John W. Yolton




The entry under ‘idea’ in the Oxford English Dictionary lists twelve different

uses of the term ‘idea’, most of which range through our two centuries. The

earlier uses in the sixteenth century are Platonic Ideas or Forms. There are

other uses derivative from this Platonic one, the sense of a standard or ideal,

and what the OED terms ‘a weakened sense’: “A conception or notion of

something to be done”. Another general use is as a pattern or preliminary

sketch; in music as a theme or phrase to be used later in a composition.

These uses find some vague reflection in later philosophical writings, but none

of them states the special and technical sense found in Locke and others in

Britain, or among the Cartesians on the Continent. A seventh use identified

by the OED begins to add features found in the philosophical use, but the

examples cited are mainly literary: “a figure, representation, likeness, image,

symbol”, even ‘picture’. The OED cites a line from Shakespeare’s Richard III:

«I did inferre your Lineaments, Being the right Idea of your Father, both in

your forme and Noblenesse of Minde».

The eighth use identified by the OED combines several of the senses that

we find associated with the philosophical use of the term: “Mental images,

conception, notion”. The justification of the phrase ‘mental image’ seems to

come from references in some of the sources cited to the faculty of imagina-

tion or fancy. For example: «Me thinks the Idea of her person represents it

selfe an object to my fantasie» (Greene, 1589), or Shakespeare (Much Ado,

1599): «The idea of her life shal sweetly creepe into his study of imagination».

Since the term ‘image’ will concern us later, I raise a question about the

OED’s use of it just because of the association of some ideas with the imagina-

tion. The faculty of the imagination is supposed to deal with or even gener-

ate images, but the notion of mental images is, I think, unclear. What these

entries in the OED do is remind us of the use of various mental faculties in

describing the mind and its acquisition of ideas. There are several features in

the context of the use of the term ‘idea’ that we need to notice, the reference

to faculties is one.


The ninth OED use again adds, mainly from literary and historical

sources, more features of the philosophical use: “Any product of mental

apprehension or activity, existing in the mind as an object of knowledge or

thought”. The phrase ‘existing in the mind’ becomes, of course, important in

those uses that interest us. Another feature of this use given by the OED is

“a thought, conception, notion, or way of thinking”. Again, the equivalence

between, or the often alternation in the use of ‘idea’ and ‘thought’ is found in

a number of the writings we shall examine.

The tenth entry in the OED refers to the previous eighth and ninth uses

for what is labeled ‘Modern philosophical developments’. There, the writers

cited as sources are Descartes, Locke, Berkeley, Hume and Watts. The OED

formulation of this use contains some of the troublesome phrases that have

plagued subsequent attempts to find precise, unequivocal meanings for the

term ‘idea’ in these writers: phrases such as ‘in the mind’, ‘directly present to

cognitive consciousness’, ‘the immediate object of thought’.

One use of our term not cited by the OED, neither in its different discer-

nible uses nor in the sources employed, is the metaphysical or, what I have

labeled on my chart (see below, p. 254) as the ontological use. We can go to

several sources for a statement of the ontological nature of ideas. One such

source is the anonymous author of a small pamphlet published in 1705, A

Philosophick Essay concerning Ideas
. It is interesting to find this author saying that

«There is hardly any Topick we shall meet with that the Learned have differ’d

more about than that of Ideas» (p. 4). This remark is interesting because in

1705, it is not easy to identify the controversies or debates to which he refers.

Of course, there was the storm created by Locke’s rejection of innate ideas and

principles, but most of the defenders of innateness in Britain (and it was these

writers against whom Locke1 mainly wrote) did not use the term ‘idea’.

Locke’s polemic had reformulated the debate in that language, so it may well

be the controversy over innateness that this author had in mind. William

Sherlock, whom this author names in the subtitle to his pamphlet, was one of

the attackers of Locke and a defender of innateness: his Digression concerning

Innate Ideas
uses Locke’s language of ideas. The 1705 writer also lists as one of

the notions about ideas that they are effluvia from bodies, but again, this Epi-

curean and scholastic doctrine did not use the term ‘idea’. The author may

have had in mind the extended debate over the nature of ideas between

Arnauld and Malebranche in the 1680s, a debate that Locke followed closely.

This 1705 author does refer to Malebranche, some of his definitions of ideas


are very close to those used by Arnauld. What is important about the remark

on the different views the learned have on the nature of ideas is its indication

of the activity over the meaning of the term, even though this author does not

name all the learned men he has in mind.

One account of ideas listed by the anonymous author can be found in at

least one writer, a follower of Malebranche, John Norris. Norris tries to

locate Locke’s ideas in the metaphysic of substance and mode. The 1705

author had said that some writers say ideas are modes, others that they are

substances. It was this schema that Norris tried to force Locke to use. In

attacking Locke for failing to explain what ideas were, Norris really meant

that Locke had not said whether his ideas were substances or modes, or even

whether they were physical or immaterial. Norris thought he had trapped

Locke within the alternatives of saying ideas are effluvia from objects, and

hence material; or saying they were immaterial substances, and hence, for

Norris the occasionalist, incapable of being caused by bodies.

By 1728, when Chambers published his Cyclopaedia, Locke’s Essay is the

main source used in characterizing and explaining the term ‘idea’. Chambers

does still mention the ‘peripatetics’ and their corporeal species that resemble

objects and the intellect that immaterializes them. Chambers also refers to

some writers who say the soul produces its own ideas, ideas that do not resem-

ble their objects. He mentions others who say the soul’s perfections reveal to

it a knowledge of the world (perhaps a reference to Leibniz), and he also tells

us that some hold to innate ideas while others locate ideas in God (clearly a

reference to Malebranche). Descartes’s classification of three kinds of ideas is

also mentioned. In the Cyclopaedia’s supplement (1753), much of the material

from Locke has been replaced with references to Berkeley, to the differences

between Berkeley and Locke on ideas. Leibniz is discussed and some attention

is given to Malebranche. This supplement also discusses another important

anonymous work, credited to Charles Mayne (the author of An Essay concerning

Rational Notions
, 1737), although the author of this work is more usually given

as a Zachary Mayne. The authorship has yet to be firmly established. This

work was Two Dissertations concerning Sense and the Imagination (1728); its central

point is to distinguish ideas from notions, the former being limited by Mayne

to sensory content.

Some examination of these various surveys and dictionary entries of the

term ‘idea’ (the OED, the 1705 pamphlet, Norris’s questions to Locke, Cham-

bers) are useful for our purposes in that they give us some indication of the

diversity in the use of that term. The OED entries under ‘Modern philosoph-

ical developments’ may reflect some assumptions made by the person who

wrote that entry, but for the most part, each of these surveys has caught some

of the actual uses of that term found in the writers we have come to include


in the philosophical canon. The term and the issues over ideas appear in

writers, both well-known and obscure, after Locke’s Essay had popularized the

way of ideas; but some books in Britain were, as indeed was Locke, influenced

by the Cartesian tradition as well. One point worth keeping in mind (a point

which the entry under ‘idée’ in the Diderot-d’Alembert Encyclopédie makes) is

that many logic books after Locke (and after the Port Royal logic) make exten-

sive use of the term. The term and the issues around it are found in books of

psychology (that is, books on the nature of the soul) and in logics2.

With this general introduction, it is time to turn to my diagram or chart

which attempts to classify the different philosophical uses of our term3. We

need to keep in mind two other features of the context. I mentioned one, the

doctrine of different faculties of the mind. The two others are the relation

between mind and body and the account of our knowledge of an external

world. Each of these general features influenced the analysis of the way of

ideas. The two latter in particular pervade eighteenth-century discussions. I

shall not deal with these important general features of the tradition of ideas,

but some references to them will occur, especially in discussing my category

III. I should also point out that the main breakdown on my chart is between

an ontic and an epistemic interpretation of ideas. The difference is obviously

not exclusive since any one of the ontic categories has an epistemic role, with

the possible exception of I(a). Moreover, I shall be suggesting later that cate-

gory III (which I shall also discuss last) may more properly belong between the

ontic and epistemic divisions, if not entirely on the epistemic side. Category

III is an attempt to merge the two sides together4.

The names under each of the entries include both those writers who cite

that use in other writers and some who used ‘idea’ in that way. Those under

I(a), ideas as modes of body, of material substance, are not users of that sense,

nor do the sources they cite use the term that way. Norris, as I mentioned,

does try to push Locke in that direction, and there were others later who

charged Locke with materialism, but there is little textual support for saying

Locke meant his ideas to be brain impressions. Thus, I(a) is on my chart for its


being mentioned rather than for its being used in the two centuries we are


‘Ideas as modes of mind’ has two different interpretations. I(b), which

talks of images, may be more akin to my category II, ideas as substances.

Chambers says that images in philosophy are «the Traces or Marks which out-

ward Objects impress on the mind, by means of the Organs of Sense», but if

these marks are on the mind, they cannot be corporeal. Chambers, like many

other writers, found it natural to use a physical vocabulary (traces, marks,

impressions) when talking about the mind and awareness. Those writers who

defended innate principles used such language liberally. The 1705 writer

remarks that one of the reasons there is so much ambiguity around the term

‘idea’ is precisely because too many

Men do not sufficiently abstract their Thoughts from Matter, but make use of

such Terms as can properly relate to Matter only, and apply them to the Mind

in the same Sense as they are spoken of Matter, such as Images and Signatures,

Marks and Impressions, Characters and Notes of Things, and Seeds of Thoughts and

Knowledge (p. 5).

When the term ‘image’ is used, as it is by many writers (both those who

use it and those who credit it to others), it is clear that what is meant is a

mental image, but there is no clear account in the literature of what a mental

image is. To decide where this use of the term ‘idea’ belongs on my chart,

we need some clarification of the phrase ‘mental image’. I place this sense of

the term under I(b) for the moment. I(c) is by far the more interesting and

more important use of ‘idea’ in our two centuries. The 1705 writer makes it

clear that there is a choice to be made. He works from a definition of ‘idea’

as «the Representation of something in the Mind» (my category IV). He

indicates that these representations can be modifications of the mind or they

can be distinct beings (p. 6). ‘Distinct beings’ are entities, substances or near-

substances. Malebranche is his example of a writer who takes this route; the

other route, ideas as modes of mind, is, he says, «the more Common and Gen-

eral Hypothesis». It is these two concepts of ideas which characterize the way

of ideas in our two centuries, the majority of writers identifying themselves

with the second concept. These two concepts are reaffirmed by Thomas Reid

late in the eighteenth century; his attack was directed against ideas as distinct


I(b) Images: In his Logick: or the Right Use of Reason in the Enquiry after Truth

(2nd ed. 1726), Isaac Watts gives several very brief statements about what

ideas are, one of which is «the Notions or Pictures of these Things, as they are

considered, or conceived in the mind»; these are, he says, «the Ideas that we


have to do with in Logick» (p. 9). The conjunction of the terms ‘notion’ and

‘picture’ is unusual, since notions are most often linked with conceiving, not

imagining. He goes on to say that «among all these Ideas, such as represent

Bodies, are generally call’d Images, especially if the Idea of the Shape be

included». He makes no attempt to explain the term ‘image’, nor is that the

main use of the term for him.

In his account of “Sense, and the Ideas of Sensation”, Peter Browne (The

Procedure, Extent, and Limits of Human Understanding
, 1728) accepts the claim that

«we cannot think, or be conscious of thinking, ‘till we have some Ideas or

Semblances of an Object to think upon» (p. 55). That his term ‘semblance’

means ‘image’ is confirmed when he continues by speaking of the imagination

as the ‘storehouse’ or ‘receptacle’ of the images «which are transmitted thro’

the Senses». While he uses the seal and wax analogy he does not always or

even often speak of images (more often it is ‘representations’), but he does use

words such as ‘likeness’, ‘similitude’ and ‘resemblance’ for sensory ideas.

If any one, not yet satisfied, shall ask farther what an Idea is? I shall desire

him to look upon a Tree, and then immediately to shut his Eyes, and try

whether he retains any Similitude or Resemblance of what he saw; and if he

finds any such within him, let him call that an Idea, till a better word can be

found (pp. 58-59).

He claims that it does not make much difference whether we say that objects

are only occasions for our having ideas, or that «the Ideas of sensible Objects

exhibit to us a true Image of their Real Nature» (p. 60). Later in this same

chapter, he is more emphatic:

Nothing is properly an Idea but what stands in the Mind for an Image or Re-

presentation of something which is not in it; the thing must be without us;

and because it cannot itself enter, the Likeness of it only is conveyed thro’ the

Senses into the Imagination (p. 65).

While Browne makes more use of the word ‘image’ (with its attendant ‘like-

ness’ and ‘similitude’) than Watts, there is no attempt made to tell us what

these mental pictures or images of the imagination are other than to suggest

they are mental contents that convey some information about objects. The

term ‘image’ is not so obviously one of those physical words applied to mind

which the 1705 author warned about, but it does have its primary home in

optics and the use of mirrors, or even in the camera obscura, where there are

visual pictures or reflections of objects.

In his Two Dissertations, Mayne does use the mirror and image analogy for

sensation (p. 61), and he says explicitly that the term ‘idea’ «signifies nothing


more ... than an Image or Copy» (p. 58). In the Dissertation on the Imagination,

the idea, image or phantasm is in the mind when the object is no longer pre-

sent (p. 69). The faculty of imagination «presents to the Mind’s view the

Images or Ideas of external Objects» (p. 69). Mayne writes as if sensing pre-

sents us with the appearance of objects, while the imagination as a ‘second

sense’ enables us to «Think of the Image or Picture, in the manner as if the

Original were present before us» (p. 71). It is better and more accurate to

«contemplate and peruse the Object it self» than its image or idea. Throughout

this chapter, he couples ideas with phantasms (e. g. pp. 82, 83, 88). Writing

about Locke, Mayne charges him with speaking indifferently of the «immedi-

ate Appearance of an Object to Sense» or of the object’s «Image in the Mind

which serves afterwards in its stead» (p. 107). Mayne insists that it is only the

image in the mind which is ordinarily referred to as an idea (pp. 108-09).

That term signifies «nothing else or more than the Image, Picture, or Repre-

sentation in the Mind of a sensible Object» (see also p. 116). The term ‘re-

semblance’ is also used by Mayne in this context (pp. 108, 110). The same

conjunction of picture and image is found in his Essay on Consciousness (attached

to his Two Dissertations), with the specification that these are in the mind when

our senses do not present us with objects (p. 163).

Mayne wants to draw a firm distinction between sensing and understand-

ing. Even the concept of object (much as Kant was to say later) is a product

of the faculty of understanding and Mayne’s special function of ‘conscious-

ness’. The senses present us with appearances. To understand those appea-

rances, we need to have what Mayne calls ‘notions’, categories really, that

make sense of what appears. Appearances (which he also calls ‘representa-

tions’) are such things as colours, sounds, tastes (p. 9). The hardness of a

body is an appearance to sense, «The Mind’s Intellectual Notion of Hardness is,

that it is a Property or Quality» belonging to some object, some ‘Being’

(p. 10). On his account of perceptual awareness (our awareness of objects)

there is the appearance (i. e., some quality or group of qualities presented to

sense), the idea or image formed from the sensible appearance by the imagina-

tion, and then the notion, a product of the understanding. But’ image’ remains

unanalyzed, Mayne being content to rest with the mirror or picture analogy.

I(c) Perceptions or Thoughts: One of the less well-known British logics, An

Introduction to Logick
, by Edward Bentham (1773; he earlier published Reflexions

upon the Nature and Usefulness of Logick
, 1740; 2nd ed. 1755), identifies the strict

and proper sense of’ idea’ as the image we have of material objects, but Ben-

tham gives as a «large sense» of the word, «the thing that is thought upon, as

it is thought upon» (pp. 2-3). This large sense is identified by Thomas Reid as

the ‘popular meaning’: «to have an idea of any thing, signifies nothing more

than to think of it» (Essays on the Intellectual Powers of Man, 1785, vol. 1, p. 173).


Reid elaborates on this popular sense: «To think of a thing, and to have a

thought of it; to believe a thing, and to have a belief of it; to see a thing, and

to have a sight of it; to conceive a thing, and to have a conception, notion, or

idea of it, are phrases perfectly synonymous” (p. 174). Reid ends this passage

by saying that, on this sense of the term, «To think without ideas would be to

think without thought, which is a manifest contradiction».

In his Notebooks, Berkeley has a number of entries using ‘thought’ or ‘per-

ception’ as a synonym for ‘idea’. For example,

Our simple ideas are so many simple thoughts or perceptions, and that a per-

ception cannot exist without a thing to perceive it or any longer than it is

perceiv’d, that a thought cannot be in an unthinking thing, that one uniform

simple thought can be like nothing but another uniform simple thought.

Complex thoughts or ideas are only an assemblage of simple ideas... (no.


Entry no. 299 also speaks of the «thought or perception I call extension», and

of «an idea i. e. perception or thought». Entry no. 378.8 says, «All our ideas

are either sensations or thoughts». Entry no. 706 starts by saying: «No Per-

ception according to Locke is active. Therefore no perception (i. e. no idea)

can be the image» of that which is active. Section 41 in the New Theory of

links thoughts with sensations: «The objects intromitted by sight would

seem to him (as in truth they are) no other than a new set of thoughts or


Hume frequently, both in the Abstract and the Enquiry, uses ‘thought’ and

‘idea’ as synonyms. For example, «All our ideas or thoughts [are] derived

from impressions» (Abstract, p. 22); «thought and sensation» as types of

impressions (ibid., p. 25); thoughts copy objects (Enquiry, p. 17); two kinds of

perceptions are again said to be thoughts or ideas and impressions (ibid.,

pp. 18, 19, 23)5.

The interpretation of ‘idea’ as ‘thought’ is found in one of the 1705

author’s close followings of Arnauld: «That Thought and Idea are the same

thing...» is plain from some of his earlier definitions (pp. 8, 9). This author

reminds us of Locke’s acceptance of this Arnauldian equivalence when Locke

says that to have ideas and to perceive are the same (Essay, 2.1.9)6. This

writer also says ideas are modes of thinking (p. 11). Later still, he denies that


thoughts are distinct beings (p. 16), and he asserts that ideas are «the natural

Operations of the Mind upon the several Objects presented to it
» (p. 20).

For the 1705 author, the contrast was sharp between ideas as distinct

beings (which he rejects) and ideas as the thoughts we have about some object,

quality or event. This distinction was the one that had been given so much

attention in the exchange between Arnauld and Malebranche, Arnauld firmly

objecting to Malebranche’s turning ideas into entities, distinct beings.

II Substances: Thomas Reid argued that most modern philosophers worked

with a definition of ‘ideas’ which made ideas special objects intervening

between the perceiver and the external world. Reid found the roots of this

modern use of the term ‘idea’ ultimately in Plato, but more specifically he saw

it as a carry-over from the scholastic theory of species and phantasms. Reid

even goes so far as to say «all the systems of perception that have been

invented» follow that tradition: «For they all suppose that we perceive not

external objects immediately, and that the immediate objects of perception are

only certain shadows of the external objects» (vol. I, p. 133), Ideas, he says,

become internal objects. These internal objects on Malebranche’s account belong

to God, who gives them to us on appropriate occasions. Malebranche’s ideas

were clearly not modes of finite minds, nor could they be substances in the

full sense, since the only proper substances for Malebranche were minds (fi-

nite and infinite) and bodies. Malebranche’s ideas might be seen as modes of

God’s mind but he writes about them more in the language of substance.

However we classify Malebranche’s ideas, Reid saw them, and the use of that

term by most British philosophers, as special objects standing in for external

objects. In Reid’s reading of the history of perception theory, he does not use

the language of substance and mode, nor do the writers he discusses. Nev-

ertheless, to talk of ideas as internal objects, as distinct beings, would contrast

with talking of them as thoughts or perceptions. If they had to choose, most

of the British writers, with the exception of Berkeley, would, I think, say ideas

were modes of mind. John Norris would be another exception, since he tries

to follow Malebranche. Malebranche’s ideas are, then, the nearest we can

come to ideas as substances. In Britain, Norris is the only writer I know of

who makes ideas substance-like. If true, this claim of mine will have impor-

tant consequences for our understanding of the way of ideas in Britain.

IV Representations: The 1705 writer defines ‘idea’ as «the Representation

of something in the Mind» (p. 6). He thinks everyone agrees with this brief

definition. Disputes arise, he says, over the question, «whether this Represen-

tation be only a Modification of the Mind, or be a Distinct Being, or Sub-

stance United to the Mind» (p. 6). Isaac Watts repeats the definition given by

the 1705 writer (without mentioning him), saying that that term was generally

defined that way (p. 8). The same definition is found in Peter Browne


(p. 55). Browne also speaks of sensible objects imprinting «some Representations

or Characters of themselves upon us», using the same term ‘representation’ for

the impressions left on wax by a seal (p. 58). Another physical analogy is

‘leaves a footstep’ on our senses. He also links together ‘image’ and ‘represen-

tation’ (p. 65). In saying that the ideas of sensation are immediate, he

explains: «when they are First obtained, [they] necessarily presuppose the Pre-

sence of the Object, and some real actual Impression of it upon the Organs of

Sense; there is an Immediate and direct Representation of the Object» (p. 103).

In this passage, the impression is made on the sense organs, an impression

which, for all Browne says to the contrary, could be a physical mark or

impulse. One would like to assume that he means the representation is available

to awareness, not just an impression on sense organs or brain, but he does not

clarify this point.

In his The Elements of Logick (1748), William Duncan is also not very care-

ful in sorting out the physical and physiological part of sensation from the

psychological and cognitive. Duncan follows the usual story of objects con-

veying impressions to the brain via the senses, arousing the attention of the

understanding (6th ed. p. 11). The ‘notices’ of objects thus conveyed to the

mind are called ‘ideas’. These notices enable the mind to take «a View of

Things, as represented to it by its own Consciousness» (pp. 21-22). Renewed

representations can be generated by the mind after the object is no longer pre-


In writing about the use of the term ‘idea’, Mayne is primarily reporting

how he thinks other writers (chiefly Locke) employed that term. Ideas, he

says, are not «perfect and exact Representations of the Originals from which

they were taken» (p. 30). He routinely links ‘image’, ‘picture’ and ‘represen-

tation’ (pp. 104, 107, 124), or ‘copy’ and ‘representation’ (p. 117). In his Essay

on Consciousness
, Mayne has consciousness perceiving or representing to the

mind its thoughts and acts, and the senses perceiving and representing «the

external Forms and Appearances of corporeal Objects» (p. 174).

There are only a few uses of the term ‘representation’ in Locke’s Essay

that are coupled with ‘image’. In discussing real and fantastical ideas, he says

all simple ideas are real (i. e., agree to the reality of things), but not all of

them are «the Images or Representations of what does exist». The ideas of

secondary qualities are real, not in being «exact Resemblances of something in

the things themselves», but because there is a «steady correspondence» with

the «distinct Constitutions of real Beings» (2.30.2). In a brief section on ade-

quate and inadequate ideas, Locke contrasts the false view of ideas of sub-

stances – that they refer to «a supposed real Essence» - with the correct view

– that «they are only designed to be Pictures and Representations in the Mind,

of Things that do exist, by Ideas of those qualities that are discoverable in


them» (2.31.6). A third coupling of ‘image’ with ‘representation’ is found in

the chapter on general terms. There, Locke says of the child’s ideas of nurse

and mother that they «are well framed in their Minds; and, like Pictures of

them there, represent only those Individuals» (3.3.7).

Other uses of the term ‘representation’ by Locke make reference to a

general (or supposed) relation between ideas and things. For example, 2.30.1

speaks of things from which ideas are taken, «or which they may be supposed

to represent». The chapter on general terms talks of abstraction making ideas

capable of «representing more Individuals than one» (3.3.6), or, the main

claim of that chapter, that «Ideas are general, when they are set up, as the

Representatives of many particular Things» (3.3.11). Later in the Essay Locke

explains the necessity of one premise in a syllogism being universal by refer-

ence to his dictum about abstraction and the ability of a particular idea to

represent more than one thing (4.17.8).

There are a number of passages dealing with the idea of substance which

use the term ‘representation’. The complex ideas of substances are «intended

to be Representations of Substances, as they really are» (2.30.5), but such ideas

are false or misleading if they are «looked upon as the Representations of the

unknown Essences of Things» (2.32.18). The names of substances, we are

told, are made use of “ultimately to represent Things” (3.11.24). Indirectly

referring to substances, he tells us that simple ideas of sense represent the

power in objects which produces those ideas (2.32.16). In 2.31.3, Locke

explains that we desire to copy and «to represent to our selves that constitu-

tion, on which» the properties of substances depend.

The term ‘representation’ also occurs in Locke’s account of personal

identity. Consciousness of some past action as mine is «a present representa-

tion of» that action (2.27.13). In that same passage, he considers the possibi-

lity that some action might be represented to the mind which never occurred.

He also speaks of representations in dreams. In 2.33.13, he talks of memory

and the representations of past events and feelings. When he discusses his

doctrine of mixed modes, ideas of actions which define what acts of stabbing,

promise-keeping, murder, etc., are, he says that these ideas represent them-

selves; that is, they do not represent anything else, as ideas of substances do

(2.31.3). More precisely, he explains that the ideas of mixed modes only re-

present a group of ideas which define the criteria or characteristics of some

action. Such ideas can be said to represent actions (2.32.17). In 4.5.4, he

says it takes time and effort to get the ideas of modes to be exactly represented

to the mind.

So far, Locke has spoken of representation in four different ways: (1) the

way images or pictures represent things or persons; (2) as particular ideas can

be made to represent many individuals; (3) as representing the collection of


qualities by which we identify and name particular substances, and (4) as

mixed mode ideas represent themselves to the mind. All four of these uses

seem to fall roughly under the two first meanings of ‘represent’ in the OED:

«to bring into presence, especially to or before a person», and «to bring clearly

and distinctly before the mind by description or by an act of imagination».

There are three other more important occurrences of this term in Locke’s

Essay, uses which relate to signs and signification.

In the chapter on true and false ideas, Locke characterizes truth as the

joining or separating signs «according to the agreement, or disagreement of

the Things they stand for» (2.32.19). Signs are either ideas or words. Both

kinds of signs are now said to be ‘representatives’: «Truth lies in so joining, or

separating these Representatives, as the Things they stand for, do, in them-

selves, agree, or disagree: and Falshood in the contrary» (2.32.19; cf. 20.25).

In the chapter on the signification of words, he speaks of words as the sensible

of ideas, the ideas being the proper and immediate significations of those

marks (3.2.1). Those word-signs can strictly only signify ideas in the mind of

the speaker, but if speaker and hearer have ideas of the same things, if my

ideas correspond with your conceptions, then I can represent to myself your

ideas, assuming that we use the same words to signify those ideas (2.32.2).

That the sign-signified relation for Locke is one of representation is confirmed

by the final chapter of the Essay, in an oft-quoted sentence: «For since the

Things, the Mind contemplates, are none of them, besides it self, present to

the Understanding, ‘tis necessary that something else, as a Sign or Representa-

tion of the thing it considers, should be present to it: And these are ideas»

(4.21.4) These sign representatives are necessary both for the communication

of thoughts to one another, and to «record our thoughts for our own use». Locke

is describing in this section the third branch of human knowledge, the doc-

trine of signs. It is logic in the epistemic sense (the sense that pervades eigh-

teenth-century logics) which is concerned with “the Nature of Signs, the Mind

makes use of for the understanding of Things, or conveying its Knowledge to

others”. Understanding and communication are what ideas and words as signs (as

representatives) enable us to do.

III Ideas as Physical Objects in the Mind: One of the more striking and bizarre

passages in Malebranche’s De la recherche de la vérité contains his argument for

ideas as special objects, based on his assertion that objects themselves cannot

enter the mind, nor can the mind walk in the sky to get close to sun, moon

and stars. The assumption is that what is known must be present to the mind.

Since, Malebranche reasons, physical objects cannot be present to the mind,

there must be some other non-physical object which is present and, which in

some way (e. g, through representation) gives us knowledge or understanding

of physical objects. Arnauld made fun of the talk of the mind walking in the


sky, even as something impossible, because he thought it revealed a confusion

in Malebranche’s thought between physical or spatial presence and cognitive

. Locke also taunted the Bishop of Worcester for rejecting ideas, since,

so Locke charged, the alternative for the Bishop was that physical objects (e.

g., the tower of the Bishop’s cathedral) are in the mind when known.

There was, of course, a tradition in accounts of perceptual awareness

which did have objects existing in the mind; not literally but, for many scho-

lastics, essentially: the form of the object could be transferred to the mind via

sensible species which were then immaterialized and rendered intelligible.

Descartes adopted and adapted this tradition (as I have argued elsewhere)7,

with his doctrine of the objective reality of ideas. That reality was the object

itself as it exists in the mind. Objects exist in minds by being known. Des-

cartes’s ideas have a dual reality, as modes of mind and as objects known or

perceived. Arnauld insisted against Malebranche that this was Descartes’s

way of getting objects into the mind; they get there cognitively. With many

scholastics, the language of sensible and intelligible species has ontological

overtones. Descartes’s use of the formal-objective reality distinction for ideas

in Meditation III also has some ontic-sounding words: the being of objects in the

mind. He does not apply this distinction widely, he only suggests its use in

the account of perceiving ordinary objects. Nowhere does Descartes develop

a full-scale theory of perception. In Arnauld’s reformulation of this doctrine,

most of the ontic references have been replaced by cognitive phrases. One of

Arnauld’s definitions in his Des vrayes et des fausses idées (no. 5) explains that to

be objectively in the mind just means to be conceived.

As I remarked earlier, the 1705 writer placed Locke with those who fol-

lowed this analysis of objective reality. There are many passages in the Essay

and in his attack on Malebranche which support a case for locating Locke

with Arnauld. I have argued for this interpretation in Perceptual Acquaintance; I

do not want to repeat that argument here, but there is one passage in the Essay

that I would like to remind you of, since it is an echo, though slightly dis-

torted, of Descartes and objective reality. This passage is found in the Epistle

to the Reader, in a section added in the fourth edition of 1700. In that pas-

sage, Locke explains that by the terms ‘determinate’ and ‘determined’, instead

of ‘clear’ and ‘distinct’, he means «some object in the mind, and consequently

determined, i. e., such as it is there seen and perceived to be». That we are

hearing Cartesian echoes in the talk of ‘some object in the mind’, and of the

object ‘there seen and perceived’ is confirmed by Locke’s elaboration: «This, I


think, may fitly be called a determinate or determined idea, when, such as it is at

any time objectively in the mind and so determined there, it is annexed and

without variation determined to a name or articulate sound».

The echoes of Descartes’s adaptation of scholastic doctrine with his objec-

tive reality of ideas are found in other British writers, even in some whose

primary goal was not to analyze or even employ ideas. For example, in a

brief reference to ‘renewed representations’, William Duncan remarks that

these renewed representations «of what we have at any time perceived and

felt» become the means by which «things are again brought under the View of

the Mind, and seem to have a kind of Existence in it» (Logick, p. 22). Similar-

ly, Isaac Watts, who devotes only two pages to “the Nature of Ideas”, explains


It is not the outward Object, or Thing which is perceived, (viz.) the Horse, the

Man, etc., nor is it the very Perceptions or Sense, and Feeling, (viz.) of Hunger, or

Cold, etc., which is called the Idea; but it is the Thing as it exists in the Mind by

Way of Conception or Representation
, that is properly called the Idea, whether the

Object be present or absent (Logick, pp. 8-9).

In the previous century, Richard Burthogge talked of the immediate object of

perception as “Sense or Meaning”, and then went on to explain that

things are nothing [to us] but as they stand in our Analogie; that is, are nothing

to us but as they are known by us; and they are not known by us but as they

are in the Sense, Imagination, or Minde; in a word, as they are in our Faculties;

and they are in our Faculties not in their Realities as they be without them; no

nor so much as by Pictures and proper Representation, but only by certain

Appearances, and Phaenomena, which their impressions on the Faculties do ei-

ther cause or occasion in them (Organum Vetus & Novum, 1678, p. 13).

In the same year as a second book by Burthogge (An Essay upon Reason, 1694,

dedicated to Locke), the Cartesian work of Le Grand was translated into

English (An Entire Body of Philosophy). Le Grand borrowed almost verbatim

Descartes’s notion of objects being in the mind.

In writers such as Duncan, Watts, and Burthogge, the use of the notion of

objects existing in the mind is far from a central feature of their logics; passing

references to this notion have simply appeared in their writings as part of an

uncritical copying of an ongoing tradition. What these brief references indicate

is that the fundamental assumption of no cognition as a distance (the need for

what is known to be present to the mind) continued to influence accounts of

perception. It is also the case that, after the extended Arnauld-Malebranche

debate over this basic principle, the alternatives were clear: either the objects

we know or perceive must literally be in the mind (an absurd alternative, but


one given prominence by Malebranche), or some other way must be found to

understand the being of objects in the mind. Arnauld tried to explain to

Malebranche that all that was needed was a notion of cognitive presence. The

scholastic alternative of talking about the form of the object existing in matter

and in the mind required more ontological baggage than most seventeenth-

century writers were prepared to accept, although the word ‘form’ does reap-

pear occasionally, e. g, in some of Descartes’s clarifications to critics. The

philosophy of the Schools was in disrepute, both Descartes and Locke had rid-

iculed its occult properties, and forces, its substantial forms, etc. What

Arnauld injected into the discussions of perceptual awareness was a clear-

headed emphasis upon turning the ontic-sounding ‘objective reality’ of Des-

cartes’s analysis into a cognitive stress: to be or exist in the mind just means to

be understood or perceived.

This explication of ‘exist in the mind’ is picked up by Locke but even

more emphatically by Berkeley. What is more important about Berkeley’s

account of perception is his break with tradition in denying that ideas are

modes of mind. The full significance of this feature of Berkeley’s analysis has

not, I think, been understood. In this rejection, there is a clear echo of Des-

cartes’s notion of objective reality. Berkeley does not give much prominence

to this feature of his analysis, a fact that may be an indication that by the early

eighteenth century, Arnauld’s lesson had been absorbed. There are, so far as I

can discover, only two brief passages where Berkeley says ideas are not modes

of mind. To reinforce the point that ‘exist in’ has no literal sense, Philonous

tells Hylas that,

when I speak of objects as, existing in the mind or imprinted on the senses; I

would not be understood in the gross literal sense, as when bodies are said to

exist, in a place, or a seal to make an impression upon wax. My meaning is

only that the mind comprehends or perceives them; and that it is affected

from without, or by some being distinct from itself (Dialogues, III, p. 250).

This passage reads as if Berkeley had just been looking over the Arnauld-Male-

branche exchange. Earlier in the Dialogues, he explained that ideas exist in the

mind «not by way of mode or property, but as a thing perceived in that which

perceives it» (p. 237). Using the term ‘qualities’ rather than ‘ideas’, Berkeley

says in the Principles that they «are in the mind only as they are perceived by it,

that is, not by way of mode or attribute, but only by way of idea» (§ 49).

The notion that sensory ideas are the objects themselves existing in, that

is, known by, the mind is present in Hume’s complex, dialectical discussion of

our belief in body. He presents this view as the belief of ordinary persons.

For them, ideas are the very things themselves: the generality of mankind per-


ceive only one being, they «can never assent to the opinion of a double exist-

ence and representation. Those very sensations... are with them the true

objects, nor can they readily conceive that this pen or paper, which is immedi-

ately perceiv’d, represents another, which is different from, but resembling it»

(Treatise, p. 202). Hume accepted the dictum (he repeats it often) that «noth-

ing is ever really present to the mind besides its own perceptions» (Treatise,

p. 197). The challenge was to make this dictum consistent with our ordinary

belief that we see objects. Since we cannot see or otherwise perceive objects

without seeing or perceiving, the ordinary belief does not rest upon some

notion of access to objects apart from perceiving. Hume’s way of making

sense of the ordinary belief and of the obvious truth about cognitive access,

was to use an old distinction between specific and numerical difference. The

notion of a double existence of ideas and objects, where the object differs in

kind, specifically
, from our ideas is, Hume agreed with Berkeley, unintelligible.

But the rejection of this kind of double existence need not lead to idealism, to

the view that all that exists are minds and their ideas. The sensory ideas we

acquire under specific conditions, the appearances that we experience, are the

objects themselves as known, as related to our ‘analogie’ as Burthogge said,

but those ideas are numerically different from objects in the world. There is for

Hume still a double existence, but ideas and objects no longer differ specifical-ly, only numerically.

As we move horizontally over my chart from left to right, we switch from

ontic to epistemic senses of our term ‘idea’: ideas as modes or substances give

way to the representative role of ideas in knowledge. As we go vertically

down the categories of the chart on the ontological side, we move from gener-

al to specific accounts of ideas. Moreover, the vertical traverse down the left

hand column ends by replacing ideas as special entities with ideas as thoughts

and the contents of thoughts. The vertical move down the middle column

(from substance to objects in the mind) reveals writers struggling to throw off

the ontic baggage which the term had carried, replacing it with a cognitive or

semantic meaning.

The historical development of the way of ideas from Descartes to Hume

reflects both the horizontal and vertical lines of my chart. The fact that the

terminology of ideas in the works of Berkeley and Hume was so directly

linked with one of those important background themes, our knowledge of the

external world, is confirmation of the strong epistemic sense of that term in

British philosophy. That sense was also bound up with the question of scepti-

cism; if ideas are the immediate objects of awareness, how can we know that

they reveal to us the world of physical objects? Historians of modern philoso-

phy have been prone to find in the talk of ideas as representative, a displacement


of physical objects in our awareness. The representative theory of perception,

from Thomas Reid to present-day writers, has been interpreted as incompati-

ble with direct realism, even as leading directly to idealism.

I have been suggesting that Arnauld focused the debate on the crucial

issue: how to understand ‘presence to mind’? It was precisely the confusion

that Arnauld charged Malebranche with, the confusion of spatial with cognitive

presence, which forced writers in Britain to work out the implications of the

cognitive nature of ideas. But the larger tradition, a tradition running at least

back to scholastic authors, contained the notion of the object known by per-

ception being in the mind. The transition from sensible to intelligible species

of earlier accounts is reflected in the writers under my categories, Ic, IV, and

especially III, but in different language. ‘Existence of objects in the mind’ for

Arnauld, Locke, and Berkeley clearly means ‘perceived by’, ‘known’ or ‘cog-

nized’. These writers did not retain the Aristotelian and scholastic notion of

the form of objects which was able to exist in material or immaterial sub-

stances (in body or in mind). They had to work only with epistemic (seman-

tic and ideational) features. They were also intent upon finding a way of for-

mulating direct realism in epistemic terms.

If my analysis of the use of the term ‘idea’ in seventeenth and eighteenth

century philosophy is accepted, then perhaps we can understand the need to

move my category III classification into the epistemological part of my chart.

This movement is aided by (1) the stress found in many writers on ideas as

thoughts or perceptions; (2) the use of ‘representation’ as ‘making present to’;

and (3) Locke’s linking of representation to signs and signification. These

three features of the use of the term ‘idea’ mark a turning away from ideas as

distinct beings. Berkeley’s freeing ideas from being modes of mind was a very

important additional step in enabling Hume to elaborate upon Berkeley’s

attempts to turn ideas into things.


Classification of the Term Idea
ontological epistemological
I Modes II Substances IV Representations
(a) Of Body 1705 Essay 1705 Essay
Malebranche Browne
1705 Essay Reid Duncan
Chambers Norris Mayne
(b) Of Mind Watts
Mayne III Physical Objects
In the Mind
(c) Of Mind
Perceptions or Berkeley
Thoughts Burthogge
1705 Essay Duncan
Bentham Hume
Berkeley Watts
Authors and Works Cited

A Philosophick Essay concerning Ideas, According to Dr. Sherlock’s Principles, 1705.

Bentham, E., Bentham, E., Reflexions upon the Nature and Usefulness of Logick, Oxford 1740.

Bentham, E., Reflexions upon Logick (2nd ed.), Oxford 1755.

Berkeley, G., An Essay Towards a New Theory of Vision, Dublin 1709.

Browne, P., The Procedure, Extent, and Limits of Human Understanding, London 1728.

Burthogge, R., Organum Vetus & Novum. Or, A discourse of Reason and Truth, London 1678.

Burthogge, R., An Essay upon Reason and the Nature of Spirits, London 1694.

Chambers, E., Cyclopaedia: Or, An Universal Dictionary of Arts and Sciences, 2 vols, London 1728.

Duncan, W., The Elements of Logick, London 1748.

Hume, D., A Treatise of Human Nature, 2 vols, London 1739.

Locke, J., An Essay concerning Human Understanding, London 1690.

Malebranche, N., De la recherche de la vérité, Paris 1674-1678.

Mayne, Z., Two Dissertations concerning Sense and the Imagination, London 1728.

Norris, J., Cursory Reflections upon a Book call’d, An Essay concerning Human Understanding, London 1690.

Reid, T., Essays on the Intellectual Powers of Man, Edinburgh 1785.

Watts, I., Logick; or, The Right Use of Reason in the Enquiry after Truth, London 1725.

See my Locke and the Way of Ideas, London 1956, chapter I for a detailed account of

those British writers against whom Locke’s polemic on innate ideas was directed.
See my Perceptual Acquaintance from Descartes to Reid, Minneapolis 1983 and Oxford 1984,

chapter VI for an elaboration of the use of the term ‘idea’ in logic manuals and in works of

See p. 254.
The heading for this category may seem paradoxical. Surely, it will be said, no one

ever thought physical objects could exist in the mind or soul. Nevertheless, there were writers

in these two centuries (and even in our own) who based their accounts of ideas on the impossi-

bility of physical objects existing in the mind. Many have used this strange notion as a reason

for rejecting direct realism.
References to Hume are to the Selby-Bigge edition, updated by P. H. Nidditch, Oxford

1975. For an analysis of Hume’s ideas, see my Perceptual Acquaintance, chapter V.
References to Locke’s Essay concerning Human Understanding are given in this form, the

numbers indicating book, chapter, and section, in that order.
Support for this reading of Descartes can be found in my Perceptual Acquaintance, chap-

ter I.

. Date:

This page is copyrighted


  • There are currently no refbacks.