[In the following essay, Kamens offers an interpretive analysis of ten memorial waka from Teika's Shui guso collection, concentrating on the allusive intertextuality of these works.]

Most Japanese poems (yamoto uta, waka) have their seeds in Japanese poems and are likely to flower forth as still more Japanese poems. When Japanese poets compose waka, those acts of production—like those of other poets in other cultures—are enabled as well as constrained by the poem-composing acts of their antecedents. Whether this condition is demonstrated or acknowledged, or one of which the poem-producer is even partially or fully aware, it is nevertheless in force in the making of the poem itself and definitive of the poem-composing act. Thus, virtually every poem in the Japanese corpus is an inscription over another or other previous inscriptions, and it creates or extends the ground over which still more poems may be inscribed. This is another feature “Japanese poetry” shares with all other poetries and many other kinds of writing.

Articulators and advocates of an ideology of Japanese poetry have advanced and celebrated for more than a millennium the tautological notion that Japanese poetry is what it is because it is Japanese, and that poem-composing acts in this tradition are best understood as transformations of the (Japanese) emotions of an identifiable moment or moments of sensory or psychic experience into readily decodable language that then becomes a permanently accessible representation of those (Japanese) emotions and moments. But much of what is to be found in the corpus of waka texts undermines these claims with evidence of a host of conditions that mediate or distance many poems from their putative originating moments of emotion. Most waka poems are far more performative than the traditional characterization of the tradition may suggest: in many cases, especially when protocols calling for the poet to address a particular topic (dai) or to adhere to a programmed sequence of topics are in force, the poem-maker and/or the poem itself mimes the act of “expressing” that which might be felt in certain emotional, social, or sensory circumstances that may themselves be more or less (if not utterly) remote from anything the poem-maker has ever known, except through the medium of poetic texts.

Some readers of this corpus may decline to privilege the nativist expressive-lyric ethos of Japanese verse as the necessary precursor to any understanding of a poem or poems in this tradition, and instead approach the waka canon in whole or part as a collectivity of texts that are open to interpretations of other kinds, just as other texts are. Such readers may then observe something such as the following as part of an emerging alternative characterization of the prevailing conditions that shape much of the production of waka: that in the great majority of poem-making moments the salient condition of the poem-maker—no matter what else we may be led to believe by ancillary texts or ideologically determined assumptions—must often have been, at some point in the production process and on some level of consciousness, an awareness of the simple but significant fact that the composer was making a poem in the likeness of other poems.

Indications of other kinds of circumstances, conditions, motives, attitudes, or postures that may have informed the making of particular poems are sometimes present and discernible in the texts of poems themselves, and they are also frequently found in ancillary texts that may accompany them, or may be inferred from context(s). The most common types of such ancillary indications are those found in prose prefaces to individual poems or groups of poems in various kinds of anthologies, which are the media in which we most often encounter waka as texts. They direct readers toward the imaginary reconstruction of an infinitely replicable narrative that treats the poem at its core as a climactic fulfillment of the potential inherent in the possibility of poem-making. The texts of the waka tradition and many others in related genres, including criticism, have taught us to shape this narrative as a tracing of the trajectory from a moment or moments of emotional experience (kokoro) toward an articulation of its impact or recollection in poetic language (kotoba). This narrative provides the foundation for countless reiterations that directly link putative sensory and psychic experience—what the poet reportedly or supposedly saw, heard, did, or felt—to literary artifacts, which are the poems the poet could not then do other than produce and which immediately become sealed but transparent containers of the image-memories of those sensory and psychic experiences. Almost every account of the making of waka poems is in one way or another a variant of this foregrounded and permeating narrative, but surely there are other ways of reading and thinking about waka production and waka reading. Only relatively recently have scholars begun to write about waka as something that is not always and not only an “expression” of emotion.¹

In addition to ancillary framing texts, the other most important medium through which we encounter what we may detect to be traces of a poem's formative process and interpret as an indication of one poem's special relationship with another or others may be one or another of a variety of allusive gestures performed within the poem. While on the one hand we may say that, given the intertextual nature of waka inscription, all poems in one sense or another allude to some other or all other poems, we can also say that allusion often operates in a narrower sense, in that the makers of some poems may explicitly signal a more direct citation of one or more particular poems (or other kinds of texts) within their poems, through various allusive strategies designed to be relatively readily recognized as such by an audience of hearers or readers of those poems.² We hear or see these signals, and follow where they lead; but in fact we cannot necessarily infer from their presence a great deal about the precise operations of the consciousness of the maker of such allusive poems at the moment of their making, other than this: that there must have been, on some level of consciousness, an awareness, and an intention, of making the “new” poem in the partial likeness of another. I have suggested elsewhere that we can also detect in such textual orchestrations the elements of a design for eliciting in the hearer or reader of the poem a spasm of (usually pleasurable) recognition of the familiar that coincides with and enables the moment of (presumably) satisfied reception of the poem by its intended audience, whoever they may have been.³

But is this really so? Just as I wonder what may happen if we attempt to read waka poems without first imposing a confining narrative-of-origins as a frame around them, I also ask what may happen if we think of allusive gestures not only as affirmative, enabling, empowering, or validating signs but perhaps also, at least in some cases, as signs of constraining, delimiting, or chafing forces at work in the process of poem-making. The meaning of the allusive gesture is not always (or perhaps ever) transparent, but it does have significance in and of itself. Reading to understand those significations, and with particular attention to instances of poem-making that appear to have taken place under specific programmatic protocols—thus distancing these particular instances to a considerable degree from any imaginable simple sequence of “feeling” finding its way unproblematically into “words”—I find myself looking beyond or away from the hoary kokoro-to-kotoba narrative toward another representation of these creative acts of the past to which none of us was witness. Toward this alternative scenario, many of the details of which we shall never be able to fill in, I would offer the suggestion that among the conditions prevailing in many if not most of the moments of traditional Japanese poem-making across time were at least these two that are likely to have been in complex tension with one another. One might be a sense of exhilaration in having the opportunity to perform the compositional act in accordance with the expectations of both poet and audience, but another condition imposed upon those same practitioners might be a sense of oppression by the burdensome but enticing task of making a new poem in the recognizable likeness of another, others, or all others.⁴ Toil in the field of waka production could produce many rewards for the toiler (such as recognition by and intimacy with the elite, and in some cases advances in status, rank, and authority that brought fiscal benefits as well), but the most conspicuous and widely circulated of the fruits of these labors were poems, poems, and more poems.

I have previously written about so-called devotional Buddhist poems, and among other things my aim was to show what such poems have in common with others in the waka canon, and to de-emphasize to that extent the things that might seem to set them apart.⁵ Here I argue, further, that in a sense all waka poems are devotional: they pay homage to all others and to the practices of waka-making and reading themselves. But is this incessant paying of homage always an affirmative and productive act of reverence, awe, and emulation? Might it be not only a sign of a triumph of continuity, survival, and successful transmission but at least sometimes the mark of a defeat, of a resigned acceptance of collective confinement in a constantly echoing sepulcher crammed full of words?

In a particular group of poems by Fujiwara no Teika (1162-1241) that I discuss here, the poet creates a program of allusive gestures designed not only to link those poems to specific others but also to embody a textual acting-out of homage for artifacts and entities that are fragmentary relics of a remembered and accessible if distant past. Should we understand this performance as what it appears to be—a votive act of resonant incantation and reinscription of the words and names of waka poets of the past carried out for the stated purpose of generating merit on their behalf? Teika's poems resuscitate the dead and their words and then seem to lay them to rest again by offering them new poems cast as prayers for their repose. But having been thus exhumed, can these resuscitated names and words find rest? Were they indeed at rest before Teika's intervention, or were they already caught up in a constant dead-end zigzag ricochet against the confining walls of the spaces of poetic memory? And in what condition are they left when Teika is finished with them—are they pacified and neutralized, or still suspended in their susceptibility to more and further disturbances? Performing as acts of liberation, as they portray the rescue of once-worldly figural fragments of earlier poems and transport them and their makers, in memoriam, to new states of static bliss in a figural paradise, do these poems really free anything or anybody? Or is this but one more set of verbal circumambulations inside a mausoleum that has no exit?

THE SOTOBA KUYō POEMS

naki hito no na wo onoono torite sotoba kuyō su tote, hito no susumeshi uta

(These are poems I was asked to compose when the names of several dead persons were selected one by one for a sotoba offering.)

Ten names (of eight women and two men) and ten poems follow the preface reproduced above in the “Buddhist poems” (Shakkyō[ka]) section of Teika's personal anthology, Shūi gusō.⁶ These poems are not found anywhere else—that is, in no other inclusive or selective anthology and, to the best of my knowledge, they have very rarely attracted comment from readers of his works. They attract my attention for several reasons. This particular group of poems (Shūi gusō numbers 2770-79) has both an adjacent framing text offering information (to be interpreted as well as we can) about the circumstances of the poems' production and a set of allusive gestures (deployed there to be recognized as such and as readily as we can) that explicitly link these poems to others. In addition, the framing text (a brief, somewhat ambiguous but in other ways highly specific kotobagaki, or prose preface) overlaps with and into the poem texts as they unfold in the context of this particular minisequence: the names of the “dead persons/poets” are explicitly present in the text, while their poems murmur in the background, making themselves heard through fragmentary citations in the foregrounded “new” poems. The allusive gestures that are these poems' most prominent features are also amplifications or realizations of the particular compositional project described in the framing text. And the project does seem to be a devotional one, or at least an acting-out of what would seem to be a Buddhist devotional act: the inscribing of names and prayers on memorial placards (sotoba), a practice that can be documented in a number of roughly contemporary sources including some other waka and kotobagaki texts (as I show below). The purpose of making, inscribing, and then offering such constructions is to generate merit, which (as many Buddhist scriptures explain) is inherent in the act of creating any kind of stūpa, and the anticipated outcome of any act of kuyō is ekō, “redirected merit.” In this case the apparent beneficiaries are the ten named naki hito themselves: the positive energy created through the proposed or realized production of these sotoba and the making of these poems is ostensibly all for them. Or is it?

Read as we now find them in the context of Teika's collected works, these sotoba kuyō poems also constitute a record of a brief but telling exercise in which Teika again exploits the potential of “allusive variation” (honkadori) for generating new signification. Most scholars of waka think of Teika as the honkadori poet par excellence, and with good reason: like many of his contemporaries, he was much inclined in various compositional circumstances to deploy this particular strategy in order to produce particular resonances between his poems and others that formed the foundation (moto) of waka discourse as they knew and made it.⁷ Teika's own use of the “honkadori strategy” is closely related to his (and his contemporaries') intensifying interest in what we now think of as the “classics” (koten), which he made manifest in many other forms, including the collection, copying, and editing of a number of core texts of both poetry and prose.⁸ The accelerating occurrence of allusive gestures toward koten texts in twelfth-century verse constituted part of this process of establishing and stabilizing the canon of texts that would serve as the foundation (moto) for subsequent literary reference.⁹

In these sotoba poems (which are the only ones in his oeuvre), Teika deployed this strategy as a means toward fulfilling the challenge or potential posed in the specific project, which inherently directed his attention to specific “dead persons” and their poems as targets and beneficiaries of his remembrance. At the same time, of course, the sequence represents and recapitulates the more generalized literary project—to which Teika contributed in so many ways—of gathering for citation and redeployment the scattered and fragmentary texts of selected “classics” in settings wherein they might be encountered again from ever renewed perspectives. Though Teika, his father Fujiwara no Shunzei, his cousin Fujiwara no Yoshitsune, his ruler Emperor Gotoba, and many other poets of their day deployed this strategy to such a wide extent that honkadori has come to be seen as one of if not the most definitive characteristic of their collective style, its deployment in this particular project was especially fitting.¹⁰ Teika's allusive gestures here are not just more honkadori in what seems to be an endless stream of reference and cross-reference, but as such they are themselves offerings to the dead (kuyō) embodied in the chanting and reinscription of their names and their words. And, as such, the offering promises to produce, through honkadori's transmutations, a transmittable store of merit, to be projected across space and time toward a realm in which the spirits of the dead hover and await such beneficent attentions.¹¹

The very act of making the sotoba placards on which these poems were to be inscribed was a small-scale version of the many other instances in Buddhist history of the construction and dedication of full-size stūpas (also sotoba)—structures of many kinds in various regional expressions of Buddhism which generally share the function of housing the holy relics of the dead (especially those of “the Buddha” or of “Buddhas”). In his notes on these poems, Kubota Jun suggests the sotoba in this kuyō were wooden placards (ita) in stūpa-like shapes on which the names of deceased persons would have been inscribed.¹² Such stūpas or sotoba—miniature replicas of a type of monument that, in many different forms, has played an extremely important role throughout the history of Buddhism (especially in its earliest stages in India)—are familiar to anyone who has seen a modern butsudan or Japanese cemetery. They often bear the posthumous Buddha-name of the memorialized individual, often inscribed in bonji, Sanskrit characters that few lay persons can decipher but which are nonetheless recognized as powerfully significant, a necessary element of the enshrinement of the mortal remains—for these unread texts function silently as ever-repeating prayers for the ōjō of the dead, their rebirth as Buddhas in another realm. Teika reports that he prepared these poems as part of the production process of a set of devotional or ritual objects that are now, as perhaps they were then, familiar if specially purposed furnishings of the spaces of everyday material life.¹³ The poems he composed on this occasion are likewise ordinary and unique, made as he made hundreds of others out of materials at hand but also specifically configured for this particular application: for reverent inscription—not in quasi-esoteric bonji but in the more readily decipherable script of waka text—on a set of real (or figurative?) sotoba dedicated to ten representatives of waka's dead.

In the form of the Japanese funerary sotoba or placard, the three-dimensional stūpa that was a housing for holy relics has become a flattened surface that contains only text but still marks the location where the dead's remains are stored and where memories of the dead are focused. Teika's sotoba poems also memorialize the dead through the creation of new forms that contain and celebrate their residual traces; and like bonji inscriptions on many sotoba, which are prayers that the mortal dead may be permanently transformed as immortal Buddhas, Teika's poems also act out a metamorphosis, through honkadori. In a reversal of the “flattening” of the stūpa as sotoba, which results in a loss of volume, the creation of these poems adds or restores surfaces or dimensions for the contemplation of the viewer or reader. They take for their foundations texts that are both “absent” (in that they are not explicitly inscribed) and “present” (in that transformed elements of them reappear in Teika's own poems), just as the memorialized subjects are absent—their spirits dwell in some other realm—and “present”—in memory and through the reinscription of their names and some of their words.¹⁴ And just as the stūpa or pagoda that contains the Buddha's relic may be regarded both as vessel for that relic and as a reembodiment of that Buddha, replete with all its powers, so does Teika's exercise produce poems, formed of language that is both of the past and of the present, that are at once dead and alive, inert and also pulsing with latent energy. Teika's intervention awakens this energy for the apparent purpose of transporting, transforming, and pacifying the spirits of the named dead, through the medium of more poems. But the cacaphonous honkadori ricochet now includes both the dead's old poems and his new ones. Can the dead find rest amidst such constant disturbance?

And why is Teika intervening in the first place? He seems to be doing so at someone else's instigation: he tells us these are hito no susumeshi uta, compositions suggested or commissioned by another. Kotobagaki that appear with several other poems and poem sequences in this same Shakkyō section of Shūi gusō also contain the phrase hito no susumeshi uta, or something quite like it.¹⁵ This seems to indicate the role of a patron or some other instigator in these exercises, one who is perhaps purposefully unnamed. It also seems possible that this other hito may have been the organizer of a group project, a collective homage-offering of poems that follow the same program as Teika's (even though we cannot find other identical sets of poems among his contemporaries' works). This is suggested by the expression ono ono torite (“selected one by one”) in the kotobagaki text: the adverbial ono ono may refer to the one-by-one choosing of the ten memorial subjects out of hundreds or perhaps thousands of koten candidates, but it also may refer to that selection by each of the members of an unnamed collective of participants who shared with Teika the plan to carry out this offering (sotoba kuyō su tote …). If only Teika's diary included an account of such a scene—a cluster of coterie poets, Teika among them, drawing lots with the names of various naki hito on them, and then composing and sharing their honkadori offerings: but the extant versions of Meigetsuki of course have no such entry, and this scenario can be nothing more than an imagined one.

As I have already noted, there are no other sotoba poems as such in Teika's oeuvre, but there are some in texts that Teika is either certain to have known or can be said to have been likely to have known about. Most obvious, perhaps, is a poem in the imperial anthology Shin kokin wakashū, of which Teika was an editor (as member of the group of court poets appointed by Gotoba to select and order its contents). Poem 843 in Shin kokin wakashū, by Hōkyō Gyōhen (1181-1264), is presented in its preface as one he composed as he inscribed not the names (na) but the numbers (kazu) of the dead on a sotoba (naku naritaru hito no kazu wo sotoba ni kakite, uta yomihaberikeru ni).¹⁶ Nothing else seems to be known about what might have been the circumstances or the occasion on which this monk, who apparently had some affiliation with the Kumano religious complex, made this sotoba and this poem.¹⁷ Teika obviously had a hand in placing Gyōhen's poem among others associated with graves and memorial rites in the Aishō (Laments) section of Shin kokin wakashū, and he had at least some personal contact with Gyōhen.¹⁸ But the absence of dates for either his sotoba kuyō project or Gyōhen's mass memorial (was it for a number of persons killed in military strife, or from infectious disease?) make it impossible to speculate about which composition pre- or postdates the other.

Examples of waka said to have been written for inscription on sotoba are few, but of those from both before and after Teika's time one can say it is more common to find examples that are memorials for single persons than for groups. The personal anthology of Kakugō (active as a poet circa 1177-89) includes a poem he inscribed on a sotoba and placed on the grave of a child or youth (warawa) “who had lived in the Hirosawa district” of the capital and had died suddenly of an unnamed illness¹⁹; and the kotobagaki for poem 437 in Shoku shika wakashū says Shin'in Kōzuke inscribed it on a sotoba at her mother's gravesite.²⁰ It is notable that all of these examples—Gyōhen's, Kakugō's, and Kōzuke's—are linked to mourning for the relatively recent dead, while Teika's sotoba kuyō seems to be a kind of retrospective “virtual” mourning, projected across time and space toward the memory of the more remote dead of waka's past: their physical remains are not present for these sotoba to mark, as were the ashes of Kakugō's warawa or Kōzuke's mother, and perhaps even in the mass grave of those for whom Gyōhen prayed as well. Teika's subjects' traces are only their names and the echoes of their poems that resound in the new poems he writes alongside them.

We can certainly find other poems among the works of Teika's contemporaries that are likewise said to have been made as retrospective offerings in memory of the poets of the past. Inpumon'in no Taifu (1131-circa 1200) was a distinguished poet in a generation just older than Teika; we know they both participated in a hyakushu project she organized in 1187. In her personal anthology, poem 295 is presented as one she wrote when Shun'e (1113-circa 1195, the central figure of a coterie known as the Karin'en, in which she was active) instigated a collective Ipponkyō kuyō—the offering of a complete copy of the Lotus Sūtra prepared one chapter at a time by each of several participants, and, in this case, dedicated to the memory of certain unnamed “poets of old” (furuki utayomi). Taifu's assigned chapter was the Yakuōbon, frequently cited as one that had special appeal to women (it contains a promise that they will overcome the handicaps of their sex and become Buddhas, like men), and her poem celebrating the dharma's saving power can be read as thanks for its particular promise to save women like herself:

yo o umi ni shizumihatete ya yaminamashi
          watasumi nori no fune nakariseba(21)

We surely would have drowned and perished in the troubled waters of this world of grief if we had had no dharma-boat to carry us over these vast seas.

Of course, the phrase furuki utayomi is not marked for gender, and the thanks given by Taifu here may be for a more general promise of salvation, as well as for her own. But another sutra-offering (kyō kuyō) in which she participated, organized by [Ukyō Gondaifu] Minamoto no Moromitsu (another Karin'en intimate), was specifically dedicated to “old female poets” (furuki onna utayomi), otherwise unnamed. Her poem on this occasion refers to the Fugengyō, which was conventionally treated as an epilogue to the Lotus Sūtra in such programs:

kokoro yori musubi okikeru shimo nareba
          omoitoku hi ni nokorazarikeri(22)

Since the encumbering frost of ignorance was of our own making, it was sure to melt in the warmth of this sun, leaving no trace.

The use of the figure of melting frost in a Fugengyō-based Shakkyōka is utterly conventional—a trope from the standard rhetorical repertoire of verses addressed to or based on passages in this sutra.²³ Taifu breaks no molds here, nor does it seem likely that anyone would have expected her to do so. Later, we see Teika using the same figure of melting frost in the first of the sotoba poems. Although he makes no explicit gesture to any particular scriptural source in any of those poems, his use of such familiar Shakkyōka tropes throughout the series is one of the means by which he links them to the conventions of Shakkyōka poesy, while at the same time his honkadori gestures link each poem, and the group as a whole, to specific “old” poems and to the traditions of waka as a whole. So one might go so far as to say that the sotoba kuyō poems (like many others, no doubt) consist of nothing but citations and references, of multiple kinds and to acts and texts variously invested with rhetorical clout, which Teika has assembled for yet another performance that does homage to the very conventions of citation and reference displayed therein. These poems mime as prayers for the dead, but they are also self-referential hymns for waka itself.

There were other modes of worshiping the waka past that paralleled and overlapped with those we see in textual spaces like Teika's sotoba kuyō. In the fifth month of 1185, Moromitsu organized another kyō kuyō at the site believed to be the grave of the revered Man'yō-era poet Kakinomoto no Hitomaro, and asked Taifu and a number of other poets to contribute poems for those observances. Known participants and contributors included the monk Jichie of Kōfukuji and the courtier Fujiwara no Nagakata (1139-91, a cousin of Teika). Kotobagaki attached to some of the poems from this same occasion as they appear in an anthology called Naranoha wakashū, compiled by Soshun in 1237, suggest that Taifu herself was the organizer and leader of the rite at Hitomaro's gravesite, and that the party placed a sotoba on the grave to commemorate the event.²⁴ In this respect, and others, the episode is reminiscent of one described in the kotobagaki for the last poem (444) in the personal anthology of the influential poet Fujiwara no Kiyosuke (1104-77):

Hearing that Hitomaro's grave is located in Isonokami in Yamato Province, in front of the temple known as Kakinomotodera, I raised a sotoba there and marked it, “Kakinomoto no Hitomaro's Grave,” and on one side I wrote:

yo wo hete mo aubekarikeru chigiri koso
                    koke no shita ni mo kuchisezarikere(25)

He may have passed on, but the vow that we would meet someday has not rotted away, even under this green moss.²⁶

“Afterward,” Kiyosuke adds, “I heard that a number of villagers there had very strange dreams”—presumably, visions of Hitomaro.

Kiyosuke's poem as well as the dreams of the Isonokami villagers suggest further associations between these activities and another form of reverence for this particular poet: the rites known as Hitomaro eigū. At these formal gatherings, poems in Hitomaro's honor were composed in the presence of what was believed to be his true portrait. Although we know that members of the Karin'en group participated in several such eigū, the anthology of poems they composed on these occasions (collected by Shun'e) has been lost.²⁷ Teika was involved as either contributor or judge (hanshi) in a cluster of eigū utaawase held between 1200 and 1203; the frequency of mention of such events in his diary Meigetsuki and other sources suggests that this particular form of antiquarian celebration of waka as cult, with the image and spirit of a founding genius, Hitomaro, presiding over the making of new poems, enjoyed a special vogue at this particular time and attracted the interest and participation of the Retired Emperor Gotoba, who hosted some of the sessions in his own Tobadono residence and in the Wakadokoro (Poetry Bureau Office) of the Imperial Palace.²⁸ By this stage in its development, the Hitomaro eigū rite, which at the time of its apparent origins in the early eleventh century seems to have been modeled on rites for Confucian sages, was in the process of becoming a more markedly Buddhist-tinged event, with the inclusion of chanting of sutra verses (kada) and the like.²⁹ This trend would seem to bring it into fairly close contextual association with a project such as Teika's (undated) sotoba kuyō, even though there is of course no evidence of a direct connection between that project and his eigū utaawase experiences.

It may thus be possible to place Teika's sotoba kuyō within a mixed context of sotoba dedication and poet worship that includes Gyōhen's, Kakugō's, and Kōzuke's poems, Taifu and Moromitsu's kyō kuyō, Kiyosuke's Hitomaro gravesite rite, the Hitomaro eigū, plus the entire array of contemporaneous kuyō practices.³⁰ But even if I were to argue that the sotoba kuyō poems should or must be seen within such a context, I would immediately have to point out several ways in which the project seems to set itself apart from virtually all of its apparent analogues. For example, the sotoba kuyō may seem to bear a strong resemblance to the sutra-offering-cum-poems for furuki onna utayomi organized by Moromitsu, in that eight out of Teika's ten naki hito are women. But it is also true that Taifu's poem for Moromitsu's project does not appear to use honkadori as its strategy for linkage to any particular “female poet of old,” whereas Teika's poems are as much about the use of that strategy as they are about anything else. Likewise, despite the interest shown by Teika and his contemporaries in paying reverence to and invoking the name, image, and spirit of Hitomaro, in particular, to preside over at least some of their poem-making ventures, Hitomaro is conspicuously absent from the sotoba kuyō subject list. So, for that matter, are Ariwara no Narihira, Ono no Komachi, Ki no Tsurayuki, and any number of other candidates who might seem likely to have been included in the sotoba kuyō if the basis of the selection was an especially high degree of conspicuousness in the waka tradition as a whole.

But this does not seem to have been the case. I cannot posit any plausible rule, rationale, or model for the basis of the naki hito list. There is no significant correlation between the sotoba kuyō selection of names and any of the various formulations of lists of “36 poetic geniuses” (sanjūrokunin kasen) and their exemplary poems that date back at least to the time of the scholar-poet Fujiwara no Kintō (966-1041). (The only concurrent poems are Saigū no Nyōgo's “koto no ne ni …” and Nakatsukasa's “sayaka ni mo. …”) Nor is there a significant correlation between the list of those named and prayed for in the sotoba kuyō and the poets and poems selected by Teika for inclusion in his Kindai shūka and Eiga no taigai (circa 1221), two of his various compendia of exemplary poems of the past: only one of the ten sotoba honka (Yukihira's “wakuraba ni”) occurs among those Teika listed in these works.³¹ Likewise, only one of the ten sotoba subjects (Saigū no Nyōgo) is one of the select hundred represented in Teika's famous anthology of “One Hundred Poems by One Hundred Poets” (Ogura Hyakunin isshu, 1235?)—and the poem selected by Teika as exemplary of her oeuvre in “One Hundred Poems” is not the one he uses as the foundation for her sotoba kuyō poem.

However, some patterns of limited correlation do appear when one cross-checks for occurrences of the poems that serve as foundation texts for the sotoba kuyō group in other kinds of collections. For example, seven of the foundation poems (i.e., those identified as honka or other cited texts in the appendix) can be found in Kokin waka rokujō, a late tenth-century anthology of some 4,500 poems. Teika is known to have possessed a version of this work: Minamoto no Ienaga made a copy of it, based on the version in Teika's hands, circa 1226-27.³² Another nine of the sotoba kuyō foundation poems were included by Teika in his anthology Hachidai chikenshō (also known as Teika hachidai shō or Nishidaishū), compiled circa 1215-16 and consisting of about 1,800 poems from the various imperial anthologies (chokusenshū).³³ (In the appendix, the numbers identifying these poems in these works are those in the versions included in the Shinpen kokka taikan editions thereof.) Given the scale of these anthologies, such correlations are not particularly impressive in numerical terms; but they do give some indication of the extent to which the poems cited in the sotoba kuyō sequence were accessible and in active circulation in the matrix of canonical poems with which Teika was in almost constant interaction throughout his poem-making career. And in this he was certainly not alone. Many of the sotoba kuyō honka can be shown to have a similar prominence in this matrix as some of his contemporaries also interacted with it: for example, the honka for the first, second, and third poems in the sotoba sequence are the readily identified foundations for poems composed by the group of poets (including Teika) who participated in the Sengohyakuban utaawase (“Poetry Competition in 1,500 Rounds”) in 1201-2.³⁴

Even if the grouping of named subjects for the sotoba kuyō poems does turn out to be unusual, Teika's deployment of the honkadori strategy in this instance is, of course, precisely the opposite; that is, it is more like than unlike his manner of composing waka in the array of situations and contexts for which we find evidence in his oeuvres. Shūi gusō poems 2754-65 comprise one set from a project that might offer useful comparison to the sotoba kuyō poems that follow not long after them in the Shakkyō section of the anthology. These were poems Teika says he inscribed on the painted frontispieces of a Hokkekyō rokubu (a six-part copy of the Lotus, Muryōgikyō, Fugengyō, and Shingyō sutras) dedicated and offered in memory of his mother on the first anniversary of her death (in 1194).³⁵ In this project, Teika certainly does deploy honkadori or honkadori-like gestures several times: he alludes at various junctures not only to fairly familiar “secular” poems and prose works while simultaneously maintaining the more explicit dynamic of citational reference to scriptural passages from each of the eight fascicles of the Lotus and each of the additional three sutras, in turn.³⁶ The Shakkyō section of Shūi gusō yields quite a few other examples of “devotional” poems in which Teika makes direct or indirect reference to honka that are not “Shakkyōka” by any means (for example, 2735, 2740, 2750, etc.); indeed, it would be surprising if this were not so. In this sense, his sotoba kuyō is simply one of a number of at least partially analogous instances in which Teika drew upon his detailed knowledge of koten texts and his command of citational strategies and precedents in carrying out an exercise defined both in its broad outlines and in the details of its performance as a Buddhist act of worship—or at least an imitation of one—yet still emphatically marked by its rhetorical gestures to other kinds of verse as one proceeding not in any way apart from but rather in the very midst of his own and many others' acts of poem-making.³⁷ …

CONCLUSION

The sotoba kuyō sets out … to tranquilize the spirits of the dead; but, paradoxically, its allusive gestures disturb their sleep—and our memories—with the noise of poems, old and new. Who then, if anyone, is pacified? Who reaps the benefits of this kuyō—someone absent or present, living or dead? Perhaps Teika, and Teika's practice as waka composer, and the collectivity of waka praxis, are the only beneficiaries. As the language of waka discourse circles back upon itself, encumbered by its own rin'e fetters, it perpetuates but also confines itself within the limits of its own house of words. Like the stūpa that enshrines the fragment of a Buddha's body, or upon which is inscribed all that may be left as keepsake of the dead—his name, or hers, transformed to mark his or her own transformation to Buddhahood—Teika's poems enshrine fragments of others, celebrating them in one more of many reconfigured recapitulations of the very act of making verse. Preserved and reanimated within these new confines, the language of these old/new poems bounces against the walls of the sepulchral housing that has been created for them, which echoes with nothing but the rich and hollow resonance of time.

In his rhetoric of offertory prayer, Teika suggests the passions of the past are now fully effaced, but at the same time, as he displays what he has done with his transformations of these tropes, his new poems reveal that the engines of these ardors are still churning forward and may never run out of energy. What he demonstrates, while acting out the pacification of these spirits, is that the ardent compulsion for making verse, the drive to show how poems beget still other poems, is immune to quelling spells and to the power of images of stasis and cessation. Chanted in the tones of Buddhist faith, masked as celebrations of the ends of lengthy processes, the sotoba kuyō poems offer still more proof that for the words of waka, the fractured fragments of its remembered forms, there shall be no end to their hurtling, spinning, colliding, congealing, and separating movements so long as reliquaries such as these are built to contain, preserve, and display them.

Notes

1. See, for example, Mark Morris, “Waka and Form, Waka and History,” Harvard Journal of Asiatic Studies, Vol. 46, No. 2 (December 1986), pp. 551-610. Thomas Lamarre's Uncovering Heian Japan: An Archaeology of Sensation and Inscription (Durham: Duke University Press, 2000), which directs attention to waka's visual forms, is a more recent example of such scholarship.

2. Edward Kamens, Utamakura, Allusion, and Intertextuality in Traditional Japanese Poetry (New Haven: Yale University Press, 1997). See also Haruo Shirane, “Lyricism and Intertextuality: An Approach to Shunzei's Poetics,” Harvard Journal of Asiatic Studies, Vol. 50, No. 1 (June 1990), pp. 71-85, and David T. Bialock, “Voice, Text, and the Question of Poetic Borrowing in Late Classical Japanese Poetry,” Harvard Journal of Asiatic Studies, Vol. 54, No. 1 (June 1994), pp. 181-231.

3. Kamens, Utamakura, Allusion, and Intertextuality, especially pp. 34-44.

4. This combination of exhilaration and oppression—along with many other sentiments, including humility and awe—surfaces and to some extent shapes the rhetoric of many passages in karon (treatises and commentaries on waka). For example, Teika discusses the challenges and the rewards of what one might call “doing allusion the right way” in Kindai shūka, Eiga no taigai, and Maigetsushō: see Hashimoto Fumio et al., eds., Karon shū, Vol. 50 of Nihon koten bungaku zenshū (Tokyo: Shōgakukan, 1975), pp. 471-72, 494, and 522-23, respectively. Further study of utaawase judgments and poets' journals should also reveal more evidence of these tensions.

5. Edward Kamens, The Buddhist Poetry of the Great Kamo Priestess: Daisaiin Senshi and Hosshin Wakashū (Ann Arbor: Center for Japanese Studies, University of Michigan, 1990); “Dragon-girl, Maidenflower, Buddha: The Transformation of a Waka Topos, ‘The Five Obstructions,’” Harvard Journal of Asiatic Studies, Vol. 54, No. 2 (December 1993), pp. 389-442.

6. Kubota Jun, ed., [Yakuchū] Fujiwara no Teika zen kashū, Vol. 1 (Tokyo: Kawade Shobō Shinsha, 1985), pp. 471-73.

7. “The honka strategy [honkateki hōhō] is a method for creating a work that can be a constituent of the unknown and as yet not living ‘new’ that treats a text that is a constituent of the already known and already extant ‘old’ [ko or kyū] as its foundation [moto]; as a technique that relies on such appropriation, this method can be said to involve the joint operation of both the desire for reciprocity and the positing of opposition [between the ‘old’ and the ‘new’].” Kawahira Hitoshi, “Honkadori to honzetsudori: ‘moto’ no kōzō,” in Waka Bungaku Ronshū Henshū linkai, ed., Shin kokinshū to sono jidai, Vol. 8 of Waka bungaku ronshū (Tokyo: Kazama Shobō, 1991), p. 198.

8. As Kawahira Hitoshi and others have pointed out, however, the term honkadori as such is one Teika never used in his own writings about waka practice or in his judgments at poem competitions (utaawase); nor did he specifically recommend its deployment as a strategy, even though he extensively demonstrated its uses by his own example. Kawahira, “Honkadori to honzetsudori,” p. 206.

9. As Kawahira argues (ibid.), reference to cited or citable texts as honka (foundation poems) or honzetsu (foundation prose passages) is a way of emphasizing the “historicity” (rekishisei) of those texts and an affirmation of their “value” (kachi). See also Nishiki Hitoshi, “Honkadori,” in Koten bungaku retorikku jiten (Kokubungaku, Vol. 37, No. 15 kaiseiban [Tokyo: Gakutōsha, 1993], pp. 142-43; and Matsumura Yūji, “Honkadori kō: seiritsu ni kansuru nōto,” in Waka Bungaku Kai, ed., Ronshū: waka to retorikku, Vol. 10 of Waka bungaku no sekai (Tokyo: Kasama Shoin, 1986), pp. 129-48.

10. Tanaka Yutaka's account of Shin kokin-era poetics focuses on, among other things, the development of honkadori practice within the context of the stylistic “reforms” (kaikaku) forcefully propounded by Shunzei, carried further by Teika, and endorsed by Gotoba. See “Kaisetsu: Shin kokin kafū ni tsuite,” in Tanaka Yutaka and Akase Shingō, eds., Shin kokin wakashū, Vol. 11 of Shin Nihon koten bungaku taikei (Tokyo: Iwanami Shoten, 1994), pp. 587-602.

11. An informal survey of canonical poems written in conjunction with various types of kuyō shows that the majority of recorded examples were composed for kuyō performed in memory of and for the benefit of the relatively recently deceased. Many are written for kechiengyō kuyō, group presentations of copies of various sutras on behalf of the mourned individual—usually, members of the imperial family, close relatives, and esteemed members of the clergy. (Like other kuyō, kechiengyō kuyō were also performed to acquire merit that would benefit their living sponsors in the hereafter; these forward-looking kuyō were called gyakushū.) Occasionally, literary figures of the past were the named beneficiaries: for example, poem 602 by Fujiwara no Muneie (1139-89) in Shinchokusen wakashū (compiled by Teika) was his contribution to a kechiengyō kuyō for Murasaski Shikibu. See Shinpen kokka taikan Henshū Iinkai, ed., Shinpen kokka taikan (Tokyo: Kadokawa Shoten, 1980-92), Vol. 1, p. 271 (hereafter cited as SKT). Other types of kuyō that were often accompanied by or commemorated in the composition of waka include jisshu kuyō (“offerings of ten kinds,” as specified in the Hōshi bon chapter of the Lotus Sūtra), dedication of new images or buildings for Buddhist monasteries, and banquets for assemblies of 1,000 or 10,000 monks. Among the more unusual types of kuyō that included or were commemorated in waka composition were Fujiwara no Michinaga's Hokke hakkō and kyō kuyō (lectures on and presentations of copies of the Lotus Sutra) on behalf of the fish in the Uji River (Gyokuyō wakashū 2712, SKT, Vol. 1, p. 479; also described in Eiga monogatari) and Fujiwara no Kinmori's presentation to a monastery of his late mother's hand mirror with the addition of a bonji (Sanskrit) inscription, acknowledged in a poem by the monk Chōken (1126-1203) (Shoku shūi wakashū 1334, SKT, Vol. 1, p. 385). Generally, waka composed on such occasions commemorate the virtuous act of performing the kuyō after the fact, whereas Teika's sotoba kuyō poems are presented as part of that act. Some additional examples of similar poems are discussed below.

12. Kubota, Teika zen kashū, p. 471.

13. The Tale of the Heike has a memorable account of a variant use of sotoba: in the chapter called “Sotoba nagashi,” the exiled Taira no Yasuyori—an accomplished poet—inscribes 1,000 sotoba with “the Sanskrit letter ‘A,’ his temporary and true names, and two poems.” (Helen McCullough, trans., The Tale of the Heike [Stanford: Stanford University Press, 1988], p. 91.) These are cast into the sea at Kumano, accompanied by prayers; one sotoba washes ashore at Itsukushima and is then taken to the capital, where it is joyously received by members of Yasuyori's family as evidence that he still lives. Thus, the typical funerary or memorial function of the sotoba is utterly reversed. See Kajihara Masaaki and Yamashita Hiroaki, eds., Heike monogatari, jō, Vol. 44 of Shin Nihon koten bungaku taikei (Tokyo: Iwanami Shoten, 1991), pp. 127-30.

14. On “presence” and “absence” in (inter)textual dimensions, see Harold Bloom, A Map of Misreading (New York: Oxford University Press, 1975), p. 71.

15. See the kotobagaki for Shūi gusō poems 2743, 2752, 2753, 2765, and 2766, all of which use variants of the verb susumu or the causative yomasu in reporting the circumstances in which Teika composed memorial or offering verses. Kubota, Teika zen kashū, Vol. 1, pp. 466, 468, 470. Such constructions are not at all uncommon in shikashū and other anthologies of this period.

16. Tanaka and Akase, eds., Shin kokin wakashū, p. 253.

17. Kubota Jun, ed., Shin kokin wakashū zen hyōshaku (Tokyo: Kōdansha, 1976-77), Vol. 4, pp. 332-34. Kubota explicitly compares the circumstances described in this kotobagaki to those of Teika's sotoba kuyō, but notes that in this “generalized” case the sotoba emplacement and inscription are for the benefit of multiple, unnamed subjects. Citing and following Kubota Utsubo's example in Kanbon Shin kokin wakashū hyōshaku (Tokyo: Tōkyōdō, 1964-65), Kubota disparages Gyōhen's poem both for this “generalization” and for its dependence on several puns and kakekotoba turns, which both critics deem inappropriate to the occasion and setting. Disparagement on such grounds is evidence of what would seem to be an ahistorical and overly conservative set of expectations about the matching of form and function in such poems.

18. In his diary entry for Genkyū 1(1204).6.15, Teika mentions that he spent a long evening in conversation with Gyōhen. Imagawa Fumio, ed., Kundoku Meigetsuki (Tokyo: Kawade Shobō Shinsha, 1977), Vol. 2, p. 117.

19. Kakugō shū 84, in SKT, Vol. 7, p. 169.

20. SKT, Vol. 2, p. 313.

21. SKT, Vol. 3, p. 576.

22. Ibid. The poem also appears as Tsukimōde shū 1062, SKT, Vol. 2, p. 354. (Tsukimōde shū is a shisenshū, or privately commissioned anthology, compiled by Kamo no Shigeyasu in 1182; it features the works of many Karin'en coterie members.) Transcription of the poem in Inpumon'in no Taifu shū is slightly corrupt; the Tsukimōde shū version is the one reproduced here. Its kotobagaki simply says that Taifu composed it on the topic “the kokoro of the Fugengyō,” a common format for Shakkyōka dai of this and other periods; here, kokoro means “the essential purport of X as it may be articulated in waka form.”

23. See Kamens, The Buddhist Poetry of the Great Kamo Priestess, pp. 130-31, for a discussion of Daisaiin Senshi's Fugengyō poem in Hosshin wakashū. The figural scheme is essentially the same as that in Taifu's and in many other Fugengyō poems.

24. This antiquarian rite had an interesting sequel: while in Yamato Province, the participants took advantage of the opportunity to visit other locales associated with poets of the past; their chief interest was in the site of what was said to have been the residence of Ariwara no Narihira at a place called Takayasu. The Naranoha wakashū version says Taifu also led the group to the grave of Shōtoku Taishi and stopped at Takayasu on the way. See Morimoto Motoko, Inpumon'in no Taifu shū zenshaku (Tokyo: Kazama Shobō, 1993), pp. 195-202. The kotobagaki to poem 216 in Inpumon'in no Taifu shū mentions another outing to what may be the same or perhaps another putative former residence of Narihira. (Ibid., pp. 178-80.) In Mumyōshō, Kamo no Chōmei reports he was unsuccessful in his own attempts to locate both Hitomaro's grave and Narihira's country residence; the date of his excursion to Yamato in search of these traces is not known. See Takahashi Kazuhiko, ed., Mumyōshō zenkai (Tokyo: Sōbunsha, 1987), pp. 82-84.

25. Kiyosuke shū, SKT, Vol. 3, p. 441a.

26. I am indebted to Ivo Smits for bringing these examples from Inpumon'in no Taifu shū and Kiyosuke shū to my attention, and for help with the translation of Kiyosuke's kotobagaki and uta.

27. Kokon chomonjū tale 5.178, which is based in large part on a document entitled Hitomaro eigū ki (1118), offers a good account of some of the origins and development of these rites. See Nagazumi Yasuaki and Shimada Isao, eds., Kokon chomonjū, Vol. 84 of Nihon koten bungaku taikei (Tokyo: Iwanami Shoten, 1966), pp. 162-66.

28. Katano Tatsurō, “‘Hitomaro eigū’ no hensen to sono wakashiteki igi,” in Katano, Nihon bungei to kaiga no sōkansei no kenkyū (Tokyo: Kasama Shoin, 1975); see especially pp. 175-79. Teika notes his participation in eigū utaawase on various dates including Shōji 2(1200).10.13 and 12.26, Kennin 1(1201).3.16, Kennin 2(1202).5.26, and Kennin 3(1203).6.16. See Imagawa Fumio, ed., Meigetsuki shō (Tokyo: Kawade Shobō Shinsha, 1986), pp. 79, 87, 90, 118, and 135.

29. Katano, “‘Hitomaro eigū’ no hensen,” and Yamada Shōzen, “Kakinomoto no Hitomaro eigū no seiritsu to tenkai: Bukkyō to bungaku to no sesshoku ni shiten o oite,” in Taishō Daigaku kenkyū kiyō, Vol. 51 (March 1966), pp. 83-124.

30. Another partly analogous kuyō project is noted in the personal anthology of Jakuren (1139-1201): Fujiwara no Suetsune (1131-1221, a court poet whose activities often intersected with those of Teika) collected poem draft sheets (waka no sō or wakagusa) and old love letters (kesōbumi) of both furuki and ima no utayomi (“poets of the past and present”), and used this recycled paper for the inscription of a kechiengyō accompanied by verses on the Lotus Sūtra by Jakuren. Thus, the most worldly jottings of waka poets were literally effaced and replaced by the words of scripture and waka poems in their praise. See Jakuren Hōshi shū 183, SKT, Vol. 4, p. 42.

31. Yukihira's “wakuraba ni …” appears in both Kindai shūka and Eiga no taigai: see Hashimoto et al., eds., Karon shū, pp. 478, 506. The honka for sotoba kuyō poems 2 (Nijō no Kisaki's “yuki no uchi ni …”) and 8 (Sotōshi no miko's “kimi ga yuki …”) both appear among the lists of exemplary poems in Korai fūteishō, Teika's father Shunzei's major treatise. (Karon shū, pp. 375, 303.)

32. Inukai Kiyoshi et al., eds., Waka daijiten (Tokyo: Meiji Shoin, 1986), p. 325.

33. Many scholars believe the Hachidai chikenshō selection of poems subsequently served as resource for many of Teika's later compendia of exemplary poems in more limited scope, including Kindai shūka, Eiga no taigai, and Ogura hyakunin isshu. For an edition with some explication, see Higuchi Yoshimaro, ed., Teika hachidai shō to kenkyū, No. 3 of Mikan kokubun shiryō, first series (Tokyo: Mikan Kokubun Shiryō Kankō Kai, 1956-57).

34. See SKT, Vol. 5, p. 489 (poem 2413 in round 1207), 422 (poem 197 in round 99), and 490 (poem 2422 in round 1212): these poems are by Ietaka, Kojijū, and Yasuhide, respectively, and the first provoked fairly extensive commentary from the judge, Kenshō. A full listing of an array of roughly contemporaneous poems that make use of the same honka as those cited by Teika in the sotoba kuyō poems would take up a great deal of space here, but the project can be carried out quite readily using the version of Shinpen kokka taikan available on CD-ROM. See Shinpen kokka taikan Henshū Iinkai, ed., Shinpen kokka taikan CD-ROM-ban (Tokyo: Kadokawa Shoten, 1996).

35. Kubota, Teika zen kashū, Vol. 1, pp. 468-70.

36. Kubota (ibid.) identifies allusive gestures to Murasaki Shikibu's poem “mezurashiki hikari sashisou …,” which can be found in Goshūi wakashū, Eiga monogatari, and Murasaki shikibu nikki, in Teika's poem 2762; to Kanesuke's “hito no oya no kokoro wa …” from Gosen wakashū and Yamato monogatari in poem 2764; and to a prose passage in Genji monogatari in poem 2755.

37. The Shakkyō section of Shūi gusō contains two other examples of poems written on the occasion of memorial kuyō; 2752 was written in conjunction with a kechiengyō kuyō on the thirteenth anniversary of his father Shunzei's death, and 2766 for a kuyō in memory of the late Hossō monk and poet Jōkei, also known as Kedatsu bō (1155-1213). (Kubota, Teika zen kashū, Vol. 1, pp. 468, 470.) As noted above, these are kuyō on behalf of the relatively recent dead, in contrast to the distance in time that separates Teika from the sotoba kuyō subjects. Teika's motivations and objectives—whatever they were—must also have varied accordingly.