into an idea or image of that whereof it is the name without the intervention of the idea of the spoken name? I answer that for the most part they do not translate them at all. The practised mind and eye passes along those signs, which it is accustomed to see combined, in exactly the same way as it passes along the sounds, or ideas of sounds, which it has frequently before met with in the same combinations. The combinations of the written signs, as of the audible, must, of course, alternately rely for their validity on the combinations of the things for which they are signs; but the mind, for the most part, is content with remembering the fact of past combination, assuming that that combination is justified by experience, and glides easily along accustomed strings of written words, only pausing to translate into terms of representative thought the conclusion of an argument, or perhaps the outward fact conveyed by a sentence.

It is to be noticed, that the mind and eye can pass over a number of written words much faster than the mind and ear could follow the same words, if it were possible to articulate them clearly in the same time. The most rapid speaker does not utter a sentence nearly so fast as we could read it; but even the clearest speaker severely strains our attention if his utterance be rapid. Now, as is the direct experience, so in every character is the representation of that experience. If

S

we can follow along a sentence of written words more quickly than we could along the same sentence of spoken words, so too, if other things were equal, we could follow along a number of mental representations of written words more quickly than along the same number of ideas of audible sounds.

At present, doubtless, it is much easier to reproduce in minds the audible sounds than the written words, simply because we and our immediate forefathers have had much wider experience of the former than the latter; but every year gives a greater proportion of the attention of mankind at large to written speech, and it is conceivable that the time may come when ideas of written words shall be substituted for those of audible sounds as the symbolic matter of thought. The change would give an increased force of thought with little or no corresponding disadvantage. There would be no greater danger of error with the one set of symbols than with the other, and the only drawback would be, that we should probably stop less often even than we now do, to ask whether our formulæ were correct. How much further, and by what other substitutions the acceleration of symbolic thought may be carried out, who shall guess?

CHAPTER VIII.

ON NECESSARY TRUTHS.

It will be scarcely necessary to say, after the foundation which I laid down in the first chapter and elsewhere, that I hold that the conception of Necessary Truth is absolutely self-contradictory. A Necessary Truth is ordinarily defined as one of which we are immediately certain as soon as the terms are presented to us, or as soon as we conceive together the ideas which the proposition connects. Now if the connection between the two ideas be independent of any direct consciousness, the proposition conceived as a simple mental conjunction of ideas will be neither true nor false; it will represent nothing. Such a mental proposition, could it exist, would be perhaps a necessary thought, but in no sense a Necessary Truth. The verbal proposition stating the conjunction of ideas in the mind would be a truth, but by no means a necessary one; for whether or no it be possible to think the reverse of such a supposed proposition, it would obviously be possible to state this reverse in words. We should only have to insert or

omit a not. Though perhaps such an unthinkable conjunction of terms ought not strictly to be called a Proposition.

It is true, that if we discovered in our minds two ideas so closely connected that we could not separate them in thought, we should immediately conclude that the mental proposition or conjunction of these two ideas was true that it did represent direct consciousness which we or others had had at some past time, although perhaps we could not at the moment remember the circumstances of such consciousness. But this con

clusion of ours would be a mere deduction from the fact which we had observed in all other cases, that every strong conjunction of ideas common to a number of sane men, represented a strong and frequently repeated conjunction of facts of direct consciousness of one of the three kinds which we have so often mentioned (Sensation, Emotion, Activity), and this defence of the truth of the proposition would rob it exactly of that quality for which its propounders contend-to wit, its necessity and immediate evidence; our belief in it would be a consequence of unbroken experience, and without that experience the belief could not exist.

But in order to make assurance doubly sure on this point, it will perhaps be well to consider separately the most important of these supposed Necessary Truths, to the end that we may see whereon their superior cer

Three Classes of Necessary Truths. 261

tainty is supposed to rest, and what right they have to arrogate to themselves this exalted position. The ground has been frequently traversed by philosophers; but since I am not quite satisfied in all cases with the arguments of the Sensational Philosophers, however much I may agree with their results; and since, moreover, I have determined throughout not to pre-suppose in my readers any wide acquaintance with the writings of the philosophers, I feel that it behoves me to discuss some of the more important portions of the subject on independent grounds.

Dr. Mansel,* the ablest modern exponent of the school which clings to Necessary Truth, informs us that Necessity is of three kinds, Logical, Metaphysical, and Mathematical. Under the head of Logical Necessity we get the three great Logical Laws of Identity, Contradiction, and the Excluded Middle (all A is A. No A is not A. Everything is either A or not A). Under Metaphysical Necessity we get as a chief law the Principle of Universal Causation. (The second law which he cites under this latter head has been so frequently disputed, that it required great boldness to adduce it as an instance of Necessary Thought). Under Mathematical Necessity we get, of course, all the mathematical axioms. He mentions a fourth class of necessary laws, to which he attributes what he calls Physical

* Prolegomena Logica,' chapter iii.