(1) One of the commonest and most obvious solutions is to identify the “substance” which has the qualities (or, to use the more general scholastic expression, the accidents) with some one group of the thing’s properties which we regard as specially important or permanent. The “substance”[“substance”] is then taken to be just this group of “primary” qualities, and is said to have or possess the less permanent “secondary qualities.” For obvious reasons, the “primary” qualities have in modern Philosophy usually been identified, as by Galileo, Descartes, and Locke, with those mathematical properties of body which are of fundamental importance for the science of mechanical Physics.[[76]] And usually, though not always, the way in which the substance, as thus defined, has the secondary qualities, has been further explained by saying that these latter are subjective changes in our sensibility produced by the action of the primary qualities upon our various sense-organs. Neither of these special views is, however, necessarily involved in the identification of the substance of things with their fundamental qualities. The essential principle of the theory consists simply in the recognition of some groups of qualities as of primary importance, and the identification of the one “substance” which has the many properties with this group.
Now, it would be impertinent for us to raise any objection to the use of such a theory as a working hypothesis in the physical sciences, so long as it does in those sciences the work for which it is required. The object of the physical sciences as a body is simply to enable us to describe and calculate the course of events in nature with the highest degree of accuracy and the least complicated set of formulæ. If this end is most successfully attained by treating a certain group of the properties of sensible things as of primary importance and all the rest as mere derivatives of them, this fact of itself affords sufficient justification for the scientific use of the distinction. For the special objects of physical science any group of properties which thus lends itself to the purposes of description and calculation is of primary importance. But it is no less true that its importance for physical purposes does not afford the least ground for regarding it as equally valuable as a solution of the metaphysical problem of the meaning of substance. For instance, one reason why the mathematical properties of body are of such supreme importance for Physics is that in respect of them bodies can be treated as differing not in kind but only in number. This is why they are of such inestimable service as the basis of our calculations as to the behaviour of things. But it might very well be that the true nature of things is most fully manifested just in those points in which they are different in kind; from the standpoint of the metaphysician, a view of non-human nature, however serviceable, which rests entirely upon the aspects in which things are most alike, may be as superficial as the statistical sociologist’s view of human nature. The true being of a concrete thing may be as inadequately expressed by its mathematical properties as the true character of an individual man by a list of anthropometrical results.[[77]]
In point of fact, we can readily see that the distinction between “primary” and “secondary” qualities, when propounded as an answer to the problem about substance, leaves us just where we were before. For (1) we ascribe the primary qualities to the “substance” of the thing in just the same fashion as the secondary. The thing is of such and such a configuration, is of such and such a mass, is solid, etc.; just as it is rough, or heavy, or green. Or, again, it has configuration, mass, solidity; just as it has weight, taste, colour. Hence the old problem breaks out again with respect to the primary qualities themselves, however the list of them may be constructed. Again we have to ask, what is the it which possesses shape, mass, velocity, etc.?
(2) Moreover, the theory fails to explain the nature of that “possession” of the secondary qualities which it ascribes to the group of primary qualities. In what way, we ask, do the primary qualities have or possess the secondary? The only serious attempt to answer this question seems to be that of the numerous philosophers (Descartes, Galileo, Locke, etc.) who treat the secondary qualities as “subjective” effects of the primary qualities upon our sense-organs. Now, this familiar solution of the problem seems deficient in logic. For the one solid argument which has been advanced in favour of the subjectivity of secondary qualities seems to be the contention that they cannot be perceived without sense-organs of a special type. Colours, it is said, exist only for an eye, sounds for an ear, taste for a tongue, and so forth. And differences of structure or temporary condition in the sense-organ lead to the perception of different secondary qualities, as when, to take the stock examples, everything looks yellow to the jaundiced eye, the same water feels warm to one hand and cold to the other, and so forth.
But these considerations seem just as applicable to the supposed “primary” as to the “secondary” qualities of things. Geometrical form, for instance, is imperceptible apart from sight or touch; motion, again, and consequently change of configuration, and similarly mass, which is a ratio of accelerations, require either sight or touch for their perception. Of course, we can think of motions and masses which we are not actually perceiving, just as we can think of an absent colour or smell, and in both cases we can in reasoning about motions or masses or colours or smells abstract altogether from the presence of a percipient. But this does not affect the fact that the mathematical qualities of body are just as dependent for perception upon the presence of a percipient with suitable sense-organs as anything else. Configurations, extensions, and motions which no one perceives by sight or touch or any other sense are exactly in the same case as a colour which no one sees or a sound which no one hears. The argument from the indispensability of a perceiving organ ought logically to tell just as much in the one case as the other.[[78]]
Again, and this is a point of the first importance, experience never gives us the “primary quality” by itself. What we get in actual experience is always the conjunction of primary and secondary qualities in a concrete perception. Thus we never perceive extension apart from some special visual or tactual filling of the “secondary” kind. The extended has always some quality of colour, or texture, or resistance. An extension which is totally devoid of colour, tactual quality, and everything which belongs to the so-called sensible, non-mathematical, or “secondary” properties of body, is an unreal abstraction, got by leaving out an aspect which in actual experience appears inseparable from it, and therefore presumably illegitimate. Illegitimate, that is, when offered as an account of the fundamental reality of body, however useful for the special purposes of natural science. Thus the attempt to take the so-called primary qualities as the unitary “substance” which has or “possesses” the secondary qualities, and to dispose of these latter as “subjective,” leads to no satisfactory result. The former, too, must be merely qualities possessed by a more ultimate substance.
§ 5. Hence it constantly happens that the same writers who treat substance as identical with the primary qualities of things, alternate this view with another according to which substance is an unknowable unit of which we can say no more than that it, whatever it may be, is what is presupposed in all propositions about the behaviour of things as the “unknown substratum” of their various qualities. According to this view, the many qualities of the thing in some inexplicable manner “flow” either from the nature of its own unknown substratum or substance, or from the relations in which this substratum stands with that of other things.[[79]] Our knowledge is then held to be confined to these consequences of the unknown ultimate character of real things; we are ignorant, it is said, of the substance both of physical and of mental existence, we know only its attributes or manifestations. Or it is otherwise phrased thus: we do not know what things really are, we know only their effects on one another and on our own senses. This is, for instance, the view represented by those portions of Locke’s Essay in which emphasis is laid upon our inability in the last resort to know the true substance of things.
Now, such a general doctrine as this is manifestly open to grave objections. (1) If we are serious in maintaining the unknowable character of the substratum of a thing’s qualities, it is hard to see how the assertion of its existence can be any addition to our knowledge of the thing. To say that we are entirely ignorant of the nature of this substratum only amounts to saying in other words, that we have really no idea how the many qualities can be qualities of a single thing. If this is so, it does not appear what we gain by talking of the single thing at all as the owner or possessor of its qualities. It would, we might think, be better to abandon the confessedly unintelligible notion of a single substratum in which the qualities “inhere,” and say that the thing, for our intellect, is simply the many qualities themselves. How this view would have to be reconciled with the tacit assumption of the thing’s unity as a substance, which underlies all the judgments in which its attributes are predicated of it, we shall have to discuss more fully in the next section.
(2) A still more serious difficulty remains behind. Not only is an “unknowable substratum of qualities” a superfluous luxury in metaphysical theory, but the nature of the supposed relation between such a substratum and the attributes which “flow” from it is unintelligible. We can understand neither what a substance or substratum totally devoid of qualities could possibly be, nor yet how the various qualities of the world of things presented to our experience could “flow” as secondary consequences from one or more such substrata. We cannot conceive how things could first “be” without this being of theirs possessing any definite character, and then subsequently, in virtue of their relations among themselves, give rise to their qualities or characteristic modes of being. Nothing can be at all without being in some determinate way, and this “being in some determinate way” is precisely what we mean by the qualities of a thing. A thing cannot be without behaving in special ways towards its environment, and these special ways of behaving are the thing’s qualities. We cannot, therefore, divorce the being or that of a thing from its determinate mode of being or what, and regard the latter as something which supervenes on or is derived from the former, or the former as something which can exist without and apart from the latter. Things are not first there and afterwards in some mysterious way clothed with qualities; their qualities are simply their special way of being there. As Lotze well puts it, all such attempts to formulate a theory of the way in which the what of things flows from a mere that, are attempts to answer the absurd question how Being is made.[[80]] The notion that things have a that or substance prior to their what or quality, and consisting simply in “being” which is not this or that determinate mode of being, is thus unmeaning as well as superfluous.
§ 6. Accordingly the whole notion of a substantial unity in things behind the multiplicity of their states or qualities has been regarded with disfavour by many students of positive science. The qualities being all that interests us in things, and the notion of an indeterminate substratum contradictory, we ought, it is argued, to identify the thing and the series of its states and qualities without more ado. From this point of view the thing ceases to be an unknown somewhat, which in some mysterious way has properties; it becomes the properties themselves thought of as a collection. It is no longer the unperceived this which has warmth, redness, etc., it is the warmth, the redness, and the rest of the sensible qualities taken collectively. For phenomenalist Metaphysics, as for associationist Psychology, the thing is a “bundle of attributes” and nothing more.