the authorship of Genji monogatari

Archive of discussion on the PMJS mailing list (Oct 15-22, 1999)

Question raised by Royall Tyler

Discussants: Lawrence Marceau, Rein Raud, Janine Beichman, Anthony Bryant, Lewis Cook, Chris Drake, Stephen Forrest, Michael Watson, Robert Khan, Gaye Rowley

Copyright of each message belongs to its author. See general note on editing.

return to: PMJS archive index / PMJS index / members / translations



From Royall Tyler (Oct 15 1999)

The list being quiet, perhaps I might raise an issue I often have on my mind, on the strenth of powerful impressions I get in the course of my passage through the Genji text. As I listen to the text, I hear many voices.

I have not studied this matter. I know a few have always been convinced that Genji monogatari is not all by the same writer, but I also have the impression that orthodox opinion recognizes Murasaki Shikibu as the sole author. Moreover, scholars and the public are, naturally, strongly inclined to champion Murasaki Shikibu. I feel exactly the same way. (If a nonspecialist may sigh when acknowledging that the homeric epics as we have them are probably not, after all, by a single genius named Homer, how much more disappointing it is to think that MS may not have written all of Genji!) Still, in my daily work I cannot evade the question. I wonder what people out there feel, think, know, or think they know.

For example, while translating most of "Yugiri" I had the strong feeling that whoever wrote what I was translating was someone new. This feeling culminated in the passage in which Murasaki no Ue reflects on the fate of women. It seemed completely out of character for the narrator's treatment of Murasaki no Ue both before and after "Yugiri," and I also found it relatively poor quality. (Strangely, immediately after this passage I began to feel at home again with the writer I assume to be MS.)

The interior monologue passage (Yadorigi) that I mentioned in connection with "waga kokoro yo" is also startling. It's very good, but it's astonishingly unlike anything in the chapters (Shiigamoto, Agemaki) on which it constitutes a kind of commentary.

A little later in "Yadorigi," (Seidensticker p. 902, top) there's a sudden, lengthy, and very surprising intervention of the narrator on the subject of whether or not Nakanokimi really deserves pity over Niou's marriage to Rokunokimi. Where did THIS come from?

These are just examples that struck me particularly.

Then there are the chapters like Niou no Miya, Kobai, Takekawa. I don't see why MS should have written them.

My most absorbing impression (wholly naive, I insist, and subject to any kind of revision or cancellation) is that the author of the main chapters and the author(s) of the Uji chapters (counting them from "Niou no Miya) cannot be the same person. The main Uji chapters are extremely accomplished, but they are so different in manner, in focus, in rhetoric, in plot construction, and so on from the main chapters that I cannot imagine how MS (my assumed main chapters author) could have written them, and certainly not AFTER the main chapters. Actually, I imagine a group of several women getting together to write the Uji chapters, in order to extend the tale. Each wrote different bits (there would have been differences of rank between them, so that a relatively inferior draft from one would have been more acceptable than from another), but they were very talented, and they got together to discuss all the drafts and to adjust them, and perhaps one acted as an overall editor. This takes place purely in my imagination.

In my mind the Uji chapters are to the main chapters as the Kokinshu is to the Man'yoshu, though to explain that would take much space and time.

Well that will do. These impressions do not represent "my position" on anything. They are something to talk about. And of course I know that they are implausible if only because, surely, if MS did not write the whole tale, some tradition or document to that effect would survive. (And I know that the works of a single writer can be surprisingly different from one another.)

Royall Tyler


From Lawrence Marceau

Royall,

Thank you for your deeply-felt and highly thought-provoking impressions regarding the authorship of Genji, based on your intimate relationship with the text over several years. Here are a few comments from a non-specialist, for what their worth.

1. Text: Before we can address issues of authorship, we need to have a firm grip on what we mean when we refer to our "text". In the case of Genji, there is precious little to depend upon prior to Teika's editing of various manuscripts and production of the "aobyoshi-bon" group of texts. These are what the various modern editions are based on. Then there are the "Kawachi-bon" group of texts, which are slightly different. Finally there are "the rest", referred to as the "beppon" group, but actually there is quite a bit of variation among these texts. It may be that the "voices" you hear in some of your readings may sound different depending on the specific text or texts you are interacting with.

2. "Murasaki Shikibu": We know very little about her, aside what we can glean from her from such sources as Murasaki Shikibu nikki and Murasaki Shikibu shu. All I can say is that, even if we knew more about the author (or authors), we must always be careful not to confuse the author with a particular narrator.

3. Near-contemporary sources and Genji reception: In Sarashina nikki the text tells us that the narrator acquired a manuscript of Genji in "50-odd" fascicles. We also learn of the narrator's fantasies about identifying with Ukifune. Given this state of events, it seems clear that editions of Genji were in circulation a few decades after Murasaki's death that included the Uji chapters. This is not to say that these chapters were written or edited by others, or that they were in the order in which we encounter them today. The earliest extant "text" of the Genji is the Genji monogatari emaki, which dates about a century after completion of the work proper. The earliest G M Emaki examples that survive in the Tokugawa and Goto Museums include both the earlier chapters and the Uji chapters together. Of course, the emaki text calligraphy is by (at least) four individuals. However, the point here is again that the text has been transmitted since soon after its completion together with the Uji chapters. I'm not aware of a collection that houses a version that leaves these chapters out...

In short, I expect that it is possible to conduct a statistical analysis of specific terms that could be used as evidence supporting a separate author or authors of the Uji chapters. It's likely that such research in fact exists.

However, as a literary historian dealing with the Edo/Tokugawa/early modern period, I am much more interested in Genji reception and interpretation. To me, the narrator is much more interesting than the author.

Best wishes,

Lawrence Marceau


From Rein Raud

Royall,

Please do not worry about wasting time and space to explain the relationships between Uji/main GM and KKS/MYS on the other hand. It would be extremely interesting to hear your opinion on this matter, although I can well understand your reluctance to wander into another "loaded" debate.


From Royall Tyler

In reply to Lawrence Marceau,

Ah, yes, I should have mentioned, too, that I know perfectly well the Sarashina diarist read those 50-odd vols. not long after Murasaki Shikibu's time. Perhaps there's no escape--MS wrote it all. But I still can't believe it. I know that a computer study of the text, done in Japan not long ago, concluded the author of the Uji chapters is different. That's all I know, though. Seriously, it seems to me that even if Takasue no Musume read those 50-odd vols., that still doesn't prove MS wrote them all. It could quite well be that her work was such a success that other people, perhaps even people who had known her, were inspired by it to try their hand at continuing or supplementing it, and that these people and their friends allowed this new work to stand as part of the old, under MS's great name. I can see nothing unnatural in this. On the contrary, it strikes me as

It may be that the "voices" you hear in some of your readings may sound different depending on the specific text or texts you are interacting with.

No, this has nothing to do with the issue. I'm of course talking about the text I read, not an unknowable original, and that text does not vary anything like enough between the editions I use to make one angstrom of difference.

All I can say is that, even if we knew more about the author (or authors), we must always be careful not to confuse the author with a particular narrator.

No doubt we must. But I reject the idea of an author who is a zero-point, a scribbling void, who can give rise, indifferently, any narrator-persona whatever, saying anything whatever. Scrupulously to avoid confusing the author with a particular narrator is (is it not?) to deny that "author" has any useful meaning in this case. I understand that some may believe that, actually. For better or worse, not me.

However, as a literary historian dealing with the Edo/Tokugawa/early modern period, I am much more interested in Genji reception and interpretation.

The matter of Genji reception in all periods, right through to 1999, is absolutely fascinating and deserves ample study. Many fair maidens, dragons, treasures, and secrets lurk in that forest. I often think (at least in my present fancy) of going on to study aspects of the reception of Genji, after I have acquired a some grasp of the work itself (I mean, as published in the currently standard texts).

To me, the narrator is much more interesting than the author.

I'm not sure I know what you mean. By narrator, do you mean "what the book (the narrator/s) says"? Or do you mean the approximate identity, social level, point of view, etc. of the narrator? In any case, the reception of Genji monogatari through history, and no doubt never more so than now, is closely connected with the reception of the idea of Murasaki Shikibu as author. There is a shifting Murasaki Shikibu-zoo that belongs to the history of the reception of the tale. In that sense, I should think that the author could be easily seen as a lot more interesting than the narrator.

Royall Tyler



From Royall Tyler

In answer to Rein Raud--

Please do not worry about wasting time and space to explain the relationships between Uji/main GM and KKS/MYS on the other hand.

I'm afraid it's more that I don't want to spend that much time writing out what flits around in my mind. I can imagine doing a long essay on the subject--sometime in the far future--unless my ideas change, which they may.

The Uji chapters as I've--how to say--tasted them???--not that many others have not--so far remind me somehow of the poetry of Ki no Tsurayuki and of the Kokinshu, whereas the main chapters in retrospect have more that Hitomaro-Akahito-Okura feel. The main chapters are (to me) varied, somewhat unruly, loosely structured, warm in feeling, wide in range. Whereas the Uji chapters (past "Takekawa") are meticulously plotted, with a seductive sort of artificiality, and narrow in range of literary concern. I once heard Mitsuta Kazunobu of Nichibunken (a wonderful man and a living database of waka and renga) describe the poetry of Tsurayuki as "metallic". I knew exactly what he meant. I don't think the Uji chapters are metallic, but they have a sort of thin, sustained literary sheen--the sheen of sophisticated artifice--than seems to me very unlike the main chapters. I can understand why people like Diane de Margerie in France (she reads Waley, not Sieffert) have been and sometimes still are keen on the main chapters of Genji as a "psychological novel", even though the idea is terribly vague and may not, strictly speaking, be appropriate at all), but psychology in the Uji chapters is usually fragmented into a stream of moment by moment sentiment, almost as in a series of waka; and the permitted causes of these emotional states fall within a narrow, strictly personal range (for example, personal treachery on the part of Niou, rather than the inevitable result, against his own will, of the restrictions placed on a future crown prince).

One thing that strikes me in the Uji chapters is the sometimes appalling isolation of the characters from one another. Even sisters, or the best of friends, seem at times to have no clue to what the other is thinking or to why the other is doing what he/she is doing. And, remarkably, the nyooboo do nothing to help. They seem in these chapters, as not in the main ones, to belong almost to a different race. For example, it is difficult enough to accept that Nakanokimi will not even hint to Niou that she is pregnant, but that not one of her women will (or is allowed to) slip the hint to him instead strikes me as horrifying. This sort of thing reminds me of some passages of Beckett (not Godot), for example one in which the the narrator is lying "dans la boue dans le noir" beside one Pym, shinnaigo-ing on and on about what it was like "la-haut dans la lumiere" (or something like that) and occasionally sticking a fork into Pym's buttocks at least to elicit a grunt from him, for company. Perhaps this effect is due to the serial way the emotional states in these chapters are evoked, and to the focus on inner states rather than on external circumstances, and to ongoing artifice of the plot.

Well, I beg everyone's indulgence. I have waxed sufficiently unintelligible for one afternoon.

Royall Tyler


From Lawrence Marceau

To Royall, in brief response, and in general agreement:

Seriously, it seems to me that even if Takasue no Musume read those 50-odd vols., that still doesn't prove MS wrote them all. It could quite well be that her work was such a success that other people, perhaps even people who had known her, were inspired by it to try their hand at continuing or supplementing it, and that these people and their friends allowed this new work to stand as part of the old, under MS's great name. I can see nothing unnatural in this.

Neither do I. In fact it would be a tribute to Murasaki Shikibu to have a coterie of gifted followers continuing the saga after she was unable to continue herself. Others since that time, including a young Motoori Norinaga, have since tried their hands at providing a sequel or addendum of some sort, with mixed results.

[LM] To me, the narrator is much more interesting than the author.
[RT] I'm not sure I know what you mean. By narrator, do you mean "what the book (the narrator/s) says"? Or do you mean the approximate identity, social level, point of view, etc. of the narrator? In any case, the reception of Genji monogatari through history, and no doubt never more so than now, is closely connected with the reception of the idea of Murasaki Shikibu as author. There is a shifting Murasaki Shikibu-zoo that belongs to the history of the reception of the tale. In that sense, I should think that the author could be easily seen as a lot more interesting than the narrator.

Sorry for the lack of clarity. I meant the narrator as a character in the narrative, with her social status, her opinions, her access to the various characters, etc. Of course the entire history of Genji reading has been to equate the narrator with Murasaki Shikibu. I only meant to suggest that the narrator is a construct of the author, just as are Genji, Yugao, Murasaki no ue, etc.

I guess, in conclusion, that I certainly didn't intend to come off as negative with regard to the issue of multiple authorship--just to stress that, ultimately, the work itself as it has been transmitted, interpreted, read, misread, misinterpreted, etc. over the last thousand years draws me more closely than trying to get back to the "original" version.

I've done some work in the transmission of Aesop's Fables to Japan, and have learned that probably none of the fables actually came from the hand of Aesop...

Best,

Lawrence M.



From Janine Beichman

Lawrence, interesting comment ---but then who *was* Aesop? Like Homer, many people?


From Anthony J. Bryant

Lawrence Marceau wrote:

Neither do I. In fact it would be a tribute to Murasaki Shikibu to have a coterie "of gifted followers continuing the saga after she was unable to continue herself. Others since that time, including a young Motoori Norinaga, have since tried their hands at providing a sequel or addendum of some sort, with mixed results.

Interesting point. For some reason, the book/film "Scarlett" -- the horrid "sequel" to "GWTW" comes immediately to mind.


From Lewis Cook

Thanks, Royall, for raising such an absorbing question (or for reopening such a notorious can of worms) and fleshing it out with first-hand reports from the front lines ;-} Can we now assume that your query on JLIT-L some months back on the Death of the Author was a prequel to this? Anyway, I hope this thread lives on until I can justify taking the time to respond with the circumspection your questions demand; till then just a few quick asides. In the intro to his translation Seidensticker, as you must recall, dismisses the multiple authorship theory (which has a history about as long as that of the recorded history of the reception of GM, prime alternative candidates in early speculation on the issue being MS's daughter and her father, not to mention a Bodhisattva or two) in memorable words: "The historical fact is that whoever wrote the Genji had no successors, and so the theory of mass genius has very little to support it. ... changes and additions in detail may have come later but the narrative points essentially to a single author working over a long period of time ..." ("Essentially?" ...well, yes but) Seems ironic to me that if there is one moment and place in *pmj* history when & where an outbreak of mass genius could be attested it would be the first decade or so of the11th century and the 'salon' in which MS was employed.

The list being quiet, perhaps I might raise an issue I often have on my mind, on the strenth of powerful impressions I get in the course of my passage through the Genji text. As I listen to the text, I hear many voices.

Keeping in mind here Lawrence M's illuminating responses (and yours to his) I don't see how you could not hear many voices. The burden of all those soushiji interventions, especially those --- "I didn't invent this, I'm just the scribe recording what this or that witness / nyoubou / narrator told me" --- which (with much interesting diversity among variant texts) tend to cap the ending of most maki in almost all mss. (and at least some of which have been omitted in Seidensticker's translation) is, on my reading, not only to deflect suspicions or attributions of 'authority' from the scriptor, but to allow her (MS?) the latitude or freedom to speak in tongues, feign a multiple personality disorder, play at shamanism, do whatever it took to exempt herself from responsibility for authoring a story which, whatever else it may be, is also (in its first many chapters at least) a fairly devastating critique of the Fujiwara program for venal control of the throne, etc.

I have not studied this matter. I know a few have always been convinced that Genji monogatari is not all by the same writer, but I also have the impression that orthodox opinion recognizes Murasaki Shikibu as the sole author.

Just an impression but I think that orthodox opinion as of the last two or three decades is divided, or better vacillating, on this question, i.e. that recent scholarship acknowledges that there isn't much basis for decisive judgements. Ishida Jouji wrote an oft-cited paper arguing from (somewhat spare) linguistic / stylistic evidence that "Takegawa" cannot have been written by MS or by the same author who wrote the first 40 some chapters anyway. What, I wonder, do you (Royall) make of the opening lines to "Takegawa"? Here more starkly than anywhere else we have a narrator making the kinds of claims one might expect from a sequel-writing author bent on rationalizing a palpable shift in style and perspective, not to mention the question about reliability of sources, but can we be confident that this isn't yet another one of MS's ploys, calculated to disarm the reader who longs for 'closure' on the authorship issue? (Vague recollection here of someone, publisher?, editor?, urging Joyce not to work so hard on sounding like a completely dfferent author / stylist / voice with each successive chapter of Ulysses. I wonder whether one of MS's accomplishments may not have been to dissimulate her own voice, manipulate styles, and so compound the tantalizating question of her identity and authorship, etc.?)

Sorry to sound so indecisive myself. One further aside, partly addressed to Lawrence. I think we should be careful to make lots of fine distinctions among putative authors, implied authors, multiple narrators, (alleged) witnesses, informants and others, since the real author(s) or scriptor(s) (MS or whoever) evidently took so much care to distinguish and (yes, also) to confuse these and thereby afford us such a measure of freedom for reading with imagination (against which literal questions about who wrote what seem rather pale).

Hoping to be able to reply with more substance next week,

Lewis Cook


From Royall Tyler

Thanks for the message, Lewis.

Can we now assume that your query on JLIT-L some months back on the Death of the Author was a prequel to this?

--No. Back then, I was just thinking MS, mostly.

In the intro to his translation Seidensticker, as you must recall, dismisses the multiple authorship theory (which has a history about as long as that of the recorded history of the reception of GM, prime alternative candidates in early speculation on the issue being MS's daughter and her father, not to mention a Bodhisattva or two) in memorable words: "The historical fact is that whoever wrote the Genji had no successors, and so the theory of mass genius has very little to support it. ... changes and additions in detail may have come later but the narrative points essentially to a single author working over a long period of time ..."

--Well, MS's father, F no Michinaga, bodhisattvas, etc., need not concern us. But I don't agree at this point that "the narrative points essentially to a single author."

("Essentially?" ...well, yes but) Seems ironic to me that if there is one moment and place in *pmj* history when & where an outbreak of mass genius could be attested it would be the first decade or so of the 11th century and the 'salon' in which MS was employed.

--Well, yes, that does come to mind. The thing is that, in principle, single authorship throughout seems to me less likely than some extent of multiple authorship for any text of anything like this length, anything like this early, from anywhere in the world. Nor is it difficult to imagine the whole thing being attributed to the great, original genius, MS. (That's just what happened with Zeami.)

Keeping in mind here Lawrence M's illuminating responses (and yours to his) I don't see how you could not hear many voices. The burden of all those _soushiji_ interventions, especially those --- "I didn't invent this, I'm just the scribe recording what this or that witness / nyoubou / narrator told me" --- which (with much interesting diversity among variant texts) tend to cap the ending of most _maki_ in almost all mss. (and at least some of which have been omitted in Seidensticker's translation)...

--Lewis, your knowledge of Genji texts clearly goes far beyond mine. But the major issue for me is not sooshiji interventions. It's in point of view, range of interest and sympathy, conception of character and motivation, plot construction, and so on.

but to allow her (MS?) the latitude or freedom to speak in tongues, feign a multiple personality disorder, play at shamanism, do whatever it took to exempt herself from responsibility for authoring a story which, whatever else it may be, is also (in its first many chapters at least) a fairly devastating critique of the Fujiwara program for venal control of the throne, etc.

--Hmm. Is there any reason to believe that those "first many chapters" were actually taken as subversive? MS's diary gives none. Perhaps one feels they should or must have been taken as subversive, but were they?

Just an impression but I think that orthodox opinion as of the last two or three decades is divided, or better vacillating, on this question, i.e. that recent scholarship acknowledges that there isn't much basis for decisive judgements.

--Undoubtedly.

Ishida Jouji wrote an oft-cited paper arguing from (somewhat spare) linguistic / stylistic evidence that _Takegawa_ cannot have been written by MS or by the same author who wrote the first 40 some chapters anyway. What, I wonder, do you (Royall) make of the opening lines to "Takegawa"?

--Well, considering what a mess these first chapters of the post-seihen sequence are, I see no compelling reason not to take them at face value. Someone else (perhaps someone with clout) is putting in her own chapter.

Here more starkly than anywhere else we have a narrator making the kinds of claims one might expect from a sequel-writing author bent on rationalizing a palpable shift in style and perspective, not to mention the question about reliability of sources, but can we be confident that this isn't yet another one of MS's ploys, calculated to disarm the reader who longs for 'closure' on the authorship issue?

--No, of course not. (Lucky MS, she had never heard of "closure.") But no, the reader does not long for (that word) on the authorship issue. Not this reader. I was going to say I don't care whether she wrote the whole thing or not, but actually, I do. I'd rather she had. It's just that I can't believe it.

Sorry to sound so indecisive myself.

--How could you not?

I think we should be careful to make lots of fine distinctions among putative authors, implied authors, multiple narrators, (alleged) witnesses, informants and others, since the _real_ author(s) or scriptor(s) (MS or whoever) evidently took so much care to distinguish and (yes, also) to confuse these and thereby afford us such a measure of freedom for reading with imagination (against which literal questions about who wrote what seem rather pale).

--By all means, Vive la liberte, but I'm afraid it's reading with imagination that started this discussion in the first place. Even authors of genius must be finite. Positing a shape-changing author who could have written the main chapters, and Takekawa, and the Uji chapters, and perhaps Finnegan's Wake as well if she'd had a mind to, leaves me feeling as though nothing has been said. Anyway, the issue is perhaps less who wrote what, since that's unknowable, than recognizing difference where there is difference--though I can already hear someone saying, Why?

I remember Fujii Sadakazu saying that until fairly recently scholars thought they had Genji monogatari pretty well down, but now they recognize again that there are all sort of things they don't know. The book is full of questions.

Royall Tyler


From Chris Drake

Royall,

Given your lack of belief in the single authorship of the GM and your sense of its stylistic diversity, are you still going to title your translation 'The Tale of Genji'? Or are you thinking of 'Tales of Genji' or even something along the lines of 'Tales From the Heian Woods' etc.? I'd be interested to know how you're translating the title.


From Royall Tyler

[...] are you still going to title your translation 'The Tale of Genji'?

If I have given anyone the impression that I might do anything else, I apologize.


From Lewis Cook

Thanks, Royall, for this thoughtful and thought-provoking reply. I'll have to make do with a very hasty and partial response for now. I'm fascinated by (and didn't mean to sound as skeptical of as I may have in my previous) your argument that we should look more closely for signs of multiple authors --- also somewhat repelled, not only because of the difficulties that's bound to entail, but on a practical (even emotional) level because so much solace is to be taken, when grappling with the difficulties of any given passage, from the assurance that a single more or less finite mind is at work, that with enough perseverance the mysteries can be resolved. I do that is share your (nostalgic) wish that MS had written the whole (with only a few qualms about conceding the three vagrant maki following the sei-hen) and that the task is to just discover how it all hangs together, something quite more manageable, finite anyway, than discerning how and where it hangs apart. You refer in your reply to Lawrence, by the way, to a statistical study which tended to confirm that the Uji-jyujou are of different authorship. Someone sent me a partial clipping from the Yomiuri --- early July, I think --- reporting this under a prominent headline. The gist seemed to be that some of the theories about the order in which the first several maki were composed, or the 'tate- / narabi' strata, were supported by the results, but I've been meaning to ask whether anyone on the list has read the actual report, where it was published, and what response it may have evoked in non-journalistic quarters, etc. One salutary effect, in my experience, of reports of the Death of the Author (c. the mid-70's or so) was to relieve us (me at least) of the notion that biographical detail is of much immediate pertinence to reading poems or stories (I, for one, would feel much more comfortable about admiring Robert Frost's poems, just for example, if we didn't have all the squalid details supplied by so many long and lugubrious biographies, and I wonder --- no parallels implied of course --- if we may not be fortunate to have as little knowledge of MS the person as we do?) Seems ironic that Roland Barthes, one of the founding advocates of the DotA doctrine, also reminds us (I doubt he was the first to do so) that in a literary text there is no such thing as an insignificant detail. But the easy inference (which I don't think can be attributed to Barthes) has been that we can replace the physiological author with the abstract but still comfortingly unified notion of a distinctive Style, underpinned by all its tacit references to 'creative genius at work.' If we let go of this, and with it the assumption of a single presiding author, the purportedly compelling significance of all those details, each little turn of phrase, etc., melts into very thin air indeed. The only practical response, I imagine, would be to take up some of those passages which rouse your suspicions, from Yugiri and Yadorigi, e.g., and scrutinize them word for word with a view to specifying their aberrations. The varous databases we have that offer e-texts of Genji might be helpful.

I've snipped most of what follows to avoid the confusion of nested quotations...

The thing is that, in principle, single authorship throughout seems to me less likely than some extent of multiple authorship for any text of anything like this length, anything like this early, from anywhere in the world. Nor is it difficult to imagine the whole thing being attributed to the great, original genius, MS. (That's just what happened with Zeami.)

I haven't looked into this, but my impression is that a distinct image of MS as sole author of GM & genius doesn't make its appearance until well into the medieval era. Her admission into the pantheon of chokusenshuu poets was belated, and the contemporary evidence identifying her as author of GM is surprisingly tenuous, no? I should be better informed, I guess. The standard argument for single authorship seems to be the remarkable consistency of detail in the chronology, designation of ranks and titles, etc. throughout the sei-hen, at least, though this is the sort of thing one can imagine an editor bringing into consistency after the fact (especially if she had use of a word processor ;-} ).

the major issue for me is not sooshiji interventions. It's in point of view, range of interest and sympathy, conception of character and motivation, plot construction, and so on.

I can appreciate this. The greatest degree of variation among early mss. does occur within the soushiji ending each chapter (just my impression, but I think it's a safe generalization) but this is not where we would look for substantial evidence of multiple authorship.

Is there any reason to believe that those "first many chapters" were actually taken as subversive? MS's diary gives none. Perhaps one feels they should or must have been taken as subversive, but were they?

Well, this is a difficult question and much hinges I think on the differences between "taken as" and (possibly) "meant as." What do you make, for example, of the proposed allusion to IseMg dan 101 towards the end of "Hana no en" (fn 26, p. 283 v. 1 of SNKBT ed., fn. 22, p. 435 of Shogakkan ed.)? (If you follow this through as far as the suggestion that "fuji" is a parasitic vine that eventually kills its host, the implications are subversive enough to satisfy the most suspicious, though subtle enough to avoid obtrusiveness, just right in short.)

Ishida Jouji wrote an oft-cited paper arguing from (somewhat spare) linguistic / stylistic evidence that _Takegawa_ cannot have been written by MS or by the same author who wrote the first 40 some chapters anyway. What, I wonder, do you (Royall) make of the opening lines to _Takegawa_? --Well, considering what a mess these first chapters of the post-seihen sequence are, I see no compelling reason not to take them at face value. Someone else (perhaps someone with clout) is putting in her own chapter.

Have (reluctantly) to agree this seems very plausible.

... (Lucky MS, she had never heard of "closure.") But no, the reader does not long for (that word) on the authorship issue. Not this reader. I was going to say I don't care whether she wrote the whole thing or not, but actually, I do. I'd rather she had. It's just that I can't believe it.

This reminds me to raise a question about "closure" in another sense, in medieval Japanese, will have to wait.

Vive la liberte but I'm afraid it's reading with imagination that started this discussion in the first place. Even authors of genius must be finite. Positing a shape-changing author who could have written the main chapters, and Takekawa, and the Uji chapters, and perhaps Finnegan's Wake as well if she'd had a mind to, leaves me feeling as though nothing has been said. Anyway, the issue is perhaps less who wrote what, since that's unknowable, than recognizing difference where there is difference--though I can already hear someone saying, Why?

I'm repeating myself but I'm entirely in favor of doing everything possible to discern and weigh differences, even if it means relinquishing our nostalgia for the singular author. Just a question of how to proceed.

I remember Fujii Sadakazu saying that until fairly recently scholars thought they had Genji monogatari pretty well down, but now they recognize again that there are all sort of things they don't know. The book is full of questions.

Recall Fujii saying (shortly after he'd finished the footnotes for vol. I of the SNKBT Genji), roughly, that "Genjimonogatari wa mada seiritsu site inai," not sure how "seiritsu" in this sense might go into English, something not quite the same as "the text / work has yet to be established." But I'm afraid he was right.

(sorry have to cut off here ---)

Lewis Cook


From Stephen Forrest

This thread has been fascinating and provocative, and I feel ill-equipped to weigh in on Genji matters, but a couple of things in Lewis' last mail leave room for clarification. He wrote (in part):

I haven't looked into this, but my impression is that a distinct image of MS as sole author of GM & genius doesn't make its appearance until well into the medieval era. Her admission into the pantheon of chokusenshuu poets was belated, and the contemporary evidence identifying her as author of GM is surprisingly tenuous, no? I should be better informed, I guess.

Period nomenclature again--do we all agree (approx.) on what "medieval" means in Japanese [literary] history? If we take it back as far as 1100 or 1150, then "well into the medieval era" would perhaps fit Teika's early 13th cent. adoration of GM and MS both. (I wonder, while we're on the subject, when "medieval" comes to an end, especially if it starts that early?)

Then again, thinking of the "pantheon of chokusenshu poets," even in the 1180s Shunzei grants MS space for nine verses in his Senzaishu (compared to, say, the three verses she has in the Goshuishu of 1086, the first chokusenshu after her death). Not a huge total, but definitely a mark of respect. Her "admission" was perhaps belated because she so clearly put her energies into her tale, and did not join in the increasingly professionalized and competitive world of (chokusen-minded) waka composition. Moreover, as we have it now the majority of her poetic oeuvre is *fictional*. There was (at first) nowhere to put such poems in a chokusenshu (am I right in thinking that eventually some GM poems appear in the latter 13 chokusenshu under her name?).

One last point: contemporary evidence. It seems to be par for the course for the 10th-11th centuries that literary figures loom far larger in the tradition than solid documentary evidence can explain--there just isn't enough surviving material for many poets and authors to talk about good historical biographies. It's enough to make you want to switch to the Edo period sometimes, just to get back into the world of solid facts and corroborating documentation. Then again, it does leave so much room for argument and imagination...

Steve Forrest


From Chris Drake

Royall,

You have no need to apologize. I was just curious after reading your interesting remarks about GM. It seems to me personally that if one doesn't consider GM to be by a single author or to be done in a single repertoire of styles, then 'Tales of Genji' or possibly some other title with 'Tales' in it would be just as appropriate in English as 'The Tale of Genji.' (This is not to say, of course, that there aren't many considerations that go into choosing the title in any particular case.) I was just wondering if you'd been thinking along these lines or not. Thanks very much for your clear reply!


From Michael Watson

I've been waiting for someone to mention Ichijo Kanera who introduces his commentary on the "Uji no maki" by saying something to the effect of "some believe that all fifty-four maki, up to and including "Yume no ukihashi", were written by MS. Others ("aru hito no moosu-haberu") say that the ten Uji maki were written by [her] daughter Daini sanmi. The proof is clear" ("sono shooko akiraka nari"). We read on in eager expectation for evidence but he rather spoils things by saying it's all very much like the question of whether Pan Piao or his son Pan Ku wrote "Shih chi"! He then goes on at boring length about the origin of the place-name "Uji"--most disappointing.

(I don't need to tell anyone that Shih chi is more usually attributed to Ssu-ma Ch'ien [d. 90? B.C.] and not to Pan Ku [d. 92 A.D.] , author of the later Han shu history.)

See "Kachoo Yosei" in Genji monogatari kochuushaku sookan, vol. 2, ed. Nakano Kooichi (Musashino Shoin, 1978), p. 316. As always when I comment on the Genji, I owe this reference to my wife Midorikawa Machiko.

Now to the main bit of "research" I've done on this topic--more Web searching, I'm afraid, and also one telephone call. But I wanted to pin down the media reports many of us have heard or read claiming statistical "proof" that the Uji section is not the work of MS. In short, the answer is: the jury is still out. Ý(In plain English and for those in countries without a jury system: no clear answer yet.)

A big project was initiated in the early 90s by a group led by MURAKAMI Masakatsu of the Tookei Suuri Kenkyuujo (Institute of Statistical Mathematics, www.ism.ac.jp). The only printed paper we have at hand is

Murakami Masakatsu, "Bunshoo bunseki to tookeigaku," Suuri kagaku, no. 389 (Nov., 1995), pp. 27-33. 村上征勝「文章分析と統計学」(数理科学)

The Suuri kagaku article by Murakami gives examples from work done on known modern writers to demonstrate differences in style through frequency in use of common particles and punctuation. It then goes on to show this technique applied to works that may or may not be by Nichiren. Pages 31-33 summarizes the work done by the group on a 380,000 word database of the Ikeda Kikan text (Genji monogatari taisei). Murakami begins by listing the proportion of parts of speech (nouns 30.7%, verbs 38.5% etc .) in the whole text of Genji and comparing this with the results of Motoori Norinaga's "Temakura", a work written in the style of Genji..

He then summarizes results of frequency analysis of parts of speech, comparing maki 1-44 with 45-54. Nouns make up 18.189% of the text of the first forty-four chapters, 16.423% of the last ten. This means a t-value (?) of 5.1624 which is regarded as significant. Similarly with jodooshi, while other parts of speech show no great difference. "Meishi to jodooshi no shiyooritsu ni wa akiraka ni sa ga mirareru koto ga wakaru". A second comparison is in frequency of use of key words--hito, mono, omoi--which again reveals significant differences.

He suggests that this kind of change in style (buntai no henka) that some have sensed (nantonaku...) is probably what lead them to argue in favour of different authorship of the ten Uji chapters (Uji juujoo ta-sakkasetsu 宇治十帖他作家説).

But he goes on to warn that at the present time (genjiten) we cannot conclude from results like these that the Uji chapters were not written by MS. It is still possible that her style changed. The final graph shows frequency of seven parts of speech plotted maki by maki. When the dots representing individual maki are grouped by the traditional three "parts" ("bu"--chapters 1-33, 34-44, 45-54), it is possible to read/interpret the results as a change in style over the course of the work ("hinshi no shiyooritsu ga henka shite iru yoo ni mo yomi-toreru"). Nice and cautious that, "...yoo ni mo..."

I've printed out but only glanced through three full papers which you can find online, with charts and footnotes. Highly recommended for the interested:
http://www.genji.co.jp/1996/ron9609.htm http://www.genji.co.jp/1997/ron9709.htm

The first is a lengthy statistical analysis of incidence of jodooshi in speech (kaiwabun) and narrative (ji no bun). The authors are Murakami Masakatsu, Ueda Hideyo (affiliation "Koten soogoo kenkyujo"--the site) and Fujita Mari (Tooden gakuen). The second paper is titled "Genji monogatari no keiryoo bunseki" (computational analysis of GM) and is a short piece written under the names of six members of the group, Murakami, Ueda, Fujita plus Imanishi Yuuichiroo (Kyushu U.), Kabashima Tadao (Kobe Gakuin U.) and Ueda Yuuichi (Motobunoge Hospital).

Imanishi Yuuichiroo is a well-known Genji specialist whom I've met at "computer & kokubungaku" meetings at the Kokubungaku Kenkyu Shiryokan (www.nijl.ac.jp), so I took the liberty of digging out his meishi and giving him a call. He was very helpful. He said that as a statistician Murakami was arguing that the figures showed a "significant difference" in comparing language use, but that this alone will not convince literature specialists.

What Professor Imanishi believes necessary now is to compare the figures for other Heian works. He mentioned Utsuho and Sagoromo. He has begun work with the Utsuho editor Muroki on an Utsuho database and analysis, and hopes to finish it in a year and a half. (He won Mombusho "kaken" funding for the project.)

Exciting stuff, to me at least. Some other Web pages of information are here:

http://www.ism.ac.jp/staff-lists/murakami.html http://www.ism.ac.jp/~murakami/jhome.htm http://www.kotengaku.bun.kyoto-u.ac.jp/letter/002/outline/3/a03-3.html all pages concerned with Murakami (who gives his email address, by the way)

There is also an interesting discussion between academics referring to the television report in spring 1998 that alerted the two of us here to the story: http://kuzan.f-edu.fukui-u.ac.jp/BBS/room_1/BBS_MSG_980409203015.html (For the uninitiated, this url is from one of the best sites for downloadable classical literary texts)

Michael Watson



From Robert Khan

For a succinct paper in English taking the lexicostatistical approach to stylistic analysis of the Genji (focusing on subjective adjective usage), I recommend Andrew Armour's article:

"Analysing an Author's Idiolect: Murasaki Shikibu. " Poetica 1985, 164-180.

My apologies if this has been already mentioned in the discussion - I don't recall having noticed it so far.

I followed a similar approach in surveying the frequency and distribution of subjective adjectives, exclamatory particles, and intensifying adverbs in my study of the language of 'Ariake no Wakare', to get a sense of how linguistically 'close' it is to the Genji relative to some other monogatari.

Thanks to all for an extremely interesting discussion.

Robert Khan


From Rein Raud

He then summarizes results of frequency analysis of parts of speech, comparing maki 1-44 with 45-54. Nouns make up 18.189% of the text of the first forty-four chapters, 16.423% of the last ten. This means a t-value (?) of 5.1624 which is regarded as significant.

A t-value, if I am not mistaken, is calculated from:

X = how many times something occurred in a part of text Y = how many times that thing occurred in the whole text, and should have occurred in that part of text also, if they were homogeneous

with the formula

T = (X-Y)/X

More on the matter in e.g. Geoff Barnbrook, Language and Computers, Edinburgh University Press 1996

He said that as a statistician Murakami was arguing that the figures showed a "significant difference" in comparing language use, but that this alone will not convince literature specialists.

I once read a study that sort of proved with statistical calculations that the sound [L] conveys sadness in the French of Racine, because it occurs significantly more often in sad passages. I remained unconvinced then, and still do.

Rein Raud


From Royall Tyler

Many thanks to everyone for your responses on this issue, and especially to Michael and Machiko for Ichijo Kanera, the state of computer analysis, and the phone call to Imanishi. I should have picked up Michael's mention of Kanera before. Now I too have read the passage. It's like what the 18th c. mathematician Fourier (wasn't it?) wrote on the margin of a page mentioning an unproved theorem, "There is a marvelous proof of this, but I have not space for it here." I think the theorem was demonstrated finally a few years ago, some two centuries later. Kanera seems to say that the justification for what "aru hito" says is obvious. But what did he mean? What was he thinking of?

It sounds so far as though various people have tried and are trying to prove the distinctness of the Uji chapters on linguistic grounds, but that no one, or no one known, has tried to argue the case of grounds of theme, approach, and style of narration. I do think it could be done, though.

Royall Tyler

* as RT points out in a message below, he meant Fermat's Last Theorem.



From Gaye Rowley

Dear Janine, Royall, and all,

Akiko did feel that all of the chapters after 'Fuji no uraba' were written by somebody else. In the afterword to her first translation of Genji, (Shin'yaku Genji monogatari, 1912-13) she writes:

The Tale of Genji can be divided into two large parts: the part in which Hikari [sic] and Murasaki are the main characters, and the part in which Kaoru and Ukifune are the main characters. When we reach the ten Uji chapters...the extreme glitter and refinement of the exquisite narrative of the first part gives way to simpler descriptive passages. This air of freshness, this sense of rejuvenation, is the product of Murasaki Shikibu's genius, ever vigorous... ("Shin'yaku Genji monogatari no nochi ni," 1913)

Later, at least by 1928, Akiko had come to the conclusion that the author of the "second part" (i.e. 'Wakana' to 'Yume no Ukihashi') was Murasaki's daughter Daini no Sanmi. Akiko's reasons are most explicitly stated in the afterword to her second translation of Genji (Shin-shin'yaku Genji monogatari, 1938-39). The passage is too long to quote here, but Akiko's reasons include differences in style; quality of poetry; her interpretation of the opening lines of 'Takekawa' ("Murasaki no yukari koyonaki niwa nizameredo", which Akiko takes to mean: "what follows will not be of the same quality as the previous chapters written by Murasaki Shikibu"); as well as Daini no Sanmi's familiarity with Uji; and evidence from the Sarashina nikki. She concludes: "It is a shame that I do not have the time to explain this in more detail." A shame indeed! Akiko's daughter Mori Fujiko writes, in an essay written shortly after her mother's death, that Akiko's only regret was that she had been unable to publish in any detail the results of her research into the authorship of Genji.

We might also note that by 1940, no less a scholar of Genji than Ikeda Kikan was convinced that Akiko was right about the structural break in Genji after 'Fuji no Uraba', but unable to agree with her theory of dual-authorship. Ikeda recalls arguing--vehemently--with her about it several times, but he was never able to persuade her otherwise.

To return briefly to metaphors of heart and the marvellous passage from 'Yadorigi' that Royall Tyler brought to our attention: --Ainashi ya, waga kokoro yo, nani shi ni yuzuri-kikoeken, Akiko doesn't use kokoro in either of her translations. Shin-shin'yaku Genji monogatari gives: --Nantaru koto de arou, futsugô na no wa jibun de aru, nan no tame ni ano hito o miya e o-yuzurishita no de arou, But Tamagami gives: --Baka datta na, watashi no kokoro wa. Dooshite miya ni o-yuzurimooshiageta no ka. (Genji monogatari hyooshaku, 11:115)

Gaye Rowley


from Royall Tyler

Thank you very much, Gaye. That's fascinating. We seem to have yet another case of Fermat's (not Fourier's) Last Theorem--I wish she'd written it all down.


From: Royall Tyler
Date: Thu, 30 Dec 1999 05:08:44 +1000
Subject: Re: Genji authorship

For anyone interested, I have done a rudimentary summary on the Genji authorship issue, which I brought up on this list some time ago, from materials that I received during the earlier discussion (thanks again to all) or that I happen to have at hand.

A.
The long article in Nihon koten bungaku daijiten lists just two premodern doubts about authorship, apart from the ideas that MS's father, Tametoki, laid out the whole plan and left his daughter to fill in the details, and
that Michinaga added to MS's work:

1. Ichijô Kanera's statement in Kachô yosei that "aru hito" says the Uji chapters are by MS's daughter, Daini no Sanmi.

2. Hanaya Gyokuei's statement in Gyokueishû (1602) that the Uji chapters are by Daini no Sanmi and Takekawa by Saiin Senshi. The NKBD article dismisses these as mere denbun and unworthy of credence.

For modern times, the NKBD article notes theories that (1) Wakana 1 and after; (2) Niou no Miya and after; (3) Niou, K ai, Takekawa; and (4) the ten Uji chapters are by someone else. It gives extensive bibliographic
information on the controversy over (3) and concludes that the issue is not resolved yet. It gives no information whatsoever on who has held positions (1), (2), and (4). Perhaps these people were/are not accredited scholars.
Certainly the omission of Yosano Akiko, whose final view (as wonderfully described by Gaye Rowley) corresponded to (1), is quite striking. In contrast, Ikeda Kikan, with whom Akiko argued so stubbornly, is frequently cited.

B.
A wrapup piece by Suzuki Hideo ("Genji monogatari no seiritsu," Kokubungaku kaishaku to kyôzai no kenkyû, vol. 40, no. 3 (Feb. 1995), pp. 40-42) concentrates on the question of what order the chapters were written in,
and it mentions the possibility of more than one author only in a summary of Origuchi Shinobu's essay, "Nihon no sôi." Suzuki does not mention the Niou no Miya-Takekawa chapters. (As far as I can see, Origuchi does not
discuss authorship in "Nihon no sôi," although one has the impression that he may have been open to the idea of multiple authorship.)

Suzuki, like others, cites as seminal two ideas put forward by Watsuji Tetsurô in a 1922 article ("Genji monogatari ni tsuite," Watsuji Tetsurô zenshû, vol. 4, Iwanami, 1962): that (1) Hahakigi was written first and that (2) there must have been a pre-GM ur-genji tale on which the present GM was built. This is not a variant of the Father-as-real-author theory in Yotsugi monogatari. Watsuji makes the suggestion (not mentioned by Suzuki)
that the present tale is the product of a sort of studio of writers under the direction of a single author, and based on an original (gen, ur-) GM solely by Murasaki Shikibu. Presumably scholars have seen that suggestion
as less seminal than terminal.

C.
In a brief squib (Kokubungaku kaishaku to ky ai no kenkyû, vol. 44, no. 5 (April 1999), p. 136) Higashihara Nobuaki says, "The hypothesis of three authors is at present highly regarded as an illustration of the splits in
the shutai of the tale." But where is this hypothesis to be found? Who is involved?

D.
These materials do not explain the assumption by René Sieffert ("Furansujin kara mita Genji monogatari," Kaikan j o sh en kinen k nkai kirokushû, Tama Shiritsu Toshokan, 1989), Edward Seidensticker ("Waga Genji zô,"
Kokubungaku 14:1, January 1969), and Setouchi Jakuchô (her recent lecture in Chicago) that many others believe the Uji chapters to have been written by someone else. There seems to be NO body of scholarly opinion that obliged these three even to mention the idea. Nonetheless, all three did,as though they nonetheless felt a marked difference between the seihen andthe later chapters. (Sieffert conceded that while a cursory reading might well leave the reader with the impression that the Uji chapters are by another author, the intensive reading required by the effort of translation makes any such view untenable.)

Nonetheless, Yosano Akiko, who began with this sort of view, changed her opinion later on. Her first translation convinced her that the entire work is by MS. As Gaye Rowley translated her: "When we reach the ten Uji
chapters...the extreme glitter and refinement of the exquisite narrative of the first part gives way to simpler descriptive passages. This air of freshness, this sense of rejuvenation, is the product of Murasaki Shikibu's
genius, ever vigorous." However, her second translation made her certain that Wakana 1 and after are by Daini no Sanmi, and according to her daughter, she always regretted not having had time to discuss the matter in
print.

I understand that colleagues should find the question of narrator and narratorial technique in GM more interesting than that of authorship. It is a much more solid academic issue. Still, Murasaki Shikibu is not just a
name, like Daini no Sanmi. She is a national heroine. People want to believe that she and Genji monogatari are, so to speak, one and the same--that GM, as we have it, is from beginning to end the unmediated expression of her individual genius. That is an interesting phenomenon in itself.

Royall Tyler



Editor: Michael Watson <watson@k.meijigakuin.ac.jp>
99/11/12. Changed to iso format 01/02/18.To Unicode format 05/01.
Later messages added 01/07/27