Roland Hall

From the beginning there has been complaint about Locke’s use of the

word ‘idea’, particularly in his main work, the Essay concerning Human Under-

. Locke was attacked in his lifetime by Stillingfleet, and also by other

contemporaries (brought to light by John Yolton in his first book on Locke),

for the way he used the term. Subsequently Thomas Reid complained

that, if the word idea, in a work where it occurs in every paragraph, is used

without any intimation of the ambiguity of the word, sometimes to signify

thought, or the operation of the mind in thinking, sometimes to signify those

internal objects of thought which philosophers suppose, this must occasion

confusion in the thoughts both of the author and of the readers. I take this to

be the greatest blemish in the “Essay on Human Understanding” (p. 227).

The complaints continue in the Nineteenth Century, for instance in Sir

William Hamilton’s Discussions, and A. C. Fraser, the early editor of the Essay,

says in his 1890 book on Locke (p. 1ll of the 1901 edition) that Locke «offers

no theory of what ideas are» and that if we call them objects of consciousness,

or perceptions, or conceptions, or states of mind, or modifications of mind,

«we are only putting one term in place of another; not explaining what any of

the terms mean». Perhaps Fraser is not complaining: he remarks, «it is

enough to say, that without ideas or conceptions of things – that is to say, in

the absence of all conscious intelligence – there cannot be knowledge, opin-

ion, or even doubt». If so, Gibson (1917) agrees:

Among the questions which the ‘plain historical method’ sought to lay on one

side was that of the metaphysical nature of ideas themselves, concerning

which, as we have seen, the followers of Descartes had found themselves in

such difficulties. The initial assumption, underlying the whole procedure of

the Essay, is that the existence of ideas may be taken for granted, and their

function in knowledge examined, without entering upon the questions which

may be raised concerning their nature as elements of reality, or their relation

to the mind as a substance (p. 26).

Aaron, in section I of ch. 3 of his well-known book, accepts «one criti-

cism which» he says «has rightly been directed against Locke», viz. «that he

has included far too much within the connotation of this one term», since


«sense-data, memories, images, concepts, abstract ideas differ from each other

greatly, and to call them all by the same name is to invite confusion». But

Aaron rejects the earlier line of criticism, found in Reid and various succes-

sors, possibly even in Gibson, that Locke allows idea to be ambiguous as

between the object and the act of perceiving or thinking, contrary to his own

definition. Aaron firmly states, «1 doubt whether a single unambiguous

instance of the explicit identification of idea with the perceiving can be found»

(p. 100). Nevertheless, in the conclusion on pp. 106-107, Aaron is troubled

by Locke’s ambiguous use of idea, partly because – and this may have been

due to Swabey’s article of 1933 – «Locke may also mean by idea a universal

meaning, a term of a proposition, a logical content».

The complaints continued: «that Locke’s way of expressing himself is

confused and careless is scarcely open to denial» (Copleston, p. 99), and

«Locke’s use of it (sc. idea) was clearly confused and confusing» (Woozley,

p. 34). But are the complaints justified? Woozley says it «must be acknowl-

edged, we cannot ascribe any single sense to ‘idea’ used consistently through-

out its thousands of occurrences in the Essay», and concludes, «we should not

interpret it merely by selection. … what we have to do is to make the best

and most general sense of it that we can which is consistent with his theory of

knowledge as a whole» (ibid.). What Woozley in fact offers is that «my idea

of a thing is what I think or would say that the thing was» (ibid.). This,

though undeniable, is not exactly informative, and I am at a loss to see it as

any improvement on Locke’s own definition of idea:

It being that term, which, I think, serves best to stand for whatsoever is the

object of the understanding when a man thinks, I have used it to express

whatever is meant by Phantasm, Notion, Species, or whatever it is, which the

mind can be employ’d about in thinking (I. i. 8).

Nor does Locke foresee any problem; he continues:

I presume it will be easily granted me, that there are such Ideas in men’s

minds; every one is conscious of them in himself, and men’s words and

actions will satisfy him, that they are in others.

Locke’s other well-known explanation of his usage is at II. viii. 8:

Whatsoever the mind perceives in it self, or is the immediate object of perception,

thought, or understanding, that I call idea.

(I shall not discuss his occasional use of idea to mean quality in the exter-

nal object, which he mentions at this point, because he seems to be apologiz-

ing for it, and also because I have already offered elsewhere – in the «Locke

Newsletter», 18 (1987) at p. 21 – an attempted explanation of this misuse of

the term.)


Locke’s own attitude is clearly that his explanations of idea in the Essay

are sufficient, and this comes out forcefully in the sarcastic words of his first

Answer to Mr Norris’s Reflections, written in 1692 and first published in 1971, in

the «Locke Newsletter»:

If you once mention ideas you must be presently called to an account what kind

of things you make these same ideas to be
though perhaps you have no design to

consider them any further than as the immediate objects of perception or if you have

you find they are a sort of sullen things which will only show them [their?]

what but will not tell you whence they came nor whither they go nor what

they are made of and yet you must be examined to all those particulars I°

whether they be real beings or no, in the next place whether they be substances or

modifications of substances and further whether they are material or immaterial sub-

and then upon their being material you must answer to an hundred

solid questions (p. 10-11).

In calling upon Locke’s discussions of such questions in the published

Remarks upon some of Mr Norris’s Books and the Examination of P. Malebranche’s

, as various commentators have done, to bolster up modern inter-

pretations of Locke’s use of ‘idea’, we have to remember that both form part

of an embittered controversy with Norris, subsequent to the publication of the

Essay, and may not be reliable as evidence on how Locke thought about ideas

during the composition of the essay. Perhaps the same goes, to a lesser

extent, for his replies to Stillingfleet, where he is stung into arguing that there

must be ideas, as representative entities, rather than claiming that their exist-

ence is to be taken for granted. From the premiss that Bishop Stillingfleet can

think of his own cathedral, actually at Worcester, or of Descartes’ vortices,

which actually «never existed anywhere in rerum natura», without having them

actually in his mind, Locke says,

I conclude, your lordship has immediate objects of your mind, which are not

the very things themselves existing in your understanding; which if, with the

academics, you will please to call representations, as I suppose you will, rather

than with me ideas, it will make no difference (Second Reply; Wks. IV. 391).

Obviously, Locke thought that when he was not challenged there was no

need to offer arguments for ideas as immediate objects of the mind, or further

considerations as to their nature: as he says in the first Answer to Norris, pre-

sumably with sarcasm:

Perhaps I was lazy and thought the plain historical method I had proposed to

myself was enough for me perhaps I had other business and could afford no

more of my time to these speculations … (p. 10).

In fact, the general point that we cannot entirely rely on later controver-

sial discussions, such as the Examination of Malebranche, to interpret the Essay,


was made by T. E. Webb in the last century, against Sir William Hamilton.

He pointed out that

The last hours of Locke’s life were devoted to the preparation of the sixth

edition of his Essay, and in the Epistle prefixed to that edition he tells the

reader he has nothing to alter or to add (p. 36)1.

In the last two decades an interpretation has come to the fore, stemming

from John Yolton, A. D. Woozley, and others, denying that Locke’s ideas are

entities. One quotation from Yolton may suffice: «I see no evidence in the

Essay that Locke thought of ideas as entities. They were... his way of charac-

terising the fact that perceptual awareness is mental» (Locke and the Compass,

p. 134). I am not happy with this very influential view, partly because some

passages seem to me to suggest the contrary, and partly because I am uneasy

with the terms of the discussion.

On the first point, passages treating ideas as pictures or images occur not-

ably in II. x, ‘Of Retention’, e.g. «keeping the idea... for some time actually in

view... is called contemplation» (sec. 1), «the pictures drawn in our minds are

laid in fading colours» (sec. 5), «viewing again the ideas, that are lodg’d in the

memory, ... the appearance of those dormant pictures, depending sometimes

on the will» (sec. 7). Also in II. xxix. 8 (the comparison with anamorphoses),

«just thus it is with our ideas, which are, as it were, the pictures of things».

Similarly, the multitude of passages about making complex ideas by putting

different elements together, sound like the construction of mental imagery.

And pictures or images, even though mental, seem to me to be entities of a

kind: efforts to deny them a kind of being as objects would appear to have

failed, even if they are obviously not ‘real beings’2 in the sense of being

physical or independent of the mind.

This brings me to the second point: perhaps, when it was denied that

Locke’s ideas were entities, ‘entity’ was being used in a stronger sense; the

term is somewhat vague. And the other terms of the debate are perhaps shift-

ing, so that the positions taken are unclear. When it is pointed out that for

Locke ideas are perceptions, as I understand this it does not detract from their

thing-like character, though of course they are not physical things, at least not

when they come before the mind. The passages that equate ideas with per-


ceptions do not imply for me that they are acts, or operations of the mind (as

Reid had it), but that they are «the result or product of perceiving» (O. E. D.

sense 8, for which the first quotations are from Essay I. iv. 20 and IV. xi. 4,

and the next, naturally, from Hume).

So it is not always clear to me how much I am in disagreement with

John Yolton’s view of Locke’s concept of idea. Of course, I agree that

ideas for Locke are not ‘real beings’ and that there are ideas of mental

acts, but not that they are themselves mental acts, if this is being main-

tained. Similarly, the Woozley account, which contains a great many de-

nials but hardly any positive elucidation, now seems to me the opposite of

the truth. When Woozley claims that Locke, writing as he did in the

Essay, was «laying himself open to the interpretation of ideas as peculiar

entities in our private... world» (p. 32), that is because Locke meant what

he said. Woozley thinks he was «the victim of the unfortunate fact that

‘idea’ is a noun» and says that «noun-proneness indeed is one of the chief

occupational diseases of philosophers; it is fatally easy... to treat all nouns as

if they named things»; but here Woozley forgets that Locke himself is not

looking for the meaning of ‘idea’; he adopts the term as the most convenient

expression for something he already believes to exist, «whatsoever is the

object of the understanding»; Woozley admits (p. 29) that the Essay is «pep-

pered with phrases» suggesting «the accepted interpretation» and gives half

a dozen examples (different from mine above), e. g. «diagrams drawn on

paper are copies of the ideas in the mind» (IV. iii. 19).

Other attempts to reinterpret Locke’s concept of idea in various articles

on the topic are too careless to invite credence. For instance, when Lewis

writes of the explanation in I. i. 8,

The most simple and natural interpretation of this passage is that Locke

is maintaining in it merely that ideas are the objects of the understanding

when a man thinks – external objects and their qualities and the operations

of our own minds are the objects of our experience when a man perceives and

reflects (p. 136)

he has totally misunderstood what Locke means by think. And when Green-

lee writes, «According to the second definition of ‘idea’, idea and perceptual

act are equated» (p. 45), he bases this on Locke saying at II. i. 9 «having ideas

and perception being the same thing», which does not in any way equate ideas

with perceptual acts, but with the objects of perceptual acts. On the next page

he further claims that there are «two uses of ‘perception’ in the Essay» and

that «the conception of ‘idea’ as act of perception is based on the broader

sense of perception». Actually, it seems quite baseless.

Now the motivation for all these attempted re-interpretations is to avoid


the supposed difficulty of representationalism, that ideas as entities that repre-

sent preclude access to the represented objects, blocking us off from the real

world and compelling scepticism about it. Objectors are fond of quoting

Locke himself in support of the difficulty, when he says against Malebranche,

in the Examination section 51, «How can I know that the picture of anything is

like that thing, when I never see that which it represents?» (It is worth men-

tioning that later in the passage Locke puts in a qualification, «upon his prin-

ciples», which suggests that in his own philosophy there would be some ground

for believing certain ideas to be «the true representation» of what exists).

Actually, there are at least two possible answers to this difficulty in the

case of Locke. One is to point out, as Peter Alexander does (pp. 187-88), that

Locke «starts from the assumption of the existence of an external world which

is independent of our perceiving it», though in saying that ideas are «the

appearances» of external objects to us, perhaps Alexander may appear to be

granting some form of reinterpretation of what ideas are. He also refers us to

IV. xi. 3, where Locke rejects scepticism about «the existence of those things

which he sees and feels», – and here Locke finds «assurance enough» from

God. Of course, I recognise that the modern epistemologist will not wish to

accept an account that relies on extra premisses from God or science, but here

we are asking if representationalism contained insuperable difficulties as seen

by Locke.

The other possible answer is that Locke was satisfied with his reliance on

simple ideas as agreeing «with things». In a well-known passage, Locke raises

the difficulty:

But what shall be here the criterion? How shall the mind, when it perceives

nothing but its own ideas, know that they agree with things themselves? (IV.

iv. 3) [IV. is ‘Of the Reality of our Knowledge’]

Locke’s main answer is that «simple ideas are not fictions of our fancies,

but the natural and regular productions of things without us» (sec. 4).

To return, then, to the main question: can idea in Locke mean, or mainly

mean, a single kind of mental entity? Is there any account that will do justice

to Locke’s assertion that «every man’s reasoning and knowledge, is only about

the ideas existing in his own mind, which are truly, every one of them, particu-

lar existences» (IV. xvii. 8; 680, 32-4)? How important is the supposed ambi-

guity of idea in Locke’s usage? Is it a harmful ambiguity? Possibly, for exam-

ple, Thought is an ambiguous term, but do philosophers complain about this?

After all, terms can always be restricted, or different senses distinguished, and

we do not always feel that the unrestricted or undistinguished term is a prob-

lem. A. C. Fraser, in his edition of the Essay, seems happy enough to com-



Idea is thus, with Locke, a term of most comprehensive generality, embracing

all that is in any way immediately apprehensible by the mind of man, –

whether as a datum of external or internal sense, a sensuous image, or an

individualized product of generalizing thought. It is difficult to find an ade-

quate synonym, but perhaps phenomenon would be the nearest (I, p. 32, note


Perhaps such great generality in the use of idea is satisfactory to Locke

because he believes the differences of sense are not important, or not


Certainly what appears to us to be a lumping-together is quite convenient

to him. As I pointed out in my Berlin lecture in 1983,

What Locke has disguised with his shifting or inclusive, umbrella-like use of

the term ‘idea’ is the important transition from perception to imagination,

and again from imagination to abstract thought. These three operations are

lumped together in 2.8.8 in the new definition of ‘idea’: “Whatsoever the

mind perceives in itself, or is the immediate object of perception, thought, or

understanding, that I call idea”. This could be convenient for him because it

leaves no special need for an explanation of how the particular experiences of

perception turn into the universals required for thinking – they are all ideas

(pp. 21-2).

But does Locke’s apparent position really make sense ? Can we go without

making distinctions between what we see, think of, and think about? Even

Peter Alexander in his recent book, though maintaining that ‘species’ in the

Lockean explanation of idea at I. i. 8 means ‘simple idea of sensation’ and that

«the most central and important way in which Locke uses ‘idea’ is to mean

‘percept’» (pp. 96-97), thinks we have to distinguish between operating with

‘species’ while looking at something, operating with a ‘phantasm’ while

remembering it, and with a ‘notion’ (i. e. concept) if I am considering it as

falling under some verbal description. He adds that «there is little evidence

that Locke uses ‘idea’ frequently to mean ‘mental image’» (p. 95).

If we really want to make sense of Locke’s «term of most comprehensive

generality» (as Fraser called it; see above), the most recent article on this top-

ic, in the «Locke Newsletter», 17 (1986), offers the most attractive suggestion.

«Locke», says Michael Ayers, «holds an imagist theory of thought» (p. 5).

Ayers manages to deal with many of the difficulties usually thought to arise

from this interpretation. Images could include concepts, as he believes they did

for Hobbes and Gassendi, if we understand abstraction as involving «partial

consideration» (II. xiii. 13) of familiar ideas (i. e. images). So, he points out,

taking the most difficult example, the idea of existence, this idea is available to

us, according to Locke, when we consider the ideas in our minds «as being

actually there», or things «to be actually without us; which is, that they exist,


or have existence» (II. vii. 7). As for the ambiguities found in having the

idea of whiteness as a sensation or as a concept, Ayers regards these as harm-

less, differing only between an occurrent and a dispositional sense. (In my

Berlin lecture I identified the simple idea with the type-idea, as opposed to the

token, and described it as a ‘concept’: «on the re-interpretation, the simple

ideas are concepts, not perceptions or images, though of course they are app-

licable to perceptions and can be represented by images», p. 16).

It may be felt that crediting Locke with an imagist theory of thinking, as

Ayers does, does not quite solve the problem of finding a single kind of men-

tal entity as the main meaning of idea. If ‘percept’ or ‘sense-datum’ is con-

trasted with ‘image’ then it would not. It is an awkwardness that we use the

term ‘image’ in a way that excludes sense-data: it is applied to cases where the

external object is absent. Perhaps this is the reason why the word ‘image’ is

not very often used by Locke himself: he feels its inappropriateness to the case

of sense-perception. And yet in sense-perception there is an immediate object

too, something we see (etc.) that is a partial appearance of the external object.

In deciding not to differentiate between a sense-datum and an image, but to

treat both as mental objects allowing us to deal with parts of the world wheth-

er present or absent, and to designate them all by the single term idea, Locke

may have obtained a convenient outlook for himself, even if it has generally

baffled his commentators.


Aaron, R. I., John Locke, London 1937 (Quoted from 3rd edn, Oxford 1971).

Alexander, P., Ideas, Qualities and Corpuscles: Locke and Boyle on the External

, Cambridge 1985.

Ayers, M., Are Locke’s “Ideas” Images, Intentional Objects, or Natural Signs?, «The Locke Newsletter», 17 (1986), pp. 3-36.

Copleston, F. C., A History of Philosophy, vol. 5, London 1959.

Fraser, A. C., John Locke, An Essay concerning Human Understanding, collated and

(etc.), 2 vols, Oxford 1894.

Fraser, A. C., Locke, Edinburgh 1890.

Gibson, J., Locke’s Theory of Knowledge and its Historical Relations, Cambridge


Greenlee, D., Locke’s Idea of “Idea”, «Theoria», 33 (1967), pp. 98-106.

(Quoted from the reprint in I. C. Tipton, Locke on Human Understanding,

Oxford 1977).


Hall, R., Locke and Sensory Experience – Another Look at Simple Ideas of Sensation,

«The Locke Newsletter», 18 (1987), pp. 11-31. (The 1983 Berlin lec-

ture; also forthcoming in a volume edited by Hans Poser).

Lewis, D., The Existence of Substances and Locke’s Way of Ideas, «Theoria», 35

(1969), pp. 124-46.

Locke, J., An Essay concerning Human Understanding (1690). (Quoted from the

edition by Peter H. Nidditch, Oxford 1975; with appropriate decapitalisa-

tion and de-italicisation by the present writer, but with all punctuation

preserved). References of the form ‘I. i. 8’ are all to the Essay.

Answer to Mr Norris’s Reflections
(Written 1692, first published, transcribed

by Richard Acworth, in 1971, «The Locke Newsletter», 2, pp. 7-11,

Locke’s First Reply to John Norris).

Mr. Locke’s Reply to the ... Bishop of Worcester’s Answer to his Second Letter

(1699) (i. e. Second Reply to Stillingfleet) in Works, London 1823, IV.

An Examination of P. Malebranche’s Opinion of seeing all Things in God (1706), in

Works, London 1823, IX.

O.E.D. = Oxford English Dictionary

Reid, T., Essays on the Intellectual Powers of Man (1785). (Quoted from The Works

of Thomas Reid
, ed. by Sir William Hamilton, 4th edn, Edinburgh-London


Swabey, W. C., Locke’s Theory of Ideas, «Philosophical Review», 42 (1933),

pp. 573-93.

Webb, T. E., The Intellectualism of Locke: An Essay, Dublin-London 1857.

Woozley, A. D., Introduction to his abridged edition of Locke’s Essay. London


Yolton, J. W., John Locke and the Way of Ideas, Oxford 1956.

Locke and the Compass of Human Understanding: A Selective Commentary on the ‘Essay’, Cambridge 1970.

I have not checked the strict correctness of this claim, but the general point remains.

Later addition: In fact the claim is incorrect. Locke does not say, as Webb claims, that he has

nothing to alter or to add, but that «there is very little added or altered». Furthermore, as Ian

Tipton has kindly pointed out to me, Locke originally made this remark in the Epistle to the

fifth edition of his Essay, he later transferred it to the sixth edition.
Locke uses this expression in IV. xvii. 8 for what ideas represent.

Roland Hall . :

This page is copyrighted


  • There are currently no refbacks.