Simply Haiku: A Quarterly Journal of Japanese Short Form Poetry
Contents Archives About Simply Haiku Submissions Search
Summer 2006, vol 4 no 2

The Formation of Allusive Resilience in Waka
and Its Relevance to Meiji Shintaishi

Dean A. Brink
Saint Martin's University

 

 …[It] is through symmetry that rectilinear systems limit repetition,
 preventing infinite progression and maintaining the organic domination
 of a central point with radiating lines, as in reflected or
 star-shaped figures. It is free action, however, which by its essence
 unleashes the power of repetition as a machinic force that multiplies
 its effect and pursues an infinite movement. Free action proceeds by
 disjunction and decentering, or at least by peripheral movement:
 disjointed polythetism instead of symmetrical antithetism. Traits of
 expression describing a smooth space and connecting with a matter-flow
 thus should not be confused with striae that convert space and make it
 a form of expression that grids and organizes matter.[1]
 Gilles Deleuze and Felix Guattari
 
 Although we might think of pre-Meiji waka[2] as limited and a
 function of a closed world, these waka may perhaps better be
 characterized as intertextually porous (transparent), open texts
 coterminous with accepted canons of waka (which varied by school). In
 this way waka were meant to be read in dialogue with a larger corpus
 in effect so as to disseminate subject-centered concentrations of
 meaning in any given poem at hand. The following paper examines how
 the waka poetics developed by Ki no Tsurayuki in its inception in the
 early tenth century intended to realize a degree of cultural
 independence vis-à-vis Chinese poetry and culture, and how this
 ideologically imbedded poetics was what mid-Meiji founders of
 new-style poetry (shintaishi) felt most compelled to displace as they
 attempted to build up a national poetry comparable to the epic and
 narrative poetry of the Western Great Powers.
 

 It is in the founding collection of waka, the Kokin(waka)shū,
 that one can see the convention of simply alluding to Chinese and
 Japanese mytho-historical prose and poetry being overtaken by a
 broader intertextuality. The Kokinshū (with its “Kana Preface”) opened
 the poetic canon to the depiction of the interaction of self and
 scene, emotions and others in ways that superseded the dominant
 Confucian approach to poetry (as explicated in the Great Preface) and
 entertained its own ranges of topics (a rubric which only skims the
 surface of conventions for the placement of imagery). Ki no
 Tsurayuki’s “Kana Preface,” as will be explored below, puts forward
 its own conventional formal expectations that reinforced the
 associative sequencing in the anthology, and in practice founding a
 new poetic lexicon and canon on the precedent set by the Man’yōshū
 (759).
 

 Especially the more contemporary waka included in the Kokinshū
 were entrenched in both the conventions and language drawn from other
 waka (equivalent to tanka in my usage) and chōka (extended waka) of
 the Man’yōshū, as well as Chinese poetry—but the allusive methods
 themselves differ from that found in Chinese poetry. Even when not
 explicitly alluding to a precedent text, so many phrases are recycled
 into poems that nearly any poem can be found to allude to a range of
 poems or build on connotations found in earlier examples. Moreover, in
 any given part of a sequence, say on spring, one will come to hold
 certain expectations, for instance, that plum blossoms will appear
 fairly early on. Poems situated in summer will likely have a cuckoo
 (hototogisu). Waka on love may be expected to include, depending on
 the stage in the affair, love before being seen, love compared to a
 river, questioning what occurred in a dream and what in reality, or
 “the autumn of the heart.”[3] This arrangement was invented in the
 Kokinshū and supported by the “Kana Preface.” Though the topics are
 arranged in books primarily on the seasons and love, many of the
 seasonal poems also deal with romance, and seasonally suggestive
 imagery is commonplace throughout the anthology.
 

 A hermeneutic reading (aimed at eliciting an interpretation “out
 of” a self-contained text) of waka, with its vitally intertextual
 poetics, is thus not effective as a mode of reading. Waka do not
 follow, at least in this intertextual respect, logocentric (nor
 metaphor-oriented) paradigms that closed, internally structured
 (autonomous) texts foreground, mirroring the “presence” of the
 reader.[4] English poetry (especially those Victorian models used by
 shintaishi poets) tends toward internal, intratextual coherence, with
 the figurative and lexical ambiguities of words comprising tensions
 that constitute the lyrical saturation or narrative continuity within
 a poem. Some form of hermeneutical reading is a matter of course, and
 interpretation focuses primarily on the poem and only secondarily on
 allusive language. This balance differs from waka, where ambiguities
 and metaphorical language exist, but in a way that tends not to make
 an issue of the hermeneutics of a given poem. Rather than reading to
 recover the intentions of the poet or the meaning of the poem, waka
 are better read within a range of expressive capabilities, including
 stylistic positioning within a long tradition of schools.
 

 This non-hermeneutic tradition can be understood in part in
 terms used to discuss waka poetics (karon), in which literal, “direct
 language” (chokugo) and “metaphorical language” do not form the
 exclusive contrasting pair one finds in Western poetics. “Direct
 language” in classical Japanese poetry also contrasts with koji (
古事,
 waka-specific conventional diction, literally “ancient matters”),
 which Amagasaki Akira describes as beginning with “pillow words
 [makura kotoba] in a broad sense,” but also “the calling up of certain
 specific phrases that have important functions.” These koji appear as
 if italicized and linked to prior uses in other waka (and chōka).[5]
 Waka, being of a non-logocentric tradition, can in part be
 understood in contemporary terms in light of a distinction made by Udo
 Hebel, who approaches allusion (out of the context of waka) as a
 tension between readers and writers, rather than the initiation of an
 encyclopedic task of investigating references, whereby the more
 obscure generates more interpretive work in elucidating texts.[6] In
 reading waka it is not the degree of covertness that makes a text
 engaging in terms of allusions, but rather the very intertextual
 ‘activation of multiple texts’.[7] The intertexts are generally to be
 understood as multiple and not hermeneutic in the sense of eliciting a
 potentially unified reading by way of intertexts.[8] The intertexts
 form an extra-text-based array of potential allusions that diffuse the
 opposition of covert and overt allusion. In the end, assuming a text
 has an autonomous inner system is not a feasible method of reading
 waka.
 

 Waka intertextuality opens multiple arrays of details, various
 allusions already figured into earlier waka. Rather than the question
 of “originality” in compositions which merely happen to “include”
 allusions secondarily, there is the open practice of composing
 variations on previous waka (honka, base waka). The consummate waka
 poet more fluidly orchestrates allusions in the phrasing of the waka,
 with recognizable echoes of combinations of words and entire 5- and
 7-syllable segments from earlier waka, or in some cases memorable
 phrases from prose works or Chinese poetry.
 

 Moreover, because waka derive from a poetics open to
 conventionally associated words (engo,
縁語) in different lines within a
 waka, and conventional pivot words (kakekotoba, 掛詞, an elaborate form
 of often only partial double-entendre employed to eke out additional
 and surplus meanings within a waka), waka texts—at least those
 employing these figures (which often appear together in waka, the
 alternate meanings of the pivot-word reinforced by the associated
 words)—are by convention already internally disintegrated, with no
 expectation of a rhetoric of self-contained “completion” appearing, as
 in much Western lyric poetry, nor of parallelism as found in Chinese
 poetry within a Confucian framework. As Amagasaki emphasizes, the
 fundamental rhetoric in waka is linking by way of associated words
 (engo).[9] It is by way of the importance of these associations (en)
 that the frame and autonomy of waka, in contrast to Western poetry, is
 diminished.[10]
 

 Given these waka-specific varieties of intertextuality,
 prominent Western approaches to intertextuality, though (or perhaps
 due to) synthesizing a broad range of studies of intertextuality,[11]
 appear overly centered on conventional Western rhetorical models
 (logocentrically situating every utterance as reflecting a
 paradigmatic “idéologème”). The intertextuality within a poetics of
 allusion, central to waka culture, cannot be fairly reduced to these
 prevalent conventions of composition and appreciation. The semantic
 and syntactic texture of waka itself might be said to reference other
 waka so as to engage not only in making an allusion, but also in the
 intertextual displacement of the given waka itself. The emotive force
 often derives from a nostalgia called upon to the degree one or more
 honka (waka alluded to, lit. “base poems”) come to express an
 intensity that displaces the scene within the waka at hand. Thus in
 various ways the waka does not favor a hermeneutic mode of reading,
 which would deduce the working parts within a working whole.
 Though restricted in syllables, waka exhibit a high degree of
 intertextual congruity within the broader socio-literary context of
 contending schools of waka. In the very rhetoric of equanimity so
 common to the experience intimated in many waka, there is a sense that
 all things, people and inanimate objects included, coexist with
 varying degrees of intimacy. It was this poetics that the
 Shintaishishō (1882) poets sought to suppress, knowing it would not
 conform to their new Western models of poetry.
 

 In terms of meaningfulness, allusions coexist with the literal,
 the intertextual constantly undermining the present in a given waka.
 Western hermeneutic concerns often reduce meaning to questions of the
 sign. Reflected in the Kokinshū “Kana Preface,” Earl Miner writes,
 “the signifier and signified [in Japanese poetry] are not taken to be
 different so much as versions of each other in another aspect. The
 distinction between signifier and signified in western thought is
 thereby blurred: the two terms merge. And Japanese [readers of waka]
 find that natural.”[12] Thomas Lamarre approaches this issue somewhat
 differently, writing: “This is the great project of Kokinwakashū: to
 fix the relations between words, emotions, occasions (not in order to
 inhibit motion but to channel it),” and, “the function of poetic signs
 in waka” is “to act at various levels and to align them.”[13] Miner
 focuses on the importance of blurring signifier and signified in
 reading waka, an approach which describes of how “pivot words” and
 “associated words” are valued in waka culture in general. Lamarre sees
 the induction of a process in the Kokinshū, a mapping of relations
 between conventional topics, words, and emotions in conjunction with
 Chinese precedents. If one concurs with Lamarre, following a line of
 scholars who see the Kokinshū as incorporating Chinese poetics and
 culture,[14] then his poetics of alignment is appropriate. Yet if one
 admits that a waka poetics was significantly shaped in resistance to
 Chinese culture, one may interpret waka as foregrounding a more
 intertextually open and playful poetics. By reading parts of the
 Kokinshū and Ki no Tsurayuki’s “Kana Preface” against the Great
 Preface after which it was putatively modeled, I hope to demonstrate
 that waka was instituted to provide a modicum of resistance to the
 highly ordered, Confucian verse culture and affirm a distinct,
 Japanese (Yamato) verse culture.
 

 How we understand the waka, with its playful intertextual
 openness within a fairly closed lexicon, informs how we situate the
 new styles of poetry in the Meiji period. The end of the “closed door
 policy” and the introduction of a wide range of discourses that
 entered into all aspects of daily life rendered Meiji Japan a more
 sociolinguistically open environment. Poetry based on waka’s closed
 poetic lexicon could not speak to the new context in this period of
 change.[15] Yet it was waka’s “open” poetics of allusion within this
 closed context for poetry that led it to be widely acclaimed as
 poetry.
 

 In Ki no Tsurayuki’s seminal statement on waka, the “Kana
 Preface” to the Kokinshū, one can outline a genealogy of ideological
 tensions incorporated into waka by his use of examples, the
 counter-examples introduced by a later editor, and how these relate to
 the original Chinese six principles of poetry (rikugi) of the Great
 Preface. Here the distinction between polythetism and antithetism
 (referred to in the above epigraph) in the context of how the subject
 is situated in poetry and poetics of early China and Heian Japan,
 Victorian England and Meiji Japan provides a contrastive point of
 reference by which one can examine a genealogy of prominent Japanese
 poetic forms. In short, the waka as founded in Heian period will be
 shown to have resisted the antithetism of Chinese poetics by asserting
 an intertextuality that allowed for a thoroughgoing polythetism, which
 in turn did not meet the needs of the founding shintaishi poets who in
 emulating English poetry needed to displace waka and ground a
 narrative, historical poetry that engaged contemporary issues in the
 service of elevating the nation.
 

 While Shirane demonstrates the unconscious analogy modern waka
 critics made with romanticist expressivism at the expense of
 understanding intertextuality in Fujiwara no Shunzei, one may add an
 argument that intertextuality (in waka-specific figures) was already a
 national cultural asset consciously cultivated by Ki no Tsurayuki.[16]
 In his preface, we can see him situating waka in terms of the politics
 of the subject, positioning Japanese conventions of waka composition
 in contrast to Chinese stylistic precedents, especially the Six
 Principles of Chinese poetry (rikugi).[17] In his own principles and
 examples one discerns a conscious divergence from the Chinese
 classifications, and suggestions that waka should rely on a
 combination of conventional allusions (such as honkadori),
 standardized phrasing (makurakotoba) and polysemic phrases (kakekotoba
 and engo). This in effect characterizes a poetics that is not simply
 “affective-expressive” (focusing on the author), but rather a writing
 process that is by convention intertextually bound up in antecedent
 texts to the point of deferring to them, and in the various ways
 described above, a process that also assumes verbal play that conceals
 and multiplies expressions of a persona.
 

 In attempts to construct a shintaishi poetics for the Meiji
 period as well, though not referred to by name of course,
 intertextuality was an integral component of Japanese waka and had to
 be addressed if one were to overcome the influence of this prominent
 form. Shintaishi poets opened poetic lines to otherness—other
 discourses, Western poetry and even the sciences—and in this process
 also to expressions of nationalism and identity politics. The old
 intertextuality in waka relied on a closed, classical lexicon, while
 the new poetry incorporated new words and ideas, as if part of an
 argument to embrace the perpetual material revolution under
 capitalism.
 

 Tsurayuki’s seminal attempt to carve out a “Japanese” realm of
 poetic discourse, distinct from Chinese poetry (and thus foreshadowing
 Nativism), articulated an ideological role for waka that relied on
 intertextual figures, and thus provides an important point of
 reference when considering the Meiji assault on waka and attempts to
 develop shintaishi in this period. Though the examples of waka types
 that Ki no Tsurayuki offered in his preface have been challenged in
 extant copies of the preface, which have editorial insertions of
 “better examples,” it should be apparent from a close reading that he,
 in fact, meant what he wrote when he gave the six examples of the six
 modes of poetry.[18] These examples not only reflect the influence of
 the Chinese principles, but are chosen so as to showcase waka poetic
 modes as surpassing the Chinese ones in splendor.
 

 As Amagasaki Akira points out, though the Chinese reference to
 the rikugi or six poetic principles from the Great Preface appears in
 the “Manyō (or Chinese) Preface” to the Kokinshū, in the corresponding
 point in the definitive “Kana Preface” the phrase “there are six forms

 of expression (sama) in poetry” appears. The word sama ( or “style”)
 becomes an important term in waka poetics (karon), and here indicates
 a conflation of the categories of “forms of expression” and
 “contents.” Thus it is safe to say that Tsurayuki sought in this first
 treatise to shape a critical discourse specific to “Japanese poetry”
 (Yamato uta), which was indeed the opening phrase and topic of this
 preface. Other critical writings had merely replicated Chinese
 treatises. Replacing “principles” (gi) with sama was a cornerstone of
 his attempt to extricate critical writings on poetry from Confucian
 thought. He seems to have worked to establish a focus on the
 immanent fusion, in sama, of what is in one’s heart (kokoro) with the
 means of expressing it in words (kotoba) selected from an
 intertextually bound range.[19] The new discursive community defining
 Japanese poetry would foreground tangled emotions, rather than
 Confucian principles.
 

 For instance, the example of “comparison” (nazuraeuta) suggests
 that Ki no Tsurayuki sought to present a refined Japanese song (Yamato
 uta) that incorporated, along the lines of the categories of Chinese
 poetry, figures specific to a Japanese poetics and which were not
 found in Chinese poetry. Three such figures are found, not by chance
 it would seem, in his example of “comparison.” These figures are a
 prefatory statement (joshi, italicized in the example), a pivot word
 (kakekotoba, underlined), and associated words (engo, in bold).[20]
 Thus he highlights differences between the poetries and languages:
 phonetic play not possible in Chinese:
 
 If in the morning
 you awake as the day’s frost
 has just settled,
 as it begins to melt,
 so will my love.
 
 kimi ni kesa
 ashita no shimo no
 okite inaba
 koishiki goto ni
 kie ya wataran
 
 
君に今朝(けさ)(あした)の霜のおきていなば
 恋しきごとに消えやわたらむ[21]
 

 The comparison is not simply in the metaphorical mode suggested
 by a later editor who thought the following a better example, where a
 cloistered male is compared straightforwardly to a silkworm in a
 cocoon:
 

 A silkworm raised
 by his mother and left
 to mope in his cocoon.
 Will the gloom ever lift,
 unable to meet my love?
 

 tarachime no
 oya no kau ko no
 mayu komori
 ibuseku mo aru ka
 imo ni awazute
 
 
 
たらちめの親のかふ蚕()の繭(まゆ)こもりいぶせくもあるか妹(いも)に逢わずて[22]
 

 The editor, with his would-be “improvement,” misses what is
 apparently Tsurayuki’s underlying point. In Tsurayuki’s example, the
 comparison is multiple, as is readily possible with pivot words, with
 the engo in relationship to multiple contexts interposed by way of the
 associations (the alba motif and morning frost melting). These
 multiple intertextual and intratextual[23] relations exist within and
 without the example, which is not closed in the sense associated with
 hermeneutical textual studies, which aim to extract interpretive
 structures and meanings. Ki no Tsurayuki’s example deliberately
 flaunts the capacity of the Japanese language, using kana, to exploit
 polysemous strings, unlike the character-based Chinese, which is not
 capable of being rendered in phonetic script and thus exploit a large
 pool of double-entendres. Choosing a phrasing that repeats the
 character for ashita in succession (in kesa and ashita) would also
 seem to underscore the doubled senses of graphic and phonetic elements
 in waka and its inherent potential for playfulness. Moreover, at the
 time of the Kokinshū, kana had not been entirely recognized as an
 acceptable alternative to Chinese or man’yōgana,[24] and thus
 foregrounding this aspect of waka also attempts to tip the scale in
 favor of kana usage, displaying its distinct virtues.
 

 Similarly, the example of a “corrective poem”[25] (tadagotouta)
 suggests a redistribution of both Chinese poetic values and a
 differentiation from Chinese poetry and poetics:
 
 If this world were free
 of falsehood,
 how joyous might
 the words of my love be.
 

 itsuwari no
 naki yo nariseba
 ika bakari
 hito no koto no ha
 ureshikaramashi
 
 
いつわりのなき世なりせばいかばかり人の言の葉うれしからまし[26]
 

 It is not presented as “elegantia,” as has been used by Laurel
 Rasplíca Rodd and John Timothy Wixted,[27] an approximation of the
 Chinese term of the fifth principle of poetry in the Great Preface to
 the Book of Songs, the category of “odes” (
ya in Chinese, ga in
 Japanese). In both Chinese and Japanese this character has a sense of
 courtly elegance (its kun’yomi being “miyabi”). Yet the example
 Tsurayuki uses replaces the principle (gi) of ga with a sama
 (composition fusing words and heart), called tadagotouta, which can be
 interpreted in two ways in light of the example. First, tadagoto
 (originally written in kana in the preface) can mean “everyday

 affairs” (徒事・唯事・只事) as suggested by the Kōjien (with an example
 from the early Heian Taketori monogatari). In this sense, the spirit

 Tsurayuki would convey in the mere use of tadagoto explicitly
 contrasts with the elegance of ga. Second, tadagoto can, following
 Ozawa Masao in his notes to the “Kana Preface,” suggest one interpret
 the poem in terms of the contents of the poem, not the form, which
 reinforces the need to comprehend sama as more than merely “style,”
 but also the fusion of kokoro and kotoba in the organic sense
 presented in the opening of the preface. Ozawa argues that the
 contents of this waka refer to “a wish to be in a world as it should
 be, devoid of deceit.”[28] The poem sings of the barrier words bring
 to any romance, as words provide both affirmations of interest and the
 seeds of doubt. In the example the poet reflects introspectively on a
 romantic encounter, while linking his own joy to the shared medium of
 words.
 

 It must have been the culmination of these treasons against the
 sense of elegance and propriety in the Great Preface that annoyed the
 later editor, who prefaces his counter-example with references to the
 Chinese poetic and political ideals of a properly ordered world,[29]
 and uses in his example itself, “a world without wind to scatter the
 petals”:
 
 I have had my fill
 of gazing at the colors
 of mountain cherry blossoms
 in a world without wind
 to scatter the petals.
 
 yamazakura
 aku made iro o
 mitsuru kana
 hana chiru beku mo
 kaze fukanu yo ni
 
 
山桜飽くまで色を見つるかな花散るべくも風吹かぬ世に[30]
 
 This is an example that, while incorporating irony, derides
 being reflexive and playful in waka while encouraging a courtly pathos
 of distance. It implies an aloof and pretentious critique of Ki no
 Tsurayuki’s example, which is personal by comparison. The editor seems
 to be conveying that, if only Tsurayuki had stayed with the ya (ga)
 category, he would have demonstrated adequate comprehension of the
 Great Preface, especially its sense of orderliness and propriety.[31]
 Tsurayuki’s example underscores the conventional assertion of
 the unreliable character of the human heart as mediated through words,
 while the counter-example avoids such ambiguity in human affairs. In
 creating this anthology, certainly after careful consideration
 Tsurayuki must have decided that unresolved threads or tensions should
 be defining elements in the dominant topic of love in the Kokinshū and
 by analogy influencing the construction of waka in general. Love and
 disorder go together in the very word midareru, to be dizzy with
 infatuation, much to the apparent irritation of the later editor, who
 seemed to have preferred a stricter adherence to the Confucian
 tradition of songs as sources of public order, as can be inferred from
 how the later editor goes out of his way to choose a waka that denies
 the very thrust of Tsurayuki’s example, namely that the world is
 characterized by falsehood, death, falling blossoms. Instead he
 inserts a poem that displays disdain (even if conventional) for
 excessive indulgence in flower-viewing, and imagines a scene without
 wind, refusing the underling theme of love and disarray (midareru) in
 Tsurayuki’s waka.
 

 Moreover, as Fujiwara no Kintō’s editorial correction includes a
 negatively situated yo, in the phrase kaze fukanu yo (a world where
 the wind does not blow), one wonders if there was a rift among poets,
 even of different generations in the tenth century, between poets more
 interested in how Buddhism encourages a predilection for multiple
 positions, the polythetic, exploration of a phenomenology of poetic
 composition, and more conservative poets who sought to preserve the
 Confucian poetics that characterized the prestigious Chinese poetry
 and Classics. Tsurayuki’s example, after all, implies a Buddhistic
 assumption of itsuwari, all being vanity (lies), in wishing: “If this

 world were free / of falsehood, / how joyous might / the words of my
 love be.”[32] Nagafuji Yasushi argues that the use of “yo no naka”
 (世の中, in this world/life [of change]), first appearing in the
 Man’yōshū, suggesting a shift in temporal consciousness accompanied
 the introduction of Buddhist thought to Japan. “Yo no naka” implicitly
 challenged the spatial world view that “yo” originally suggested. We
 can extend this to objections to the highly spatially regulated views
 of society and the self within Confucian thought. Tanaka Gen, who
 focuses on the Kokinshū, sees the Man’yōshū as “substantive texts”
 based primarily on the presence or absence of things: objects, people
 ideas, emotions. Though these “substantive texts are essentially
 expressions of decisions, and negotiate no temporal relation,” waka in
 the Kokinshū are “predicative texts that have an essentially deep
 relation to time,” he writes. In predicative waka the occurrence,
 progress, and continuance are expressed according to the speaker’s
 choices.[33] Given the various temporal gradations classical
 conjugation affords depiction of the relation of poet to his or her
 fiction, Japanese waka, written with kana, in contrast to kanshi, had
 an opportunity to resituate the expressive technology of the kanshi
 culture as articulated in the Great Preface, as Tsurayuki’s examples
 in the “Kana Preface” show.
 

 Another example written by Ki no Tsurayuki and included in the
 Kokinshū explicitly uses midareru to describe a spring scene of
 blossoms, probably cherry, blown about by a profusion of hanging
 willow branches. Clarifying a major difference from Fujiwara no
 Kintō’s poetics, here external “confusion” is central to the
 excitement in the poem, suggesting an ironic appreciation without need
 of carefully balancing inner and outer realms in the veiling and
 unveiling that surely pleased Fujiwara no Kintō.
 
 Threads of willow
 tangled in the wind
 this very spring,
 in the frenzy of full bloom
 the cherry blossoms are unraveled.
 

 aoyagi no
 ito yori kakuru
 haru shi mo zo
 midarete hana no
 hokorobinikeru
 
 
青柳(あをやぎ)の糸よりかくる春しもぞ乱れて花のほこるびにけ[34]
 

 Thus from the very origins of the imperial waka tradition, an
 aesthetics and ideology of countermanding the influence of other types
 of poetry, whether Chinese (beginning with Tsurayuki) or Western
 lyricism in the Meiji period, are available in the very dispersive
 effects inherent to waka poetic discourse. Rhetorically, waka poetics
 (karon) can be seen as inherently sustaining a mode of ideological
 displacement of other poetics (Chinese and later Western).
 Historically, these intratextually “associated words” and “pivot
 words” (engo and kakekotoba) and intertextually recycled phrases
 (“pillow words,” “prefatory words,” honkadori and allusions, and so
 forth) became integral to the demonstration of mastery of the form and
 participation in elite cultural and political circles. Waka may be
 seen as in part a form of poetry founded within a closed poetic
 lexicon expressly designed to replicate the playful dispersion of
 meaning, Confucian gravity and its ideal of the mindful subject, and
 to in this way remain somewhat distinct from Chinese poetry and
 culture. Deconstructing this aspect of waka helps to explain the
 genealogy of its role in Japanese nationalism.
 

 In the context of the Great Preface’s Confucian poetics, with
 its metaphysics of recovering a utopian socio-political order, waka in
 the Kokinshū as well as the Shinkokinshū establish a cyclical sense of
 time moving forward seasonally and/or in accord with matters of love,
 not bound to an essentialist metaphysics.[35] The poetics is grounded
 in dominant imagery and themes, such as “early spring” as it
 transitions to “spring proper.” As Konishi Jin’ichi et al describe the
 sequence in the seasonal poems:
 
 The overall time progression from early to late spring is very subtly
 handled in order to create a total effect of the passage of the season
 which is in harmony with the physical world. This effect is achieved
 through the conscious manipulation of certain dominant images; the key
 “spring” image in one poem is juxtaposed to the key images in the
 preceding and following verses in such a way that the reader is
 carried along through the vicissitudes of early spring weather….[36]
 
 The effect is that of situating each waka in an implicit drama of life
 larger than the poem itself, which is dependent on the poems preceding
 and following it. This ordering of waka became the foundation for all
 pre-Meiji poetics, for the themes of seasons and love would dominate
 not only waka, but be codified even more strictly in renga (linked
 poetry), haikai no renga (non-standard linked poetry) and haiku.
 The progression of romantic entanglements and relationships too
 was codified in the Kokinshū, in which the six books on love dominate
 explicitly in poems 469 thru 828, 359 of the 1111 poems in the
 collection. For instance, the sequences on love can be broken down
 roughly into, according to Konishi: “The development of an affair from
 ‘love yet undeclared’ (469–551), through such phases as courtship
 (552–615), love after first meeting (616–704), the lover’s growing
 coolness (705-746), and, the ending of the affair in bitterness and
 misery (747–828).”[37] Teele breaks down five such generally grouped
 waka sequences on stages of a romance into between 14 and 29
 sub-topics each (differing with Konishi only slightly in his general
 grouping).[38]
 

 The point is that any waka, given the importance of the Kokinshū
 and later imperial waka anthologies that emulated its ordering, such
 as the Shinkokinshū, affirmed and defined not only a de-Sinified
 orthographically (primarily kun’yomi or Japanese reading in lieu of
 only kanji in kanshi) but also a de-Confucianized poetics, which
 partially exempted itself from the focus on the metaphysical,
 socio-political relationship of intent and landscape, and instead
 focused on an more open horizon of possibilities in the correspondence
 of human affairs and natural affairs, inner heart and outer landscape.

  The new scheme would in itself constitute a poetics by which all
 waka refer, not with the aim of addressing the matter of one’s intent
 and social standing, as Chinese poetry ultimately did, and in doing so
 encouraged a place for hermeneutically unpacking a given poem in terms
 of the historical and allusive literary command the poet demonstrated
 within Confucian values of uprightness in expression that naturally
 flows from the morally engaged poet.[39] On the contrary, waka would
 increasingly become opaque to such hermeneutic attempts to read them,
 and merely reinforce in the performance of writing itself (not its
 socio-political “contents”) the courtly status of those already within
 aristocratic families. They would resist this foreign system that
 included poetry as part of its civil service examinations, as well as
 resist stylistically the parallel structuring integral to Chinese
 poetry, which was predicated on couplets and generally reflected the
 recovery and exhibition of Confucian values in society and self
 (putting aside the poetry inspired by Taoism and Buddhism, since the
 poetics resisted in the “Kana Preface” of the Kokinshū was clearly
 Confucian).
 

 The intertextual poetics of allusion referred to in this paper,
 though still part of the reading of waka today, would not survive in
 its intertextually open form as material changes transformed society
 and rendered the established poetic diction antiquated. In the Meiji
 period in particular, new contexts emerged by which the ideological
 usefulness of waka intertexts and the restricted poetic lexicon
 changed dramatically. Though some shintaishi poets, such as those
 contributing to the Shintaishishō, sought to eliminate waka from the
 literary horizon, for others its apparent disappearance would be an
 error to be corrected (in the development of modern tanka by poets
 appearing in the journal Myōjō, especially Yosano Akiko). In general,
 “Bit by bit a past work of literature will come to refer to one
 environment while its readers refer to another.”[40] As the Meiji
 changes were so abrupt, waka composition in the mode sustained for
 over a thousand years would become alienated from the new Japan.
 Moreover, while some shintaishi would exhibit the occasional
 waka figure, most of the early poetry presented as shintaishi, as
 Kamei Hideo observes, being narrative poetry (monogatari-shi), was not
 conducive to such waka figures of polysemy,[41] which function by
 convention best in very short poems that rely on such allusive
 resilience and intratextual play, and that tend to untangle the linear
 “presence” of a given voice, whether directed at Confucian, Western or
 Japanese thought that strays from the established diction of waka.
 
 More on Waka Intertextuality and Translated Poetry


 One fairly random example, Kokinshū 532, from the first
 book on love, topic unknown and anonymous, exemplifies the aspects of
 waka that make it important to understand in terms of the critical
 issues raised in this section:
 
 Sleek seaweed
 tangled in waves heading
 neither inshore nor out to sea,
 dizzy with infatuation,
 will my love be kept drifting too?
 

 okihe ni mo
 yoranu tamamo no
 nami no ue ni
 midarete nomi ya
 koiwatarinamu
 
沖辺(おきへ)にも寄らぬ玉藻(たまも)の波のうへに乱れてのみや恋ひわたりなむ[42]
 

 The opening image of the sea and shore, with oki (the sea, the
 offing), unlike its English counterpart’s sense of “just over the
 horizon” (with a sense of “just beyond the rainbow”), in Japanese
 poetry at this time and still today there is a ruggedness associated
 with such seashore and offing scenes, particularly because many
 striking chōka as well as waka in the Manyōshū include imagery of
 laboring on the shore, and partly because of uses such as in the
 makurakotoba (pillow word) okitsu nami (waves in the offing) which
 modify are (rough). An example of this is found in one of the few
 chōka in the Kokinshū, the opening here translated by Rodd as, “I too
 who now fish / the wide waters of Ise / lived for years on end /
 within that palace now grown / desolate and still / like a ship adrift
 in rough / waves in the offing / no mooring can I find.”[43] Already
 the similarity of feeling adrift in association with “waves in the
 offing” reinforces a specific scene and emotional experience. The
 first three phrases (the upper verse) is a poetic preface (jo or
 jokotoba) creating an analogical relationship with midarete, which has
 already been discussed above for its multiple sense of confusion,
 entanglement, and infatuation. With the preface—“Sleek seaweed / … in
 waves heading / neither inshore nor out to sea”—modifying midarete
 creates substantial complication in this meaning-based (ushin) joining
 of the preface to the rest of the waka, grounding the emotions in a
 very specific image that, in being applied to a specific word as if it
 were always meant to be there, has a gravity of a poetic epithet or
 makurakotoba.
 

 Also in this poem is the poetic word “Jeweled seaweed,”
 which has, by prior uses in the Manyōshū, sexual connotations, as in
 the image of “cutting / the sleek seaweed” in a poem “by Yuge thinking
 of Princess Ki.”[44] Also in book two of the Manyōshū, poem 131, a
 chōka with the headnote attributing it to “Kakinomoto Hitomaro when he
 parted from his wife in the land of Iwami and came up to the capital,”
 includes lines which further clarify the sexual aspects of “sleek
 seaweed”:
 
 
 
 Though it has no good inlets,
 I don’t care.
 Though it has no good lagoons,
 the wind, with morning wings,
 carry over those whale-hunted seas
 to the desolate beach in Nikita harbor
 green, sleek seaweed,
 
 Seaweed from the offing,
 and the memory of my wife,
 whom I left there
 as mist and frost
 are left on the ground,
 who swayed to my side in sleep
 like sleek seaweed
 swaying to and fro with the waves.[45]
 
 
よしゑやし 潟はなくとも 鯨魚取り 海辺を指して 柔田津の 荒礒の上に か青なる 玉藻沖つ藻 朝羽振る 風こそ寄せめ 夕羽振る 波こそ来寄れ
 波のむた か寄りかく寄り 玉藻 なす 寄り寝し妹を露霜の 置きてし来れば[46]
 
 
 Thus many if not most of the words used and reused in waka
 sustain not specific allusive relations to antecedent texts but a
 generalized intertextuality that supports what otherwise might be a
 confusing joining of words and phrases. Waka increasingly demonstrated
 elegance and wit by way of the varied reiteration of existing phrases,
 which along with the generic use of topics from the mid-ninth century
 engendered an intertextual poetics within the semantic limits of
 poetic expressions and topics.[47] One can see an emergence of a sense
 of concrete expression that yet allows a critical drama in a given
 short waka poem that is intertextually situated within a larger,
 emerging body of waka. Moreover, the Manyōshū had already provided an
 intertextual model for the centrality of love in conjunction with
 seasonal progression, thus also providing the means for a
 sophisticated tailoring of a poetics, as Ki no Tsurayuki did in his
 preface, so as to not be overshadowed by the love-eschewing Confucian
 precursor of waka poetics.
 

 Teele points out that the 539 poems of book ten of the
 Manyōshū are arranged: “spring poems, spring love poems, summer poems,
 summer love poems, autumn poems, autumn love poems, winter poems,
 winter love poems.”[48] This interweaving of nature and human affairs
 writ large engenders a meaningful context for writing and reading
 waka, and even though the context is not explicit, the poems are not
 abstract. Meaningfulness and complex tensions emerge intertextually,
 by way of a fullness of meaning that is not simply a matter of
 previous uses that semantically augment the meaning of words and
 phrases, nor of allusions that refer to antecedent literary or
 historical situations. The appreciation of waka depends on
 conventional associations and implications in words and phrases that
 through repetition have come to be part of a larger seasonal, romantic
 or other series of expected occurrences. Exploring one waka’s
 integration of multiple images in set phrases I hope will sufficiently
 demonstrate how this intertextual dependency in part defines waka
 poetry ultimately in order to show why shintaishi poets felt compelled
 to suppress this poetics in the Meiji period.
 

 To turn briefly to an early Meiji waka, one can see how
 complicated writing poetry becomes in this period. The intertextual
 poetics of allusion referred to in this study, though still part of
 the reading of waka, would not survive in its intertextually open form
 as material changes transformed society so as to render much of the
 poetic diction antiquated. In the Meiji period in particular, new
 contexts emerged by which the ideological usefulness of waka
 intertexts and the restricted poetic lexicon changed dramatically.
 Though some, as is expressed in the Shintaishishō, sought to eliminate
 waka from the literary horizon, for others its apparent disappearance
 would be an error to be corrected (which lead to the evolution of
 tanka in the 1890s, lead by Myōjo poets such as Yosano Akiko and
 Tekkan). In general, the historical context is always changing, and
 very rapidly in the early and mid Meiji period.[49] As the Meiji
 changes were so abrupt, waka composition became alienated from the new
 Japan. This displacement of how poetry in Japanese (wabun) was read in
 the Meiji period is ideologically important; not only had waka poetry
 been politicized by the National Studies scholars, but evocations of
 waka-like poetic enunciations also permeated shōka (discussed in a
 later chapter), as well as chōka (or nagauta), which resemble extended
 waka.
 

 An example of the hybridization (or compromising) of traditional waka
 and karon and its slippage into kyōka (“satirical poems,” non-standard
 if not satirical waka) may be taken from the 1878 (Meiji 11)
 collection, Kaika shindai kashū. Its diction seems unremarkable, at
 first glance:
 
 

 Along with the wind
 heard in heaven—

 though not the sounds of visitations
 from the world of gods
 I keep thinking about them.
 
 (Endo) Sugamori
 kaze no muta
 ten ni kikoeshi
 kami no yo no
 otodurezu tomo
 omoiyamarenu
 
 
 風のむた天にきこえし神の世のおとつれずとも思ひやまれぬ[50]
 猿渡
 

 What makes this waka unusual is the topic. It is presented under the
 topical heading “the telegraph” (電信機). Given this topic, the “double
 meaning” one would expect in kyōka or senryū materializes in the
 figure known as the “dig” (ugachi). The absurdity in this waka is
 commensurate with the seamless transposition of the closed poetics and
 diction of waka into a context with the then futuristic technologies
 being introduced as part of the nation-building project of the
 Enlightenment. The phrase “though not the sounds of visitations”
 suggests superstitious premonitions of divine encounters. Yet with the
 topical context of the telegraph, these words become a scientific
 description of magical expectations as the poet playfully suggested an
 alternative explanation rooted in poetic speculation. Moreover, the
 phoneme oto does, as often is the case in classical waka—especially in
 the more polysemous ones emphasized above— contains the pun of
 oto-dure, meaning both “no sounds trailing” and “not visiting.” This
 pun suggests the sounds of a telegraph as well as the attempt to
 discern a visiting god (or similar paranormal phenomenon) while the
 wind is blowing. The wind importantly sets the stage for dissonance in
 the opening phrase. Along the lines of the tension between scientific
 explanation and folk beliefs is that between critically developed
 knowledge and the common use of gods to explain whatever appears
 incomprehensible.[51] This split in waka would not be maintained in
 most waka of the Meiji period with such acute tension; there would be
 a crisis in waka precipitated by the insistent assault on waka in
 Shintaishishō prefaces.
 

 While some shintaishi would exhibit the occasional waka figure, as
 Kamei Hideo observes, most of the early poetry presented as
 shintaishi, being narrative poetry (monogatari-shi), was not conducive
 to such waka figures of polysemy,[52] which function by convention
 best in very short poems, where there may be balanced intratextual
 slippage and/or intertextual points of reference that undo the linear
 “presence” of a given waka. This is true of all the earlier shintaishi
 anthologies: the Shintaishishō, Yamada Bimyō’s Shintaishisen, and the
 five Shintaishika volumes. The seminal Shintaishishō translations
 included for the most part poems based on events of the past: battles,
 general tales derived from a branch of the military, or
 pseudo-narrative poems such as “Grey’s Elegy.” The original poems too
 began with one event and concluded with a succeeding, related event
 presented as “temporal development itself.”[53] Bimyō drew heavily on
 narrative tales in his poetry as well as in some of his fiction.
 Though there are lyrical vignettes as well, both the poetry espousing
 people’s rights and pro-government poems (beginning with the
 Shintaishishō) took narrative forms.
 

 In sorting out the ideological roles of poetry in Japan, one might
 examine the suggestion that Western epic poetry evolved into a
 structural worldview, into “history,” itself shaped by the narrative
 sequencing in epic poetry. National histories around the world can be
 seen as having been produced by adapting to the threat of Western
 colonialism by constructing their own “histories” built on this
 structured narrative model originating in the West, even if their own
 worldview and model of society, epistemology, and history had been
 quite different.[54] A long, epic form of poetry did not exist in
 Japan. Even though there were within lyric writing localized
 associations of imagery and themes (central to waka sequences, renga
 and nō), they were not written with epic-narrative continuity. Lyrical
 poems were subordinated to the contexts of narrative in the Kojiki,
 Heian monogatari and nikki (literary diaries), and in haibun (prose
 poems including haiku and written in the spirit of haikai). Premodern
 Japanese poetry did not include a poetry imagined as an extended
 “poetry of constructed thoughts”[55] per se (what the Shintaishishō
 poets advocated). Kanshi (poetry in Chinese) too had been written in
 Japan mostly in short verses, and being bound up with a Confucian
 metaphysics tended to reflect various antithetical pairings of
 compliments (such as principle and substance). With their inclusion of
 broad ideals, one finds here greater similarity with Western poetry
 available in the 1880s, and a kanshi elements are evident in many
 poems of the Shintaishishō.
 

 Chōka most closely resemble shintaishi, but they were in most respects
 another version of waka, though because of chōka’s very capacity for
 narrative exposition it usually lacked the intertextual tension that
 waka conveys in the balance of human affairs and natural scene. The
 poetry of the Shintaishiika (volume one, August, 1882), in part, would
 challenge the Shintaishishō (July, 1882) poets’ unanimous declaration
 that traditional poetry (waka) and poetics (karon) stood in the way of
 national interests. Some volumes even attempted to find shintaishi
 prototypes in earlier chōka by writers, such as Fujita Tōko and Sakuma
 Shōzan.[56] These appear to be introduced more as afterthoughts,
 putative antecedents, rather than as models for the new poetry.
 Nevertheless, other poets would later attempt to revive a new chōka
 (shin-chōka) as an alternative to shintaishi, but the influence of
 chōka per se on shintaishi is questionable both in terms of the
 context that gave rise to shintaishi (as a response to the political
 uses of hayariuta) and the association of chōka with a traditionally
 restricted poetic lexicon that would seem to inhibit engagement in
 sustained and wide-ranging thought. Though some of the less famous
 shintaishi of the 1890s emulated chōka, indeed might be termed chōka,
 the apparent impact of this genre on the initial formation of
 shintaishi appears to be very limited.[57]
 

 Waka were too lyrical, in the sense of containing action within the
 realm of the sensory and emotional, for the needs of the time of great
 action and changes. They were absurdly removed from daily concerns in
 the early Meiji period. As short poems found in traditional Japanese
 poetry left very little room for the development of ideas and
 narratives, they did not fit the needs of nation building, which
 itself was predicated on constructing grand historical narratives of
 the creation of modern Japan. It is no surprise that many of the
 translations and original poems were lengthy narratives. One can only
 wonder how the locally linked images and topics of renga and waka
 sequences might survive the challenge of this new ideology and
 aesthetic. Renga by its very structured composition is disjunctive:
 every other line should break from the scene evoked by the preceding
 two verses and link only with the immediately prior verse. Resistance
 to narrative development is part of the genre (whether composed alone
 or, as was usual, in a group). The rhetorical effects found in waka
 can occasionally be found in shintaishi, but length itself demands
 organization around ideas or repetition of words or sounds that
 sustain lyrical engagement with a topic, but, as in chōka (which
 exists in the shadow of waka, definable as an “extended waka”), the
 process of prolonging the 5-7 phrasing in waka and adhering to its
 poetic lexicon tends to render many of the waka effects, which rely on
 the briefly poised interactive engagement of sentiment and scene, as
 passing anomalies.
 

 The “composed thoughts” that would-be Meiji poets found in examples of
 Western poetry exhibited narrative organization not seen in poetry in
 Japan, and structuring metaphors—juxtaposing words associated with
 realms usually considered disparate, so that one semantic domain is
 transferred onto another so as to produce various effects, such as the
 justification of one order in terms of another. For instance, in the
 lines “The meteor flag of England / Shall yet terrific burn,” in

 Thomas Campbell’s “Ye Mariners of England: A Naval Ode”[58] is
 translated as “国の光とたてし旗 / 益光り輝きて” (“the flag raised in the
 country’s honor [lit. light], / resplendent with readiness”). Though one might
 say all the translations in the Shintaishishō were chosen so as to
 best match what Japanese verse would accommodate as “poetry” (shīka),
 one sees even in this marginally original metaphor born of the
 juxtaposition of the celestial realm of “meteor” with the national
 flag of England the metaphor has been rejected as inappropriate or
 untenable in shintaishi: “the meteor flag of England” becomes “flag
 raised in the country’s light.” In the original the meteor metaphor
 reinforces the intense perseverance in the heat of the battle, yet it
 has been omitted in lieu of a more general “light” in the translation.
 

 From the perspective of shintaishi poets, such English poems would
 seem to inhabit a new mode of lyricism that denied temporal rupture,
 as well as figures such as kakekotoba and associations internal to the
 poem (engo). The wordplay of waka was in light of these foreign poems
 seen as superficial, conventional and thus exhibiting the malady of
 “dead” metaphors. The life of an English poem was its fresh turns of
 phrase and metaphors. In addition, scientific positivism reinforced
 the premise that connections between ideas and things could be mapped
 out and clarified in toto, and that metaphors, subordinated to this
 overarching belief, had incorporated a teleological faith that the
 cognitive dissonance produced by skillful yet discordant metaphors
 would achieve a satisfactory degree of resolution. From a broader
 vantage, one can see how in resituating poetry along the lines of
 English verse the very socio-linguistic expectations and acceptable
 modes of representing the country were being renegotiated in the Meiji
 period as part of the project of resituating Japan as a world-class
 nation.
 

 The Westward-looking Shintaishishō poets thus found themselves
 advocating “connected thoughts” (連続したる思想)[59] in lieu of the “linked
 images” of renga or Noh, or the intertextual poetics of allusion that
 characterizes pre-Meiji waka. Though the relatively heavy-handed use
 of discordant, evocative metaphors in translations of European poetry
 would continue to be omitted or often sound somewhat clumsy in
 shintaishi, the more prevalent use of narrative was incorporated with
 immediate broad success in shintaishi and became one of its defining
 characteristics.
 
 
 [1] Gilles Deleuze, tr. and foreword by Brian Massumi. A Thousand
 Plateaus: Capitalism and Schizophrenia. Minneapolis: University of
 Minnesota Press, 1987, 498. I want to thank especially Bill Sibley and
 Susan Klein as well as Norma Field, Rein Raud and Lewis Cook for
 helpful comments on an earlier version of this or for discussing
 related issues with me. An abridged version of this essay appeared as
 “The Formation of Allusive Resilience in Waka and Its Relevance to
 Meiji Shintaishi.” Japanese Poeticity and Narrativity Revisited,
 Proceedings of the Midwest Association for Japanese Literary Studies
,
 Vol. 4 (2003), 166-83.
 

 [2] Although I treat waka primarily in terms of the Kokinshū and its
 “Kana Preface,” glossing over a long tradition of waka criticism
 (karon), this is sufficient for the purpose and scope of the primary
 topic, Meiji shintaishi. Moreover, as Thomas Lamarre has pointed out,
 the Kokinshū and prefaces initiated the primary poetics for waka over
 the centuries to follow. See Thomas Lamarre. Uncovering Heian Japan:
 An Archaeology of Sensation and Inscription. Durham: Duke University
 Press, 2000, especially 14.
 

 [3] See Nicholas John Teele. The Love Poems of the “Kokinshū”: A
 Translation, with Commentary, and Study of the Influence of Chinese
 and Earlier Japanese Poetry. Austin: The University of Texas at
 Austin, 1980, 178–195.
 

 [4] The relationship of metaphor and philosophy has been explored as a
 central figure in Western thought in Jacques Derrida, “White
 Mythology,” trans. by Alan Bass, in Margins of Philosophy, Chicago:
 University of Chicago Press, 1982, 207–271. For an outline of various
 figures and terms comparable to “metaphor” in classical Japanese
 literature and many examples, see Kokubungaku henshūbu, Koten bungaku
 retorikku jiten. Gakutō, 1993.
 

 [5] See Amagasaki Akira, En no bigaku—Uta no michi no shigaku, II.
 Keikō Shobō, 1995, 22. He also points out that Fujiwara no Hamanari
 (724–790) had situated this distinction as one of three aspects of a

 waka, along with “fresh meaning” (shin-i, also understandable as “new
 intention or sense”) that presents the image (butsuzō, 物像), and kekku
 (
結句), concluding verse, which was for him the delivery, the singing
 the “the thoughts of the author.”
 

 [6] Hebel, Udo J. “Towards a Descriptive Poetics of Allusion,” in
 Heinrich F. Plett, ed., Intertextuality, New York: de Gruyter, 1991,
 135–164. Hebel also cites Harold Bloom in Maps of Misreading, writing
 that “Bloom’s conclusive dichotomy draws attention to a significant
 point of controversy among scholars of allusion as it foregrounds the
 definitional opposition ‘covert’ vs. ‘overt’.” Hebel, 1991, 136.
 

 [7] I am paraphrasing Ziva Ben-Porat, cited in Hebel, 1991, 136.
 

 [8] Citing Riffaterre, Hebel writes, “any allusion acts as a
 ‘stumbling block’, drawing the reader’s attention to the text’s
 intertextual relationships. From this perspective, the allusive system
 of a text becomes the verifiable cross section where text and
 intertext meet, and where the intertextual background of the text
 becomes tangible for the reader. In terms of intertextual theory,
 allusions are manifestations of the text’s idéologème that marks the
 text’s historical and social coordinates.” Hebel, 1991, 139. In waka,
 the allusions are as likely to be points of dispersion in a chain of
 intertexts as to be references that ground the waka at hand in a
 specific literary scene in a specific earlier work.
 

 [9] Amagasaki, 1995, 19.
 

 [10] Amagasaki, 1995, 20.
 

 [11] See, for instance, Hebel, 1991.
 

 [12] Miner, 1990, 93.
 

 [13] Lamarre, 2000, 58 and 62.
 

 [14] Lamarre, 2000, 145.
 

 [15] The famous scholar Yoshida Seiichi links kanshi to a feudal
 character, and cites Tsubouchi Shōyō, who called tanka and chōka “the
 poetry (shiika) of a closed world.” See Yoshida Seiichi, 7–8. It is
 this closed aspect that complicates consideration of how we might
 situate its intertextuality, which is precisely what makes old and new
 forms of poetry so divergent and politically contentious.
 

 [16] See Shirane, 1994, 90, 93.
 

 [17] Accounts of the relation of the six Chinese principles of poetry
 and the principles and examples given in Ki no Tsurayuki’s preface
 tend to emphasize the passive reception of ideas, not his own agenda
 of defining Yamato. See John Timothy Wixted, “Chinese Influences on
 the Kokinshū Prefaces,” in Rodd, Laurel Rasplica, with Henkenius, Mary
 Catherine. Kokinshū: A Collection of Poems Ancient and Modern. Boston:
 Cheng and Tsui Company, 1996, 387–400.
 

 [18] The anonymous editor is thought to be Fujiwara no Kintō. Thanks
 to Susan Klein for pointing this out. See Rodd, 1996, 36; and
 Nishishita Kyōichi and Sanekata Kiyoshi, Kokinshū, Shinkokinshū, in
 Zōhō kokugo kokubungaku kenkyūshi taisei, vol. 7. Sanseidō, 1977, 80.
 Also see Wixted in Rodd, 1996, 387–400.
 

 [19] Amagasaki Akira. Hanatori no tsukai—uta no michi no shigaku, I.
 Keisō Shobō, 1983, 34–39; on separation from Chinese poetics and
 Confucian thought, see 43–47; in “Kana Preface,” Ozawa Masao.
 Kokinshū, in Nihon koten bungaku zenshū, vol. 7. Shogakukan, 1971, 51,
 and note 19.
 

 [20] Pivot words (kakekotoba, 掛詞) - kana characters that can be read
 in two or more ways, not only so as to present semantic ambiguity, but
 also multiple syntactical ways of reading a string of characters
 (usually, a noun is also read as a verb or adjective with a very
 different meaning, such as using matsu to mean both “(I) wait” and

 “pine tree”). Associated words (engo 縁語) are words within waka
 (usually different phrases) that are intratextually related by way of
 common context (such as the sea, freezing weather, spring occurrences,
 etc.), often suggesting this context as a vision or metaphorical
 backdrop for the more immediate scene depicted in the poem. A
 prefatory statement or preface (jokotoba or joshi,
序詞), resembles the
 common makurakotoba (
枕詞) usually translated as “pillow words”
 (epithets), but was not fixed, often extended for two or three lines,
 and connected to the rest of the poem by either sound or sense:
 repetition of sounds to bind the conjoined parts of the poem, or by
 sense, setting up an analogy of natural occurrence or scene with the
 social, for instance using nature imagery to suggest some aspect of a
 love affair. Those phrases bound by sense would in the late Heian
 period be distinguished as “meaningful prefaces” (ushin no jo).
 

 [21] Ozawa Masao. Kokinshū, in Nihon koten bungaku zenshū, vol. 7.
 Shogakukan, 1971, 52.
 

 [22] Ozawa, 1971, 52.
 

 [23] Joseph Allen’s use of “intratextuality” in his study of Yuefu
 poetry contributed to my understanding of the importance of
 distinguishing various forms of this mode. See Joseph Allen. In the
 Voice of Others. Ann Arbor: Center for Chinese Studies Publications,
 1992.
 

 [24] Teele, 1980, 38.
 

 [25] Rein Raud interprets tadagotouta as poems using “plain language,”
 Kubota Shōichirō as poems using “everyday words just as they are.”
 Both of these approaches avoid the mimetic connotation possible if one
 were to associate ordinary words with description. See Rein Raud, The
 Role
Of Poetry in Classical Japanese Literature: A Code and
 Discursivity Analysis. Tallinn: Eesti Humanitaarinstituut, 1994, 99.
 Kubota Shōichirō, Kokinwakashū. Kadogawa Shoten, 1973, 11.
 

 [26] Ozawa, 1971, 53.
 

 [27] See Rodd, 1996, 39 and 393.
 

 [28] Ozawa, 1971, 53.
 

 [29] See Ozawa, 1971, 53, note 16.
 

 [30] Ozawa, 1971, 53. Notice how the image of “last three lines” was
 used with a similar eye on “national” interests in Yoshida Shōin’s
 waka (cited in an earlier chapter).
 

 [31] Stephen Owen, Readings in Chinese Literary Thought, Cambridge,
 Mass.: Harvard University Press, 1992, 49.
 

 [32] Yo appears 71 times in the Manyōshū, while “yo no naka” appears
 44 times. See Nagafuji Yasushi, Kodai Nihonjin no jikan ishiki—Sono
 kōzō to tenkai. Miraisha, 1979, 66.
 

 [33] See Tanaka Gen, Kodai nihonjin no jikan ishiki: sono kozo to
 tenkai. Yoshikawako bunkan, 1975, 121. Similarly, the famous early
 twentieth century scholar of Nietzsche and Hegel, Watsuji Tetsujirō
 wrote that lyric time in the waka of the Kokinshū entailed avoiding
 outright exclamation, maintaining a measure of composure, analyzing
 the psychology of love or describing the atmosphere of love, and
 overall paying attention to “the processes of emotions.” Cited in
 Tanaka, 1979, 65–66.
 

 [34] Ozawa, 1971, 71. See very helpful notes in Kubota, 1973, 23.
 

 [35] Konishi (1989, 99) calls this a “retrospective orientation” in
 describing the centrality of allusion in Chinese poetics.
 

 [36] Konishi Jin’ichi, tr. and adapted by Robert Brower and Earl
 Miner, “Association and Progression: Principles of Integration in
 Anthologies and Sequences of Japanese Court Poetry, A. D. 900-1350.”
 Harvard Journal of Asiatic Studies, Vol. 21 (Dec., 1958), 74.
 

 [37] Cited in Teele, 1980, 177. Konishi, 1958, 100.
 

 [38] See Teele, 1980, 178–195.
 

 [39] See Owen, 1992, 37–49.
 

 [40] See J. R. de J. Jackson, Historical Criticism and the Meaning of
 Texts. London and New York: Routledge, 1989, 39. Jackson provides
 numerous diagrams of the historically changing relationships between
 the socio-literary environment, the language, the text at hand, and
 the readers of different time periods. See Jackson, 24, 40, 45–49, and
 52. Though paying less attention to the historical dimension, a more
 nuanced approach to intertextual differences of context for pretexts
 and texts at hand is suggested by Hebel’s summary of Ben-Porat’s four
 stages of processing allusion; these are: “recognition of a marker,”
 which is a “directional signal ... always identifiable as an element
 or pattern belonging to another independent text,” “identification of
 the evoked text, modification of the initial interpretation of the
 signal, activation of the evoked text as a whole in an attempt to form
 a maximum of intertextual patterns,” Hebel, 1991, 137–38.
 

 [41] This is true of all the earlier shintaishi anthologies: the
 Shintaishishō, Yamada Bimyō’s Shintaishisen, and the five Shintaishika
 volumes. Yamada Bimyō uses pivot words in the narratives poems in
 Shōnen sugata, though even these can be attributed to gesaku shōsetsu

 models, not waka, and are themselves atypical shintaishi in a style
 that Bimyō himself would soon abandon in favor of experimental shōka
 (唱歌) and shintaishi. The self-published Yuasa Hangetsu, though not
 very influential, had some training in waka, and lines in his
 monogatari-shi occasionally include kakekotoba, and often mix waka
 diction with his Christian subject matter. See Yamamiya, 1950, I:
 115–133.
 
 [42] Ozawa, 1971, 229.
 
 [43] See Rodd, 1996, 346-47.
 
 [44] See Ian Hideo Levy, The Ten Thousand Leaves. Princeton: Princeton
 University Press, 1981, 93-94.
 
 [45] Levy, 1981, 98.
 
 [46] See University of Virginia Library, Japanese Text Initiative,
 http://etext.lib.virginia.edu/japanese/manyoshu/AnoMany.html. Accessed
 1/13/03.
 
 [47] On the importance of topics from the mid-ninth century, see
 Konishi, 1989, 188-199.
 
 [48] Teele, 1980, 88-89.
 
 [49] “Bit by bit a past work of literature will come to refer to one
 environment while its readers refer to another.” See J. R. de J.
 Jackson, Historical Criticism and the Meaning of Texts. London and New
 York: Routledge, 1989, 39. Jackson provides numerous diagrams of the
 historically changing relationships between the socio-literary
 environment, the language, the text at hand, and the readers of
 different time periods. See Jackson, 24, 40, 45-49, and 52. Though
 paying less attention to the historical dimension, a more nuanced
 approach to intertextual differences of context for pretexts and texts
 at hand is suggested by Hebel’s summary of Ben-Porat’s four stages of
 processing allusion; these are: “recognition of a marker,” which is a
 “directional signal ... always identifiable as an element or pattern
 belonging to another independent text,” “identification of the evoked
 text, modification of the initial interpretation of the signal,
 activation of the evoked text as a whole in an attempt to form a
 maximum of intertextual patterns,” Hebel, 1991, 137-38.
 
 [50] Ōkubo Tamotsu, ed. Kaika shindai kashū. Kinkadō, 1878, 2.
 
 [51] See Todorov’s The Fantastic for an outline of this mechanism in
 literature.
 
 [52] Yamada Bimyō uses pivot words in the narratives poems in Shōnen
 sugata, though even these can be attributed to gesaku shōsetsu models,
 not waka, and are themselves atypical shintaishi. The self-published
 Yuasa Hangetsu, though not very influential, had some training in
 waka, and lines in his monogatari-shi occasionally include kakekotoba,
 and often mix waka diction with his Christian subject matter. See
 Yamamiya, 1950, I: 115-133.
 
 [53] Kamei Hideo, “Kindaishi sōki ni okeru kōsei no mondai,” Bungaku,
 January 1984, 6.
 
 [54] See Kenmochi Takehiko. Chōshikei to tanshikei, in Kōza: Nihon
 gendai shishi, Vol. I, ed by Hara Shirō, et al. Ubun Shoin, 1973,
 163-64.
 
 [55] Kenmochi, 1973, 164.
 
 [56] See Yamamiya, 1950, I: 59 and 75.
 
 [57] For an example see the well known long narrative chōka/shintaishi
 by Ochiai Naobumi, Kōjo Shiragiku no uta (Song of the Dutiful Daughter
 Shiragiku, based on a kanshi by Inoue Tetsujirō), published in several
 installments in the journal Tōyō gakkai zasshi in 1889-90 and included
 in Ōmiya Shūji, ed., Shinsen chōkashū. Tōkyōdō, 1892, 49-72. It is
 also in Yamamiya, 1950, I:145-158.
 
 [58] Thomas Campbell, The Complete Poetical Works of Thomas Campbell,
 With a Memoir of His Life. New York: T.Y. Crowell, 1877, 147-148.
 
 [59] Toyama Masakazu uses the phrase “connected thoughts” twice in his
 preface to the Shintaishishō. See Yamamiya, 1950, I:25.

 Versions of this piece appeared in the author's dissertation, Intertexts for a
 National Poetry: The Ideological Origins of Shintaishi (New-
style Poetry). University of Chicago, March 2003, and in Japanese Poeticity  and Narrativity Revisited, Proceedings of the Midwest Association for Japanese Literary Studies, Vol. 4 (2003), 166-83.