words makes composition ridiculous, so the affectation of easy and common ones may make it unmanly. But not to digress. With respect to grammar, we must sometimes content ourselves with such explications of its customary terms, as cannot claim to be perfect definitions; for the most common and familiar things are not always those which it is the most easy to define. When Dr. Johnson was asked, "What is poetry?" he replied, "Why, sir, it is easier to tell what it is not. We all know what light is: but it is not easy to tell what it is."--Boswell's Life of Johnson, Vol. iii, p. 402. This was thought by the biographer to have been well and ingeniously said.
12. But whenever we encounter difficulties of this sort, it may be worth while to seek for their cause. If we find it, the understanding is no longer puzzled. Dr. Johnson seemed to his biographer, to show, by this ready answer, the acuteness of his wit and discernment. But did not the wit consist in adroitly excusing himself, by an illusory comparison? What analogy is there between the things which he compares? Of the difficulty of defining poetry, and the difficulty of defining light, the reasons are as different as are the two things themselves, poetry and light. The former is something so various and complex that it is hard to distinguish its essence from its accidents; the latter presents an idea so perfectly simple and unique that all men conceive of it exactly in the same way, while none can show wherein it essentially consists. But is it true, that, "We all know what light is?" Is it not rather true, that we know nothing at all about it, but what it is just as easy to tell as to think? We know it is that reflexible medium which enables us to see; and this is definition enough for all but the natively blind, to whom no definition perhaps can ever convey an adequate notion of its use in respect to sight.
13. If a person cannot tell what a thing is, it is commonly considered to be a fair inference, that he does not know. Will any grammarian say, "I know well enough what the thing is, but I cannot tell?" Yet, taken upon this common principle, the authors of our English grammars, (if in framing their definitions they have not been grossly wanting to themselves in the exercise of their own art,) may be charged, I think, with great ignorance, or great indistinctness of apprehension; and that, too, in relation to many things among the very simplest elements of their science. For example: Is it not a disgrace to a man of letters, to be unable to tell accurately what a letter is? Yet to say, with Lowth, Murray, Churchill, and a hundred others of inferior name, that, "A letter is the first principle or least part of a word," is to utter what is neither good English nor true doctrine. The two articles a and the are here inconsistent with each other. "A letter" is one letter, any letter; but "the first principle of a word" is, surely, not one or any principle taken indefinitely. Equivocal as the phrase is, it must mean either some particular principle, or some particular first principle, of a word; and, taken either way, the assertion is false. For it is manifest, that in no sense can we affirm of each of the letters of a word, that it is "the first principle" of that word. Take, for instance, the word man. Is m the first principle of this word? You may answer, "Yes; for it is the first letter." Is a the first principle? "No; it is the second." But n too is a letter; and is n the first principle? "No; it is the last!" This grammatical error might have been avoided by saying, "Letters are the first principles, or least parts, of words." But still the definition would not be true, nor would it answer the question, What is a letter? The true answer to which is: "A letter is an alphabetic character, which commonly represents some elementary sound of human articulation, or speech."
14. This true definition sufficiently distinguishes letters from the marks used in punctuation, because the latter are not alphabetic, and they represent silence, rather than sound; and also from the Arabic figures used for numbers, because these are no part of any alphabet, and they represent certain entire words, no one of which consists only of one letter, or of a single element of articulation. The same may be said of all the characters used for abbreviation; as, & for and, $ for dollars, or the marks peculiar to mathematicians, to astronomers, to druggists, &c. None of these are alphabetic, and they represent significant words, and not single elementary
