Locke on Our Knowledge of Substances

In book IV of his Essay Concerning Humasn Understanding, Locke argues that our knowledge of substances is inherently limited, and that there is very little about the world that we may come to know. In making this claim, Locke is in complete opposition to Bacon, who saw no inherent limits to our scientific knowledge. I will herein examine Locke's argument, and point out where Locke went wrong, for it has turned out that Bacon was right, and that we have made scientific advances Locke would have thought impossible.

At the beginning of book IV of the Essay, Locke asserts that knowledge is "the perception of the connection of and agreement, or disagreement and repugnancy of any of our ideas" (bk. IV, ch. I, §2). Locke then gives a twofold classification of knowledge: knowledge may be classified by its object, i.e. what sort of agreement or disagreement of ideas that one perceives; 1 or by the way we come to have that knowledge. 2 When classified by its object, there are four varieties of knowledge (bk. IV, ch. I, §3): (1) perception of identity or diversity of ideas, (2) perception of relations between ideas, 3 (3) perception of coexistence, or necessary connection, between ideas, or (4) perception of the real existence of ideas. Under category (1) falls that knowledge that Kant calls "analytic"; the knowledge of the other categories is what Kant would call "synthetic." The knowledge that falls under category (4) is knowledge of simple existential propositions: "God is," or "There is something black." The knowledge that falls under category (3) concerns whether two ideas exist in the same subject, and may be either existential, "There is something that is both F and G," or universal, "All Fs are G." Category (2) is the most puzzling, and appears to be merely a catch-all category for all knowledge that doesn't fit nicely under the any of the other three categories. Some geometrical knowledge, such as knowledge of the proposition "Two triangles upon equal bases between two parallels are equal," falls under category (2) (bk. IV, ch. I, §6), as does all our knowledge of moral truths (bk. IV, ch. III, §18). Category (4) "is the largest field of our knowledge," according to Locke (ibid.), but Locke has relatively little to say about this sort of knowledge, perhaps because "it is a hard matter to tell when we are at an end of such discoveries" (ibid.).

In chapter II of book IV, Locke gives us a classification of knowledge according to its origin, or, as Locke puts it, by its "degree": knowledge may be either intuitive, demonstrative, or sensitive. We have intuitive knowledge whenever "the mind perceives the agreement or disagreement of two ideas immediately by themselves, without the intervention of any other" (bk. IV, ch. II, §1). Knowledge of all the truths of category (1), those Kant would call "analytic," is intuitive (bk. IV, ch. III, §8), as is knowledge of certain simple mathematical truths, such as "three are more than two and equal to one and two" (bk. IV, ch. II, §1), as well as certain simple moral truths (bk. IV, ch. III, §18). Demonstrative knowledge is that known by the intervention of other ideas, that is by means of a proof. Locke gives the following geometrical example of demonstrative but non-intuitive knowledge, referring to the classic proof that the internal angles of a triangle add up to 180 degrees:

Thus, the mind being willing to know the agreement or disagreement in bigness between the three angles of a triangle and two right ones, cannot by an immediate view and comparing them do it: because the three angles of a triangle cannot be brought at once, and be compared with any other one, or two, angles; and so of this the mind has no immediate, no intuitive knowledge. In this case the mind is fain to find out some other angles, to which the three angles of a triangle have an equality; and finding those angles equal to two right ones, comes to know their equality to two right ones. Each premise of a proof must be intuitively known to be true, and each inference intuitive known to be truth-preserving, in order for the proof to yield demonstrative knowledge (bk. IV, ch. II, §7). All of our intuitive and demonstrative knowledge is certain (bk. IV, ch. II, §§1 and 4). Thus Locke's account of intuitive and demonstrative knowledge is essentially the account Descartes gives of all knowledge in Rules for the Direction of the Mind.

Some of our knowledge of category (3) co-existence and category (4) real existence is intuitive or demonstrative. My knowledge of my own real existence is intuitive, according to Locke, and my knowledge God's real existence is demonstrative (bk. IV, ch. III, §21). Our knowledge of the necessary connection between triangularity and trilaterality is presumably either intuitive or demonstrative. But many of our category (3) and (4) knowledge will be neither intuitive nor demonstrative; it will be of a third type, sensitive knowledge (bk. IV, ch. II, §14). My knowledge that there is something red before me, knowledge of category (3), or my knowledge that there is something that is both red and solid, knowledge of category (4), is sensitive knowledge. Sensitive knowledge, according to Locke, does not reach the degree of certainty had by intuitive or demonstrative knowledge, but somehow manages to go "beyond bare probability" (ibid.). 4 Sensitive knowledge reaches "no further than the existence of things actually present to our senses" (bk. IV, ch. III, §5), and since only a finite number of particular things can be present to our senses (including, presumably, memory), universal truths, or truths that entail the existence of an infinite number of things, cannot be the objects of sensitive knowledge.

Any belief we have that is not intuitively known, demonstratively known, or sensitively known, does not constitute knowledge at all, but is mere faith or opinion, according to Locke, although it may be "probable."

We may now see Locke's argument that our knowledge of substances, that knowledge that Bacon claims that his method of induction will lead to, will forever remain quite limited, as an argument by elimination: certain propositions about substances can be neither intuitively known, demonstratively known, or sensitively known, and thus any belief of this sort we have is mere faith or opinion. The sort of knowledge of substances that Locke claims we will never have is knowledge of category (3), perception of coexistence of ideas, particularly necessary coexistence or "incoexistence" of "secondary" qualities, that is, knowledge of universal propositions of the form "All Fs are Gs" or "Nothing is both F and G," where F or G is the name of a secondary quality. Thus Locke is denying the Baconian claim that we can ever attain "true and certain axioms of knowledge," which give us "another nature . . . that is interchangeable with a given nature" (Novum Organum, bk. II, §4). Locke has no wish to deny that we sometimes have knowledge of particular claims of coexistence, e.g. "Something is both red and solid," only that we will never have knowledge of such universal propositions as "Everything that is hot has insensible parts in rapid motion."

Ruling out the possibility of sensitive knowledge of universal propositions is easy. As I pointed out above, Locke holds that sensitive knowledge reaches "no further than the existence of things actually present to our senses" (bk. IV, ch. III, §5), so any knowledge of a universal proposition cannot be sensitive knowledge.

Ruling out the possibility of intuitive or demonstrative knowledge of substances of the sort in question requires that we first make two distinctions that Locke relies heavily upon. First, we must make a distinction between primary and secondary qualities; second, we must distinguish between the nominal and real essences of types of substances. The distinction between primary and secondary qualities is first made in book II, chapter VIII, section 9. Primary qualities are those that are in the substance whether we perceive the substance or not: these include "bulk, figure, number, situation, and motion or rest of their solid parts" (bk. II, ch. VIII, §23) Secondary qualities are those that are merely powers to produce some sensation in a perceiving mind: these include colors, sounds, smells, and tastes. Secondary qualities are dependent upon a perceiving mind, and when a thing is not being perceived, it ceases to have these qualities: "Can anyone think any real alterations are made in the porphyry by the presence or absence of light; and that those ideas of whiteness and redness are really in porphyry in the light, when it is plain it has no colour in the dark?" (bk. II, ch. VIII, §19). Locke says several things about this distinction, two of which it is difficult to take with seriousness, but a third of which shows a genuine insight on Locke's part, and is important to his distinction between nominal and real essence. Locke claims that the primary qualities are "utterly inseparable from the body, in what state soever it be; and such as in all the alterations and changes it suffers, all the force can be used on it, it constantly keeps" (bk. II, ch. VIII, §9). If we took this as the criterion for making the primary/secondary distinction, this would rule out shape as a primary quality; and it is clear that Locke admits shape as one of the primary qualities. Locke also claims that our ideas of primary qualities bear a genuine resemblance to the objects that have the primary qualities in question, whereas our ideas of secondary qualities do not (bk. II, ch. VIII, §15). This cannot be taken very seriously either, as Berkeley pointed out. For if we suppose that substance dualism is true, then there can be nothing that less resembles a physical object than an idea. If, on the other hand, we accept materialism, and hold perhaps that an idea is just a bunch of neurons firing in the brain, my idea of neurons firing in a brain will bear a genuine resemblance to its object, but my idea of a cube 10 feet on each side will be neither cubical nor 10 ft. on each side.

Locke's genuine insight, however, was his view that the secondary qualities of a thing are dependent on the primary qualities of a thing (and, we should add, on the primary qualities of the mind that it interacts with) (bk. II, ch. VIII, §14). This view of Locke's anticipates the distinction contemporary materialists make between the fundamental physical properties and other properties: the fundamental physical properties are those upon which all other properties supervene. The contemporary materialist says, "There can be no difference without a physical difference"; Locke would say, "There can be no difference in secondary qualities without a difference in primary qualities." Contemporary materialists would add to Locke's list, of course: in addition to spatial properties and solidity, i.e. mass, we add electric charge, spin, and perhaps those properties that physicists tell us that some quarks have, such as "charm" and "strangeness."

With the primary/secondary quality distinction made, we can make sense of Locke's distinction between nominal and real essences of types of substance (bk. III, ch. VI, §2). Every general idea of a type of substance is a complex idea made up of some simple ideas, usually at least some of which are ideas of secondary qualities. The idea of gold, for example, might be made up of the ideas of yellowness, heaviness, and malleability, along with the idea of a substratum in which these qualities inhere. 5 The nominal essence of gold, then, is just that set of simple ideas that make up our compound idea of gold. We may say that gold is essentially yellow, meaning that yellowness is part of the nominal essence of gold. 6

The real essence of a kind of substance, on the other hand, is the set of primary qualities, in all likelihood insensible ones, on which the secondary qualities in the nominal essence depend (along with any primary qualities that are a part of the nominal essence). Let us suppose that the secondary quality of yellowness depends on primary quality Y (possibly some texture of the surface of an object), and that malleability depends on primary quality M (some property of the molecular structure of an object), and that heaviness depends on primary quality H. The real essence of gold, then, consists of Y, M, and H. We may say that gold is essentially Y, meaning that Y is part of the real essence of gold. (In Bacon's terminology, the real essence of a kind of substance is a latent schematism had by every substance of that kind.)

Locke's argument that we can have no intuitive or demonstrative knowledge of substances, i.e. of what secondary qualities necessarily coexist in substances, is as follows. We cannot have knowledge of coexistence of secondary qualities by mere knowledge of nominal essences; we could have knowledge of coexistence of secondary qualities by our knowledge of real essences, if only we had this knowledge; but we can have no such knowledge. Therefore, we can have no knowledge of the coexistence of secondary qualities.

To see that we can have no knowledge of the coexistence of secondary qualities by our knowledge of nominal essences, consider the case of gold, and ask whether gold has the property of solubility in aqua regia (nitric acid). Solubility in aqua regia is no part of the nominal essence of gold, and there is no conceptual connection between any part of the nominal essence of gold and solubility in aqua regia. We might find that some sample of gold is soluble in aqua regia, or indeed that all samples we have examined are soluble in aqua regia, but this makes it no more than probable that other samples of gold are soluble in aqua regia. We may, of course, use our nominal essence of gold to determine that all gold is yellow, or that no gold is white, but this is not genuine knowledge of coexistence of qualities, knowledge of category (3), but mere knowledge of the identity and diversity of ideas, knowledge of category (1). "All gold is yellow" is, we would now say, a mere analytic truth, and does not extend in any way our knowledge of the world.

If, on the other hand, we had knowledge of the real essence of gold, those primary qualities on which yellowness, heaviness, and malleability depend, we might be able to demonstrate that all gold is soluble in aqua regia, by using some proof similar to the proofs of geometry (which deals with the shapes of objects, which may be parts of the real essence of things), or similar to the deductions of Newton's system in the Principia, which Locke apparently thought of as demonstrative knowledge (bk. IV, ch. I, §9). Locke, however, denies that we could ever have such knowledge, for two reasons: (1) we can never know what the insensible primary qualities of a body are (bk. IV, ch. III, §11), and (2) we can never know what sort of connection there is between a primary quality and a secondary quality (bk. IV, ch. III, §12).

This knowledge of real essences, however, was exactly the sort of knowledge that Bacon claimed his method would yield, and is exactly the sort of knowledge that we take ourselves to have now, at least in certain cases. Gold has as its real essence a certain atomic structure: its nucleus is made up of 79 protons and 121 neutrons, and has a certain number of electrons in each "shell" surrounding the nucleus. These properties dictate which wavelengths of light will be reflected and absorbed by gold, giving gold its characteristic yellow color. They dictate that gold has the weight it does, and that it will interact chemically with nitric acid. Locke was quite impressed by the physics of Newton, but he did not live long enough to witness the chemical and atomic revolution, in which people finally came to have knowledge of Locke's "real essences" and Bacon's "latent schematism." Bacon saw no inherent limits to scientific progress, whereas Locke was quite the pessimist about the future of science. History has come down squarely on Bacon's side.

1. This may be read, roughly, as the type of proposition one has knowledge of.

2. Knowledge may also be classified as either actual or habitual (bk. IV, ch. I, §8), which corresponds to the contemporary distinction between occurrent and non-occurrent knowledge, but this is of relatively little interest.

3. Locke admits that categories (1) and (3) are also perceptions of relations between ideas, for "identity and co-existence are truly nothing but relations", but "they are such peculiar ways of agreement or disagreement of our ideas, that they deserve well to be considered as distinct heads, and not under relation in general" (bk. IV, ch. I, §6).

4. Locke's attempt to reply in ch. II, §14, to Cartesian skepticism about knowledge derived from the senses, which would have it that all sensitive knowledge is at best merely probable, is admittedly unimpressive.

5. The concept of a substratum in which all qualities, primary or secondary, must inhere, is one of the most problematic in Locke; but we might excise this from Locke's theory without doing any violence to the primary/secondary quality distinction, or the distinction between nominal and real essence.

6. We should note that essences, according to Locke, are of types of things, and not of individual things. If we say that an individual thing is essentially F, we are always thinking of it as being of some kind or other (bk. III, ch. VI, §4). This statue is essentially yellow, if we are considering it as being gold, but it is not essentially yellow if we are considering it as being metal, for yellowness is not a part of the nominal essence of metal. This doctrine of Locke's is echoed in the twentieth century by Quine, when he objects to quantified modal logic. It makes perfect sense, Quine claims, to say that it is necessarily the case that all bachelors are unmarried, but absurd to say that there is something which is a bachelor and is necessarily unmarried.

Back to Brock's Philosophy Page


Copyright © 1997 Carl Brock Sides.
Permission granted to distribute in any medium, commercial or non-commercial, provided all copyright notices remain intact.