II. No true poet ever wrote limping, imperfect, or inharmonious verses.
We may add to these the plain facts, that printers are not, and never were, infallible, and that those works of which the authors themselves read the proof-sheets are in general more correct than those of which those sheets were read by others.
3.
Now the plays of Shakespeare are, as is well known, full of passages of which it is nearly impossible to make any good sense, and they abound in imperfect and inharmonious verses. On the other hand, the poems of Venus and Adonis, and of Lucrece, which he himself probably saw through the press, are almost entirely free from error, and do not contain a single unmelodious verse. The natural inference, then, is, that the defects of the plays are all owing to the transcribers and printers, and that the correction of them and the restoration of sense and melody, when possible, is the legitimate office of sound emendatory criticism. The truth of this has been felt from the very beginning; for in the 2nd folio, published in 1632, only nine years after the first, there are numerous corrections, which must have been made by the editor solely on his own authority, for, many of them being very bad, he could not have derived them from any manuscripts. And the same is the case in the subsequent folios, of all of which there are many copies in existence, like that which Mr. Collier met with to his misfortune, and which has excited such a storm in a puddle,—of which, by the way, my own opinion is that the corrections in it were made between 1744, the date of Hanmer's, and 1765, that of Johnson's edition; whence I only cite it for the corrections later than the former date. These contain manuscript corrections, some, like that copy, anonymously, others by Southern and other men of repute. In the beginning of the last century Rowe gave the first example of an annotated edition of Shakespeare's works, and from that time to the present edition has succeeded edition bearing the names of critics of various degrees of ability and eminence, but all agreeing in the necessity of revising the text and rendering it as correct as may be possible.
Of the early editors, Rowe and Pope made little more than the most obvious corrections, Warburton, always ingenious and almost always wrong, made notwithstanding some that were very good, as also did Hanmer. But they were all eclipsed by Theobald, one of the acutest emendatory critics that this country has produced, whose merits, though long clouded through the malignity of Pope, are now fully acknowledged. Capell, the next in order of time, also rendered good service; but Johnson, Steevens, and Malone have done much less than might have been expected. Of these the last was a native of Ireland, the only emendatory critic that country has produced; and, in my opinion, he is not at all inferior to his English rival Steevens. That true critic Tyrwhitt should also be noticed as an emendator of a high order. It is surprising how little has really been done in the present century; and I was perfectly astonished to find what a number of passages still remained in a corrupt or imperfect state when I ventured on the task of emendation. It would seem as if critical sagacity, low in rank as it may be, is one of those talents most rarely bestowed; and besides, print appears to exercise almost a magic power over most persons; they seem to think that what is in print cannot be wrong: it is in fact only very few minds that can fully emancipate themselves from its influence. It must, however, be remembered that alteration and critical emendation are widely different. The former is in the power of almost any one; the latter, as I said, requires a peculiar faculty. It is a curious fact that in nothing does critical sagacity show itself more than in, as I may say, seeing what is before the eyes. For an instance see the first note on Henry V. Another curious fact is this, that poets (Coleridge for example) are rarely good emendators. It is, in fine, the merest folly, and a proof of the grossest ignorance, to sneer at such labours and represent them as needless, if not mischievous. Editors, however, ought to be very cautious about introducing their conjectures into the text, and should place them only in notes, unless when they are such as must almost command assent, and the place corrected had previously yielded no tolerable sense or metric melody. It is a good rule to let well alone, as we say, and not to alter what gives tolerable good sense. I have, I believe, transgressed this rule only in a couple of places.
4.
The following instance is a convincing proof of the evil consequences of the proof-sheets not having passed under the author's own eye. Sismondi, the celebrated historian of France, wrote an historic tale named Julia Severa, which was of course written in French, and it was printed at one of the best offices in Paris; but the author himself did not see the proofs, whence, though it makes only two rather slender duodecimo volumes, there are actually whole pages of errata! I know, by the way, no better mode of weaning oneself from what I shall presently describe as printer-worship than a habit of examining the errata in books printed in the two last centuries. Further, in good printing-offices, at least in this country, it is the practice that the proofs, after they come from the compositor, should be read in the office and corrected before they are sent to the author or editor; while, even at the present day, in the Royal printing-office in Copenhagen there are no readers, and the sheets are sent out just as they come from the compositor, and what that state is must be known to any one conversant with printing-offices. Now it is very possible, or rather highly probable, if not certain, that such was the case in England also in former times; and supposing that it was the editors, and not Blount, that read the proofs, as the wretched state of the punctuation might seem to indicate, we need not feel any surprise at the very unsatisfactory state of the text of the folio of 1623. In opposition, however, to this opinion, the Cambridge editors think that "there were no proof-sheets, in those days, sent either to author or editor," and that "after a MS. had been sent to press it was seen only by the printers, and one or more correctors of the press employed by the publishers for that purpose." But on this hypothesis how are we to account for the great correctness of our author's Poems, Spenser's Shepherd's Calendar, the Works of Jonson, Drayton, and others? The following words on the title-page of Marston's Fawne, 2nd edition, offer decisive evidence that authors did read the proof-sheets of their works:—"Now corrected of many faults, which, by reason of the author's absence, were let slip in the first edition."
5.
Notwithstanding all this, there are editors of Shakespeare who, rejecting the evidence of sense, grammar, and logic, obstinately adhere to the printed text, terming that alone authority, and even holding the second and later folios to be such; and it is really pitiable to see them superstitiously retaining for instance 'tis or it's when the metre requires it is, and syncopated forms as lov'd, own'd, etc., when the verse has need of the complete word, and vice versâ. These I denominate Printer-worshipers; for it is in reality to the authority of the printer, not of the poet, that they bow. The most extraordinary instance of this propensity in existence is the retention of "strain at a gnat" in our authorized version of the New Testament, a most manifest printer's error; for no schoolboy could have made it in translating, and in all the previous English versions the word is out. Yet there it has stood uncorrected for two centuries and a half! Hardly inferior as a piece of printer-worship is the following. A stanza of a song in Fletcher's Spanish Curate (ii. 5) ends thus:
From that breath, whose native smell