be able to tell directly and clearly what they are. The improvement of our grammatical code in this respect, was one of the principal objects which I thought it needful to attempt, when I first took up the pen as a grammarian. I cannot pretend to have seen, of course, every definition and rule which has been published on this subject; but, if I do not misjudge a service too humble for boasting, I have myself framed a greater number of new or improved ones, than all other English grammarians together. And not a few of them have, since their first publication in 1823, been complimented to a place in other grammars than my own. This is in good keeping with the authorship which has been spoken of in an other chapter; but I am constrained to say, it affords no proof that they were well written. If it did, the definitions and rules in Murray's grammar must undoubtedly be thought the most correct that ever have been given: they have been more frequently copied than any others.
20. But I have ventured to suggest, that nine tenths of this author's definitions are bad, or at least susceptible of some amendment. If this can be shown to the satisfaction of the reader, will he hope to find an other English grammar in which the eye of criticism may not detect errors and deficiencies with the same ease? My object is, to enforce attention to the proprieties of speech; and this is the very purpose of all grammar. To exhibit here all Murray's definitions, with criticisms upon them, would detain us too long. We must therefore be content to take a part of them as a sample. And, not to be accused of fixing only upon the worst, we will take a series. Let us then consider in their order his definitions of the nine parts of speech;--for, calling the participle a verb, he reduces the sorts of words to that number. And though not one of his nine definitions now stands exactly as it did in his early editions, I think it may be said, that not one of them is now, if it ever has been, expressed grammatically.
21. FIRST DEFINITION:--"An Article is a word prefixed to substantives, to point them out, and to show how far their[68] signification extends."--Murray, and others, from, Lowth's Gram., p. 10. This is obscure. In what manner, or in what respect, does an article point out substantives? To point them out as such, or to show which words are substantives, seems at first view to be the meaning intended; but it is said soon after, "A or an is used in a vague sense, to point out one single thing of the kind, in other respects indeterminate; as, 'Give me a book;' 'Bring me an apple.'"--Lowth, p. 11; Murray, p. 31. And again: "It is of the nature of both the articles to determine or limit the thing spoken of."--Murray's Gram., 8vo, p. 170. Now to point out nouns among the parts of speech, and to point out things as individuals of their class, are very different matters; and which of these is the purpose for which articles are used, according to Lowth and Murray? Their definition says the former, their explanations imply the latter; and I am unable to determine which they really meant. The term placed before would have been better than "prefixed;" because the latter commonly implies junction, as well as location. The word "indeterminate" is not a very easy one for a boy; and, when he has found out what it means, he may possibly not know to which of the four preceding nouns it ought to be referred:--"in a vague sense, to point out one single thing of the kind, in other respects indeterminate." What is this "vague sense?" and what is it, that is "indeterminate?"
22. SECOND DEFINITION:--"A Substantive or Noun is the name of any thing that exists, or of which we have any notion."--Murray, and others. According to his own syntax, this sentence of Murray's is wrong; for he himself suggests, that when two or more relative clauses refer to the same antecedent, the same pronoun should be used in each. Of clauses connected like these, this is true. He should therefore have said, "A Substantive, or Noun, is the name of any thing which exists, or of which we have any notion." His rule, however, though good against a text like this, is utterly wrong in regard to many others, and not very accurate in taking two for a "series" thus: "Whatever relative is used, in one of a series of clauses relating to the same antecedent, the same relative ought, generally to be us
