« 上一頁繼續 »
I. THE MANUSCRIPT
A small quarto volume in the British Museum, Cotton MS. Nero Ax+4 (new numbering) contains, bound between two Latin manuscripts, the unique manuscript of the four poems generally attributed to the author of Sir Gawain and the Green Knight. Purity, which occupies folios 61a-86a, follows The Pearl, and precedes Patience and Gawain. Several crude pictures illustrate episodes in
Of the two which precede Purity, the first represents Noah and his family in the ark, and the second shows Daniel expounding the writing on the wall to Belshazzar and the queen.
The manuscript is written in a small, sharp handwriting, which varies considerably in size, of the late fourteenth century. It is in many places very difficult to read, owing partly to the paleness of the ink, which has often faded so much that passages are hardly legible, and partly to the fact that certain lines have been blotted on the pages opposite them. Often the words have been so fully printed on the opposite page that one can read them plainly with the aid of a mirror. Dr. Knott pointed out the existence and value of these 'offsets,' as he calls them, in the text of Gawain, where they furnish in some cases the only
* The best description of the MS. is that by Sir Frederick Madden in his edition of Sir Gawayne (London, 1839), pp. xlvii-1. For the history of the MS. see also Gollancz's preface to his edition of Patience (London, 1915).
2 Madden, p. 301, ‘reign of Richard II’; Ward, Catalogue of Romances in the British Museum 1. 387, ‘end of the fourteenth century'; Gollancz, preface to Patience, ‘end of fourteenth or early part of fifteenth century.'
evidence for the original reading. In Purity, in addition to the offsets of a few letters on folios 746 and 75b, unimportant because the text is here perfectly legible, almost all the initial words in the lines of fol. 64a, 11. 217-52, have been partly impressed on fol. 63b. Since these words on 64a are often extremely faint and hardly decipherable, the offset is, in this last case, of some slight value in establishing and confirming the readings of the text, for instance, the initial bot of l. 226; but here too the offset is for the most part even less distinct that the original words.?
The offsets are less important for the text than the additions and corrections to the manuscript by a second hand, which sometimes obscure the original reading. In his second edition Morris noted the fact that sorewe of 1. 778 and broßer of 1. 924 (see notes on both these lines) were written by a later hand over the original. But there are traces of what is probably the hand of this same corrector in a great many other words and passages. In some cases the original scribe's letters have been merely retraced; but in others the corrector's hand is more certainly betrayed by letters of a type that the original scribe never uses. The following letters most strikingly distinguish the corrector's hand from that of the scribe: the corrector's a is like a modern printed a, whereas the scribe's is formed by two converging upright strokes and a cross-stroke; the corrector's e is a curved e made with one stroke and usually very flat, whereas the scribe's is made sharply with two strokes; the corrector's d has at the top a marked curl to the right, which the scribe's lacks; the corrector uses a
1 Mod. Lang. Notes 30. 102-8.
Offsets are very frequent in The Pearl, where every large initial letter is visible, some very distinctly, on the opposite page. With a mirror considerable sections may be easily read off, e. g., on folios 47a and 48b (so at least in Osgood's photographs deposited in the Yale University Library), but they are of no value, as the original is also easily legible.
Greek s never used by the scribe. For the sake of convenience I give here a list of those words and passages in which the writing of the scribe has been tampered with. The letters which cannot be the scribe's are italicized, and it is fairly certain that the words or passages in which they occur were corrected by the same man; but in other cases, though evidence of retracing is plain, it is possible that the attempts to make the faded parts of the manuscript more legible were not all due to this corrector: 108 sw (elt); 245-52 the ends of all these lines, and possibly more, on fol. 64a have been partly retraced; 245 towched; 247 þe vengiaunce (prob. written over vengaunce); 248 make had never; 249 forprast al þat þryve schuld; 250 (m)ercyles (and) mawgre much scheued; 251 fylþe upon folde pat þe folk used; 252 wythouten any maysterz; 257 ?ffor (and possibly more at the top of fol. 64b); 322 boskez; 323 I schal waken; 324 alle þat; 431 ?(was)ted; 778 sorewe (see note); 918 (foo)schip; 922 (for)sake; 923 out of; 924 broßer; 928 wore, and probably a few other letters on fol. 73b; 1015 þer faur (see note), is (inserted above line); 1664 þat weldes; 1669 one (added to end of line, see note).
According to Dr. Knott (Mod. Lang. Notes 30. 108), the words in the second hand, which appear in Gaw. 43, 81, etc., are written in a dark brown ink. This can naturally not be seen in my rotographs, and I have no means of determining whether the same corrector is at work in both poems. An examination of the manuscript would probably settle this, and would also, if the same difference in ink appears in Purity, lead to a more precise delimitation of retraced passages than I am able to give.
· Four other instances of insertions above the line occur, on, 432; synne, 520; wont, 739; the el of Daniel, 1756; but the writing is in each case so small that it is impossible to tell whether or not it is the scribe's. Omissions were undoubtedly made in each case, and I think that, with the exception of synne, 520, all these insertions are correct, whether by the scribe or not. Even is of 1015, which is almost certainly inserted by the second hand, because of the peculiar s, seems indispensable.
The scribe's own handwriting, even where there is no question of revision, offers difficulties. In addition to frequent repetitions and omissions due to carelessness, it is very hard to distinguish some of his letters: there is usually no difference between u and n; a t whose crossstroke is careless frequently looks like a c; bo, because the two letters are combined, cannot be distinguished from lo; nor ha from la for the same reason, if the second stroke of the h is not distinct below the line.
In printing the text, peculiarities of the manuscript in the division of words and capitalization have been disregarded. I and j, u and v have been normalized, and 3, when it was written for 2, so printed. The ordinary abbreviations for and, with, þou, þat, n, -e, -er, -es, -us, -(u)r,3
* For the scribe's peculiarities in the division of words, see Osgood, The Pearl, p. x, n. 1.
? It is difficult to determine when the strokes through long letters are intended for abbreviations and when they are mere flourishes. I have followed Morris in considering a stroke starting from the first stroke of h and with a decided upward crook an abbreviation for e in bilooghe, 116; innoghe, 297, 669, 1303; loghe, 366; also wyrle, 475. But there is frequently a straight stroke from a long letter which is certainly a mere flourish, since in some cases, e. g., ho 1126, an e would be out of the question. Only one of these, kyth, 912, is noted by Morris and expanded to kythe, but the stroke here resembles that in the words below, and not at all the stroke with the crook which I have considered an abbreviation. The following words have this meaningless flourish: after b in be, 123, 173; brentest, 379; bryngez, 636; biseged, 1180; after h in kyth, 912; ho, 1126; heged, 1584; after l in leve, 401, 1114; whyl, 1493.
The curl above 0, which previous editors of poems of this manuscript expanded consistently ur, was apparently used by the scribe more generally to represent simply an r (cf. Cook, Mod. Phil. 6. 199 on the rhymes of The Pearl). I have regularly expanded the abbreviation r, since such a word as corte never has ur when it is actually written out (191, 1109, 1530, 1562, 1751) and yor (once expanded yor, 715) is elsewhere consistently written yõ (942, 618, 620, 801). The expansion to r, not ur, is further justified by the occurrence of the abbreviation in such words as for, 756,