(3) Now consider the way in which we exercise this power of mental synthesis in relation to those complex subjects which are in great part our own production. These correspond to what Locke called "mixed modes", and which he contrasted so sharply with substances. He greatly exaggerated their arbitrary character when he maintained that, whereas nature sets before us types of substances so clearly that we cannot but recognize them, we find ourselves at liberty to exercise our own arbitrary choice about the mixed modes. If we are to think and to use names to any purpose we must do so in accordance with the actual conditions of life. So regarded, many of these so-called artificial entities give us little more option than we feel when confronted with a real substance. 'Murder', for instance, is commoner than many metals, so far as we are concerned, and is quite as definite and determinate as many more material conceptions. It presents itself to human notice; it had already got a name long before any one of us was born; and if there is any question as to the correct use of the name we may appeal to the experience of accepted instances in order to decide our usage.

None the less, however, instances of this kind, taken as a class, do certainly represent a decided advance beyond the simple substances and persistent attributes with which we supposed ourselves to start. They remind us of the continual synthetic process which is required in order to describe the ever increasing complexity which the march of evolution entails. We of the present time require, as subjects of our propositions, such entities as 'The Christian Dispensation', or 'The social status of woman'; and to these subjects we may require to apply such predicates as suited to the needs of Western civilization', or 'much discussed in certain circles at the present day'. As we have already had occasion to point out, the reader must not confound verbal or formal multiplicity with real intricacy. The most complicated subject may be indicated by a very simple term, and probably will come to be so indicated if we have very frequent occasion to refer to it. A much compounded term is generally the symbol, not merely of a complicated object but of one which is either but rarely referred to, or which has come into recognition too lately to have acquired a more terse and familiar appellation.

(4) In the preceding cases the group of attributes which

constituted the subject,-whether or not there was a substance underlying it, was sufficiently large and stable for us to be easily able to conceive any assigned attribute as being separated from it or re-attached to it, without seriously compromising the integrity of the group. Its mass, so to say, was sufficiently great for it to be able to bear such slight losses and gains without its centre of gravity being seriously shifted. And such a state of things might almost seem to be a condition requisite to the possibility of a sentence being cast into the subject and predicate form.

It may, however, very well happen that the synthetic group is composed of very few attributes, so that the detachment or addition of one of them to form a predicate whilst the remainder are left to constitute the subject, will break rather seriously into its integrity. Take an example. What do we mean by the weather", in common discourse? Presumably nothing more than the general condition of the heat, moisture, and wind, experienced out of doors. Hence to predicate any one of these qualities of the weather' comes very near to counting it twice over, since the subject without it would scarcely be able to claim the name. Still more if we were to go on to predicate all three qualities at once, by saying that the weather is cold, wet and gusty; in which case the subject seems reduced to a sort of phantom or fiction'.

The fact is that assertions of this kind fall more appropriately into the form of existential propositions, and it seems to be owing merely to our familiarity with the subject-and-predicate form of statement, and to most of our language being cast in that mould, that we ever apply it to such extreme cases as this. Indeed we should quite as naturally say, It is cold, wet and gusty, as introduce a sort of fictitious subject to which these attributes are to be attached. The former is intrinsically the more appropriate; for in it we simply predicate the presence of

"There was a severe frost in the metropolis, and this, coupled with bitter winds from the East and North-east, rendered the weather extremely cold" (Times, 19. 1. 88). Contrast this with a precisely analogous verbal form where we are dealing with some material object and its group of attributes: e.g. There were many passengers inside, and this coupled with the luggage on the roof, rendered the coach extremely heavy.' It seems obvious that the 'subject and predicate' form has much better justification in the latter case than in the former,

three attributes or occurrences, whereas in the latter we feign a subject which really consists of scarcely anything else than the very three things which are assigned to it.

As this seems the most appropriate place, a few words more may be added as to the nature of these impersonal propositions. It seems to me that the most convenient way of regarding them is to consider them as constituting the extreme opposite of those which predicate comparatively accidental and trifling properties of some fixed substance, and to which the subject and predicate form seems the most appropriate. On the other hand, when we are considering some group of events with nothing in the nature of a fixed substance underlying them, it would seem that the most natural form would be some single expression or term indicative of that group, with of course some kind of inflection indicative of the fact that it was actually occurring at the time in question. It is no doubt an easy task to say what ought to be the convention adopted when we know what actually is adopted, but having made this admission we may claim that the appropriate form is one closely analogous to a mere term. What is meant, for instance, by the term 'thaw'? A certain group of events, the melting of ice and snow and the softening of what was hardened before. The utterance of this term does not tell us that such an occurrence is taking place at this or any other particular time, though it raises a strong presumption that it is known to occur some time. If we want to express the fact that it is going on at this particular time we should naturally adopt some closely analogous form, and this we find in the impersonal proposition, "It thaws". Whether or not this at all corresponds to the origin of such expressions,-whether, that is, the term or the proposition was actually prior,-there can be little doubt that this form is both convenient and suitable. In other words just as certain natural occurrences are appropriately indicated by the subject and predicate form, others are equally appropriately indicated by the simple impersonal form.

One caution must be insisted on here, which is often needed when we are engaged in defining current forms of speech, and which will recur again when we come to deal with Hypotheticals.

To speak of one kind of occurrence as being appropriate to the predicative form and another to the impersonal does not at all imply that they are exclusively so employed. As a mere

matter of actual usage, there are many cases in which we can employ either form, or in which different forms are adopted by different nations; and as a matter of possible usage it is hard to invent a case in which we could not adopt at will whichever form we please.

For instance, that convenient abstraction 'the weather' stands as supposititious subject to many a predicate which has little right to anything of the kind. Not the slightest difference can be detected in the meaning of the expressions,' It blows hard' and The weather is boisterous'. Still more instructive is the comparison between the German and the English languages. The former, as Prof. Sigwart has pointed out1, has a special partiality for the impersonal form and uses it in a multitude of cases in which we should prefer the personal. For instance, if there is any one thing which to us would seem to have a right to stand as a subject it is our own personality; and on the strength of English usage we could argue that no other form was reasonable than I am cold', when I want to express the fact of experiencing that sensation. But to the German it seems quite as appropriate to regard the sensation as an event or occurrence by itself, so to say, and to put it Es friert mich'. Again we make 'the ghost' an object and state that it 'haunts the house': the German prefers the impersonal phrase 'Es spukt im Hause'; and so on.

(5) There still remains one more case, which must be reserved however for discussion in a separate chapter, and is only noticed here because its entire omission might mislead. In all the preceding cases, in which the predicative form was habitually or occasionally used, we found that there were two elements connected together in the mind. That one which constituted the subject was in almost every case the most important or substantial one, but the same form was still occasionally employed even when the subject had dwindled down into insignificant proportions; and it could even be employed, on occasion, when that element had disappeared as a separate one. Suppose now that we have two elements which are obviously of comparatively equal importance and solidity, so to say; can we connect these together in the same way, i.e. by predication, as we connect a substance and its attributes? Suppose that they 1 See his very interesting and instructive Essay Die Impersonalien, 1888,

are two events,-simultaneous or successive, proximate or remote, when we want to make a synthesis of them by mentally connecting them in the way in which we believe them to be physically connected in nature, is there anything to prevent us from employing the predicative form? If one of them be symbolized by X and the other by Y, may we express the relation in any such form as X is Y'? Common usage is so against our doing this that it would not be easy perhaps to find a case in point of the employment of the predicative form here.

We will not therefore discuss them at the present juncture, inasmuch as the topic immediately before us was confined to the cases in which this predicative form is naturally used. I merely remark that the essential characteristic, viz. that of a mental synthesis presumed to be in harmony with the external junction of phenomena, seeming to exist here, there must be some special reason for the almost entire rejection of the predicative form of speech. We will examine the most important class of these propositions in a separate chapter.

The foregoing remarks will have served in some part to furnish an answer to our next enquiry, viz. What is the nature of the distinction between the subject and the predicate of a proposition: how are we to determine which is the subject. in any given case? Various answers are given as to the nature and characteristics of this distinction; but I need hardly remind the reader that all which here concerns us is the logician's account of the matter, not that which the grammarian may find it convenient to adopt.

One account of the distinction is to the effect that the subject is that of which something is affirmed, and the predicate is that which is affirmed of it. This is substantially the account which Mill gives, but of course without professing that it amounts to a real definition. Indeed, if it aimed at being a definition it would be something of a circular one, for we should find it difficult to understand what is meant by affirmation unless we had already distinguished between the subject and predicate; and in any case we should find it of little use to appeal to such a test in any doubtful cases. Hamilton gives an account which, with due allowance for his notional phraseology, seems to me to be somewhat nearer what we want. He holds that the