locke on primary and secondary qualities

Abstract:

In this paper, I argue that Book II, Chapter viii of Locke’s

Essayis a unified, self-consistent whole, and that the appearance of inconsistencyis due largely to anachronistic misreadings and misunderstandings. The keyto the distinction between primary and secondary qualities is that the formerare, while the latter are not, real properties, i.e., properties that exist inbodies independently of being perceived. Once the distinction is properlyunderstood, it becomes clear that Locke’s arguments for it are simple, validand (in one case) persuasive as well.

Introduction

In Book II, Chapter viii of An Essay Concerning Human Understanding,Locke distinguishes between primary qualities and secondary qualities.Among the former, Locke counts the determinables, Solidity, Size, Shape,Mobility, Number, and (sometimes) Texture and Situation (II.viii.9–10),and their determinates (i.e., particular sizes, shapes, etc.) (II.viii.23);among the latter, he counts colors, smells, tastes, sounds, light andtangible qualities (such as heat and coldness) (II.viii.10, 14, 17). Carefulreaders of the Essay have noticed that Locke makes a number of puzzlingclaims about the nature of this distinction. At the beginning of hisdiscussion, Locke defines qualities as powers to produce ideas in ourminds (II.viii.8), which would suggest that both primary and secondaryqualities are powers; but, although he often states that secondary qualitiesare powers (II.viii.10, 14, 22, 23), he sometimes implicitly distinguishesbetween powers and primary qualities (II.viii.23) and never explicitlystates that primary qualities are powers. At first, primary qualities are

297

LOCKE ON PRIMARYAND SECONDARY

QUALITIES

BY

SAMUEL C. RICKLESS

Pacific Philosophical Quarterly 78 (1997) 297–319 0031–5621/97/0300–0000© 1997 University of Southern California and Blackwell Publishers Ltd. Published by

Blackwell Publishers, 108 Cowley Road, Oxford OX4 1JF, UK and350 Main Street, Malden, MA 02148, USA.

defined as being “utterly inseparable” from bodies “in what estate soever[they] be” (II.viii.9); but later, Locke seems to ignore this way of drawingthe distinction, choosing instead to characterize primary qualities as beingsuch that our ideas of them resemble the bodies that possess them(II.viii.15). On occasion, the claims he makes appear to be flatlyinconsistent. For example, he tells us on the one hand that secondaryqualities are powers in bodies (II.viii.10), and, on the other hand, thatsecondary qualities aren’t really in bodies at all (II.viii.17).

It is not surprising, then, that scholars are deeply divided over how bestto interpret Locke’s remarks in this chapter. Concerning the questionwhether primary qualities are powers, one commentator writes that “thedistinction between primary and secondary qualities is meant to be adistinction between different kinds of powers, not a distinction betweenpowers and qualities or relations and qualities” (Curley, p. 445 – see alsoCampbell, p. 572). Others, by contrast, deny that the primary qualitiesof bodies are powers (Jackson, pp. 59–60, Mackie, pp. 14–15). As towhether Locke’s primary qualities should be thought of, in the firstinstance, as inseparable from bodies or as being such that bodies resembleour ideas of them, some opt for the former alternative (Alexander, p. 118), others for the latter (Mackie, p. 16). Finally, Locke’s claim thatsecondary qualities are in, but not really in, bodies continues to befuddle.Some interpret the claim that secondary qualities are not really in bodiesas equivalent to the statement that secondary qualities are mere powers(Bennett, p. 107, Mackie p. 13). But perhaps the most common reactionis to say that Locke simply chose a sloppy way of expressing the truththat our ideas of secondary qualities are not in bodies (Bennett, pp. 108–109, Curley p. 440).

Disagreement in regard to such fundamental issues merely sets the stagefor lack of consensus concerning other matters. For example, opinionsdiffer as to whether the examples at II.viii.16–21 are designed asarguments for (Mackie, pp. 17–23, Bolton, pp. 360–67), or as applicationsand explanations of (Alexander, p. 124), the primary/secondary qualitydistinction. And even those who agree that the examples should be treatedas arguments differ over how the arguments should be represented. Tosome, the arguments are inductive (Curley, p. 458, Mackie, pp. 17–23);to others, not (Bolton, pp. 360–67). And there seems to be no agreementon what the conclusions of these arguments are supposed to be.

Now it is tempting to see this lack of unanimity as revealing “thoseunfortunate inconsistencies for which Locke’s work is so famous” (Curley,p. 440). It may be that scholars have fastened onto different strands inLocke’s thought, strands which he simply did not recognize as beingmutually inconsistent. But it should strike us as strange that a man ofLocke’s intellect should produce a string of loosely organized remarks orembrace contradictory propositions in the very same chapter.

298 PACIFIC PHILOSOPHICAL QUARTERLY

© 1997 University of Southern California and Blackwell Publishers Ltd

The purpose of this paper is to argue that II.viii of Locke’s Essay is aunified, self-consistent whole, and that the appearance of inconsistency isdue largely to misreadings and misunderstandings. In Section 1, I attemptto clarify the distinction Locke drew. I argue that the key to the distinctionbetween primary and secondary qualities is that the former are, while thelatter are not, real properties, i.e., properties that exist in bodiesindependently of being perceived. In doing so, I defend the followingclaims: (a) that primary qualities are not powers, (b) that it is perfectlyconsistent for Locke to claim that secondary qualities are in, but not reallyin, bodies, and (c) that the thesis that primary qualities do, whereassecondary qualities do not, resemble the bodies that possess them, whenproperly understood, is (as Locke himself indicated) a trivial consequenceof his way of drawing the distinction. In Section 2, I reconstruct thearguments Locke employed to establish the reality of the distinction. Oncethe nature of the distinction is properly understood, it becomes clear thatLocke’s arguments for it are simple, valid, and (in one case) persuasiveas well.

1. The Primary/Secondary Quality Distinction

It is impossible to understand Locke’s way of distinguishing betweenprimary and secondary qualities without understanding the terms in whichit is drawn, the most important of these being “idea”, “power”, and “quality”.

An idea, for Locke, is “whatsoever the Mind perceives in itself, or isthe immediate object of Perception, Thought, or Understanding”(II.viii.8). Our minds are populated by countless ideas, many of whichrepresent mind-independent things. Ideas are immediately perceived, inthe sense that there is nothing mediating between a mind and the ideasit perceives: it is not by virtue of perceiving anything else that a mindperceives the ideas it perceives. By contrast, mind-independent entities aremediately perceived: for any mind-independent entity E and any mind Msuch that M perceives E, it is by virtue of perceiving an idea of E that Mperceives E.

Locke’s world of mind-independent entities includes substances (apples,gold, human beings), qualities (color, shape), and relations (being thesame as, being taller than). The ideas that represent these things are takento be either simple or complex. A simple idea is an idea which, “being initself uncompounded, contains in it nothing but one uniform appearance,or Conception in the mind, and is not distinguishable into different Ideas”(II.ii.1). Complex ideas, then, are composed of other ideas, themselveseither simple or complex. Locke holds, for example, that an idea of anapple is “distinguishable” into various other ideas, such as the (simple)ideas of redness, roundness, sweetness, and crunchiness.

LOCKE ON PRIMARY AND SECONDARY QUALITIES 299


Primary and secondary qualities are, one and all, perceptible qualitiesof bodies, i.e., solid extended substances. In accord with Locke’s theoryof perception, minds mediately perceive these qualities by immediatelyperceiving ideas that represent them. Thus, it is by (immediately)perceiving the ideas of squareness and redness that minds (mediately)perceive squareness and redness.

Among the relations that obtain between substances, Locke lists thepowers of substances to causally affect, or to be causally affected by, othersubstances. He calls the former powers “active” and the latter powers“passive”. Thus, whereas “Fire has a power to melt Gold”, “Gold has apower to be melted [by Fire]” (II.xxi.1). In general, if a substance S is ableto causally affect another substance T, the active power relating S and Tis located in S, while the passive power relating S and T is located in T;whereas the power to blanch wax is located in the sun, the power to beblanched by the sun is located in wax.

Locke claims that there are two kinds of active powers relating bodiesand the minds that perceive them. In the first place, there are powers“whereby [bodies] are fitted [...] by immediately operating on our Bodies,to produce several different Ideas in us” (II.viii.26): these are thesecondary qualities. For example, redness is the power “to operate aftera peculiar manner on any of our Senses, and thereby produce in us” theidea of redness (II.viii.23). In the second place, there are powers wherebybodies are fitted, “by operating on other Bodies [...] to render them capableof producing Ideas in us, different from what before they did” (II.viii.26):these are bare powers. Thus, fire has the bare power to melt gold, i.e.,has an active power to causally affect gold in such a way that gold losesthe power to produce the idea of hardness, but gains the power to producethe idea of fluidity, in our minds.

Now Locke famously distinguishes between ideas and qualities:

Whatsoever the Mind perceives in itself, or is the immediate object of Perception, Thought,or Understanding, that I call Idea; and the Power to produce any Idea in our mind, I callQuality of the Subject wherein that power is. (II.viii.8)

Given that qualities are defined to be powers to produce ideas, and giventhat colors, etc. are powers to produce ideas, it makes perfect sense forLocke to say that colors, etc. are qualities in the sense just defined. It alsomakes perfect sense for Locke to say that colors, etc. are secondary, sinceLocke holds that a body has the power to produce a particular idea inour minds by virtue of the fact that it has a particular number of partsthat possess particular shapes, sizes, and degrees of motion. Qualities ofbodies are called “secondary” because they depend on (arise from) theprimary qualities of their parts (II.viii.10, 14, 23).



With few exceptions, most Locke scholars are likely to agree with whatI have said thus far. Unfortunately, they agree about little else. In thefirst place, there is disagreement as to why Locke called Shape, Size, etc.“qualities”. The disagreement arises out of the following conundrum.Shape, for Locke, is a primary quality. But, if Shape is a primary quality,then Shape is a quality. Moreover, if Shape is a quality, then, inaccordance with Locke’s definition of “quality”, Shape must be a powerto cause an idea (presumably, the idea of Shape) in our minds. But, sincesecondary qualities are powers to produce ideas in our minds, it followsthat primary and secondary qualities are both powers to produce ideasin our minds. Hence, there is no distinction in kind between primary andsecondary qualities. But Locke insists that there is such a distinction. (Thisreasoning could be reproduced for every one of the ‘primary’ qualities.)

Scholars have reacted to this piece of reasoning by denying one or otherof its various assumptions. One commentator claims that primaryqualities are not qualities, at least not in the sense that Locke defines theword “quality”: they are, rather, intrinsic (i.e., inseparable) properties ofbodies (Mackie, p. 13). Another maintains that primary qualities arequalities (in Locke’s sense of the word), and that the difference betweenprimary and secondary qualities is a difference between different sorts ofpowers (Curley, p. 445 – see also Campbell, p. 572). Yet another arguesthat the difference between primary and secondary qualities is a differencebetween different sorts of intrinsic properties (Alexander, p. 118).

In order to solve this exegetical problem, we need to take a close lookat the way in which Locke first characterises the distinction betweenprimary and secondary qualities. Here is what he says:

Qualities thus considered in Bodies are, First such as are utterly inseparable from the body,in what estate soever it be; such as in all the alterations and changes it suffers, all the forcecan be used upon it, it constantly keeps; and such as Sense constantly finds in every particleof Matter, which has bulk enough to be perceived, and the Mind finds inseparable fromevery particle of Matter, though less than to make itself singly be perceived by our Senses... These I call original or primary Qualities of Body, which I think we may observe toproduce simple Ideas in us, viz. Solidity, Extension, Figure, Motion, or Rest, and Number.Secondly, Such Qualities, which in truth are nothing in the Objects themselves, but Powersto produce various Sensations in us by their primary Qualities, i.e. by the Bulk, Figure,Texture, and Motion of their insensible parts, as Colours, Sounds, Tasts, etc. These I callsecondary Qualities. (II.viii.9–10)

This way of contrasting primary and secondary qualities strongly suggests,first, that primary qualities are, but secondary qualities are not,inseparable properties of bodies, and, second, that secondary qualitiesare, but primary qualities are not, powers to produce ideas in our minds.For, if Locke had meant to distinguish between different sorts of powers,



he would surely have said that primary qualities are powers, and if hehad meant to distinguish between different sorts of inseparable properties,he would surely have said that secondary qualities are inseparableproperties.

Admittedly, this still leaves us with our original problem, namelyLocke’s accepting both that primary qualities are qualities and thatprimary qualities are not qualities (since qualities are powers and primaryqualities, it seems, are not powers). The solution, I submit, is that Lockeuses the word ‘quality’ in two different senses. Let us say that X is aquality1 iff X is a property.1 Thus, primary qualities, such as Shape andSize, and secondary qualities, such as redness and sweetness, are qualities1.Among the qualities1, Shape and Size are counted as primary by virtueof the fact that they are inseparable from bodies. But Locke recognizesthat “the general term Quality, in its ordinary acception, comprehendsColours, Sounds, Tastes, Smells, and tangible Qualities, with distinctionfrom Extension, Number [and] Motion” (III.iv.16), decides to “complywith the common way of speaking” (II.viii.10), and, in line with thisdecision, defines qualities to be powers to produce ideas in our minds. Inso doing, Locke expressly uses the word ‘quality’ in a different sense, asense which might be more perspicuously represented by means of adifferent subscript. We might say, then, that X is a quality2 iff X is apower to cause an idea in our minds. Thus, secondary qualities arequalities2, but primary qualities are not. On this analysis, it is misleadingto say, along with some scholars (Jackson, p. 55), that redness is a power,but not a quality. One should say instead that redness is a quality2, butnot a primary quality1. Moreover, rather than saying that Shape is aquality, but not a power, one should say that Shape is a primary quality1,but not a quality2.

Although this way of reading Locke avoids a number of possiblemisunderstandings, it does not avoid all. For example, Locke’s way ofdescribing the distinction between primary and secondary qualities atII.viii.9–10 has been criticized on the grounds that it depends on “anunfair comparison between determinable qualities on the primary sideand determinate ones on the secondary side” (Mackie, p. 20). Thus, it issaid that, although determinable primary qualities (such as Shape andSize) are inseparable from bodies, so are determinable secondary qualities(such as Color and Taste); and, although determinate secondary qualities(such as redness and sweetness) are separable from bodies, so aredeterminate primary qualities (such as squareness and the property ofbeing at rest). Those who accept this criticism are drawn to the descriptionof the distinction in II.viii.15, where Locke says that “the Ideas of primaryQualities of Bodies, are Resemblances of them [...] but the Ideas, producedin us by these Secondary Qualities, have no resemblance of them at all.” Andthis leads to statements about Locke’s distinction such as the following:



The principle of this primary/secondary distinction is that the ideas of primary qualitiesresemble the grounds of the powers to produce them while the ideas of secondary qualitiesdo not. (Mackie, p. 16).

But to read Locke as holding that determinable secondary qualities areinseparable from bodies is to miss the point. It is clear that Locke believedthat all bodies necessarily possess some shape, some size, and some degreeof motion. It follows, by Locke’s definition of ‘primary quality’, that thedeterminables Shape, Size, and Motion-or-Rest are primary qualities. Butit is also clear that Locke did not believe that all bodies necessarily possesssome color, taste, smell or sound. For example, at II.viii.19, Locke saysthat porphyry “has no colour in the dark”. By Locke’s definition of‘primary quality’, then, it follows that the determinable Color is not aprimary quality.

It is difficult not to explain this sort of criticism as the product of well-intentioned anachronism. I suspect that those who dismiss Locke’s firstway of describing the primary/secondary quality distinction do so on thegrounds that it would have been highly unreasonable for him to denysomething that currently seems obvious to them, namely that bodiesalways retain some color, some taste, some smell, and some sound. But,counter-intuitive as it may seem, this is precisely what Locke was at painsto deny.

So Locke gives us a way of distinguishing among determinableproperties: whereas Shape, say, is a property which is inseparable frombodies, Color is not. But where does this leave the determinate properties?Surely the critics are right to point out that redness and squareness areboth separable from bodies: although it is possible for a red body to loseits redness, it is also possible for a square body to lose its squareness.

The truth is that there is more to the distinction than the issue ofseparability. To see this, we need to concentrate on Locke’s statementsabout the secondary qualities. As we’ve seen, Locke defines secondaryqualities to be powers to produce ideas in our minds. But Locke insistsin numerous places (II.viii.10, 14, 22, 23, 24, 26) that secondary qualitiesare nothing but powers, and claims in at least one place (II.viii.24) thatthe primary qualities existing in an object are “something more than merePowers in it.” This is an important part of the distinction, and one whichwill enable us to clarify the relation between Locke’s various statementsabout it.

Let us begin by taking a closer look at the passage from II.viii.24:

The Idea of Heat, or Light, which we receive by our Eyes, or touch from the Sun, arecommonly thought real Qualities, existing in the Sun, and something more than mere Powersin it.



It is plausible to read this sentence as containing an implicit ‘as such’between the words ‘and’ and ‘something’. If the addition is made, thesentence reads:

The Idea of Heat, or Light, which we receive by our Eyes, or touch from the Sun, arecommonly thought real Qualities, existing in the Sun, and, as such, something more thanmere Powers in it.

What I am suggesting is that, for Locke, primary qualities are somethingmore than mere powers because they are real qualities1 (i.e., realproperties). If this is right, then secondary qualities are nothing but powersbecause, although they are powers, they are not real qualities1. On thisreading, the answer to the question whether a property is a primary orsecondary quality hinges, at least in part, on the answer to the questionwhether the property is a real quality1.

We are left with the following picture. First, Locke draws a distinctionbetween inseparable and separable determinable qualities1 of bodies. Hecalls the former ‘primary qualities’, and counts Shape, Size, Mobility, etc.among them. Second, Locke draws a distinction between real and non-real determinate qualities1 of bodies. He calls the former ‘primaryqualities’, and counts particular shapes, sizes, etc. among them. Withinthe category of non-real determinate qualities1 of bodies, he draws afurther distinction between powers to cause ideas in our minds and powersto cause perceivable changes in bodies. He calls the former (i.e., the non-real determinate powers to cause ideas in our minds) ‘secondary qualities’,and the latter (i.e., the non-real determinate powers to cause perceivablechanges in bodies ) ‘bare powers’. Among the secondary qualities, Lockecounts particular colors, tastes, sounds, smells, light, and tangibleproperties (such as heat and coldness). The secondary qualities are nothingbut powers by virtue of the fact that they are powers, but not realqualities1.

I hope we are beginning to see what Locke is trying to do. Theinseparable/separable property distinction takes center stage at the levelof the determinables, but disappears at the level of the determinates. Bycontrast, the real/non-real property distinction takes center stage at thelevel of the determinates, but is not emphasized at the level of thedeterminables. Still, the picture I am painting is far from complete, forthe following important questions have yet to be answered. First, whatis it for a quality1 to be real? And, second, is there any reason to thinkthat Locke accepted the claim that determinate shapes, etc. are, butdeterminate colors, etc. are not, real qualities1?

Taking the first question first, let us begin by looking at how Lockedefines the term ‘real quality’. The definition appears at II.viii.17:



The particular Bulk, Number, Figure, and Motion of the parts of Fire, or Snow, are really inthem, whether any ones Senses perceive them or no: and therefore they may be called realQualities, because they really exist in those Bodies.

I understand Locke to be saying that reality among qualities1 is a matterof independence from perceivers: a property P which is in an object O isa real quality1 in O (or, as Locke sometimes puts it, is really in O) iff Pexists in O independently of anyone’s perceiving P to be in O. It followsthat, if P’s being in O depends on someone’s perceiving P to be in O, thenP is not a real quality1 in O, even if P is located in O. Thus, squarenesscounts as a real quality1 by virtue of the fact that squareness is a propertythat exists in square objects independently of anyone’s perceiving theseobjects to be square. (Notice that this understanding of ‘really in’ explainswhy the statement that some properties are in, but not really in, bodiesis not self-contradictory.)

But what of, say, redness and sweetness? Is there any evidence thatLocke took these qualities1 to be non-real? Indeed there is. For example,consider the following statement at II.viii.17:

Take away the Sensation of them; let not the Eyes see Light, or Colours, nor the Ears hearSounds; let the Palate not Taste, nor the Nose Smell, and all Colours, Tastes, Odors, andSounds, as they are such particular Ideas, vanish and cease …

Now some read this passage as saying merely that our ideas of colors, etc.vanish when there are no perceivers to perceive them. But that Locke issaying more than this is evident from his way of repeating the same pointelsewhere. Thus, at II.xxiii.11, he writes:

Had we Senses acute enough to discern the minute particles of Bodies, and the realConstitution on which their sensible Qualities depend, I doubt not but they would producequite different Ideas in us; and that which is now the yellow Colour of Gold, would thendisappear … [Moreover,] Sand, or pounded Glass, which is opaque, and white to the nakedEye, is pellucid in a Microscope; and a Hair seen this way, loses its former Colour, and isin a great measure pellucid, with a mixture of some bright sparkling Colours … (emphasisadded)

But the most striking passage to make the same point (this time, aboutlight and heat) appears at II.xxxi.2:

Were there no fit Organs to receive the impressions Fire makes on the Sight and Touch;nor a Mind joined to those Organs to receive the Ideas of Light and Heat, by thoseimpressions from the Fire, or the Sun, there would yet be no more Light, or Heat in theWorld, than there would be Pain if there were no sensible Creature to feel it, though theSun should continue just as it is now, and Mount Ætna flame higher than ever it did.



Counter-intuitive as it may seem to many of us, there is no doubt thatLocke believed that an object’s being red/sweet/hot depends on someone’sperceiving it to be red/sweet/hot. Remove all perceivers, and objects thatwere previously red/sweet/hot are no longer so.

Now I said above that Locke defines a real quality1 to be a propertywhich exists in objects independently of anyone’s perceiving it to be inthem. It should not surprise us, then, that scholars who refuse to believethat Locke denied that bodies lose their colors, tastes, etc. when there areno perceivers either ignore Locke’s definition of ‘real quality’ or interpretit differently. Thus, one commentator writes:

Locke sometimes says that the secondary qualities are not really in external objects and donot exist at all whenever, for one reason or another, they are not being perceived(II.viii.17–19). But when he says this he seems to be identifying the quality with the idea itproduces and forgetting his official doctrine about secondary qualities, which is that theyare powers which bodies have to produce sensations (Curley, p. 440).

In defense of this interpretation, it is common to focus on II.viii.17(quoted above), and the following passage from II.viii.19 (Alexander, pp. 168–69):

Hinder light but from striking on [porphyry], and its Colours Vanish; it no longer producesany such Ideas in us: Upon the return of Light, it produces these appearances on us again.Can any one think any real alterations are made in the Porphyre, by the presence or absenceof Light; and that those Ideas of whiteness and redness, are really in Porphyre in the light,when ’tis plain it has no colour in the dark?

These passages are cited as evidence for the claim that, in saying thatcolors, tastes, etc. vanish when they are not perceived, Locke is sayingno more than that no ideas of colors are produced under these conditions.

This interpretation would be quite persuasive, were it not for the factthat Locke admits that, when speaking of ideas as being in objects, heuses the word ‘idea’ to mean the same as ‘quality’. Thus, at II.viii.8, hesays:

which Ideas, if I speak of sometimes, as in the things themselves, I would be understood tomean those Qualities in the Objects which produce them in us.

(Notice that Locke is not saying that, when he speaks of ideas as beingin things, he should sometimes be understood to mean those qualities inthe objects which produce ideas in us.) Therefore, when Locke suggestsat II.viii.19 that “those Ideas of whiteness and redness, are [not] really inPorphyre”, he ought to be read as saying that the qualities1 of whitenessand redness are not really in porphyry. And when he says at II.viii.17that “all Colours, Tastes, Odors, and Sounds, as they are such particular



Ideas, vanish and cease” when they are not perceived, he ought to be readas saying that the relevant qualities1 disappear under those conditions.

It is worth mentioning that some commentators have fallen for anotheranachronism that may explain their refusal to accept that Locke basedhis distinction between determinate shapes and colors on the thesis thatcolors do, but shapes do not, vanish when all perceivers are removed. Itis natural to analyze statements about powers counterfactually. On thissort of analysis, to say that gold is soluble (i.e., has the power to dissolve)in Aqua Regia is just to say that, if gold were placed in Aqua Regia, itwould dissolve. Thus, given Locke’s claim that colors are powers, it istempting to read him as putting forward a counterfactual analysis ofcolor-statements. Thus, one influential scholar writes that Locke takes ‘xis green’ to mean “roughly the same as ‘If x were sunlit and were in theline of vision of a normal open-eyed human, he would have a visual fieldof such and such a kind’” (Bennett, p. 94). But anyone who accepts thisanalysis of ‘x is green’ must also accept that porphyry retains its rednessin the dark, since it remains true in the dark that porphyry would producean idea of red in the mind of a normal open-eyed human if it were sunlitand placed in his line of vision. Thus, if Locke is understood to accept thisanalysis, he cannot be read as holding that colors disappear when notperceived without also being read as inconsistent, which (it should notsurprise us to learn) is precisely how some have read him (Curley, p. 440).

But there is no evidence to suggest that Locke would have accepted acounterfactual analysis of color-statements. Locke does say that colorsare powers to produce ideas in our minds. But he also says that the mind“comes by” the idea of power when it “considers in one thing thepossibility of having any of its simple Ideas changed, and in another thepossibility of making that change” (II.xxi.1). Thus, his statement thatcolors are powers may be taken to mean no more than the following: forany color C and object O, O is (or has) C under circumstances W iff itis possible for O to produce ideas of C in the minds of humans undercircumstances W. In that case, inconsistency can be avoided. For thisaccount of color entails that objects aren’t green when placed in the dark,since objects in the dark cannot produce an idea of greenness in our mindsunder those conditions. And a similar account of heat entails that the sunisn’t hot when there are no perceivers to perceive it, since the sun can’tproduce ideas of heat when there are no minds to receive them.

As against this, one scholar has pointed to what he takes to be textualevidence for the claim that Locke “recognises that the power [to producesensations in our minds] continues to exist even when the conditions forits manifestation are not fulfilled” (Curley, p. 440 – see also McCann, p.66). The relevant passage appears at II.viii.19:



[Porphyry] has, indeed, such a Configuration of Particles, both Night and Day, as are aptby the Rays of Light rebounding from some parts of that hard Stone, to produce in us theIdea of redness, and from others the Idea of whiteness …

But it would be a mistake to read this sentence as saying that porphyryretains its redness and whiteness in the dark. For Locke is merely sayingthat, at night, porphyry retains the same configuration of particles thatproduces the ideas of redness and whiteness in us by day. But, from thefact that porphyry has the same configuration of particles, both night andday, it does not follow that porphyry has the same powers, both nightand day.

Let us take stock. On the interpretation of II.viii that I have beendefending, the determinable property, Shape, is a primary quality byvirtue of the fact that it is a quality1 which is inseparable from bodies,while the determinate property, squareness, is a primary quality becauseit is real, in the sense that its existence in objects does not depend on itsbeing perceived to exist in them. By contrast, the determinate property,redness, is a secondary quality because it is a non-real quality2 (i.e., powerto cause ideas in our minds).

The careful reader will no doubt have noticed that the determinables,Color, Taste, etc. have been left out of the picture thus far. We havealready noted that Locke takes these properties to be separable frombodies: porphyry, he tells us, has no color in the dark. But it should alsobe noted that these determinables are not real properties, and this for thevery same reason that they are separable: porphyry has no color in thedark because it is not perceived as having any color under theseconditions. Does it then follow that Color, Taste, etc. are secondaryqualities? The answer is “no”. For secondary qualities are defined to bepowers to produce ideas in our minds, and there is no evidence that Locketook Color, Taste, etc. to be powers of any sort. In fact, there is evidenceagainst the supposition. For Locke says that secondary qualities “serveprincipally to distinguish Substances one from another” (II.xxiii.8), andit is clear that it is impossible to distinguish between two substances bypointing out that one does, but the other does not, have Color, Taste,and so on. Thus, although Locke never directly compares determinableswith determinates, it must be admitted that he leaves some determinables(namely, Color, Taste, etc.) out of his classificatory scheme entirely.

We are now in a position to see that Locke had the conceptual resourcesto draw the primary/secondary quality distinction differently. For, sinceall properties that are inseparable from bodies are real (i.e., exist in objectsindependently of anyone’s perceiving them to be in those objects), Lockemight simply have said (i) that P is a primary quality iff P is a(determinable or determinate) real property, and (ii) that P is a secondaryquality iff P is a determinate non-real property which depends for its



existence on real properties. On this account, Shape counts as a primaryquality, not by virtue of the fact that it is a property inseparable frombodies, but rather because its existence in bodies does not depend onanyone’s perceiving it to exist in those bodies. Although Locke himselfdoes not present the primary/secondary quality distinction at the level ofthe determinables as hinging on the real/non-real distinction, it is notunreasonable to suppose that it is this way of drawing the distinction atwhich he is aiming.

With respect to Locke’s way of drawing the distinction, there is onemore issue to be settled. I said earlier that Locke sometimes describes thedifference between primary and secondary qualities in terms ofresemblance: ideas of primary qualities do, but ideas of secondaryqualities do not, resemble the bodies in which the qualities are located.And several scholars (Curley, pp. 450–51, Mackie, p. 16) have taken this“resemblance thesis” to lie at the heart of Locke’s distinction. What, then,is the relation between the distinction as I have described it and theresemblance thesis?

In order to answer this question, we need to examine the way in whichLocke introduces the thesis. Here is what he says:

What I have said concerning Colours and Smells, may be understood also of Tastes andSounds, and other the like sensible Qualities; which, whatever reality we, by mistake, attributeto them, are in truth nothing in the Objects themselves, but Powers to produce variousSensations in us, and depend on those primary Qualities, viz. Bulk, Figure, Texture, andMotion of parts; as I have said. From whence I think it is easie to draw this Observation,That the Ideas of primary Qualities of Bodies, are Resemblances of them, and their Patternsdo really exist in the Bodies themselves; but the Ideas, produced in us by these SecondaryQualities, have no resemblance of them at all. There is nothing like our Ideas, existing in theBodies themselves. (II.viii.14–15)

This passage reveals that Locke does not introduce the resemblance thesisas the key to understanding the distinction between primary andsecondary qualities. Rather, he introduces it as an “observation” that issupposed to follow from the statement that secondary qualities in objects“are in truth nothing in the Objects themselves, but Powers to producevarious Sensations in us.” That he does so is consistent with theinterpretation of the distinction that I have reconstructed on his behalf.But it also raises a further question: on what grounds does Locke takethe resemblance thesis to follow from his previous description of theprimary/secondary quality distinction?

Look again at Locke’s way of putting the resemblance thesis. AtII.viii.15, Locke makes two separate points about the primary qualities.First, he says that “the Ideas of primary Qualities of Bodies, areResemblances of them.” Second, he says that “Patterns [of primary



qualities] do really exist in the Bodies themselves.” Next, he makes twocorresponding points about the secondary qualities: first, that “the Ideas,produced in us by these Secondary Qualities, have no resemblance of themat all,” and, second, that “there is nothing like our Ideas [of secondaryqualities] existing in the Bodies themselves.” Putting these points together,we obtain the following thesis: ideas of primary qualities of bodies do,but ideas of secondary qualities of bodies do not, resemble real qualities1

in these bodies. This is Locke’s resemblance thesis.It should be noted that Locke repeats this thesis in several places. Thus,

at II.viii.22, he writes that, with the primary/secondary quality distinctionin hand, “we also may come to know what Ideas are, and what are notResemblances of something really existing in the Bodies, we denominatefrom them” (emphasis added). And, at II.viii.25, Locke writes that we aretempted to suppose, albeit mistakenly, that ideas of secondary qualities“are the resemblances of something really existing in the Objectsthemselves” (emphasis added). In these passages, Locke is saying morethan that ideas of secondary qualities of bodies do not resemble (qualities1

in) these bodies. He is saying, in addition, that there are no real qualities1

in bodies to which ideas of secondary qualities bear any resemblance.If this is the proper interpretation of Locke’s resemblance thesis, then

it becomes easy to see why Locke would have thought it to be aconsequence of the primary/secondary quality distinction. I have alreadyargued that it is part of this distinction that determinate primary qualitiesare, but secondary qualities and bare powers are not, real qualities1, i.e.,properties which are such that their existence in objects does not dependon their being perceived to be in those objects. Now consider the ideasof secondary qualities produced in us by certain bodies. If these ideas areto resemble real qualities1 in the bodies that produce them in us, then,given that secondary qualities and bare powers are not real qualities1, theideas must resemble the primary qualities in these bodies. But, on anyconception of what resemblance might amount to, ideas of secondaryqualities do not resemble any primary qualities. It follows that ideas ofsecondary qualities of bodies do not resemble any real qualities1 in thosebodies.

That the resemblance thesis (introduced at II.viii.15) follows from theprimary/secondary quality distinction drawn earlier (at II.viii.9–10) turnsout to be a fairly trivial point. But this should count for, rather thanagainst, my interpretation of the various claims that Locke was trying tomake. For Locke presents the resemblance thesis as an “observation”which it is “easie to draw” from his statement that secondary qualitiesare “nothing […] but Powers to produce various Sensations in us.” Andit is difficult to see how Locke could have thought the thesis to be a triflingconsequence of the distinction, if the interpretation of Locke’s remarksfor which I have argued is rejected.



“But,” I may be asked, “if the resemblance thesis is merely a corollaryof the primary/secondary quality distinction, then why did Locke makesuch heavy weather of it?” I imagine that the answer to this question isthat II.viii was written in part as a response to the Schoolmen of Locke’stime, whose Aristotelian theory of perception relied on both parts of theresemblance thesis. According to this theory, shapes and colors are“sensible forms” that can be transmitted through some medium to ourvarious sense-organs, whence they are communicated to, and therebycome to “inform”, the soul. These “sensible forms” are, in Locke’sparlance, real qualities1, i.e., properties which are such that their existencein objects does not depend on their being perceived to be in those objects.Moreover, given that the same “sensible form” may be found in a bodyand in any soul that perceives that body, there is a fairly straightforwardsense in which there is something in the soul which resembles somethingin the body it perceives. Thus, if Locke is right, the scholastic theory ofperception is mistaken. And I suspect that it is precisely in order to makethis point that Locke gives the resemblance thesis such a prominent placein his discussion of the primary/secondary quality distinction (see alsoMcCann, pp. 63–64).

2. Locke’s Arguments for the Primary/Secondary QualityDistinction

If Locke’s distinction between primary and secondary qualities is as Ihave described it, then how are we to understand the various examplesand cases Locke uses at II.viii.16–21? Are these cases meant to beapplications of, or arguments for, the distinction?

The answer, I believe, is that there are four arguments at II.viii.16–21.The first of these arguments, contained in II.viii.16 and II.viii.18, isdesigned to show that there are no good reasons to deny the existence ofthe primary/secondary quality distinction. The other three arguments,contained in II.viii.19–21, are designed to establish that the distinctionexists.

Let us begin with the passage at II.viii.16:

Flame is denominated Hot and Light; Snow White and Cold; and Manna White and Sweet,from the Ideas they produce in us. Which Qualities are commonly thought to be the samein those Bodies, that those Ideas are in us, the one the perfect resemblance of the other, asthey are in a Mirror; and it would by most Men be judged very extravagant, if one shouldsay otherwise. And yet he, that will consider, that the same Fire, that at one distance producesin us the Sensation of Warmth, does at a nearer approach, produce in us the far differentSensation of Pain, ought to bethink himself, what Reason he has to say, That his Idea ofWarmth, which was produced in him by the Fire, is actually in the Fire; and his Idea of



Pain, which the same Fire produced in him the same way, is not in the Fire. Why is Whitenessand Coldness in Snow, and Pain not, when it produces the one and the other Idea in us;and can do neither, but by the Bulk, Figure, Number, and Motion of its solid Parts?

I take it that Locke’s point is this. It is tempting to think that warmth isa real quality1 existing in fire, on the grounds that it is a quality1 whichis transmitted from fire to our senses, and thereafter to our minds. (Infact, it is in part for this reason that the Scholastic theory of perceptiongained such widespread acceptance.) But this reason for thinking thatwarmth is a real quality1 is, in fact, a poor one. For, given that fireproduces the idea of pain in us in much the same way that it producesthe idea of warmth in us, it is possible to make a parallel case for theabsurd, and widely rejected, claim that pain is a real quality1 in fire.However, shape, size, etc. are universally acknowledged to be realqualities1. Therefore, unless we are given other grounds for thinking thatwarmth is a real quality1, there is no good reason for us to deny that thereis a significant difference between warmth, on the one hand, and shape,size, etc., on the other. Substantially the same point is made at II.viii.18,where Locke argues that there is no more reason for us to think thatwhiteness and sweetness are real qualities1 in manna than there is to thinkthat pain and sickness are real qualities1 in the very same substance.

It is not until II.viii.19 that Locke provides a direct argument for theprimary/secondary quality distinction. The reasoning is contained in thefollowing excerpt from that section:

Let us consider the red and white colours in Porphyre: Hinder light but from striking onit, and its Colours Vanish; it no longer produces any such Ideas in us: Upon the return ofLight, it produces these appearances on us again. Can any one think any real alterationsare made in the Porphyre, by the presence or absence of Light; and that those Ideas ofwhiteness and redness, are really in Porphyre in the light, when ’tis plain it has no colour inthe dark?

Concerning how best to interpret these remarks, there is a marked lackof consensus. One commentator sees the passage as providing support forthe “plausible hypothesis that different surface textures merely reflectdifferent rays of light” (Mackie, p. 21). Another reads Locke as supporting“his belief that the appearance to sight of both the colour and theextension in the light, the appearance to touch of extension in the darkand the failure of sight to show us colour in the dark, can all be explainedin terms of the action upon us of configurations of corpuscles” (Alexander,p. 127). Yet another reads the passage as an argument for the thesis that“the ideas [of whiteness and redness] are not like any quality in the stone”(Bolton, p. 360).



Now there is no doubt that Locke accepted each of the theses that thesescholars attribute to him: Locke certainly believed that surface texturesreflect rays of light, that the appearance to sight of color in the light(among other things) is due to the action of insensible corpuscles, andthat ideas of color do not resemble qualities1 of objects. But I would arguethat the porphyry passage is not meant to establish, or even providesupport for, any of these beliefs. Rather, Locke should be read as arguingfor the claim that the colors of porphyry are not real qualities1 in it.

The porphyry argument may be reconstructed as follows:

(1) In the dark, porphyry is neither red nor white.(2) The real qualities1 possessed by porphyry in the dark are the same

as the real qualities1 possessed by porphyry in the light.

∴(3) Redness and whiteness are not real qualities1 in porphyry, whetherin the light or in the dark.

That this argument is valid is easily seen. For suppose that redness andwhiteness are real qualities1 in porphyry in the dark. In that case, porphyrymust be red and white in the dark. But it isn’t, if (1) is true. So rednessand whiteness are not real qualities1 in porphyry in the dark. Now supposethat redness and whiteness are real qualities1 in porphyry in the light. Inthat case, by (2), redness and whiteness are also real qualities1 in porphyryin the dark. But then porphyry must be red and white in the dark. But,again, it isn’t, if (1) is true. Therefore, (1) and (2) both entail that rednessand whiteness are not real qualities1 in porphyry, whether in the light orin the dark.

Now this argument, on its own, does not establish that there is adistinction between properties that are, and properties that are not, realqualities1 in objects. But the argument does not require much in the wayof supplementation in order to count as an argument for theprimary/secondary quality distinction. Locke need only add somethingthat his contemporaries all accepted, namely that shapes, sizes, etc. arereal qualities1, and point out that a parallel argument cannot be used toprove that shapes, sizes, etc. are not real qualities1. For it is clear thatporphyry retains its shape, size, etc. in the dark. Thus, the main reasonfor differentiating between, say, redness and squareness is that there areconditions under which objects lose the former, conditions under whichthe same objects retain the latter. Porphyry loses its redness, but retainsits shape, when the lights are turned off.

It is important to notice that the claim that porphyry has no color inthe dark is an essential premise in Locke’s argument at II.viii.19. Withoutthis premise, Locke’s argument is invalid. Those scholars who aim toattribute cogent arguments to Locke, but who cannot bring themselvesto accept that Locke would have endorsed this premise, must (and do)



read the passage as making a completely different point. But it should beclear that the alternatives canvassed above stray from the wording of thetext: Locke’s argument at II.viii.19 contains no reference to surfacetextures or rays of light, no reference to configurations of corpuscles, andno reference to the relation of resemblance.

The interpretation of the porphyry argument that comes closer thanany other to the reconstruction I have offered appears in Bolton (pp. 360–61). Bolton offers the following outline of the argument:

(1') Porphyry looks red in the light, but its color disappears in thedark.

(2') The “real” qualities of porphyry are the same whether it is in thelight or in the dark.

(3') It is inconceivable that the porphyry should have a propertyresembling the idea of red at the same time that its color hasdisappeared.

∴(4') The porphyry does not have a “real” quality resembling the ideaof red.

This argument resembles the one given earlier, in that (1') says much thesame as (1) and (2') says much the same as (2). But the conclusions ofthe two arguments differ, and the latter argument contains an additionalpremise.

Bolton’s reconstruction of the porphyry argument faces the followingdifficulties. First, there is no reason to think that (4') is the argument’sconclusion. Locke is surely arguing for the claim that the “ideas” ofredness and whiteness are not really in porphyry. Now, remarkablyenough, Bolton agrees. For, just before setting out her version of theargument, she writes:

The question [at the end of the passage excerpted from II.viii.19] is evidently intended asan argument in support of the claim, as Locke puts it here, that “the ideas of whiteness andredness are [not] really in porphyry.” (p. 360).

But she adds:

Strictly speaking, [Locke’s] own distinction between qualities and ideas has it that no ideais in a body; Locke considers it more accurate to say that the ideas are not like any qualityin the stone. (p. 360).

So Bolton reads Locke’s claim that the “ideas” of whiteness and rednessare not really in porphyry as a claim about the ideas of whiteness andredness produced in our minds by porphyry in the light. But, as we sawabove, Locke himself admits to using the word ‘idea’ to mean the same



as ‘quality’ when speaking of ideas as being in bodies (II.viii.8), and thisis a case in point. When Locke suggests that the “ideas” of whiteness andredness are not really in porphyry, he surely means that the qualities1 ofwhiteness and redness are not really in porphyry. If we read him in theway that he himself suggests, then we need not understand him to bearguing for the claim that our ideas of whiteness and redness do notresemble any quality1 in porphyry.

But if this is correct, then the porphyry argument is an argument for(3), not an argument for (4'). At this point, problems multiply. If (3) issubstituted for (4') in Bolton’s reconstruction of the argument, it followseither that (3') is superfluous, since (1') and (2') entail (3), or that (2') issuperfluous, since (1') and (3') entail (3). But Bolton would be the first torecognize that, of the two potentially superfluous premises, only (2') isexplicitly stated. Since the argument remains valid when (3') is removed,we are left with our original reconstruction of it.

How good is the argument thus represented? It should be clear that theargument is only as good as its first premise, a premise which Locke’scontemporaries might have accepted, but which few of ourcontemporaries would be prepared to embrace. The view that objectsretain their colors in the dark is now treated as a datum which any theoryof color must accommodate. This fact may explain why Locke scholarshave drawn the primary/secondary quality distinction incorrectly andhaven’t been able to reconstruct arguments for the distinction that arefaithful to the text.

Let us now turn to the almond example at II.viii.20. There Locke writes:

Pound an Almond, and the clear white Colour will be altered into a dirty one, and the sweetTaste into an oily one. What real Alteration can the beating of the Pestle make in any Body,but an Alteration of the Texture of it?

The argument in this passage is rather condensed. So it should not surpriseus that it has been understood in different ways. To one scholar, thealmond passage “argues that an almond’s colour and taste are mereupshots or symptoms of its primary-quality ‘texture’, since the latter isall that can be altered by pounding” (Bennett, p. 103). To another, thepassage “may be thought of as issuing a challenge: show me how thepounding could produce a change of colour except through change oftexture” (Alexander, p. 127).But surely the most plausible hypothesis is that Locke was challenging

the (common) view that “the colour and taste of the almond are real, thatis, qualities the almond has whether or not it is perceived” (Bolton, p. 364). On this interpretaton, Locke’s almond argument may berepresented as follows:



(1) The pounding of an almond changes its color and taste.(2) The only real quality1 in an almond that it is possible to change

by pounding it is its texture.(3) Neither the color nor the taste of an almond is identical to its

texture.∴(4) The color and taste of an almond are not real qualities1 in it.

This argument is valid, and, since it does not rely on the claim that“pounding can cause only primary-quality changes in the object pounded”(Bennett, p. 103), cannot be faulted on the grounds that this claim is false.But it is not clear that the argument is sound, for someone might wellobject, as against (2), that “if there were intrinsic colour-as-we-see-it andtaste-as-we-taste-it qualities, the beating could alter them” (Mackie, p. 22). In reply to this criticism, Locke might insist that (2) is simply alogical consequence of mechanism, the view that “all changes of state ofbodies are due to a change in texture […] and all changes in texture arethe result of impact or contact action of one body upon another”(McCann, p. 56). But this would be a mistake. For although mechanismentails that pounding an object changes its texture, it does not entail thattexture is the only real quality1 that the pounding can change, unless itis assumed that colors, tastes, etc. are not real qualities1. Unfortunately,this is an assumption that Locke cannot make, since it is clearly identicalto the proposition he is attempting to prove.2

The third and final argument for the primary/secondary qualitydistinction is contained in the hot-and-cold water case, which Lockedescribes at II.viii.21:

Ideas being thus distinguished and understood, we may be able to give an Account, howthe same Water, at the same time, may produce the Idea of Cold by one Hand, and of Heatby the other: Whereas it is impossible, that the same Water, if those Ideas were really in it,should at the same time be both Hot and Cold.

It is common to interpret this passage (together with the rest of the section)as containing an inductive argument for the mechanical hypothesis thatthe idea of heat is produced in us by an “increase […] of the motion ofthe minute Parts of our Bodies” and the idea of cold is produced in usby a “decrease” of the motion of the very same corpuscles (II.viii.21). Forthis hypothesis explains why “the same Water, at the same time, mayproduce the Idea of Cold by one Hand, and of Heat by the other” whenit is combined with the verifiable empirical claims that (i) the motion ofthe corpuscles in the water is greater than the motion of the corpusclesin the hand that is causally involved in the production of the idea of heat,and (ii) the motion of the corpuscles in the water is less than the motionof the corpuscles in the hand that is causally involved in the production



of the idea of cold (Curley, p. 458, Mackie, p. 22, Alexander, p. 128). Butthe passage has also been read as an argument for the claim that the samewater cannot resemble the ideas of hot and cold that it simultaneouslyproduces in our minds (Bolton, p. 365).

Now, again, there is no doubt that Locke accepted the hypothesis thatthe idea of heat is produced in us by an “increase […] of the motion ofthe minute Parts of our Bodies,” and that he rejected the view that ourideas of hot and cold resemble the objects that produce them in our minds.But it is not all clear that the passage in question contains an argumentfor either of these claims. Rather, Locke appears to be arguing for the(now familiar) conclusion that heat and coldness are not real qualities1.The argument may be represented as follows (where W is the water inquestion):

(1) W, at the same time, produces an idea of Heat when touched byone hand and produces an idea of Coldness when touched by theother hand.

(2) Water that produces an idea of Heat when touched by a hand has[the quality1] Heat.

(3) Water that produces an idea of Coldness when touched by a handhas [the quality1] Coldness.

∴(4) W, at the same time, has Heat and Coldness. (From 1, 2, 3)(5) Heat and Coldness are opposites.(6) It is impossible for opposite real qualities1 to exist in the same

substance at the same time.(7) Heat is a real quality1 if and only if Coldness is a real quality1.

∴(8) Heat and Coldness are not real qualities1. (From 4, 5, 6, 7)

Again, this argument, considered on its own, does not establish theexistence of a fundamental distinction between, say, shape and heat. ButLocke adds that the same object, at the same time, never produces “theIdea of a square by one Hand” and “the Idea of a Globe by another”. Ifthis is correct, then there is no sound parallel argument to the conclusionthat determinate shapes are not real qualities1 of objects. Therefore, ifthere are good independent reasons for thinking that shapes are realqualities1, then Locke’s hot-and-cold water argument provides us with apowerful reason for accepting the primary/secondary quality distinction.

To summarize: Locke first employs II.viii.16 and II.viii.18 to discreditthe reasons for thinking that heat, tastes, and colors are real qualities1,and then employs II.viii.19–21 to argue directly that heat, tastes, andcolors are not real qualities1. Although the porphyry and almondarguments of II.viii.19–20 are inconclusive, the water argument ofII.viii.21 is, I think, persuasive. Thus, if shapes, sizes, etc. are (as Locke



and his Scholastic opponents assumed) real qualities1, it follows that thereis indeed a fundamental distinction between primary and secondaryqualities.

Conclusion

I think it fair to say that Locke is widely viewed as a valiant and thoughtfulchampion of common sense who did not worry much about whether hisviews were clearly expressed or mutually consistent. Scholars speakconfidently of “those unfortunate inconsistencies for which Locke’s workis so famous” (Curley, p. 440), and of his inability to extricate himselffrom “self-contradictory situation[s]” (Bennett, p. 110). Less generally,some describe Locke’s way of drawing the primary/secondary qualitydistinction as “somewhat misleading” (Mackie, p. 8). Others dismissLocke’s own statements about the distinction as “indefensible” (Bennett,p. 90), or as “fumbled attempts” to express true claims (Bennett, p. 106).His use of words is characterized as “sloppy” (Bennett, p. 107),“deplorable” (Bennett, p. 111), and “inconstant” (Mackie, p. 8). And,finally, while some read his arguments for the distinction as providing“something of a case for [it], but not […] a very strong one” (Mackie, p. 23), others read his arguments as providing a relatively strong case forpositions that Locke clearly did not hold (Curley, p. 445, Alexander, p. 118).

I believe that a careful and sympathetic reading of Book II, Section viii(as well as other parts) of the Essay reveals that, despite occasionalcarelessness, Locke’s way of drawing the primary/secondary qualitydistinction is internally consistent and inherently plausible, that all of hisarguments for the distinction are simple and valid, and that one of hisarguments for the distinction is convincing as well. The ultimate irony isthat Locke is much less confused about the distinction and his ownarguments for it than are many of his interpreters.3

NOTES

1Here, and in what follows, I use the term ‘property’ as we ordinarily use it, not as Lockeused it. As an anonymous reviewer for this journal rightly points out, Locke uses the term‘property’ as we would use the phrase ‘essential property’.

2The same anonymous reviewer mentioned in note 1 points out (again, rightly) that if,as I have argued, determinate shapes and sizes are real qualities1, then Premise (2) of myreconstruction of Locke’s almond argument is false, for the simple reason that it is possibleto change the shape and size of an almond by pounding it. It follows that Locke’s almondargument, as I read it, is unsound. In Locke’s defense (and in defense of my reconstructionof his argument), it should be noted that Locke had the means with which to avoid thiscriticism without sacrificing anything in the way of plausibility. For consider the following



statements:

(2'') The only real qualities1 in an almond that it is possible to change by pounding it areits texture, shape and size.

(3'') Neither the color nor the taste of an almond is identical to its texture, shape or size.

It should be clear that my reconstruction of the almond argument remains valid if (2) isreplaced by (2'') and (3) is replaced by (3''). And it should also be clear that the amendedversion of the almond argument obtained by replacing (2) with (2'') and (3) with (3'') avoidsthe criticism raised above. (Needless to say, however, this maneuver does not avoid thecriticism discussed in the text.)

3 I was first introduced to the issues discussed in this paper by Keith Donnellan, whotaught a stimulating course on Locke that I attended as a graduate student at UCLA in1989. The paper itself grew out of ideas I presented in a Spring 1997 Graduate Seminar onLocke’s Metaphysics and Epistemology at Florida State University. I wish to thank mystudents in that seminar (Justin Barnard, Brad Hadaway, Greg Smith, and Chip Summey)for helping me to sharpen my thoughts in various ways. In February 1997, I presented aversion of the first part of this paper at a Philosophy Department Colloquium at theUniversity of Alabama at Tuscaloosa. I would like to thank those who attended thatColloquium for their incisive comments, especially Torin Alter and Norvin Richards. I amespecially grateful to an anonymous reviewer for this journal and to my colleagues, PatMatthews and Dana Nelkin, each of whom read a previous version of the paper and helpedme avoid serious mistakes as a result. Needless to say, none of them is responsible for anyerrors that remain. I dedicate this paper to my father, Elwood A. Rickless, and to my father-in-law, Norton Nelkin, both of whom encouraged me in more ways than they will everknow.

REFERENCES

Alexander, P. Ideas, Qualities and Corpuscles: Locke and Boyle on the External World(Cambridge: Cambridge University Press, 1985).

Bennett, J. Locke, Berkeley, Hume: Central Themes (Oxford: Oxford University Press, 1971).Bolton, M.B. “Locke and Pyrrhonism: The Doctrine of Primary and Secondary Qualities,”

in The Skeptical Tradition, ed. M. Burnyeat (Berkeley and Los Angeles: University ofCalifornia Press, 1983), pp. 353–75.

Campbell, J. “Locke on Qualities,” Canadian Journal of Philosophy 10, 1980, pp. 567–85.Curley, E.M. “Locke, Boyle, and the Distinction between Primary and Secondary Qualities,”

Philosophical Review 81, 1972, pp. 438–64.Jackson, R. “Locke’s Distinction between Primary and Secondary Qualities,” Mind 38,

1929, pp. 56–76. Reprinted in Locke and Berkeley: A Collection of Critical Essays, ed.C.B. Martin and D.M. Armstrong (Garden City, NY: Doubleday, 1968), pp. 125–54.

Locke, J. An Essay Concerning Human Understanding, ed. P.H. Nidditch (Oxford: OxfordUniversity Press, 1975). References in the text are to the 5th edition, published in 1706.

McCann, E. “Locke’s Philosophy of Body,” in The Cambridge Companion to Locke, ed. V.Chappell (Cambridge: Cambridge University Press, 1994), pp. 56–88.

Mackie, J. Problems From Locke (Oxford: Oxford University Press, 1976).



locke on primary and secondary qualities

Documents