Late Edo-period painting by Kubo Shunman of a woman at a writing desk.

Courtesy Metropolitan Museum of Art and Marjorie H. Holden Gift, 2012.

 

INTRODUCTION

Poets asked to define poetry, especially good poetry, often throw up their hands. “I can no more define poetry,” said A. E. Housman, “than a terrier can define a rat.”1 The medieval Japanese poet Shōtetsu (1381–1459) says something similar, albeit less colorfully: “A truly excellent poem is beyond logic.… One cannot explain it in words; it can only be experienced of itself.”2 With that in mind, my title, How to Read a Japanese Poem, may seem a bit audacious. How does one dare to tell people how to read poetry if it is hard to even define what poetry is?

So perhaps my project would be better stated as, “How to begin to read a Japanese poem.” My purpose is not to dictate a destination but to give a few signposts as readers walk the road. This book, then, offers my advice on how to approach Japanese poems in traditional forms, according largely to analytical methods established by Japanese poets, scholars, and critics over the centuries. Obviously there are other ways to approach such a task, but this one seems to me a wise one for beginners.

Each of my seven chapters focuses on separate genres of Japanese poetry and proceeds by analyzing examples in chronological order. In each case, I first give short notes about authorship, along with other details of context, including variously the time of composition, physical setting, social occasion, and textual setting—things to which particular attention has habitually been given by participants in Japanese poetic discourse from the earliest times. Then I move on to a short commentary. My decision to proceed from context to commentary is dictated by a central feature of Japanese poetic discourse—namely, that poems are so often occasional. In other words, Japanese poems were often written in specific situations—social, political, and historical situations, in the broad sense—that need to be described for them to be understood in their own milieu.

Throughout the book, I employ the technical vocabulary of Japanese poetic discourse to frame my comments. Thus readers not familiar with the subject will encounter new technical terms such as kakekotoba (pivot word) and honkadori (allusive variation), the meanings of which should become clear through usage. (For those who want more detail, I have included appendixes.) Beyond that, however, I have not been bashful about using other terms that are of broader scope, such as “metaphor,” “symbol,” and “prosody.” While reducing Japanese poems to some Platonic notion of poetry writ large would be a mistake, not showing ways in which Japanese poetry is similar to poetry in other cultures would be equally foolish.

GENRES

Terminologies always do some harm to what they are meant to represent, and this is particularly true in the study of Japanese poetry. Throughout history, for instance, Japanese poets, critics, and scholars have often used the generic terms waka and uta almost interchangeably in reference to Japanese poetry in a general sense, while confusingly using those same terms to refer specifically to the highly canonical 5-7-5-7-7 genre that remained central to poetic discourse from the 700s to the 1800s. For the sake of clarity, in this book I use uta in reference to the general category of Japanese poetry and waka in the narrower sense of thirty-one-syllable poem—a defensible choice, I believe, however much it rankles the medievalist in me personally.3 The first six of my chapters are in that sense devoted to what might be termed subgenres within uta as an overarching but untidy discourse, in the rough chronological order in which they emerge in history: kodai kayō (ancient song), chōka (long poem), waka (short poem), kayō (popular song), renga (linked verse), haikai (unorthodox poems), kyōka (comic waka), and senryū (comic haikai). I have also included a chapter on Chinese poems by Japanese poets (kanshi; also karauta, in contrast to yamatouta), a genre that was intertwined with Japanese-language poems throughout its long history. If this conception seems confusing I can only say that it is no more so than the reality I am trying to represent.

Kodai kayō 古代歌謡 (ancient song) refers to our earliest Japanese poetic texts, usually called songs (the Japanese term used is, again, and frustratingly, uta) because they were often sung to instrumental accompaniment. The reader will notice, however, that on the page they often look no different from chōka or waka and use many of the same rhetorical techniques. By grouping all these forms—songs, chōka, and waka—together in a beginning chapter I am suggesting that in the earliest times the boundaries between them were not yet solid. However, as it is common practice to refer to these forms as they appear in early texts like Kojiki (Record of Ancient Matters, 712) as “songs” while referring to later examples in Man’yōshū (Collection of Ten Thousand Leaves, ca. 759) and later texts as not songs but chōka or waka, I follow that practice here.

Chōka 長歌 (also read nagauta; literally, “long poem”) are poems of indeterminate length, made up of alternating 5- and 7-syllable lines, concluding with a final 7-7-syllable couplet and often accompanied by a separate hanka, or “envoy,” in 5-7-5-7-7 format. Poems of irregular prosody mostly predate the standardized form as employed in its heyday by Kakinomoto no Hitomaro (fl. ca. 680–700), Yamabe no Akahito (early eighth century), and others around the turn of the eighth century. At its zenith, the chōka allowed for more variety of subject matter than the waka form, but from the late ninth century on the chōka was employed primarily for elegies. Our primary sources for chōka are early histories and chronicles, the Man’yōshū and later imperial anthologies (chokusenshū), and the collections (shū) of individual poets. Poets writing in the form tend to employ parallelism and other kinds of repetitive syntax, formulaic phrasing such as fixed epithets or modifiers (makurakotoba, “pillow words”), and homophones and puns (kakekotoba) that function as syntactic pivots joining two predicates or clauses.

Waka 和歌 (also referred to as tanka 短歌, meaning literally “short poem”) are poems of five lines, following the syllabic pattern 5-7-5-7-7, with a caesura appearing at the end of any line but most often at the end of the third line. Our earliest examples of waka come from the same time as our earliest chōka, and most of the writers of the latter wrote in both forms. The general topics of love and the four seasons dominate the waka canon from the eighth century onward, but other broad topics also appear, such as travel, Buddhism, and lamentation, and poems were also written in the context of correspondence and for ritual occasions. Our primary sources for waka are various kinds and sizes of collections, including those of individual poets, salons, or other kinds of groups, the largest being an entire “court” memorialized in an imperial anthology. Also important are poems inscribed on screen paintings, poems composed for poem contests, and poems contained in prose works (histories, tales, diaries, travel records, critical writings, etc.). Poets writing in the waka form employ parallelism only rarely but do use pillow words, pivot words, and other kinds of wordplay, while also using elliptical phrasing and other devices that allow for semantic expansion, such as borrowing phrases or lines from earlier poems.

Kayō 歌謡 (popular song): Throughout history, poems in Chinese or Japanese were often chanted. However, we also know that popular songs—i.e., lyrics written to be sung melodically or to instrumental accompaniment (kayō)—were a feature of Japanese culture from the earliest times. Four of the major subgenres are saibara (folk song), some of which may have had Chinese origins; kagurauta (sacred song), associated with Shinto rituals; taueuta (rice-planting song); and imayō (modern song) of the sort popular among the nobility from the mid-1100s onward. Later, beginning in the late medieval era, came kouta, or “little songs,” a term that is also used to refer to songs to samisen accompaniment in the Edo period, which for reasons of length I have not included here. Songs often consist of alternating 5- and 7-syllable lines but are of varying length. That we rarely know their authorship hints that they evolved over time and changed greatly as they circulated. Topics are also various, but in a general way Japanese songs tend to express emotions of all sorts, from elation or passion to amazement or frustration. As is the case in other forms, love is a common subtheme, but all one can say beyond that is that songs tend to document the vicissitudes of human existence, often humorously and in dramatic terms.

Renga 連歌 (also read tsuraneuta; literally, “linked verse”) are made up of alternating long (5-7-5) and short (7-7) stanzas, any two of which would make up a waka, in formal terms. At its zenith in the late fourteenth through the mid-sixteenth centuries, the standard form of renga was one hundred verses (a form known as the hyakuin). Generally speaking, hyakuin were composed by a group of poets at a sitting, although for practice and sometimes on social occasions poets often composed just tsukeku (linked “couplets” or simply “links”), which is in that sense the basic unit of renga composition. Renga collections, whether made up of works by one poet or many, generally offer only tsukeku and hokku, or “first verses.” In a broad sense, the subject matter of the subgenre is the same as the subject matter for waka, although some variations in vocabulary and thematic material did occur over time. There were elaborate rules (shikimoku) involved in directing the composition of a sequence, the aesthetic and social purposes of which include thematic variety and highly regulated change. Many full hyakuin remain from the early 1400s onward, and the record also includes two imperially commissioned anthologies and the collections of the works by individual poets. Also important are tsukeku and hokku included in diaries and critical works. Renga poets use all the literary devices mentioned for waka and, as one might expect given the shortness of the form, go farther in the direction of enjambment and elliptical phrasing.

Haikai 俳諧 is a word used to refer to both “unorthodox” and “humorous” sequences and to the first verses (hokku) of such sequences. Links or first verses involving wordplay or bawdy subject matter were produced in renga meetings from the earliest days of that genre, but they began to be recorded and anthologized only around 1500, and then usually without their author’s names being recorded. Not long after that, some renga masters became known for their haikai efforts and developed a new subgenre—haikai renku—“haikai linked verse”—that in time gained its own canons that departed from those for waka and renga and were in that sense unorthodox. The standard venue of composition was the hyakuin, although from around the time of Matsuo Bashō (1644–1694) the kasen, or thirty-six-verse form, gained acceptance. Composition of tsukeku for practice (maekuzuke) was common; later on, as first verses gained independent artistic status, hokku contests were also important. Our source for haikai is again collections of various kinds (although there are no imperial anthologies of the form), along with the collections of individual poets and poems appearing in prose works, especially travel journals and critical writings. Since compiling and publishing (in woodblock-print editions) collections including the work of one’s disciples was a major practice of haikai masters, the written record for the genre dwarfs the written record for earlier forms. Ellipsis, enjambment, and punning and other kinds of wordplay are mainstays of haikai, as are various forms of (sometimes) outlandish metaphor, while allusive variations on lines from classical poems are less frequent. In the work of Bashō and many other haikai poets, the word “unorthodox” applies only in the way the genre allowed common vocabulary unacceptable in formal waka and renga.

Kyōka 狂歌 and senryū 川柳 (the latter also being called zappai 雑俳 and kyōku 狂句) are comic forms of waka and haikai. Humorous poems in Japanese appear from the earliest times, in all genres, but it is in the medieval era that various identifiable genres begin to appear in the historical record. Kyōka refers to comic waka, a subgenre of some prominence already in the 1500s that reached its heyday in the eighteenth and nineteenth centuries. Senryū is a genre that developed within the larger tradition of haikai and refers specifically to haikai tsukeku and hokku without the obligatory season words of their parent genre, composed from the early 1700s to the present. Elaborate punning, parody, and risqué humor are the mainstays of comic poetry, which comes down to us mostly in the form of large collections. Poets writing humorous poems tend to rely heavily on punning and other kinds of wordplay.

Kanshi 漢詩 is used in this book in reference to Chinese poems written by Japanese poets. From early times competence in written Chinese was a virtual requirement for most official service, and the Japanese imperial court (and later samurai governments) used Chinese in court documents. This was also true of Buddhist institutions. It should come as no surprise, then, that some of our earliest collections of poetry by Japanese authors are written not in Japanese but in Chinese or that many poets writing waka also wrote poems in Chinese. This tradition continued among government officials and clerics even during times when for political reasons contact with the continent was minimal. During the medieval period, Chinese poems by Zen monks were particularly prominent, while the same could be said for Chinese poems by Confucian scholars during the Edo period. As to subgenres, Japanese poets wrote in all forms, including the ancient shi (poems of indeterminate length made up of four-character lines), but favored mostly the lüshi (“regulated verse,” containing four or eight lines of five, six, or seven characters) and the jueju (quatrains of two couplets of five or seven characters). Our sources for kanshi include large collections from all periods of Japanese history, as well as the collections of individual poets, poems quoted in prose works, inscriptions, etc. The devices of parallelism and repetition are major features of kanshi, but other devices basic to Japanese poetry—pillow words and pivot words, for example—do not appear.

The canon of Japanese poetry is immense and includes forms not introduced here, such as the poetic sections of Japanese Noh or jōruri plays or epic tales. But I have aimed for variety in every other way. Rather than concentrating only on canonical figures, I have also included lesser-known poets; and rather than only the most well-known genres—the thirty-one-syllable waka and the seventeen-syllable hokku (what we now call haiku)—I have also included poems in Chinese by Japanese poets (kanshi), long poems (chōka), linked verse (renga), comic poems, and a short book of popular songs. Though each of these subgenres has its own discourse, the resonations among them are numerous. My hope is that readers will come away with a sense of the great variety of the canon and the richness and diversity of Japanese poetic expression.

KEYWORDS

Readers will find an abundance of further introductory material in my analyses of specific poems. As an encouragement to those who want to begin with a greater sense of the broader characteristics of Japanese poetry, however, I offer the following seven musings, each illustrated by poems, as enticements to reflection. Basically, my thoughts concentrate on seven keywords that I find myself returning to again and again when reading Japanese poems: courtly elegance, ellipsis, context, intertextuality, Buddhist stoicism, natural icons, and rhetorical play.

Our earliest examples of Japanese poetry were collected by officers of the imperial court, and courtly ideals of rhetoric, decorum, diction, and proper subject matter would remain important throughout poetic history, just as the court would remain a potent cultural force long after it had lost much of its political power. Although the term miyabi, usually translated as “courtly elegance,” does not appear as often as many other aesthetic terms in commentaries and critical writings, there can be no doubt that the ideals behind it—decorous language and subject matter and avoidance of anything smacking of vulgarity—were important in one way or another for nearly all poets writing well into the Edo period. Thus even the warrior-poet Kinoshita Chōshōshi (1569–1649), who is known for being somewhat unconventional, produced mostly poems like this one from his personal collection Kyohakushū (The Kyohaku collection, 1649; no. 1511):

“From among ten poems he composed at a hut named Toba View in the Eastern Hills”

Far and near	ochikochi no
the low slopes of the mountains	yamamoto shiroku
are white with rising mist,	tatsu kiri ni
but on the paddies at Toba	tobata no omo wa
the moon is shining clear.	tsuki zo sayakeki

Chōshōshi was known as an independent figure, but this poem shows that when presented with a famous landscape—the paddies at Toba, now Minami-ku, between Kyoto and Fushimi—he resorted automatically to established courtly rhetoric and imagery from the imperial anthologies. There, a half dozen poems include the phrase tobata no omo (on the surface of the Toba paddies). Furthermore, every phrase in his poem except for yamamoto shiroku tatsu (mountains / white with rising mist) has numerous precedents.

Another poem by the same poet drives the point home. This one, from his Unaimatsu (Grave-marking pine, no. 297), is about the death of a daughter, the sort of event for which we might anticipate a less-conventional response. As it turns out, however, Chōshōshi again turns to courtly habits:

I fell asleep	omoitsutsu
thinking about her, and yet—	nuru yo mo au to
she did not appear.	mienu kana
In this world of ours, the dead	yume made hito ya
go even from our dreams.	naki yo naruran

Here the poet draws on an old trope more common in love poems, where thinking of someone while falling asleep might lead to meeting in dreams. But there is nothing iconoclastic in his conception, which is not a gut-wrenching cry of grief but a model of understatement that invokes courtly habits of mind.

In the late-medieval period and on into the Edo period, courtly ideals faded in significance in many ways. Yet the dai (topics) of new genres like haikai and various forms of comic poetry can often be traced back to courtly precedents, and poets frequently employed parody—always a gesture toward established convention, as a poem from the early haikai collection Enokoshū (Mongrel-puppy collection, 1633; no. 1129) by the haikai master Matsunaga Teitoku (1571–1653) from the early haikai collection demonstrates.

It is the cause	minahito no
of all these people napping—	hirune no tane ya
the autumn moon.	aki no tsuki

Moon viewing was not just a courtly practice, of course, but the centrality of the image of the autumn moon in court poetry cannot be denied, and the humor of the poem obviously plays on a contrast between elegant cause and vulgar effect.

Japanese poems are relatively short, the most famous genres being just seventeen or thirty-one syllables in length. Even the so-called long poems are not truly long. (The longest of the long poems presented in this book, Hitomaro’s poem on passing the ruins at Ōmi, is just forty-seven lines.) Scholars point to linguistic reasons for this. Whatever the cause(s), the fact of brevity remains and is connected to a number of phenomena, the most prominent being a trend toward ellipsis and economy of expression, and minimalist rhetoric. For instance, the many suffixes relating to tense, mood, gender, and status that are so important in classical Japanese prose are mostly absent in poetry. Even grammatical subject and point of view must often be inferred from context. But that is not all. In addition to syntactic and grammatical ellipsis, Japanese poetry often displays semantic ellipsis, as is apparent in even a quite ordinary hokku from Taigi kusen (A selection of hokku by Taigi, 1707; no. 707) by Tan Taigi (1709–1771).

For a thief	nusubito ni
they ring the temple bell.	kane tsuku tera ya
Winter grove.	fuyu kodachi

Here the particle ni (a versatile particle that can be a locative or directional case particle, a conjunctive indicating cause or contrast, a copular particle or an intensifier, and other things) is rather vague. It is historical knowledge that tells us it probably means that someone at the temple is ringing the bell as an alarm, “because of” a thief. And the last line is literally a noun fragment with no predicate. To arrive at an interpretation of the poem one must thus deal with an elliptical statement: because of a thief, someone strikes the alarm bell, at a temple in a grove of barren trees. But so what? Is the poem just description, perhaps of an actual event? The best guess is that the poet wanted to create a human portrait of a thief finding little cover as he attempts to escape, perhaps as a way to symbolize the basic human predicament, and to show hot desperation in the cold of winter—articulating the essence of a topic being one of the chief purposes of poetry. Beyond that, one might claim that the poem is lightly humorous (something we expect of Edo haikai), evoking the image of people in the town around the temple being called out into the bleak winter landscape to protect community interests.

But much remains unresolved. Should we make something of the syntactic break between the first two lines and the last? Do the first two lines indicate that the neighbors know why the bell is being rung, or just the monks? And might it really imply that the neighbors are not yet outside, that perhaps they are only waking, listening, having trouble leaving their warm beds? One must conclude that one product of elliptical expression is a correspondingly high tolerance for uncertainty, as well as for values such as understatement, subtlety, suggestion, and ambiguity. The poem gives us a sound, an image, and a gesture toward the human world but leaves it at that.

As noted, reading Japanese poetry is often helped by a knowledge of context—i.e., authorship, time, social setting and place, and in many cases conventional topic (dai). This does not mean that poems cannot be made to signify without such information, but it does mean that without it some poems may seem opaque and superficial. Knowing context usually adds depth and triggers recognition of metaphor and symbolism. For instance, just the fact that the following Chinese poem from Kūgeshū (Ephemeral flowers, before 1388, p. 10) is by the Zen monk Gidō Shūshin (1325–1388) is enough to prepare us for an allegorical reading:

Sparrow in the Bamboos

He doesn’t go for the grain in the storehouse,

nor does he peck holes in the landlord’s place.

In the mountain groves he can make his living,

perching for the night on a tall bamboo.

Zen poems so often focus on everyday things that the sparrow is no surprise. However, knowing that Zen monks often use allegory makes it easy to see the sparrow as standing for the monk himself or Zen monks in a more general way: not men involved in the worldly affairs of “storehouses” and “landlords” but monks content to make a meager living in “mountain groves,” the latter being a conventional metaphor for Zen temples. (Poetry by Zen monks is in fact called Gozan poetry, or “poetry of the Five Mountains,” in reference to the governing temples of the Rinzai Zen sect.) And if we know that Gidō was in fact not some obscure mountain dweller but a high-ranking cleric, we can go even further and surmise that his poem is not meant literally at all but as a highly literary portrait of the poet as above the administrative affairs that were in fact the substance of his everyday life.

In addition to authorship, time, and place, many Japanese poems also rely on intertextual references (often in the form of explicit allusion, or honkadori) that may be seen as another kind of context. Perhaps this is a universal feature of poetry in the sense that Wordsworth’s sonnets gesture toward Shakespeare and Shakespeare’s toward Petrarch. In the case of Japanese poetry, however, the gestures are frequently so obvious that we are meant to apprehend them as part of our first experience of the poem. Also, in a general way such gestures toward past tradition are a constant of Japanese poetic culture, in the way that certain famous gardens incorporate the “borrowed landscape” around them as part of their expression. Thus a two-verse renga link from Chikurinshō (Poems from the bamboo grove, 1476; no. 1264) by the renga master Shinkei (1406–1475) “borrows” as a backdrop the following passage from The Ten-Foot Square Hut (Hōjōki, p. 22) by Kamo no Chōmei (1153–1216) in which he describes his life in the mountains outside Kyoto. “When the weather is good, I climb up to a mountain ridge where I can see the skies far off over my former home [in the city]”:

Now it is my heart	kokoro no kayou
that journeys at evening.	yūbe ni zo naru
Going out, I gaze—	tachiidete
and cannot forget Kyoto	miyako wasurenu
in my mountain hut.	mine no io

The unknown author of the first verse here was probably not thinking of Chōmei specifically but only of someone or somewhere as evening comes on, alluding most likely to a lover. But in a comment in Shibakusa (Grasses on the wayside, 1470, p. 42) about his “linking poem” Shinkei tells us he was thinking of Chōmei, who wrote so thoughtfully, he says, of how difficult it is to leave memories behind. Hundreds of years had passed since Chōmei’s time, and countless people had retired to the mountains around the city. The shadows of all of them crowd around Shinkei’s lines, all finding it hard to put thoughts of their former lives from their minds.

In Japanese poetry, especially before the Edo period, the basic assumptions about life, its meaning, its challenges and purposes, may in the main be described as Buddhist stoicism, even when a poem is not explicitly religious in content. One reason for the prominence of natural imagery across all genres of Japanese poetry is the belief that the cycles of the natural world teach us fundamental Buddhist concepts such as constant change and transience. This is not to say that Japanese poetry is universally dark and pessimistic. (As Earl Miner pointed out nearly half a century ago, the sense of desolation is often tempered by a spirit of celebration deriving from Shinto animism.)4 But it is to say that the appreciation of beauty or emotion is often tinged with sadness that comes from knowledge that time is fleeting and all human experience illusory (the relevant term is mujō). It is in this sense that one may say that irony is a constituting feature of Japanese poetry. To appreciate the splendor of the full moon is also to recognize that it must wax and wane. A hokku from an early hokku compendium (Hokkuchō, 1614, p. 313) by the linked-verse poet Shūkei (d. 1544) makes the point explicitly.

Falling onto moss	chirite nao
they glow again below:	koke no shitateru
autumn leaves.	momiji kana

Days of glory may end and still come again, but the cycles of the seasons are inevitable: the leaves will fade, as will the moss, in time. Even if nature occasionally tricks us, we know the exception from the rule. The glow of life remains a temporary thing, if more precious for that reason. Rather than despair we confront stoicism.

Anyone reading Japanese poetry quickly notices the prominence of certain recurring natural images, particularly cherry blossoms, the moon, autumn leaves, and snow, that might be termed iconic in the way they point beyond themselves toward large currents of tradition and socio-aesthetic value. And in the same way, one cannot help but notice that the range of subject matter in Japanese poetry, especially before the Edo period, seems rather narrow, involving primarily the four seasons, love, and a few miscellaneous categories such as travel and laments—usually with some human situation or predicament present, whether in the foreground or the background. Rather than dismissing these features of the canon as signifying a lack of creativity, it is worthwhile to see them as revealing the Buddhist idea of change within constancy and to appreciate the challenge of poets to take an established image or idea and somehow “make it new,” even if only slightly, as the following waka illustrates.

“On ‘lingering heat,’ for a ten-poem sequence at the house of the Mikohidari major counselor”

Autumn has come,	aki kite mo
yet still I must use a fan	ōgi no kaze o
to stir the air.	narasu kana
This year the season will begin	kotoshi wa tsuyu ya
with dewfall first of all.	saki ni okuran

In the canons, cool breezes are often the first harbinger of autumn, but the author, Tonna (1289–1372), and his friends at a poetic gathering—who were the first audience for his poem (Sōanshū [Grass hut collection], 1359?; no. 424)—knew that sometimes the natural and poetic worlds are out of sync. Another icon would have to do. The topic “lingering heat” is suggested only by the fan, but that is enough. Such subtle manipulations of convention are an essential value in Japanese poetic discourse, where, as in this case, they become a theme in themselves.

Not surprisingly, given the prominence of Buddhist culture and concepts in the Japanese tradition, there are certain Buddhist-inspired thematic constants that arc through both form and content. The most obvious of these is wordplay, specifically punning, which expresses a strong sense of the artificiality of language. And rhetorically this same emphasis of the ultimate indeterminacy of meaning is evident in the techniques of allusion, punning, recasting, and recontextualizing, this last term referring to taking a poem from an “original” source and placing it another textual setting. Usually the sort of relativism suggested by this scenario was not truly subversive in either intent or effect. Rather, it seems to have encouraged a sense of stoicism that comes across in poems in mild rather than stark ways. It is also noteworthy that in a corpus that involves so much rhetorical play, one of the highest terms of praise encountered in poetic commentary and criticism is ushin, “sincere feeling.” The two terms are not always antithetical, but the tension between them is again a constituting feature of much Japanese poetry. Sometimes the tension is literally the subject of a poem, as in a hokku from Sarumino (The monkey’s straw raincoat, 1691; no. 1861) by the haikai poet Nozawa Bonchō (d. 1714):

Something makes a sound.	mono no oto
Fallen over, by himself—	hitori taoruru
a scarecrow.	kakashi kana

There is something innately humorous (and spooky) about a scarecrow: a trickster and impostor that looks like us but is not. And Bonchō’s poem involves humor as well, arising from the personification involved in saying the scarecrow falls “by himself” (hitori). But can one claim personification when an object was created to stand for a person to begin with? Whatever our response, the brief smile of the speaker ends in lonely silence. If he had hoped the sound might signal a visitor, he is in fact left as solitary as the scarecrow, a creature of wood and cloth who cannot share his feelings. The metaphor is unmasked as metaphor, a likeness and nothing more, creating a poem that is playful on the surface but full of feeling at its core.

Of course the preceding list of key words and concepts is not exhaustive. But looking at a final poem through the lens they provide shows how keeping them in mind can be useful. It is a thirty-one-syllable poem from Kakanshū (The hazy gate collection, 1793?; no. 469) by an eighteenth-century woman we know as the Wife of Yoshimasa, a woman of samurai background married to a bureaucrat.

“Written on the fifteenth day of the Eighth Month”

From a past unseen	minu mukashi
and on into a future	mizaran nochi mo
I shall never see—	kawaraji na
ever unchanging is the light	nadataru aki no
of the renowned autumn moon.	tsuki no hikari wa

A simple poem? Yes, in a way, for it may be read as praise for the beauty of the full moon that contains no metaphors, indeed no figurative language at all. But attention to the key words and concepts listed in the preceding helps tease more signification out of the poem. The references to an undifferentiated past and future, for instance, are obviously elliptical, even with the contextual information provided by a headnote that gives us crucial temporal orientation. Furthermore, there is a natural icon from the courtly tradition of elegant expression at the center of the poem, the moon—in this case the full moon, which gestures intertextually toward a long tradition of poems about that image. Nor can it be denied that the poem expresses a tension between permanence and impermanence that is at the heart of the Buddhist worldview. And finally, the poem is also rhetorically playful in the way it draws attention to how the moon may be unchanging, but we—who are in a moment in time between the past and the present, both beyond our sight—are not. Beautiful but also cold and aloof, the moon shines down from above, renowned because it is not subject to the cycles of life that human beings must endure below.

A NOTE ABOUT TRANSLATION

Japanese poems have customarily been recorded in a variety of formats. Here I have followed practices of lineation that have been dominant among translators for many years. In other words, I break poems into lines using an alternating scheme of long and short lines based on the genre in question—five lines in the case of waka, three in the case of haiku, and so on. (Historically, Japanese poems were recorded on paper in a variety of formats that practical considerations make too difficult for me to reproduce.) Whenever possible, I also try to adhere to the image order of the originals. In order to show continuity across time and genres, I anchor each of the lines of my translations on the left, with no indentation. For the benefit of students who want to see the poems in Japanese, I have included both Romanized and original kanji-kana texts of the poems I analyze in the individual entries that make up the body of the book. Rather than embark on the quixotic task of guessing at the pronunciation of the kanji and kana in each historical period, I have Romanized texts as one would modern Japanese.

A NOTE ON NAMES

In most scholarly and reference works, Japanese poets are referred to not by their full names (clan or family name plus given name) but by their given names or pen names: thus the medieval poet Fujiwara no Teika is usually referred to as Teika, his given name (which confusingly can also be read Sadaie); and the haikai master Matsuo Bashō is usually referred to as Bashō, one of his sobriquets (he had others: Tōsei being the most prominent one). But it should be noted that even a cursory survey of standard works reveals a diversity of practices that makes it impossible to establish a useful rule. (An example: the renga master Chiun, who was of the Ninagawa lineage, is nonetheless usually referred to not as Ninagawa Chiun but as simply Chiun, while the modern poet Masaoka Shiki is often referred to by his full name.) To compensate for the confusion, I have used cross-references in the index.

I should also note that there is not uniformity concerning the pronunciation of many names. Until recently, for example, the name of the renga poet that I have here rendered as Kenzai tended to be read as Kensai; and in the case of another renga poet, Kyūzei, there are two other variants—Kyūsei and Gusai. Insofar as possible, I have followed examples in the most recent reference works.

In the body of this book (chapters 1–7), I give full names in the first instance and thereafter allude to them by their given names or pen names, and I do the same for other people mentioned in the text. In the appendixes, I follow that same practice in each separate entry.

HEADNOTES, TOPICS, AND TITLES

In most original sources, Japanese poems are introduced with headnotes (kotobagaki) that offer various kinds of contextual information, including such things as time and place of composition, along with the poem’s topic (dai) or title, whenever relevant. The distinction between topics and titles is important. Dai were conventional, prescribed topics, usually assigned to the poet to treat before actual composition, whereas titles were often assigned after composition. In this book, I translate full headnotes as often as possible, placing them in quotation marks (“Written on the fifteenth day of the Eighth Month”). I also place dai in quotation marks: “Snow in front of a shrine.” In the case of a dai appearing as just part of a headnote, the dai is given in single quotation marks, as in “From his many poems on ‘blossoms.’ ” For more details on dai, see the appendixes.