the mind with a liveliness that resembles ·that of· an immediate impression, must have a considerable effect on all
the operations of the mind, easily distinguishing itself from
mere fictions of the imagination. Of these impressions or
ideas of the memory we form a kind of system, incorporating
into it whatever we remember having been present to our
internal perception or ·our external· senses; and whenever
some particular item in that system is joined to a present
impressions, we choose to call it ‘a reality’. But the mind
doesn’t stop at that. Finding that •this system of perceptions
is connected by custom—or, if you like, by the relation of
cause and effect—with •another system, it proceeds to consider the ideas of items in the latter system. It feels itself to
be somehow forced to view these particular ideas, and finds
that the custom or relation which does the forcing can’t be
changed in the slightest; so it forms them—·this second set
of ideas·—into a new system, which it likewise dignifies with
the title of ‘realities’. •The former of these two systems is the
object of the memory and senses, the •latter of the judgment.
The judgment is what populates and furnishes the world,
acquainting us with things that are too remote in time or
space for our senses or memory to reach them. [Hume goes
on to describe his beliefs about the history of Rome. Then:]
These and all my other beliefs are nothing but ideas, though
by their force and settled order, arising from custom and
the relation of cause and effect, they distinguish themselves
from ideas that are merely the offspring of the imagination.
As to the influence of •contiguity and •resemblance, we
may observe that if the contiguous and resembling object is
included in this system of realities, there is no doubt that
these two relations will assist that of cause and effect, and
fix the related idea with more force into the imagination. . . .

iii: Knowledge and probability

But though I can’t entirely exclude the relations of resemblance and contiguity from operating on the imagination in
this way, it is observable that when they occur on their own
their influence is very feeble and uncertain. The cause-effect
relation is needed to persuade us of any real existence, and
its persuasion is also needed to give power to these other
relations. ·Here is why·. Take a case where the appearance
of an impression leads us not only •to feign another object
but quite arbitrarily •to give the latter a particular relation
to the impression: this can’t have any great effect on the
mind, and there is no reason why if the same impression
returns we should be led to place the same object in the
same relation to it. [The word ‘feign’ comes from a Latin word that
is also the source for ‘fiction’. Hume is talking about fictions, inventions,

It is in no way necessary for the mind
to feign any resembling or contiguous objects; and if it does
feign them it needn’t always do it in the same way. Indeed,
such a fiction is based on so little reason that nothing but
pure whim can lead the mind to form it; and whim being
fluctuating and uncertain, it can’t possibly operate with any
considerable degree of force and constancy. . . . The relation
of cause and effect has all the opposite advantages. The objects it presents are fixed and unalterable. The impressions
of memory never change in any considerable degree; and
each impression draws along with it a precise idea, which
takes its place in the imagination as something solid and
real, certain and invariable. The thought is always made to
pass from the impression to the idea—and from that particular impression to that particular idea—without any choice or
hesitation.
stories we tell ourselves.]

Not content with removing this objection, however, I shall
try to extract from it an argument for my doctrine. Contiguity
and resemblance have much less effect than does causation;
but they still have some effect, and strengthen the confi62

Treatise, Book 1

David Hume

dence of any opinion and the liveliness of any conception. If
I can show this with various new examples in addition to the
ones I have already noted, you will grant that that will be a
considerable further argument that belief is nothing but a
lively idea related to a present impression.
To begin with contiguity: it has been remarked that
Moslem pilgrims who have seen Mecca, and Christians who
have seen the Holy Land, are from then on more faithful
and zealous believers than those who haven’t had that advantage. A man whose memory presents him with a lively
image of the Red Sea, the desert, Jerusalem, and Galilee
can never doubt any miraculous events that are related by
Moses or by the evangelists. His lively idea of those places
passes by an easy transition to the events that are supposed
to have been related to them by contiguity, and increases
his belief by increasing the liveliness of his conception. A
memory of these fields and rivers has the same influence on
ordinary people as a new argument would— and from the
same causes! Something similar holds for resemblance. I
have remarked that •the conclusion we draw from a present
object to its absent cause or effect is never based on any
qualities that we observe in that object considered in itself;
or, in other words, that it is only through experience that one
can determine what a given event resulted from or what will
result from it. But though this is so obvious that it didn’t
seem to need supporting argument, some philosophers have
imagined that there is a visible cause for the communication
of motion, and that a reasonable man could immediately
infer the motion of one body from the impact of another,
without appealing to any past observation. It is easy to prove
that this is false, thus:
If such an inference can be drawn merely from the
ideas of body, motion, and impact, it must amount
to a demonstration, and must imply the absolute im-

iii: Knowledge and probability

possibility of any contrary supposition. From this it
would follow that ‘A case of impact caused something
other than the communication of motion’ implies a
formal contradiction: not merely that it can’t possibly
be true but that it can’t even be conceived. But we can
quickly satisfy ourselves that this is wrong by forming
a clear and consistent idea of one body’s colliding with
another, and
immediately coming to rest, or
going back in the same line in which it came, or
going out of existence, or
moving in a circle or an ellipse,
or—cutting it short—going through any one of countless other changes.
These suppositions are all consistent and natural; and the
reason why some philosophers imagine the communication
of motion to be more consistent and natural, not only than
those suppositions but also than any other natural effect,
is based on the resemblance between the cause and the
effect—·motion into the collision, motion out from it·. In this
case the •resemblance combines with •experience ·of motion
in, motion out·—and binds the objects in the closest and
most intimate manner to each other, so as to make those
philosophers imagine them to be absolutely inseparable. Resemblance, then, has the same influence as experience, or
anyway a parallel one; and as the only immediate effect of
experience is to associate our ideas together, it follows that
all belief arises from the association of ideas—which is what
my hypothesis says.
Writers on optics all agree that the eye at all times sees
the same number of physical points, and that a man on a
mountain-top has no larger an image presented to his senses
than when he is cooped up in the smallest room. It is only
through experience that he infers from some special quali63

Treatise, Book 1

David Hume

ties of the image the largeness of the object ·he is seeing·;
and here as in other contexts he confuses this inference
of his •judgment with a •sensation. [Hume develops this
point in some detail, giving a special role to the relation of
resemblance; omitted here.] No weakness of human nature
is more universal and conspicuous than what we commonly
call ‘credulity’, or too easily believing what others say; and
this weakness is also very naturally accounted for by the
influence of resemblance. When we accept any matter of
fact on the strength of human testimony, our belief comes
from the very same source as our inferences from causes
to effects, and from effects to causes. Our experience of
the dominant drives in human nature is the only possible
basis for any confidence we may have in the veracity of men.
But though experience is the true standard for this as of
all other judgments, we seldom regulate ourselves entirely
by it, and have a remarkable propensity to believe whatever
we are told—even about apparitions, enchantments, and
wonders, however contrary to daily experience and observation. The words or discourses of other people have an
intimate •connection with certain ideas in their minds, and
these ideas have a •connection with the facts or objects that
they represent. This latter •connection is generally much
overrated, and commands our assent beyond what experience will justify; and the explanation for this must lie in the
resemblance between the ideas ·of the speakers· and the
·supposed· facts. Other effects indicate their causes only
in an oblique manner; but the testimony of men does it directly, and is to be considered as a likeness as well as an
effect. So it is not surprising that we are so rash in drawing
inferences from it, and are less guided by experience in our
judgments about it than we are in our judgments about any
other subject.

iii: Knowledge and probability

strengthens our reasonings, so a considerable lack of resemblance can almost entirely destroy them. A remarkable
example of this is the universal carelessness and stupidity of
men with regard to a future state, a topic in which they show
as obstinate an incredulity as they do a blind credulity about
other things. There is indeed no richer source of material for
a studious man’s wonder, and a pious man’s regret, than the
negligence of the bulk of mankind concerning their after-life;
and it is with reason that many eminent theologians have
been so bold as to say that though common people don’t
explicitly assent to any form of unbelief they are really unbelievers in their hearts and have nothing like what we could
call a belief that their souls are eternal. Let us consider
on the one hand •what divines have presented with such
eloquence about the importance of eternity, ·which is to be
spent either in heaven or in hell·; and in estimating this
let us reflect that though in matters of rhetoric we can expect some exaggeration, in this case we must allow that the
strongest figures of speech fall infinitely short of the subject.
Then let us view on the other hand •how prodigiously safe
men feel about this! Do these people really believe what they
are taught, and what they claim to affirm? Obviously not.
·And I shall now explain why·.
Given that belief is an act of the mind arising from custom, it isn’t surprising that a lack of resemblance should
overthrow what custom has established, and lessen the
force of the idea as much as custom increases it. A future
state is so far removed from our comprehension, and we
have so obscure an idea of how we shall exist after our bodies have disintegrated, that all the reasons we can devise—
however strong in themselves and however much assisted by
education—can never, in people with slow imaginations, surmount this difficulty and bestow a sufficient authority and
force on the idea. I ascribe this incredulity to the faintness

Just as resemblance when combined with causation
64

Treatise, Book 1

David Hume

iii: Knowledge and probability

linkage with something else·, should frequently appear to
the mind, this idea must gradually become easier to have
and more forceful when it does occur; and this facility and
force—this easy introduction and firm hold on the mind—
distinguish this recurring idea from any new and unusual
idea. This is the only respect in which these two kinds of
custom agree; and if it turns out that their effects on judgment are similar, we can certainly conclude that my account
of judgment ·or belief· is satisfactory. Well, is their influence
on judgment similar? Who can doubt it when we consider
the nature and effects of education?
All the opinions and notions of things to which we have
been accustomed from our infancy take such deep root that
it is impossible for us, by all the powers of reason and experience, to eradicate them; and this habit has an influence that
is as strong as the influence arising from the constant and
inseparable union of causes and effects. Indeed, it is sometimes stronger, and overcomes the latter influence. Don’t say
that the vividness of the idea produces the belief; the vivid
idea is the belief. The frequent repetition of an idea fixes it
in the imagination, but such a repetition couldn’t possibly
produce belief all by itself if we were so built that belief could
come only through reasoning and comparison of ideas. . . .
(Here are three parallel instances. •Someone who has lost
a leg or an arm by amputation tries for a long time afterwards
to use the lost limb. •After someone’s death it is common
for members of his household, especially the servants, to
say that they can hardly believe he is dead, but still imagine
him to be in his study or wherever else in the house they
were accustomed to find him. •In conversation about some
celebrated person, I have often heard something like this:
‘I never saw him, but I almost fancy that I have, because I
have so often heard talk of him.’)
If we look at this argument from education in the right

of our idea of our future condition, derived from its •lack of
resemblance to the present life, rather than to faintness derived from the after-life’s •remoteness in time. For I observe
that men are everywhere concerned about what may happen
in this world after their death, and that nearly everyone has
some care for his ·post mortem· reputation, his family, his
friends, and his country.
[Then a paragraph continuing this theme, adducing other
evidence that hardly anyone really believes that he is at risk
of eternal damnation. Then:] I would further remark that in
matters of religion men take a pleasure in being terrified, and
that no preachers are as popular as those who arouse the
most dismal and gloomy emotions. In the common affairs of
life, where we feel and are penetrated with the reality of the
subject, nothing can be nastier than fear and terror; it is only
in •dramatic performances and •religious discourses that
they ever give pleasure. In these latter cases the imagination
lazily admits the idea; and the emotion, being softened by
the lack of belief in what is said, has merely the agreeable
effect of enlivening the mind and fixing the attention.
My hypothesis will be further confirmed if we examine the
effects of other kinds of custom as well as of other relations.
Custom, to which I attribute all belief and reasoning, can
operate on the mind in invigorating an idea in two different
ways. ·One is the way I have been describing·. •If in all past
experience we have found two ·kinds of· objects to be always
conjoined together, the appearance of one of these objects
in an impression leads us, through custom, to move easily
to the idea of the ·kind of· object that usually accompanies
it; and the present impression and the easy transition make
us conceive that idea in a stronger and livelier manner than
we do any loose floating image of the imagination. But let
us next •suppose that a mere idea alone, without any of
this curious and almost artificial preparation ·of experienced
65

Treatise, Book 1

David Hume

way, it will appear very convincing; and all the more so
from being based on one of the most common phenomena
that is to be met with anywhere. ·We are all familiar with
education; and I am contending that the core of education
is the production of beliefs through sheer repetition of certain ideas—that is, through creating customs of the second
of the two kinds I have mentioned·. I am convinced that
more than half of the opinions that prevail among mankind
are products of education, and that the principles that are
implicitly embraced from this cause over-balance the ones
that come either from abstract reasoning or from experience.

iii: Knowledge and probability

come at last to remember them, so our judgment, or rather
our imagination, can through similar repetition have ideas
so strongly and brightly imprinted on it that they operate
on the mind in the same way as do the perceptions that
reach us through the senses, memory, or reason.6 But as
education is an artificial and not a natural cause, and as
its maxims are frequently contrary to reason and even to
one another in different times and places, philosophers don’t
take account of it ·in their theorizing about belief·, though in
reality it is built on almost the same foundation of custom
and repetition as are our ·natural· reasonings from causes
and effects.

[Hume’s word ‘over-balance’ might mean ‘outnumber’ or ‘overpower’ or
both.] As liars through the frequent repetition of their lies

10: Influence of belief
[This section discusses, with examples, ways in which imagination and belief interact with one another, always with an eye to
confirming Hume’s own theory about what belief is.]

11: The probability of chances
•knowledge and •probability, and have defined knowledge to

In order to give this system its full force and convincingness,
we should turn briefly from it to its consequences, using the
same principles to explain some other kinds of reasoning
that are derived from the same origin.

be the evidentness that arises from the comparison of ideas,
have to bring all our arguments from causes or effects under
the general label ‘probability’. I have followed suit earlier in
this book (everyone is entitled to use words as he sees fit);

Philosophers who have divided human reason into
6

I should remark that as our assent to all probable reasonings is based on the liveliness of ideas, it resembles many of the whimsies and prejudices
that are rejected as ‘mere offspring of the imagination’. From this way of talking we learn that ‘imagination’ is commonly used in two different senses;
and in the following reasonings I have used it in both of them (I know that nothing is more contrary to true philosophy than this sort of inaccuracy).
When I contrast •imagination with •memory, I mean (·broad sense·) the faculty by which we form our fainter ideas. When I contrast it with •reason,
I mean (·narrower sense·) the same faculty but excluding our demonstrative and probable reasonings. When I am not contrasting it with either
memory or reason, it doesn’t matter whether you take it in the broader or narrower sense, or at least the context will sufficiently explain the meaning.

66

Treatise, Book 1

David Hume

but really it is certain that in everyday talk we regard many
arguments from causation as having conclusions that are
certain enough to count as more than merely ‘probable’. It
would seem ridiculous to say that it is only probable that
•the sun will rise tomorrow, or that •all men must die; yet
clearly we have no further assurance of these propositions
than what experience gives us. For this reason it might be
better, in order to preserve the common meanings of words
while also marking the different levels of evidentness, to distinguish human reason into three kinds: knowledge, proofs,
and probabilities. By •‘knowledge’ I mean the assurance
arising from the comparison of ideas. By •‘proofs’ I mean
arguments that are derived from the relation of cause and
effect, and are entirely free from doubt and uncertainty. By
•‘probability’ I mean the evidentness that is still accompanied by uncertainty. It is this third sort of reasoning that I
proceed to examine ·in the present section·.
Probability—or reasoning from conjecture—can be divided
into two kinds, one based on •chance, the other on •causes.
I shall consider these in order.
The idea of cause and effect is derived from experience,
which presents us with certain ·kinds of· objects constantly
conjoined with each other, and from this produces a habit of
surveying them in that relation—a habit so strong that we
must do violence to our thoughts to ·break it and· consider
objects of those kinds in any other way. In contrast with this,
chance is nothing real in itself; strictly speaking, it is merely
the negation of a cause. So its influence on the mind is contrary to that of causation: and it is essential to chance that
it leaves the imagination perfectly free to consider either the
existence or the non-existence of the object that is regarded
as contingent ·or dependent on chance·. A cause shows our
thought the path to follow; in a way, it forces us to regard
certain objects in certain relations. All that chance does is

iii: Knowledge and probability

to destroy this compulsion of thought, leaving the mind in
its original state of indifference, ·that is, evenly balanced
between assent and dissent to the proposition· . . . .
Since it is of the essence of chance to produce complete
indifference, the only way one chance can be greater than
another is by being composed of a •greater number of •equal
chances. If we said on any other basis that one chance could
be greater than another, we would be saying that something about it made it superior to the other, pushing the
outcome to its side more than to the other’s. That is, we
would be allowing a cause into the story, thus negating what
we had started out with, namely the supposition that we
were dealing with chance. A perfect and total indifference
is essential to chance, and one total indifference can never
in itself be either greater or lesser that another. This truth
is not special to my system. It is accepted by everyone who
does calculations about chances.
This combination of chances that is needed to make one
risk greater than another brings up a remarkable fact about
•chance and •causation. The two are directly contrary, yet
we can’t conceive the combination I have mentioned without
supposing that •causes are mixed in among the •chances—
supposing •necessity in some details and total •indifference
in others. When nothing constrains the chances, every notion that the most extravagant fancy can form is on an equal
footing with every other, and there can’t be any circumstance
that could give one an advantage over the others. If we don’t
allow that there are some causes to make the dice fall, to
keep their shape when doing so, and to come to rest on one
of their sides, we can’t make any calculation about the laws
of chance. But if we suppose that those causes operate, and
suppose that all the rest is indifferent and determined by
chance, we can easily arrive at a notion of a superior combination of chances. A die that has four sides marked with a
67

Treatise, Book 1

David Hume

certain number of spots, and only two with another number,
affords us an obvious and easy instance of this superiority.
The mind is here limited by the causes to a precise number
and quality of upshots—·specifically to six possible upshots,
each consisting in the die’s coming to rest on one side·—and
at the same time it is undetermined in its choice of any of
the six.

iii: Knowledge and probability

can say for sure that it is more likely and probable that it
will fall on the side that has the greater number of chances
than that it will fall where there is a smaller number. If this
is said, I reply:
What do you mean by ‘likelihood and probability’? The
likelihood and probability of chances is a greater number of equal chances; so when you say that it is ‘likely’
that the outcome will fall on the side which has the
greater number, rather than on one having a lesser
number of chances, all you are saying is that where
there is a greater number of chances there is actually
a greater, and where there is an lesser there is a lesser.
These are identical propositions [= ‘tautologies’], and of
no significance.
So the question remains: how does a greater number of equal
chances operate on the mind to produce belief or assent? Apparently it’s not by arguments derived from demonstration,
or by ones from probability.
In order to clear up this difficulty, consider the following
case:
Someone takes a die that has a circle on four of its
sides and a square on the other two; he puts this die
into a box, intending to throw it.
Obviously, he must consider a circle to be more probable
than a square; that a circle will fall uppermost is the prediction that he must prefer. In a way he believes that a circle
will come uppermost, but with hesitation and doubt in proportion to the number of chances of a square; and if the
number of ‘square’ chances were lessened, thus increasing
the gap between it and the number of ‘circle’ chances, his
belief would become less hesitant and more confident. This
belief arises from his mind’s operations on the simple and
limited object before us, so we ought to be able to discover
and explain it. We have nothing but one single die to think

In our reasoning so far we have advanced three steps;
that •chance is merely the negation of a cause, and produces
total indifference in the mind; that •one negation of a cause
and one total indifference can never be greater or lesser than
another; and that •there must always be a mixture of causes
among the chances if any reasoning about chances is to have
a basis. Now we must move on, and consider what effect a
greater combination of chances has on our mind—how does
it influence our judgment and opinion? Here I can repeat
all same arguments that I employed in examining the belief
that arises from causes; and can prove in the same way
that neither •demonstration nor •probability has any role in
getting a greater number of chances to produce our assent .
·I shall take these one at a time·.
·Regarding •demonstration·: It is indeed obvious that
mere comparison of ideas can never reveal to us anything
relevant to our present question: it is impossible to prove
with certainty that any outcome must fall on the side that
has the greater number of chances. To suppose there is
any certainty about this would be to overthrow what I have
established about the perfect equality of opposing ·single·
chances and the indifference of the mind with respect to
them.
·Regarding •probability·: It might be said that though in
an opposition of chances it is impossible to determine with
certainty on which side the outcome will fall, we nevertheless
68

Treatise, Book 1

David Hume

about, in order to grasp one of the most curious operations
of the understanding. [By ‘curious’ Hume probably means something like ‘intricate and challenging’.] We should attend to three
facts about the die that I have described. •First, certain
causes—gravity, solidity, cubic shape, etc.—will cause it to
fall, remain unaltered during the fall, and come down with
one side uppermost. •Secondly, it has a particular number
of sides, which are supposed indifferent—·that is, which are
supposed to be such that there is no reason to expect any
one rather than other to fall uppermost·. •Thirdly, on each
side a certain figure is inscribed. These three facts constitute the whole nature of the die, so far as we are concerned
here, and so they are the only things the mind can go by
when forming a judgment about how the die will fall. So let
us consider slowly and carefully what influence these facts
must be having on our thought and imagination.
•First, I have already observed that custom makes the
mind pass from any cause to its effect, and that when one
appears it is almost impossible for the mind not to form an
idea of the other. . . . When it thinks of the die as no longer
supported by the box, the mind can’t without •violence ·to
itself· regard it as suspended in the air. Rather, it •naturally
imagines it as lying on the table with one of its sides uppermost. This is an effect of the admixture of causes that is
needed if we are to make any calculation about chances.
•Secondly, we are supposing that though the die must fall
and turn up one of its sides, there is nothing to fix the particular side, this being determined entirely by chance. The
very nature and essence of chance is a negation of causes
and leaving the mind in complete indifference among those
outcomes that are supposed to be contingent, ·i.e. at the
mercy of chance·. So when the causes make our thought
consider the die as falling and turning up one of its sides,
the chances present all these sides as equal, and make us

iii: Knowledge and probability

regard each of them as being just as probable and possible
as each of the others. The imagination passes from the cause
to the effect—from the throwing of the die to the turning up
one of the six sides—and feels itself as somehow unable to
make this process stop short or terminate in some other
idea. But only one side can lie uppermost at a time, and
the causal factors don’t make us think of the sides as all
lying uppermost together, which we regard as impossible;
nor do they direct us with their entire force to any particular
side, for if they did, the chosen side would be considered as
certain and inevitable. Rather, the causal factors direct us
to the whole six sides in such a way as to divide their force
equally among them. We conclude in general that some one
of them must result from the throw; we run all of them over
in our minds; the forces acting on our thought are common
to all of them; but what they exert with respect to any one
outcome is no more than what is suitable given what proportion of the whole it makes. This is how the original impulse,
and consequently the liveliness of thought arising from the
causes, is divided and split in pieces by the intermingled
chances.
So now we have seen the influence of the two first aspects
of the die—the causes, and the number and indifference of
the sides—and have learned how they give a push to our
thought, and divide that push into as many parts as there
are sides. We must now look into the effects of •the third
factor, namely the figures inscribed on the sides. Obviously,
where several sides have the same figure inscribed on them,
they must work together in their influence on the mind,
bringing to bear on one image or idea of the figure all those
divided pushes that were scattered over the several sides
that have that figure on them. If we were asking ‘Which
side will fall uppermost?’, all the sides would be perfectly
equal, and no-one could have any advantage over any other.
69

Treatise, Book 1

David Hume

But the question is ‘Which figure will fall uppermost?’; and
as the same figure is exhibited by more than one side, it is
obvious that the pushes belonging to all those sides must
come together on that one figure, and become stronger and
more forcible by their union. In our example, four sides
have a circle, two have a square. The pushes on the circle are therefore more numerous than the pushes on the
square. But as the outcomes are contrary—it can’t happen

iii: Knowledge and probability

that circle and square both turn up in a single throw—the
pushes likewise become contrary; the weaker force destroys
the stronger as far as it has strength to do so; ·and what
remains of the stronger one after the weaker has expended
itself is the mind’s probability- judgment about the outcome·.
The liveliness of the idea is always proportional to the degrees of the push or tendency to make the transition; and
according to my doctrine that liveliness of the idea is belief.

12: The probability of causes
The only use for what I have said about the probability of
•chances is to help us explain the probability of •causes,
since it is commonly allowed by philosophers and scientists
that what plain people call ‘chance’ is really a secret and
concealed cause. The latter sort of probability, therefore, is
what we must chiefly examine.

Although this sort of probability comes before proof, and
naturally takes place before any entire proof can exist, when
people reach maturity they no longer have anything to do
with it. It often happens of course that someone with the
most advanced knowledge achieves only an imperfect experience of some particular conjunctions of events, which
naturally produces ·in him· only an imperfect habit and
transition; but then we must consider that the mind, having
formed another observation concerning the connection of
causes and effects, gives new force to its reasoning from
that observation [Hume’s exact words from ‘;’ to here]; and by this
means the mind can build an argument on one single experiment if it is properly prepared and examined. What we
have found once to follow from an object ·of some kind· we
conclude will always follow from it [= ‘from objects of that kind’];
and if we don’t always build on this maxim as a certainty,
it is not because •we haven’t observed a large enough number of experiments but because •we have often met with
instances to the contrary. And that leads us to ·the topic of
this section·, namely the second kind of probability, where
there is a contrariety in our experience and observation.

The probabilities of causes are of several kinds, but all
come from the same source, namely the association between
a present impression and certain ideas. As the habit that
produces the association comes from the frequent conjunction of ·kinds of· objects, it ·can’t spring into existence all at
once, but· must arrive at its full force gradually, gaining new
force from each instance that we observe. The first instance
has little or no force, the second adds a little to it, the third
becomes still more noticeable; and it is by these slow steps
that our judgment arrives at full confidence. But before it
reaches such completeness it passes through several lower
degrees, and in all of them it is to be regarded as only a
presumption or probability. So the gradation from probabilities to proofs is in many cases imperceptible, and large
differences between these kinds of confidence are easier to
perceive than small ones.

It would be very happy for men in the conduct of their
70

Treatise, Book 1

David Hume

lives and actions if the same ·kinds of· objects were always
conjoined, and we had nothing to fear but the mistakes of our
own judgment, with no reason to allow for the uncertainty of
Nature. But as it is often found that one observation conflicts
with another, and that causes and effects don’t follow in the
same way that we have experienced in the past, I have to
modify my theory so as to take into account this uncertainty,
paying attention to the contrariety of outcomes. I start with
the question of the nature and causes of this contrariety.

iii: Knowledge and probability

But however philosophers and scientists may differ from
common folk in how they •explain the contrariety of outcomes, their •inferences from it are always of the same kind
and based on the same principles. A contrariety of outcomes
in the past may give us a kind of hesitating belief for the
future, in either of two ways. First, by producing an imperfect habit and transition from the present impression to
the related idea. When the conjunction of any two objects
is frequent but not entirely constant, the mind is pushed
towards passing from one object to the other, but not with
such a complete habit as when the conjunction has been
without exceptions and all the instances we have ever met
with are uniform and of a piece. . . . There is no doubt that
this is sometimes what happens, producing the ·tentative·
inferences we draw from contrary phenomena; but I am
convinced that it isn’t what mainly influences the mind in
this sort of reasoning. When our mind is moved purely by
our habit of transition, we make the transition without any
reflection, and don’t have a moment’s delay between seeing
one object and believing in the other that is often found to accompany it. The custom doesn’t depend on any deliberation,
so it operates immediately, without allowing time to think.
But it is very seldom like this in our probable reasonings. . . .
In the latter usually take account of the contrariety of past
outcomes, knowing that we are doing so: we compare the
different sides of the contrariety, and carefully weigh the evidence that we have on each side. From this we can conclude
that our reasonings of this kind arise from habit not directly
but in an oblique manner which I must now try to explain.

Common folk, who judge things according to their first
appearance, attribute the uncertainty of outcomes to an uncertainty in the causes—they think that the causes often fail
to have their usual influence even when they don’t meet with
any obstacle to their operation. But philosophers and scientists, observing that almost every part of Nature contains a
vast variety of mechanisms and forces that are hidden from
us because they are so small or so distant, think it at least
possible that the contrariety of outcomes may come not from
any contingency [here = ‘unreliability’] in the cause but rather
from the secret operation of contrary causes. This possibility
becomes certainty when they bear in mind that when any
contrariety of effects is studied carefully it always turns out
that it does come from a contrariety of causes, and proceeds
from their mutual hindrance and opposition. A peasant can
give no better reason for a clock’s stopping than to say ‘It
often doesn’t go right’; but a clockmaker easily sees that
the same force in the spring or pendulum always has the
same influence on the wheels, but has failed of its usual
effect because of a grain of dust that puts a stop to the whole
movement. Having observed various cases of this general
kind, philosophers and scientists form a maxim that the connection between all causes and effects is equally necessary,
and that its seeming unreliability in some cases comes from
the secret opposition of contrary causes.

Obviously, when a kind of object has contrary effects ·at
different times·, we base our opinions about them purely on
our past experience, and always consider as possible any effects that we have observed to follow from this kind of object.
And just as past experience regulates our judgments about
71

Treatise, Book 1

David Hume

the possibility of these effects, so it also regulates what we
think about their probability; and we always take to be the
most likely the effect that has been the most common. So we
have two things to think about here: •why we treat the past
as a standard for the future, and •how we extract a single
judgment from a contrariety of past outcomes.
First ·the question of why·: The supposition that the
future resembles the past isn’t based on arguments of any
kind, and comes solely from a habit that makes us expect
for the future the same sequence of events as we have been
accustomed to ·in the past·. This habit or push to transfer
the past to the future is full and perfect; and therefore the
first impulse of the imagination in this kind of reasoning is
full and perfect too.
Secondly ·the question of how·: When we look back on
past experiences and find them to be contrary, this push ·to
transfer the past to the future·, though full and perfect in
itself, doesn’t take us to any one steady object, but offers us
a number of disagreeing images in a certain order and proportion. So in this case the first impulse ·of the imagination·
is split up and diffuses itself over all those images, each of
them having an equal share of the force and liveliness that
the impulse gives. Any of these past outcomes may happen
again, and we think that when they do happen they will be
mixed in the same proportion as in the past.
[A long paragraph spelling this out in more detail. A
notable episode is this:] Each new experience ·of a causeeffect pair· is like a new brush-stroke, which gives additional
liveliness to the colours without altering any of the shapes.
Summing up, then: experiences with contrary outcomes
produce an imperfect belief, either •by weakening the habit,
or by •dividing and then recombining the perfect habit that
makes us conclude in general that instances of which we
have no experience must resemble those of which we have.

iii: Knowledge and probability

To justify still further this account of the second sort of
probability, where we reason with knowledge and reflection
from a contrariety of past experiences, I shall propose some
further considerations. (They have an air of subtlety, but
don’t hold that against them. Sound reasoning oughtn’t to
lose any of its force through being subtle; just as matter
retains its solidity in air and fire and animal spirits, as well
as in larger and more perceptible forms.) [Two points about that
sentence. •It involves a half-suppressed pun: it was standardly said that
air etc. differ from rocks etc. in being more ‘subtle’, meaning more finely
divided. •When Hume implies that air is as ‘solid’ as rock, he means that
it won’t share its space with any other bodies, any more than rock will.]

[The two-page argument that follows is subtle and ingenious, but it is exhausting to read and follow, and seems not
to add much to what Hume has already said. He follows it
with something else, equally demanding, that he describes as
‘almost the same argument in a different light’. This material
is omitted from the present version.]
I am aware of how abstruse all this reasoning must appear to the general run of readers—people who aren’t accustomed to going so deeply into the intellectual faculties of the
mind, and so will be apt to reject as fanciful anything that
doesn’t fit with common received notions and with the easiest
and most obvious principles of philosophy. You do have to
take some trouble to follow these arguments of mine, though
it takes very little trouble to see to see ·how bad the rival
accounts are·—to see the imperfection of every plain-man
hypothesis on this subject, and how little light philosophy
has so far been able to cast in these elevated and challenging
inquiries. If you can once be fully convinced that
•Nothing in any object, considered in itself, can give
us a reason for drawing a conclusion about anything
other than that object, and
•Even after observing the frequent or constant con72

Treatise, Book 1

David Hume

junction of objects, we have no reason to draw any
inference about any object other than those of which
we have had experience,
these two principles will throw you so loose from all common
systems that you will have no trouble accepting other theses
that may appear very extraordinary. These principles proved
to be sufficiently convincing when applied to our most certain
reasonings from causation; but I venture to say that they
become even more believable when applied to the conjectural
or probable reasonings that are our present topic.
[•Hume then goes again through his account of probabilistic reasoning, bringing out how it requires (and makes
plausible) the two principles in question. •Then two paragraphs in which he presents ‘two reflections which may
deserve our attention’. One concerns the difference between
experiencing contrary outcomes and merely imagining them.
The other concerns (in effect) the mathematics of adding
belief-strengths, which Hume says has ‘a parallel instance
in the affections’. The core of his view about the latter is
that ‘a man who desires a thousand pounds has in reality a
thousand or more desires which unite together and seem to
make only one passion’.]
Beside these two sorts of probability—derived from
•imperfect experience and from •contrary outcomes—there
is a third arising from •analogy, which differs from them in

iii: Knowledge and probability

some significant respects. According to the account I have
given, all kinds of reasoning from causes or effects are based
on two things: •the constant conjunction of any two ·kinds
of· objects in all past experience, and •the resemblance of a
present object to one of the kinds. These have the effect that
•the present object invigorates and enlivens the imagination,
and •the resemblance together with •the constant union conveys this force and liveliness to the related idea, which we
are therefore said to believe. If you weaken either the •union
or the •resemblance, you weaken the force of transition and
thereby weaken the belief that arises from it. The liveliness
of the first impression can’t be fully transferred to the related idea unless •the conjunction of objects of their kinds
has been constant and •the present impression perfectly
resembles the past ones whose union we have been accustomed to observe. In probabilities of chance and of causes
(discussed above) it is •the constancy of the union that is
diminished; and in the probability derived from analogy it
is only •the resemblance that is diminished. Without some
degree of resemblance there can’t be any reasoning. But this
resemblance can be greater or smaller, and the reasoning
is proportionally more or less firm and certain. An experience loses some of its force when transferred to instances
that don’t exactly resemble it; but as long as there is some
resemblance remaining there is still a basis for probability.

13: Unphilosophical probability
four of them·.

The three kinds of probability that I have described are all
accepted by philosophers as reasonable bases for belief and
opinion. But there are other kinds that are derived from
the same principles but haven’t had the good fortune to be
accepted in the same way. ·In this section I shall discuss

The first probability of this kind can be described like
this. The vividness of the inferred idea may be lessened by a
lessening of •the union or of •the resemblance, and also—I
now add—by a lessening of •the impression. . . . The argu73

Treatise, Book 1

David Hume

ment that we base on a remembered matter of fact is more
or less convincing according to whether the fact is recent
or remote in time. This source for difference in degrees of
evidentness is not accepted by philosophy as solid and legitimate, because ·if it is accepted, then· an argument must
have more force today than it will have in a month’s time.
But despite the opposition of philosophy, the remotenessin-time aspect certainly has a considerable influence on the
understanding, and secretly changes the authority of an
argument, depending on when it is put to us. . . .
A second source of difference in our degrees of belief and
assurance, always disclaimed by philosophers but always
effective, is this. An experience that is recent and fresh
in the memory affects us more, having a greater influence
on judgment as well as on the passions than one that is
in some measure obliterated. A lively impression produces
more assurance than a faint one, because it has more initial
force to pass on to the related idea, which thereby gets more
force and liveliness. Similarly with a recent observation: the
custom and transition is more complete in that case, and
preserves better the initial force of what is transferred. Thus
a drunkard who has seen his companion die from a drinkingspree is struck with that instance for some time, and dreads
having such an accident himself; but as the memory of it
gradually decays, his former sense of security returns and
the danger comes to seem less certain and real.
I add as a third instance ·of unphilosophical probability·
the following. Although our reasonings from •proofs are considerably different from our reasonings from •probabilities,
the former kind of reasoning often slides imperceptibly into
the latter simply because the proof in question involves so
many connected arguments. When an inference is drawn
immediately from an object without any intermediate cause
or effect, the conviction is much stronger . . . . than when the

iii: Knowledge and probability

imagination is carried through a long chain of connected arguments, however infallible the connection of each link may
be thought to be. The liveliness of all the ideas comes from
the original impression, through the customary transition of
the imagination; and it is obvious that this liveliness must
be gradually lessened in proportion to the distance that the
transition has to cover. Sometimes this distance does more
to reduce conviction than even contrary experiences would
have done; and a man may receive a livelier conviction from
a probable reasoning that is brief and immediate than from
a long chain of consequences, even if the latter is sound and
conclusive in each part. Indeed, reasons of the latter kind
seldom produce any conviction: one must have a very strong
and firm imagination to preserve the evidentness through so
many stages right to the end!
An odd point arises here, ·which I shall state in the form
of an objection to what I have been saying·:
There is no point of ancient history of which we can
have any assurance except through many millions of
causes and effects, and through a chain of inferences
of an almost immeasurable length. Before the knowledge of the fact could come to the first historian, it
must be conveyed through many mouths; and after it
is committed to writing, each new copy is a new object
whose connection with the previous one is known only
by experience and observation. From what you have
been saying about strength of belief it seems to follow
that the evidentness of all ancient history must now
be lost, or at least will be lost in time as the chain
of causes gets ever longer. But it seems contrary to
common sense to think that if the world of scholarship
and the art of printing continue in the same way that
they do now, our descendants will some day come
to doubt that there ever was such a man as Julius
74

Treatise, Book 1

David Hume

Caesar. So this looks like an objection to the account
you have been giving. If belief consisted (·as you say
it does·) only in a certain liveliness conveyed from
an original impression, it would fade in accordance
with the length of the transition, and would eventually have to be utterly extinguished. And if belief is
sometimes not capable of such an extinction, it must
be something different from that liveliness.
(Before I answer this objection I should remark that this
line of thought has generated a very celebrated argument
against the Christian religion, with just one difference: ·in
the anti-Christianity argument· it is supposed that each link
of the chain of human testimony is only probabilistically
sound, and to be ·in itself· liable to some doubt and uncertainty. And it must be admitted that in this way of looking
at the subject—which is not the correct one—every history
and tradition must indeed eventually lose all its force and
convincingness. Every new probability lessens the original
conviction; and however great that conviction may be, it can’t
continue under such repeated lessenings. This is true in
general, though we shall find in 1iv that there is one very
memorable exception, a vastly important one for our present
topic of the understanding.)
Meanwhile, to answer the preceding objection on the
supposition that historical evidence amounts initially to a
complete proof, bear in mind that though the links connecting any historical fact with a present impression are very
numerous, they are all of the same kind, depending ·only·
on the reliability of printers and copyists. One edition is
succeeded by another, and that by a third, and so on, till the
chain reaches the history book we are now reading. There is
no variation in the steps. After we know one, we know them
all; and after we have taken one ·inferential step· we can’t
hesitate to take all the others. This is enough to preserve the

iii: Knowledge and probability

convincingness of history. . . .
A fourth unphilosophical sort of probability, ·which will
be the topic of the remainder of this section·, is derived from
general rules that we rashly form to ourselves—rules that
are the source of what we properly call prejudice [the Latin
root of which means ‘pre- judgment’]. An Irishman can’t have wit,
and a Frenchman can’t have solidity; so even in particular cases where the Irishman talks entertainingly and the
Frenchman talks judiciously, we have held such a prejudice
against them that ·we think· they must be a dunce and a fop
·respectively·, in spite of sense and reason.
Human nature is very given to errors of this kind, and
perhaps this nation as much as any other! Why do men
form general rules and allow them to influence their judgment, even contrary to present observation and experience?
I think that it comes from the very same sources as to all
judgments about causes and effects. [In the rest of this
paragraph, Hume reminds us of his account of causal and
probabilistic reasoning, especially stressing how the latter
may be weakened by imperfect resemblances amongst the
instances.]
Although custom is •the basis of all our judgments, sometimes it has an effect on the imagination •in opposition to
the judgment, and produces a contrariety in our views about
the same object. Let me explain. In most kinds of causes
there is a complication of factors, some essential and others
superfluous, some absolutely required for the production of
the effect and others present only by accident. Now, when
these superfluous factors are numerous and remarkable and
frequently conjoined with the essential factors, they influence the imagination so much that even in the absence of
something essential they carry us on to the idea of the usual
effect, giving it a force and liveliness that make it superior
to the mere fictions of the imagination. We can correct this
75

Treatise, Book 1

David Hume

propensity by reflecting on the nature of the factors on which
it is based; but it is still certain that custom starts it off and
gives a bias to the imagination.
To illustrate this by a familiar example: a man who is
hung out from a high tower in a cage of iron can’t help trembling when he sees the drop below him, even though his
present experience of the solidity of the iron that supports
him tells him that he is perfectly safe from falling, and the
idea of falling and harm and death come only from custom
and ·past · experience. That custom goes beyond the instances from which it is derived and to which it perfectly
corresponds—·instances in which heavy things are released
without support and fall to the ground·—and influences his
ideas of objects that resemble the others in some respects
but don’t precisely fit the same rule. The factors of •depth
and descent impress him so strongly that their influence
can’t be destroyed by the contrary factors of •support and
solidity, which ought to make him feel perfectly safe. His
imagination runs away with its object, ·the thought of falling·,
and arouses a passion (·fear·) proportional to it. That passion reacts back on the imagination, and enlivens the idea;
this newly enlivened idea has a new influence on the passion,
increasing its force and violence; so his imagination and
his feelings mutually support each other, causing the whole
·situation· to have a very great influence upon him.
But why need we look for other instances, when the
present subject of unphilosophical probabilities offers us
such an obvious one, in the conflict between judgment and
imagination that arises from custom? ·I shall explain this by
presenting an apparent difficulty for my account·:
According to my theory, reasonings are merely effects
of custom, and custom’s only influence is to enliven
the imagination and give us a strong conception of
some object. So it seems to follow that our judgment

iii: Knowledge and probability

and our imagination can never be in conflict—that
custom can’t operate on the imagination in such a
way as to put it in opposition to the judgment. ·But
we have seen that they do sometimes conflict with one
another; so this is a problem for my theory·.
The only solution for this difficulty is to bring in the influence of general rules. In section 15 I shall call attention to
some general rules by which we ought to regulate our judgment about causes and effects; and these rules are based
on the nature of our understanding, and on our experience
of how it operates in our judgments about objects. Through
those rules we learn to distinguish accidental circumstances
from effective causes; and when we find that an effect can
be produced in the absence of a certain factor we conclude
that that factor is not part of the effective cause, however
often it is conjoined with it. But this frequent conjunction
necessarily makes the factor in question have some effect
on the imagination, in spite of the opposite conclusion from
general rules; and so the opposition of these two principles
produces a contrariety in our thoughts, and makes us ascribe •one inference to our judgment, and •the other to our
imagination. The general rule is attributed to our judgment
because it is more extensive and constant; the exception to
the general rule is credited to the imagination because it is
more capricious and uncertain.
Thus our general rules are in a way set in opposition to
each other. When an object appears that •resembles some
cause in very considerable respects, the imagination naturally carries us to a lively conception of the usual effect, even
if the object •differs from that cause in the most significant
and effective respects. Here—·in this wrong transition to
an idea of the usual effect·—is the first influence of general
rules. But when we review this act of the mind and compare
it with the more general and authentic operations of the un76

Treatise, Book 1

David Hume

iii: Knowledge and probability

two cases is that •in openly revealing his sentiments he
uses signs that are general and universal, while •in secretly
indicating them he uses signs that are more singular and uncommon. And when the imagination runs from the present
impression ·of the man’s words or behaviour· to the absent
idea ·of his hostility or contempt·, it makes the transition
more easily—and so conceives the object with greater force—
when the connection is •common and universal than when
it is more •rare and particular. . . .
[Hume adds a further paragraph and a half, adding detail
to this, and offering a reflection on reasons why sometimes
‘scurrility is less displeasing than delicate satire’.]
To this account of the different influence of open and
concealed flattery or satire, I shall add the consideration of
another phenomenon that is analogous to it. There are many
violations of codes of honour that the world —though not excusing them—is more apt to overlook when the appearances
are saved and the transgression is secret and concealed.
(This holds for both men and women.) People who know perfectly well that the fault has been committed pardon it more
easily when the proofs seem somewhat indirect and ambiguous than when they are direct and undeniable. In both cases
the same idea is presented, and strictly speaking is equally
assented to by the judgment; but its influence is different
because of the different ways in which it is presented. . . . The
difference is just this: in the first case the sign from which
we infer the blamable action is single, and suffices all on its
own to be the basis for our reasoning and judgment; whereas
in the second case the signs are numerous, and decide little
or nothing when taken alone and not accompanied by many
minute and almost imperceptible factors. Any reasoning is
convincing in proportion as it is single and united to the
eye, and gives less work to the imagination in collecting its
parts and going from them to the correlative idea that is the

derstanding, we find it to be irregular and destructive of all
the most established principles of reasoning, which causes
us to reject it. This is a second influence of general rules, and
implies the condemnation of the first one. Sometimes one
prevails, sometimes the other, according to the disposition
and character of the person. Ordinary folk are commonly
guided by the first, and wise men by the second. Meanwhile
sceptics can enjoy this prospect of a new and notable contradiction in our reason, and of seeing all philosophy ready to
be subverted by a force in human nature and then saved by
giving a new direction to the very same force! The following
of general rules is a very unphilosophical sort of probability,
but it is only by following them that we can correct this and
all other unphilosophical probabilities.
Since we have instances where general rules act on the
imagination contrary to the judgment, we needn’t be surprised to see their effects increase when they combine with
the judgment, presenting to us ideas that have more force
than any others. Everyone knows there is an indirect manner
of insinuating praise or blame, which is much less shocking than the open flattery or censure of any person. Even
if someone does communicate his sentiments by such secret insinuations, making them known just as certainly as
openly revealing them would, their influence is not equally
strong and powerful. Someone who lashes me with concealed
strokes of satire doesn’t move me to indignation as intensely
as if he had flatly told me I was a fool and coxcomb, though
I understand his meaning just as well as I would if he had
done that. This difference is to be attributed to the influence
of general rules.
Whether a person •openly abuses me or •slyly indicates
his contempt, in neither case do I immediately perceive his
sentiment or opinion; I become aware of it only by signs,
that is, by its effects. So the only difference between these
77

Treatise, Book 1

David Hume

conclusion. . . .
[In a final pair of paragraphs Hume re-states his main
conclusions in sections 11–13, contending that they are con-

iii: Knowledge and probability

firmed by their ability to interlock and solve problems, and
that their success helps to confirm his account of belief.]

14: The idea of necessary connection
me the idea ·of necessity· which I am at present examining.
For after a frequent repetition I find that on the appearance
of one of the objects, custom makes the mind think of its
usual attendant, and think of it more vividly on account of
its relation to the first object. So it is this impression, this
being-made-to-think-of-the-effect, that gives me the idea of
necessity.

Having thus explained how we reason beyond our immediate
impressions, and conclude that such and such causes must
have such and such effects, we must now retrace our steps
and pick up again the question that first occurred to us, and
that we dropped along the way (near the end of section 2).
The question is: What is our idea of necessity, when we say
that two objects are necessarily connected? As I have often
said already, if we claim to have such an idea we must find
some impression that gives rise to it, because we have no
idea that isn’t derived from an impression. So I ask myself:
In what objects is necessity commonly supposed to lie? And
finding that it is always ascribed to causes and effects, I
turn my attention to two objects that are supposed to be
related as cause and effect, and examine them in all the
situations in which they can occur. I see at once that they
are contiguous in time and place, and that the one we call
‘cause’ precedes the one we call ‘effect’. In no instance can I
go any further: I can’t find any third relation between these
objects. So I take a broader view, and consider a number
of instances in which I find objects of one kind always existing in relations of contiguity and succession with objects
of another kind. At first sight this seems to be pointless:
the reflection on several instances only repeats the same
objects, so it can’t give rise to any new idea. But on further
enquiry I find that the repetition is not the same in every
respect. It produces a new impression ·that I don’t get from
any single instance·, and through that impression it gives

I’m sure that you will have no trouble accepting this result, as being an obvious consequence of principles that I
have already established and have often employed in my
reasonings. This obviousness, both of the first principles
and of the inferences from them, may seduce you into incautiously accepting the conclusion, making you imagine that it
contains nothing extraordinary or worth thinking about. But
although such casualness may make my reasoning easier
to accept, it will also make it easier to forget; so I think I
should warn you that I have just now examined one of the
most elevated questions in philosophy, the one that seems
to involve the interests of all the sciences—namely the question about the power and efficacy of causes. That warning
will naturally rouse your attention and make you ask for a
fuller account of my doctrine, as well as of the arguments on
which it is based. This request is so reasonable that I can’t
refuse to comply with it, especially because I have hopes that
the more my principles are examined the more forceful and
convincing they will be.
There is no question which, on account of its importance
78

Treatise, Book 1

David Hume

as well as its difficulty, has caused more disputes among
both ancient and modern philosophers than this one about
the •efficacy of causes, •the quality that makes an effect
follow a cause. But before they embarked on these disputes,
I think, they would have done well to examined what idea we
have of the •efficacy they are arguing about. This is what I
find principally lacking in their reasonings, and what I shall
here try to provide.
I begin by observing that the words ‘efficacy’, ‘agency’,
‘power’, ‘force’, ‘energy’, ‘necessity’, ‘connection’, and ‘productive quality’, are all nearly synonymous, which makes it
absurd to employ any of them in defining any of the others.
This observation rejects at once all the common definitions
that philosophers have given of ‘power’ and ‘efficacy’. Our
search for the idea must be directed not to these definitions
but to the impressions from which it was originally derived.
If it is a compound idea, it must arise from compound impressions. If simple, from simple ones.
I believe that the most widely accepted and most popular

iii: Knowledge and probability

of existence. I have explained these two points already, so I
shan’t go on about them here.
I shall only infer from them that since •reason can never
give rise to the idea of efficacy, that idea must be derived
from •experience—from particular instances of this efficacy
which get into the mind through the common channels of
sensation or reflection. . . . If we claim to have a sound idea
of this efficacy, we must produce some instance in which the
efficacy is plainly revealed to the mind and its operations
are obvious to our consciousness or sensation. If we evade
this demand, we are admitting that the ·so-called· idea ·of
efficacy· is impossible and imaginary; since the only other
escape is to plead that the idea is an innate one, and ·that
escape-route is blocked because· the theory of innate ideas
has been already refuted and is now almost universally rejected in the learned world. What we have to do, then, is
to find some natural cause-effect pair in which the mind
can grasp—clearly, unambiguously, and securely—how the
cause operates and what gives it its efficacy.
We don’t get much encouragement in this from the enormous variation that we find in the opinions of philosophers
who have claimed to explain the secret force and energy of
causes. Various philosophers have variously contended that
bodies operate by
their substantial form,
their accidents or qualities,
their matter and form,
their form and accidents,
certain powers and faculties distinct from all the above.
Further, all these opinions are mixed and varied in a thousand different ways, creating a strong presumption that none
of them is solid or credible, and that there are simply no
grounds for thinking that any of the known qualities of matter has any kind of efficacy. This presumption gains strength

[here = ‘appropriate for ordinary folk who lack philosophical skills and
knowledge’] explanation of our idea of power is to say this:

We find from experience that various new productions
occur in the world of matter, such as the motions and
variations of bodies; and we conclude that there must
somewhere be a power capable of producing them;
and this reasoning brings us at last to the idea of
power and efficacy. (Thus Mr Locke, in his chapter on
Power) [Essay Concerning Human Understanding II.xxi.1])
But to be convinced that this explanation is more popular
than philosophical we need only to remember two very obvious principles. First, •that reason alone can never give rise
to any original idea, and secondly •that reason, as distinct
from experience, can never make us conclude that a cause or
productive quality is absolutely required for every beginning
79

Treatise, Book 1

David Hume

when we consider that substantial forms and accidents and
faculties are not really among the known properties of bodies,
but are perfectly unintelligible and inexplicable. Obviously
philosophers would never have had recourse to such obscure
and uncertain notions if they had met with any satisfaction
in ideas that are clear and intelligible; especially in such
an affair as this, which must be an object of •the simplest
understanding if not of •the senses. The bottom line is this:
we can conclude that it is impossible in any one instance ·of
a cause-effect pair· to show what it is that contains the force
and agency of the cause; and that in this respect the most
refined understandings are on a par with the plain man in
the street. If you think you can refute this assertion, you
needn’t take the trouble to invent any long arguments; all
you need to do is to show us an instance of a cause where
we discover the power or operating force. We often have to
use this kind of challenge, as being almost the only means
of proving a negative in philosophy.

iii: Knowledge and probability

This conclusion leads them to another which they regard
as entirely inescapable. ·They argue like this·:
Matter is in itself entirely inactive and deprived of any
power to produce or continue or communicate motion; but these effects are evident to our senses, and
the power that produces them must be somewhere.
So it must lie in God, the divine being who contains
in his nature all excellency and perfection. So God
is the first mover of the universe: he not only first
created matter and gave it its initial push, but also
through a continuing exertion of his omnipotence he
keeps it in existence and gives it all its motions and
configurations and qualities.
This opinion is certainly very interesting, and well worth
our attention; but if you think for a moment about why
it has come up for us in our present inquiry, you will see
that we needn’t examine it in detail here. We have settled
it as a principle that, because all ideas are derived from
some previous perceptions, we can’t have any idea of •power
and efficacy unless instances can be produced in which this
•power is perceived to exert itself. These instances can never
be discovered in body, so the Cartesians have relied on their
principle of innate ideas and had recourse to a God whom
they think to be the only active being in the universe, and
the immediate cause of every alteration in matter. But given
the falsity of the principle of innate ideas, the supposition
of a God can’t be of any use to us in accounting for the
idea of agency which we can’t find among the objects that
are presented to our senses or those that we are internally
conscious of in our own minds. For if every idea is derived
from an impression, the idea of a God must come from the
same origin; and if no impression, either of sensation or reflection, implies any force or efficacy, it is equally impossible
to discover or even imagine any such active force in God. So

The failures of their attempts to pin down this power has
finally obliged philosophers to conclude that the ultimate
force and efficacy of Nature is perfectly unknown to us, and
that it is no use looking for it among the known qualities
of matter. They are almost unanimous about this; where
their opinions differ it is in what they infer from it. Some of
them, especially the Cartesians, have satisfied themselves
that we are acquainted with the whole essence of matter,
which they say consists in extension. Now, extension doesn’t
imply actual motion, but only mobility; so they naturally conclude that when matters moves, the energy that produces
the motion can’t lie in the extension, ·which means (for them)
that it can’t lie in the matter·. So, they conclude, matter is
not endowed with any efficacy, and can’t possibly (unaided)
communicate motion or produce any of the effects that we
ascribe to it.
80

Treatise, Book 1

David Hume

when these ·Cartesian· philosophers argue that
No efficacious force can be discovered in matter, so
no such force should be attributed to matter,
they ought by parity of reasoning to argue
No efficacious force can be discovered in God, so no
such force should be attributed to God.
If they regard that conclusion as absurd and impious, as
indeed it is, I shall tell them how they can avoid it—namely,
admitting at the outset that they have no adequate idea of
power or efficacy in any object, since they can’t discover a
single instance of it in bodies or in minds, in divine natures
or in creaturely ones.
The same conclusion is unavoidable on the hypothesis
of those who maintain the efficacy of subordinate causes,
and credit matter with having a power or energy that is real
but derivative. For they grant that this energy doesn’t lie in
any of the known qualities of matter, so ·for them as for the
Cartesians· the difficulty still remains about the origin of the
idea of it. If we really have an idea of power we can attribute
power to an •unknown quality; but
the idea couldn’t be derived from a quality that we
don’t know, and there is nothing in •known qualities
that could produce the idea,
so it follows that it is mere self-deception for us to imagine
we have any idea of this kind in the way we ordinarily think
we do. All ideas are derived from and represent impressions.
We never have any impression that contains any power or
efficacy. So we never have any idea of power.
Some have asserted that we feel an energy or power in
our own mind, and that having acquired the idea of power in
this way we transfer that quality to matter, where we can’t
7

iii: Knowledge and probability

immediately discover it. The motions of our body and the
thoughts and sentiments of our mind (they say) obey the will,
and we needn’t look beyond that for a sound notion of force
or power. But to convince us of how fallacious this reasoning
is, we need only notice that the will—which they are taking
to be a cause—doesn’t have a discoverable connection with
its effects any more than any material cause has one with
its effect. We are so far from perceiving the connection between •an act of volition and •a bodily movement that it is
generally agreed that the powers and essence of thought and
matter come nowhere near to providing an explanation for
the relation between willing to make a certain movement and
making it. And the will’s power over our mind is no more
intelligible. In that case ·too· the effect is distinguishable
and separable from the cause, and couldn’t be foreseen without the experience of their constant conjunction. We can
effectively command our thoughts up to a certain point, but
not beyond that; and it is only by consulting experience that
can know where the boundaries to our authority lie. (·For
example, I can think about horses just by choosing to think
about horses; but I can’t rapidly run through thoughts of the
first nineteen prime numbers or believe that the earth is flat
just by choosing to do so; and it is only from experience
that I know what I can do just by choosing to and what I
can’t—none of it ‘stands to reason’, none of it can be seen to
be expectable given the nature of the will’s command over
thoughts·.) In short, so far as our present topic goes, the
actions of the mind are like the actions of matter: all we perceive is constant conjunction, and we can’t reason beyond
it. . . . We have no chance of attaining an idea of force by
consulting our own minds.7

Our ideas of God are similarly imperfect, but this can’t have any effect on either religion or morals. The order of the universe proves that there is an
omnipotent mind, that is, a mind whose will is constantly accompanied by the obedience of every creature and being. That’s all that is needed as a
basis for all the articles of religion; we don’t need to form a distinct idea of God’s force and energy.

81

Treatise, Book 1

David Hume

It has been established as a certain principle that general
or abstract ideas are nothing but individual ones looked at
in a certain way, and that when we reflect on any object
we have to bring into our thought its particular degrees of
quantity and quality—just as the object itself has to have
particular degrees of quantity and quality. So if we have any
idea of power in general we must also be able to conceive
some specific kind of power; and as power can’t exist alone
but is always regarded as an attribute of some existing thing,
we must be able to place this power in some particular thing
and to conceive that thing as having a real force and energy
by which such and such a particular effect necessarily results from its operation. We must •conceive the connection
between the cause and the effect distinctly and in detail,
and •see from a simple view of one of them that it must be
followed or preceded by the other. This is the true manner of
conceiving a particular power in a particular body; . . . . and
it is perfectly obvious that the human mind ·can’t do any
such thing, that is, it· can’t form an idea of two objects that
will enable it to conceive any connection between them, or
comprehend distinctly the power or efficacy by which they
are united. Such a connection would amount to a demonstration, and would imply the absolute impossibility for the one
object not to follow, or to be conceived not to follow on the
other; and that kind of connection has already been rejected
in all cases. If you disagree, and think you have acquired
a notion of power in some particular object, please point
out to me the object. Until someone does that—and nobody
will!—I have to conclude that since we can never distinctly
conceive how any •particular power can possibly reside in
any particular object, we deceive ourselves in imagining we
can form any such •general idea.
From all this we may infer that when we
•talk of any being, whether divine or creaturely, as

iii: Knowledge and probability

having a ‘power’ or ‘force’ that is exactly right for some
effect, or •speak of a ‘necessary connection’ between
objects, and suppose that this connection depends
on an ‘efficacy’ or ‘energy’ that some of these objects
possess,
we really have no clear meaning for any of these expressions, and are merely using common words without any
clear and determinate ideas. Perhaps the expressions never
have meanings; but it is more probable that they do have
proper meanings which they lose in these contexts through
being wrongly used. So let us return to our subject, to see if
we can discover the nature and origin of the ideas that we
attach to the expressions ·when we are using them properly·.
As we confront a particular cause-effect pair, we can’t
just by considering either or both of those objects •perceive
the tie that unites them, or •say for sure that there is a
connection between them. So it is not from any one instance
that we arrive at the idea of cause and effect, of a necessary
connection, of power, of force, of energy, of efficacy. •If all we
ever saw were particular conjunctions of objects, each conjoined pair being entirely different from each of the others,
we could never form any such ideas.
But •when we observe numerous instances in which the
same ·kinds of· objects are conjoined, we immediately conceive a connection between them, and begin to draw an inference from one to another. So this •multiplicity of resembling
instances constitutes the essence of power or connection,
and is the source from which the idea of it arises. To understand the idea of power, then, we must consider this
•multiplicity—and that is all I require for a solution of the
difficulty we have been wrestling with. I reason thus: The
repetition of perfectly similar instances can’t on its own give
rise to an original idea different from what is to be found in
any particular instance; I have pointed this out already, and
82

Treatise, Book 1

David Hume

it obviously follows from my basic principle that all ideas
are copied from impressions. But the idea of power is a new
original idea that isn’t to be found in any one instance, and
yet it arises from the repetition of numerous instances; so it
follows that the repetition doesn’t have that effect on its own,
but must either (1) reveal or (2) produce something new that
is the source of that idea. . . . (1) But the repetition of similar
objects in similar relations of succession and contiguity obviously doesn’t •reveal anything new in any one of them, since
we can’t draw any inference from it or make it a subject of
either demonstrative or probable reasonings (as I proved in
section 6). Indeed, even if we could draw an inference, it
wouldn’t make any difference in the present case. That is
because no kind of reasoning can give rise to a new idea such
as the idea of power is; when we reason we must already
have clear ideas to serve as the objects of our reasoning. The
conception always precedes the understanding; and where
one is obscure the other is uncertain, where one fails the
other must fail also.
(2) It is certain that this repetition of similar objects in
similar situations •produces nothing new in these objects
or in any external body. For you will readily agree that the
different instances we have of the conjunction of resembling
causes and effects are in themselves entirely independent
·of one another·, and that the passing on of motion that I
see result from the present collision of two billiard balls is
totally distinct from what I saw result from such a collision
a year ago. These collisions have no influence on each other:
they are entirely separated by time and place, and one of
them could have existed and communicated motion even if
the other had never occurred. So:
Nothing new is either •revealed or •produced in any
objects by their constant conjunction, and by the uninterrupted resemblance of their relations of succession

iii: Knowledge and probability

and contiguity. Yet it is from this resemblance that
the ideas of necessity, of power, and of efficacy are
derived. So these ideas don’t represent anything that
does or can belong to the objects that are constantly
conjoined.
Look at this argument from any angle you like—you will
find it to be perfectly unanswerable. Similar instances are
the first source of our idea of power or necessity; but their
similarity doesn’t give them any influence on each other or
on any external object. We must therefore look in some other
direction to find the origin of that idea.
Though the numerous resembling instances that give
rise to the idea of power have no influence on each other,
and can never produce in the object any new quality that
could be the model for that idea, our observation of this
resemblance produces a new impression in our mind, and
that is the idea’s real model. For after we have observed
the resemblance in a sufficient number of instances, we immediately feel a determination of the mind to pass from one
object to its usual attendant, and to conceive the latter in
a stronger light on account of that determination. [Feeling a
‘determination’ to form a certain idea is just feeling oneself being made
to form the idea. Most of Hume’s uses of ‘determine’ etc. have been rendered here by ‘make’ etc., but in the present section ‘determination’ is

This determination is the only effect of the
resemblance, and so it must be the power or efficacy the idea
of which is derived from the resemblance. The numerous
instances of resembling conjunctions lead us into the notion
of power and necessity. These instances are in themselves
totally distinct from each other and have no union except
in our mind, which observes them and collects their ideas.
So necessity is the effect of this observation, and is nothing
but an internal impression of the mind—a determination to
carry our thoughts from one object to another. If we don’t
allowed to stand.]

83

Treatise, Book 1

David Hume

view it in this way we can never arrive at the most distant
notion of it, or be able to attribute it either to external or
internal objects, to spirit or body, to causes or effects.
•The necessary connection between causes and effects is
the basis of our inference from one to the other. The basis of
our inference is •the transition ·in our minds· arising from
the accustomed union. These, therefore, are the same: ·the
necessary connection between causes and effects is the move
our mind makes from an impression of the cause to a lively
idea of the effect, or perhaps it is not the move itself but
rather our being made or determined to make the move·.
The idea of necessity arises from some impression. No
impression conveyed by our ·outer· senses can give rise to
it. So it must be derived from some internal impression,
some impression of reflection. The only internal impression
that has anything to do with the present business is ·the
impression of· the propensity that custom produces in us to
pass from an object to the idea of its usual attendant. This,
therefore, is the essence of necessity. The bottom line is this:
necessity is something that exists in the mind, not in objects,
and we can’t ever form the remotest idea of it considered as
a quality in bodies. Either we have no idea of necessity, or
necessity is nothing but the determination of the thought
to pass from causes to effects (and vice versa) according to
their experienced union.
Thus, just as •the necessity that makes twice two equal
four . . . . lies only in •the act of the understanding by which
we consider and compare these ideas, so also •the necessity
or power that unites causes with effects lies in •the determination of the mind to pass from the one to the other. The
efficacy or energy of causes doesn’t belong to the causes
themselves or to God or to the two together; it belongs entirely to the mind that considers the union of two or more
objects in all past instances. It is here that the real power of

iii: Knowledge and probability

causes is placed, along with their connection and necessity.
I am aware that this is the most violent of all the paradoxes that I have advanced or will advance in the course
of this Treatise, and that only through solid proof and reasoning can I hope to get it accepted and to overcome the
ingrained prejudices of mankind. Before people are reconciled to this doctrine, they will have often to repeat to
themselves ·the central line of argument·:
•The simple view of any two objects or actions, however
they are related, can never give us any idea of power
or of a connection between them.
•This idea arises from the repetition of their union.
•The repetition doesn’t reveal anything or cause anything in the objects; its only influence is on the mind,
through the customary transition that it produces.
Therefore:
•this customary transition is the same as the power
and necessity, which are therefore qualities of perceptions rather than of objects, and are internally felt by
the soul rather than perceived externally in bodies.
Any extraordinary claim is usually met with astonishment,
which immediately changes into the highest degree of admiration or contempt, depending on whether we approve or
disapprove of what is said. I am much afraid that although
the above reasoning seems to me the shortest and most
decisive imaginable, the bias of the mind will persist in the
general run of readers, giving them a prejudice against the
present doctrine.
This bias against it is easily accounted for. It is widely
recognized that the mind has a great propensity to spread
itself on external objects: when some objects cause internal
impressions that always occur at the same time that the
objects appear to the senses, the mind conjoins these impressions with the objects. For example, as certain sounds
84

Treatise, Book 1

David Hume

and smells are always found to accompany certain visible
objects, we naturally imagine that the sounds and smells are
in the objects, even being in the same place, though in fact
the qualities are the wrong sorts of thing to be conjoined with
objects, and really don’t exist in any place. I shall return to
this in 5iv . All I need say here is that this propensity ·that
the mind has for spreading itself on external objects· is what
makes us suppose necessity and power to lie in the objects
we consider, not in our mind that considers them. . . .
But although this is the only reasonable account we can
give of necessity, the contrary notion is so riveted in the mind
by the forces I have mentioned that I am sure my views will
be treated by many as extravagant and ridiculous.
What! the efficacy of causes lies in the determination
of the mind? As if causes didn’t operate entirely independently of the mind, and wouldn’t continue their
operation even if no minds existed to think about them
or reason about them! •Thought may well depend on
•causes for its operation, but •causes don’t depend
on •thought. ·To suppose otherwise· is to reverse the
order of Nature and give a secondary role to what
is really primary. To every operation there is an appropriate power, which must belong to the body that
operates. If we remove the power from one cause, we
must ascribe it to another; but to remove it from all
causes and bestow it on a being that relates to the
cause and the effect only by perceiving them is a gross
absurdity and contrary to the most certain principles
of human reason.
All I can say in reply to these arguments is that they are like
a blind man’s claiming to find a great many absurdities in
the supposition that the colour of scarlet is not the same
as the sound of a trumpet, or that light is not the same as
solidity! If we really have no idea of power or efficacy in any

iii: Knowledge and probability

object, or of any real connection between causes and effects,
it won’t do much good to ‘prove’ that efficacy is necessary in
all operations. People who say such things don’t understand
their own meanings, and ignorantly run together ideas that
are entirely distinct from each other. I willingly allow that
both material and immaterial objects may have various qualities of which we know nothing; and if we choose to call these
‘power’ or ‘efficacy’, that won’t matter much to the world. But
when we use the terms ‘power’ and ‘efficacy’ not as •meaning
those unknown qualities, but rather as •signifying something
of which we do have a clear idea, and which is incompatible
with the objects to which we attribute it, obscurity and error
begin to occur and we are led astray by a false philosophy.
That is what happens when we transfer •the determination
of the thought to •external objects and credit them with a
real intelligible connection between them, this being ·an objectivised analogue of· a quality that can belong only to the
observing mind.
As for the point that the operations of Nature are independent of our thought and reasoning, I agree; which is why
I have remarked
•that objects have the relations of contiguity and succession to each other,
•that similar objects can be observed to have similar
relations in many instances, and
•that all this is independent of the operations of the
understanding.
But if we go beyond that and ascribe a power or necessary
connection to these objects, we are ascribing something that
we can never observe in them, and have to derive the idea of
it from what we feel internally when we think about them. I
carry this doctrine so far that I am ready to apply it to ·the
causal claim involved in· my present line of thought. ·I do
that in the following paragraph·.
85

Treatise, Book 1

David Hume

iii: Knowledge and probability

a •philosophical relation (a mere comparison of two ideas),
the other makes us consider it as a •natural relation (an
association between two ideas). [See note on page 8.] We can
define a ‘cause’ to be
An object precedent and contiguous to another, and
where all the objects resembling the former are similarly precedent and contiguous to objects that resemble the latter.
If you find this to be defective because in addition to the
cause and the effect it brings in something extraneous
(·namely, other objects that resemble them·), we can substitute this other definition in its place:
A cause is an object precedent and contiguous to another, and united with it in such a way that the idea
of one determines the mind to form the idea of the
other, and the impression of one to form a livelier idea
of the other.
If you reject this too for the same reason—·because in addition to the cause and the effect it brings something extraneous (namely our impressions and ideas of them)·—I can only
ask you to replace it by a better definition. I have to admit
that I can’t do that. [Hume then goes on to repeat his theory
and his reasons for it, concluding:] However extraordinary
my views ·about cause-effect· may appear, I think it is useless to trouble myself with any further enquiry or reasoning
on the subject, and shall now rely on them as on established
maxims.
Before leaving this subject I shall draw some corollaries
from my theory—ones that will enable us to remove four prejudices and popular errors that have held sway in philosophy.
(1) We can learn from my doctrine that all causes are of
the same kind, and that there is no basis for distinguishing
•making causes from •enabling causes, or for sorting out
causes according to whether they are

When an object is presented to us, it immediately gives
the mind a lively idea of the object that is usually found
to accompany it, and this determination of the mind forms
the necessary connection of these objects. But when we
step back and attend not to •the objects but to •our perceptions of them, we still have a causal claim to consider,
namely that the impression (of one object) is the cause and
the lively idea (of another object) is the effect; and their necessary connection is the new determination that we feel to
pass from the idea of the impression to the idea of the lively
idea. The force that unites our internal perceptions is as
unintelligible—·as incapable of being seen as necessitating,
just by hard thinking·—as is the force that unites external
objects, and is known to us only by experience. Now, I have
already sufficiently examined and explained the nature and
effects of experience: it never gives us any insight into the
internal structure or operating force of objects, but only accustoms the mind to pass from ·an impression of· one to ·a
lively idea of· another.
It is now time to gather up all the parts of this reasoning,
and assemble them into an exact definition of the relation
of cause and effect, which is our present topic. This order
of exposition—first examining our inference from the causeeffect relation and then explaining the relation itself—would
have been inexcusable if it had been possible to proceed in
any other way. But as the nature of •the relation depends so
much on that of •the inference, I have had to advance in this
seemingly preposterous manner, using certain terms before
being able exactly to define them or fix their meaning. I shall
now correct this fault by giving a precise definition of cause
and effect.
There are two definitions we can give for this relation,
which differ only in that they present different views of
the same object; one makes us consider cause-effect as
86

Treatise, Book 1

David Hume

efficient,
formal,
material,
exemplary, or
final.

iii: Knowledge and probability

weakening this conjunction and determination. Even in the
operation of bodies there are different degrees of constancy
·of going-together·, and different degrees of force ·exerted on
the mind in its movement from impression to idea·, without
producing different kinds of causality.
The distinction that is often made between ·having· power
and exercising it is equally baseless.
(3) Perhaps I can now fully overcome all the natural reluctance to accept my earlier arguments in which I tried to prove
that the necessity of a cause to every beginning of existence
has no demonstrative or intuitive support. That conclusion
won’t appear strange in the light of my definitions. If we
define a ‘cause’ to be
An object precedent and contiguous to another, and
where all the objects resembling the former are similarly precedent and contiguous to objects that resemble the latter,
we can easily grasp that there is no absolute or metaphysical necessity that every beginning of existence should be
preceded by such an object. And if we define a ‘cause’ to be
An object precedent and contiguous to another, and
united with it in the imagination in such a way that
the idea of one determines the mind to form the idea of
the other, and the impression of one to form a livelier
idea of the other,
we shall have even less difficulty in assenting to my opinion.
Such an influence on the mind—·so far from being something
we can be sure must go with every beginning of existence—is
in itself perfectly extraordinary and incomprehensible, and it
is only from experience and observation that we are certain
that it ever occurs.
(4) We can never have reason to believe in the existence
of something of which we can’t form an idea. All our reasonings about existence are derived from causation, so they are

[The efficient cause of a coin is the stamping of a die on hot metal, its
formal cause is its roundness etc., its material cause is the metal it is
made of, and its final cause is the commercial end for which the coin
was made. The notion of ‘exemplary cause’, employed by some mediaeval
philosophers wishing to combine Plato with Christianity, can’t be briefly
explained here.] Our idea of efficiency ·or making· is derived
from the constant conjunction of two ·kinds of· objects; when
this is observed the cause is efficient; and where it is not,
there is no cause of any kind. For the same reason we must
deny that there is any essential difference between cause
and occasion. If constant conjunction is implied in what we
call ‘occasion’, it is a real cause. If not, it isn’t a ·natural·
relation at all, and can’t give rise to any argument or reasoning. [Some philosophers, notably Malebranche, held that created things
cannot really act on one another, and that what happens in billiards (for
example) is that God causes the cue-ball to move on the occasion of its
being struck by the cue.]

(2) The same course of reasoning will make us conclude
that just as there is only one kind of cause, so also there is
only one kind of necessity, and that the common distinction
between ‘moral’ and ‘physical’ necessity has no basis. This
account I have given of necessity makes this clear. The constant conjunction of objects, along with the determination
of the mind, constitutes physical necessity; and when these
are absent what you have is chance. As objects must either
be conjoined or not, and as the mind must either be determined or not determined to pass from one object to another,
there can’t be any middle case between chance and absolute
necessity. You don’t change the nature of the necessity by
87

Treatise, Book 1

David Hume

iii: Knowledge and probability

can’t be any mystery in our conclusions—·that is, we can’t
soundly argue for the existence of an I-know-not-what of
which we don’t have an idea·. . . .

derived from the experienced conjunction of objects and not
from any exercise of pure thinking. So the same experience
·that grounds our causal reasoning· must give us a notion
of these objects ·whose existence we reason to·; so there

15: Rules by which to judge of causes and effects
According to my doctrine, there are no objects which we can,
by merely surveying them and without consulting experience,
discover to be the causes of anything else; and no objects
that we can certainly discover in the same manner not to
be the causes ·of specified other things·. Anything can
produce anything. Creation, annihilation, motion, reason,
volition—all these can arise from one another, or from any
other object we can imagine. You won’t find this strange if
you hold in your mind together two principles that I have
explained: •that the constant conjunction of objects determines their causation, and •that strictly speaking no objects
are contrary to each other but existence and non-existence
(see i.5). Where objects are not contrary, nothing hinders
them from having the constant conjunction on which the
relation of cause and effect totally depends.
Since it is thus possible for any object to be a cause or
effect of any other, it may be proper to fix some general rules
by which we can know when the cause-effect really does
obtain. ·I shall offer eight such rules·.
1. The cause and effect must be contiguous in space and
time.
2. The cause must be prior to the effect.
3. There must be a constant union between the cause
and effect. This is what chiefly constitutes the cause-effect
relation.
4. The same cause always produces the same effect, and

the same effect always comes from the same cause. We
derive this principle from experience. And it’s the source
of most of our philosophical reasonings. For when by any
clear experience we have discovered the causes or effects of
any phenomenon, we immediately extend our observation to
every phenomenon of the same kind, without waiting for the
constant repetition from which the idea of the cause-effect
relation was originally derived.
5. (This rule depends on rule 4.) Where several different
objects produce the same effect, it must be by means of
some quality that we find to be common to them all. For as
like effects imply like causes, we must always ascribe the
causation to the respect in which the causes are alike.
6. (Another rule stemming from 4.) The difference in the
effects of two similar objects must come from a respect in
which the objects are not alike. For as like causes always
produce like effects, when in any instance we find that this
seems not to hold we must conclude that this irregularity
proceeds from some ·not-yet-discovered· difference between
the causes.
7. When an object increases or diminishes with the increase or diminution of its cause, it is to be regarded as
a compounded effect, derived from the union of different
effects arising from different parts of the cause. The absence
(or presence) of one part of the cause is here supposed to be
always followed by the absence (or presence) of a correspond88

Treatise, Book 1

David Hume

ing part of the effect. This constant conjunction sufficiently
proves that one part is the cause of the other. But we must
not rashly draw such a conclusion from a few instances. A
certain degree of heat gives pleasure; if you reduce the heat,
the pleasure lessens; but it doesn’t follow that if you raise the
heat beyond a certain degree the pleasure will increase correspondingly; for we find that ·on the contrary· it degenerates
into pain.
8. An object which exists for any time in its full perfection
without any effect is not the sole cause of that effect, but
needs to be assisted by some other force that can forward
its influence and operation. For as like effects necessarily
follow from like causes, and in a contiguous time and place,
their separation for a moment shows that these causes are
not complete ones.
Those eight rules contain all the logic that I think proper
to use in my reasoning; and perhaps even they weren’t much
needed: the logic they contain might have been supplied by
the natural workings of our understanding. Our Aristotelian
intellectuals and logicians don’t exhibit so much superiority
over ordinary folk in their reason and ability that I want to
imitate them by delivering a long system of rules and precepts to direct our judgment in philosophy! All the rules of
this sort are very easy to discover, but extremely difficult to
apply; and even empirical science, which seems the most nat-

iii: Knowledge and probability

ural and simple of all, requires the utmost stretch of human
judgment. Every phenomenon in Nature is compounded and
modified in so many details that in order to arrive at the
decisive point we must carefully separate whatever is superfluous and investigate through new experiments whether
every detail of the first experiment was essential to it. These
new experiments are open to critical examination of the same
kind; so that we need the utmost constancy to persevere in
our enquiry, and the utmost skill to choose the right way
among so many that present themselves. If this is the case
even in •physical science, how much more in •the sciences
of human nature, where there is a much greater complication of details, and where the beliefs and feelings that are
essential to any action of the mind are so unconscious and
obscure that they often escape our strictest attention, and
are not only unaccountable in their causes but not even
known to exist! I greatly fear that the small success I meet
with in my enquiries will make this remark sound like an
apology rather than—·what it really is·—a boast! If anything
can give me confidence that I am proceeding on the right
lines, it will be the widening of my range of empirical data
as much as possible; so it may be proper at this point to
examine the reasoning faculty of non- human animals as
well as that of human creatures.

16: The reason of animals
It is ridiculous to deny an obvious truth, and almost as
ridiculous to take much trouble to defend one; and no truth
appears to me more obvious than that beasts are endowed
with thought and reason as well as men. The evidence for
this is so obvious that it never escapes the most stupid and

ignorant.
We are conscious that we ourselves, in adapting means
to ends, are guided by reason and design, and that we don’t
ignorantly or casually perform the actions that tend to selfpreservation, and to getting pleasure and avoiding pain. So
89

Treatise, Book 1

David Hume

when we see other creatures in millions of instances perform
•similar actions directed to •similar ends, all our principles
of reason and probability carry us with an irresistible force
to believe in the existence of a •similar cause. I don’t think
I need to illustrate this argument with particular examples;
the smallest attention ·to the non-human part of the animal kingdom· will supply us with more than enough. The
resemblance between the actions of animals and those of
men is so complete in this respect that the first action of
the first animal we happen to choose will provide us with
incontestable evidence for the present doctrine.
This doctrine is as useful as it is obvious, and furnishes
us with a kind of touchstone by which to test every theory
in this area of philosophy. The resemblance of the •external
actions of animals to our own actions leads us to judge that
their •internal actions also resemble ours; and that same line
of reasoning, carried one step further, will make us conclude
that since their internal actions resemble ours, the causes
must also be alike. So when any hypothesis is advanced
to explain a mental operation that is common to men and
beasts, we must apply the same hypothesis to both; and just
as every true hypothesis will survive this test, I venture to
say that no false one will do so. In the systems that philosophers have employed to account for the actions of the mind,
the common defect has been that they presuppose so much
subtlety and refinement of thought that the thought they
describe is out of reach not only of mere animals but even of
children and common people in our own species, although
they are capable of the same emotions and affections as
people of the most accomplished genius and understanding.
Such •subtle complexity is a clear proof of the falsehood ·of
a theory of mind·, just as •simplicity is proof of its truth.
Let us, therefore, put our present system about the nature of the understanding to this decisive trial, and see

iii: Knowledge and probability

whether it will equally account for the reasonings of beasts
as for these of the human species.
I need to distinguish •the actions of animals that are of a
down-to-earth kind and seem to be on a level with their common capacities from •those more extraordinary instances
of wisdom that they sometimes display in the interests of
their own preservation and the propagation of their species.
A dog that avoids fire and precipices, that shuns strangers
and caresses his master, gives us an instance of the •first
kind. A bird that chooses with such care and precision the
place and materials of her nest, and sits on her eggs for an
appropriate time in a suitable season, . . . . provides us with
a lively instance of the •second.
As to actions of the former kind, I assert that they come
from a reasoning that is not different—in itself or in the
forces behind it—from what appears in human nature. It is
necessary in the first place that there be some impression
immediately present to their memory or senses, to be the
basis for their judgment. From the tone of voice the dog
infers his master’s anger and foresees his own punishment.
From a certain sensation affecting his smell he judges that
his prey is not far away.
The inference he draws from the present impression is
built on experience, and on his observation of the conjunction of objects in past instances. As you vary this experience,
he varies his reasoning. Make a beating follow on one sign
or motion for some time, and afterwards on another; and he
will successively draw different conclusions in line with his
most recent experience.
Now, let any philosopher try to explain the act of the
mind we call ‘belief’, giving an account of its causes without
bringing in the influence of custom on the imagination, and
let his hypothesis be equally applicable to beasts as to the
human species; when he has done this, I promise to accept
90

Treatise, Book 1

David Hume

the result! But at the same time I demand that if my system
is the only one that can do this, it should in fairness be
accepted as entirely satisfactory and convincing. That it is
the only one is evident almost without any reasoning.
•Beasts certainly never perceive any real connection
among objects. So
•it is by experience that they infer one from another.
•They can’t by any argument reach the general conclusion that objects of which they have had no experience
resemble those of which they have. So
•it is through custom alone that experience operates
on them.
All this was obvious enough with respect to man. When
applied to beasts there can’t be the least suspicion of mistake; which must be admitted to be a strong confirmation,
or rather an invincible proof, of my system.
The force of habit in reconciling us to a phenomenon

iii: Knowledge and probability

shows nowhere more strikingly than in this: men are not
astonished at the operations of their own reason, yet they
wonder at the instinct of animals, and find it hard to explain
because it can’t be traced back to the very same sources as
their own reason·. To consider the matter rightly, reason
·itself· is nothing but a wonderful and unintelligible instinct
in our souls, which carries us along a certain sequence of
ideas and endows them with particular qualities according
to their particular situations and relations. This instinct,
admittedly, arises from past observation and experience; but
can anyone give the ultimate reason why •past experience
and observation produce such an effect, any more than why
•Nature alone should produce it? Nature can certainly produce ·without help from habit· anything that can arise from
habit; indeed, habit is merely one of the forces of Nature,
getting all its power from Nature.

91

Treatise, Book 1

David Hume

iv: Scepticism and other systems

Part iv: The sceptical and other systems of philosophy
1: Scepticism with regard to reason
In all demonstrative sciences the •rules are certain and infallible; but when we •apply them, our fallible and uncertain
faculties are very apt to depart from them and fall into error.
So we must in every reasoning form a new judgment, as a
check or control on our first judgment or belief; and ·as a
basis for the new judgment· we must enlarge our view to take
in a kind of history of all the cases where our understanding
has deceived us, compared those with the ones where its testimony was sound and true. Our reason must be considered
as a kind of cause of which truth is the natural effect; but
a cause that can often be prevented from having its natural
effect by the intrusion of other causes and by the inconstancy of our mental powers. In this way, all •knowledge
degenerates into •probability; and this probability is greater
or less depending on our experience of the truthfulness or
deceitfulness of our understanding, and on how simple or
complex the question is.

for their security. Rather, they structure their accounts
in a manner that gives their results a greater probability
than what is derived from the skill and experience of the
accountant. For it is clear that skill and experience do yield
some probability ·of accuracy·, though what the probability
is varies according to how experienced the accountant is and
how long his account is. Now, nobody will maintain that the
result of a long calculation can be more than probable. Yet
it is safe to say that there is hardly any proposition about
numbers of which we can be more sure; for it is easy to
break the longest series of additions down into steps in each
of which one number less than 10 is added to another—the
simplest operation that can be done with numbers. So we
shall find it impracticable to show the precise limits of knowledge and of probability, or discover the particular number of
steps at which knowledge stops and probability begins. But
knowledge and probability can’t shade into each other: they
are of contrary and disagreeing natures, and they can’t be
split up—each of them must be either entirely present, or
entirely absent. Furthermore, if •any single addition were
certain ·and a case of knowledge·, •every one would be so,
and consequently the total sum would be certain—unless the
whole can be different from all its parts. I had almost said
‘This is certain’, but I reflect that what I am saying applies
to itself as well as to every other reasoning, and thus must
therefore slide from knowledge down into probability.

No algebraist or mathematician is so expert in his science
that he places complete confidence in any truth immediately
on discovering it, or regards it ·initially· as more than merely
probable. Every time he runs over his proofs, his confidence
increases; but still more by the approval of his friends; and
it is brought to full perfection by the universal assent and
applause of the learned world. And this gradual increase in
confidence is obviously nothing but the addition of new probabilities, and is derived from the constant union of causes
and effects according to past experience and observation.

So all knowledge resolves itself into probability, and eventually comes to be of the same nature as the kind of assurance that we have in common life. Let us, then, examine

In financial accounts of any length or importance, merchants seldom rely on the infallible certainty of numbers
92

Treatise, Book 1

David Hume

our common-life sort of reasoning, to see what foundation it
stands on.
In every judgment that we can form about probability, as
well as about knowledge, we ought always to correct the •first
judgment derived from the nature of the object by a •second
judgment derived from the nature of the understanding. A
man of solid sense and long experience certainly should and
usually does have more confidence in his opinions than a
man who is foolish and ignorant. . . . But even in someone
with the best sense and longest experience this confidence is
never complete, because such a person must be conscious
of many errors in the past, and must still fear making more.
So now there arises a new sort of probability to correct and
regulate the first, assigning to it its proper level of confidence.
Just as demonstration is subject to the control of probability, so also ·this· probability admits of further adjustment
through an act of the mind in which we reflect on the nature
of our understanding and on the reasoning that took us to
the first probability.
Now we have found in every probability •the original uncertainty inherent in the subject and also •a second uncertainty derived from the weakness of our judgment ·in arriving
at the first probability·. When we have put two together ·to
get a single over-all probability·, we are obliged by our reason
to add •a third doubt derived from the possibility of error ·at
the second stage· where we estimated the reliability of our
faculties. This third doubt is one that immediately occurs
to us, and if we want to track our reason closely we can’t
get out of reaching a conclusion about it. But even if this
conclusion is favourable to our second judgment, it is itself
based only on probability and must weaken still further our
first level of confidence. And it must itself be weakened by
a •fourth doubt of the same kind, and so on ad infinitum;
till at last nothing remains of the first probability, however

iv: Scepticism and other systems

great we may have supposed it to be, and however small
the lessening of it by every new uncertainty. Nothing that
is finite can survive an infinity of repeated decreases; and
even the vastest quantity that we can imagine must in this
manner be reduced to nothing. However strong our first belief is, it is bound to perish when it passes through so many
new examinations, each of which somewhat lessens its force
and vigour. When I reflect on the natural fallibility of my
judgment, I have less confidence in my opinions than when I
consider only the topic that I am reasoning about; and when
I go still further and scrutinize every successive estimation
that I make of my faculties, all the rules of logic require a
continual lessening and eventually a total extinction of belief
and evidentness.
‘Do you sincerely assent to this argument that you seem
to take such trouble to persuade us of? Are you really one
of those sceptics who hold that everything is uncertain, and
that our judgment doesn’t have measures of truth and falsehood on any topic?’ I reply that this question is entirely
superfluous, and that neither I nor anyone else was ever
sincerely and constantly of that ·sceptical· opinion. Nature,
by an absolute and uncontrollable necessity, makes us judge
as well as breathe and feel; and we can’t prevent ourselves
from •viewing certain objects in a stronger and fuller light
on account of their customary connection with a present
impression, any more than we can prevent ourselves from
•thinking as long as we are awake, or from •seeing nearby
bodies when we turn our eyes towards them in broad sunlight. Whoever has taken trouble to refute the objections
of this total scepticism has really been disputing without
an antagonist, trying to establish •by arguments a faculty
that •Nature has already implanted in the mind and made
unavoidable.
Then why did I display so carefully the arguments of that
93

Treatise, Book 1

David Hume

fantastic sect (·the total sceptics·)? It was to make you aware
of the truth of my hypotheses that •all our reasonings about
causes and effects are derived from nothing but custom,
and that •belief is strictly an act of the feeling part of our
natures rather than of the thinking part. ·I now proceed to
connect the second of these hypotheses with what I have
shown about reasoning and probability·.
Concerning the elements in our make-up that make us
reach a conclusion on any subject, and correct that conclusion in the light of thoughts about our intellectual limits and about the situation of our mind when we reached
the conclusion, I have proved that they—these very same
elements—when carried further and applied to every new
judgment on ourselves, must by continually lessening our
original confidence eventually reduce it to nothing, utterly
subverting all belief and opinion. So if belief were a simple
act of thought, not involving any special manner of conception such as conceiving in a forceful and lively way, it would
be bound to destroy itself and in every case terminate in a
total suspense of judgment. But experience will sufficiently
convince you (if you think it worthwhile to try this) that although you can’t find anything wrong with my arguments
you still continue to believe, think, and reason as usual; so
you can safely conclude that your reasoning and belief is
some sensation or peculiar manner of conception that can’t
be destroyed by mere ideas and reflections.
But here a further question may be raised:
Even on your hypothesis ·about what belief is·, how
does it happen that your arguments early in this section don’t produce a total suspension of judgment?
How does the mind ever retain any degree of assurance on any subject? These new probabilities whose
repetition perpetually lessens the original confidence
are based on the very same principles as the first judg-

iv: Scepticism and other systems

ment in the series, and it makes no difference whether
the principles have to do with thought (·which you
deny·) or with sensation (·which you assert·). Either
way, it seems unavoidable that they must subvert belief, through the opposition either of contrary thoughts
or of contrary sensations, reducing the mind to a total
uncertainty. Some question is proposed to me, and
after going over the impressions of my memory and
senses, and carrying my thoughts from them to objects of the kinds commonly conjoined with them, I
feel a stronger and more forcible conception on one
side ·of the question· than on the other. This strong
conception (·according to you·) constitutes my first
conclusion ·or belief·. Next, I examine my judgment
itself and, observing from experience that it is sometimes sound and sometimes erroneous, I see it as
governed by contrary forces or causes, of which some
lead to truth and others to error; and in balancing
these contrary causes I arrive at a new probability
which lessens the assurance of my first conclusion.
This new probability is open to being lessened in the
same way as the previous one was, and so on, ad
infinitum. So how does it happen that even after all
that we retain a degree of belief that is sufficient for
our purpose in philosophy or in common life?
I answer that after the first and second conclusions the action of the mind becomes forced and unnatural, and the
ideas become faint and obscure. The •principlesc of judgment and the balancing of opposite causes is the same as at
the very beginning, but their •influence on the imagination
and the difference they make to the vigour of the thought
is by no means the same. When the mind doesn’t grasp its
objects with easy smoothness, the same sources of activity
don’t have the same effect as they do in a more natural
94

Treatise, Book 1

David Hume

conception of the ideas; and the imagination doesn’t feel a
sensation anything like the one that comes from its everyday
judgments and opinions. The attention is on the stretch;
the posture of the mind is uncomfortable, and the ·animal·
spirits, being diverted from their natural course, are not
governed in their movements by the same laws as when they
flow in their usual channel—or at any rate are not governed
by them to the same degree.

iv: Scepticism and other systems

can never be complete when it is not founded on something
natural and easy.
I take this to be the true state of the question, and cannot
approve of the way in which some people try dispose of the
sceptics by rejecting all their arguments at once, without
enquiry or examination. They argue like this:
If the sceptical reasonings are strong, that is a proof
that reason can have some force and authority; if they
are weak, they can never be sufficient to invalidate all
the conclusions of our understanding.
This argument is not sound, and here is why. If the sceptical reasonings could exist and not be destroyed by their
own subtlety, they would be successively strong and weak,
according to the successive dispositions of the mind. Reason first appears in possession of the throne, prescribing
laws with absolute authority. So her enemy (·the sceptical
argument·) has to take shelter under her protection and by
using rational arguments to prove reason’s incompetence
and liability to error, her enemy produces a sort of warrant
of authenticity signed and sealed by reason. This warrant
initially has some authority in proportion to the present and
immediate authority of reason from which it is derived. But
as it is supposed to be contradictory to reason, it gradually
lessens •the force of that governing power and •its own force
at the same time; till at last they both vanish away into
nothing through regular and proper lessenings. ·Here is
how·. The sceptical and anti-sceptical reasons are of the
same kind, though working in contrary directions, so that
when the anti- sceptical case is strong it has to reckon with
an enemy of equal force in the sceptical case; and as they
started out with equal force, they continue like that for as
long as either of them exists; and neither loses any force in
the contest without taking as much from its opponent. So
it is fortunate that Nature eventually breaks the force of all

It isn’t difficult to provide other examples of the same phenomenon; the present subject of metaphysics supplies them
in abundance. An argument that would have been found
convincing in a reasoning about history or politics has little
or no influence in abstruser subjects such as metaphysics,
even when it is perfectly understood; and that is because
understanding it requires a study and an effort of thought,
which disturbs the operation of our sentiments on which
the belief depends. The case is the same in other subjects.
The straining of the imagination always hinders the regular
flowing of the passions and sentiments. A tragic poet who
represented his heroes as talking cleverly and inventively in
their misfortunes would never touch the passions. Just as
the •emotions of the soul prevent any •subtle reasoning and
reflection, so •reflective thinking tends to quell •emotions.
The mind, as well as the body, seems to be endowed with a
certain definite amount of force and activity which it employs
in one action only at the expense of all the rest. This is more
evidently true where the actions are of quite different kinds;
for then the force of the mind is not only redirected but its
disposition is changed, making us incapable of a sudden
switch from one action to the other, let alone of performing
both at once. No wonder, then, that the belief arising from
a subtle reasoning lessens in proportion to the efforts that
the imagination makes to enter into the reasoning and to
conceive it in all its parts. Belief, being a lively conception,
95

Treatise, Book 1

David Hume

sceptical arguments, keeping them from having much influence on the understanding. If we put all our trust in their
destroying themselves, ·as alleged in the above argument·,

iv: Scepticism and other systems

we would be relying on something that can never take place
until they (·the sceptical arguments·) have first subverted all
belief and totally destroyed human reason.

2: Scepticism with regard to the senses
Thus the sceptic still continues to •reason and believe, even
though he asserts that he can’t defend his reason by reason; and by the same rule he must •assent to the principle
concerning the existence of body, though he can’t claim to
maintain its truth by any arguments of philosophy. Nature
hasn’t left this to his choice, and has doubtless thought
it too important to be trusted to our uncertain reasonings
and speculations. We may well ask ‘What causes induce us
to believe in the existence of body?’ but it is pointless to
ask ‘Is there body or not?’, because that is something we
must—·being compelled by Nature·—take for granted in all
our reasonings.
So the subject of our present enquiry is the causes that
induce us to believe in the existence of body. I start with a
distinction that at first sight may seem superfluous, but
which will contribute greatly to the perfect understanding of
what follows. Two questions that are commonly run together
ought to be examined separately. They are:
•Why do we attribute a continued existence to objects
even when they aren’t present to the senses? and
•Why do we suppose objects to have an existence distinct from the mind and perception?
In the second question, I ·am using ‘distinct from’ to· refer to
object’s spatial position as well as its ·causal· relations—•its
external position as well as •the independence of its existence
and operation. These two questions, about the continued
and distinct existence of body, are intimately connected. For

if the objects of our senses •continue to exist even when they
are not perceived, their existence is of course independent
of the perception and ·in that sense· •distinct from it; and
conversely, if their existence is independent of the perception
and ·in that sense· •distinct from it, they must •continue
to exist even when they are not perceived. But though a
decision on either of the questions also decides the other as
well, it will be easier for us to discover the sources in human
nature from which the decision arises if we treat •continuity
separately from •distinctness. So I shall inquire whether the
opinion that bodies have a •continued existence is produced
by the senses, by reason, or by the imagination, and shall
inquire into the analogous question regarding the opinion
that bodies exist •distinct from the mind. These are the only
questions that are intelligible on the present subject. As for
the notion of external existence, when understood to mean
that bodies exist and are of a categorially different sort from
our perceptions, I have already shown its absurdity in 6ii .

·T HE

SENSES ·

Obviously the senses can’t give rise to the view that objects •continue to exist after they have stopped appearing
to the senses. For them to do that would be for them to
continue to operate even after they have entirely stopped
operating, which is a contradiction in terms. So if the senses
have any influence in the present case, it must be in producing the opinion that bodies have a •distinct (not a continued)
existence. If they were to do that, it would have to be either
96

Treatise, Book 1

David Hume

iv: Scepticism and other systems

of the uniting principlec that ·makes a •number of items·
constitute •one person. So far from being able to answer
it merely through our senses, we must—·and in section 6
I shall·—have recourse to the most profound metaphysics
to give a satisfactory answer to it; and in common life it is
obvious that these ideas of self and person are never very
fixed or determinate. So it is absurd to suggest that the
senses can ever distinguish ourselves from external objects.
And a further point: All impressions (external and internal), passions, affections, sensations, pains, and pleasures
are originally on the same footing; and whatever differences
we may observe among them, they all appear in their true
colours as impressions or perceptions ·and not as objects
distinct from ourselves·. Indeed, it is hardly possible that
it should be otherwise: it isn’t conceivable that our senses
should be able to deceive us about the •situation and relations of our impressions, any more than about their •nature.
For since all the actions and sensations of the mind are
known to us by consciousness, they must in every detail
appear to be what they are, and be what they appear. It is
impossible that something that enters the mind as really a
perception should appear to be something different. If that
could happen, it would mean that we might be mistaken
even about what we are most intimately conscious of.
Rather than spending more time examining whether our
senses possibly could deceive us by representing our perceptions as distinct from ourselves (that is, as •external to and
•independent of us), let us consider whether they really do
so. . . . Here is an argument that might be used:
My own body evidently belongs to me, and as various impressions appear exterior to my body I suppose
them to be exterior to me. (Let’s set aside the metaphysical question about the identity of a thinking
substance, ·which may be tied up with the question of

by presenting their impressions as •images [= ‘likenesses’] and
representations ·of bodies existing distinct from the mind· or
by presenting their impressions as •themselves being these
distinct and external existences. ·Let us look at these separately·.
It is obvious that our senses don’t offer their impressions
as the images of something distinct (i.e. independent and
external), because all they convey to us is a single perception,
with not the slightest hint of anything beyond it. A single perception can’t produce the idea of two existing things except
through some inference of either reason or imagination (·and
I shall come to them later·). When the mind looks further
than what immediately appears to it, its conclusions can
never be attributed to the senses; and it certainly does look
further when from a single perception it infers two existing
things and supposes relations of resemblance and causation
between them.
So if our senses suggest any idea of distinct existences,
they must do it by presenting their impressions as being
those very existences, this being a kind of fallacy and illusion.
In this connection I point out that all sensations are felt by
the mind as what they really are; when we wonder whether
they present themselves as distinct objects or only as impressions, we aren’t asking about their nature but about their
•relations and •situation—·specifically, about whether they
are •related to us by causation or resemblance, and whether
they are •located somewhere other than where we are·. Now,
if the senses presented our impressions as being objects that
are external to and independent of ourselves, they must be
able to relate the objects to ourselves, which means that we
ourselves must appear to our senses. So that is the question
we now have to face: how far are we ourselves the objects
of our senses? No question in philosophy is more abstruse
than the one about ·personal· identity—about the nature
97

Treatise, Book 1

David Hume

what I am·.) The paper on which I am now writing is
beyond my hand. The table is beyond the paper. The
walls of the room beyond the table. And in looking
towards the window I see a great stretch of fields and
buildings beyond my room. From all this it can be
inferred that all I need are my senses, with no help
from any other faculty, to be convinced of the external
existence of body.
This inference is blocked by the three following considerations. (1) Properly speaking, when we look at our limbs and
other body-parts what we perceive isn’t •our body but rather
•certain impressions that come to us through the senses;
so when we treat these impressions as being (or as being
impressions of ) real bodies, that is an act of the mind that’s
as hard to explain as the one we are now examining. (2)
Sounds, tastes, and smells, though commonly regarded by
the mind as •continued •independent qualities, don’t appear
to have any existence in the extended realm, so that they
can’t appear to the senses as situated outside the body. The
reason why we ascribe a place to them will be considered in
section 5. (3) Even our sight doesn’t inform us of distance or
outerness immediately and without a certain reasoning and
experience, as is agreed by the most rational philosophers
·under the lead of Berkeley·.
As to the •independence of our perceptions from ourselves, this can never be given to us by the senses; any
opinion we form about it must be derived from experience
and observation; and we’ll see later that our conclusions
from experience are far from being favourable to the doctrine
of the independence of our perceptions. Anyway, I would
point out that when we talk of real ‘distinct’ existents, we
are usually thinking more of their •independence than of
their •external position; we think an object has sufficient reality if its existence is uninterrupted, and independent of the

iv: Scepticism and other systems

incessant revolutions that we are conscious of in ourselves.
Summing up what I have said about the senses: They
give us no notion of •continued existence because they can’t
operate beyond the limits within which they really operate.
No more do they produce the opinion of a •distinct existence,
because they can’t offer that to the mind as represented or
as original. To offer it as represented, they must present
both an object and an image. To make it appear as original,
they would have to convey a falsehood, . . . . but in fact they
don’t and can’t deceive us. So we can conclude with certainty
that the senses don’t give rise to the opinion of a •continued
existence or of a •distinct one.
I shall confirm this ·with an argument that will run to
the end of the next paragraph·. Three different kinds of
impressions are conveyed by the senses:
•those of bodies’ shape, size, motion, and solidity,
•those of colours, tastes, smells, sounds, heat, and

cold; and
•pains and pleasures that arise from the application

of objects to our bodies, for example by the cutting of
our flesh with steel.
Both philosophers and ordinary folk suppose the first of
these to have a distinct continued existence. Only common
people regard the second in that way. Both philosophers
and common folk, again, regard the third as merely perceptions and thus as being interrupted and dependent in their
existence.
Now, whatever our philosophical opinion may be, it is obvious that so far as the senses can tell colours, sounds, heat,
and cold exist in the same way as do motion and solidity;
and that the mere perception ·of them· isn’t what makes us
distinguish them in this respect, ·by attributing independent
existence to the latter group and not the former·.
98

Treatise, Book 1

David Hume

·On the contrary, many people think their senses tell
them that colours etc. do have an independent existence·.
The prejudice in favour of assigning a distinct continued
existence to colours etc. is so strong that when the contrary
opinion is advanced by modern philosophers, people think
they can almost refute it by appealing only to their feeling
and experience; their very senses, they think, contradict this
philosophy! It is also obvious that colours etc. are originally on the same footing as the pain that arises from steel
and pleasure that comes from a fire, and that the difference
between them is based not on perception or reason but on
the imagination. Both lots—·colour etc. and pain etc.·—are
agreed to be nothing but perceptions arising from the particular configurations and motions of the parts of body, so
how could they possibly differ? Taking all this into account,
we can conclude that, as far as the senses are judges, all
perceptions are the same in their manner of existence.

iv: Scepticism and other systems

the very things they feel or see (·perceptions·). This opinion
is entirely unreasonable, therefore, and so it must come from
some faculty other than the understanding, ·i.e. other than
reason ·. To which I would add this: As long as we take our
perceptions and objects to be the same, we can’t infer the
existence of the objects from the existence of the perceptions,
or form any argument from the relation of cause and effect,
which is the only one that can assure us of any matter of fact.
And even after we distinguish perceptions from objects, it
will soon appear that we still can’t reason from the existence
of one to the existence of the other. All this shows that our
reason doesn’t and couldn’t possibly, on any supposition,
give us an assurance of the continued and distinct existence
of body. That opinion must be entirely owing to the imagination, which must now be the subject of our enquiry. ·The
discussion of the imagination’s role in producing the belief in
continued bodies that are distinct from us will occupy more
than half of the length of this section·.

·R EASON·

·I MAGINATION :

Notice that when people attribute a distinct continued existence to sounds and colours, they do this without ever consulting reason or testing our opinions by any philosophical
principles. Indeed, whatever convincing arguments philosophers may think they can produce to establish the belief in
objects that are independent of the mind, these arguments
are known to only a very few; it is not by them that children,
peasants, and most of mankind are induced to attribute
·independent· objects to some impressions and deny them to
others. Thus, we find that all the conclusions that common
people arrive at about this are directly contrary to those that
are confirmed by philosophy! For philosophy informs us
that everything that appears to the mind is nothing but a
perception, and is interrupted and dependent on the mind;
whereas common people confuse •perceptions with •objects,
and attribute a •distinct continued existence (·objects·) to

FIRST ATTEMPTS ·

Since all impressions are internal and perishing things,
and appear as such, •the notion of their distinct and continued existence ·can’t arise from them alone; so it· must
arise from some of their qualities aided by qualities of the
imagination; and since •this notion doesn’t extend to all of
them, it must arise from qualities that only some impressions possess. So we can easily discover what these qualities
are by comparing the impressions to which we attribute a
distinct and continued existence with those that we regard
as internal and perishing.
It has commonly been supposed that we attribute a reality and continued existence to some impressions because
they are involuntary (·as I look up from this table with my
eyes open I can’t help seeing the window, whereas with my
eyes closed I can choose whether to imagine the window·);
99

Treatise, Book 1

David Hume

and another suggestion is that we attribute a reality and
continued existence to some perceptions because they have
greater force and violence than the others (·my perception
when I see the window is more forceful than the one I have
when I imagine the window·). These are both wrong. It is
obvious that some impressions that we never suppose to
have any existence beyond our perception are just as involuntary as, and are more violent than, the impressions of
shape and extension, colour and sound that we suppose to
be permanent beings; for example our pains and pleasures,
our passions and affections. . . .
Having rejected these common opinions, we must search
for some other theory revealing the special qualities in some
impressions that makes us attribute to them a distinct and
continued existence.
After a little examination we shall find that all the objects
to which we attribute a continued existence have a peculiar
constancy that distinguishes them from the impressions
·that we don’t regard as existing continuously, through gaps
in our perception, because we think that their· existence
depends on our perception. The mountains and houses and
trees that I see at this moment have always appeared to me
in the same order, and when I lose sight of them by shutting
my eyes or turning my head I soon after find them return to
me without the least alteration. My bed and table, my books
and papers, present themselves in the same uniform manner,
and don’t change because of interruptions in my seeing or
perceiving them. This is the case with all the impressions
whose objects are supposed to have an external existence,
and it doesn’t hold for any other impressions, whether gentle
or violent, voluntary or involuntary.
But this constancy is not perfect, and admits of considerable exceptions: bodies often change their position and
qualities, and after a little absence or interruption they may

iv: Scepticism and other systems

be hardly knowable. But we can see that even in these
changes they preserve a •coherence, and have a regular
•dependence on each other, which is the basis for a kind
of reasoning from causation that produces the opinion of
their continued existence. When I return to my room after
an hour’s absence, I don’t find my fire in the same state as
when I left it; but then in other cases I have been accustomed
to seeing a similar alteration produced in a similar period
of time, whether I am present or absent. (·Similar initial
states of the fire have regularly been followed by similar subsequent states; this makes me think that the former cause
the latter; and that requires that the fire stayed in existence
throughout. This is the ‘kind of reasoning from causation’
to which I referred·.) So this coherence in their changes is
one of the characteristics of external objects, as well as their
constancy.
Having found that the belief in the continued existence
of body depends on the coherence and constancy of certain
impressions, I now ask how these qualities give rise to this
extraordinary opinion. To begin with coherence: although
the internal impressions that we regard as fleeting and perishing also have a certain coherence or regularity in their
appearances, it is of a somewhat different kind from what we
find in bodies. We find by experience that our •passions have
a mutual connection with and dependence on each other;
but we never find ourselves having to suppose that they have
existed and operated when they were not perceived, in order
to preserve the same dependence and connection of which
we have had experience. It is not like that with •external
objects. They require a continued existence if they are not to
lose much of the regularity of their operation. I am sitting
here in my room with my face to the fire, and all the objects
that strike my senses are within a few yards of me. (It is
true that my memory informs me of the existence of many
100

Treatise, Book 1

David Hume

·other· objects; but what it tells me is only about their past

iv: Scepticism and other systems

convincing through being the only one that lets me reconcile
the contradiction. At almost every moment of my life there is
a similar instance presented to me, leading me to suppose
the continued existence of objects in order to connect their
past appearances with their present ones, giving them such
a union with each other as I have found by experience to
be suitable to their particular natures and circumstances.
Thus I am naturally led to regard the world as something
real and durable, and as preserving its existence even when
I don’t perceive it.
•This inference from the coherence of appearances may
seem to be of the same nature with •our reasonings about
causes and effects, because both are derived from custom
and regulated by past experience. But we shall find that they
are ultimately quite different from one another, and that our
present inference arises from the understanding and from
custom ·not in the direct way that causal reasoning does,
but· in an indirect and oblique manner. You will agree that
since nothing is ever really present to the mind except its
own perceptions,
it is impossible that •any habit should ever be acquired other than through the regular succession of
these perceptions, and impossible that •any habit
should ever exceed that degree of regularity.
So a certain degree of regularity in our perceptions can’t be
a basis for us to infer a greater degree of regularity in some
objects that are not perceived. To suppose that it could is to
suppose a contradiction—namely, a habit acquired by something that was never present to the mind. But when we infer
the continued existence of the objects of sense from their
coherence and the frequency of their union, we obviously do
this so as to give them a greater regularity than has been
observed in our mere perceptions. ·To make this clearer, I
shall redescribe the situation in slightly different terms·. We

existence, and neither it nor my senses tell me that those
things have continued in existence until now.) So here I am,
turning over these thoughts, when suddenly I hear a noise
as of a door turning on its hinges, and a moment later I see
a porter coming towards me. This gives rise to many new
reflections and reasonings ·in which three things predominate·. •I have never observed that this ·kind of· noise could
come from anything but the motion of a door; so I conclude
that the present phenomenon is a contradiction to all past
experience unless the door that I remember on the other
side of the room still exists. •I have always found that human bodies have a quality that I call ‘gravity’ which prevents
them from floating in the air, which is what this porter must
have done to arrive at my chamber unless the stairs that I
remember have survived my absence ·from them·. •I receive
a letter which, when I open it, I see by the handwriting and
signature to have come from a friend, and in it he says he
is six hundred miles away. Obviously I can’t account for
this phenomenon, consistently with my experience in other
instances, without spreading out in my mind the whole sea
and continent between us, and supposing the effects and
continued existence of coaches and ferries, according to my
memory and observation. Looked at in a certain way, these
phenomena of the porter and letter are contradictions to
common experience, and may be regarded as objections to
the maxims we form about the connections of causes and
effects. I am accustomed to hearing a certain sound and
at the same time seeing a certain object in motion. On this
occasion I have received one of these impressions without
the other. These observations are contrary unless I suppose
that the door still exists and that it was opened without my
perceiving it; and this supposition, which at first was entirely arbitrary and hypothetical, becomes more strong and
101

Treatise, Book 1

David Hume

notice a connection between two kinds of objects in their
past appearance to the senses, but we don’t see this connection to be perfectly constant, because we can break it by
turning our head or shutting our eyes. So what we suppose
in this case is that these objects still continue their usual
connection, despite their apparent interruption, and that the
irregular appearances ·of them· are joined by something that
we don’t perceive. But as all reasoning about matters of fact
arises purely from custom, and custom can only be the effect
of repeated perceptions, extending custom and reasoning
beyond the perceptions can never be the direct and natural
effect of the constant repetition and connection. It must,
therefore, arise from the cooperation of some other forces.
I have already observed in examining the foundation of
mathematics (in 4ii ) that when the imagination embarks on
any line of thinking it is apt to continue even when its object
fails it; like a galley put in motion by the oars, it carries
on its course without any new impulse. I gave this as the
reason why, after considering several rough standards of
equality and correcting them by each other, we proceed to
imagine a standard of equality that is so correct and exact that it can’t admit of the least error or variation. The
same tendency makes us easily entertain this opinion of the
continued existence of body:
Objects have a certain •coherence even as they appear
to our senses; but this coherence is much greater
and more uniform if we suppose the objects to have a
continued existence; and once the mind is engaged in
observing a uniformity among objects, it naturally continues this until it renders the uniformity as complete
as possible. The simple supposition of their continued
existence suffices for this purpose, and gives us a
notion of a much greater regularity ·or •coherence·
among objects than they have when we look no further

iv: Scepticism and other systems

than our senses.
But whatever force we may ascribe to this tendency, I am
afraid it is too weak to support unaided such a vast edifice as
the continued existence of all external bodies. To give a satisfactory account of that opinion, I think, we must bring in
not only the •coherence of objects but also their •constancy.
There is an inference from the constancy of our perceptions
which, like the preceding one from their coherence, gives rise
to the opinion of the continued existence of body. (·Notice
that I am still focussing on objects’ continued existence·; the
belief in that is prior to, and a cause of, the belief in their
distinct existence.) Explaining this will lead me into a considerable range of very profound reasoning, and I want to avoid
confusion; so I think it worthwhile to give a short sketch or
abridged version of my system before proceeding to lay out
its parts in detail.

·I MAGINATION :

SKETCH OF THE SYSTEM ·

When we have been accustomed to observe a constancy
in certain impressions, and have found that the perception
of the sun or ocean (for instance) returns to us after an
absence or annihilation with similar parts and in a similar
order to its first appearance, we aren’t apt to regard these
interrupted perceptions as different, which they really are;
on the contrary, we consider them as individually the same—
·thinking that my present impression that I now have is
the very one, the same individual impression, that I had an
hour ago·—on account of their resemblance. But ·we are
pulled also in the opposite direction·: the interruption of
the existence of the impressions is contrary to their perfect
·individual· identity, and makes us think that the first impression was annihilated and a second one created later;
so we find ourselves somewhat at a loss, and are involved
in a kind of contradiction. In order to free ourselves from
this difficulty, we disguise the interruption as much as we
102

Treatise, Book 1

David Hume

can, or rather we abolish it by supposing that these interrupted perceptions are connected by a real existence that
we don’t perceive. This supposition—this idea of continued
existence—acquires force and liveliness from the memory
of the broken impressions and from that propensity they
give us to suppose them to be ·individually· the same; and
according to my theory of belief, the very essence of belief
consists in the force and liveliness of the conception.
In order to justify this system, four things are needed. •To
explain the principium individuationis, or principle of identity; •to explain why the resemblance of our broken and
interrupted perceptions induces us to attribute an identity
to them; •to explain why this illusion—·this false attribution of identity·—gives us a propensity to unite these broken
appearances by ·supposing· a continued existence; and •to
explain the force and liveliness of conception that arises from
the propensity.

·I MAGINATION :

iv: Scepticism and other systems

ther of them. At first sight this seems quite impossible: there
can’t be something intermediate between unity and number,
any more than there can between existence and non- existence. Given one object, we either have another, in which
case we have the idea of number; or we don’t have any other,
in which case the object remains at unity.
To remove this difficulty, let us get help from the idea of
time or duration. I have already observed in 5ii that time in a
strict sense implies change, and that when we apply the idea
of time to any unchanging object, supposing it to participate
in the changes of the coexisting objects and in particular
of the changes in our perceptions, this is only a fiction of
the imagination. This fiction, which almost universally takes
place, is the means by which we get a notion of identity from
a single object that we survey for a period of time without
observing in it any interruption or variation. ·Here is how it
does that·. We can consider any two points in this period in
either of two ways: we can
•survey them at the very same instant, in which case
they give us the idea of number: both as being two
points in time, and as containing perceptions of two
objects, for the objects must be multiplied in order to
be conceived as existing in these two different points
of time;
or we can
•trace the succession of time by a matching succession of ideas, conceiving first one moment along with
the object at that time, then imagine a change in
the time without any variation or interruption in the
object; and so we get the idea of unity.
Here then is an idea that is intermediate between unity and
number, or—more properly speaking—is either of them, according to how we look at it; and this is the idea that we call
the idea of identity. We can’t in propriety of speech say that

FIRST PAR T OF THE SYSTEM ·

First, as to the principle of individuation, notice that the
view of a single object is not sufficient to convey the idea
of identity. Consider the proposition An object is the same
as itself. If the idea expressed by ‘object’ is exactly the one
meant by ‘itself’, the proposition really means nothing; and
in that case it doesn’t contain a predicate and a subject,
though the sentence purports to do so. One single object
conveys the idea of unity, not of identity.
On the other hand, a number of objects can never convey
the idea of identity, however alike they may be. The mind
always pronounces this one not to be that or the other, and
considers them as forming two, three or some higher number of objects, whose existences are entirely distinct and
independent.
Since number and unity are thus both incompatible with
it, the relation of identity must lie in something that is nei103

Treatise, Book 1

David Hume

an object is the same as itself unless we mean that the object
existent at one time is the same as itself existent at another.
In this way we make a difference between the idea meant
by ‘object’ and that meant by ‘itself’, without going as far
as number yet without confining ourselves to a strict and
absolute unity.
Thus the principle of individuation is nothing but the
invariableness and uninterruptedness of an object through
a supposed variation of time, by which the mind can trace it
in the different periods of its existence, without any break
in the view, and without being obliged to form the idea of
multiplicity or number.

·I MAGINATION :

iv: Scepticism and other systems

shall call ‘object’ or ‘perception’ as seems best for my purpose
·in the given context·, understanding each word to stand for
what any common man means by ‘hat’ or ‘shoe’ or ‘stone’ or
any other impression that his senses bring to him. I shall
be sure to warn you when I return to a more philosophical
way of speaking and thinking. [See page 109.] Now we face the
question about the source of the error and deception that
we are prey to when we attribute identity to our resembling
perceptions, despite their interruption. Here I must recall
something that I proved and explained in 5ii , namely that
what is most apt to make us mistake one idea for another is
a relation between them that links them in the imagination
so that it passes easily from one to the other. The relation
that does this the most effectively is resemblance, because it
causes an association not only of ideas but also of dispositions: when some act or operation of the mind leads us to
have a certain idea, it will be led also to have a similar idea
through a similar act or operation. I have commented on the
importance of this. We can take it as a general rule that any
•two ideas that put the mind into the same disposition, or
into similar ones, are very apt to be confounded—·and thus
to be thought to be •one idea·. The mind readily passes
from one to the other and doesn’t notice the change unless
it attends very closely—and that is something of which most
people’s minds are wholly incapable.
In order to apply this general maxim, we must first examine
•the disposition of the mind when it views an object
that preserves a perfect identity,
and then find
•some other object that we wrongly identify with the
former one because it causes in us a similar disposition.
When we fix our thought on some object and suppose it

SECOND PAR T OF THE SYSTEM ·

I now proceed to show why the constancy of our perceptions makes us ascribe to them a perfect numerical identity, even though there are very long intervals between their
appearances, and even though they have only one of the
essential qualities of identity, namely invariableness. To
avoid all ambiguity and confusion about this, I explain that I
am here going to account for the opinions of common people
with regard to the existence of bodies; so I must entirely conform to their manner of thinking and talking. Now, ·some·
philosophers distinguish sense-perceptions from objects of
the senses, and suppose that the objects coexist with the
perceptions and resemble them; but, as I have already remarked, this distinction is not recognized by the general
run of people, who perceive only one thing and wouldn’t
assent to the opinion that there really are two, of which one
represents the other. For them, the very sensations that
enter by the eye or ear are the true objects, and they can’t
make much sense of the suggestion that •this pen that is
immediately perceived represents •another pen that is like it.
To accommodate myself to their notions, therefore, I shall at
first suppose that there is only a single existing thing that I
104

Treatise, Book 1

David Hume

to continue the same for some time, it’s clear that we are
supposing that only the time is changing, and we don’t put
ourselves to the trouble of producing any new image or idea
of the object. The mind’s faculties in this case are not put
to any work beyond what is necessary to continue the idea
we formerly had, which goes on existing without variation
or interruption. The passage from one moment to the next
is hardly felt, and the conception of it doesn’t involve any
difference of perception or idea. . . . That is the disposition of
the mind when it contemplates a perfectly identical object.
Now we have to discover what other objects can put the mind
in that same disposition when it considers them, causing
the same uninterrupted passage of the imagination from one
idea to another. This is of the highest importance. For if
we find any such objects, we can certainly conclude (from
the foregoing principle) that it is very natural for them to be
wrongly identified with identical objects, and are taken to be
such in most of our reasonings. But though this question
is very important, it is not very difficult or doubtful. For
I immediately reply that a sequence of related objects puts
the mind into this disposition: such a •sequence is contemplated with the same smooth and uninterrupted progress of
the imagination as accompanies a view of a •single invariable
object. The very nature and essence of ·natural· relations
is to connect our ideas with each other, and when one idea
appears to facilitate the move to the related one. The move
between related ideas is therefore so smooth and easy that
it produces little alteration in the mind, and seems like a

iv: Scepticism and other systems

continuation of a single action; and as the continuation of a
single action is an effect of the continued view of a single object, this is why we attribute singleness to every succession
of related objects, treating them as though they were a single
object. The thought slides along the succession as easily as
if it were considering only one object; and so it confounds
the succession with the identity.
We shall later see many instances of this tendency of
relations to make us ·wrongly· identify different objects with
one another, but here I shall stay with the present subject.
We find by experience that there is so much constancy in
most of the impressions of the senses that their interruption
produces no alteration in them, allowing them to returning
·to our senses· with the same appearance and situation as
they had before. I survey the furniture in my room; I shut my
eyes and then re-open them; and I find my new perceptions
to resemble perfectly the previous ones. I observe this resemblance ·across interruptions· in a thousand instances, and it
naturally connects my ideas of these interrupted perceptions
by the strongest relation, conveying the mind easily from
one to another. An easy passage of the imagination along
the ideas of these •different and interrupted perceptions is
almost the same disposition of mind as that in which we
contemplate •one constant and uninterrupted perception. It
is therefore very natural for us to mistake the one for the
other.8

·I MAGINATION :

THIRD PAR T OF THE SYSTEM ·

The people who have this opinion about the identity of
8

This reasoning is admittedly rather abstruse and hard to understand; but the remarkable fact is that this very difficulty can be turned into an argument for the reasoning! We can see that there are two resemblances that contribute to our mistaking •the sequence of our interrupted perceptions
for •an identical object. The first is the resemblance of the perceptions that are involved in each; the second is the resemblance of the acts of the
mind that are involved in each. Now we are apt to confound these resemblances with each other, ·and that is what makes this whole piece of theory
hard to get straight in one’s mind·. It is also what it is natural for us to do, according to this very theory. If you can only keep the two resemblances
distinct, you’ll have no difficulty in following my argument.

105

Treatise, Book 1

David Hume

our resembling perceptions are in general all the uthinking
and unphilosophical part of mankind, (that is, all of us at
one time or another); so they are the ones who (·as I said earlier·) suppose their perceptions to be their only objects, and
never think of a double existence: ·perception and external
object·, internal and external, representing and represented.
The very image that is present to the senses is for them (for
us!) the real body, and it is to these interrupted images we
ascribe a perfect identity. But the interruption of the appearance seems contrary to their identity, and that naturally
leads us to regard the resembling perceptions as different
from each other ·after all·. Here we find ourselves at a loss
how to reconcile such opposite opinions.
•The smooth passage of the imagination along the
ideas of the resembling perceptions makes us ascribe
to them a perfect identity. •The interrupted manner of
their appearance makes us consider them as a number of distinct though similar things that appear after
certain intervals.
The perplexity arising from this contradiction inclines us to
unite these broken appearances by the fiction of a continued
existence, which is the third part of the system I offered to
explain.
Our experience shows us—as certainly as it shows
anything—that whatever contradicts either our opinions or
our passions generates a noticeable uneasiness, whether the
contradiction comes from without or from within—from the
opposition of external objects or from the conflict of forces
inside us. On the other hand, anything that chimes with
our natural propensities, and either externally advances
their satisfaction or internally goes along with their turns of
thought and feeling, is sure to give us conscious pleasure.
Now, we have here an opposition between •the notion of the
identity of resembling perceptions and •the interruption in

iv: Scepticism and other systems

their appearance, so the mind is bound to be uneasy and
to seek relief from that uneasiness. Since the uneasiness
arises from the opposition of two contrary forces, the mind
must look for relief by sacrificing one to the other. But as
the smooth passage of our thought along our resembling
perceptions makes us ascribe an identity to them, we are
very reluctant to give up that opinion. So we must turn to
the other side ·of the dilemma·, and suppose that our perceptions are not interrupted after all, that their existence is not
only invariable but continuous, and that this enables them
to be entirely the same, strictly identical. But appearances
of these perceptions are interrupted so often and for such
long periods that we can’t overlook the interruptions; and
they seem to imply that the perceptions didn’t exist during
those periods. The alternative is to suppose that they existed but weren’t present to the mind; but this looks like a
flat contradiction that we couldn’t ever swallow, because a
perception’s existing seems at first sight to be the very same
thing as its appearing to a mind. To clear this matter up,
and to learn how an interruption in the appearance of a
perception doesn’t necessarily imply an interruption in its
existence, I need to touch on some principles that I’ll have
occasion to explain more fully in section 6.
I begin by observing that our present difficulty is not
about the factual question of whether the mind does form
such a conclusion about the continued existence of its perceptions, but only about how it does so, about what forces
are at work in this. It is certain that almost all mankind—and
even philosophers most of the time—take their perceptions
to be their only objects, and suppose that the very thing that
is intimately present to the mind is the real body or material
thing. It is also certain that this very perception or object is
supposed to have a continued uninterrupted existence, and
to be neither annihilated by our absence nor brought into
106

Treatise, Book 1

David Hume

existence by our presence. We say:
When we are absent from it, it still exists, but we don’t
feel, we don’t see it. When we are present, we feel or
see it.
So two questions arise. •How can we get ourselves to be
satisfied in supposing a perception to be absent from the
mind without being annihilated? •How do we conceive an
object to become present to the mind, without some new
creation of a perception or image; and what do we mean by
‘seeing’ and ‘feeling’ and ‘perceiving’ an object? As to the
first question, I would remark that what we call ‘a mind’ is
nothing but a heap or collection of different perceptions, held
together by certain relations and wrongly supposed to be
endowed with a perfect simplicity and identity. Now, every
perception is distinguishable from every other, and can be
considered as existing separately from any other; from which
it clearly follows that there is no absurdity in separating any
particular perception from the mind—that is, in breaking off
all its relations with that heap of connected perceptions that
constitute a thinking being.
The same reasoning gives us an answer to the second
question. If the label ‘perception’ doesn’t make this separation from a mind absurd and contradictory, the label ‘object’,
standing for the very same thing, can’t make a presence to
the mind impossible. External objects are seen and felt and
become present to the mind; that is, they acquire such a relation to a connected heap of perceptions as to influence them
very considerably in augmenting their number by present
reflections and passions, and in storing the memory with
ideas. The same continued and uninterrupted being can
therefore be sometimes present to the mind and sometimes
absent from it, without any real or essential change in the
being itself. An interrupted appearance to the senses doesn’t
necessarily imply an interruption in the existence; the sup-

iv: Scepticism and other systems

position of the continued existence of perceptible objects or
perceptions involves no contradiction; we can easily go along
with out inclination to make that supposition. When the
exact resemblance of our perceptions makes us ascribe to
them an identity, we can remove the seeming interruption
by feigning a continued being that can fill those intervals
and preserve a perfect and entire identity to our perceptions.
[Feigning is creating a fiction; the two words come from a single Latin
word.]

·I MAGINATION :

FOUR TH PAR T OF THE SYSTEM ·

But we don’t just feign this continued existence—we believe in it. Where does this belief come from? This question
leads us to the fourth part of my system. I have already
shown that belief in general consists in nothing but the liveliness of an idea, and that an idea can acquire this liveliness
by its •relation to some present impression. Impressions are
naturally the most vivid perceptions of the mind, and some
of this vividness is conveyed by the •relation to every connected idea. The relation •disposes the mind to go from the
impression to the idea, and causes the move to be a •smooth
one. The mind goes so easily from the one perception to
the other that it hardly notices the change, and retains in
the second perception (·the idea·) a considerable share of
the liveliness of the first (·the impression·). It is aroused by
the lively impression, and this liveliness is conveyed without much loss to the related idea, because of the •smooth
transition and the •disposition of the imagination.
Even if this disposition arises from something other than
the influence of relations, its source—whatever it is—must
obviously have the same effect ·as I have been attributing to
relations·, and must convey the liveliness from the impression to the idea. And that is what we have in our present case.
Our memory presents us with a vast number of instances
of perceptions perfectly resembling each other that return
107

Treatise, Book 1

David Hume

at different distances of time and after considerable interruptions. This resemblance disposes us to consider these
·different· interrupted perceptions as ·being· •the same; and
also disposes us to connect them by a continued existence
in order to justify •this identity and avoid its seeming contradiction with the interrupted appearance of these perceptions.
So we are disposed to feign the continued existence of all
perceptible objects; and as this disposition arises from some
lively impressions of the memory it gives liveliness to that
fiction—which is to say that it makes us believe in the continued existence of the body. If we sometimes ascribe a
continued existence to objects that are perfectly new to us,
and of whose constancy and coherence we have no experience, it is because they present themselves to our senses
in a manner that resembles that of constant and coherent
objects; and this resemblance is a source of reasoning and
analogy, leading us to attribute the same qualities to objects
that are similar.
I believe a thinking reader will find it easier to •assent to
this system than to •grasp it fully and clearly, and after a
little thought will agree that every part carries its own proof
along with it. ·I shall now run through the argument again
in a slightly different way·. It is obvious that as common
people
•suppose their perceptions to be their only objects,
and at the same time
•believe in the continued existence of matter,
we have to explain how •that belief can arise for people who
make •that supposition. Now, on that supposition it is not
true that any of our objects (or perceptions) is identically
the same after an interruption; and consequently the opinion of their identity can never arise from reason, but must
arise from the imagination. The imagination is seduced into
this opinion only by the resemblance of certain perceptions

iv: Scepticism and other systems

(evidence for this: our resembling perceptions are the only
ones that we are disposed to suppose the same). This disposition to confer an identity on our resembling perceptions
produces the fiction of a continued existence. That fiction ·is
properly so-called· because it, as well as the identity, really
is false, as all philosophers agree, and its only effect is to
remedy the interruption of our perceptions which is the only
obstacle to their identity. Finally, this disposition causes
belief by means of the present impressions of the memory
(evidence: without memories of former sensations we would
obviously never have any belief in the continued existence of
body). Thus, in examining all these parts, we find that each
of them is supported by the strongest proofs; and that all
of them together form a consistent system that is perfectly
convincing. . . .
But although the natural disposition of the imagination
leads us in this way to ascribe a continued existence to those
perceptible objects or perceptions that we find to resemble
each other in their interrupted appearance, a very little reflection and philosophy is sufficient to make us see the fallacy
of that opinion. I have already remarked that there is an
intimate connection between the two theses, of a •continued
existence and of a •distinct or independent existence, and
that we no sooner establish one than the other follows as
a necessary consequence. It is the belief in a continued
existence that comes first, and without much study or reflection pulls the other along with it. . . . But when we compare
experiments and think about them a little, we quickly see
that the doctrine of the independent existence of our sensory
perceptions is contrary to the plainest experience. This leads
us to retrace our steps and perceive our error in attributing
a continued existence to our perceptions. It is the origin of
many very curious opinions that I shall here try to account
for.
108

Treatise, Book 1

David Hume

the identity and continuity of our interrupted •perceptions.
If we weren’t first convinced that our perceptions are our
only objects, and continue to exist even when they no longer
appear to the senses, we would never be led to think that our
perceptions and our objects are different, and that it is only
our objects that have a continued existence. ·I contend·:
The philosophical hypothesis •has no primary recommendation either to reason or the imagination, and
•acquires all its influence on the imagination from the
common hypothesis.
This ·displayed· proposition contains two parts, which I shall
try to prove as distinctly and clearly as such abstruse subjects will permit.
As to the first part of the proposition that this philosophical hypothesis has no primary recommendation either to
reason or the imagination, we can soon satisfy ourselves
with regard to reason, by the following reflections. The only
existences of which we are certain are perceptions that, being immediately present to us in consciousness, command
our strongest assent and are the ultimate basis of all our
conclusions. The only conclusion we can draw from the
existence of one thing to the existence of another is through
the relation of cause and effect, showing that there is a connection between them and that the existence of one depends
on that of the other. The idea of the cause-effect relation is
derived from past experience in which we find that two ·kinds
of· beings are constantly conjoined and are always present
together to the mind. But no beings are ever present to the
mind except perceptions; so we can observe a conjunction or
cause-effect relation •between different perceptions, but can
never observe it •between perceptions and objects. So it is
impossible that from any fact about perceptions we can ever
validly form any conclusion about the existence of objects
·when these are understood, as they are in the philosophical

First I should mention a few of those experiential episodes
that convince us that our perceptions don’t have any independent existence. When we press one eye with a finger, we
immediately perceive all the objects to become double, and
half of them to be removed from their usual position. But as
we don’t attribute a continued existence to both these perceptions, and as they are both of the same nature, we clearly
perceive that all our perceptions depend on our organs and
the disposition of our nerves and animal spirits. This is
confirmed by the seeming growth and shrinkage of objects
according to how far away they are, by the apparent alterations in their shapes, by the changes in their colour and
other qualities, when we are ill, and by countless other experiences of the same kind—from all which we learn that our
sensible perceptions don’t have any distinct or independent
existence.

·T HE

iv: Scepticism and other systems

NEW PHILOSOPHICAL SYSTEM ·

The natural consequence of this reasoning should be that
our perceptions don’t have a continued existence either; and
indeed philosophers have reached this view so thoroughly
that they change their system, and distinguish (as I shall do
from here on) between •perceptions and •objects. They hold
that perceptions are interrupted and perishing, and different
at every different return ·to our senses·; and that objects
are uninterrupted and preserve a continued existence and
identity. But however philosophical this new system may
be thought to be, I contend that it is only a superficial remedy, and that it contains all the difficulties of the common
system along with some others that are all its own. There
are no drives in either the understanding or the imagination
that lead us directly to embrace this opinion of the double
existence of •perceptions and •objects, and we can’t arrive
at it except by passing through the common hypothesis of
109

Treatise, Book 1

David Hume

hypothesis, as different from perceptions·.
It is no less certain that this philosophical system has no
primary recommendation to the imagination, which would
never have arrived at such a view on its own and through
forces that are intrinsic to it. It will be somewhat difficult
to prove this to your full satisfaction, I admit, because it
implies a negative, and negatives very often don’t admit of
any positive proof. If someone would take the trouble to look
into this question and invent a system ·claiming· to account
for how this opinion does arise directly from the imagination,
we could by examining that system reach a certain judgment
on the present topic. Thus:
Let it be taken for granted that •our perceptions are
broken and interrupted, and that •however alike they
are they are still different from each other; and let anyone on this basis show why •the imagination directly
and immediately (·not through the indirect mechanism I have proposed·) proceeds to the belief in another existing thing that resembles these perceptions
in their nature but ·differs from them in being· continuous and uninterrupted and identical.
When someone has done this to my satisfaction, I promise
to renounce my present opinion. Meanwhile I can’t help
thinking that this, because of the very abstractedness and
difficulty of the first supposition [Hume’s phrase], is not fit
material for the imagination to work on. Whoever wants to
explain the origin of the common opinion about the continued and distinct existence of body must focus on the mind
as it commonly is, and proceed on the supposition that our
perceptions are our only objects and continue to exist even
when not perceived. This opinion is false, but it is the most
natural of any, and is the only one that has any primary
recommendation to the imagination. As to the second part of
the proposition ·that is displayed a page back·, that the philo-

iv: Scepticism and other systems

sophical system acquires all its influence on the imagination
from the common one: this is a natural and unavoidable
consequence of the foregoing conclusion that the philosophical system has no primary recommendation to reason or the
imagination. We find by experience that the philosophical
system does take hold of many minds, especially of all those
who reflect even a little on this subject; so it must derive all
its authority from the common system, as it has no authority
of its own. These two systems, though directly contrary, are
connected together and here is how.
The imagination naturally thinks along the following
lines:
•Our perceptions are our only objects.
•Resembling perceptions are the same, however broken
or uninterrupted in their appearance.
•This apparent interruption is contrary to the identity.
•So it is only an apparent interruption, and the perception or object really continues to exist even when
absent from us.
•So our sensory perceptions have a continued and
uninterrupted existence.
But as a little reflection destroys this conclusion that our
perceptions have a continued existence by showing that they
have a dependent one—·and I have shown that they couldn’t
be continuous unless they were independent·—it would naturally be expected that we should altogether reject the opinion
that Nature contains any such thing as a continued existence that is preserved even when it no longer appears to the
senses. But that is not what has happened! Philosophers
don’t in general infer from
Our sensory perceptions are dependent and not continuous
that
Nothing has a continued existence ·through gaps in
110

Treatise, Book 1

David Hume

our perceptions·.
Indeed, they are so far from making that inference that although all philosophical sects agree with the former view,
the latter—which is in a way its necessary consequence—has
been the property only of a few extravagant sceptics; and
even they have maintained it in words only, and were never
able to bring themselves sincerely to believe it.
There is a great difference between opinions that we form
after calm deep thought and ones that we embrace by a
kind of instinct or natural impulse because of their suitableness and conformity to the mind. When opinions of these
two kinds come into conflict, it is easy to foresee which will
win! As long as our attention is focussed on the subject,
the philosophical and studied principle may prevail; but
the moment we relax our thoughts, Nature will display herself and pull us back to our former ·instinctive or natural·
opinion. Indeed, Nature sometimes has so much influence
that she can stop us in our tracks, even in the middle of
our deepest reflections, and keep us from running on into
all the consequences of some philosophical opinion. Thus,
though we clearly perceive the dependence and interruption
of our perceptions, we come to an abrupt halt and don’t
infer that there is nothing independent and continuous. The
opinion that there are such things has taken such deep root
in the imagination that it is impossible ever to eradicate it;
no tenuous metaphysical conviction of the dependence of
our perceptions is sufficient for that purpose.
But though our natural and obvious drives here prevail over our studied reflections, there must surely be some
struggle and opposition over this, at least so long as these
reflections retain any force or liveliness. In order to set
ourselves at ease in this respect, we contrive a new hypothesis that seems to take in both these influences—of reason
and of imagination. This is the philosophical hypothesis of

iv: Scepticism and other systems

the double existence of perceptions and objects: it pleases
our reason by allowing that our dependent perceptions are
interrupted and different, and it is also agreeable to the
imagination because it attributes a continued existence to
something else that we call ‘objects’. This philosophical system, therefore, is the misshaped offspring of two principles
that are •contrary to each other, are •both at once embraced
by the mind, and are •unable mutually to destroy each other.
The imagination tells us that our resembling perceptions
•have a continued and uninterrupted existence, and
are not annihilated by being absent from us.
Reflection tells us that even our resembling perceptions
•are interrupted in their existence, and are different
from each other.
We escape the contradiction between these opinions by a
new fiction that squares with the hypotheses of both reflection and imagination by ascribing these contrary qualities to
different existences—•the interruption to perceptions, and
•the continuity to objects. Nature is obstinate, and refuses
to give up, however strongly it is attacked by reason; and
at the same time reason is so clear about this matter that
there is no possibility of disguising it ·by muffling and then
evading its message·. Not being able to reconcile these two
enemies, we try to set ourselves at ease as much as possible by successively granting to each whatever it demands,
and by feigning a double existence in which each can find
something that meets all the conditions it lays down. ·Look
at how we get ourselves into this·:
If we were fully convinced that our resembling perceptions are continued and identical and independent, we
would never go for this opinion of a double existence;
because in that case we would find satisfaction in our
first supposition, and would not look beyond.
On the other hand,
111

Treatise, Book 1

David Hume

If we were fully convinced that our perceptions are
dependent and interrupted and different, we would be
equally disinclined to embrace the opinion of a double
existence; because in that case we would clearly perceive the error of our first supposition of a continued
existence, and give it no further thought.
So the opinion of a double existence arises from the half-way
situation of the mind—from adhering to these two contrary
principles in such a way as to seek some pretext to justify accepting both; which (happily!) is found at last in the system
of a double existence.
Another advantage of this philosophical system is its similarity to the common one: it enables us to humour our
reason for a moment when it becomes troublesome and anxious, but as soon as reason’s attention flags, the system
makes it easy to us to return to our common and natural
notions. Sure enough, we find that philosophers make use of
this advantage: as soon as they leave their studies they join
with the rest of mankind in those exploded opinions that our
perceptions are our only objects, and continue identically
and uninterruptedly the same through all their interrupted
appearances.
Other aspects of the philosophical system show very conspicuously its dependence on the imagination. I shall note
two of them. First, ·in the philosophical system· we suppose
external objects to resemble internal perceptions. I have
already shown that the relation of cause and effect can never
let us soundly infer the existence of external continuous
objects from the existence or qualities of our perceptions;
and I now add that even if we could justify such an inference,
we should never have any reason to infer that our objects
resemble our perceptions. So that opinion is comes purely
from the quality of the imagination that I have explained
above, namely that it borrows all its ideas from some earlier

iv: Scepticism and other systems

perception. We never can conceive anything but perceptions,
so ·in our imagination· we must make everything resemble
them.
Secondly, ·in the philosophical system· we don’t merely
suppose our objects to resemble our perceptions in a general way; we also take it for granted that each particular
object resembles the perception that it causes. The relation
of cause and effect makes us bring in that other relation,
resemblance; and since the ideas of these items—·the object
and the perception of it·—are already united together in the
imagination by the former relation (·cause-effect·), we naturally add the latter (·resemblance·) to complete the union. We
have a strong disposition to complete every union by joining
new relations to those that we have before observed between
any ideas, as I shall have occasion to remark in section 5.

·F INAL

REMARKS ·

Having thus given an account of all the systems, both
popular and philosophical, with regard to external existents,
I can’t help expressing a certain attitude that arises in me
when I review those systems. I began this subject by laying
it down that we ought to have an unquestioning faith in our
senses, and that this would be the conclusion I would draw
from the whole of my reasoning. Frankly, however, I feel
myself at present in a quite contrary frame of mind, and am
more inclined to put •no faith at all in my senses (or rather
my imagination) than to place in it such an •unquestioning
confidence. I can’t conceive how such trivial qualities of the
imagination, guided by such false suppositions, can ever
lead to any solid and rational system. I mean the qualities of
the coherence and constancy of our perceptions, which produce the opinion of their continued existence, although these
qualities of perceptions have no perceivable connection with
such an existence. The constancy of our perceptions has the
most considerable effect, and yet it is the one that brings
112

Treatise, Book 1

David Hume

the greatest difficulties. It is a gross illusion to suppose that
our resembling perceptions are numerically the same ·after
an interruption·; and it is this illusion that leads us to the
view that these perceptions are not interrupted and still exist
when not present to our senses. So much for our popular
system! As to our philosophical system: it suffers from the
same difficulties, and in addition to them it is loaded with
the absurdity of at once •denying and •asserting the common
supposition! Philosophers say that our resembling perceptions are not identically the same and uninterrupted; yet
they have so great a disposition to believe that they are that
they arbitrarily invent a new set of perceptions to which they
attribute these qualities. (I say ‘a new set of perceptions’ ·for
a good reason·. We can in a general ·vague· way suppose
there are objects that are not perceptions, but it is impossible
for us to think clearly and sharply about objects as being in
their nature anything but exactly the same as perceptions.)
What then can we look for from this confusion of groundless
and extraordinary opinions but error and falsehood? And
how can we justify to ourselves any belief in them? This sceptical doubt, with respect to both reason and the senses, is

iv: Scepticism and other systems

an illness that can never be thoroughly cured; it is bound to
return upon us every moment, even if we chase it away and
sometimes seem to be entirely free from it. On no system is it
possible to defend either our understanding (·i.e. reason·) or
our senses, and when we try to justify them in that manner
·that I have been discussing· we merely expose their defects
further. As the sceptical doubt arises naturally from deep
and hard thought about those subjects, it always increases
as we think longer and harder, whether our thoughts are
in opposition to sceptical doubt or conformity with it. Only
•carelessness and •inattention can give us any remedy. For
this reason I rely entirely on •them; and I take it for granted
that whatever you may think at this present moment, in an
hour from now you will be convinced that there is both an
external and internal world; and on that supposition—·that
there is an external as well as an internal world·—I intend
now to examine some general systems, ancient and modern,
that have been proposed regarding both ‘worlds’, before I
proceed ·in section 5· to a more particular enquiry about our
impressions. This may eventually be found to be relevant to
the subject of the present section.

3: The ancient philosophy
Several moralists have recommended, as an excellent method
of becoming acquainted with our own hearts and knowing
our progress in virtue, to recollect our dreams in the morning and examine them as severely as we would our most
serious and deliberate actions. Our character is the same
sleeping as waking, they say, and it shows up most clearly
when deliberation, fear, and scheming have no place, and
when men can’t try to deceive themselves or others. The
generosity or baseness of our character, our mildness or

cruelty, our courage or cowardice, are quite uninhibited in
their influence on the •fictions of the imagination, revealing
themselves in the most glaring colours. In a similar way I
believe that we might make some useful discoveries through
a criticism of the •fictions of ancient philosophy concerning
substances, substantial forms, accidents, and occult qualities; those fictions, however unreasonable and capricious
they may be, have a very intimate connection with the forces
at work in human nature.
113

Treatise, Book 1

David Hume

iv: Scepticism and other systems

not •tracking it gradually through the successive moments,
but instead •surveying at once any two distinct periods of
its duration, and •laying its qualities at those two moments
side by side in our minds, then the variations that we didn’t
notice when they arose gradually appear significant, and
seem entirely to destroy the identity. Thus there comes to be
a kind of contrariety in our method of thinking, because of
the different •points of view from which we survey the object
and the different •lengths of time between the moments that
we consider together. ·Here is the essential contrast·:
•When we gradually follow an object through its successive changes, the smooth progress of our thought
makes us ascribe an identity to the sequence, because this smooth progress is similar to our act of the
mind when we consider an unchanging object. •When
we compare its situation after a considerable change
·with its situation before·, the progress of the thought
is broken, so that we are presented with the idea of
diversity, i.e. non-identity.
To reconcile these contradictory positions the imagination is
apt to feign something unknown and invisible which it supposes to continue the same under all these variations; and
this unintelligible something it calls a ‘substance’, or ‘original
and first matter’.
We have a similar notion with regard to the simplicity
of substances, and from similar causes. Suppose that •a
perfectly simple and indivisible object is presented, along
with •another object—·a composite one·—whose coexistent
parts are linked by a strong relation. Obviously the actions
of the mind in considering these two objects are not very
different. The imagination conceives the simple object
at once, easily, by a single effort of thought, without
change or variation.
The connection of parts in the composite object has almost

The most judicious philosophers agree that our ideas of
bodies are nothing but
collections formed by the mind of the ideas of the
various distinct perceptible qualities of which objects
are composed and which we find to have a constant
union with each other.
Although these qualities are in themselves entirely distinct
·from one another·, it is certain that we commonly regard the
compound that they form as one thing and as continuing to
be that thing while it undergoes very considerable alterations.
The admitted •compositeness is obviously contrary to this
supposed •simplicity, just as the •alteration is contrary to
the •identity. So it may be worthwhile to consider the causes
that make us almost universally fall into such evident contradictions, and also the means by which we try to conceal
them. [In this context, ‘simple’ means ‘without parts’. In equating ‘x is
one thing’ with ‘x is simple’, Hume is assuming that an item with parts—
a ‘composite’ item—is really a collection of its parts, not really one thing.]

The ideas of the various different qualities that an object has
one after another are linked by a very close relation; so when
the mind looks along the series it is carried from one part of
it to another by an easy transition, and doesn’t perceive the
change any more than it would perceive a change when contemplating a single unchanging object. This easy transition
is an effect . . . . of the relation ·between each quality and its
successor·; and as the imagination readily identifies one idea
with another when their influence on the mind is similar, it
comes about that the mind considers any such •sequence of
related qualities as •one continuous object, existing without
any alteration. The smooth and uninterrupted movement of
thought, being alike in both cases, easily deceives the mind
and makes us ascribe an identity to the changing sequence
of connected qualities.
But when we look at the sequence in a different way,
114

Treatise, Book 1

David Hume

the same effect ·on the contemplating mind·: it unites the
object within itself in such a way that the imagination doesn’t
feel the transition when it passes from one part to another.
Thus the colour, taste, shape, solidity, and other qualities
that are combined in a peach or a melon are thought of as
forming one thing; and this happens because of their close
relation, which makes them affect our thought in the same
way as if the object were perfectly uncompounded—·i.e. had
no parts at all·. But the mind doesn’t stop at that. When
it views the object in a different way it finds that all these
qualities are different, distinguishable, and separable from
each other; that view of things destroys the mind’s primary
and more natural notions, and obliges the imagination to
feign an unknown something—an original •substance and
•matter—as a source of the union or cohesion among these
qualities, and as what may entitle the composite object to be
called one thing, despite its diversity and compositeness.
The Aristotelian philosophy says that the ‘original’ matter
is absolutely the same in all bodies, and it considers fire,
water, earth, and air as being of the very same substance
because of their gradual changes into each other. At the
same time it assigns to each of these sorts of objects a distinct substantial form that it supposes to be the source of
all the different qualities the objects possess, and to be a
new basis for simplicity and identity for each particular sort.
All depends on how we look at the objects. •When we look
along the imperceptible changes of bodies, we suppose all
of them to be of the same substance or essence. •When we
consider their perceptible differences, we attribute to each of
them a substantial and essential difference. •And to allow
ourselves to keep both these ways of considering our objects,
we suppose all bodies to have at once a substance and a
substantial form.
The notion of accidents [= ‘qualities’] is an unavoidable con-

iv: Scepticism and other systems

sequence of this way of thinking about substances and substantial forms. [Hume uses ‘quality’ freely throughout the Treatise.
He uses ‘accident’ for qualities thought of as existing things that have to
be kept in existence by other things, namely the substances that have

We can’t help thinking of colours, sounds, tastes,
shapes, and other properties of bodies as existents that can’t
exist on their own and have to be supported by something
in which they inhere. For whenever we have discovered any
of these perceptible qualities we have, for the reasons mentioned above, imagined a substance to exist also; the same
habit that makes us infer •a connection between cause and
effect here makes us infer •a dependence of every quality on
an unknown substance.
The custom of imagining a dependence has the same effect as the custom of observing it would have. But this turn
of thought is no more reasonable than any of the previous
ones. Every quality is distinct from every other, and can be
conceived to exist on its own—apart from every other quality
and also from that unintelligible chimera of a substance.
But these philosophers carry their fictions still further in
their opinions about occult qualities: they suppose a supporting substance, which they don’t understand, and a supported accident, of which they have no better an idea. The
whole system, therefore, is entirely incomprehensible, and
yet is derived from principles as natural as any of the ones I
have explained.
In considering this subject, we can see that as the people
concerned acquire new degrees of reason and knowledge,
their opinions rise up through three levels. These opinions
are •that of the common people, •that of a false philosophy, and •that of the true philosophy—and we shall find
when we look into it that the true philosophy is closer to the
views of the common people than it is to those of a mistaken
knowledge ·such as many philosophers have·. It is natuthem.]

115

Treatise, Book 1

David Hume

ral for men in their common and careless way of thinking
to imagine that they perceive a connection between objects
that they have constantly found united together; and because custom has made it hard for them to separate the
ideas, they are apt to imagine such a separation to be in
itself impossible and absurd. ·Thus, for example: Someone
observes—for things (x) like middle-sized physical objects—
that •x-is-left-unsupported is almost always followed immediately by •x-falls-to-the-ground; this creates in him a custom
of expectation, in which an impression of •x-unsupported
leads quickly and smoothly and easily to an idea of •x-falling;
and this inclines him to think that the idea of •non-support
is absolutely tied to the idea of •falling in the way that the
idea of being square is tied to the idea of being rectangular;
which means that he is inclined to think he can see that it is
absolutely (logically) impossible for an unsupported object
of the relevant kind not to fall·. But philosophers, who set
aside the effects of custom and look for relations between
the ideas of objects, immediately see the falsehood of these
common opinions and discover that there is no known connection among objects—·that is, none of the kind involving a
connection between the ideas of the objects·. Every object
appears to them entirely distinct and separate from every
other; and they see that when we infer one from another, our
basis is not a view of the nature and qualities of the objects
but only an experience of having often observed ·objects of
those kinds· to have been constantly conjoined. But these
philosophers, instead of soundly inferring from this that we
don’t have any idea of mind-independent objective power
or agency, frequently search for the qualities in which this
agency consists, and are displeased with every account of
it that their reason suggests to them. Their intellects are
sharp enough to keep from the common error that there
is a natural and perceivable connection ·of ideas· between

iv: Scepticism and other systems

matter’s various perceptible qualities and how it behaves,
but not sharp enough to keep them from looking for such
a connection in matter itself—in the causes themselves. If
they had found their way to the right conclusion, they would
have turned back to the situation of the common people, and
would have adopted a lazy ‘don’t care’ attitude to all these
long investigations ·into the causal tie·. As things are, they
seem to be in a very lamentable condition, much worse that
the poets present in their descriptions of the punishments of
Sisyphus and Tantalus. For what could be more tormenting
than to seek eagerly something that always flies away from
us, and to seek it in a place where it can’t possibly be?
But as Nature seems to have observed a kind of justice and compensation in everything, she hasn’t neglected
philosophers more than the rest of the creation, but has provided them with a consolation amid all their disappointments
and afflictions. This consolation principally consists in their
invention of the words ‘faculty’ and ‘occult quality’. After the
frequent use of a term that is significant and intelligible, we
often omit the idea that we mean to express by it, and preserve only the custom by which we recall the idea when we
want to; so it naturally happens that after the frequent use
of a term that is wholly insignificant and unintelligible, we
fancy it to be on the same footing with the meaningful ones
and to have a meaning that we don’t actually have in mind
but that we could bring to mind if we thought about it. . . . By
this means these philosophers set themselves at ease, and
eventually arrive through •an illusion at the same ‘don’t care’
attitude that common people achieve through •their stupidity, and true philosophers achieve through •their moderate
scepticism. They need only to say that a phenomenon that
puzzles them arises from a ‘faculty’ or an ‘occult quality’ and
there’s an end of all dispute and enquiry about it!
But among all the examples of the ancient Aristotelians’
116

Treatise, Book 1

David Hume

showing they were guided by every trivial twist of the imagination, none is more remarkable than their ‘sympathies’,
‘antipathies’, and ‘horrors of a vacuum’! There is a very
remarkable inclination in human nature to attribute to external objects the same emotions that it observes in itself,
and to find everywhere those ideas [here = ‘qualities’] that are
most present to it. This inclination is suppressed by a little
reflection, and it occurs only in children, poets, and the
ancient philosophers. It appears in children when they want

iv: Scepticism and other systems

to kick the stones that hurt them; in poets by their readiness
to personify everything; and in the ancient philosophers by
these fictions of ‘sympathy’ and ‘antipathy’. We must pardon
•children because of their age, and •poets because they are
openly obedient to the promptings of their imagination; but
what excuse shall we find to justify our •philosophers—·the
ancients and their modern disciples·—in such a striking
weakness?

4: The modern philosophy
You may want to object:

the latter rejected. Someone who hears an articulate voice in
the dark and concludes that there is someone there reasons
soundly and naturally, even though his inference is derived
from nothing but custom, which brings him a lively idea of
a human creature because of his usual conjunction of that
with the present impression ·of the voice·. But someone who
is tormented—he knows not why—with the fear of spectres
in the dark may perhaps be said to reason, and indeed to
reason ‘naturally’; but then it must be in the same sense
that a malady is said to be ‘natural’ because it arises from
natural causes, even though it is contrary to health, which
is the most agreeable and most natural condition for a man
to be in.

You say yourself that the imagination is the ultimate
judge of all systems of philosophy. So you are unjust
in blaming the ancient philosophers for making use of
their imagination, and letting themselves be entirely
guided by it in their reasonings.
In order to justify myself, I must distinguish two kinds of
forces that are at work in the imagination: •those that are
permanent, irresistible, and universal, such as the customary transition from causes to effects and from effects to
causes, and •those that are changing, weak, and irregular;
such as those on which I have just been commenting. •The
former are the foundation of all our thoughts and actions,
so that if they were lost human nature would immediately
perish and go to ruin. •The latter are not ones that must
be at work in mankind, and they are not necessary for the
conduct of life or even useful in it. On the contrary, we see
them at work only in weak minds, and because they are
opposite to the former forces of custom and reasoning they
can easily be overthrown when confronted by the opposition.
For this reason, the former are accepted by philosophy and

The opinions of the ancient philosophers, their fictions
of substance and accident, and their reasonings about substantial forms and occult qualities, are like spectres in the
dark! They are driven by forces which, however common,
are neither universal nor unavoidable in human nature. The
modern philosophy claims to be entirely free from this defect,
and to arise only from the solid, permanent, and consistent
principles of the imagination. We must now look into the
117

Treatise, Book 1

David Hume

grounds for this claim.
The fundamental principle of that philosophy is the opinion about colours, sounds, tastes, smells, heat, and cold,
which it asserts to be nothing but impressions in the mind,
derived from the operation of external objects and without
any resemblance to the qualities of the objects. Having examined the reasons commonly produced for this opinion, I find
only one of them to be satisfactory, namely the one based on
the variations of those impressions even while the external
object seems to remain unaltered. These variations depend
on various factors. •Upon the different states of our health:
a sick man feels a disagreeable taste in food that used to
please him the most. •Upon the different conditions and
constitutions of men: stuff that seems bitter to one man is
sweet to another. •Upon differences in location and distance:
colours reflected from the clouds change according to the
distance of the clouds, and according to the angle they make
with the eye and the luminous body. Fire also communicates the sensation of pleasure at one distance and of pain
at another. Instances of this kind are very numerous and
frequent.
The conclusion drawn from them is also utterly satisfactory. When different impressions of the same sense come
from an object, it certainly can’t be that each of these impressions resembles a quality that exists in the object. (Why? Because one object can’t, at one time have different qualities of
the same sense, and one quality can’t resemble impressions
that are entirely different from one another.) It evidently
follows that many of our impressions have no external model
or archetype [= ‘thing from which something is copied’]. Now, from
similar effects we presume similar causes. Many of our impressions of colour, sound, etc., are admittedly nothing but
internal existences ·with no archetypes in Nature·, arising
from causes that don’t resemble them in the slightest. These

iv: Scepticism and other systems

impressions are in appearance in no way different from the
other impressions of colour, sound, etc. So we conclude that
they all have causes of that sort.
Once this principle has been accepted, all the other doctrines of the modern philosophy seem to follow by an easy
inference:
Once we have removed sounds, colours, heat, cold,
and other perceptible qualities from the category of
continuous independent existents, we are left with
only what are called ‘primary qualities’, as the only
real ones of which we have any adequate notion.
These primary qualities are extension and solidity,
with their different mixtures and special cases: shape,
motion, gravity, and cohesion. The generation, growth,
decline, and death of animals and vegetables are nothing but changes of shape and motion, as are all the
operations of bodies on each other, and the operations
of fire, light, water, air, earth and all the elements and
powers of Nature. One shape and motion produces another shape and motion; and we can’t form even the
remotest idea of any force or drive (active or passive)
among systems of matter other than that one.
I think that many objections could be made to this system,
but at present I shall confine myself to one that I think is
very decisive. I contend that instead of explaining the operations of external objects by means of this system, we
utterly annihilate all these objects and reduce ourselves to
the opinions of the most extravagant scepticism about them.
If colours, sounds, tastes, and smells are merely perceptions,
nothing that we can conceive has a real, continuous, and
independent existence—not even motion, extension, and solidity, which are the primary qualities emphasized most ·in
the modern philosophy·.
To start with motion: obviously this quality is altogether
118

Treatise, Book 1

David Hume

iv: Scepticism and other systems

depends on the idea of •solidity. So the idea of solidity can’t
possibly depend on either of those two ideas (·motion and
extension·), for that would be to run in a circle, make one
idea depend on another which at the same time depends on
it. Our modern philosophy, therefore, provides us with no
sound or satisfactory idea of solidity or, therefore, of matter.

inconceivable except when thought of as the motion of an
object: the idea of motion necessarily supposes that of a
moving body. Now, what is our idea of the moving body,
without which motion is incomprehensible? It must come
down to the idea of •extension or of •solidity; so the reality of
motion depends on the reality of those other two qualities.
Everyone agrees with this opinion about motion, ·namely
that it is conceivable only as the motion of something·; and
I have proved that it holds also with regard to extension,
·which is conceivable only as the extension of something·—I
have shown that it is impossible to conceive extension except as composed of parts that have either colour or solidity.
The idea of extension is a compound idea; but it isn’t compounded out of infinitely many parts or lesser ideas, so it
must eventually be made up of parts that are perfectly simple and indivisible ·and thus don’t have parts in their turn·.
These simple and indivisible parts are not themselves ideas
of extension ·because extension must have parts·, so they
must be non-entities, nothings, unless they are conceived as
coloured or solid. Colour is excluded from any real existence
·by the modern philosophy which I am now examining·. The
reality of our •idea of extension therefore depends on the
reality of •our idea of solidity; the former can’t be sound if
the latter is chimerical. Let us look, then, into the idea of
solidity.
The idea of •solidity is the idea of •two objects which,
however hard they are pushed, can’t penetrate each other,
but still maintain a separate and distinct existence. So solidity is perfectly incomprehensible taken on its own, without
the conception of some bodies that are solid and maintain
this separate and distinct existence. Now, what idea do we
have of these bodies? The ideas of colours, sounds, and
other ‘secondary qualities’ are excluded. The idea of •motion
depends on the idea of •extension, and the idea of •extension

This argument will appear entirely conclusive to anyone
who understands it; but it may seem abstruse and complicated to the general run of readers, so I shall try to make
it obvious by some changes of wording. To form an idea
of solidity we must conceive two bodies pressing on each
other without any penetration; and we can’t do that if we
confine ourselves to one object. (And still less if we don’t
conceive any: two non-entities can’t exclude each other from
their places, because they don’t have places and don’t have
qualities.) What idea do we form of these bodies or objects
to which we attribute solidity? To say that we conceive
them merely as solid is to run on ad infinitum. To affirm
that we depict them to ourselves as extended either •bases
everything on a false idea or •brings us around in a circle. Extension must necessarily be considered either as •coloured,
which is a false idea ·according to the modern philosophy,
which says that nothing out there in the world is coloured·,
or as •solid, which brings us back to where we started. The
same argument applies regarding mobility and shape; and
so ultimately we have to conclude that after the exclusion
of colour, sounds, heat, and cold from the category of external existents there remains nothing that can give us a
sound and consistent idea of body. . . . Let us remember here
our accustomed method of examining ideas by considering
the impressions from which they came. The impressions
that enter through the sight and hearing, smell and taste,
are affirmed by modern philosophy to have no resembling
·external· objects; so the idea of solidity, which is supposed
119

Treatise, Book 1

David Hume

to be real—·i.e. to resemble external objects·—can’t be derived from any of those senses. So all that remains is the
sense of touch as a conveyor of the impression that is the
ultimate source of the idea of solidity; and indeed we do
naturally imagine that we feel the solidity of bodies, and
need only to touch an object to perceive its solidity. But this
is a layman’s way of thinking rather than a philosopher’s, as
will appear from the following ·two· reflections.
First, it is easy to observe that although bodies are felt
by means of their solidity, the feeling doesn’t resemble the
solidity. A man with no feeling in one hand has as perfect an
idea of impenetrability when he •sees that hand supported
by the table as when he •feels the table with the other hand.
An object pressing on any part of our bodies meets with resistance; and that resistance, through the motion it gives to the
nerves and animal spirits, conveys a certain sensation to the
mind; but it doesn’t follow that there are any resemblances
among the sensation, the motion, and the resistance.
Secondly, the impressions of touch are simple impressions (except with regard to their extent, which is irrelevant
to the present purpose); and from this simplicity I infer that
they don’t represent solidity or any real object. Consider
these two cases ·in which solidity is manifested·:
•A man presses a stone or other solid body with his
hand;

iv: Scepticism and other systems

•Two stones press each other.
You will agree that these two cases are not in every respect
alike, because the former involves not just solidity but also
a feeling or sensation that doesn’t appear in the latter. So
to bring out the likeness between these two cases alike we
must remove ·at least· some part of the impression that
the man feels by his hand; but a simple impression doesn’t
have parts, so we have to remove the whole impression;
which proves that this whole impression has no archetype or
model in external objects. To which we may add that solidity
necessarily involves •two bodies along with •contiguity [=
‘nextness’] and •impact; but that ·trio· is a compound object,
and can’t possibly be represented by a simple impression.
Not to mention the fact that though •solidity is always the
same, •tactual impressions keep changing, which is a clear
proof that •the latter are not representations of •the former.
Thus there is a direct and total opposition between our
reason and our senses; or, more properly speaking, between
the conclusions we form from cause and effect and those
that convince us of the continued and independent existence
of body. When we reason from cause and effect, we conclude
that neither colour, sound, taste, nor smell have a continued
and independent existence. When we exclude these perceptible qualities there is nothing left in the universe that does
have such an existence.

5: The immateriality of the soul
nature of the •mind (which we are apt to imagine so much
more obscure and uncertain). But in this we would be wrong.
The intellectual world, though involved in infinite obscurities,
is not tangled in contradictions such as we discovered in the
natural world. What is known about it is self-consistent, and

Having found such contradictions and difficulties in every system concerning •external objects, and in the idea
of •matter (which we imagine is so clear and determinate),
we would expect still greater difficulties and contradictions
in every hypothesis about our •internal perceptions, and the
120

Treatise, Book 1

David Hume

what is unknown we must be content to leave so.
Certain philosophers promise to lessen our ignorance if
we will listen to them, but I’m afraid that in doing so we
would risk running into contradictions from which the subject itself is free. These philosophers are the reasoners who
probe the question of whether the ‘substances’ in which they
suppose our perceptions to ‘inhere’ are material or immaterial. In order to put a stop to this endless point-scoring
on both sides, I know no better method than to ask these
philosophers ‘What do you mean by “substance” and by
“inhere”?’ It will be reasonable to enter seriously into the
dispute after they have answered this question, but not until
then.
We have found the question impossible to answer with
regard to matter and body; and when it comes to mind there
are all the same difficulties and some additional ones that
are special to that subject. As every idea is derived from a
preceding impression, if we had any idea of the substance
of our minds we must also have an impression of it; and
it is hard if not impossible to conceive what such an impression could be. For how can an impression •represent
a substance otherwise than by •resembling it? And how
can an impression resemble a substance, given that (according to the philosophy I am examining) it isn’t a substance
and has none of the special qualities or characteristics of
a substance? But leaving the question of what may or may
not be, and turning to the question of what actually is, I
ask the philosophers who claim that we have an idea of the
substance of our minds to point out the impression that
produces it, and say clearly how the impression operates
and from what object it is derived. Is it an impression of
sensation or of reflection? Is it pleasant, or painful, or neither? Do we have it all the time, or does it only return at
intervals? If at intervals, when does it principally return,

iv: Scepticism and other systems

and what causes produce it? If, instead of answering these
questions, anyone should evade the difficulty by saying that
the definition of ‘a substance’ is something that can exist by
itself, and that this definition ought to satisfy us, I would
reply that this definition fits everything that can possibly be
conceived, and can’t possibly serve to distinguish substance
from accident, or the soul from its perceptions. Here is why.
This is a principle:
Everything can be distinguished from everything else;
and if two things can be distinguished, they can be
separated by the imagination—·which is to say that
they can be conceived as separate from one another·.
Another principle that has been already acknowledged is
this:
Anything that is clearly conceived can exist, and anything that can be clearly conceived as being thus-andso can exist in that way—·for example, things that can
be conceived as existing separately from one another
can exist separately from one another·.
My conclusion from these two principles is that since all
our perceptions are different from each other, and from
everything else in the universe, they are also distinct and
separable, and may be considered ·or conceived· as separately existent, and ·therefore· can exist separately and have
no need of anything else to support their existence. So they
are substances according to this definition.
So we can’t arrive at any satisfactory notion of substance,
whether by looking for an originating impression or by means
of a definition; and that seems to me a sufficient reason for
abandoning utterly the dispute about whether the soul is
material or not, and makes me absolutely condemn the very
question. We have no perfect idea of anything except perceptions. A substance is entirely different from a perception.
So we have no idea of a substance. It is thought ·by some
121

Treatise, Book 1

David Hume

philosophers· that our perceptions can exist only if they ‘inhere in’ something that supports them; but nothing seems
to be needed to support the existence of a perception. So
we have no idea of ‘inhesion’. That being the case, how can
we possibly answer the question ‘Do perceptions inhere in a
material substance or in immaterial one?’ when we don’t so
much as understand the meaning of the question?

·T HE

iv: Scepticism and other systems

stance of the soul, but only the question about its being in
the same place as matter; so it may be worthwhile to consider in general what objects are capable of being in places
and what ones are not. This is an interesting and challenging question, which may lead us to some discoveries of
considerable importance.
Our first •notion of space and extension is derived solely
from the senses of sight and touch; only things that are
coloured or tangible can have parts that are arranged in
such a way as to convey •that idea. ·You might say that
a taste has parts, because it can be lessened or increased;
but· increasing or lessening a taste is not like lessening or
increasing a visible object. ·Again, you might say that we
experience distance—and thus extension—through the sense
of hearing; but· when several sounds strike our hearing at
once, it is only through custom and reflection that we form
an idea of spatial relations among the bodies from which the
sounds are derived. Anything that exists somewhere must
either •be extended or •be a mathematical point having no
parts or inner complexity. Something extended must have a
particular shape—square, round, triangular—none of which
can be true of a desire, or indeed of any impression or idea
except ones belonging to sight and touch. And although a
desire is indivisible, it oughtn’t to be considered as a mathematical point. If it were one, it could be arranged along
with three or four other desires in such a way as to make
a complex with a determinate length, width, and thickness;
which is obviously absurd.
In the light of these remarks, you won’t be surprised
when I affirm something that is condemned by many metaphysicians, and regarded as contrary to the most certain
principles of human reason. It is that an object can exist,
and yet be nowhere. And I assert that this is not only possible but that most existing things do and ·indeed· must exist

LOCATION OF OUR PERCEPTIONS ·

One argument that is commonly employed for the immateriality of the soul seems to me remarkable:
Whatever is extended consists of parts; and whatever
consists of parts can be divided, if not in reality then
at least in the imagination. But something that is
divisible can’t possibly be conjoined to a thought or
perception, which is altogether indivisible. If such a
conjunction did occur, would the indivisible thought
exist on the left or on the right side of this extended
divisible body? On the surface or in the middle? On
the back or on the front side of it? ·If you aren’t convinced by those rhetorical questions, consider instead
this sober argument·. If the thought or perception
is conjoined with something extended, it must exist
somewhere within that thing’s boundaries—either •in
one particular part or •in every part. In •the former
case, that particular part is indivisible, and the perception is conjoined only with it and not with the
extended thing; and in •the latter case, the thought
must also be extended and separable and divisible,
just as the body is, which is utterly absurd and contradictory. Can anyone conceive a passion that is a
yard long, a foot wide, and an inch thick? So thought
and extension are wholly incompatible qualities, and
can never come together in one subject.
This argument doesn’t bear on the question about the sub122

Treatise, Book 1

David Hume

in that way. An object can be said to ‘be nowhere’ when
•its parts are not related to one another in such a way as
to form any shape or size, and •it as a whole isn’t related
to other bodies in such a way as to fit our notions of closeness or distance. Now this is obviously the case with all our
perceptions and objects except those of the sight and touch.
•A smell or a sound can’t be either circular or square; •a
moral reflection can’t be situated to the right or to the left of
a passion. These objects and perceptions, so far from requiring any particular place, are absolutely incompatible with
it; we can’t even imagine their being located. . . . Perceptions
that have no parts and exist nowhere cannot be spatially
conjoined with matter or body—i.e. with something extended
and divisible —because any relation has to be based on some
common quality. But there is no need for me now to press
this argument. It may be better worth our while to remark
that this question of the placing of objects comes up not
only in metaphysical disputes about the nature of the soul
but even in everyday life. Consider a fig at one end of the
table and an olive at the other: when we form the complex
ideas of these substances, one of the most obvious is that of
their different tastes, and clearly we incorporate and conjoin
these qualities with ones that are coloured and tangible. The
bitter taste of one and sweet taste of the other are supposed
to lie in the visible bodies and ·thus· to be separated from
each other by the whole length of the table. This illusion is
so remarkable and ·yet· so natural that it may be proper to
consider its causes.

iv: Scepticism and other systems

And it’s not just that they coexist in some general way—
·their coexistence exhibits two relations that we have seen
to have a powerful effect on our minds·. The taste •appears
in the mind at the same time as the smell; and it is when
the extended body comes within reach of our senses that we
perceive its particular taste and smell—·so we naturally infer
that the body •causes the taste and smell·. So we have the
relations of •causation and •contiguity in the time of their
appearance between the extended object and the quality that
exists nowhere; and this must have such an effect on the
mind that when one of the related items appears the mind
will immediately turn to the conception of the other. And
this is not all. As well as turning our thought from one to
the other on account of their relation, we try to give them a
further relation—namely, being in the same place—so as to
make the transition more easy and natural. For it is a quality
in human nature that I shall often have occasion to mention,
and shall explain more fully in its proper place, that when
objects are united by some relation we are strongly disposed
to add some further relation to them in order to complete
their union. . . . But whatever confused notions we may form
of a union in place between (say) a fig and its particular
taste, when we think about it we have to see that there is
something altogether unintelligible and contradictory about
such a union. Let us ask ourselves one obvious question:
The taste that we conceive to be contained within the
boundary of the fig—is it in every part of the fig, or in
only one part?
Faced with this, we must quickly find ourselves at a loss, and
see the impossibility of ever giving a satisfactory answer. We
can’t reply that it is only in one part, for experience convinces
us that every part has the same taste. And it’s no better to
reply that it exists in every part, for then we must suppose
the taste to have shape and size, which is absurd and incom-

Although things that exist without any place or extension
can’t be •joined in space by something extended, they can
enter into many •other relations. Thus the taste and smell
of a piece of fruit are inseparable from its other qualities of
colour and tangibility; and whichever of them is the cause
and whichever the effect, they certainly always exist together.
123

Treatise, Book 1

David Hume

prehensible. So here we are pulled in opposite directions by
two forces—•the inclination of our imagination, which makes
us incorporate the taste into the fig, and •our reason, which
shows us the impossibility of such a union. Being divided
between these opposing pulls, we don’t renounce either of
them, but instead involve the subject in so much confusion
and obscurity that we no longer see the opposition. We suppose that the taste exists within the boundary of the fig, but
in such a way that it •fills the whole thing without being
extended, and •exists complete in every part of it without
being divided! In short, in our most ordinary everyday way of
thinking we use a principle of the Aristotelian philosophers
which seems shocking when it is expressed crudely: totum in
toto, et totum in qualibet parte—which is about the same as
saying that a thing is in a certain place and yet is not there.

iv: Scepticism and other systems

and that a certain number of smells can combine with a
certain number of sounds to compose a body of twelve cubic
inches; the mere mention of which shows it to be ridiculous.
But though in this view of things we can’t refuse to condemn the materialists, who conjoin all thought with •an extended body, a little thought will show us an equally strong
reason for blaming their opponents, who conjoin all thought
with •a simple and indivisible substance. The plainest and
most down-to-earth philosophy informs us that an external
object can’t make itself known to the mind immediately; it
has to appear through the interposition of an image or perception. The table that appears to me right now is only a
perception, and all its qualities are qualities of a perception.
Now, the most obvious of all its qualities is extendedness.
The perception consists of parts. These parts are arranged
in such a way as to give us the notion of distance and closeness, of length, width, and thickness. The termini of these
three dimensions create what we call shape. This shape is
movable, separate, and divisible. Mobility and divisibility are
the distinguishing properties of extended objects. And to cut
short all disputes, the very idea of extendedness is copied
from nothing but an impression, with which it must therefore perfectly agree. To say that the idea of extension ‘agrees
with’ something is to say that the ‘something’ is extended.
The ·materialist· free-thinker can now have his turn to
triumph. Having found that some impressions and ideas
are really extended, he can ask his opponents ‘How can you
bring a simple and indivisible subject together with an extended perception?’ All the arguments of the theologians can
here be turned back against them. ·They have demanded
of the materialist ‘Is the unextended perception on the lefthand or the right-hand part of the extended body?’, but now
the materialist can demand·: ‘Is the unextended subject (or
immaterial substance, if you like) on the left-hand or the

[The Latin means, literally, ‘The whole in the whole, and the whole in
each part’.] All this absurdity comes from our trying to assign

a place to something that is utterly incapable of it; and that
attempt comes from our inclination to complete a union that
is based on causation and contiguity of time, by crediting the
objects with being in the same place. But if reason is ever
strong enough to overcome prejudice, it must surely prevail
here. For here are our only choices ·regarding such items as
passions and tastes and smells·:
•They exist without being in any place. •They have
shapes and sizes. •They are incorporated with extended objects, and then the whole is in the whole
and the whole is in every part.
The absurdity of the second and third suppositions proves
sufficiently the truth of the first. And there is no fourth
opinion. What about the supposition that these items exist
in the way mathematical points do? That ·isn’t a genuine
fourth option, because· it boils down to the second opinion:
it supposes that various passions may be placed in a circle,
124

Treatise, Book 1

David Hume

right hand part of the ·extended· perception? Is it in this
particular part, or in that other? Is it in every part without
being extended? Or is it complete in any one part without
deserting the rest?’ It is impossible to give to these questions any answer that won’t both •be absurd in itself and
•·be available (if it weren’t absurd) for the materialists to use
for their purposes, that is, to· account for the union of our
unextended perceptions with an extended substance.

·T HE

iv: Scepticism and other systems

the one substance·. All the passions of the soul, all the
configurations of matter however different and various, inhere in the same substance ; they can be distinguished from
one another, without their distinctness bringing it about that
they inhere in distinct substances. The same substratum [=
‘underlay’], if I may so speak, supports the most different qualities without any difference in itself, and varies them without
itself varying. Neither time, nor place, nor all the diversity of
Nature are able to produce any composition or change in the
perfect simplicity and identity of the one substance.
This brief exposition of the principles of that famous atheist will, I think, be sufficient for the present purpose. Without
our having to enter further into these gloomy and obscure regions, I shall be able to show that •this hideous hypothesis of
Spinoza’s is almost the same as •the doctrine of the immateriality of the soul, which has become so popular. To make this
evident, let us remember (from 6ii ) that because every idea is
derived from a preceding perception, it follows that ·we can’t
have an idea of something that it is radically different in kind
from a perception; from which it follows in turn that· our
idea of an externally existing object can’t possibly represent
anything radically different in kind from every perception.
Whatever difference we may suppose between perceptions
and external objects, it is still incomprehensible to us; and
we are obliged either to make external objects the same as
perceptions or to conceive an external object merely as a
relation without a relative—·that is, to conceive it emptily as
whatever-it-is-that- some-perceptions-are-perceptions-of ·.
The conclusion I shall draw from this may at first sight
appear to be a cheat, but a very little thought will show it to
solid and satisfactory. I start with this:
We can suppose there to be a radical difference in kind
between an object and an impression, but we cannot
conceive such a difference; so when we reach any con-

SUBSTANTIAL UNDERLAY OF OUR PERCEPTIONS —

(S PINOZA )·
This is my opportunity to take up again the question
about the substance of the soul. Though I have condemned
that question as utterly unintelligible, I can’t refrain from
offering some further reflections on it. I assert this:
The doctrine of a thinking substance that is immaterial, simple and indivisible is a true atheism. From it
we can infer all the ·atheistic· views for which Spinoza
is so universally infamous
. From this line of thought I hope at least to reap one
advantage, that my adversaries won’t have any excuse for
rendering my doctrine odious by accusations that can be so
easily turned back against them. The fundamental principle
of Spinoza’s atheism is the doctrine of the simplicity of the
universe—·that is, the universe’s not having parts·—and the
unity of the substance in which he supposes both thought
and matter to inhere. There is only one substance in the
world, says Spinoza, and that substance is perfectly simple
and indivisible, and doesn’t have any particular position because it exists everywhere. Whatever we discover externally
by sensation, whatever we feel internally by reflection —all
these are nothing but qualities of that one simple and necessarily existent being, and don’t have any separate or distinct
existence. ·This table and that chair are not two distinct
things, they are just two qualities of the one and only thing—
125

Treatise, Book 1

David Hume

iv: Scepticism and other systems

After this I consider •the other system of beings, namely the
universe of thought, or of my impressions and ideas. There
I observe another sun, moon, and stars, an earth and seas,
covered and inhabited by plants and animals; towns, houses,
mountains, rivers—and in short everything I can discover
or conceive in the first system. When I ask about these,
theologians present themselves and tell me that
these also are qualities, and indeed qualities of one
simple, uncompounded, and indivisible substance.
Then I am deafened by the noise of a hundred voices that
treat Spinoza’s hypothesis with detestation and scorn, and
the theologians’ view with applause and veneration! I look
into these hypotheses to see what may be the reason for
such a strong preference for one of them, and I find that
•they share the fault of being unintelligible, and that •as
far as we can understand them they are so much alike
that we can’t find any absurdity in one that isn’t shared
by the other. Because all our ideas are derived from our
impressions, we have no idea of a quality in an object that
doesn’t match and can’t represent a quality in an impression.
So if we can ·against Spinoza· find a conflict between an
extended object as a quality and something simple and uncompounded which is the substance in which it inheres, then
there must (·against the theologians·) be the same conflict
between the perception or impression of an extended object
and something simple and uncompounded which is the substance ·in which it inheres·. Every idea of a quality in an
object passes through an impression, so every perceivable
relation, whether of connection or incompatibility, must be
common to both objects and impressions.
Looked at in a general way, this argument seems obvious
beyond all doubt and contradiction. Still, to make it clearer
and more intuitive, let us go through it in detail, and see
whether all the absurdities that have been found in Spinoza’s

clusion about impressions that are inter-connected
or incompatible we shan’t know for certain that it
will apply also to objects; but any such conclusion
that we form about objects will certainly apply also to
impressions.
The reason is not difficult. An object is supposed to be different from an impression; so if in our reasoning we start
with the impression, we can’t be sure that the details ·of the
impression· that we are going by are shared by the object;
it may ·for all we know· be that the object differs from the
impression in that respect. But ·the converse doesn’t hold·:
our reasoning, if it starts with the object, certainly must
hold also for the impression. Why? Because the quality of
the object on which the reasoning is based must at least be
conceived by the mind (·otherwise it couldn’t be reasoned
about·), and it couldn’t be conceived unless it were a quality also possessed by an impression, because all our ideas
are derived from impressions. So we can lay it down as a
certain maxim that we can never . . . . discover a connection
or incompatibility between objects that doesn’t hold also
for impressions; though the converse proposition—that all
the discoverable relations between impressions hold also for
objects—may not be equally true, .
Let us now apply this to the present case. I am presented
with two different systems of existing things for which—I
am supposing ·for purposes of argument·—I have to assign
some substance or ground of inhesion. •I observe first the
universe of objects or of bodies—the sun, moon, and stars,
the earth, seas, plants, animals, men, ships, houses, and
other productions of art or of nature. Here Spinoza appears,
and tells me that
these are only qualities, and the subject in which they
inhere—·the substance that has them·—is simple, uncompounded, and indivisible.
126

Treatise, Book 1

David Hume

system may not also be found in that of the theologians.
First, this has been said against Spinoza:
Because according to Spinoza a mode [= ‘quality’] is not
a distinct or separate existent—·something over and
above the one substance·—it must be its substance.
So the extended universe, which is supposed to inhere
·as a mode or quality· in a simple, uncompounded
substance, must be in a manner identified with that
substance. But this is utterly impossible and inconceivable, unless the indivisible substance expands so
as to correspond to the extended world, or the extended world contracts so as to match the indivisible
substance.
This argument (·against Spinoza·) seems sound, as far as we
can understand it; and it is clear that with some change in
the wording it applies equally (·against the theologians·) to
our extended perceptions and the simple substance of the
soul. For the ideas of objects and of perceptions are in every
respect the same, except for the supposition of a difference
that is unknown and incomprehensible.
Secondly, it has been said ·against Spinoza· that
we have no idea of substance that isn’t applicable to
matter, and no idea of a distinct substance that isn’t
applicable to every distinct portion of matter. So matter
is not a mode ·or quality· but a substance, and each
part of matter is not a distinct mode but a distinct
substance.
I have already proved that we have no perfect idea of substance, but that taking ‘substance’ to mean ‘something that
can exist by itself’ it is obvious that every perception is a
substance and every distinct part of a perception is a distinct
substance. So in this respect each hypothesis labours under
the same difficulties as does the other.
Thirdly, it has been objected to the system of one simple

iv: Scepticism and other systems

substance in the universe that
this substance, being the support or substratum of
everything, must at the very same instant be modified into forms that are contrary and incompatible.
The round and square figures are incompatible in the
same substance at the same time. How then is it
possible for one substance to be modified into that
square table and into this round one?
I ask the same question about the impressions of these tables, and I find that the answer is no more satisfactory in
one case than in the other. ·So any embarrassment for
Spinoza along these lines is equally an embarrassment for
the theologians·.
It appears, then, that whichever way we turn the same
difficulties follow us, and that we can’t advance one step
towards the establishing the simplicity and immateriality
of the soul without preparing the way for a dangerous and
incurable atheism. The situation is the same if, instead of
calling thought a modification ·or quality· of the soul, we
give it the more ancient and yet more fashionable name of
‘action’. By an action we mean much the same thing as what
is commonly called an ‘abstract mode’— that is, something
that strictly speaking isn’t distinguishable or separable from
its substance, and is conceived only through a distinction
of reason, that is, an abstraction. ·For example, a dance is
not distinguishable or separable from the dancer, but from
the totality that is the dancer we abstract one aspect, which
we call her dance ·. But nothing is gained by this switch
from ‘modification’ to ‘action’: it doesn’t free us from a single
difficulty. . . . [Hume explains and defends this claim in two
paragraphs which are not included here.]

·T HE

CAUSE OF OUR PERCEPTIONS ·

From these hypotheses about the •location and the
•substance of our perceptions, let us pass to another that is
127

Treatise, Book 1

David Hume

more intelligible than •the latter and more important than
•the former, namely concerning the cause of our perceptions.
The Aristotelians say this:
Matter and motion, however varied, are still ·only·
matter and motion, and cause only differences in
where bodies are and how they are oriented. Divide
a body as often as you please, it is still body. Give it
any shape and nothing will result but shape (which
is the relation of parts). Move it in any way and all
you will get is motion (which is a change of relation
·to other bodies·). It is absurd to imagine that motion in a circle should be merely •motion in a circle
while motion in an ellipse should also be •a passion or
moral reflection; or that the collision of two spherical
particles should become •a sensation of pain while
the collisions of two triangular ones yields •pleasure.
Now, as these different collisions and variations and
mixtures are the only changes of which matter is capable, and as they never give us any idea of thought
or perception, it follows that thought cannot possibly
be caused by matter.
Few have been able to resist the seeming force of this argument, yet nothing in the world is easier than to refute
it! We need only reflect on what I have proved in general,
namely we never sense any connection between causes and
effects, and that it is only through our experience of their
constant conjunction that we can arrive at any knowledge of
the causal relation. Now,

iv: Scepticism and other systems

duce anything, and we shall never discover a reason
why any object may or may not be the cause of any
other, however alike or unalike they may be.
This obviously destroys the foregoing reasoning about the
cause of thought or perception. For though no connection
between motion or thought appears to us, neither does any
connection between any other causes and effects. Place one
body of a pound weight on one end of a lever, and another
body of the same weight on another end; you will never find
in these bodies any •movement-force that depends on their
distances from the centre, any more than a •force of thought
and perception. So if you claim to prove a priori that •a
position of bodies can never cause thought because, turn it
which way you will, it is nothing but a position of bodies, you
must by the same line of reasoning conclude that •a position
of bodies can never produce motion, since there is no more
apparent connection in that case than in the other. But the
latter conclusion is contrary to evident experience, ·which
shows that how a body moves may depend on how it is situated·; and we could come to have similar experiences in the
operations of the mind, perceiving a constant conjunction of
thought with motion. So you reason too hastily when you
conclude, from merely attending to the ideas, that motion
cannot possibly produce thought and that a different position of parts cannot produce a different passion or reflection.
Indeed, it is not only possible for us to have such an experience, but it is certain that we do have it, for everyone can
perceive that the different dispositions of his body change his
thoughts and sentiments. You might say: ’This ·is a special
case, because it· depends on the union of soul and body.’ To
that I reply that we must separate the question about •the
substance of the mind from the one about •the cause of its
thought; and that if we take the latter question on its own,
we find by comparing their ideas that thought and

•no two real objects are contrary to one another, and
•objects that are not contrary are capable of being

constantly conjoined,
and from these two principles I have inferred in 15iii that
•to consider the matter a priori, anything could pro128

Treatise, Book 1

David Hume

iv: Scepticism and other systems

•Supposing that God were the great and effective force that

motion are different from each other,

makes up for what is lacking in all ·other· causes, this leads
us into the grossest impieties and absurdities. It involves
having recourse to God in natural operations, saying that
matter can’t of itself communicate motion or produce thought
because matter has no perceptible connection with motion
or thought, ·so that when matter seems to cause something
it is really God at work·; and I say that on this account we
must acknowledge that God is the author of all our volitions
and perceptions, for they also have no perceptible connection
with one another or with the supposed but unknown substance of the soul. Father Malebranche and other Cartesians
have taken this view of all the actions of the mind, except for
volition, or rather an inconsiderable part of volition—though
it’s easy to see that this exception is a mere dodge to avoid
the dangerous consequences of their doctrine. If nothing is
active except what has a perceptible apparent power, thought
is never any more active than matter; and if this inactivity
must make us fall back on a Deity ·to explain what seem to
be cause-effect relations·, God is the real cause of all our
actions, bad as well as good, vicious as well as virtuous.
So we are necessarily brought to the other horn of the
dilemma, namely that all objects that are found to be constantly conjoined are—for that reason and only for that
reason—to be regarded as causes and effects. Now, as all
objects that are not contrary are capable of being constantly
conjoined, and as no real objects are contrary, it follows that
(for all we can tell by considering the mere ideas of things)
anything could be the cause or effect of anything; which
obviously gives the advantage to the materialists ·who let
matter do all the causing· over their antagonists ·who say
that God must be brought into the causal story·.
The final verdict, then, must be this: •the question concerning the substance of the soul is absolutely unintelligible;

and
we find by experience that thought and motion are
constantly united.
Such constant uniting is all we demand for the causal relation when we are considering the effects of matter on matter;
so we can confidently conclude that motion can be and actually is the cause of thought and perception.
We seem to be left with a dilemma. Either •nothing can be
the cause of something else unless the mind can perceive a
connection between the ideas of the two items, or •all objects
that we find constantly conjoined are on that account to be
regarded as causes and effects. If we choose the first horn
of the dilemma, the consequences are as follows. First, we
are really saying that there is no such thing in the universe
as a cause or productive force, not even God himself, since
our idea of that Supreme Being is derived from particular
impressions, none of the ideas of which have any perceptible
connection with ·the idea of· any other existent. You may
object: ‘The connection between the idea of an infinitely powerful being and that of any effect that he wills is necessary
and unavoidable.’ To this I make two replies. •We have no
idea of a being endowed with any power, much less of one
endowed with infinite power. And if ·in order to avoid this
point· you seek to define ‘power’, you will have to do it in
terms of ‘connection’; and then in saying that
the idea of an infinitely powerful being is connected
with that of every effect that he wills
you are really saying only that
a being whose volition is connected with every effect
is connected with every effect;
which is an identical proposition—·a tautology·—and gives
us no insight into the nature of this power or connection.
129

Treatise, Book 1

David Hume

•some of our perceptions are unextended, so they can’t all be

iv: Scepticism and other systems

to religion, I hope the following explanation will remove his
worries.
There is no foundation for any a priori conclusion about
either the •operations or the •duration of any object that the
human mind can conceive. Any object can be imagined suddenly to become entirely •inactive, or to be •annihilated, and
it is an obvious principle that whatever we can imagine is
possible. Now this is no more true of matter than of mind; no
more true of an extended compounded substance than of a
simple and unextended one. In both cases the •metaphysical
arguments for the immortality of the soul are equally inconclusive; and in both cases the •moral arguments and those
derived from the analogy of Nature are equally strong and
convincing. If my philosophy doesn’t add to the arguments
for religion, I have at least the satisfaction of thinking that it
doesn’t take anything from them either. Everything remains
precisely as before.

located in the same place as something extended, and some
of them are extended, so they can’t all be co-located with
something unextended; and as the constant conjunction
of objects constitutes the very essence of cause and effect,
matter and motion can often be regarded as the causes of
thought, as far as we have any notion of the causal relation.
Philosophy’s sovereign authority ought to be acknowledged everywhere; so it is a kind of indignity to oblige her
on every occasion to apologize for her conclusions and justify herself to every particular art and science that may be
offended by her. It’s like a king being arraigned for high
treason against his subjects! The only occasion when philosophy will think it necessary and even honourable to justify
herself is when religion may seem to be in the least offended;
for the rights of religion are as dear to philosophy as her
own, and are indeed the same. So if anyone imagines that
the arguments I have presented are in any way dangerous

6: Personal identity
Some philosophers believe this:

cause no fact we could use as evidence is as intimately
present to our consciousness as is the existence of
our self. If we doubt the latter, we can’t be certain of
anything.

We are every moment intimately conscious of what we
call our self ; we feel its existence and its continuing
to exist, and are certain—more even than any demonstration could make us—both of its perfect identity
and of its simplicity. The strongest sensations and
most violent emotions, instead of distracting us from
this view ·of our self·, only focus it all the more intensely, making us think about how these sensations
and emotions affect our self by bringing it pain or
pleasure. To offer further evidence of the existence of
one’s self would make it less evident, not more, be-

Unfortunately, all these forthright assertions are in conflict
with the very experience that is supposed to support them.
We don’t so much as have an idea of self of the kind that
is here described. From what impression could this idea be
derived? This question can’t be answered without obvious
contradiction and absurdity; yet it must be answered if the
idea of self is to qualify as clear and intelligible. Every real
idea must arise from some one impression. But self or per130

Treatise, Book 1

David Hume

son is not any one impression, but is rather that to which
all our many impressions and ideas are supposed to be related. If the idea of self came from an impression, it would
have to be an impression that remained invariably the same
throughout our lives, because the self is supposed to exist
in that way. But no impression is constant and invariable.
Pain and pleasure, grief and joy, passions and sensations
follow one other and never all exist at the same time. So it
can’t be from any of these impressions or from any other
that the idea of self is derived. So there is no such idea.
Furthermore, if we retain this hypothesis about the self,
what are we to say about all our particular perceptions?
They are all different, distinguishable, and separable from
one other—they can be separately thought about, and can
exist separately—with no need for anything to support their
existence. In what way do they belong to self? How are they
connected with it? For my part, when I look inward at what I
call myself, I always stumble on some particular perception
of heat or cold, light or shade, love or hatred, pain or pleasure, or the like. I never catch myself without a perception,
and never observe anything but the perception. When I am
without perceptions for a while, as in sound sleep, for that
period I am not aware of myself and can truly be said not
to exist. If all my perceptions were removed by death, and
I could not think, feel, see, love or hate after my body had
decayed, I would be entirely annihilated—I cannot see that
anything more would be needed to turn me into nothing. If
anyone seriously and thoughtfully claims to have a different
notion of himself, I can’t reason with him any longer. I have
to admit that he may be right about himself, as I am about
myself. He may perceive something simple and continued
that he calls himself, though I am certain there is no such
thing in me.
But setting aside metaphysicians of this kind, I am willing

iv: Scepticism and other systems

to affirm of the rest of mankind that each of us is nothing
but a bundle or collection of different perceptions that follow
each other enormously quickly and are in a perpetual flux
and movement. Our eyes can’t turn in their sockets without
varying our perceptions; our thought is even more variable
than our sight; and all our other senses and faculties contribute to this change in our perceptions, with no one of
them remaining unaltered for a moment. The mind is a kind
of stage on which many perceptions successively make their
appearance: they pass back and forth, glide away, and mingle in an infinite variety of positions and situations. Strictly
speaking, there is no •simplicity in the mind at one time and
no •identity through different times, no matter what natural
inclination we may have to imagine that simplicity and identity. ·That is to say: It is not strictly true that •when a blue
colour is seen and a whistling sound heard at the same time,
one single unified mind has both these perceptions; nor is
it strictly true that •the mind that has a certain perception
at one time is the very same mind that has a perception
at another time·. The ‘stage’ comparison must not mislead
us. What constitutes the mind is just the successive perceptions; we haven’t the faintest conception of the place where
these scenes are represented or of the materials of which it
is composed.
What, then, makes us so inclined to ascribe an identity to
these successive perceptions, and to suppose that we have
an invariable and uninterrupted existence through the whole
course of our lives? To answer this question we must distinguish what we think and imagine about personal identity
from the role of personal identity in our emotions and desires.
The former is our present subject. To explain it perfectly we
must dig fairly deep: first we must account for the identity
that we attribute to plants and animals, because there is
a great analogy between that and the identity of a self or
131

Treatise, Book 1

David Hume

iv: Scepticism and other systems

we think of it as (1) variable or interrupted, ·which it is·,
yet the very next moment we ·wrongly· think of it as (2) a
single, identical, unchanging and uninterrupted thing. ·That
completes the explanation·. The resemblance that I have
mentioned ·between the two acts of the mind· gives us such
a strong tendency to make this mistake that we make it
without being aware of what we are doing; and though we
repeatedly correct ourselves and return to a more accurate
and philosophical way of thinking, we can’t keep this up
for long, and we fall back once more into the mistake. Our
only way out ·of this oscillation between truth and error· is
to give in to the error and boldly assert that these different
related objects are really the same, even though they are interrupted and variable. To justify this absurdity to ourselves,
we often feign [= ‘create a fiction of’] some new and unintelligible
thing that connects the objects together and prevents them
from being interrupted and variable. The perceptions of our
senses are intermittent—·there are gaps between them·—but
we disguise this by feigning that they exist continuously; and
they vary, but we disguise this by bringing in the notion of
a soul or self or substance ·which stays the same under all
the variation·. Even in contexts where we don’t indulge in
such fictions, we are so strongly inclined to confuse identity
with relatedness that we are apt to imagine something unknown and mysterious connecting the parts, other than the
relations between them; and this is what I think happens
when we ascribe identity to plants. When even this ·kind of
fiction-making· doesn’t take place, we still feel impelled to
confuse these ideas with one another, though we can’t give a
satisfactory account of what we are doing or find anything
invariable and uninterrupted to justify our notion of identity.

person.
We have a clear idea of an object that remains invariable
and uninterrupted while time supposedly passes. We call
this the idea of identity or sameness. We have also a clear
idea of many different objects existing successively in a close
relation to one another; and this, properly understood, is
just as good an example of diversity as it would be if the
objects were not related to one another in any way. ·As the
sand runs in the hour-glass, this grain is distinct from that
one that falls a tenth of a second later and a micromillimetre
behind; they are diverse from one another, which is simply
to say that they are two grains, not one; and the fact that
they are closely related to one another (in space, in time,
and in being alike) makes no difference to that. They are as
distinct from one another—they are as clearly two—as the
Taj Mahal and the Grand Canyon·. But though these two
ideas of identity and a sequence of related objects are perfectly distinct from one another and even contrary, yet in our
everyday thinking they are often confused with one another,
treated as though they were the same. ·I now explain what
leads us into that confusion·. Here are two mental activities:
(1) thinking about a sequence of related objects, and
(2) thinking about one uninterrupted and invariable object.
Although these are distinct, and involve different activities
of the imagination, they feel the same. The activity in (1)
doesn’t require much more effort than the activity in (2): in
(1) the relation between the objects helps the mind to move
easily from one to the next, making its mental journey as
smooth as if it were contemplating one continued object as
in (2). This resemblance between these two kinds of thought
generates the confusion in which we mistakenly substitute
the notion of (2) identity for that of (1) related objects. When
contemplating a sequence of related objects, at one moment

Thus the controversy about identity is not a merely verbal dispute. For when we attribute identity in an improper
sense to variable or interrupted objects, we are not just
132

Treatise, Book 1

David Hume

using words wrongly but are engaging in a fiction, a false
thought, either of something •invariable and uninterrupted
or of something •mysterious and inexplicable. To convince
a fair-minded person that this is so, we need only to show
him through his own daily experience that when variable or
interrupted objects are supposed to continue the same, they
really consist only in a sequence of parts, connected together
by resemblance, contiguity [= ‘nextness’], or causation. Such
a sequence obviously fits our notion of diversity, so it can
only be by mistake that we attribute an identity to it; and
this mistake must arise from the fact that when the imagination moves from one of the related parts to the next, this
act of the mind resembles the act in which we contemplate
one continued object. What I mainly have to prove, then, is
that whenever we ascribe identity to something that we do
not observe to be unchanging and uninterrupted, what we
are really talking about is ·not a single object, but rather· a
sequence of related objects.
To get started on this, suppose we have in front of us a
mass of matter whose parts are contiguous and connected;
clearly we have to attribute a perfect identity to this mass
so long as it continues uninterruptedly to contain the very
same parts, even if those parts move around within it. Now
suppose that some very small or inconspicuous part is added
to the mass or removed from it. Strictly speaking, it is no
longer the same mass of matter; but we—not being accustomed to think so accurately—don’t hesitate to say that a
mass of matter is still ‘the same’ if it changes only in such
a trivial way. Our thought moves from the object before
the change to the object after it so smoothly and easily that
we are hardly aware that there is any movement; and this
tempts us to think that it is nothing but a continued survey
of the same object.
One aspect of this phenomenon is well worth noticing.

iv: Scepticism and other systems

Although a turnover in any large part of a mass of matter destroys the identity of the whole, ·that is, makes us unwilling
to say that it continues to be the same thing·, what we count
as large in this context depends not on the actual size of the
part but rather on how big a proportion it is of the whole. We
would count a planet as still ‘the same’ if it acquired or lost
a mountain, but the change of a few inches could destroy
the identity of some bodies. The only way to explain this
is by supposing that objects interrupt the continuity of the
mind’s actions not according to their real size but according
to their proportion to each other; and therefore, since this
interruption makes an object cease to appear ‘the same’, it
must be the uninterrupted movement of the thought that
constitutes the imperfect identity, ·that is, that leads us to
say that something is ‘the same’ when, strictly speaking, it
is not the same·.
This is confirmed by another phenomenon. Although a
change in any considerable part of a body destroys its identity, if the change is produced gradually and imperceptibly
we are less apt to see it as destroying the identity. The reason
for this must be that the mind, in following the successive
changes of the body, slides easily along from surveying its
condition at one moment to surveying it at another, and is
never aware of any interruption in its actions.
However careful we are to introduce changes gradually
and to make each a small proportion of the whole, when
eventually they add up to a considerable change we hesitate to attribute identity to such different objects. But we
have a device through which we can induce the imagination
to go one step further ·in attributing identity where really
there is none·—namely, relating the parts to one another
through some common end or purpose. A ship of which a
considerable part has been changed by frequent repairs is
still considered ‘the same’ even if the materials of which it
133

Treatise, Book 1

David Hume

is composed have come to be quite different. Through all
the variations of the parts, they still serve the same common
purpose; and that makes it easy for the imagination to move
from the ship before the repairs to the ship after.
This happens even more strikingly when we see the parts
as being causally related to one another in everything they
do, in ways that reflect their common end. This ·is not the
case with ships, but it· is the case with all animals and vegetables: not only are the parts taken to have some over-all
purpose, but also they depend on and are connected with
one another ·in ways that further that purpose·. The effect of
this relation is that, although in a very few years both plants
and animals go through a total change, with their form, size
and substance being entirely altered, yet we still attribute
identity to them. An oak that grows from a small plant to a
large tree is still the same oak, ·we say·, though there is not
one particle of matter or shape of its parts that is the same.
An infant becomes a man, and is sometimes fat, sometimes
thin, without any change in his identity.
We should also consider two further noteworthy facts.
The first is that though we can usually distinguish quite exactly between numerical and specific identity, yet sometimes
we mix them up and use one in place of the other in our
thinking and reasoning. [Numerical identity is real identity, or being

iv: Scepticism and other systems

architectural style. Here neither the form nor the materials
are the same; the buildings have nothing in common except
their relation to the inhabitants of the parish; and yet this
alone is enough to make us call them ‘the same’. It is relevant that in these cases ·of the noises and the churches·
the first object is in a manner annihilated before the second
comes into existence. That protects us from being presented
at any one time with the idea of difference and multiplicity;
·that is, we are not in a position to pick out both noises
(or both churches) at the same time, and have the thought
‘This is one and that is another’·; and that increases our
willingness to call them ‘the same’.
Secondly, although in general we don’t attribute identity
across a sequence of related objects unless the change of
parts is gradual and only partial, with objects that are by
nature changing and inconstant we will say they are ‘the
same’ even if the changes are quite sudden. For example, the
nature of a river consists in the motion and change of parts,
so that there is a total turnover of these in less than twentyfour hours, but this does not stop the river from being ‘the
same’ for centuries. What is natural and essential to a thing
is expected, and what is expected makes less impression and
appears less significant than what is unusual and extraordinary. A big change of an expected kind looks smaller to the
imagination than the most trivial unexpected alteration; and
by making less of a break in the continuity of the thought it
has less influence in destroying the ·supposition of· identity.
I now proceed to explain the nature of personal identity,
which has become such a great issue in philosophy. The line
of reasoning that has so successfully explained the identity
of plants and animals, of ships and houses, and of all changing complex things—natural and artificial—must be applied
to personal identity too. The identity that we ascribe to the
mind of man is fictitious; it is like the identity we ascribe to

the very same thing. It is called ‘numerical’ because it affects counting:
if x is not numerically identical with y, then x and y are two. By ‘specific
identity’ Hume means similarity, qualitative likeness, being of the same

Thus, a man who hears a noise that is
frequently interrupted and renewed says it is still ‘the same
noise’, though clearly the sounds have only a specific identity, that is, a resemblance, and there is nothing numerically
the same but the cause that produced them. Similarly, when
an old brick church fell to ruin, we may say that the parish
rebuilt ‘the same church’ out of sandstone and in a modern
species, sort, or kind.]

134

Treatise, Book 1

David Hume

plants and animals. So it can’t have a different origin from
the latter, but must come from a similar operation of the
imagination on similar objects.

iv: Scepticism and other systems

in the world of ideas; without them every distinct object is
separable by the mind and can be separately thought about,
and seems to be disconnected from every other object, not
merely from ones that are very dissimilar or distant. So
identity must depend on some of the three relations of resemblance, contiguity, and causation. Now, the very essence
of these relations consists in their making ideas follow one
another easily; so our notions of personal identity must proceed entirely from the smooth and uninterrupted movement
of thought along a sequence of connected ideas, in the way I
have explained.
The only remaining question is: Which of the three relations produce this uninterrupted movement of our thought
when we consider the successively existing perceptions that
we take to constitute a mind or thinking person? Obviously
contiguity has little or nothing to do with it; so we must
attend to resemblance and causation.
Let us take resemblance first. If someone always remembers a large proportion of his past perceptions, this will
contribute greatly to the holding of a certain relation within
the sequence of his perceptions, varied as they may be. For
memory is just a faculty by which we raise up images of
past perceptions; and an image of something must resemble
it. So ·each memory involves a perception that resembles
some past perception the person has had; and· the frequent
occurrence of these resembling ·pairs of· perceptions in the
chain of thought makes it easier for the imagination to move
from one link in the chain to another, making the whole
sequence seem like the continuation of a single object. In
this way, therefore, memory doesn’t merely show the identity
but also helps to create it, by bringing it about that many of
the perceptions resemble one another. The account given in
this paragraph applies equally to one’s sense of one’s own
identity and to one’s thoughts about the identity of others.

That argument strikes me as perfectly conclusive, but if
you aren’t convinced by it you should consider the following
even tighter and more direct argument. It is obvious that the
identity we attribute to the human mind, however perfect
we may imagine it to be, cannot make many different perceptions become one by making them lose the distinctness
and difference that are essential to them. Every distinct
perception that enters into the mind’s make-up is a distinct
existence, and is different and distinguishable and separable from every other perception (whether occurring at the
same time or at other times). Yet we suppose the whole
sequence of perceptions to be united by identity—·we say
that the members of the sequence are all perceptions of a
single person·—which naturally raises a question about this
relation of identity. Is it something that really binds together
our various perceptions themselves, or does it only associate
the ideas of them in the imagination? In other words, when
we speak about the identity of a person, do we observe some
real bond among his perceptions, or do we merely feel a
bond among the ideas we form of those perceptions? The
question is easy to answer, if we remember what I have already proved, namely that the understanding never observes
any real connection among objects, and that even the causeeffect relation, when strictly examined, comes down to a
customary association of ideas. For that clearly implies that
identity doesn’t really belong to these different perceptions,
holding them together, but is merely a quality that we attribute to them because of how the ideas of them are united
in the imagination when we think about them. Now, the
only qualities that can unite ideas in the imagination are
the three I have mentioned. They are the uniting principles
135

Treatise, Book 1

David Hume

Causation also has a role. The true idea of the human
mind is the idea of a system of different perceptions that
are linked by the cause-effect relation, through which they
mutually produce, destroy, influence, and modify each other.
Our impressions give rise to corresponding ideas, which in
their turn produce other impressions. One thought chases
another and draws after it a third by which it is expelled
in its turn. In this respect the soul is very like a republic
or commonwealth, in which the members are united by the
links that connect rulers with subjects; these members cause
others to come into existence ·by begetting or giving birth to
them·, and these in their turn keep the same republic continuously in existence throughout all the unceasing changes
of its parts. And just as the same individual republic may
change not only its members but also its laws and constitution, so also the same person can vary his character and
disposition as well as his impressions and ideas. Whatever
changes he undergoes, his various parts are still connected
by causation. Our emotions contribute to our identity just
as our impressions and ideas do, by making some of our perceptions influence others that occur at very different times.
This is what happens when we have a present concern for
our past or future pains or pleasures.

iv: Scepticism and other systems

grounds to have existed. How many of our past actions do
we actually remember? Who can tell me, for instance, what
he thought and did on the 1st of January 1715, the 11th of
March 1719 and the 3rd of August 1733? Or will he overturn all the most established notions of personal identity
by saying that because he has forgotten the incidents of
those days his present self is not the same person as the
self of that time? Looked at from this angle, memory can be
seen not so much to create personal identity as to reveal it,
by showing us the relation of cause and effect among our
different perceptions. Those who contend that memory alone
produces our personal identity ought to explain how we can
in this way extend our identity beyond our memory.
The whole of this doctrine leads us to the very important
conclusion that all the precise, subtle questions about personal identity can never be settled, and should be seen as
verbal difficulties rather than philosophical ones. Identity depends on the relations of ideas; and these relations produce
identity by means of that easy movement of thought that
they give rise to. But the relations in question are matters of
degree, and so is the easiness of the mental movement that
depends on them; so we have no correct standard by which
to settle when they acquire or lose their entitlement to the
name ‘identity’. ·Just because the basis of our identity judgments consists in matters of degree, there can be borderline
cases— just as there are borderlines for baldness, tallness
and so on·. All the disputes about the identity of connected
objects are merely verbal, except in so far as the relation of
parts gives rise to some fiction—some imaginary source of
union—such as I have described.
What I have said about the origin and the uncertainty
of our notion of the identity of the human mind can also be
applied—with little or no change—to our notion of simplicity,
·that is, the notion of a thing’s not having parts·. An object

Memory should be regarded as the source of personal
identity, mainly because without it we wouldn’t know of
the existence of this lengthy and continuous sequence of
perceptions. If we had no memory, we would never have
any notion of causation or, consequently, of the chain of
causes and effects that constitute our self or person. Once
we have acquired this notion of causation from our memory,
we can extend the same chain of causes—and consequently
the identity of our persons—beyond our memory, stretching
it out to include times, circumstances and actions that we
have entirely forgotten but which we suppose on general
136

Treatise, Book 1

David Hume

whose different coexistent parts are closely related strikes
the mind in much the same way as one that is perfectly
simple and indivisible, and the thought of it doesn’t require
a much greater mental stretch. Because contemplating it is
like contemplating something simple, we regard as though
it were simple, and we invent a principle of union as the
support of this simplicity and as the centre of all the different
parts and qualities of the object.

(2) Whatever is distinct is distinguishable, and whatever
is distinguishable is separable by the thought or imagination.
All perceptions are distinct. They are, therefore, distinguishable, and separable, and may be thought of as separately
existent, and may exist separately, without any contradiction
or absurdity.
When I view this table and that chimney, nothing is
present to me but particular perceptions that are of the
same kind as all other perceptions. This is the doctrine of
philosophers. But this table and that chimney can and do exist separately. This is the doctrine of the common man, and
it implies no contradiction. So there is no contradiction in
extending the same doctrine to all perceptions—·that is, the
doctrine that they can exist separately. The next paragraph
gives an argument for this·.
The following reasoning seems satisfactory on the whole.
All ideas are borrowed from previous perceptions. So our
ideas of objects are derived from that source. Therefore any
proposition that is intelligible and consistent with regard
to objects must be equally so when applied to perceptions.
But it is intelligible and consistent to say that objects exist independently, without having to inhere in any common
simple substance. So it can’t be absurd to say the same
thing about perceptions. ·We are therefore not entitled to
insist that there must be some self or substance in which
our perceptions exist·.
(3) When I look in on myself, I can never perceive this
self without some one or more perceptions; nor can I ever
perceive anything but the perceptions. It is a complex of
these perceptions, therefore, that constitutes the self.
(4) We can conceive a thinking being to have as few perceptions as we like—even to be reduced to the level (below
that of an oyster) of having only one perception, such as
that of thirst or hunger. In considering such a mind, do you

*****
[After Book I of the Treatise of Human Nature had been published, Hume
had some afterthoughts that were published in an Appendix to Book III.
Here is the afterthought that he asks us to insert at this point.]

·S TAR T

iv: Scepticism and other systems

OF THE APPENDIX PASSAGE ·

I had hoped that however deficient my theory of the intellectual world might be, it would at least be free from those
contradictions and absurdities that seem to infect every explanation that human reason can give of the material world.
But reconsidering more carefully the section on personal
identity I find myself involved in such a labyrinth that I don’t
know how to correct my former opinions, nor do I know how
to make them consistent. If this is not a good general reason
for scepticism, it is at least a sufficient one (as if I didn’t
already have plenty) for me to be cautious and modest in all
my conclusions. I shall present the arguments on both sides,
starting with those that led me to deny the strict and proper
identity and simplicity of a self or thinking being. ·I offer
seven of these, each pretty much independent of the others·.
(1) When we talk of self or substance we must associate
ideas with these terms, otherwise they would be meaningless. Every idea is derived from previous impressions; and
we have no impression of self or substance as something
simple and individual. We have, therefore, no idea of them
in that sense.
137

Treatise, Book 1

David Hume

conceive anything more than merely that one perception?
Have you any notion of self or substance? If not, the addition
of other perceptions can never give you that notion.
(5) The annihilation that some people suppose to follow
on death, and which entirely destroys this self, is nothing but
an extinction of all particular perceptions—love and hatred,
pain and pleasure, thought and sensation. So these must be
the same as the self, since the one cannot survive the other.
(6) Is self the same as substance? If it is, then there can
be no question of the same self remaining when there is a
change of substance. If on the other hand self and substance
are distinct, what is the difference between them? For my
part, I have no notion of either when they are conceived as
distinct from particular perceptions.
(7) Philosophers are beginning to be reconciled to the principle that we have no idea of external substance distinct from
the ideas of particular qualities. This should pave the way
for a similar principle regarding the mind, namely that we
have no notion of it distinct from the particular perceptions.
All of this seems clear and true. But having started my
account with our particular perceptions all loose and separate, when I proceed to explain the principle of connection
that binds them together, making us attribute to them a real
simplicity and identity, I come to realize that my account
is very defective, and that I wouldn’t have accepted it if it
weren’t for the seeming power of the foregoing arguments.

iv: Scepticism and other systems

our thought reflects on the sequence of past perceptions that
compose a mind, and feels its the ideas of them to be interconnected and to follow on from one another in a natural
way. Extraordinary though it is, this conclusion need not
surprise us. Most philosophers today seem inclined to think
that personal identity arises from consciousness, and consciousness is nothing but a thought or perception directed
inwards towards oneself. To that extent, this present philosophy of mine looks promising. [Now comes the trouble.] But
all my hopes vanish when I come to explain the principlesc
that unite our successive perceptions in our thought or consciousness. I cannot discover any satisfactory theory about
this.
In short, there are two principles that I cannot render consistent, nor can I give either of them up: (1) all our distinct
perceptions are distinct existences, and (2) the mind never
perceives any real connection among distinct existences. If
our perceptions either inhered in something simple and individual, or if the mind perceived some real connection among
them, there would be no difficulty. For my part, I must plead
the privilege of a sceptic and confess that this problem is too
hard for my understanding. I don’t say outright that it is
absolutely insoluble. Perhaps someone else—or even myself
after further reflection—will discover some hypothesis that
will reconcile those contradictions.
·N OW BACK TO SECTION 6·
Thus I have finished my examination of the various systems of philosophy, both of the intellectual and the moral
world; and, in my miscellaneous way of reasoning I have
been led into several topics that will either illustrate and
confirm some of what I have been saying or prepare the way
for what is to come. It is now time to return to a closer
examination of our subject, and to proceed in the accurate
anatomy of human nature, having fully explained the nature
of our judgment and understanding.

[Hume now re-states his own theory of personal identity, in a manner that is favourable to it. His subsequent worries and doubts start to

If perceptions are distinct existences, they form a whole only by being connected
together. But the human understanding can never discover
connections among distinct existences; we only feel a connection in our mind when our thought is compelled to pass
from one object to another. It follows, then, that personal
identity is merely felt by our thought: this happens when
surface only at the end of this paragraph.]

138

Treatise, Book 1

David Hume

iv: Scepticism and other systems

7: Conclusion of this book
But before I launch out into those immense depths of philosophy that lie before me ·in Books II and III of this Treatise ·, I
want to stop for a moment at the place I have so far reached,
and to ponder the voyage I have undertaken, which undoubtedly requires the utmost skill and hard work to be brought to
a happy conclusion. I am like a man who, having grounded
his ship on many shoals and nearly wrecked it in passing
a small island, still has the nerve to put out to sea in the
same leaky weather-beaten vessel, and even carries his ambition so far as to think of going around the globe in it. My
memory of past errors and perplexities makes me unsure
about the future. The wretched condition, the weakness and
disorder, of the ·intellectual· faculties that I have to employ
in my enquiries increase my anxiety. And the impossibility
of amending or correcting these faculties reduces me almost
to despair, and makes me resolve to die on the barren rock
where I am now rather than to venture into that boundless
ocean that goes on to infinity. This sudden view of my danger
makes me gloomy; and as that above all is the passion that
indulges itself, I can’t help feeding my despair with all those
down-casting reflections that the present subject provides in
such abundance.

that beats upon me from every side. I have exposed myself to
the enmity of all metaphysicians, logicians, mathematicians,
and even theologians—can I wonder at the insults I must
suffer? I have declared my rejection of their systems—can
I be surprised if they express a hatred of mine and of me?
When I look outwards ·and ahead· I foresee on every side
dispute, contradiction, anger, slander, and detraction. When
I look inwards I find nothing but doubt and ignorance. All
the world conspires to oppose and contradict me; and I am
so weak that when •my opinions are not supported by the
approval of others I feel •them loosen and fall away. I take
every step with hesitation, and every new reflection makes
me dread an error and absurdity in my reasoning.

·This is not unreasonable·; for what confidence can I
have in venturing on such bold enterprises when, beside the
countless infirmities that I personally have, I find so many
that are common to human nature? Can I be sure that when
I leave all established opinions I am following truth? and by
what criterion shall I recognize her [= truth] even if fortune
should at last guide me onto her path? After the most accurate and exact of my reasonings, I can give no reason why I
should assent to it [= my conclusion]; I merely feel a strong disposition to consider objects strongly in the manner in which
they appear to me ·as a result of that reasoning·. Experience
is a force that instructs me in the various conjunctions of
objects in the past; habit is another force that makes me
expect the same in the future; and the two work together on
the imagination, making me form certain ideas in a more
intense and lively manner than other ideas that don’t have
the same advantages. This quality by which the mind enlivens some ideas more than others seems trivial, and has

First, I am frightened and confused by the forlorn solitude
in which my philosophy places me, and see myself as some
strange uncouth monster who, not being able to mingle and
unite in society, has been expelled from all human society
and left utterly abandoned and disconsolate. I would like to
run into the crowd for shelter and warmth, but I can’t get myself to mix with such deformity. I call on others to join me so
that we can make our own separate society, but no-one will
listen. Everyone keeps at a distance, and dreads the storm
139

Treatise, Book 1

David Hume

no basis in reason; yet without it we could never assent to
any argument, or carry our view beyond •the few objects that
are present to our senses. Indeed, even to those objects we
could never attribute any existence but •what was dependent on the senses, and must ·therefore· bring them entirely
into that sequence of perceptions that constitutes our self or
person. And even in relation to that sequence, we could ·at
any given time· only accept the existence of •the perceptions
that are immediately present to our consciousness ·at that
moment·; the lively images with which the memory presents
us could never be accepted as true pictures of past perceptions. The memory, senses, and understanding are therefore
all founded on the imagination, or the liveliness of our ideas.

iv: Scepticism and other systems

us into opinions that seem to make a mockery of all our
past trouble and work, and to discourage us from future
enquiries. Nothing is more assiduously enquired into by the
mind of man than the causes of every phenomenon; and we
aren’t content with knowing the immediate causes, but push
our enquiries on until we arrive at the basic ultimate cause.
We aren’t willing to stop until we are acquainted with the
energy in the cause by which it operates on its effect—the
tie that connects cause and effect together—and the effective
quality on which that tie depends. This is our aim in all
our studies and reflections; so how disappointed we must
be when we learn that this connection, tie, or energy lies
merely in ourselves, and is nothing but that set of mind that
custom creates, which causes us to make a transition from
the impression of an object to the lively idea of its usual
accompaniment! Such a discovery not only cuts off all hope
of ever attaining satisfaction, but won’t even let us wish for
it; for it appears that when we say that we want to know
‘the ultimate and operating force’, regarding this as something that resides in the external object, we either contradict
ourselves or talk without a meaning.

No wonder a force that is so inconstant and fallacious
should lead us into errors when uncritically followed (as it
must be) in all its variations. It is this force that •makes us
reason from causes and effects, and that •convinces us of
the continued existence of external objects when they are
absent from the senses. But though these two operations
are equally natural and necessary in the human mind, in
some circumstances they are directly contrary to one another (section 4); so we can’t reason soundly and regularly
from causes and effects while at the same time believing in
the continued existence of matter. How then shall we relate
those two forces to one another? Which of them shall we
prefer? Or if we prefer neither of them, and (as philosophers
usually do) go sometimes with one and at other times with
the other, how confidently can we give ourselves the glorious title of ‘philosopher’ when we thus knowingly accept an
obvious contradiction? This contradiction (see 14iii ) would
be more excusable if it were compensated by any degree of
solidity and satisfaction in the other parts of our reasoning.
But that is not how things stand. When we trace human
understanding back to its first sources, we find that it leads

This deficiency in our ideas is not indeed perceived in
common life. Indeed, we are not in general aware that in
the most usual conjunctions of cause and effect we are as
ignorant of the ultimate force that binds them together as
we are in the most unusual and extraordinary cases. But
this ·unawareness· comes merely from an illusion of the
imagination; and the question is ‘How far ought we to yield
to these illusions?’. This question is very difficult, and the
choice of answers forces us to confront a very dangerous
dilemma. One option is to assent to every trivial suggestion
of the imagination. But these suggestions are often contrary
to one another; and anyway they lead us into such errors,
absurdities, and obscurities that we must eventually become
140

Treatise, Book 1

David Hume

ashamed of our credulity. Nothing is more dangerous to
reason than the flights of the imagination, and nothing has
led to more mistakes among philosophers. Men with bright
imaginations may in this respect be compared to the angels
whom the Scripture represents as covering their eyes with
their wings! I have already shown so many instances of this
that I can spare myself the trouble of going on about it any
more.

iv: Scepticism and other systems

principle and condemn all refined reasoning, we run into the
most manifest absurdities. If we reject it in favour of these
reasonings, we entirely subvert the human understanding.
We are left with a choice between •a false reason and •no
reason at all. For my part, I don’t know what ought to be
done in the present case. I can only observe what commonly
is done, namely: this difficulty is seldom or never thought of,
and even when it is present to the mind it is quickly forgotten
and leaves only a small impression behind it. Very refined
reflections have little or no influence on us; and yet we don’t
and can’t accept the rule that they ought not to have any
influence, for that implies a manifest contradiction.
But what have I just said? That very refined and metaphysical reflections have little or no influence on us? I can
scarcely refrain from retracting ·even· this opinion, and condemning it on the basis of my present feeling and experience.
The intense view of all these contradictions and imperfections in human reason has so heated my brain that I am
ready to reject all belief and reasoning, and can’t see any
opinion ·as true, or· even as more probable or likely than
another.
Where am I?
What am I?
What has caused me to exist, and to what condition shall
I return ·after death·?
Whose favour shall I court, and whose anger must I
dread?
What beings surround me? Which ones can I influence,
and which have any influence on me?
I am bemused by all these questions, and begin to
fancy myself in the most deplorable condition imaginable—
surrounded by the deepest darkness, and utterly deprived of
the use of every skill of body and mind.
Most fortunately it happens that since reason can’t scat-

The consideration of these troubles might make us resolve to reject all the trivial suggestions of the imagination,
and adhere to the understanding—that is, to the imagination’s general and more established properties. But even
this resolution, if steadily kept to, would be dangerous and
would bring the most fatal consequences. For I have already
shown in section 1 that the understanding, when it acts
alone and according to its most general principles, entirely
subverts itself and leaves us without even the lowest level
of conviction about any proposition, either in philosophy or
common life. We save ourselves from this total scepticism
only by means of a special and seemingly trivial property
of the imagination—namely, its making it difficult for us to
enter into remote views of things, not being able to accompany them with as strong an impression as we do things
that are more easy and natural. Shall we, then, adopt it as a
general maxim that no refined or elaborate reasoning is ever
to be accepted? Consider well the consequences of such a
principle! It cuts you off entirely from all science and philosophy; you proceed on the basis of one special quality of the
imagination, and by parity of reasoning you should embrace
them all; and you explicitly contradict yourself, because this
maxim must be based on the preceding reasoning, which
you must admit is sufficiently refined and metaphysical ·to
fall under the principle and thus be rejected by it·! What side
shall we choose among these difficulties? If we embrace this
141

Treatise, Book 1

David Hume

ter these clouds, Nature herself suffices for that purpose and
cures me of this philosophical gloom and frenzy, either by
reducing the intensity of these thoughts or by some pastime
that makes lively impressions on my senses that obliterate
all these chimeras. I dine, I play a game of backgammon, I
converse cheerfully with my friends; and when after three
or four hours’ amusement I turn back to these speculations,
they appear so cold, strained, and ridiculous that I can’t find
in my heart to enter into them any further.

iv: Scepticism and other systems

my follies shall at least be natural and agreeable! Where I
strive against my inclination, I shall have a good reason for
my resistance; and will no more be led to wander into such
dreary solitudes and rough passages as I have so far met
with.
These are the sentiments of my depression and slackness;
and indeed I must confess that philosophy has nothing to
bring against them, and expects a victory more from the benefits of a serious good-humoured disposition than from the
force of reason and conviction. In all the incidents of life, we
ought still to preserve our scepticism. If we believe that fire
warms or water refreshes, it is only because it is too much
trouble to think otherwise. Indeed, if we are philosophers, it
ought only to be on sceptical principles—·not in the hope of
arriving at assured truths, but only· because we feel inclined
to employ ourselves in that way. Where reason is lively, and
mixes itself with some disposition, it ought to be assented to.
Where it doesn’t, it can’t have any right to operate on us.
Thus, at a time when I am tired with amusement and
company, and have allowed myself a daydream in my room
or in a solitary walk by a river-side, I feel my mind all collected within itself, and am naturally inclined to think about
all those subjects about which I have met with so many disputes in the course of my reading and conversation. I can’t
help wanting to know the sources of moral good and evil,
the nature and foundation of government, and the cause of
the various passions and inclinations that move and govern
me. I am •not contented with the thought that I approve of
one thing and disapprove of another, call one thing beautiful
and another ugly, and make decisions concerning truth and
falsehood, reason and folly, without knowing what principles
I am going by in all this. I am •concerned for the condition
of the learned world that is so deplorably ignorant about all
this. I •feel an ambition arising in me to contribute to the

Here, then, I find myself absolutely and necessarily made
to live and talk and act like other people in the common
affairs of life. But although my natural disposition and the
course of my animal spirits and passions bring me to this
lazy acceptance of the general maxims of the world, I still feel
such remains of my earlier frame of mind that I am ready
to throw all my books and papers into the fire, and resolve
never again to turn away from the pleasures of life in order
to resume reasoning and philosophy. For that’s how I feel
in the depressed mood that governs me at present. I may—
I must—go with the current of Nature in my dealings with
my senses and understanding, and in this blind obedience I
show most perfectly my sceptical disposition and principles.
But does it follow that I must go against the current of Nature that leads me to laziness and pleasure? that I must to
some extent shut myself away from dealings with and the
society of men that is so agreeable? that I must torture my
brain with subtleties and sophistries, doing this at the very
time when I can’t satisfy myself that this painful activity is a
reasonable thing to do and can’t have any tolerable prospect
of arriving through it at truth and certainty? ·Why must I?·
What obliges me to misuse my time in that way? And what
purpose can it serve, either for the service of mankind or
for my own personal interests? No: if I must be a fool (and
all those who reason or believe anything certainly are fools),
142

Treatise, Book 1

David Hume

instruction of mankind, and to make myself known through
my discoveries. These •feelings spring up naturally in my
present frame of mind; and if I tried to get rid of them by
applying myself to any other activity or pastime, I feel I would
be a loser in point of pleasure; and this is the origin of my
philosophy.

iv: Scepticism and other systems

or dervish that ever was in the world. Generally speaking,
the errors in religion are dangerous; those in philosophy are
only ridiculous.
I am aware that these two cases of the strength and weakness of the mind—·that is, philosophy and superstition·—
don’t cover all mankind, and that in England in particular
there are many honest gentlemen who are always engaged
in their domestic affairs, or amusing themselves in common recreations, and so have carried their thoughts very
little beyond the objects that are every day exposed to their
senses. I don’t purport to make philosophers of these, and I
don’t expect them either to join in these researches or listen
to their results. Such people do well to keep themselves
in their present situation; and, rather than refining them
into philosophers, I ·would like to make philosophers more
like them; that is, I· wish we could give our founders of
·philosophical· systems a share of this gross earthy mixture,
as an ingredient that they commonly need and don’t have,
an ingredient that would damp down those fiery particles
of which they are composed! As long as philosophy makes
room for a lively imagination and for hypotheses that are
embraced merely because they are glittering and agreeable,
we can never have any steady principles or any opinions that
will square with common practice and experience. If such hypotheses were removed from philosophy, then we might hope
to establish a system or set of opinions which—if not true
(for that may be too much to hope for)—might at least be satisfactory to the human mind, and might stand the test of the
most critical examination. Many flimsy systems that have
arisen and then died, but this shouldn’t make us despair
of attaining this goal; consider the shortness of the period
in which these questions have been the subjects of enquiry
and reasoning. Two thousand years, with long interruptions
and under mighty discouragements, are a small stretch of

But if this curiosity and ambition didn’t carry me into
speculations outside the sphere of common life, I would still
inevitably be led into them by my own weakness. ·Let me
explain·. It is certain that •superstition is much bolder in
its systems and hypotheses than •philosophy is: whereas
•philosophy contents itself with assigning new causes and
explanations for the phenomena that appear in the visible world, •superstition opens up a world of its own, and
presents us with scenes and beings and states of affairs that
are altogether new. Now, it is almost impossible for the mind
of man to stay—like the minds of lower animals—within the
narrow circle of items that are the subject of daily conversation and action; ·so we are bound to stray outside that circle·,
and all we have to deliberate about is our choice of guide
·when we do so·, looking for the one that is safest and most
agreeable. In this respect I venture to recommend philosophy, and I don’t hesitate to prefer it to superstition of every
kind. For as superstition arises naturally and easily from the
popular opinions of mankind, it seizes more strongly on the
mind and is often able to disturb us in the conduct of our
lives. Philosophy stands in contrast to that. Sound philosophy can present us only with mild and moderate sentiments;
and the opinions offered by false and extravagant philosophy are merely the objects of cool generalizing thought, and
seldom go so far as to interrupt the course of our natural
inclinations. The Cynics are an extraordinary instance of
philosophers who, from purely philosophical reasonings, entered into extravagances of conduct as great as any monk
143

Treatise, Book 1

David Hume

time to bring the sciences to anything like completion; and
perhaps the world is still too young for us to discover any
principles that will stand up under examination by our remote descendants. Speaking for myself, my only hope is that
I may contribute a little to the advancement of knowledge by
giving in some respects a different turn to the speculations
of philosophers, and more clearly indicating to them the only
subjects in which they can expect assurance and conviction.
Human nature is the only science of man; and yet it has
been until now been the most neglected. I will be satisfied if
I can bring it a little more into fashion; and the hope of this
serves to bring me out of the depression and slackness that
sometimes take me over. If you find yourself in the same easy
disposition, follow me in my future speculations ·in Books II
and III·. If not, follow your own inclination, and wait for the
return of good humour and industriousness. The conduct
of a man who studies philosophy in this careless manner
is more truly sceptical than the conduct of one who, feeling
in himself an inclination to it, nevertheless totally rejects it
because he is overwhelmed with doubts and worries. A true
sceptic will be cautious about his philosophical doubts as
well as about his philosophical convictions; and he will never

iv: Scepticism and other systems

refuse any innocent satisfaction that offers itself on account
of either of them.
It is proper that we should •in general indulge our inclination in the most elaborate philosophical researches,
notwithstanding our sceptical principles, and also that we
should •give rein to our inclination to be positive and certain
about particular points, according to how we see them at any
particular instant. It is easier •to give up examination and
enquiry altogether than •to restrain such a natural disposition in ourselves and guard against the confidence that
always arises from an exact and full survey of an object. At
those moments we are apt to forget not only our scepticism
but even our modesty, and make use of such expressions as
‘it is evident’, ‘it is certain’, ‘it is undeniable’, which a due
deference to the public ought perhaps to prevent. I may have
followed others into committing this fault, but in face of any
objections that may be made against me on that account I
declare that such expressions were dragged out of me by my
view of the object at that moment; they don’t imply any dogmatic spirit or conceited idea of my own judgment—attitudes
that I am aware are not suitable for anybody, least of all a
sceptic.

144