The Cambridge History of English and American Literature in 18 Volumes (1907–21).
Volume IV. Prose and Poetry: Sir Thomas North to Michael Drayton.
§ 6. Evidence as to Order of Plays
Before generalising on what this is, we may turn to the individual plays themselves, to which we have now come in well grounded chronological advance. The Meres list is well known; it is as follows: Gentlemen of Verona, [Comedy of] Errors, Love labors Lost, Love labours wonne, Midsummer night dreame, Merchant of Venice, Richard II, Richard III, Henry IV, King John, Titus Andronicus and Romeo and Juliet. Of these, we know all—for the proposed rejection of Titus Andronicus will be dealt with presently—except Loves Labour’s Won, which has been identified, as plainly as mere conjecture can identify anything, with All’s Well that Ends Well. It is, however, all-important to observe that Meres gives no order of sequence; and that so large a bulk of work as this, greater than the whole theatre of some considerable dramatists, must have taken no short time to write, especially when we consider that the writer, during four years unquestionably and, beyond reasonable doubt, for a good deal longer, had been busily employed in acting. Twelve years possibly, since the baptism of Hamnet and Judith, six at least, if we accept the Greene reference, may be suggested as not conjectural items in the problem; eight or ten as a plausible splitting of the difference. To the fruits of this time we may add, fairly enough, if no certainty be insisted upon, Shakespeare’s part, whatever it was, in Henry VI (see below and the chapter on the doubtful plays) as well as portions or first sketches of others and, perhaps, some whole plays. But the Meres list, from its solidity, affords such an invaluable basis for investigation and classification that it is wise, in the first place, not to travel outside of it in quest of either external or internal evidence of order, or characteristics of quality.
The external evidence is of the smallest. No one of the plays except Titus was published till the year before Meres wrote, and some not till the folio of 1623. A Comedy of Errors was acted near the close of 1594. The Greene reference quotes a line of Henry VI—not a Meres play. Several, Romeo and Juliet, Richard II, Richard III, were printed in 1597; Love’s Labour’s Lost (with alterations) in the next year. Titus Andronicus was acted in January, 1593/4 and printed in the latter year, in which The Merchant of Venice, as The Venetian Comedy, may have appeared. This is all, and it will be observed; first, that much of it comes close up to the Meres date itself; secondly, that it concerns only a few of the plays. We have, therefore, to fall back on internal evidence, as it is called. But internal evidence is of very different kinds; and it is important to distinguish them from each other with the greatest possible care. One kind—or, rather, group of kinds—has figured very largely, indeed, in Shakespearean study. It is based on what may be broadly called “allusions”—passages in the plays which seem to refer to contemporaneous and known events, coincidence of the general subject of them with such events, or, sometimes, references in other more or less certainly dated work to them. It cannot be too strongly asserted, from the point of view of the present survey, that this class of evidence is open to the gravest suspicion. It ought not, of course, to be judged from its caricatures, as in the case where the mention of “pepper” is supposed to be connected with a known capture of a large cargo of that comforting spice. But, in almost all cases, it is exceedingly difficult to be sure that the coincidences are not purely imaginary. Nor is this the worst part of the matter. Admit that they are not purely imaginary—that the actual cited passages may have had some connection with the actual known events. How are we possibly to be certain that these passages were parts of the play as originally acted, much more as originally written? “Those who live to please must please to live”: the topical insertion or “gag” is one of the best known features of theatrical composition and is probably as old as Thespis in ancient times or Boileau’s imaginary pilgrims in modern. Some of Shakespeare’s plays, we know, were not printed till nearly thirty years after they were first acted; it is not impossible that, in some cases, the interval may have been even longer. Even if you can date the passage, it will give you no right whatever to date the play accordingly. If, therefore, this whole class of “evidence” is not to be ruled out bodily, it must be relegated to the utmost margin—kept strictly in the court of the Gentiles.
The other kind of internal evidence is not itself quite homogeneous, except that it is, or should be, always and entirely concerned with literary matters—with the quality, style, construction, form, character generally, of the work. Even here, there are dangers—and quite as fantastic tricks have been played in this way as in the other. By judging piecemeal, by adopting arbitrary standards of judgment and, above all, by considering, not what Shakespeare wrote but what we should like Shakespeare to have written, or think he ought to have written, it is possible to go as far wrong in this as in any way whatever. In no way, however, is it possible to reach so far and so safely, if due precaution be observed and if there be brought to the enterprise, in the first place, a sufficient study of the whole of Shakespeare’s work, and, in the second, a competent knowledge of preceding and contemporary English literature.