Of these two theories I have no hesitation in adopting the former. My reasons are these. The intention of language is to communicate our sentiments; to express what we think, feel, perceive, or desire. Hence its general character is indicative or assertive. “I believe,” “I wish,” “I see,” are affirmative sentences; and whatever variety of forms the phraseology may assume, they are all strictly significant of assertion, and all resolvable into the language of affirmation. “Go,” “teach,” “read,” are equivalent to, “I desire you to go,” “to teach,” “to read.” “Have you finished your task?” means, when the sentiment is fully expressed, “I desire to know, whether you have finished your task.” Ellipses of this kind are natural. They spring from an eagerness to impart to the vehicle of our thoughts a degree of celerity, suited to the promptitude with which the mind conceives them. Vehemence or passion, impatient of delay, uniformly resorts to them. The assertive form of expression I therefore conceive to be the parent whence every other is derived, and to which it is reducible. If this be the case, no interrogative, conceived purely as such, can claim so early an origin as definite or affirmative terms. Hence we may conclude, that who, which, when, where, were at first used as relatives, and came afterwards, by implication, to denote interrogations.
Again, we know that the meaning of an expression is frequently collected, not so much from the strict import of the terms, as from the tone or manner in which it is delivered. If I say, “he did it,” the sentence is affirmative; yet, by the tone of voice or manner of the speaker, this affirmative sentence may denote an interrogation. Thus, “he did it?” by an elevation of the voice, or the mode of notation, maybe rendered equivalent to “did he do it?” “Who did it” is in like manner an affirmative clause; but it is obvious that this form of expression, like the other now adduced, may be likewise employed to note an interrogation, thus, “Who did it?” And it is evident, that, if the ellipsis be supplied, the sentence would read thus, “I want to know who did it.” The preceding clause, however, is sufficiently supplied by the manner of the speaker. An ellipsis of this kind seems to be involved in every interrogation. If I say, “did he do it?” it is equivalent to “tell me, if he did it.” Accordingly, we find that the Latins, in such interrogations, employed only the latter clause; for an (whether), which is termed an interrogative, is, in fact, nothing but the Greek ἂν, synonymous with si (if) among the Latins. “An fecit,” did he do it? is therefore strictly equivalent to “si fecit” if he did it, the former clause “tell me,” being understood, and its import supplied by the manner of the speaker, or the mode of notation.
Besides, let any person ask himself what idea he annexes to the word who, considered as an interrogative, and I am persuaded he will be sensible that he cannot form any distinct conception of its import.
I am inclined therefore to think that interrogatives are strictly relatives: and that these relatives, by the aid of voice, gesture, or some explanatory circumstance, answer the purpose of interrogation.
In using these pronouns interrogatively, it is to be observed, that who and which are each applied to persons, which is not the case when they are employed as relatives. This difference, however, is to be observed, that when the pronoun which is used interrogatively, and applied to persons, it is generally, if not always, understood that the character of the individual, who is the object of inquiry, is in presence of the inquirer, or is in some degree known. Who is more indefinite. If I say, “which is the man?” I mean “who of those now before me?” or of those who have been described? Agreeably to this notion, we say, “which of the two,” not “who of the two,” was guilty of this crime?
If I say, “Who is the man that will dare to affirm?” it implies that I am entirely a stranger to him, and that I even doubt his existence. “Which is the man?” not only implies his existence, but also that the aggregate of individuals, whence the selection is made, is known to me.
What is also used interrogatively, and is employed in introducing questions, whether the subject be persons or things, as, “What man is that?” “What book is this?” When no substantive is subjoined, it is then wholly indefinite, as, “What is man that thou art mindful of him?” When we inquire, therefore, into the character of any person, and not for the individual himself, it is to be remembered that we employ this pronoun, and not who or which.
There seems to be the same difference between who and what definite, as between who and which. If I say, “What man will dare to affirm this?” and “Which man will dare?” &c., it is obvious that the former interrogatory is more indefinite than the latter; the one implying a total ignorance of the individual, and some doubt of his existence; the other, that he is one of a number in some degree known to the inquirer.
When any defining clause is subjoined, either may be used, as, “What, or which, man among you, having a hundred sheep, and losing one, would not leave the ninety and nine?”