Esser, Frederick: John Locke‘s Investigation into our Knowledge of Bodies


Kapitel 5. Locke‘s Theory of Qualities

In the chapter on ideas of substances, one of Locke‘s chief contentions is that our everyday ideas of bodies include mainly secondary qualities, which are non-real or nothing but powers, and only few primary qualities, which are real or in the things themselves.<223> This account is based on previous results, namely on the analysis of resemblance and qualities in II.viii. Locke defines here primary and secondary qualities and advances an argument to establish both that sensible qualities are non-real, or respectively that our ideas of them are not resemblances, and that corpuscularian properties are primary qualities and real, or respectively that our ideas of corpuscularian features are resemblances. In addition, as the comments in the chapter on ideas of substances display, Locke takes it for granted that everyone who reflects on his or her ideas will concede that our ideas of bodies include only few corpuscularian, primary qualities and for the most part secondary qualities, especially sensible qualities. Likewise, he assumes consent that the features he regards as non-sensible, secondary qualities are genuinely non-real. Thus, though Locke does present an argument that corpuscularian properties are primary qualities, i. e. that we do partly depict bodies by real properties, the focus is clearly on the non-reality of sensible qualities. More importantly, given these presumptions and his argument in II.viii, Locke can justify his chief contention that our everyday ideas of bodies inlcude only few real, primary qualities and mostly non-real, secondary qualities. I will call this whole account Locke‘s theory of qualities.

In the light of the so far developed reading of the reality status of qualities, the theory of qualities is an assessment of our everyday grasp of bodies from a perception-neutral perspective. That is: Locke‘s claims on which type of qualities are included in our common ideas of bodies specify the cognitive content of the ideas with respect to a scientific understanding of the represented properties. In principle, this account of our everyday knowledge of bodies can still be reconciled with Alexander‘s general view that


the comments on qualities relate to corpuscularian theory.<224> For, given Alexander were right, the causal perspective would be bounded up with corpuscularian theory and would therefore have to be conceived as a specific corpuscularian standpoint. On the contrary, according to the here advanced interpretation, the perception-neutral perspective is theory-neutral, i. e. it is not confined to or bounded up with a specific account of mental representation. One leading question therefore is whether Locke examines our common ideas of bodies from a specific, corpuscularian viewpoint or from an ideal scientific one.

Since the theory of qualities includes the reality doctrine and since the doctrine corresponds to the chief assertions on resemblance, the difference between these topics is, by and large, that the focus is either on features qua primary or secondary qualities, or on features qua real or non-real qualities, or on ideas being resemblances or dissemblances. The theory of qualities, the reality doctrine and the resemblance theory concern assertions corresponding to each other: for instance, that sensible qualities are secondary qualities, that sensible qualities are non-real, and that their ideas are not resemblances. Thus, Locke‘s argument on qualities establishes all three accounts.

An interpretation of Locke‘s reasoning on qualities first affords a clarification of the definition of primary and secondary qualities. it will become plain that the distinction is not confined to corpuscularian theory, but is rather related to an account of bodies from an ideal perception-neutral viewpoint. Subsequently, the full argument will be reconstructed in the context of resemblance theory and reality doctrine. Here the issue will finally be settled whether Locke‘s reasoning intends to establish corpuscularian theory or a theory-neutral assessment of our ideas of bodies from an ideal perception-neutral viewpoint.

a. Locke‘s Definition of Primary and Secondary Qualities

We begin with Locke‘s notion of secondary qualities. In the discussion of the reality doctrine we have seen that the striking similarity between Locke‘s wording in §10 and §14


of II.viii indicates that secondary qualities are defined as non-real properties, namely as features being nothing but dispositions which are to be explained in terms of primary qualities in a perception-neutral account of bodies and their properties.<225> More precisely, in §10 secondary qualities are not identified with primary qualities, but with corpuscularian properties, i. e. with features in terms of which dispositions like secondary qualities are elucidated in corpuscularian theory, e. g. form and solidity. On the face of it, this suggests that secondary qualities are not defined as the properties which are reducible according to a correct or ideal theory, but as the properties which are reducible according to corpuscularian theory. This would mean, first, ’secondary quality‘ is a technical term of corpuscularian theory and, second, there would be in fact no secondary qualities if corpuscularian theory were false. If in a true, comprehensive or ideal theory of bodies, their dispositions were not to be identified with corpuscularian qualities, there would be no secondary qualities since no dispositions would fulfill the definition of secondary qualities. But there is one passage which displays that, if pressed, Locke understands secondary qualities solely by being non-real and not necessarily as reducible to primary qualities.

Locke asserts that (ideas of) secondary qualities might be in fact to be explained by features which are not primary qualities or corpuscularian properties respectively and of which one does not have yet any comprehension: (ideas of) secondary qualities “[depend] all (as has been shewn) upon the primary Qualities of their minute and insensible parts; or if not upon them, upon something yet more remote from our Comprehension“.<226> What Locke evidently wants to express is that the features he has determined as secondary qualities are truly secondary qualities whether they are to be elucidated in terms of corpuscularian properties or other features. In either case, he insists, the so-called secondary qualities are reducible, i. e. they have to be explained in terms of properties which are of a different conceptual type. This shows that, if pressed, Locke portrays secondary qualities not as features which are to be identified with corpuscularian properties, but as features which are dispositions in an ideal theory. ’Secondary quality‘ is


not a technical term of corpuscularian theory which has an extension only if corpuscularian theory is largely right. To the contrary, a secondary quality is understood as a feature that is reduced to other features in an ideal scientific account of bodies. Consequently, one has to read §10 as defining secondary qualities as being non-real features, i. e. as being of a conceptual type in terms of which bodies are not conceived from an ideal perception-neutral viewpoint.

We now turn to primary qualities. In his definition of primary qualities Locke first gives three abstract characterizations (which are syntactically seperated by semicolons) and then makes a thought experiment which identifies corpuscularian properties as primary qualities, e. g. bulk and form. In his abstract portrayals Locke defines primary qualities as those properties no body can do without: in whatever state a body is, whatever changes it suffers, whether a body is perceivable or not, these features are inseparable from the body. Primary qualities are the properties every body has:

“§9. 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 forces can be used upon it, it constantly keeps; and such as Sense constantly finds in every particle of Matter, which has bulk enough to be perceived, and the Mind finds inseparable from every particle of Matter, though less than to make it self singly be perceived by our Senses. v.g. Take a grain of Wheat, divide it into two parts, each part has still the same qualities; and so divide it on, till the parts become insensible, they must retain still each of them all those qualities. For devision (which is all that a Mill, or Pestel, or any other Body, does upon another, in reducing it to insensible parts) can never take away either Solidity, Extension, Figure, or Mobility from any Body, but only makes two, or more distinct separate masses of Matter, of that which has but one before, all which distinct masses, reckon‘d as so many distinct Bodies, after division make a certain Number. These I call original or primary Qualities of Body, which I think we may observe to produce simple Ideas in us, viz. Solidity, Extension, Figure, Motion, or Rest, and Number.“<227>

One should not understand Lockean primary qualities to be determinable properties, as some commentators do.<228> The notion of shape, for example, is a determinable property in the sense that the feature of having a shape has to be determined or specified with


respect to each body, since bodies have always a determined or specific shape. ’Shape‘ does not denote to every body the same particular shape, but only some particular shape which varies amongst bodies. However, this interpretation does not work out, because solidity is not a determinable property. According to Locke‘s technical term of ’solidity‘, which is roughly aquivalent to ’impenetrability‘,<229> there are not various kinds of solidity as there are different kinds of shape. There is simply solidity. Every body, whether it is water, a diamond or a single corpuscle, is called in the same sense solid. The example of solidity shows that primary qualities are not determinable properties. Primary qualities are simply features being general enough to be properties of all bodies.<230> The feature of having the shape of a globe, by contrast, is too specific, i. e. not every body has it.

After having advanced the three abstract portrayals of primary qualities, Locke expounds a thought experiment in the course of which he identifies primary qualities with corpuscularian properties. Given this structure of the reasoning, it is natural to read the thought experiment as an argument for specifying which features are primary qualities, i. e. which properties fulfill the previously given definition. This “grain-of-wheat argument“ determines with which features a body cannot do without: even if a body is devided in unperceivable parts, each resulting body has still solidity, form, etc. In the last sentence of §9 Locke then sums up that primary qualities are the properties mentioned in the grain-of-wheat argument. The thought experiment is a reasoning validating by analogy that the mentioned corpuscularian properties are primary. The analogy is between perceivable and unperceivable bodies, arguing that the latter always retain solidity, extension, form and mobility despite any possible kind of alteration just as the former do. The analogy thus implies that all bodies, whether perceivable or not, have corpuscularian properties.

In the light of Locke‘s use of the thought experiment to establish that corpuscularian properties are primary qualities, ’primary quality‘ seems like ’secondary quality‘ not to be a technical term confined to corpuscularian theory in the sense that it has extension only if corpuscularian theory is largely correct. Primary qualities are not defined as being


corpuscularian properties; rather, a reasoning is presented to identify the former with the latter. This shows, both concepts can have extension even if Boyle‘s theory is fundamentally flawed. This strongly suggests that the quality distinction is meant to relate to an ideal account of bodies.

This conclusion is confirmed, even if §10 is differently interpreted in order to justify a “corpuscularian reading“ of Locke‘s notion of primary qualities. Alternatively, one could insist that the thought experiment does not determine which properties fulfill a previously given definition, but is rather part of this foregoing portrayal. Furthermore, in the last sentence Locke is conceived to finally define primary qualities as being corpuscularian properties fulfilling the preceding depiction. However, since ‚secondary quality’ is not regarded as a specifically corpuscularian term, as argued above, corpuscularian theory would effectively be reduced to the assumption that every body has corpuscularian features if primary qualities were understood to be defined as features which are possessed by every body according to corpuscularian theory. But this conception of primary qualities would certainly not do justice to Boyle‘s hypothesis. And this last consequence is rather implausible since Locke is so well acquainted with the theory.

Correspondingly, the distinction between being real and being non-real relates to an ideal theory of mental representation as well. Locke‘s epistemological perspective is an ideal one being not confined to corpuscularian theory. Locke believes of course in the corpuscularian claim that the properties he calls secondary qualities are truly to be explained by corpuscularian properties, i. e. by features which corpuscularian theory has determined as real qualities. This is the reason why Locke usually portrays secondary qualities as reducible to corpuscularian, real properties, e. g. in his definition of secondary qualities in §10. Yet, when engaged in scientific speculation, as we have just seen, Locke is prepared to concede that corpuscularian theory might be false and that secondary qualities are then not to be identified with corpuscularian, real qualities.

One can enhance the analysis. On the backdrop of §§7-8,<231> Locke introduces the quality distinction in §§9-10 in order to clarify the relationship between an idea, i. e. an intentional state representing a property, and the property causing the mental


representation and being the content of an idea. In other words, Locke draws the distinction in order to specify which simple idea resembles the property it is about, namely that only ideas of corpuscularian properties are resemblances but not ideas of sensible qualities. Moreover, the reasoning in the chapter on ideas of substances shows that Locke is interested in resemblance and qualities to determine the cognitive content of our ideas of bodies: which of the features included in our ideas are real. Given this backdrop, the thrust of the quality distinction is clearly an epistemological one. The distinction serves to determine which of the properties, by which we depict bodies, are of the same conceptual type in terms of which one conceives bodies as independent of how they are mentally represented. Locke‘s understanding of qualities thus relates to an ideal account of bodies from a perception-neutral viewpoint. In this sense, the conception of primary and secondary qualities is intended to be theory neutral, i. e. to be appropriate for an ideal account and not to be confined in its employment to the truth of any particular theory.

Correspondingly, if the quality distinction is not confined to Boyle‘s hypothesis, then certainly is so the notion of resemblance either. This implies, the viewpoint of resemblance is an ideal scientific one. Locke therefore does not aim at establishing corpuscularian theory, as Alexanders contends. Rather, he transforms Boyle‘s scientific issue of explaining human sense perception of bodies in the genuinely epistemological issue of assessing our conceptions of bodies from an ideal scientific, perception-neutral standpoint.

Coming back to the topic of qualities, at first sight, there are two curious aspects about Locke‘s quality distinction. Locke treats the quality distinction as if it were complete, i. e. as if it classified all the properties bodies have in being either primary or secondary. However, given his definitions, there might be properties being neither primary nor secondary. For, if there were explanatory basic or real qualities which are not shared by all bodies, these real qualities would neither be primary nor secondary since they are neither common to every body nor non-real. In other words, there is a “conceptual gap“ between primary and secondary qualities. It seems to me, Locke has not recognized this due to his belief in corpuscularian theory. According to the theory, on the one hand, corpuscularian properties are primary and real, since all bodies are supposed to have them and since all


features of bodies are assumed to be reducible by corpuscularian properties. And, on the other hand, corpuscularian properties are regarded as the sole features in terms of which bodies have to be understood from a scientific viewpoint. If one therefore identifies primary qualities with corpuscularian properties in the light of corpuscularian theory, primary qualities are the only real qualities bodies have. On this background, it becomes plain why Locke conceives the quality distinction as complete. Corpuscularian theory bridges the conceptual gap: primary qualities add up to real qualities so that primary qualities are opposed to secondary qualities which are defined as non-real properties. Features of bodies are seen as being either real or non-real, namely as being either primary or secondary.

Another reason why Locke does not draw a complete distinction of qualities might be his objective. After all, he wants to examine our ideas of bodies as to which of the entailed features are real. He is thus concerned with corpuscularian properties, sensible qualities and other (secondary) qualities. In addition, Locke apparently believes that corpuscularian properties are real. This matches his grain-of-wheat argument which despite its simplicity is taken to indicate strongly that corpuscularian properties are primary and thus real. Moreover, according to Locke there is no question that all the other, known qualities, which are not corpuscularian features or sensible qualities, are secondary. And with respect to sensible qualities Locke clearly holds, as the quotation above shows, that they are secondary in any case whether they are to be reduced to corpuscularian properties or to other ones of which we do not even have a conception. Thus, the range of features Locke has in mind when discussing qualities are properties of whose quality status he has strong convictions. Corpuscularian properties are primary, all the other features are secondary. Locke holds this belief of course in the light of corpuscularian theory. Given therefore corpuscularian theory, the quality distinction files, despite its “incompleteness“, all the properties being contained in our ideas of bodies.

There is another striking aspect of the quality distinction. After having defined ’primary quality‘, this depiction of primary qualities virtually disappears in his further comments. Instead, primary qualities are portrayed as being real, as being in the things themselves and by their explanatory role in a theory about human mental representations of bodies. And these characterizations have turned out to be aquivalent. It seems to me, the reason


for the disappearance of the defining portrayal of primary qualities is again Locke‘s leaning towards corpuscularian theory. According to this account, the only qualities which are real are primary, namely corpuscularian properties. Identifying primary qualities with corpuscularian features, it is natural for Locke to equate primary qualities with real qualities.

But having said all that, should one not expect Locke to define primary qualities as real properties in the first place instead of determining them as features all bodies have? This would indeed be the conceptually straight forward way, since in the subsequent passages Locke is interested in properties which are real, but not in features which are possessed by every body. However, Locke‘s train of argument affords first the notion of primary quality as he defines it and then the notion of real quality. On the basis of his definition Locke specifies first the set of primary qualities by his grain-of-wheat argument in §9. He then outlines a corpuscularian theory of sense perception which entails the assumption that corpuscularian properties are primary. Locke‘s depiction of corpuscularian properties as primary serves here as a premise to make intelligible a corpuscularian theory of perception. And it is thereby where corpuscularian properties acquire the feature of being real, since ideas of primary qualities and ideas of secondary qualities are said to be produced by bodies that are understood in terms of corpuscularian properties.

Of course, Locke could have disentangled his concepts by defining first primary and secondary qualities as real respectively non-real and by introducing subsequently the notion of a property being shared by all bodies to get his argument off the ground. However, Locke does not present his theory in this conceptually more matured form because he has corpuscularian theory before his eyes according to which real qualities are aquivalent to the properties being possessed by every body. Yet, as the discussion of secondary qualites has revealed above, if pressed, Locke is prepared to concede that there might be other, yet unknown features which are real and in terms of which one has to explain secondary qualities.

One can add a further, critical remark in this context. Given Locke‘s definition of primary qualities, there might be none. For it is possible that smaller bodies have properties of a conceptual type which is not also instantiated by bigger bodies. In the extreme case, atoms might be conceptually conceived radically different than macrophysical objects. Again,


one sees that despite its theory neutrality the quality distinction is designed in the light of corpuscularian theory in the very end.

b. The Argument

Having clarified the quality distinction, we can move on to assess the argument. It has already been indicated, Locke develops the quality distinction in the context of his resemblance theory.<232> His topic is the assessment of which idea resembles the property that causes the idea (§7). As the reality doctrine has shown, the notion of being a real quality corresponds to his conception of an idea being a resemblance. A property is real if it is of the same conceptual type in terms of which bodies are depicted as being independently understood of the way the feature is mentally represented. An idea is a resemblance if the idea represents the property in the same terms in which the feature is portrayed as being independently grasped of the manner the idea represents the property. To come to grips with resemblance, Locke first clarifies his general use of ’idea‘ and ’quality‘ and then introduces ’primary quality‘ and ’secondary quality‘ (§§8-10). As we have seen, the quality distinction is drawn in the light of corpuscularian theory. Yet, the pair of concepts is neutral to or not part of the corpuscularian hypothesis. It is an epistemological distinction going along the lines of the resemblance status of ideas and the reality status of qualities. Moreover, while defining primary qualities, Locke presents the grain-of-wheat argument to establish corpuscularian properties as primary.

In §11 Locke maintains that we have to conceive bodies to interact by impulse when producing ideas in us: “§11. The next thing to be consider‘d, is how Bodies produces Ideas in us, and that is manifestly by impulse, the only way which we can conceive Bodies operate in.“<233> Locke states thus a requirement for every plausible hypothesis of sense perception, namely that causal interaction between bodies and human minds has to be explained in terms of impulse. Of course, corpuscularian theory fulfills this requirement. As Rogers has pointed out, in the face of Newton‘s theory of gravitational forces, Locke


softened the original version (in the first three editions of the Essay) of this paragraph which had excluded the possibility that bodies could operate upon each in another way than by impulse.<234> But although Locke allows implicitly from the fourth edition onwards (as just quoted) the existence of gravitation, the way gravitation works is for him still inconceivable.<235> For him only the conception of impulse and motion makes comprehensible how bodies act upon each other. §11 sets thus the stage for the introduction of a corpuscularian theory of perception in §§12-13.

In §§12-13 a corpuscularian theory of the genesis of mental representations is outlined. Locke does not go into details just as he does not do in other passages as well.<236> Apparently, for him it is sufficient to make intelligible the kind of explanation corpuscularians give. More specifically, Locke is interested in making plain that in a theory of perception features causing ideas of primary and of secondary qualities have to be understood in terms of (microphysical) primary qualities. In §§12-13 Locke focuses on primary qualities and on some sensible qualities, namely on colours and smells. §13 ends with the conclusion that ideas of colours and smells are in the same way not a resemblance as the idea of pain is.<237> Given the advanced interpretation of resemblance and of the pain analogy, in §13 Locke simply concludes the dissemblance status of ideas of colours and smells from the non-reality status of colours and smells which they have according to the scetched corpuscularian theory of perception. Colours and smells are understood from the causal viewpoint in terms of primary qualities, i. e. in terms of other


concepts as the ones by which they are grasped as being represented by our ideas of colours and smells.

In §§14-15 Locke moves on to generalize his point. §14 contends that not only colours and smells but also other sensible qualities like tastes and sounds are non-real and have to be identified in terms of primary qualities from the explanatory perspective. In §15 Locke then concludes from the non-reality status of sensible qualities the dissemblance status of their ideas. Likewise, he deduces that ideas of primary qualities are resemblances. The subsequent paragraphs of II.viii highlight the principal claim that only ideas of primary qualities are resemblances and that ideas of sensible qualities - like the ideas of other secondary qualities - are dissemblances. As will be delineated, he thereby appeals to experiments, or everyday experience respectively, to undermine an Aristotelian theory of sense perception and of real qualities which thus serve as additional support for the corpuscularian account. And since Locke assumes consent that non-sensible secondary qualities are not real, in II.viii Locke‘s reasoning focuses clearly on sensible qualities. The aim of the argument is to refute the popular belief that sensible qualities are real and that our ideas of them are resemblances.<238>

On the basis of this argument in II.viii, Locke can easily assess our ideas of bodies in II.xxiii. Our ordinary ideas, Locke maintains, portray bodies mainly only by non-real features or respectively secondary qualities. Bodies, he contends, are primarily depicted by powers, namely by mediately perceivable secondary qualities or by sensible qualities.<239> And since according to the argument not only mediately perceivable but also immediately perceivable or sensible qualities are non-real, we grasp bodies notably in terms of non-real properties. As Locke‘s remarks show, this is most obvious as to chemical substances where even form and bulk are disregarded for characterizing bodies.<240> In other words, our everyday knowledge does not conceive bodies chiefly from the ideal epistemological-perspective, i. e. from the perception-neutral standpoint. This account of our knowledge is even more disillusioning if one considers that for Locke ideal knowledge of bodies would depict them not by macrophysical, but by microphysical primary qualities, i. e. by


real essences.<241>

Evidently, Locke‘s claims as to qualities and his corresponding contentions of resemblances are related to a corpuscularian-mechanical theory of bodies, their qualities, and our perception of them. But how is his reasoning exactly to be understood?

Importantly, one should distinguish between arguments relating to a general mechanical understanding of bodies and specific claims relating to corpuscularian theory. In §11 Locke asserts that a mechanical understanding of perception is the only kind of explanation at hand which makes the process of generating ideas or mental representations intelligible. One can conceive interaction between bodies and human minds only in terms of impulse. Moreover, this mechanical conception of perception can be spelled out differently, i. e. not only by corpuscularian theory. Similarly, given the simplicity of the grain-of-wheat argument, it effectively asserts that bodies can be comprehended only in terms of the mentioned corpuscularian properties, e. g. solidity. Every body, whether perceivable or not, has to be understood as having these features. And, again, the conception of bodies represents only a framework for a mechanical account of bodies which can be spelled out differently, since it leaves open how to comprehend the relationship between properties, the inner structure of bodies, etc. The thought experiment establishes effectively that in any mechanical account of bodies the mentioned properties are primary qualities. Thus, in §§9 and 11 Locke does not demonstrate Boyle‘s hypothesis. Instead, characterizing bodies in the only way we can conceive of them, the two arguments represent a general mechanical account of bodies and their relation to ideas. Locke‘s contentions about impulse as well as about corpuscularian features being primary qualities make up the only conceivable conception to grasp bodies in relation to our ideas of them, namely a general mechanical one.

Up to this point, it has only been asserted that as a matter of fact in §§9 and 11 Locke argues for a general mechanical account, not specifically for Boyle‘s hypothesis. But does Locke also view his claims in this way? Given the evident generality of the contention that bodies interact by impulse, Locke cannot believe that the claim establishes corpuscularian theory. As to the grain-of-wheat argument, Locke certainly echoes Boyle


who has previously used a similar reasoning in his corpuscularian theory to establish that bodies have to be conceived in terms of corpuscularian properties.<242> At first sight, one could therefore take the thought experiment to demonstrate Boyle‘s hypothesis. On the other hand, since for Locke the only conceivable conception of bodies is a mechanical one and since the thought experiment evidently determines only which features are primary, Locke need not to regard the argument as establishing corpuscularian theory in particular. Moreover, corpuscularian features coincide by and large with the set of physical properties which have generally been regarded as primary or explanatory basic in the mechanical tradition going back to Galileo.<243> Thus, an argument that establishes corpuscularian features as primary qualities can likewise be seen as determining generally which properties have to be primary according to a mechanical account. More importantly and decisively, Locke seems to identify corpuscularian theory with more specific claims than that bodies interact by impulse or that corpuscularian properties are primary. He refers explicitly to corpuscularian theory more or less only when he outlines more specific theses of the hypothesis, e. g. that animal spirits transfer motion from the senses to the brain. Locke apparently identifies corpuscularian theory with its particular claims as to the details of the constitution of bodies and their causal interaction on the human senses. In this light, the grain-of-wheat argument serves to establish for any kind of mechanical account which features are primary.


Locke argues for a minimal account of bodies which represents the only way we can conceive of them, and signifies explicitly further claims as more doubtful hypotheses. But, as his identification of primary qualities with corpuscularian features shows, he tends to view a general mechanical account already in the perspective of Boyle‘s theory. If pressed, Locke would certainly maintain that the grain-of-wheat argument applies to any mechanical account. Yet, on the other hand, the thought experiment effectively determines primary qualities as those features which, in Locke‘s terms, Boyle has specified as real qualities. Thus, in Locke‘s comments there is a distinction between a general mechanical account, on the one hand, and corpuscularian theory which spells out this conception, on the other hand, but this distinction is not clear-cut. The reason apparently is that for him Boyle‘s hypothesis is unmatched. Corpuscularian theory is for Locke the most convincing, unrivaled mechanical theory. Speaking of a mechanical account of bodies, their qualities and their relation to ideas, is conceiving bodies in corpuscularian terms and properties. It is therefore natural for Locke to regard corpuscularian theory as the only conceivable account at hand and thus to treat qualities in accordance with Boyle‘s hypothesis.

In short, the two arguments in §§9 and 11 respectively establish a general framework for a mechanical theory of bodies and their relation to perception. He then fills out this account by the assumption of corpuscularian theory in §§12-13. For Locke a mechanical account is the only one providing us with a scientific conception of bodies which is best spelled out by the unmatched corpuscularian theory. This corresponds to the structure of the argument. After having advanced the grain-of-wheat argument and his claim as to impulse, Locke assumes explicitly the corpuscularian hypothesis to make intelligible how one has to understand perception in mechanical terms. Crucially, it shows up that there is no particular reasoning for corpuscularian theory itself.

In the same light, one has to read Locke‘s declarations that the programme of the Essay does not consist in scientific investigations, e. g. in developing a theory of sense perception. At various places, he distinguishes his epistemological project to determine the extent and origin of knowledge from natural philosophy. These remarks stand usually in connection with excursions into natural philosophy to ensure the reader that


his reasoning does not require a discussion of scientific questions.<244> Given the use of corpuscularian theory in his argument, the punchline apparently is that Locke regards his analysis as not requiring a debate whether corpuscularian theory is really true. For Locke it is sufficient to argue for an outline of the only satisfying model making sense perception conceivable, namely a corpuscularian-mechanical one. This aim of the argument fits with Locke‘s contention that we can imagine secondary qualities to be explained in terms other than corpuscularian qualities, but that we do not have any conception of these other, unknown features. Since Locke does not present any argument why in any case the so-called secondary qualities are truly secondary, one might, however, regard Locke to be a bit dogmatic on this subject matter. Here Locke adheres to a characteristic claim of Boyle‘s corpuscularian theory, namely that sensible qualities and other dispositional properties are reducible. To sum up, Locke argues for his contentions about qualities by outlining a corpuscularian-mechanical account making plausible that his claims hold.

Moreover, Locke‘s comments on a corpuscularian-mechanical theory cannot mean to prove its truth. First, the grain-of-wheat argument serves only as a good reason for believing an hypothesis to be true, but not as a proof yielding knowledge. This becomes manifest in Locke‘s account of hypotheses. Analogies can only make plausible, but do not demonstrate the truth of hypotheses.<245> This means, for Locke his analogy serves as strong evidence for the hypothesis that corpuscularian properties are primary. One should therefore not mistake Locke‘s thought experiment as a proof for the existence of corpuscularian properties and, eventually, of primary qualities. The aim of the grain-of-wheat argument is to make plausible that corpuscularian properties are features every body has.

Second, when saying that we cannot conceive bodies acting on our senses otherwise than by impulse, Locke just points out that we do not have any other conception of interaction and does not attempt to prove that bodies have to act on us by impulse. He should be understood in this way, since he concedes as to secondary qualities that they might have


to be identified with properties other than corpuscularian features, namely with properties of which we do not have a conception.

Third, though corpuscularian theory provides the most convincing explanation, Locke calls in these and other passages the corpuscularian theory explicitly a hypothesis.<246> Again, this is confirmed by and is in accordance with his theory of knowledge and hypotheses. Corpuscularian theory is a hypothesis whose truth is uncertain. This epistemic status of corpuscularian theory is not surprising, however, given Locke‘s manysided criticism of the theory. He points to the limitations of the explanatory power of corpuscularian theory with respect to: the cohesion of corpuscles which make up a compound, the acting of bodies on one another by impulse (not to speak of gravitational forces),<247> and the causal interaction between mind and body.<248> - All this indicates, when advancing corpuscularian theory, Locke assumes the most reasonable and available account of bodies to assess the reality status of qualities and the resemblance status of our ideas.

It becomes manifest why Locke presents only a sketch of corpuscularian theory. For, if a corpuscularian theory of perception is more or less the unrivaled hypothesis and if the reasoning cannot yield knowledge in the first place, one needs to delineate corpuscularian theory only to the extent that it becomes sufficiently conspicuous how the argument runs in principal, namely an argument establishing the reality doctrine and the resemblance theory on the basis of assuming corpuscularian theory. Given the alternatives, Locke has neither to delineate Boyle‘s hypothesis nor to justify it in detail.<249>

This leaves us with the two arguments which are supposed to demonstrate a mechanical framework for any account of bodies. Are they good? First, the grain-of-wheat argument is certainly far from being convincing even though Locke is in distinguished company, namely with Boyle.<250> All bodies we perceive do not only have the corpuscularian


properties Locke mentions but also sensible qualities, e. g. colours. If the analogy were sound, it would prove as well, for example, that colours are primary: since all perceivable bodies retain colour whatever change they undergo, every unperceivable body is coloured. Second, given Locke‘s belief that gravitation is inconceivable, it is natural for him to assume as a mechanist that impulse is the only conceivable manner in which bodies interact with human senses. Given Locke‘s historic context, his arguments and assumptions are certainly not inappropriate, but they are not satisfying either.

As indicated, Locke adds further reasonings to his main argument which, pace McCann, render implausible an alternative, Aristotelian account of perception and real qualities.<251> One can distinguish between four experiments, or types of everyday experience respectively, that are directed against Aristotelian theory and that thus supercede those passages in which Locke insists emphatically on sensible qualities being on par with non-sensible secondary qualities.<252> According to Aritsotelian theory, at least as Locke conceives it,<253> the perception of sensible qualities involves the communication of a substantial form, or a real quality respectively. As we have seen, a real quality is conceived as a property which bodies have from an explanatory viewpoint. Change in the perception of the features of a body indicates change of its constitution, i. e. of its properties, and change of the latter causes a corresponding change in perception. On this backdrop, one can naturally highlight §16, and §§19-21, as Heyd has shown.<254>

First, in §16 Locke argues that, depending on the distance, the same fire causes the idea of warmth and the idea of pain. This contradicts Aristotelian theory which implies that the


change of the idea of warmth to the idea of pain represents an alteration in the constitution of the fire. But the fire is the same, only our sensation of it changed because the perceptual conditions are different. Second, in §19 porphyry is said to cause sensations of colour when striking by light, but not to do so in darkness. Locke similarly presumes here that turning on or off the light does not alter the constitution of porphyry. Hence, this contradicts Aristotelian theory which implies that “real alterations are made in the Porphyre<255> because the change of perception is not explained by light, but by the transmission of real qualities. Third, in §20 Locke refers to the change of the sensible qualities of almonds when they are pounded. Here it is presupposed that poundering can only alter the physical structure of bodies, but not any Aristotelian real qualities. This contradicts Aristotelian theory which implies that the change in sensible qualities goes hand in hand with an exchange of sensible qualities qua real qualities, e. g. that there is privation of sweetness and acquisition of oiliness. By contrast, according to the corpuscularian story, there is only a change in texture, i. e. in the arrangement of corpuscles, but no process of privation and acquisition of real qualities. Fourth, in §21 the same water is described to cause different sensations of temperature (when one hand has been previously cooled), i. e. one hand feels warmth, the other one coldness. Given however Aristotelian theory, this implies that the same water has contrary properties which is impossible. From an explanatory perspective, water cannot be both cold and warm, if both properties are taken as real qualities.

In turn, the experiments represent empirical support for a corpuscularian account of sense perception and real qualities since it can convincingly explain the phenomena, as Locke makes plain. On the other hand, the appeal to everyday experience is of course no conclusive evidence for the truth of corpuscularian theory. But given the alternatives, the corpuscularian account is clearly established as the only conceivable explanation.

To conclude, the issue of Locke‘s theory of qualities is to assess qualities contained in our ideas of bodies from a perception-neutral perspective. This is his philosophical programme. The quality distinction is designed as part of a conceptual framework for an ultimate, scientific account of bodies and their properties. Locke‘s notions of qualites are


not part of corpuscularian theory. Yet, Boyle‘s hypothesis illustrates the quality distinction and, more importantly, makes it intelligible, as the conceptual gap between primary and secondary qualities has revealed. McCann is therefore wrong to take Locke to advance a distinction that simply reflects our common sense conception of bodies and their causality.<256> In fact, McCann‘s view introduces incoherences into Locke‘s argument since, according to him, the common sense view of sensible qualities is that they are real. On the contrary, the topic of resemblance and of real qualities, and thus of the quality distinction, relates to an explanatory account of bodies, their properties, and our perception of them.

Moreover, it becomes manifest that Locke does not genuinely intend to develop an argument for corpuscularian theory. Pace Alexander, the Essay is not a grand epistemological argument for Boyle‘s theory by showing that the assumption of the hypothesis leads to an adequate and plausible account of our everyday experience.<257> True, Locke does claim that corpuscularian theory coheres with or, perhaps, even naturally grows out of everyday experience, and that it is intelligible on a general mechanical backdrop which is argued for. But notwithstanding the corpuscularian hypothesis is assumed as a premisse to establish epistemological claims as to the reality status of qualities and corresponding contentions of the resemblance status of ideas. And for this reason only Boyle‘s hypothesis is justified in the light of a general mechanical conception of bodies and everyday experience. The issue of Locke‘s analysis that our ideas portray bodies mainly by non-real features is an epistemological claim which is rendered


plausible by corpuscularian theory.



300, II.xxiii.8; 301, II.xxiii.10; 317, II.xxiii.37.


This view is implied by Alexander‘s contention that “Locke took over and developed the distinction from Boyle, as an essential part of the corpuscular philosophy“. Cp. Alexander (1985), 118.


Cp. 4b.


544, IV.iii.11.


134f, II.viii.9.


Cp. Mackie (1976), 14.


122f, II.iv.1.


Cp. Kienzle (1997), 97.


Cp. 3a.


Cp. 3a.


135f, II.viii.11.


Cp. §11 of the fourth edition with the first sentence of the first three editions: “The next thing to be consider‘d, is how Bodies operate one upon another, and that is manifestly by impulse, and nothing else.“ Cp. Rogers (1978), 225f. But despite this change of views on Locke‘s side by Newton‘s natural philosophy, Locke might have influenced Newton‘s philosophical, or epistemological, views much more than is traditionally thought. Rogers argues this point in a series of interesting articles as to Newton‘s “methodological empiricism“ and “epistemological empiricism“. Cp. Rogers (1979), 195-205.


Works IV, 467f. Cp. Rogers (1978), 225f.


133, II.viii.4f; 140, II.viii.22; 287, II.xxi.73; 547f, IV.iii.16.


Implicitly, Locke makes another point on dissemblance. According to the Aristotelian theory of sense perception, a quality affects the senses not by corpuscles, but by a form being specific for the property. Given Locke‘s corpuscularian account, however, no forms are involved in perception. Hence, there cannot be the alleged Aristotelian resemblance relationship between a secondary quality and its idea which is supposed to be intelligible in the light of a form which is transmitted from the quality to the mind. Cp. McCann (1994), 64.


142, II.viii.25.


317, II.xxiv.37.


381, II.xxxi.8.


Cp. 10c.


Boyle presents the following thought experiment in The Origin of Forms and Qualities: “[...] if we should conceive that all the rest of the universe were annihilated, except any [scil. any one] of these intire and undivided corpuscles ... it is hard to say what could be attributed to it, besides matter, motion (or rest), bulk, and shape. Whence by the way you may take notice that bulk, though usually taken in a comparative sense, is in our sense an absolute thing, since a body would have it, though there were no other in the world.“ Boyle (1772), III, 22. Locke‘s deviding of a grain of wheat, if continued, would eventually yield the same result, namely the set of properties which are ascribed to a single corpuscle.


Galileo (1968), VI, 347: „Per tanto io dico che ben sento tirarmi dalla necessità, subito che concepisco una materia o sostanza corporea, a concepire insieme ch‘ ella è terminata e figurata di questa o di quella figura, ch‘ ella in relazione ad altre è grande o piccola, ch‘ ella è in questo o quel luogo, in questo o quel tempo, ch‘ ella si muove o sta ferma, ch‘ ella tocca o non tocca un altro corpo, ch‘ ella è una, poche o molte, nè per veruna imaginazione posso separarla da queste condizioni; [...]“. That is: „Whenever I conceive any material or corporeal substance, I immediately feel the need to think of it as bounded, and as having this or that shape; as being large or small inrelation to other things, and in some specific place at any given time; as being in motion or at rest; as touching or not touching some other body; and as being one in number, or few, or many. From these conditions I cannot separate such a substance by any stretch of my imagination.“ Cp. Drake (1957), 274


133, II.viii.4f ; 140, II.viii.22; 287, II.xxi.73; 547f, IV.iii.16.


666f, IV.xxvi.12.


Cp. 136, II.viii.13: “Let us suppose at present, [...] [scil. the corpsucularian theory]“. Cp. 547, IV.iii.16: Locke calls the corpuscularian theory the “corpuscularian Hypothesis“.


For Locke, Newton‘s gravitational force is like a mystery for which one cannot even can propose a hypothesis, i. e. which is inconceivable. Cp. Works IV, 467f. Cp. Rogers (1978), 225f.


Cp. 10e.


139f, II.viii.22.


Boyle presents a similar thought experiment in The Origin of Forms and Qualities: “[...] if we should conceive that all the rest of the universe were annihilated, except any [scil. any one] of these intire and undivided corpuscles ... it is hard to say what could be attributed to it, besides matter, motion (or rest), bulk, and shape. Whence by the way you may take notice that bulk, though usually taken in a comparative sense, is in our sense an absolute thing, since a body would have it, though there were no other in the world.“ Boyle (1772), III, 22. Locke‘s deviding of a grain of wheat, if continued, would eventually yield the same result, namely the set of properties which are ascribed to a single corpuscle.


McCann downplays the impact of Locke‘s considerations on the intelligibility of Aristotelian theory because he believes that the quality distinction is entailed in our notion or common sense understanding of bodies and their causality. Cp. McCann (1994), 65ff. According to McCann, the mentioned experiments “are not supposed to be decisive counterexamples to the Aristotelian theory of qualities“. Cp. McCann (1994), 66.


137ff, II.viii.17f; 141f, II.viii.24f.


Cp. Works, IX, 215.


Heyd (1994), 19-27.


139, II.viii.19.


Cp. McCann (1994), 65ff. McCann (1994), 66f: “On ths view of Locke‘s arguments, the distinction between primary and secondary qualities is [...] a natural consequence of the ways we ordinary people think of the world, for better or worse.“


Alexander (1985), 7: “I believe this form of argument figures largely in Locke‘s Essay, in particular, in the way in which it uses the corpuscular hypothesis: if an account of our everyday experience and description of the world based on the best available scientific hypothesis were adequate and plausible then this would provide powerful indirect support for that hypothesis.“

[Titelseite] [Widmung] [Danksagung] [Einleitung] [1-1] [1-2] [1-3] [1-4] [1-5] [1-6] [2-1] [2-2] [2-3] [2-4] [3-1] [3-2] [3-3] [Zusammenfassung] [Bibliographie] [Selbständigkeitserklärung]

© Die inhaltliche Zusammenstellung und Aufmachung dieser Publikation sowie die elektronische Verarbeitung sind urheberrechtlich geschützt. Jede Verwertung, die nicht ausdrücklich vom Urheberrechtsgesetz zugelassen ist, bedarf der vorherigen Zustimmung. Das gilt insbesondere für die Vervielfältigung, die Bearbeitung und Einspeicherung und Verarbeitung in elektronische Systeme.

DiML DTD Version 2.0
Zertifizierter Dokumentenserver
der Humboldt-Universität zu Berlin
HTML - Version erstellt am:
Fri Sep 27 15:57:21 2002