the Eiga monogatari is that the compiler made a conscious attempt to cut out all the more personal aspects of Murasaki’s account, and much of this fragment consists of precisely that. It would also seem highly likely that the compiler of the ‘Nikkiuta’ appendix, probably Teika himself, had this larger diary in his possession. Extra evidence for this is supplied by a passage in Teika’s diary, the Meigetsuki , for Tenpuku 1 (1233).3.20. This entry records that a picture scroll consisting of paintings and poems illustrating the twelve months of the year ( tsukinami-e ), originally a gift to Teika’s daughter from Princess Shikishi, is now to be given to the reigning Empress. The poem for the fifth month was recorded as ‘Diary of Murasaki Shikibu: a dawn scene’. As no such poem appears in the diary that we have today, this must be one of the poems above, probably ‘Now I see my face floating there’.
DATE OF COMPOSITION
What if we accept the above hypothesis that the diary as we have it is incomplete? As the extra fragments identified above predate the present beginning of the diary, it follows that the beginning must be considered lost. The only reason that a large number of scholars continue to argue against this idea is the fact that the present beginning constitutes such a superb introductory passage, so fitting indeed that it is hard to believe that anything could have preceded it. In this case, the only way we can explain away the existence of the Kankō 5.5.5 passage is to posit either an Ur -diary, out of which the present one was presumably extracted and rewritten, or to posit another separate diary. But this is all entirely supposition, based on a certain ‘feeling’ about the inviolability of the beginning as we have it now. One sympathizes, but as yet no one has come up with a satisfactory argument for denying the theory that the beginning of the diary has been lost. It may, of course, be that much more has in fact been lost, but there is no way of knowing.
What of Section B? In contrast to Section A, which is usually known as the ‘record’ part, Section B has traditionally been given the name ‘letter’ part ( shōsokubumi ). Why this is so should be clearfrom the last passages of this section, where it is obvious that there is a specific addressee in mind. Obvious, that is, unless you consider that this may actually be a narrative technique, a fictional letter. There is also the ubiquitous presence of the polite auxiliary verb haberi , which has the effect of stressing the existence of a receiver. Indeed for some time Section B was considered to be a real intruder, a letter that had been interpolated into the diary by mistake. This is an attractive idea at first sight but cannot be accepted as a feasible explanation, for one main reason: the smoothness with which A runs into B. The transition from record to personal comment takes place gradually over a number of paragraphs. If we insist on seeing B as an interpolation, then we are driven to the somewhat unsatisfactory conclusion that a second hand deliberately rewrote the beginning of the letter so that the join would be invisible.
While the smoothness of the join is undeniable, there still remain a number of elements that suggest that Sections A and B are by no means an organic unit. In the first place, the contrast in content and style is striking, and the presence of the occasional personal comment and passage of self-analysis in Section A does nothing to dispel such an impression. In the second place, there is a marked imbalance in the frequency of the auxiliary verb haberi . In Section B haberi appears 136 times, acting in its normal role as an auxiliary expressing politeness to the person being addressed, quite in character for a letter being written to someone specific. It also appears, however, in Section A, where one would not normally expect it to be present at all. Admittedly, it only occurs here twenty-five times and then subject