Archive of discussion on the PMJS mailing list (Jan 21 - Feb 8, 2000). All rights reserved.
Question raised by Lewis Cook
Discussants: Lewis Cook, Robert E. Morrell, Rein Raud, Richard Bowring, Paul S. Atkins, Noel Pinnington, Janine Beichman, Royall Tyler, Stephen Miller, Sonja Arntzen, Michael Watson, Lawrence Marceau, William Bodiford
The messages are listed in the order they were posted. The disparity in dates reflects the widely separated time zones in which members live.
See general note on editing. Japanese text given on separate page, with cross-reference.
return to: PMJS archive index
/ PMJS index / members
/
Date: Jan 21
2000
From: Lewis Cook
Subject: Extreme kakekotoba?
I've been perplexed for some time now by proposals for what strikes me as an extreme form or degree of kakekotoba in some commentaries on canonical texts, and hope that list members might have suggestions on how to help me overcome this perplexity.
The clearest example I can cite occurs in Komachi's poem "iro miete utsurou mono wa yo no naka no hito no kokoro no hana ni zo arikeru" (KKS No. 797). [Japanese]
(Jane Hirshfield's appealing translation of this, in Ink Dark Moon, reads: "How invisibly / it changes color / in this world, /the flower / of the human heart.")
The interpretive crux of this poem, for several centuries, has been whether "mie-te" is to be taken as affirmative ("miete") or as negative ("miede").
Hirshfield's translation assumes it is to be read "miede," and this has been the predominant interpretation since c. the 12th cent., but there have always been dissenters who prefer "iro miete..." (On this latter reading, Hirshfield's translation would have to be amended to "How visibly / it changes color...") (A fair enough summary of the debate and its terms as things stood c. 1970 is available in Takeoka Masao's KKS Zenhyoushaku)
The complication, and cause of my perplexity, is that at least some exegetes have proposed that this "-te / -de" is a kakekotoba and that both the affirmative and negative readings are to be understood at once. (The most recent example, far as I'm aware, is Arai and Kojima, eds., Shin Nihon Koten Bungaku Taikei Kokinshuu. 1989. In a footnote, they merely assert, without citing precedents or arguments, that this is a kakekotoba. At least one 14th c. KKS commentary makes this claim as well, however.)
This proposal is intriguing to me but also baffling. (It is not a fluke: there are other such instances --- one occurs in a comment by Kanera in Kachou Yosei on "Fujinouraba" in Genji, in which he claims that "aratamete / aratamede" is a kakekotoba.) Yet it seems to defy reason in a way that kakekotoba of the order of "matsu" as "pine" and "wait" do not. (Am I the only one troubled by this?)
Lewis Cook
Queens College, CUNY
To at least one literary archaeologist, the issue draws attention to the corrosive insistence on current Western Eurocentric analyses of Japanese (or any) non-European literature. . . Until we seriously try to come to terms with the philosophic bases for THEIR OWN literary practice, we are merely practicing intellectual colonialism.
A related issue. Current Western practice is against footnotes. Without footnotes, one can simply imply that one is not an academic pedant. Or, we can simply ignore the fact that such conventions as kakekotoba are much more prevalent in Japanese than they are in English, and then pretend that they really make no difference to the translation. I agree that, to me at least, many explanations/interpretations of ambiguous Japanese texts often seem forced. (Take, for example, the first three waka early on in the Taketori monogatari. The standard headnotes squeeze them dry. But is all this word-play REALLY intended? Who is to say?)
So what do we do in English? My own sense is that too little explication can often be more misleading than too much. But I also suppose that the issue will be debated for at least the next three centuries.
Bob
Why do we say that this instance (te ~de) of kakekotoba is extreme, whereas the standard "pine"/"wait" type is not? If any "logic" is valid in Komachi's world, it is certainly not Aristotelian, and there is no need to oppose affirmation/negation. For me, the particular charm of this poem by Komachi has always been in the fact that it posits two incompatible views of "the human heart" at once, and both are correct. And this precisely is the message of the poem. It is also corroborated by the double meaning of "iro" (colour/passion) and what "fade away" would mean in the context - it is one thing if the passion fades away, another if it applies to the whole heart-flower. As it is well known, the last two lines of Komachi's more chrestomatic KKS 113 do not even break into a single grammatical sentence, which also demonstrates that she is able to use the very fact that linguistic expressions has it limits in order to refer to the "extremeness" of the emotional reality beyond.
Rein Raud
I don't think I am as worried as you are about this being called
kakekotoba, which simply refers to one word being used in more
than one sense. Visually this is indeed one word. The positive-negative
switch is just as valid, although I suppose you could call it
more 'radical'. To put it another way, what else would you call
it in Japanese? The really interesting thing here is one of possible
over-interpretation. The interpretive gap offered by the quite
normal habit of omitting diacritics is surely omnipresent and
it is left to the reader whether s/he wishes to be creative or
not. But then I seem to remember somewhere in Brower and Miner
that describes one occasion when judge X ridiculed judge Y for
having made the 'stupid' mistake of having read 'tazu' instead
of 'tatsu'. Had Y really made a mistake, or was he being creative?
And was X just riding roughshod over him into order to show 'superior'
knowledge? I'm afraid I suspect that the majority of these courtiers
were probably a fairly literal and unimaginative lot.
R. Bowring
Prof. Cook, thank you for raising this interesting problem.
I found a similar mention of the -te/-de ambiguity in Yamashita Michiyo's Kokinshuu: koi no uta (Chikuma, 1987), but such a reading does not appear in the earlier Iwanami edition of the Kokinshuu, first published in 1958. One gets the feeling that this is an interpretation that has gained currency among Japanese scholars only in the last forty years or so, perhaps even in the last 15-20 years.
Using the Japanese Text Initiative's index, I was able to find a few instances of "miete" in the Shin Kokinshuu, but none seemed to carry the second meaning of -de. Kakekotoba are supposed to be established usages, so it's quite odd. Also if -te/-de were a kakekotoba, it would be somewhat unusual, because most kakekotoba involve two words of different grammatical classes. (e.g. matsu/pine is a noun, matsu/wait is a verb). But -te and -de are both particles appended to verbs.
Perhaps someone versed in historical linguistics could tell us if a distinction was made in speech in Heian Japanese between -te and -de. Would someone reading this poem out loud read it -te no matter what the interpretation was? Or would she read it as -de if that were her interpretation?
Let me go out on a limb and say that I don't think the two readings are equally meaningful. If the poem is read,
iro miete utsurou mono wa yo no naka no hito no kokoro no hana ni zo arikeru
it means simply that something that shows its color (passion) and then fades is the flower that is the human heart. That's not so interesting, because ordinary flowers do the same thing.
If however, the poem is read,
iro miede utsurou mono wa yo no naka no hito no kokoro no hana ni zo arikeru
it means that what fades without a perceptive change in color is the flower that is the human heart. Now that's interesting. A man loses interest in her, but she hasn't the slightest clue from him. We may note the contrastive wa, which implies a comparison between the flower of the human heart and ordinary flowers.
I don't have access to the KKS zen hyoushaku, so this may be an old argument.
I'm not sure to what extent this message is intended as a reply to my query, but since you mention intentions as well as footnotes, let me add a footnote to my own query correcting one error of transmission and (I hope) clarifying some of my intentions.
The error: in my query as it came back to me from the list server there is a string of 11 or so question marks following my citation of the Komachi poem in romaji. In the message as input by me (under Win98-J), this string consisted of kanji and kana for the first half of Komachi's poem, which I meant to cite partly just because it can be done (in principle, anyway). Evidently, I chose the wrong one of several options for encoding when I sent the message.
I menton this only because I'm afraid that all those question
marks may have given the impression that I find the whole of Komachi's
poem very perplexing, which I do not. Otherwise, a little context
to clarify what I was perplexed by:
My initial response to this poem was to accept the "negative"construal ("iro miede...") --- which as I noted has traditionally been the predominant reading --- on the grounds that the poem seems designed to focus on the irony of the fading of an invisible flower, etc. (The case for this reading is presented convincingly enough in Takeoka's review of the tradition of debate on "miete" vs. "miede.") In the early 1970's (roughly) there was a renewal of this debate among Heian lit. scholars, and it happened that the one participant in this debate who argued in favor of the affirmative reading ("miete") was the professor (Nomura Seiichi) who, a decade or so later, was directing my research in Japan. So I invested a certain amount of effort in trying to understand the terms of the debate, and came away convinced that the affirmative reading was indeed more persuasive because it entails the somewhat finer irony of the all too visible fading of what should be an invisible flower.
My surprise, and perplexity, was thus a response to the suggestion (which I first encountered in the SNKBT KKS, in 1989) that, in effect, this debate need never have taken place since the crux was not a crux but a kakekotoba.
As far as I am aware, none of the participants in recent (post-Muromachi, say) debates over this poem, indeed no commentator since c. the 14th century, has (until Arai & Kojima's edition) ever proposed the suggestion that this was a kakekotoba. Which compounds my surprise. (Arai has spent much of his career studying pre-Edo commentaries on KKS so I assume he encountered this suggestion himself in the 14th c. Reizei family commentary which seems to be the unique <? precedent for this reading.)
I don't mean to back-pedal on this question, but perhaps it was a tactical error to mention evidence (e.g. from Kanera) that the supposition of a "te-/de" kakekotoba in Komachi's poem is not a fluke. (I know of only 2 or 3 other cases where similar claims for an affirmative / negative kakekotoba are made and all are much later.) Are there any other waka from the 9th c. or before for which it is arguable that this particular (or even an analogous) kakekotoba is to be accepted?
About intentions again, with certain specific kinds of exceptions (usually involving the editorial intentions of Tsurayuki et al. in arranging the poems in KKS as they did) medieval KKS commentators rarely if ever speculate on the putative intentions of authors of poems.
Re: "intellectual colonialism," I'm probably misunderstanding
here but not sure I see the point. If you wish to use the word
"colonize" in that sense, and with respect for its etymological
co-origins with the word "culture," you could certainly
argue that any effort to understand a culture other than "my
own" (although I don't believe the notion of "my own"
much less "our own" culture makes all that much sense,
especially not as opposed to "their own") is a form
of colonialism, and so much the worse for that word's force.
Don't mean to sound antagonistic, just hoping to provoke clarification.
I am probably missing something. Surely a kakekotoba involves a sequential swivel; it ends one phrase (joshi) with one significance, and then kicks off another phrase within which it has a second meaning. No one is suggesting that there is a phrase ending: miete and then beginning: de utsurou... are they?
I too thought a kakekotoba involved a sequential swivel, to
borrow Noel's
pithy phrase. Could we take a few steps back and define a kakekotoba?
Miete/miede seems close to an ordinary pun. But then perhaps the
definition of a kakekotoba has itself undergone evolution? I wonder
what the history of the term itself is. Perhaps the definition
of kakekotoba has changed since Japanese critics became aware
of its likenesses and differences to wordplay in Western languages,
i.e. since the Meiji period, but especially since the post-war
period...?
Date: Jan 22 2000
From: Lewis Cook
I am probably missing something. Surely a kakekotoba
involves a sequential swivel; it ends one phrase (joshi)
with one significance, and then kicks off another
phrase within which it has a second meaning. No
one is suggesting that there is a phrase ending:
miete and then beginning: de utsurou... are they?
Noel Pinnington
No, the suggestion is just that "miete" and "miede"
are kakekotoba calling for two antithetical but simultaneous readings
of the same grammatical phrase (no swivel). There is a concise
and, I think, reliable entry ("Kakekotoba") in the Waka
Daijiten (Meiji Shoin revised ed., 1986) which addresses this
question, stating that the word "kakekotoba" has historically
(meaning, here, not just post-Edo or post-Meiji) been used to
refer both to cases of 'one word' (not the "same"word
but two homophonous words or two strings of identical phonemes
not counted, for etymological reasons or whatever, as the same
lexeme, e.g. nagame as "endless rain" and nagame
as "gaze") used at once -- within roughly the same grammatical
construction, that is -- in two different senses (as in the Komachi
poem KKS No. 113 referred to by Rein) -- puns, in short -- and
also to cases of "pivot"words, as in SKKS No.
952 (izuku ni ka koyoi wa yado wo kari-goromo himo yuu-gure no
mine no arashi ni) where "kari" and "yuu"are
both puns and pivot words (syllepses or zeugmas), i.e. involve
"sequential swivels"and not simply (non-sequential)
superimpositions of disparate lexemes. I appreciate the temptation
to assume that the latter cases might more properly be called
kakekotoba but am not aware of any historical basis for such a
distinction. (Medieval commentaries on waka and renga regularly
use the word "kakekotoba" in both of these senses.)
As far as I can see, the judgment that iro miete/de is a kakekotoba is irreducibly arbitrary. Since the writing system requires no distinction between the two, and since no canonical authority (I assume) recognizes the te/de ambiguity as a device for producing kakekotoba, the judgment that miete/det IS a kakekotoba can appeal only to authorial intention, which is unknowable, or to kokubungaku fashion, which is mutable.
Date: Jan 23 2000
From: Robert E. Morrell
Dear Professor Cook,
You wrote:
[section omitted]
About intentions again, with certain specific kinds of exceptions (usually involving the editorial intentions of Tsurayuki et al. in arranging the poems in KKS as they did) medieval KKS commentators rarely if ever speculate on the putative intentions of authors of poems.
Well, they may rarely overtly "if ever [have] speculate[d] on the putative intentions of authors of poems." But they must surely have covertly done so. When poets used such devices as honkadori, would they not necessarily have had to speculate on the intentions of the writer of the original phrase? And when, over time, a dramatic shift occurred in poetic goals and techniques -- say, from the KKS kokoro/kotoba balance (sama?) to the SKKS's kotoba sukunaki kokoro amashi -- could this have ever happened if the community of poets was not continually speculating on and judging the merits of earlier writing, and then looking for ways to infuse new life into old words? True, they rarely stated these judgments in any formal way, but they must have made them.
Re: "intellectual colonialism," I'm probably misunderstanding here but not sure I see the point. If you wish to use the word "colonize" in that sense, and with respect for its etymological co-origins with the word "culture," you could certainly argue that any effort to understand a culture other than "my own" (although I don't believe the notion of "my own" much less "our own" culture makes all that much sense, especially not as opposed to "their own") is a form of colonialism, and so much the worse for that word's force. Don't mean to sound antagonistic, just hoping to provoke clarification.
I really only had a couple of simple observations to make,
and I made them rather clumsily. The first is that since most
of us are not Heian Japanese, one of the best places to look for
clues might be in their underlying philosophies rather than in
methods of analysis derived from Western literature and its underlying
presuppositions ("intellectual colonialism"). . . My
second point was just that since there will always be differences
of interpretation, we will always need footnotes -- however they
may be disfavor these days, especially among poets.
Sorry for the confusion.
Bob
Rein Raud writes:
Why do we say that this instance (te ~de) of kakekotoba is extreme, whereas the standard "pine"/"wait" type is not? If any "logic" is valid in Komachi's world, it is certainly not Aristotelian, and there is no need to oppose affirmation/negation. For me, the particular charm of this poem by Komachi has always been in the fact that it posits two incompatible views of "the human heart" at once, and both are correct. Ant this precisely is the message of the poem. It is also corroborated by the double meaning of "iro" (colour/passion) and what "fade away" would mean in the context - it is one thing if the passion fades away, another if it applies to the whole heart-flower. As it is well known, the last two lines of Komachi's more chrestomatic KKS 113 do not even break into a single grammatical sentence, which also demonstrates that she is able to use the very fact that linguistic expressions has it limits in order to refer to the "extremeness" of the emotional reality beyond.
I'm sorry it's taken me so long to reply to this (and to several replies from other list members, for which I am grateful --- I will try to get to these soon).
I don't have any very good answers to your question -- why this particular kakekotoba seems 'extreme' (maybe not the best choice of word) and perhaps, as Richard Bowring suggests, I am worrying it needlessly.
My rough sense of what gives a kakekotoba (or a pun) its peculiar force is that two words (matsu as "wait" and as "pine" say) which we know perfectly well are related only by coincidence (a meaningless accident of homophony) are joined in a relation that catches us by surprise, creates meaning where there was none, is somehow both more and less 'natural' than what our common sense of language leads us to expect. (Not a very elegant statement but will have to do for now.)
One reason I suppose this effect to be more intense in the pair "-te / -de" is that (unlike matsu as 'pine" and as "wait,") the superimposed words are not just different (semantically unrelated) lexemes but are logically antithetical. I don't mean to quibble, since I am very much in agreement with your account of the charm of Komachi's poem, but I don't think one has to be an Aristotelian to oppose affirmation to negation, and you seem to agree in accepting that this kakekotoba posits mutually "incompatible views." Much of Komachi's poetry builds on ironic mediations of oppositional terms, and of course on elaborate kakekotoba, and I did not mean to suggest that I am skeptical of the punning interpretation of the 'iro miete' poem, only surprised, and perplexed partly because I know of no precedent for making of this "-te/-de" pair a kakekotoba and only a handful of later occurrences. (There is a similar crux for "omohoete/-de" in KKS 351. As far as I know no one has ever proposed resolving this into a kakekotoba, though the context is somewhat less hospitable to such a reading than in the case of Komachi's poem.) But I wonder how many of Komachi's kakekotoba have well-attested precedents?
Royall Tyler writes:
As far as I can see, the judgment that iro miete/de is a kakekotoba is irreducibly arbitrary. Since the writing system requires no distinction between the two, and since no canonical authority (I assume) recognizes the te/de ambiguity as a device for producing kakekotoba, the judgment that miete/det IS a kakekotoba can appeal only to authorial intention, which is unknowable, or to kokubungaku fashion, which is mutable.
Thanks very much for this response, Royall. (Again I apologize for belated replies to this and other comments.)
You are quite right to suggest that judgements about kakekotoba
are arbitrary; we seem bound either to appeal to canonical authorities,
if these exist, or fall prey to the whims of fashion (presiding
interpretive communities, etc.) But canonical authorities are
only immutable while they last ('fashion' in very slow motion?);
the foundations of their authority is ultimately no less arbitrary,
after all.
(One laborious aside, pertinent here because it involves a similarly
"antithetical" reading of the 'same' word, this one
thoroughly canonical in its time: throughout the medieval era,
Nijo Family / School commentators on KKS maintained that the phrase
in the Kana Preface "ima wa, fujinoyama mo keburi tatazu
nari" was to be read "even now, the smoke of Mt. Fuji
has not ceased [rising].' They accomplished this feat by taking
the word or string of phonemes "tatazu" to represent
the negative form of the verb "tatsu" meaning to "cut
off." This doesn't work too smoothly from a grammatical point
of view, but it had the virtue of establishing a clear differendum
vis a vis the Reizei Family reading, i.e. that "tatazu"
means "no longer rises." I doubt if anyone since the
19th century has accepted the Nijo School's gloss on this, but
it is noteworthy that according to To no Tsuneyori this "tatazu"
was taken to be an 'antithetical' kakekotoba, to be understood
both as "no longer rises" and as "has not ceased
[to rise]" by cognoscenti of Tameie's generation, before
the schism between the Nijo and Reizei families, that is.)
Isn't the point though just that questions about whether a certain
word in a given poem or text "is" a kakekotoba are not
questions about matters of fact (nor certainly about such inscrutables
as authorial intentions) but about more or less persuasive or
acceptable interpretations? When we claim to 'see' a pun on matsu
as "pine" and "wait" in any number of passages
in, say, Matsukaze, we are appealing to (pretty overwhelming)
evidence from context and from precedent and convention, but not
to anything quite like a grammatical rule, and not necessarily
to an explicit canonical ruling. The problem with Komachi's
"iro miete/de" poem is that there is a shortage of context
and a seeming absence of precedent, and the usual suspects (Teika,
Kensho, and heirs) were (for the most part) silent on this crux.
I don't think this means we need to give up in despair, though.
There are a number of other poems in KKS which include words which may or may not be (taken as) kakekotoba, for which there are apparently no precedents, and which have thus provoked not futile but I think fruitful debate. E.g., "munashiki kara no" in No. 571 --- where kara seems to demand to be read as "corpse" as well as "because." Or gani in 349, presumably a devious pun on the archaic suffix gani /gane expressing a wish, and on the name of the genre the poem was required by decorum to fit. Or the se ni in 990 (which virtually every medieval, Edo and modern commentary, with 2 or 3 exceptions, accepts as a pun on zeni [money], despite the complete lack of precedent and its manifest indecorousness). Or --- a good example of canonical authorities in disagreement --- the ushi in Sosei's KKS 803 ("aki no ta no ine chou koto mo kakenaku ni nani wo ushi to ka hito no karuramu"), which Kensho of the Rokujo Family wants to read not only as "displeasing" but as "ox," to which Teika objects "amari ni ya?" (in Kenchu Mikkan). Almost every other substantive in this poem (KKS 803) is a kakekotoba, if you want to see it as such; Sosei even works in a reflexive pun on the word kake-koto[ba]. Who is to say, or rather how are we to decide, whether this ushi is not also an ox?
I don't think (to repeat) the fact that we cannot say with certainty makes this question any less worth pursuing. Most of what we do by way of reading & writing is perforce interpretation rather than asseveration, no?
(Forgive me for going on and on. I admit I find these problems irresistible.)
A wonderful reply, Lewis, from a true scholar. As for me, I didn't mean that the te/de question is uninteresting. I only meant that it is ultimately undecidable according to any principle outside one's own judgment. Is that not so? To read miete and miede simultaneously certainly gives the poem a new zing. Presumably anyone for the last thousand years has been free to do that privately. However, people want rules (as you said, most Genji readers want answers, not questions), and so there's a lot to be said for AUTHORIZING them to do it, ex cathedra SNKBT; except that some may then be disappointed because they really did prefer miede alone, and now they have to take both coz SNKBT said so.
It remains true, anyway, that double meanings will inevitably crop up from time to time (in any language, but certainly in J) as an effect of chance. The poet might even edit out such a chance double meaning if s/he noticed it. Sometimes it may be so unlikely that no one could take it seriously, but as the gap between far-out accident and obvious intention narrows, more and more people will start making the connection, more or less outrageously. It's a question of interpretation, it is indeed; and I suppose of measuring the degree of improbability.
Royall Tyler
Date: 26 Jan 2000
From: Lewis Cook
I hope I'm not violating netiquette by persisting with this thread, but I'd like to reply to some outstanding messages (and then would be happy to pursue the details off-list with anyone interested).
Paul Atkins writes:
I haven't seen the book by Yamashita but from your account I'll assume you mean that Yamashita notes the ambiguity --- the possibility of two alternative readings --- but does not specifically propose the _kakekotoba_ interpretation here. Right? And I'll also assume that the recently current interpretation you refer to is just that there is an ambiguity re: te / de in KKS 797.
One very relevant point which both Richard Bowring and Royall Tyler remarked in passing --- the ambiguity itself is not new. The earliest extant manuscripts of KKS (as of most everything else in 'kanabun') do not contain diacritical marks (seiten); if they do these are usually later additions. The ambiguity is there from the start, as an effect of the writing system. (I'd imagine there were oral traditions of recitation of this poem which would have had to resolve the ambiguity, but we don't have any evidence for these of course.)
There is a good historical explanation for the renewal of attention to this ambiguity between, say, 1958, date of Saeki's NKBT KKS, and 1989, when Arai & Kojima's SNKBT KKS was published. The intervening decades saw an abrupt revival of interest in "medieval learning," including pre-Kokugaku commentary traditions on the canon. (One striking example: all the 'discoveries,' during those decades, of allusions in noh plays by Zeami et al. to 13th and 14th c. esoteric commentaries on KKS and IseMg, facilitated by the publication of the first typographic editions of some of these commentaries, which had been almost completely ignored for the preceding 2 centuries, by Katagiri Yoichi and others in the 1970's.) This revival was in part, I think, a by-product of the post-war critique of Kokugaku ideology, but the effects hadn't really set in by 1958; at that time, very few KKS scholars paid much attention to pre-Kokugaku or pre-Keichu~ scholarship on KKS.
So it's not all that surprising that Saeki (who in any case was a grammarian) didn't bother to mention that certain medievals had been troubled by the "te / de" ambiguity in Komachi's poem.
That said, most of the medieval exegetes who comment on Komachi's "iro miete" poem do prefer the "-de" reading, though many of them note the alternative "-te" reading, attributing it to "other schools." (One of those 'others' appears to have been the Asukai Family: Masachika's Kokin Eigasho~ [1496] comes out in favor of the "-te" reading.) And I have to confess, now, that the 14th c. Reizei Family commentary (the Tamesuke KKSChu~) I've referred to more than once I think as the earliest source of the kakekotoba reading actually dates from the last decade of the 13th c. and unambiguously prefers the "-de" reading. (So much for working from memory.)
Kanera's KKS Do~myo~sho~ (1476) seems to endorse the kakekotoba reading, but the locus classicus is apparently the To~ Family transmission, which (reportedly) asserts that "both readings are to be accepted at once." (Earliest written evidence for this from c. 1470?)
My point is just that this ambiguity is not news, and that I take a certain assurance from the evidence that the difficulties I experience in deciding what a poem such as Komachi's may have meant are least one thing I have in common with ancient adepts who knew a lot more about this business than I do.
This has been my impression as well, but it doesn't seem to stand up to scrutiny. There is a brief (very incomplete) table of 'standard' kakekotoba in the appendices to the _Fukutake Kogo Jiten_. Roughly half do belong to different grammatical classes, but I don't think this is enough to make this count as a rule. (Didn't someone, a Swedish scholar, write a dissertation in English a decade or two ago on kakekotoba in KKS?)
The contrastive "wa" is indeed the nexus on which Takeoka hangs his grammatical argument in favor of the "miede" reading. (Let me reply off-list and try to argue in favor of the "miete" reading. Which I think is a prerequisite to entertaining the kakekotoba reading. More "zing," as Royall says. Isn't zing at least part of what we are after?)
Lewis Cook
Date: Jan 28 2000
From: Stephen Miller
Lewis,
Which kogo jiten is the "Fukutake Kogo Jiten?" Who is the publisher? I have six kogo jiten in front of me and none of them is edited or compiled by anyone named Fukutake.... Just wondering. I'd like to look at the list of kakekotoba. This discussion has been fascinating. I can only hope it continues for some time. Thanks for initiating it, Lewis. Stephen
Date: Jan 28 2000
From: Janine Beichman
As far as I am concerned you are certainly not violating netiquette
by persisting with this thread. In fact it is getting more and
more interesting. Just a plea that, just when it has gotten so
interesting, that it not suddenly go underground/off-line. I look
forward to further arguments for 'miete' as opposed to 'miede'.
And to have some idea why medieval commentaries on Ise and KKS
were ignored for so long was a gem....Please continue!
Date: Jan 28 2000
From: Sonja Arntzen
I would like to echo Janine's comment. Even though I have not participated directly in the discussion, I have enjoyed following it very much. [Continues with self-introduction /ed]
Date: Jan 28 2000
From: Richard Bowring
A few short comments here.
Certainly we do not know much about oral recitation but we do know that it was taken extremely seriously. Murasaki herself comments in the nikki that you had to 'watch out' when reciting in front of a master like Kinto: Shijou no Dainagon ni sashiidemu hodo, uta oba sarumono nite, kowazukai youi arubeshi. I had assumed this was a matter of intonation and pattern but perhaps in extreme cases 'misreading' diacritics was another hazard for the unwary. The ambiguities we are talking of here were surely treasured, since they actually fuelled the proliferation of secret traditions and special readings: grist for the mills of medieval commentary.
Not quite right. I assume you are referring to Lindberg-Wada 1983, but this was on the use of KKS as a source of allusion in GM and dealt with hikiuta etc.
Date: Jan 28 2000
From: Michael Watson
Listening fascinated on the sidelines, I can guess the answer
to Stephen
Miller's query:
Fukutake is the publisher, and the title is also part of the
title. The
authors are Inoue Muneo and Nakamura Yukihiro. Webcat gives
Fukutake kogo jiten .... Fukutake shoten, 1988. (title in Japanese)
I see that Richard Bowring has already replied to the point
I'll give the biblio. anyway --from the biblio. of RB's CUP book on Genji:
G. Lindberg-Wada, Poetic Allusion: Some Aspects of the Role
of Kokin Wakashu as a Source of Poetic Allusion in Genji Monogatari,
Japanological Studies 4, Stockholm: Univ. of Stockholm.
Date: Jan 28 2000
From: Lewis Cook
I was looking forward to replying to a question I can answer with some confidence, for once, and see that our webmaster has gotten there first. ;-} I would like to hear which of your half dozen kogojiten you've found most useful. To Michael's bibliographical data, let me add that the Fukutake Kogojiten is, in my experience, the most informative of the many short kogojiten on the market, for medieval (post-Heian, pre-Edo) Japanese. I'm told that Inoue Muneo (titular editor for pre-Edo entries) actually had a hand in the editng. The appendices are invaluable.
Thanks very much, Stephen. I've learned much from the many generous and learned responses (some off-list) received, and hope there will be more, perhaps on related topics.
Date: Jan 29 2000
From: Lawrence Marceau
Fukutake Shoten changed their name a couple of years ago to
"Benesse
Kooporeeshon" ("benesse" means "prosperity,
I believe). They have
reissued Inoue's and Nakamura's kogo jiten under a new name, Benesse
kogo jiten as well. TRC's
database provides the following information
(apologies to those who lack Japanese encoding for their e-mail):
[...] ISBN 4-8288-0442-0 [--> Japanese notes]
With regard to the "te/de" interpretation possibilities, it is certainly intriguing to consider affirmation and negation as two alternative readings to the Komachi (and other) poems. I don't believe that we are displaying a system of logic alien to Heian court sensibilities, though, when our temptation is to exclude opposites as equally acceptable alternatives.
I have in mind one example of an instance in which it seems that the "unacceptable" alternative is now the "standard", while the "correct" (in my opinion, at least) interpretation seems to be secondary. This is in the famous Iroha poem/song. The last part goes as follows:
Asaki yume miji / Wei mo sezu
Shallow dreams unseen / Nor does intoxication occur
In the reading made popular by the manga series (and now feature film), the first line is read as "Asaki yume mishi" with a possible English equivalent of "Having seen a shallow dream". As far as I know, this reading does not fit with the last line, which includes the particle "mo" ("also"), and is also unambiguously in the negative. If the poet has "crossed the deep mountains of existence and non-existence" it does not seem likely that s/he would still be dreaming, especially if no longer in a deluded state.
One last point -- it does not seem that the te/de dichotomy fits the sense of "kakekotoba" as I am aware of it. The closest example I can find of a verb ending or particle serving in a kakekotoba function would seem to be in an example such as "omo-hi" ("thinking of") or "ko-hi" ("longing for") being alluded to with the noun "hi" ("fire"). Metaphor here is the key, I think. If the "mie-te/de" were somehow linked to "te" ("hand(s)") or "de" ("?"), then this would fall within the standard (generally agreed upon) realm of kakekotoba usage. If you have the Shin Meikai kogo jiten (Kindaichi Haruhiko ed., Sanseido), there is a handy list of kakekotoba on page 1230. Te/de are not listed.
Date: Jan 29 2000
From: William Bodiford
Dear Lewis Cook:
Thank you for your "laborious aside." My seminar was just discussing how Japanese Buddhists during the medieval period taught rather forced interpretations of Chinese texts. It is extremely relevant to know that other Japanese forced interpretations on Japanese texts as well.
Date: Jan 29 2000
From: Rein Raud
In his first comment, Richard Bowring raised a very important point also echoed by Royall Tyler: overinterpretation and authorial intent. I guess in our age we are not supposed to appeal to auhtorial intent any longer, i.e. the kokoro of a waka is, for our purposes, only the uta no kokoro. But can we speak about over-interpretation when this was the norm in the Confucian treatment of Shijing on the continent, and, we have to suppose, in the circles of Heian Confucian scholars as well? Especially during Saga's times? Perhaps the linguistic awareness characteristic of Heian waka is a result of the combination of this attitude and the semi-playful deconstruction of each others' poems in poetic exchanges. On the other hand, Richard's example shows that "overinterpretation" could be used as a critical argument in poetic debates - thus, probably, in our current debates on waka as well. If we then let the allowed limits of interpretation to be fixed by tradition, won't we then give up our inalienable rights as readers? I have to say I have always liked interesting readings (old and new) that are backed by textual material only, and not the commentary tradition - but it is certainly true that in this way it is always tempting to "over-interpret". It even seems that at the time of the practice itself the allowed limits of interpretation were much broader than in later times, and therefore closer to the criteria of our age than f.e.x the Edo period.
Date: Jan 31 2000
From: Lewis Cook
Thanks very much for these (and previous) comments. Intonation as well as possible ambiguities were certainly issues also on formal occasions, utaawase and kakai (for which protocols for reading aloud are taken up in some detail in Fukurozoushi. The utaawase you referred to previously (discussed in Brower & Miner, pp. 249 ff) is interesting because it involved a misreading of near homophones (tazu and tatsu) where, evidently, no kakekotoba was intended. For anyone really curious, there is a wealth of information on diacritical marks in late Heian and medieval texts of KKS in the multi-volume study by Akinaga Kazue, Kokinwakashuu Seiten-bon no Kenkyuu, which might give some clue to how KKS 797 was performed (I don't have any but the last or most recent volume of this series myself and the sole mention of this poem therein refers to pitch levels.)
A tangent, but I wonder if the "iro miete/de" poem is never cited in the many Komachi-related noh plays, in which case the ambiguity would have to have been dealt with in performance. (I don't see an entry for this poem in the index to the Youkyoku 250-ban shuu, but have no idea how comprehensive that is.)
I appreciate this correction. Perhaps a dissertation is waiting to be written on kakekotoba in KKS.
Date: Feb 4 2000
From: Noel Pinnington
The -te -de possibility must have been one to which KKS poets were sensitive. An interesting case of a different kind of play on te and de occurs in 833:
nete mo miyu, nede mo miekeri, ookata wa, utsusemi no yo zo, yume ni wa ari sleeping
I see him, awake and still I see that vision before me, ah truly do they say this locust shell world is a dream (using Laurel Rodd's translation).
I remember when I first saw it wondering how 'nete' and 'nede' were distinguished. The announcement that we can never know authorial intent always seems a bit extreme to me. At least we can ask whether what we observe as a possible poetic device contributes or is tangential to other aspects of our reading of a poem. In that sense the first two phrases can be read as deliberate play on -te and -de. The general sense of the poem, that waking and sleeping, whilst apparently quite antithetical are actually in another sense not different, is paralleled in -te and -de (in meaning opposed, but orthographically similar / the same).
Date: Feb 4 2000
From: Lewis Cook
This poem, and Noel's comments, raise several very interesting questions. I wish I could take the time to reply with care; instead a few very hasty remarks --
I quite agree that the play on -te / -de here seems calculated to work against the common-sensical antithesis of waking and sleeping, (This in itself is familiar enough, and seems to have been a conventional trope of classical waka from at least early Heian times; calls to mind the formulaic "yume-utsutsu" and cliches such as "okite-mo nete-mo," "nete-mo samete-mo," etc., which occur frequently in Heian and medieval waka, and more generally phrases in the form of "whether X or not X" -- typically, I suppose, with hyperbolic or ironic force.) But in this case the effect is, as Noel remarks, intensified by the fortuitous graphic similarity of 'nete' and 'nede' (especially without diacritics), and leads me to wonder whether the kakekotoba reading of 'miete / miede' isn't a logical extension, a further degree of compression, of this orthographic potential for combining opposite meanings in the 'same' word. (In an off-list message, Kendon Stubbs -- I hope he doesn't mind my citing him here -- remarked the relevance of Empson's discussion, in Seven Types of Ambiguity, of the 7th [or 'extreme'] cases in which the 'same' word carries antithetical meanings.)
One aside: I mentioned before (in passing) that Teika (and Kenchu) were silent on Komachi's (KKS 797) poem, meaning that neither of them seems to have left comments on the miete/de crux. I'd forgotten that Teika did compose at least one poem taking Komachi's as a honka, and in this case clearly prefers the miede reading. (Shuuigusou 305: "iro miede haru ni utsurohu kokoro kana / yami wa ayanaki ume no niohi ni"). Not decisive evidence that he didn't accept the alternative or even the kakekotoba reading, though.
A further aside: Noel writes, quoting again, "the announcement that we can never know authorial intent always seems a bit extreme..." (I'm tempted to respond "yes and no," miete / miede). Rein Raud's most recent post also raises this spectre, as have several earlier messages. I don't think even Brooks & Warren (so frequently cited without having been read) can be credited with this extremity. The argument, at least as far as poetry goes, is not that we cannot fathom authorial intent, but that everything we need to know about intentions is just there in the words of the poem, or should be. (Though kotobagaki or headnotes to poems, and other contexts, certainly complicate the issue, especialy since they were often ignored or altered by editors of chokusenshuu, for example.) If there are ambiguities, they are there because the author put them there or because the language (or script) affords (or perhaps imposes) them. I don't think it's reasonable to imagine that Komachi wasn't aware of the manifest ambiguity of iro miete, Better suppose she left it to the reader to decide.
The notion that there is something like a singular authorial intention underlying or preceding the words of a poem (which would be the verbal excresence thereof) seems to me closely analogous to the phlogiston theory of combustion.
[later correction ]
-- that should be "Kenshou" not "Kenchu" --
(sorry if this seems hyperfastidious.)
[Lewis Cook also wrote off-list to correct his attribution to Brooks & Warren of Beardsley & Wimsatt's "Intentional Fallacy." /ed]
Date: Feb 6 2000
From: Janine Beichman
That comparison of the notion of a single authorial intent to the phlogiston theory of combustion is interesting. I looked up 'phlogiston' but can't make the jump to the theory of combustion based on it --could you explain?
Also, the idea that ambiguities are there because the author put them there--what about the poem written in a white heat? Surely there are times when things get in without the author having time or space to think about them? And maybe that's how it was for Komachi when she wrote this poem. I just don't see how we can know what she meant unless we have evidence from outside the poem, which we don't. Which leaves the evidence for the ambiguity as coming from what the language/script affords/imposes, to borrow your words.
Date: Feb 7 2000
From: Lawrence Marceau
I believe it was noted earlier on this list, but the early modern philologists (wagakusha), including Keichu, Kamo no Mabuchi, and Motoori Norinaga, seem to be unanimous in reading KKS 797 (Komachi) "miete" as "miede."
Keichu, in Kokin yozai sho (*) is most specific, and even advances the "de" reading, only to reject it: "While blossoms, as blossoms, have colors that change visibly, only the blossom of the hearts of people in this world change without being visible--this is how it is read. While the poem says the greater world, it refers to people. There is a theory that takes the "te" in the first line as being clear (=unvoiced), meaning, who can see the color of the blossom of the heart? The poem could be read in this way, but it is more honest (=sunao) to read it as voiced (=nigoru)."
Mabuchi, in Kokin wakashu uchigiki (*) follows Keichu, it seems. He says in the headnote: "It is wrong (=waroshi) to read 'Iro miete' in the clear/unvoiced pronunciation."
Norinaga, in Kokin wakashu tookagami (*) translates into funky contemporary Japanese prose: "Kusa ya konohana wa / Iro ga aru yue ni utsurou ja ga / Iro wa aru to mo miezu ni / Utsurikawaru mono wa / Yo no naka no hito no hanabanashii kokoro no hana de (sa) gozarimasu wai!"
He goes on to say that "iro miede" means the absence of color, and that Keichu's comment on the first two lines is no good (chuu waroshi). Norinaga's observation comes from his particular understanding of the nature of the human heart as something beyond physical appearance, and sensual comprehension. (I like his modern translations more than his interpretations in many cases, nonetheless.)
It seems that, by the late 17th century, the philological approach dictated against incorporating multiple readings into a poem, but rather, working through linguistic and historical methodology, to "pin down" a single, authoritative reading.
Date: Feb 8 2000
From: Lewis Cook
Thanks very much, Janine, for your comments.
I had in mind here (intended) only to suggest that the notion that there is some mysterious hypothetical mental element ("singular [not single] authorial intention") to which the meaning or reading of a given poem should be held accountable, despite its (intention's) unfathomability, strikes me as about as useless as the notion that we (medieval alchemists) must postulate the existence of some mysterious element called phlogiston to account for the occurrence of fire.
I certainly don't mean to deny that poets must have various things in mind before and while composing poems, and it is quite possible that Komachi unambiguously meant (intended) her poem to be read iro miede.... But I think it is undeniable that Komachi's intentions are strictly irrelevant, in this case (a case, that is, in which we have no further context, e.g. a Komachi diary, an epigraph, etc., against which to weigh our judgements --- though even in cases where we do have further context, these are not revelations of ultimate intentions [hidden mental processes] but simply more marks on a page to be interpreted), because we know that the rule is that a poem (a 9th c. waka, anyway) is more or less required to mean as much as it seems to say, and I think we can also assume that Komachi knew the rules (and, to repeat, was aware of the ambiguities) when she published this poem. (Things could have been otherwise -- if it had been submitted in an utaawase, she might have been expected [not likely in the 9th c. but certainly later, as long as we're just speculating] to supply an account of her intentions, or rather of how she meant her words to be read.) There are exceptions to this rule, in which poets were interrogated on the spot about their intentions, and not only in utaawase. One example: IseMg 101, Narihira's poem (saku hana no shita ni kakuru hito wa...). The poem (especially if you accept the pun arishi = the Ari[hara] clan as opposed to the Fujiwara) is pretty clearly a cynical comment on the political hegemony of the Fujiwara. But are we expected to believe Nairhira's account of his intentions when he explains this poem away? (A cheap shot, granted, but I don't think any author's account of intentions is quite worth a poem.)
Also, the idea that ambiguities are there because the author put them there--what about the poem written in a white heat? Surely there are times when things get in without the author having time or space to think about them?
Yes, exactly, and all the more reason not to get hung up on abstractions about intentions. Examples, please. For white heat, how about Saikaku's final 'Ooyakazu'? Wasn't that 23,000 or so haikai verses in 24 hours? (Chris Drake, give us the facts, please?) Even if the scribes had been able to record this feat, what could we have expected to infer about Saikaku's intentions? I think we'd have to allow at least a little room for the unconscious, here (and once we admit that, most of our talk about intentions is compromised, no?)
as of 2000/03/01, minor revisions 2001/01/27. Changed to ISO
format 2001/03/03
edited for pmjs website by Michael Watson <watson@k.meijigakuin.ac.jp>