The Cambridge History of English and American Literature in 18 Volumes (1907–21).
Volume II. The End of the Middle Ages.
§ 4. Tyrwhitt’s Recension
The true restorer of Chaucer, and the founder of all intelligent study of his work, was Thomas Tyrwhitt (1730–86), fellow of Merton College, Oxford, who, in 1775, published an edition of The Canterbury Tales with prefatory matter, and a glossary dealing with the whole subject. Tyrwhitt had no theory to serve and no arbitrary standard to apply; but he had a combined knowledge of classical and medieval literature then probably unequalled in Europe, a correct ear, a sense of poetry and a singularly sane judgment strengthened and directed by legal training. He did not proceed by electing certain of the works to a position of canon and determining the reprobation of others by reference to this—a proceeding itself reprobated by the best principles of law, logic and literature. He knew, doubtless, that although The Canterbury Tales themselves are Chaucer’s beyond all reasonable doubt, no testimony that we have, from Lydgate’s onward, authenticates any particular form of them like an autograph MS., or a modern printed book issued by the author. He knew, also, doubtless, that it cannot be safe to assume that an author, especially in such days as Chaucer’, must have rigidly observed the same standard of grammar, diction and prosody at all times of his life—that, for instance, if we did so, we should, on the evidence of one edition of The Essay of Dramatic Poesy, assume that Dryden preferred to put the proposition at the end of the clause, and on that of another decide that he avoided this. He, therefore, proceeded on the only sound plan—that of sifting out, first, things certainly, and then, things probably, false—of gathering first the tares according to the advice of the parable—and so, by successive degrees, winnowing a surer and purer wheat for garnering after it had been itself threshed and cleansed from offal and impurity.
The beginning of the process was easy enough: for some things had been expressly included by Thynne in the original collection as not Chaucer’, and these or others were, in some cases, known, practically beyond doubt, to be the work of actual and identified persons. Such was the case with Gower’s and Scogan’s verses above referred to, with Lydgate’s Tale of Thebes, etc., and with the very remarkable and beautiful Testament of Cresseid, which, on the clearest internal showing, could not be Chaucer’s and which had been printed earlier as the work of the Scottish poet Henryson. The Letter of Cupid is not only acknowledged by Occleve, but actually dated after Chaucer’s death; and La Belle Dame sans Merci is not only attributed in MS. to Sir Richard Ros, but is adapted from Alain Chartier, who belonged to the next century. Other pieces Trywhitt rejected for different reasons, all valid—Gamelyn, The Plowman’s Tale, that of Beryn, The Pardoner and the Tapster, The Lamentation of Mary Magdalen, The Assembly of Ladies, etc.—while he brushed away contemptuously at a sweep “the heap of rubbish” added by Stow. He left the following verse, besides The Canterbury Tales, the two undoubtedly genuine prose works and The Testament of Love (which he had evidently not had time to examine carefully):—The Romaunt of the Roase, Troilus and Criseyde, The Court of Love, The Complaint unto Pity, Anelida and Arcite, The Assembly [Parliament] of Fowls, The Complainant of the Black Knight (which had not then been identified as Lydgate’), the A B C, Chaucer’s Dream, The Flower and the Leaf, The Legend of Good Women, The Complaints of Mars and Venus and The Cuckoo and the Nightingale, with nine shorter poems. It is, however, very important to observe that, though Tyrwhitt had read all these pieces for his glossary, he did not edit their text; and, therefore, cannot be taken as vouching fully for their authenticity. It is, for instance, pretty certain that if he had so edited The Testament of Love he would have discovered that it was not Chaucer’, whether he did or did not discover whose it actually was.
But great as was the service which Trywhitt did in sweeping out of the Chaucerian treasury much, if not all, of what had no business to be there, it was still greater in respect of the principal genuine treasure, which alone he subjected to thorough critical editing. It is quite astonishing, a century and a quarter after his work, to find how far he was in advance not merely of all his predecessors in the study of Chaucer but —in one of the most important points—of many who have followed. Whether it was in consequence of Chaucer’s uniquely clear understanding of English versification as shown in his predecessors, or of his setting a standard too high for his contemporaries, or merely of a tyrannous change in the language, it is certain that even his immediate successors (in some cases actually contemporary with him) failed to reproduce the harmony of his verse in the very act of imitating it, and that following generations misunderstood it altogether. Some have thought that this misunderstanding extended even to Spenser; but, while disagreeing with them as to this, one may doubt whether Spenser’s understanding of it was not more instinctive than analytic. Dryden frankly scouted the notion of Chaucer’s metre being regular: though it is nearly as much so, even on Dryden’s own principles, as his own. Tyrwhitt at once laid his finger on the cause of the strange delusion of nearly three centuries by pointing out what he calls “the pronunciation of the feminine-e”; and, though in following up the hint which he thus gave he may have failed to notice some of the abnormalities of the metre (such as the presence of lines of nine syllables only) and so have patched unnecessarily here and there, these cases are very exceptional. He may not have elaborated for Chaucer a system of grammar so complete and so complex as that which has been elaborated for him by subsequent ingenuity, to amend the errors of contemporary script. But his text was based upon a considerable collation of MSS. in the first place; in the second, on an actual reading—astonishing for the time when we remember that this also had to be mostly in MS.—of Chaucer’s English, as well as foreign, predecessors and contemporaries; and, in the third, on careful examination of the poems themselves with, for guide, an ear originally sensitive and subsequently well-trained. Of the result, it is enough to borrow the—in the original—rather absurd hyperbole applied earlier to Kynaston’s Troilus in the words “None sees Chaucer but in Kynaston.” It was hardly possible for the ordinary reader to “see Chaucer” till he saw him in Tyrwhitt; and in Tyrwhitt he saw him, as far as The Canterbury Tales were concerned, in something very like a sufficient presentment.
But, just as Chaucer himself had gone so far beyond his contemporaries in the practice of poesy, that they were unable fully to avail themselves of what he did, so Tyrwhitt was too far in advance of the English scholarship of his age for very much use to be immediately made of his labours. For some half-century, or even longer, after his first edition, little was done in regard to the text or study of Chaucer, though the researches of Sir Harris Nicolas threw much light on the facts of his life. But the increasing study of Middle English language and literature could not fail to concentrate itself on the greatest of Middle English writers; and a succession of scholars of whom Wright and Morris were the most remarkable among the earlier generation, and Skeat and Furnivall among the later, have devoted themselves to the subject, while, of the societies founded by the last named, the Early English Text Society is accumulating, for the first time in an accessible form, the literature which has to be compared with Chaucer, and the Chaucer Society has performed the even greater service of giving a large proportion of the MSS. themselves, with apparatus criticus for their understanding and appreciation. Complete agreement, indeed, has not been—and, perhaps, can never be expected to be—reached on the question how far grammatical and other variables are to be left open or subjected to a norm, arrived at according to the adjuster’s construing of the documents of the period; but the differences resulting are rarely, if ever, of strictly literary importance.