The poem The Bruce, by John Barbour, is preserved in only two manuscripts, one in the library of St. John’s College, Cambridge, and the other in the Advocates’ Library, Edinburgh. The former is hereafter denoted by the letter C, the latter by E. Of these E alone is complete in the sense of having both beginning and end; the first three Books and Book IV. 1-56 are missing in C. On the other hand, C bears to have been completed in 1487, E in 1489. Other things being equal, the earlier MS. must, of course, be preferred. Here, however, intervenes a series of extracts, numbering 280 lines from Books I. and II., embedded in Wyntoun’s Orygynale Cronykil of Scotland, and the two MSS. of the Cronykil are actually older than those of The Bruce. This raises a difficulty, as Wyntoun’s extracts show a goodly proportion of variations in language from the corresponding passages in E, the only other MS. which covers the same ground. Professor Skeat considers that Wyntoun’s lines are “in a better form (in the main)” than those of E;[1] but, on the other hand, we do not know Wyntoun’s method of working in such a case—how far he transcribed verbatim, how far “he modified the language of the MS. which he must have had before him.”[2] Many lines he omits, and others he obviously paraphrases; he incorporates matter from another source; and his version of The Bruce lines may quite well be due to memorial reproduction after a hurried reading. It is not otherwise easy to account for scraps of a few lines of the poem being here and there embedded in narrative independently worded or derived. There is thus no warrant for erecting this chopped-up, second-hand version of the lines in question into a canon or standard for a purely scribal transcript made for its own sake. It is needful to enter this plea in view of the separatist theory of Mr. J. T. T. Brown, for whom the passages in Wyntoun represent so much of the original or ur-Bruce, out of which our MS. and printed versions have been elaborated by a fifteenth-century editor, who, to do so, borrowed freely and with no great cunning from the works of contemporary authors.[3]

The earliest printed versions of The Bruce raise yet another issue bearing on the purity of the text. The first is apparently of the year 1571, and only one copy is known to exist.[4] It does not, however, differ materially from that of “Andro Hart” (H), published at Edinburgh in 1616. In this the language is modernized; still more so is it in the edition of four years later from the same publisher. And these seem to have been the basis of the gradually worsening issues so common in the eighteenth century, until in 1790 Pinkerton reverted to the sound critical method of having a transcript made directly from the Edinburgh MS. This again was the origin of Jamieson’s more careful edition of 1820, reprinted with a few corrections in 1869. Meantime Cosmo Innes had prepared for the Spalding Club (1856) a version which, for the first time, introduced readings from the Cambridge MS., but which, in being dressed up in a “consistent orthography,” so far reverted to the evil example of Hart. Subsequently, for the Early English Text Society, and later, for the Scottish Text Society, Professor Skeat, basing on C, but also utilizing E and H with a few readings from Wyntoun and Anderson’s issue of 1670, produced, for the first time, a full and in all respects competent text. To Skeat’s edition the present one is essentially indebted.

The main point about Hart’s edition (H) is that it supplies 39 lines not found in either MS., with an expansion of two others into eight, 45 new lines in all. The expanded portion Skeat perforce relegates to the footnotes. Twelve lines from Hart in the last book he at first accepted as genuine, but finally discarded as an interpolation.[5] He might justifiably have gone further, for he seems to me to have erred in attaching undue importance to Hart’s unsupported contributions.

This is made clear by considering the question as between C and E. Each MS. has portions not found in the other. The scribe of E furnishes his own excuse; his copy was “hurriedly written” (raptim scriptus). Consequently we are not surprised to find that he has dropped 81 lines found in C. On the other hand, the more careful Cambridge scribe has overlooked, as the best of copyists might, 39 lines preserved in the Edinburgh version. Upon analysis of these two groups a satisfactory test of character emerges. In one case only—C, Bk. VI. *85-*92, E, Bk. VI. 101-106—do we find an unexplained confusion, traces of two alternative accounts of one incident, a possibility to which Barbour refers in several instances. One line from C Skeat rejects because it results in a triple rhyme.[6] Having eliminated these, we find that of the remaining omissions in E two lines are the result of the misplacing of one;[7] eight lines are couplets which have been overlooked; four lines are necessary to complete couplets, so that their loss is due to sheer carelessness; while the bulk of the missing lines, 57 out of 80, is accounted for by the recurrence of the same word or words at the beginning or end of the line, whereby the eye of the scribe has run on from, e.g., “Toward the toun” in Bk. IX. *374 to “Toward the toun” in 374, and from “thai fand” in Bk. XIII. 446 to “thai fand,” in *450, missing all between. A parallel result is given by analysis of the 39 lines wanting in C but present in E. Six are involved in the mutual perplexity of Bk. VI.; one is merely a careless oversight, and the remaining 32 come under the main category of omission through recurrence, within a short space, of the same word or rhyme. On the whole, then, with the reservation noted above, the condition of things as between the two MSS. is quite normal; the omissions are explicable on ordinary grounds, and as the missing lines, with but one real exception, take their places again without disturbance of their neighbours, we may conclude that C and E are individual versions of a single original poem, and complementary to each other. But copyists were only mortal; an author too might see cause to alter a MS.; and the variations of reading, even with those of Wyntoun thrown in, after all supply a less serious illustration of such possibilities than do the MSS. of the Canterbury Tales from the Ellesmere to the so disturbing Harleian.

As for the lines found only in Hart’s edition, their every feature arouses distrust and suspicion. Skeat’s judgment of “almost certainly genuine” he has had to retract for 18 out of the total of 45, including the eight-line version of a couple in the MSS.[8] Those on the heart-throwing episode, Bk. XX. *421-*432, have been referred to above. Not a single example of the remaining accretions meets the test of repetition operative in the case of the MSS., or suggests its own explanation. The couplet in Bk. II., p. 25, is nothing either way; that on p. 283 is awkward; the intrusive lines on p. 321 are neither sense nor grammar; those on pp. 215, 216 can find a place only by an unwarrantable alteration of the succeeding line in both MSS., a liberty which Mr. Brown, on purely speculative grounds, lightly accepts from the very passage in question.[9] On the untimeous harangue into which Bruce is made to pass on p. 239 I have spoken in its place. In general it may be said that Hart’s contributions are clear misfits. Moreover, the circumstantial evidence seems to clinch the main conclusion. Hart, or his editor, had a turn for rhyme: to him are due the rhyming rubrics, and he added at the close of the poem a halting colophon of six lines, which in the later corrupt editions was simply merged in the poem, and is quoted as a specimen thereof in a critical historical work of 1702.[10] In XX. 610 he has barefacedly substituted a line for that of the MSS., which introduces a detail not found before the time of Bower and no doubt taken from him.[11] This throws a strong light on the origin of other lines in the same Book.[12] Thus we prove capability and inclination. Hart “modernized” the language of The Bruce, and from “modernization” to “improvement” is a tempting transition.

2. The Scribes.

The Cambridge MS. bears witness that it was completed on August 18, 1487, by the hand of “John de R., chaplain”; the Edinburgh MS. that it was “hurriedly written” by “John Ramsay” in 1489, for a Fife vicar; and the latter signature is attached to the only MS. of The Wallace, which accompanies that of The Bruce but was transcribed two years earlier. Skeat immediately pronounces that the names signify but one person, that “John de R.” is also “John Ramsay,” apparently on the logic of Wonderland, because both surnames begin with the same letter.[13] Mr. Brown, however, points out that this equation of alternative forms was highly improbable for fifteenth-century Scotland, and substitutes a reading of his own whereby the scribes are still merged in one personality as “John Ramsay” otherwise “Sir John the Ross,” one of Dunbar’s makars, the real author of The Wallace, and the wholesale redactor of The Bruce. The details of Mr. Brown’s argument and all that flows therefrom must be read in The Wallace and The Bruce Restudied.[14] Mr. Brown (if I may say so) never fails to be suggestive and interesting, and even the light which led him astray was real critical illumination; but John Ramsay, who, “as a chaplain”—which he does not claim to have been—“was entitled to the courtesy title of Sir,”[15] and took his alternative name from his office as “Ross Herald or Ross Secretary”;[16] who lightened the toil of transcribing Acts of Parliament by dropping into verse on the margin—an unjustifiable accusation;[17] and who, from the seed of Blind Harry’s “gests,” raised the prickly bloom of The Wallace, and grafted enough borrowed material on to the rough stock of the original Bruce to make it something substantially different, and did all this without leaving even a cipher as a hint to posterity—of this complex and composite personage Mr. Brown is the only begetter, and his brief and inglorious career may be followed in The AthenÆum, November 17-December 8, 1900, February 9, 1901. Mr. Brown, of course, can still claim that the problem of late redaction remains, whoever the guilty one may have been.[18] On this understanding I deal with it elsewhere.[19]

For the MSS., it needs but a slight examination to show that they are from different hands. The fifteenth century had no “consistent orthography,” but a scribe would probably have of himself; would not, at the least, exhibit the systematic differences that mark the MSS. in question. That the differences are due to the scribes is indicated by their occurrence even in proper names where E is, on the whole, much more accurate than C.[20] Add that C offers more traces of southern English influences; that it invariably gives the weak form I for the Ic or Ik of E, and substitutes can for the latter’s gan; that it regularly prefers of to the off which distinguishes E and in certain positions i for y—these with other minor peculiarities, not being vital in character, are certainly due to individual idiosyncrasies in spelling. Ramsay is an honest scribe, who, at places, cannot read his original, and leaves a blank which must be supplied from the copy of the chaplain.[21] There is thus not the faintest reason for supposing but one scribe to have been at work. At the same time the essential agreement of the two transcripts shows that we are dealing with a single, complete, familiar poem which has suffered in precision of copying from the usual mishaps incident to its manner of publication and preservation.

3. The Present Edition.

The present edition of The Bruce is based upon the printed text of the Cambridge MS., collated throughout with that of E—that is, upon the versions of Skeat and Jamieson. I have, however, adopted rather more readings from E than does Skeat, also a few more from Wyntoun, and offer some slight emendations—e.g., luffys for liffys in Bk. II. 527, oft for off in III. 194, Fyn all for Fyngall in II. 69, etc. I have profited, too, by criticism of the published text as in the adoption of Dr. Neilson’s corn-but in Bk. II. 438. The question of Hart’s version has been discussed above; it is valid only as a check upon the MSS. Variants of any interest or importance are given in the footnotes.

There has been no modernization of the language save in the case of the rubrics, which are no part of the text proper and have been contributed by the scribes or editors in order to facilitate the understanding of the poem. I have thus adhered to the spirit while modifying the letter of their work. But while avoiding any change in the language of the poem or even any attempt at a uniform spelling, I have taken a few harmless liberties with its alphabet and restricted certain of the letters to their modern values, substituting for others a modern equivalent. Skeat did this in the matter of the ancient “thorn” letter = th. In consideration of the general reader, I have gone somewhat farther, viz.:

1 The s with the ornamental curl I read as merely s; Jamieson and Skeat take it as, generally, = ss. But such alternative forms as Parys,[22] purches,[23] and purpos,[24] on the one hand, and the actual use of the tailed letter following the ordinary type in dress, press,[25] fix the usage I have adopted.[26] There are a few exceptions in which this letter is probably a contraction for is—e.g., II. 366, 459.

2. I have distributed their modern values to i, j, u, v, w. There is no advantage in preserving such forms as iugis, Evrope, wndyr: the hedge of the language—to use Lowell’s simile—is prickly enough without these accessories. Moreover, I have throughout written Edward for Eduuard or Eduard and Inglis for Ynglis (C).

3. As Skeat has substituted “th” for the “thorn” (Þ), I have done likewise with the ancient English g (?), the “yok” letter, resolving it into the digraph yh. As ultimately, in almost every case, significant of the consonantal y, I might have simply replaced it by that letter. But alternative forms, nearly without exception, show the digraph, both in The Bruce and in Wyntoun, giving yhe, yhet, yharnit, fenyhe, etc., and in Wyntoun’s extracts feyhnnyng, senyhoury, yhystir-day, bayhllys, etc. Even with the original letter the h is added as often as not. Apparently the usage, which had practically disappeared from the southern practice, was in a transitional stage on its way to its full revival in later Scots, where it became fixed, at the hands of the printers, as z, and survives in such forms as Cadzow, capercailzie, etc.[27] In I. 16, however, it has been read as g in forget, though foryhet is to be found in Ratis Raving, and in XV. 75 it is obviously z in Fi(t)z-Waryne.

4. The placing of the capital letters and the punctuation are, of course, modern.

Further, the poem in MSS. is not divided into Books, but paragraphs are denoted by the insertion of a large capital; these, as in C, are similarly marked in the text. The division into twenty Books was first made by Pinkerton, and, as the most convenient, has been adopted by Skeat in his editions. From Pinkerton also Skeat adopts the numbering of the lines. Jamieson, however, made a division into fourteen Books with a numbering to suit. Cosmo Innes gave up the Books in favour of Cantos, with a fresh renumbering. To avoid confusion I have adhered to Skeat’s divisions and numbering, which are those of Pinkerton; inconvenient though the duplicate numbers certainly are, a totally new and fourth arrangement would be much more so. To break up and make more accessible the matter, I have also introduced, where possible, the paragraphs of Jamieson distinguished by spaces, some of these, however, being found in C. They are merely for the convenience of the reader. I may, perhaps, draw attention to the critical treatment of The Bruce as an historic document without which we move greatly in the dark. The historical notes of the early editors are few and superficial. Skeat does not profess to deal with the work strictly on this line (note, vol. ii., p. 224), though he does not fail to pass unnecessary censure at several places. But some such examination as I have tried to make seems necessary in the interests of Scottish historiography.