It will be noticed that the conception of this battle, alike as to position and tactics, elaborated in the notes in strict conformity with Barbour, differs entirely from that now universally accepted. The engagements of the first day (Sunday) were the outcome of attempts to clear the two paths of approach to Stirling—that through the New Park, and the other on the level below St. Ninians. Both failed, and the means by which their failure was brought about determined the operations of the following day (Monday). This main engagement, however, it has been hitherto held, took place on the banks of the Burn, below or in the neighbourhood of Brock’s Brae, with the Burn separating the forces. This is pure misconception. There can be no doubt that the battle was fought on a position roughly at right angles to this—on “the playne,” “the hard feld,” or level ground east of St. Ninians, reaching back into the angle formed by the Forth and the Bannock. The main data for such a conclusion are these: (1) The English passed the night on the Carse, having crossed the Bannock; (2) the Scots attacked early next morning, and to do this “tuk the playne,” leaving their camp-followers in the Park, so that they astonished the English by their audacity; (3) in the rout many English were drowned in the Forth and in the Bannock; (4) Edward II., unable to get away, fled to the castle; (5) so did many of his men, as the castle “wes ner.” These facts, fully substantiated from both sides, are wholly inconsistent with a site of battle south of St. Ninians, and fix its position between the Forth and the Bannock. Barbour’s “pools” are the “polles” in which, according to Hemingburgh,[57] the English baggage was bogged and captured after the battle of Stirling Bridge. The English and French (and Irish) chroniclers invariably speak of the battle as that of “Stirling,” and Trokelowe calls it the Battle of Bannockmoor. For a full discussion of the matter, see my paper on “The Real Bannockburn” in Proceedings of the Glasgow ArchÆological Society, 1908-1909.

APPENDIX B
BRUCE’S SPEECH AT BANNOCKBURN

Book XII. 210-327

It is the privilege of early historians to equip their leading personages with speeches, and in its pertinent, practical character the speech here provided for King Robert is a good example of such—so good, indeed, as to suggest the probability that Barbour is working up some transmitted material. There is on record another speech attributed to Bruce, which formed part of a Latin poem on Bannockburn by Abbot Bernard of Arbroath, Bruce’s Chancellor, portions of which are quoted in the Scotichronicon.[58] This speech consists of twenty-five hexameter lines, and is a rhetorical flourish on Scottish liberty, the miseries inflicted by the English on the country, and the hapless condition of “mother Church,” closing in strains of ecclesiastical exhortation. Moreover, it immediately precedes the opening of the battle, while Barbour’s version is of the evening before. In the latter a special interest attaches to lines 263-268 and 303-317, which may be compared with the following extracts from a speech by Alexander the Great in The Vowes, one of the three romances which make up the Scottish Buik of Alexander, the translation of which from the French was probably the work of Barbour himself.[59] Alexander says:

To the same purport as these latter lines is a portion of a subsequent address;[61] and lines 325, 334 find a similar parallel in:

Of the cardinal sentiment in the speech, the origin is probably to be found in the familiar story of the Maccabees, referred to more than once in The Bruce. Judas Maccabeus was one of the typical heroes of French romance, and had one metrical romance, at least, devoted to his career. And in 1 Maccabees, chap. iv., we have:

“17. (Judas) said to the people, Be not greedy of the spoils, inasmuch as there is a battle before us.

“18. And Gorgias and his host are here by us in the mountain; but stand ye now against our enemies, and overcome them, and after this ye may boldly take the spoils.”[63]

APPENDIX C
THE NUMBERS AT BANNOCKBURN

English: One hundred thousand men and ma.

Scots: Thretty thousand, and sum deill mare.

These figures have given rise to much discussion, without any very certain result. Yet official data are not wanting—sufficient, at least, to check what is only another example of the wild conjectures of mediÆval chroniclers when dealing with numbers. Hemingburgh gives Wallace at Falkirk “about three hundred thousand men”[64]—rather more, probably, than the whole male population of Scotland. We need not be surprised, then, at how all such estimates shrink in the cold light of Exchequer figures.

Edward II. summoned all owing him military service,[65] which corroborates the statement of the author of the Vita Edw. Sec. that “the King exacted from all the service due,”[66] as well as that of Barbour—“of England hale the chivalry.” The Earls of Lancaster, Warenne, Arundel, and Warwick did not attend, for a particular reason, but sent their contingents.[67] Now, by Mr. Round’s calculations, the whole number of knights’ fees in England did not exceed 5,000;[68] Mr. Morris raises the figure to something short of 7,000.[69] The important point is, however, that in practice the assessment was only a nominal or conventional one. Thus Gloucester, with 455 fees, was assessed at ten knights.[70] Including all grades of horsemen, Mr. Morris puts “the maximum of the cavalry arm” at “about 8,000”; but, all things considered, no such number could ever take the field.[71] Edward I. had summoned his full feudal array (omnes sui fideles) for the Falkirk campaign, and Hemingburgh says that, when counted, it came to 3,000 men on armoured horses (Barbour’s “helit hors”), and more than 4,000 on unarmoured horses—say, roughly, 7,000 in all.[72] Mr. Morris, however, by a generous calculation from the rolls, arrives at 2,400 as the highest possible figure.[73] Now, it is to be noted that the author of the Vita Edw. Sec., while lauding the size and magnificence of the host that went to Bannockburn, gives 2,000 men-at-arms as apparently the total of the cavalry, since he simply adds “a considerable body of footmen.”[74] On the whole, 3,000 to 4,000 English horse is a higher limit for Bannockburn, when we consider all the difficulties of sufficient armour, remounts, and forage. Mr. Morris thinks 10,000 “impossible,” though he is here calculating on yards of frontage on a site where the battle was not fought.[75] About 7,000 is Mr. Round’s free estimate, adopting Hemingburgh’s figure for Falkirk.[76] Bain accepts Barbour’s 3,000 heavy horsemen, and suggests 10,000 light horse, but proceeds on no data.[77] Mr. Oman calculates that “three thousand ‘equites coperti,’ men-at-arms on barded horses,” means, probably, 10,000 for the whole cavalry,[78] but this traverses his Falkirk figures. England never put, nor could maintain, on the field such a mounted force, to say nothing of the difficulty of handling and manoeuvring it.

For the foot we have, fortunately, exact figures in the Foedera[79]—21,540 men all told, which would include the archers. Only the northern counties—but not all—and Wales are drawn upon, as those of the south would be for a French campaign.[80] Such had been the practice of Edward I., whose levies from the northern counties and Wales ranged from 29,400 foot in 1297 to 12,000 in 1301.[81] Mr. Morris contends that not till 1322 were infantry drawn from all England for a Scottish campaign (as cited), but in this he is wrong. It was done by a special vote of Parliament, and according to a prescribed form, as early as March, 1316, when every township, with some special exceptions, furnished one soldier,[82] and again in 1318.[83] These are clearly new and special arrangements, and there is thus no reason to believe that the list in Foedera, etc., is not complete, as Mr. Oman suggests, adding, accordingly, a southern contingent of about 30,000 men, though he doubts if “the extreme South” sent its full muster.[84] This is quite gratuitous. Lord Hailes, too, contended that the official records are imperfect, and that the numbers given by Barbour “are within the limits of probability.”[85] Bain’s authoritative reply is that, “as a rule, the writs were always enrolled, and the Patent Rolls of the time are not defective.”[86] This, however, is not always true, and Bain, applying this principle absolutely, is once, at least, led to a wrong conclusion.[87]

An important question now suggests itself, but no one has so far raised it: did the levies in these full numbers turn up? They are allotted in round figures: what proportion was actually furnished? That there would be some trouble in securing the conscripts is anticipated and provided for in severe measures for the contumacious.[88] This was usual, and even the strong hand of Edward I. could not prevent men from deserting after they had received their wages.[89] Here we have, also, a sufficient basis for an estimate. On May 12, 1301, Edward I. summoned for midsummer 12,000 men from nine of the counties included in the Bannockburn levy—York, as in that case, being assessed at 4,000.[90] On July 12 we have the numbers from these counties as they appear on the pay-roll, when it is stated that they had contributed in proportions which give only 5,501 all told; York having sent only 1,193, and Northumberland, assessed at 2,700, providing the largest proportion—2,019.[91] The numbers vary slightly on other days, but seem never to have exceeded, if they reached, 50 per cent. of the nominal levy. Mr. Morris works out the same result for the Caerlaverock Campaign of 1300.[92] There are no grounds for assuming that things went differently in 1314, and thus over 21,540 men are reduced by about half. It is quite a fair conclusion that not more than 12,000 English foot—which exceeds the proportion above—were actually present at Bannockburn.

For the foreign contingents no figures exist. Bain thinks they were not “more than a few thousands.”[93] The Gascon corps in the Falkirk army should have been 106 mounted men.[94] The Hainault and Flanders auxiliaries who shared in the campaign of 1327 amounted to 550 men-at-arms, and were an expensive item.[95] The Irish contingent which came to Edward I. in 1304 amounted at most, for a few weeks only, to 3,500 men,[96] but to merely 361 in the army of 1300.[97]

I would suggest, therefore, for the English army the following round numbers: 3,000 to 4,000 horse of all sorts, 12,000 English and Welsh foot, 3,000 (?) Irish, 1,500 (?) foreigners, or, in a lump sum, 20,000 men of all arms, to which must be added a crowd of non-combatants—servants, traders, and camp-followers generally. Bain (as cited) proposes 50,000; Round, 30,000; Oman, 60,000 to 70,000. I consider 18,000 to 20,000 the most probable range. With even the lower of these numbers, the English commanders in organization and commissariat would have rather more than they could manage.

Barbour’s figure for the Scottish army must be similarly reduced. More than 30,000 would be a huge proportion of the Scottish population of that time, especially as the whole does not seem to have been drawn upon, and of that, as Barbour insists, a good many were still hostile.[98] William the Lion was credited in 1173 with a national host of 1,000 armoured horsemen, and 30,000 unarmoured footmen,[99] and the latter unit is surely over the score. At Halidon Hill, 1333, the Scots are said to have had 1,174 knights and men-at-arms and 13,500 light-armed men or foot;[100] and this chronicler consistently exaggerates. Yet these figures represent a united kingdom. Forty thousand at Bannockburn is the estimate for the Scots of the Vita Edw. writer, but the English writers, on their side, grossly overstate the numbers of the enemy, as witness what is said of Hemingburgh above. Bain’s figure of 15,000 to 16,000 is no doubt nearer the mark; “perhaps twenty-five thousand men in all” is Mr. Oman’s conjecture.[101] Possibly 6,000 to 7,000 is as near as we can go, adopting Barbour’s ratio, which gives a proportion of 1 to 3 of the English army. The non-combatants here, too, would be numerous. Up to this time Bruce’s men in the field could be numbered only in hundreds, so that as many thousands would represent a very special effort. And note that after Murray’s success over Clifford nearly the whole Scots army gathered round him to see him and do him honour—a fact which is suggestive[102] as to its size.

APPENDIX D
THE THROWING OF THE HEART

Book XX. *421-*432

These lines are found only in Hart’s printed edition. Pinkerton thought there was “no reason to view them as an interpolation,” and Jamieson regarded their agreement with the account in the Howlat[103] “a strong presumption of authenticity.” By Skeat they were at first accepted as genuine, but afterwards, influenced by the reasoning on Barbour’s rhymes of P. Buss in Anglia,[104] he surrendered them as an interpolation. In the passage of twelve lines three rhymes occur, which are unusual—more strongly, impossible—for Barbour on the basis of his admitted work. These are battell—tell, to be—de, ho—to. In the first case, Barbour, it is claimed, elsewhere always uses the “liquid” form bataill (battalyhe) to rhyme with another word of the same character as assaile or travaill (travailyhe).[105] In the second, he “never rhymes be with de (correctly dey),” as Skeat puts it, for de (Icel. deyja) was still influenced by the terminal semi-guttural, giving it an “impure” sound, whereas “be,” with no ghostly after-sound, is quite “pure.” The final example brings together two different values of “o,” and, it may be added, in the four cases in which Barbour uses the word, it is in the form hoyne.[106] These rhyme-tests had also been applied to the same result by Mr. W. A. Craigie.[107]

With this conclusion Mr. Brown agrees, “although on slightly different grounds.”[108] Hart’s edition, of course, takes a place in his general scheme of redaction. But he would “hesitate to reject the lines on the rimes alone,” and “The be, de test” seems to him “quite untrustworthy.”[109] Skeat thinks it unanswerable.[110] Mr. Neilson pleads “that this canon begs the whole question of the text of the Bruce ... first you find your canon; then you edit out of your text all that is disconform.”[111] Arguing specially on its application to The Legends of the Saints, he points out that “There are not a few metrical and other solecisms in the Bruce,” and that the “exceptional e-rhyme” is the stamp of transition.[112] It is to be observed also that Chaucer, Barbour’s contemporary, and more careful in such matters than he, rhymes ho, y-do in the Knight’s Tale.[113] In the Alexander occurs the tell—battell rhyme.[114] On the whole, the test is perhaps not so conclusive—out of Germany—as Skeat imagines. Further, from the indubitable reference in the Howlat to the Bruce, Neilson accepts the latter as the sole source of its digression, and the lines as therefore authentic.[115]

If, however, what has already been said of the passages from Hart hold good,[116] then this one must go with the rest. Fortunately, in this specific case that argument can be greatly strengthened, for the lines have never been tried by their relation to the context and their historic implications, and that obvious and indisputable test puts the question beyond doubt. They have but an outside connection with the narrative of Barbour, and otherwise are in flat contradiction thereto. So much is at once evident from the closing couplet:

It is a series of performances of this kind that is contemplated, not a single example, which is all that Barbour’s account gives room for. Douglas is credited with a habit of this sort, “ever in field”; while Barbour, like Froissart, knows of only one battle in which Douglas fought while bearing the heart of Bruce.[117] Nor is Barbour likely to have omitted such a “point of chivalry” on the part of his twin hero, had a valid tradition of it existed in his day.

The problem becomes clearer when we consider alternative and later accounts of the expedition of Douglas, for which see note on Book XX. 191, 192. Evidently the idea of his going to the Holy Land, as Froissart explains the commission,[118] and as it occurs in Bower, gave an opening for embellishment, which expands in the hands of Boece to the extent of thirteen victories achieved by Douglas over the Turks! This, however, is only to give more precision to a composite account contained in the Buke of the Howlat of the middle of the fifteenth century, a poem written in glorification of the Douglases. The author, supposed to be Richard Holland, speaks of the great friendship Bruce had for Douglas: “Reid the writ of thar work to your witness”[119]—a clear reference to the Bruce, especially as in xxxv. and xxxvi. he paraphrases the reply of Douglas to the King in Book XX. 223, 234. Thereafter, however, he strikes off from Barbour. Douglas goes to “the haly graif,” where—

The story then proceeds:

An extension of the original commission, be it noted, and a motive for what follows:

So we go back to Barbour (“the writ”), but in the final scene there is no mention of throwing the heart, any more than in the genuine Bruce, though it is stated that “His hardy men tuk the hart syne upon hand.”[122]

Obviously we have in these stanzas, and especially in the words underlined, the source of the lines in the Bruce, which are further in express contradiction to Barbour’s narrative, and have no place in it. The threefold argument leads inevitably to the one conclusion that these lines are an interpolation, and, as a corollary, that their source is the Howlat. Mr. Amours, in editing that poem,[123] has gone so far as to say that this is “almost certain.” I would remove the qualification.[124]

APPENDIX E
THE “ALEXANDER” AND THE “BRUCE”

The Buik of the Most Noble and Vailyeand Conquerour, Alexander the Great is an anonymous Scots translation of three French romances in the Alexander cycle, dated, in a rhyming colophon, 1438, and published for the Bannatyne Club in 1831. Between this translation and the Bruce there is a remarkably intimate and undisguised connection, not only in spirit and method, but in “the diction as a whole, the choice of words and the arrangement of the sentences, (and) the abundant use of alliteration,” to such an extent that “in reading the Buik of Alexander one would often think that he discerned the singer of the Bruce.”[125] A few examples have been given in the notes, but for a full survey of this literary phenomenon the reader must go to the dissertation quoted from above, or to Mr. J. T. T. Brown’s The Wallace and the Bruce Restudied, pp. 100-112 (Bonn, 1900), or Mr. Neilson’s John Barbour, Poet and Translator (London, 1900), which is devoted to the subject; or, for the parallels in the Bannockburn account, to Mr. Neilson’s article on Barbour in Chambers’s CyclopÆdia of English Literature, vol. i.

On the facts there is no dispute; for explanation three hypotheses have been put forward. Hermann, accepting the 1438 date, concludes that the translator of the Alexander was so familiar with the language of the Bruce—“here and there, indeed, knew it by heart” (stellenweise es wohl auswendig wusste)—that his translation was necessarily strongly influenced thereby.[126] This is inadmissible; the French poems are earlier than the Bruce, and to these the links of connection ultimately go back. The relationship is really deeper than the mere language of the translation, as Hermann himself indicates. Mr. Neilson, accordingly, in a detailed and forcible argument, claims Barbour himself as the translator of the Alexander, arguing that, the literary proofs being so conclusive, the date given must be an error, “scribal or printer’s.”[127] Given Roman numerals to begin with, such a slip is not in the least unlikely; variations of this sort occur in the Bruce itself,[128] and 1438 may have been a misreading of 1338, or the date may be that of the scribe’s copy, not of the actual work. Mr. Neilson has an ingenious section on the wayward fortunes of dates.[129] Thus, reversing Hermann’s thesis, he holds that “Barbour’s mind and memory had been steeped in the Alexander when he wrote the Bruce.”[130] Mr. Neilson’s argument and conclusion are vigorously contested by Mr. Brown in a Postscript to the work cited. His more elaborate hypothesis is that David Rate translated the Alexander in 1437, and that “John Ramsay, Sir John the Ross, wishful to improve the plain song of John Barbour, used the translation of the Alexander extensively, taking freely whatever he required.”[131] Mr. Brown’s negative criticism is independent of this proposition which is involved in his wider theory regarding the construction of the Bruce. The eclectic conclusion of the writer in the Cambridge History of English Literature, vol. ii., is: “Either the book (i.e., the Alexander) is the work of Barbour preserved in a somewhat later form, or the author was saturated with Barbour’s diction, so that he continually repeats his phrases.”[132]

In the dust of the conflict a crucial fact has gone unobserved—namely, that one of the parallel lines enumerated by Brown and Hermann appears in the portion of the Bruce incorporated in his own work by Wyntoun.[133] Here, then, we have a line of the alleged translation of 1438 occurring in the “Bruce” as it existed before 1420. Thus the only outstanding difficulty of Mr. Neilson’s proposition disappears. The effect on the rival propositions is obvious.

APPENDIX F
MR. BROWN’S “SOURCES” FOR THE “BRUCE.”

In pursuance of his “hypothesis of fifteenth-century redaction” of the Bruce, Mr. Brown applies what he claims to be “fair and ordinary tests” to six “selected examples,” in order to show that his hypothesis “has a basis in demonstrable fact.”[134] I shall notice such of these very briefly, premising that I do not consider Mr. Brown’s use of his tests either “fair” or “ordinary.” So much, I think, will appear.

1. The Trojan War, Alexander the Great, Julius CÆsar, and King Arthur.

(a) The only thing urged against the Trojan War passage[135] is that it is in the suspicious company of the others, and these, Mr. Brown suggests, are derived from Chaucer’s Monk’s Tale, from which he produces a selection of lines to parallel those in the Bruce. It may be urged,[136] at the outset, that two contemporary poets dealing with the same set of historical events are very likely to display similarities. As Chaucer himself begins by saying—

But even in Mr. Brown’s “selected” lines it is the divergencies rather than the similarities that stand out. Chaucer says nothing of “Babilony’s tour”; “his awyne hows” is no parallel to “of thyne owne folk”; and “Bot, ar he deit, his land delt he” has no equivalent whatsoever in the Monk’s Tale. Yet Chaucer has forty lines to Barbour’s eight. In fact, the “example” is a stock one, even to its phraseology, as witness these excerpts from sources half a century and more earlier than either poet: Commendatio Lamentabilis on Edward I. in 1307, “magnus Alexander ... Nam ille annos regnans duodecim veneno hausto vita defungitur (15); Vita Edwardi Sec. (c. 1326). Sed ille magnus GrÆcorum imperator Alexander, totius orbis domitor, cum cunctas nationes orbis subicit, per familiares proditores toxicatus occubuit.” Do such close parallels prove that either Barbour or Chaucer borrowed from chronicles which they never saw?

(b) Mr. Brown argues that the Julius CÆsar parallels are “not less remarkable for significant agreement, as regards the sequence of the narrative,”[137] and that “so far as concerns the diction it (the Bruce passage) approaches even nearer the Chaucerian original.”[138] But where Chaucer speaks of CÆsar conquering “thoccident” and “the orient,” Barbour enumerates the countries. Is this similarity of diction? According to Mr. Brown, he is giving “simply an expansion of the Chaucerian phrases.”[139] Elsewhere, in such a case, Barbour is convicted of “summarising” or “assimilating,” here of “expanding.” This is Mr. Brown’s “fork” from which no author could escape. In twelve lines Barbour comments on CÆsar’s conquests, his imperial position, and his death, and in forty-eight Chaucer gives a detailed biography introducing Pompey, of whom Barbour says nothing, and Barbour is thereupon charged with following “the sequence of the narrative”—as if he could avoid doing so! That CÆsar by both poets is styled “Emperor” goes for nothing; that was the medieval way; as also was the statement that he was killed in “the Capitol,” as Shakespeare, too, believed. But the most striking note of difference remains. Barbour says of CÆsar—

Now Chaucer remarks: “Of honestee yet had he remembrance,”[141] and Mr. Brown enrols the word “honestee” among the things “not to be explained either as commonplaces or as mere coincidences.”[142] We see Barbour’s idea of his “honeste”; this is Chaucer’s:

Mr. Brown here seems to have followed CÆsar’s example and “closit hys eyn”!

(c) The only point made with regard to the Arthur lines[144] is the calling of Lucius “Emperor,” and regarding this see my note on the passage. Geoffrey of Monmouth does the same. That Wyntoun corrects Huchown, and not Barbour, in this usage—well, Mr. Brown can make all he can of that. Barbour’s dozen lines on the familiar Arthur story is charged with being “an excellent summary of the Morte Arthure,”[145] a poem of 4,364 lines! No “expansion” here!

2. The Alexander allusions in Bks. III., X. With reference to these see Appendix E. But why should Mr. Brown speak of “the famous grey palfrey of Lord Douglas” on the strength of one notice in Bk. II. 118? There is nothing to justify the epithet “famous”; and Ferrand was no more an unusual name for a “grey” horse than Blanchard for a white one. Bishop Lamberton might have had “a grey” as well as Eumynedus.[146]

3. The Ferumbras Romance. I have analyzed this passage in my notes to Bk. III., 435-462. Mr. Brown contends that the adoption of the form Ferumbras for the French Fierabras “points to a knowledge on the part of the Scottish poet” of either the Syr Ferumbras or the Sowdone of Babylone, in both of which this form occurs. It is possibly an adaptation of the spelling Fierenbras, which occurs in The Destruction of Rome and also in a fragment of the romance.[147] But there is nothing more remarkable in Barbour’s intrusion of the “m” here than in his calling Lubaut or Lybaut, Lumbard in Bk. X. 324,[148] or Capaneus, Campaneus.[149] More significant is the name Lavyne for Balan, which is the normal form for all the existing texts, French and English, save the Sowdone, where we have Laban, Lavane and Lavyne, and The Destruction of Rome (French), which has Balan and Laban, the latter of which has given the spelling in “v.” Mr. Brown, however, rules out the Destruction for Barbour, as being “merely related to the Charlemagne cycle,”[150] though Dr. Hausknecht accepts it as the original of the first part of the Sowdone.[151] Why, then, should Barbour, too, not have known it? Mr. Brown’s conclusion is that the name, with other material, points to the whole passage being based on the Sowdone, and the Sowdone being post-Chaucerian—in which opinion he follows, and even goes beyond, Hausknecht—of the beginning of the fifteenth century, after Barbour’s death, it obviously follows that the lines cannot have been his work.[152] Against this may be put Hausknecht’s own conclusion, to which Mr. Brown makes no reference: “It is worthy of notice that the account of the Fierabras romance, as given by Barbour, may be considered, on the whole, as identical with the subject of the French Fierabras or the English Syr Ferumbras, but not with the Sowdan, as there is no mention made of the combat before Rome, nor any trace of what makes up the first part of the Sowdan.”[153] A few additional facts will substantiate Hausknecht’s statement.

In the Sowdone the twelve peers are shut up in “Egrymor”; Barbour gives “bot eleven.”[154] Mr. Brown says of the Bruce account that, “Every line is traceable in the Sowdone.”[155] There is no trace of line 452, or of Lavyne’s “flot” in 456; Charlemagne, instead of “being joyful” at the news that his peers were alive, there bursts out in anger against the traitor Ganelon.[156] Conclusive, on Mr. Brown’s own case, is the fact, noted by Hausknecht, that the relics mentioned by Barbour “differ from all other versions.”[157] It is not “off the croice a gret party”[158] in the Sowdone, but simply “the crosse, the crown, the nailes bente”;[159] and there is no mention of “the sper.”[160] Nothing is said of the cross in the Fierabras or the Syr Ferumbras, and the Destruction has the crown of thorns, the cross, the nails, and the “signe”[161] or shroud. Nor did Barbour invent “the sper,” for it is spoken of in this connection, though not in the final distribution, in the ProvenÇal version of the Fierabras: “e del fer d’una lansa,” the spear-head.[162] It would seem, then, that Barbour was using a version of the romance different, in certain particulars, from any we now possess. At any rate, Mr. Brown has to get round these awkward corners before he can transfix Barbour on Lavyne.

4. The Tydeus Episode, Bk. VI. 179-268. It may be granted that Barbour here is not, as Skeat supposed, drawing directly on the Thebaid of Statius, even in memory. But then no more is Lydgate in his Story of Thebes;[163] having used, according to Dr. Koeppel, some French prose epic based on the classical story,[164] now lost, or giving in his poem “a transcript from a French rendering of Statius.”[165] There was ample material, including even, as Gaston Paris has suggested, an abridgment of Statius in Latin prose. Mr. Brown contends that Barbour’s “redactor” borrowed from Lydgate. But here, again, the differences are sufficiently vital to thwart such a conclusion. “Betwix ane hye crag and the se”[166] has no equivalent either in Statius or Lydgate, as is clear from Mr. Brown’s own citations;[167] nor has the detail that the “gret stane ... throu the gret anciente, Wes lowsyt reddy for to fall,”[168] for which Lydgate has:

In an inductive literary argument differences must be accounted for as well as similarities, and any other possible sources must be satisfactorily eliminated. Opinions as to “the classical parsimony of independent translators”[170] are purely speculative.

5. The Hannibal lines, Bk. III. 207-248.—Here Mr. Brown’s argument needs only to be summarily stated to show its inadequacy. Wyntoun borrowed from Barbour in some instances, but not in this; therefore, since there are similarities in the way the same story is told, the Bruce lines are drawn from the Cronykil.[171] Wyntoun confesses to having translated from Orosius, Mr. Brown says through Martinus Polonus, and he follows his author closely. Barbour’s narrative is brief and, in several details, faulty, in which he diverges from both the Latin author and his Scots translator.[172] The startling error in chronology, especially, “would not be readily made by one using Martinus at first hand,” but with the Cronikyl before him the writer was quite likely to do so![173] The peculiar “touches” in the Bruce passage are noted, but unexplained. It is assumed that “The agreement with Martinus is much too close for it to be considered to be written from memory;”[174] and “there is nothing in Martinus that should lead us to expect independent translators to ascribe such a victory to God and in terms so similar:

But the terms are not “similar,” and both are a mediÆval rendering of divina miseratio in Orosius and Polonus. The whole argument in this connection is involved and haphazard. What is there illegitimate in the simple hypothesis that Barbour wrote from a well-stored but not perfectly accurate memory, and that Wyntoun preferred the first-hand to the second-hand source? So we explain both similarities and differences. Wyntoun’s lines are a plain following of his Latin author; Barbour’s a brief summary, with expansions and comments of his own leading up to the moral he wishes to enforce; and in this case Mr. Brown cannot point to a single line in common.

6. Froissart and the Bruce.—Here Mr. Brown prints his parallels in full, and the matter can be safely left to the judgment of the unprejudiced reader. It need only be remarked that Froissart (1) calls Douglas William instead of James; (2) sends the heart to the Holy Sepulchre[177]; (3) makes Bruce choose Douglas; (4) embarks him at Montrose; (5) takes him to Sluys; (6) lands him at Seville: while Barbour (1) is right as to Douglas; (2) sends the heart against “Goddis fayis”; (3) makes the barons choose Douglas; (4) embarks him at Berwick; (5) says nothing of Sluys; and (6) lands him at Valentia. Of Barbour’s “motive” lines, of the repentance for blood-spilling and the saving of his sins,[178] Froissart (or Le Bel) has no trace. The two narratives are independent, though, in their trend, alike.

APPENDIX G
LANGUAGE AND ORTHOGRAPHY

The language of the Bruce is Northern English, the dialect spoken north of the Humber. Barbour himself calls it “Inglis” (Bk. IV. 253), and Scottish writers down to the sixteenth century do the same. The name “Scots” is therefore a term of pure convenience, signifying the English spoken within the political borders of Scotland, which continued to be an independent literary medium after the Northern English of England had ceased to be such, and had yielded place to the standard dialect of Chaucer and his successors. But the language of the Aberdeen Barbour is substantially that of the Yorkshire Richard Rolle.

The most obvious characteristic of the northern tongue is that in development it was far ahead of its southern contemporary, in so far as it had absolutely got rid of inflections, not even retaining the final e which casts its shadow over Chaucer. Where the final e occurs in the Bruce it is non-significant and unsounded. Like the variable spelling, it is the result of the writing of English by scribes accustomed to the sounds and orthography of contemporary French, which was rich in final e’s[179]—a process which had seriously complicated the straightforward phonetic spelling of Old English. A peculiar Scots fashion, however, was the representation of the long vowels by a combination with y or i, sometimes in addition to a final e, which came to be regarded as the sign of a long vowel. Thus we have such groups as mar, mair, mayr: done, doyn, doyne. In these and similar cases y is simply an alternative form of i; at first it would be written before letters like m, n, u, where i, having no dot, might be mistaken for a part of the succeeding letter; then in practice the two would become interchangeable according to the caprice of the writer. In this matter the scribe of C is rather more regular than that of E. Now, such a diphthong as ai was in time reduced to a simple sound, when the i became superfluous, and came to be looked upon as a sign of length. With this function it was afterwards, at the pleasure of the writer, transferred to the company of other long vowels where there was no original diphthong. Thus we have the long vowel sound represented in three ways—simple, with following i (y), or with terminal e. In cases like tais, gais, however, the i is part of the termination of the present tense of the verb, of which Barbour is particularly fond.[180]

But while Scots—in the sense indicated above—had thus early parted with its inflections, it was, on the other hand, more conservative than the southern dialects in its treatment of the vowel sounds. It retained, for example, the Old English a, which in Southern English was rounded into broad o (oa)—cf. ga, stane (O.E. or A.S. gan, stan) for go, stone. In the Bruce (X. 199; XII. 299) the rhyme more, before may show the rounding influence of r.[181] Vowels in words of French origin show slight modification. Anglo-French nasal a before m, n tends to become au, whence daunger, etc., and o or u to become ou as in baroun, felloun. Ai and ei incline to merge in long e as feble (O.F. faible), and sesyt (O.F. seisir); but ai may also become a—e.g., tratour for O.F. traitor.

AphÆresis occurs in such shortened forms as stroy (destroy), semble (assemble), etc.

The following peculiarities may be noticed among the consonants:

ch sometimes = “tch” in middle and end of word—e.g., fech = fetch (II. 532).

h is silent in hoost, as in all Anglo-French words of Latin origin, and sometimes drops out—e.g., ost (II. 559, etc., O.F. ost), also in ayris (heirs, V. 520).

b in combination with m is dropped both in pronunciation and spelling. Cf. chamyr (II. 97), tumlit (IV. 182, etc.). C Hummyr for Humber; E Humbre (XV. 538).

d usually dropped after n—e.g., henmast = hindmost (XII. 268, etc.); sometimes intrusive—e.g., suddandly (VI. 11), but suddanly (VI. 220). In avantage, aventure from French the English d is intrusive.

f represents final ve in pruf or preif, leiff, driff, knaf, haf, etc., retaining the voiced sound (v) which it had in A.S. between vowels, and r l and vowels.

g in words of French origin has before e, i, the English sound = da(j), which was also then the French one—e.g., liege, jugis.

l as indicating merely a long vowel is unsounded in walknyt (VI. 296), walk (wake, VII. 179, etc.). So we have also falt (VI. 345, etc.) and fawt (III. 298); realte and reawte, etc. Skeat holds that the l in the first case represents but a way of writing the first k, so that the words should be wakk, etc., but this does not cover the second case.[182]

m and n are often interchanged, as in confort, manteme, etc.

ng becomes n before th as lenth, strenth, etc. Cf. also Grammar. The nasal ng may take the place of the liquid n, as in ryng (I. 78, etc.) for “reign” (Fr. regner). The French nasal nce is represented by ng or nch—e.g., Vallang (II. 201) or Vallanch (V. 472) for “Valence.”

r. Metathesis of r is very characteristic of the Northern dialect—e.g., fryst (first), brist (burst, A.S. berstan), tursit (O.F. trusser), etc.

s is changed to sch in schir, isch (issue), etc.; but on the other side we have sall, suld (shall, should) and Inglis.

z has in Fiz the Anglo-French sound ts, as in rats.

APPENDIX H
GRAMMAR

Nouns

Singular—Possessive Case.—1. The inflected genitive or possessive is formed by adding ys or is; MS. E, on the whole, shows a preference for the former. Sometimes the final consonant is doubled. Cf. a manis hand (VII. 580), to mannys fude (X. 189).

2. Nouns denoting relationship are uninflected, as in Anglo-Saxon: Mordreyt his sistir son (I. 557). Till his brothiris pes (XI. 652) is an exception found in both MSS. For And sperit of his brotheris fair (XVI. 21) in C, E has brodyr.

Note.—Wode-syde (E woddis sid, IX. 139), sonne-rising, hous end (VII. 163), all common combinations. Editors, however, usually read the termination of the last as housis, though it seems to fall under the same category.

Plural.—1. The plural also is in ys or is, more seldom es, representing A.S. as. It “formed a distinct syllable in monosyllables and words accented on the final syllable.”[183] The latter part of the statement is not, however, uniformly true. Cf. battalis (XI. 122), bischoppes and prelatis (XX. 294), etc., which are dissyllables.

2. Simple s occurs only after r and y (i)—galays, werriours, etc.—but spurys, baneris, towris also occur. Feys (I. 58) and fays (I. 223) are to be read as two syllables.

3. Final f (ff) is generally changed to v, as theif, thevis; but wif preserves the f throughout, as wifis, wiffys, etc. (E, however, in XII. 246, has wyvis); and lif has lyvis (XI. 590), or livis and lyffis (X. 106), or liffis (IV. 137).

4. Words ending in er, ir, yr, yn, ill for the most part reject the final vowel before the ending of the plural: dochtrys, lettres or lettrys, wapnys, etc. Where the vowel is retained, it is silent, as in schulderis (IX. 356), letteris (XVII. 31, 39), etc.

Note.—Burges in XVI. 80, XVII. 236, is clearly unaltered in the plural, and this would seem to apply also to marras (VIII. 35), hous (VIII. 514, etc.), and mos (VIII. 173), which are usually read as three and two syllables respectively. Cf. on Singular (2).

5. Of strong plurals, besides men and its compounds, ky and brethir by change of vowel (Umlaut), we have eyn(e) (A.S. eagan), with singular e or ey (VI. 523), oxin or oxyne (X. 381, 388), schoyne (II. 510). These (with hosen and fan (faes)) are the only plurals in n preserved in Northern English.

6. Some old neuters (A.S.) take no inflection in the plural: deir, hors, etc.; hous has hous and housis (X. 60, etc.), but cf. on (4) (note); thing has thing (XI. 27, etc.) and thingis. Cf. also gudis and gude (XVII. 517, 521). One form is in er or ir, childer (XII. 246) or childir (XVII. 515), A.S. cildru.

7. Nouns indicating time, space, quantity, weight, and number are unchanged in the plural: fifty yheir (in I. 522 occurs ten yheris, quite an exceptional case), tuelf moneth, six and fourty wyntir (A.S. plural also winter), twa myle, tuenty thousand pund, etc. For paris in C (XIII. 463), E reads payr. In XII. 234 C has thre gret avantage, where E gives avantagis, but reads vasselagis (!) to rhyme.

8. Only men suffers inflection in the plural possessive: the Inglis menis fewte (VIII. 19), of othir mennis landis (XI. 148). In till Scottis men possessioune (XVII. 202) we may have, as Henschel suggests,[184] a piece of “scribal carelessness;” but Hampole undoubtedly uses this as a valid form:

Proper Nouns.

In general, these follow the common nouns in their forms, but note:

1. Two names ending in s have no inflection for the genitive: King Adrastus men (II. 529), Thomas prophecy (II. 86). This occurs also in Chaucer.

2. Douglas has both flectionless and inflected forms: the Douglas men (X. 398), the Lord Douglassis men (XX. 481).

3. Possessive or genitive formed by his: Hannyball his mekill mycht (III. 232). Cf. also VI. 435, 446. Originally, as here, confined to proper names. A special example is in the reading of C (XVII. 940): Berwyk his toune; E Berwykis.

Adjectives.

1. There is no distinction in the Bruce between strong and weak adjectives, the latter in Chaucer being indicated by a final e, and used after the definite article, demonstrative and possessive pronouns, etc., nor is the plural similarly distinguished from the singular. In all these cases the Scots adjective is invariable; the final e, which often occurs, is of no grammatical significance. There is no difference between the gud Erll (X. 686) and the gude Erll (XX. 603), while the plural is gud werriours (XX. 416).

2. The only example of inflection is the form aller, genitive plural (er = -ra, A.S. eal-ra). It occurs once: throuch thar aller hale assent (I. 137).

3. The comparative and superlative are formed by adding ar or er (seldom ir) and est or ast. In words of two and more syllables mar (mair) and mast (maist) also occur. The same rule applies to adverbs. The final consonant of a monosyllable may be doubled: thikkest (VIII. 81), but also thikast (XVII. 156).

4. Change of vowel (Umlaut) appears in two cases—sterkar (E starkar) (XV. 491) and eldest (I. 51, IV. 71). Ill or evill is the Northern form for bad, compared with wer or war, werst (III. 192). Mor, for mar, occurs twice to rhyme with befor (X. 199, XII. 299). Er gives erar.

5. After the comparative comes usually than, which is always used by Rolle, but there are four examples of na—wes starkar fer na he (VI. 538), also III. 229, II. 519, X. 637.

Numerals.

1. Ane before a consonant sometimes becomes a: in a nycht and a day (III. 429). The tane and the ta are compounds of this numeral, with the old neuter thaet (thet) of the definite article, really thet-ane, thet-a. The former is a substantive, the latter an adjective: the tane suld be kyng (VI. 186), the ta part (III. 239). Similarly the tothir = thet othir (I. 7, etc.), which is throughout used for the ordinal “the second,” as formast is sometimes used for “the first”: Fule-hardyment the formast is, And the tothir is cowardis (VI. 337, 338).

2. Note thresum, fiffsum, sex sum, where “sum denotes conjunction” (Jamieson’s Dictionary), three, etc., together; half deill (A.S. dÆl, a part); twa part = two parts; yneuch, singular (usually adverb), used for quantity, and ynew (enew), plural, for number. Cf. XX. 337, and I. 558.

N.B.—For demonstrative, etc., adjectives, see also Pronouns, Demonstrative, etc. Note sam(m)yn (same) as adjective and adverb (V. 72, XII. 49, etc.). Distinguish ilk(e) = same (A.S. ylc) from the distributive ilk (A.S. Æle) = each. Others are sik (syk), swilc, etc.

Pronouns.

Personal.—1. In MS. E besides I is found the strong form Ik or Ic (A.S. Ic). The latter is entirely wanting in C, which has everywhere I.

2. The third person feminine is scho (A.S. seo), once spelled sche (XIII. 635), a Southern effect, where, however, E has scho.

3. Note the gen. sing. mas. and gen. plur. in magre his, in spite of him, and magre thairis, in spite of them.

Possessive.—1. The genitive cases of the personal pronouns are most often adjectives. Before words beginning with a vowel they are mine, thine; before words beginning with a consonant they become my, thi; words beginning with h may have either form.

2. The forms used absolutely are myn, his, ours, yhouris, thaires: That suld be myn (V. 223), etc.

Reflexive.—1. The personal pronouns serve as reflexives: I will me spied (XI. 638); And went hyr hame (V. 177).

2. There is also the form in self, more seldom selvyn, selvin. When the subject is a personal pronoun this may drop out: that himself suld wele Kepe the entre (XI. 445).

3. Self is both singular and plural: That thai mycht help thameself (X. 619).

Demonstrative.—1. The Northern plural of this is thir, which supplanted thas as thise did in the Midland. Its origin is obscure.[185]

2. The plural of that is tha(i). Distinguish from thai (they), 3rd pers. plur.

3. Yhon(e), more often an adjective, has also a substantive use in three cases only (IV. 502, 506; XIV. 280).

Relative and Interrogative.—1. The relative pronoun is that or at, never quha. At is a purely Northern form. Quhais and quham occur occasionally as oblique cases. In quhom, as in words like so, mor, etc., there is evidence of Southern influence. Quha is used as an interrogative. The simple quhilk occurs but once (XVIII. 225, see note), where E has a different reading; otherwise it is in the form the quilk that.

2. The relative in the oblique cases is often omitted: the small folk thai had thar (IX. 261). This is a marked feature of the Wallace.

Indefinite.—1. None for nane occurs only once (IX. 485); and so, too, ilkone for ilkane, rhyming with a proper name (XI. 303). Nane is used both as substantive and as adjective. As adjective it comes before a word beginning with a vowel. Before consonants and h, na or no is used. In two cases no precedes a vowel: XIII. 145; XVI. 249.

2. Othir, often = “each other,” as Thai dang on othir (X. 680).

3. The plural of man has a pronominal use = one, German “man”: as men in the Bibill seys (I. 466); men mycht se (X. 678), etc.

Verbs.

1. Note the periphrastic form with gan in E (can in a few cases), can always in C: all the land gan occupy (I. 184); all can thai cry (XII. 200). The past tense of can is couth: thar vittale all fast couth fale (VIII. 460): whence, by analogy, the false form begouth for past of gan, itself the past of ginnen, to begin.

2. Weak verbs form their past tense and past participle in it or yt, the latter being more common in E. In certain cases the vowel drops out, as in dwelt. After r, l (ll), n, the ending is often in d, as herd, ansuerd, etc.

Note.—Felt for the ordinary feld occurs once, in III. 119. Cald, too, for callit, is forced to a rhyme (XIII. 61). The proper past of send is send.

3. The present tense indicative is, I spek, thou spekis, he spekis, we spek, etc. But when the personal pronoun is separated from the verb, or when the subject is a noun or relative pronoun or other form, the ending is in is or ys throughout. Cf. yhow that takis (XVI. 592); yhe that this redis (VI. 269); all men fleis the did (IX. 90), etc.

4. The imperative plural is in is (ys): departis us (VI. 543). It is more rare without the ending (XI 309; XII. 227, etc). This, however, is the rule when the imperative is followed by its pronoun: luk yhe (XII. 217).

5. The present participle ends in and. Doubling of the consonant may occur after a short vowel—e.g., wonnand (X. 160).

6. The verbal noun (so-called) ends in yng or ing. In many examples the g is dropped, as the tendency was and is in Scots, and we have a form in yn(e). Cf. fechting (XII. 119) and fechtyn (III. 241). And these may rhyme, which shows that in sound the difference was not apparent (cf. IX. 120). Sometimes the particular form depends on the MS.—e.g., VI. 520, C supposing: E supposyn. The same thing is found in proper names with a similar termination.

7. The past participle of strong verbs has ending yn or in, seldom en. Clymen in X. 648, where the correct form, clummyn, is given in 606, is either really a surviving or intruded infinitive, or has been assimilated to it.

N.B.—VII. 524, C cum: E cummyn; X. 506, C won: E wonnyn.

8. The infinitive has dropped all trace of an ending. The e which sometimes appears is silent. It is formed by prefixing to or till, in certain cases with for in addition. For one case we have a reduced infinitive with at: a-do (X. 348).

Adverbs.

1. him allane is a strengthened form of him ane (VI. 272, 320), and is the more frequent of the two; equivalent to German “allein” and analogous in construction to the pronoun with self. In Early Scots alane is attached to the dative, in later and modern Scots to the possessive, his ane, etc.: the fuller form, from a confusion with lone, is now his lane, etc.

2. The surviving phrase, the morn (XIV. 478) occurs once for the older to-morn. Mr. Gregory Smith says that the former “begins to appear in M. Sc.” (Middle Scots, 1450-1600),[186] and MS. E indeed reads to-morn in the passage above.

3. Note the compounds with gat (a way), thusgat, swagat, etc.: in XIX. 253; C swagatis; E swagat: also howgatis, etc., preserving Genitive ending.

4. The forms on liff (XVIII. 154), on slepe (VII. 192), on stray (XIII. 195), etc., are peculiarly Northern forms for “alive,” “asleep,” etc. Cf. also on fer, on flot.

Prepositions.

1. Till = to, the former being a Northern fashion. C frequently extends to on-to, on-till, as in VI. 622, XVII. 29, etc.

2. Note the verbal preposition that I of tell, etc.

3. Be and by are common to both MSS., C more frequently the latter, probably a Southern influence. Strictly in Scots be alone denotes agency; by usually = beyond, as in by ordinar, etc.

Conjunctions.

1. The verb suppos is several times used as a conjunction.

2. the-quhethir = thohquhether = thohquether, “nevertheless,” unknown to the Southern dialects.[187]