Since the significatory function of words is especially vital for the achievement and expression of human knowledge, Locke devoted an entire Book of the Essay to his careful examination of "the Nature, Use, and Signification of Language." [Essay II xxxiii 19] On Locke's view, language is the instrument of all human social interaction, for the employment of which we are provided with three crucial requirements: First, we have appropriate physical organs with which to form a great variety of articulate sounds. Second, we have the capacity for using these sounds as "Signs of internal Conceptions," with which we communicate our thoughts to each other. Third, in order to avoid the inconvenience of trying to have unique names for everything, we have the ability to employ words as general terms applied commonly to many particular things. [Essay III i 1-3] Lesser creatures lack the third requisite and superior beings somehow communicate (in ways we cannot imagine) without the first, Locke supposed, but human beings rely upon spoken language as a primary vehicle for communication. Like thought itself, language is designed to serve the practical needs of human life.
The question, then, is how we achieve our communicative goals, and Locke proposed a deceptively simple theory: words signify ideas. Since my own ideas are inaccessible to others, I employ the articulate sounds of human speech as extrinsic sensible objects by means of which to convey my thoughts to others. The absence of any universal human language, Locke argued, shows that the connection between each word and the idea it signifies is not natural but purely conventional, an association established by "voluntary Imposition." [Essay III ii 1-2] Frequent repetition renders this association so intimate that some ideas may be induced more easily by words than by their referential objects, yet the freedom with which individuals form the requisite association often makes it difficult for a speaker to be sure that the appropriate idea has actually been induced in a hearer. [Essay III ii 6-8] Both the equivocal use of terms and the invention of needless jargon by the learned, Locke supposed, amount to violations of the conventional agreements upon which language is properly founded. [Essay III x 5, 11]
There is a fundamental problem here: since ideas differ from person to person, and even in a single individual at different times, and since the association of word to idea is purely voluntary, even when secured by conventional agreement, it follows that the correct signification of a particular use of any word depends wholly upon the particular idea in the mind of its speaker, to which the hearer has no access except through the mediation of the word. [Essay III ii 2-4] On this view of language, it is always possible that two people interact with each other verbally even though they do not achieve genuine communication because they do not associate similar ideas with the words they employ. In principle, I can never be sure that the sensory idea you experience in association with the word "red" is at all like the one I have when I use that word.
Locke officially granted only two kinds of exception to the general principle that every meaningful word signifies some idea. The first kind of exceptional case includes negative terms"nothing" or "ignorant," for example. We don't need a distinct idea in mind for each of these words, Locke held, since we already have the positive idea whose absence each negative term signifies. [Essay III i 4] The other, more significant, exception comprises syncategorematic terms, what Locke called "particles"such words as "is," "of," and "but." Such words are not associated with particular ideas, Locke supposed, but "are all marks of some Action, or Intimation of the Mind." Thus, on theoretical grounds, these terms are techinically insignificant, but serve the important role of providing helpful guidance for interpreting the mental dispositions upon which human reasoning so often depends. [Essay III vii]
The use of general terms is one of the most vital features of linguistic competence. Even though existing thingsincluding words themselves are invariably particular, Locke held that three considerations demand that some words be employed in a general signification: having a distinct word for each and every particular thing would exceed our linguistic and intellectual capacities; unique names for particular ideas would make it impossible to communicate with others who do not share exactly the same ideas; and the most significant varieties of knowledge are precisely those that comprehend many particulars under some general rule. Thus, we have a natural prejudice in favor of general terms. Only under special circumstances will the practical needs of life override it in favor of having distinct names for particular things. [Essay III iii 1-5] Since the signification of any word is an idea, Locke supposed that the signification of every general term must be an abstract idea, employed by the mind in reference to many particular things.
General and Universal, belong not to the real existence of Things; but are the Inventions and Creatures of the Understanding; made by it for its own use, and concern only Signs, whether Words, or Ideas. Words are general, as has been said, when used, for Signs of general Ideas; and so are applicable indifferently to many particular Things. [Essay III iii 11]Thus, on Locke's view, the classification of particular things into sorts or kinds, denominated by general terms, has no direct foundation in nature; it is, instead, the end result of our own complex process of abstraction. [Essay III iii 12-14]
Essences are nothing more than abstract ideas formed by the mind to provide significance for general terms, and their presumed immutability derives from our own power to retain the precise content of an abstract idea even in the absence of conforming instances. [Essay III iii 19-20] Formed by the mind, abstract ideas may differ from person to person, resulting in disputes over the applicability of the general terms that signify them. For "ordinary Conversation," however Locke held that minor variations of this kind may be tolerable, even for instances when one person's conception happens to be far more comprehensive and accurate than the other's. [Essay III vi 20, 31]
Even though our urge to classify things into sorts arises from the observation of genuine similarities among them, then, it is satisfied only by the provision for practical communication with general terms that designate abstract ideas. Every distinct abstract idea creates a distinct species of things by governing the applicability of a distinct general term. [Essay III vi 38] As Locke pointed out, this makes it easier for us to classify artifacts than to discover the presumed genera of natural things. Having devised the artificial things themselves, we already have a conception of their structure and operations, and this provides a rich basis for the abstract ideas by means of which we sort them. [Essay III vi 40-43] It's easier to be clear about what "clocks" are than it is to decide whether or not "cardinals" belong to the "finch" family.
In general, Locke denied that human classificatory activity is ever governed by reference to natural kinds. The species of mixed modes by means of which we differentiate among human actions, for example, are devised entirely for our convenience in communication; the abstract ideas that represent such actions are held together precisely by their association with the general terms that signify them in ordinary speech. [Essay III v 9-11] Even in the case of substances (to which we will return shortly) Locke argued that a mistaken belief that our classification must be drawn by reference to natural kinds or substantial forms amounts to an abuse of language, since it is founded upon a pair of false suppositions: that there are such natural kinds and that we have knowledge of them. [Essay III x 20-21] The false expectation that classification is natural can only lead to profound skepticism about the possiblity of general knowledge.
Locke drew a careful distinction between two senses of the word "essence." Understood etymologically, as the very being of a thing, the real essence can only be that particular internal constitution from which all of the perceived qualities of a thing causally flow. Understood scholastically, as the ground for human classification of things into species and genera, on the other hand, the nominal essence is nothing more than the abstract idea, conformity with which justifies application of the associated general term. [Essay III iii 15] For both simple ideas and modes, he supposed, these two "essences" invariably coincide, but for substances they are always different. Sortal terms for kinds of substances signify abstract ideas whose content rarely even appears to coincide with an internal stucture or compositions from which their features might be presumed to flow. The nominal essence signified by the general term "human being," for example, would just be the idea by comparison with which we recognize individual instantiations of our own species, not the unknown (and perhaps unknowable) genetic structure that generates the development of a human life. [Essay III vi 1-3]
Locke held out for the possibility that there is some real essence (perhaps a set of primary qualities) that do in fact generate all of the observable features (many of them secondary qualities and powers) that make individual things what they are. But since our knowledge includes neither a direct awareness of that inner essence nor an understanding of the causal processes by which it produces ideas of its features in us, it would be pointless and counter-productive to suppose that such knowledge is required for the practical task of classifying them. [Essay III x 17-18] When we try to use general terms to speak about the true nature of either material or thinking substances in themselves, then, we are pursuing a fool's errand and must of necessity fail to accomplish what we set out to do.
The practical use of general terms to sort out varieties of familiar substances is a different process altogether. Here, Locke supposed, a speaker observes the natural regularity with which certain features are observed to occur together, combines and abstracts the ideas of these features to form the abstract idea of a kind of substance, associates a general term with this abstract idea, and then applies the term to new instances of the same experience, without claiming thereby to fathom the true nature or real essence of such things. [Essay III vi 46-48] In order to serve our practical communicative needs, Locke supposed, the process must employ nominal rather than real essences. We find it convenient to associate a single general term to each abstract idea, rather than repeatedly generating a tedious list of specific features. But it follows that the distinction of species is entirely a product of the human understanding rather than a fixed aspect of nature itself. [Essay III vi 34-35]
On the one hand, this makes the (nominal) essence genuinely ingenerable and incorruptible: each general term designates an abstract idea whose pristine purity is unaffected by the existence or non-existence of any things that conform to it. [Essay III vi 19-20] On the other hand, it also makes the distinction of species perfectly arbitrary, framed for the convenience of human discourse rather than out of any respect for the real essences of natural things. [Essay III vi 38-41] For each sort of substances, we choose a few "leading Qualities" as characteristic marks of such things, to which we commonly add a few other featuresincluding especially the active and passive powers that we have found often to coexist with these marks, forming the nominal essence by virtue of which we determine the applicability of a general term. [Essay III xi 19-21] When it comes to the classification of substances into sorts, it's all about us.
Although simple ideas carry with them a presupposition of real existence, Locke held, the names of simple ideas signify both the real and the nominal essences of the qualities they represent. [Essay III iv 2-3] Like the simple ideas themselves, which are involuntarily received in perception, the names of simple ideas have non-arbitrary content. We find it difficult (and, usually, unnecessary) to classify them into sorts, and when we do soas, for example, with colors or soundsit is typically by reference to their perceptual origins rather than any supposed ideational similarity. [Essay III iv 16-17] Similarly, because each simple idea is a uniform perception, easily retained and intended to conform only to itself as an archetype, our names for simple ideas are rarely vulnerable to imperfection; the more simple the idea, the less likely we are to misuse the word that expresses it. [Essay III ix 18-19]
One consequence of all this is that the names of simple ideas are indefinable. At peril of an infinite regress, the provision of verbal definitions must ground out on some indefinable terms, and since simple ideas have no component parts that could be assembled under the direction of an appropriate definition, their names are perfectly suited to that role. Scholastic efforts to offer definition of such simple ideas as motion or light, Locke argued, are ludicrous precisely because they so patently fail to produce any new idea in the minds of those who hear them. Imagine the comparable folly of trying to provide auditory definitions of visual ideas, or vice versa. [Essay III iv 4-11]
The names of mixed modes, on the other hand, are general terms that signify abstract ideas governing species of human actions. These names clearly are definable, since it is always possible in principle to offer a comprehensive list of those simple ideas which, assembled together by the mind, would constitute the appropriate complex idea. That's why we don't need to witness an act of sacrilege or experience the resurrection of the dead in order to understand what kind of events those terms would signify. [Essay III v 1-5] Thus, Locke supposed that the initial formation of the abstract idea (and the stipulation of the name associated with this mixed mode) can be freely performed without any possibility of error. It is only the subsequent use of the same word by other members of the same language-community that become vulnerable to mistake because they may fail to associate the word with the already-intended idea. [Essay III vi 43-45]
Here, even more obviously than in the case of substances, our acts of naming and classification are purely arbitrary, lacking any natural foundation. We do not discover the categories of human action (or their moral significance) as existing patterns of the natural world, but rather invent them in accordance with our own decisions about how to pursue practical life amongst one another.
[Essay III v 6-9]
The vocabularies of morality, religion, and law are entirely of our own devising. This renders the use of such terms less problematic, Locke noted, and makes their misuse and confusion even less excusable.