expressed by a transitive verb or a phrase involving one. Such names are usually found in pairs called correlatives, one for each party to the relation, e.g., 'master', servant'; ' landlord', ' tenant' ; 'debtor', 'creditor'; 'victor',

We might avoid the use of such terms by saying that one person works for another, or has rented his house, or has borrowed money from him and not yet paid it back, and so forth. In a world in which nothing affected anything else this first kind of relative terms would have no place. They express a relation existing only between two active beings. Relative terms of the second kind, on the other hand—of which we are about to speak-express no action of one thing upon another, but merely the fact that some one has compared them in a given respect.

Second sense.

In the second sense of the word terms are called 'Relative' when, like such words as 'larger', they indicate the result of a comparison with some standard which the term itself does not indicate. The most obvious examples of such terms are adjectives of the comparative degree, for with them it is necessary to name the standard in some added words in order to give the term any meaning, e.g., A robin is larger than a sparrow'. With superlatives we recognize that a standard is involved though we do not always mention it, e.g., 'I saw the loveliest picture'. Such expressions as rather beautiful', 'very beautiful', 'extremely beautiful', most beautiful' (not distinguished in Latin from ordinary comparatives and superlatives) also involve some reference to a standard. By a rather beautiful' thing we mean one that is perhaps somewhat more beautiful than the average of its kind; by an excellent' piece of work, one that is exceptionally good.

my beloved. They are perhaps occasionally adjectives: e.g., parental, though I am not sure whether this should really be called a relative term or not. They are almost always applied to persons, though not always : e.g., reagent.

It is less obvious, but no less true, that an immense number of terms 'positive' in form are really used in a comparative or relative sense. A large thing is larger than a small thing, a clean thing cleaner than a dirty thing, an intelligent creature more intelligent than a dull one; but a large rat is not necessarily larger than a small elephant, a clean stable cleaner than a dirty table-cloth, or an intelligent horse more intelligent than a dull man. In each of these cases the name of the thing to which the adjective is applied indicates the standard of comparison. A large rat is one larger than most other rats; an intelligent horse, one more intelligent than most other horses; a clean stable, one cleaner than most others, or as clean as one could reasonably hope to keep a stable. A rat is an animal, but we cannot say that a large rat is a large animal, because the standard of largeness changes as we pass from the consideration of rats to that of animals in general. What we have a right to say is that a large rat is an animal larger than most rats. In this way we retain the meaning of the word 'large' with which we started.

Dangers.

Amongst other kinds of inference Jevons mentions Immediate Inference by added Determinants, which "consists in joining some adjective or similar qualification. both to the subject and predicate of a proposition." Hyslop says of it that terms expressing quantity, such as 'large', 'long', 'small', 'short', must be used carefully; but that terms expressing quality "can be used with perfect. freedom, provided they are not used equivocally".* If more examples are needed to show that this is not the case there are plenty to be had. An Australian Bushman is a man, but an intelligent Bushman is not an intelligent man, a respectable saloon is hardly a respectable place, an energetic snail an energetic animal, nor a fast mule-car a fast means of transportation.

This "inference by added determinants" gives a good

* Elements of Logic, p. 168 (Scribners, 1894. Third Edition),

illustration of the trouble we get into when we substitute rules of verbal manipulation for thought about the things that the words are intended to denote.

The following passage from Schopenhauer shows how even a first-class author is likely to deceive himself and his readers when he makes too much use of relative terms. The passage seems to be full of meaning, but it turns out on analysis to be absolutely empty.

"This human world is the kingdom of chance and error. [1] Everything better only struggles through with difficulty; what is noble and wise seldom attains to expression, becomes effective and claims attention, but [2] the absurd and perverse in the sphere of thought, the dull and tasteless in the sphere of art, the wicked and deceitful in the sphere of action, really assert a supremacy, only disturbed by short interruptions. On the other hand [3] everything that is excellent is always a mere exception, one case in millions, and therefore if it presents itself in a lasting work, this, when it has outlived the enmity of its contemporaries, exists in isolation, is preserved like a meteoric stone, sprung from an order of things different from that which prevails here."'*

These words mean something like this: I am dissatisfied with the world because [1] the exceptionally good is exceptional, [2] what is no better [and no worse] than the common is common, and [3] anything good that is exceptional enough for me to call it excellent is very exceptional indeed.' If the value of everything in the world were increased a thousandfold a new philosopher who happened to feel dissatisfied could use Schopenhauer's very words, for 'good' and 'bad' would be interpreted then as they are now with reference to the average.

"The World as Will and Idea", Vol. I, p. 417 (Kegan Paul, Trench, Trübner & Co., 1891).

CHAPTER VI

THE RELATIONS EXPRESSED IN PROPOSITIONS

THE meaning of words and statements has been discussed enough in the foregoing chapters to guard against various kinds of rather gross blunders. Our next task is to inquire what fundamental relations of things statements of various kinds imply. This is the work of the present chapter. In the next we shall carry the same subject further by inquiring what any proposition implies about the existence of the things it mentions. After these more general discussions we shall come back and consider propositions from a more formal standpoint, taking up the difference between various forms of statement, the significance of each, and the way in which a statement in one form of words may imply the truth or falsity of a statement in some other form of words.

'This is John'; ' John is happy'; 'John is riding a horse'. It is perfectly evident that in these propositions the relations expressed by the verb 'is' are wholly different and incomparable. In the first case it helps to identify

Five

relations.

a person; in the second, it helps to tell something fundamental about his state of mind; in the third, it helps to

tell his relation to something else. How many absolutely different kinds of relation we can think of and express in propositions it is hard to say. The question is one whose full discussion belongs to metaphysics rather than to logic; and it is one about which metaphysicians have not agreed. Nevertheless, something should be said about it here; for the

very first rule of logic is to understand the meaning of the words we use.

In thinking about the question it is well to remember that in reality nothing exists but things (including persons) with their various attributes and acts. The following list is perhaps as good as any.

1. We are able, first of all, to distinguish between various individual things and to recognize the fact that through all the changes they undergo each one of them remains the same: the Paul who preached Christ was identical with the Saul who was present at Stephen's death, in spite of his change of heart and name in the meantime; the rock on Emerson's grave to-day is the very same rock that was there yesterday and that was carried there years ago. Precisely what it is that makes a thing to be the same throughout all its changes of state and circumstances,—whether the ship of Argos, repaired so often that at last none of the original material remained, was really the same ship or not,—this is a detail of metaphysics which we cannot discuss here. But in a general way every one knows what it is to recognize an old friend, or to say that he himself is the very person who performed such and such an act, or that the book in his hand is not the one he bought yesterday, though the two look precisely alike.

The first kind of proposition, then, is that which asserts or denies this unity or individual identity of a thing that then was there or did that, with one that afterwards was here or did this.

It is evident that any proposition of this kind must involve two demonstrative terms connected by the verb to be, e.g., This-is-the man who was here yesterday.

2. The second kind of relation affirmed or denied by propositions is that which exists between a thing and its qualities, states, or activities (whether we suppose it to be conscious of these or not); in short, it is a relation of subject and attribute, e.g.: Bulldogs are courageous'; this iron is