SOURCE: Yiu, Angela. “The Category of Metaphorical Poems (Hiyuka) in the Man'yōshū: Its Characteristics and Chinese Origins.” Journal of the Association of Teachers of Japanese 23, no. 2 (November 1989): 7-33.

[In the following essay, Yiu classifies and analyzes hiyuka poetry as found in the Man'yōshū, defining such works as highly metaphorical poems of Chinese origin but removed from their political dimension.]

INTRODUCTION

Hiyuka (metaphorical poems) as a category appears in the Man'yôshû and disappears altogether in later anthologies. Its existence raises interesting questions concerning the classification of poetry at the very beginning of a long tradition of anthologization of Japanese poetry. The three major categories of poems in the Man'yôshû, zôka (miscellaneous poems), sômon (love exchanges), and banka (elegies), are classified by content, while hiyuka and two other small categories of poems, seijutsu shincho (direct expression of feelings) and kibutsu chinshi (relying on things to express thoughts), are classified by technique of expression.¹ The immediate problem that arises with this system of categorization is that poems classified by content unavoidably exhibit the same kinds of literary devices used in poems classified by technique and thereby challenge the inherent independence of the categories. I single out the category of hiyuka in this paper to examine how a category is formed, what constitutes a poetic category, and why hiyuka as a category should disappear in later anthologies while others survived.

The anthologization of the Man'yôshû was largely an experimental process which involved an attempt to define Japanese poetry in terms of existing Chinese categories. Given that the poetry and the exegetical traditions of the two countries are fundamentally different, an interesting process of interpretation occurs along with the borrowing and influence of Chinese poetics. The compulsion to emulate the continental poetic tradition even though it is not perfectly suited to the Japanese mode of poetic expression is discernible in the categorization of the Man'yôshû. Hiyuka exemplifies the problem. The term hiyu, as I will show in later discussion, originates from the most prestigious and authoritative Han and Six Dynasties Confucian exegetical texts, and it is fundamentally a rhetorical device instrumental in realizing the political function of poetry. In the context of the Man'yôshû, the technique of hiyu designates different levels of metaphorical usages which serve to enhance the complexity of the poetic language of love and longing. In other words, the Chinese poetic term of piyu (J. hiyu) is borrowed, forced to fit with the existing Japanese writing, and eventually naturalized.

In the following sections, after a discussion of hiyuka, I will describe the Chinese origins of the term hiyu and analyze its function in the Japanese poetic tradition as opposed to the Chinese context. My purpose is to show that in the early stage of categorization, the Man'yôshû compilers borrowed one of the most prominent terms in the Confucian exegetical texts, took note of its rhetorical quality of entrusting thoughts to objects, and stripped away its essential political nature. That is to say, the compilers omitted the potentially caustic and unyielding political character of the Chinese concept of piyu and concentrated instead on its charm and the flexibility of indirect expression.

THE CHARACTERISTICS OF HIYUKA

The one hundred and sixty-four hiyuka in the Man'yôshû are unevenly distributed among six books, with twenty-five of them in Book 3 (Poems 390-414), one hundred and eight in Book 7 (Poems 1296-1403), three in Book 10 (Poems 1889, 1978, 2309), thirteen in Book 11 (Poems 2828-2840), one in Book 13 (Poem 3323), and fourteen in Book 14 (Poems 3429-37, 3572-76). Hiyuka are mostly tanka except for Poem 1403, which is a sedôka, and Poem 3323, a chôka.² The category of hiyuka itself seems to be a little different in arrangement every time it appears. For instance, the hiyuka in Book 3, the first instance where hiyuka as a category appears in the Man'yôshû, are the only group of hiyuka whose composers are identified. They are mostly composed by Ôtomo no Yakamochi (718?-785) and members of his family, including Sakanoue Iratsume (fl ca 728-746) and Surugamaro (?-776), as well as members of the Imperial household, such as Ki no Himemiko (?-699) and Ichiwara no Ôkimi (fl ca 743-763). The compilers even provide the historical context for some of the poems. This makes the reading of hiyuka an experience in contextualization. That is to say, the reader is furnished with the information regarding who wrote the poem, to whom it was sent, and under what circumstances or for what purpose it was written. The reading of the hiyuka, then, involves linking the metaphor in the poem to the historical context. The hiyuka in Book 7 are presented differently and thus suggest a different kind of reading. The historical context is not provided but the metaphors are identified. Under the category of hiyuka, encyclopedic subcategories are formed, with headings such as “entrusting [thoughts] to clothing,” “entrusting [thoughts] to mountains” (kinu ni yosu, yama ni yosu), and so on. The subcategories of the hiyuka overlap totally with the category of kibutsu chinshi, and this raises the question of why two presumably identical categories should be established. The three hiyuka in Book 10 belong respectively to the categories of “Spring miscellany,” “Summer miscellany,” and “Autumn love,” while the thirteen in Book 11 exist side by side with the categories of kibutsu chinshi and seijutsu shincho. Following each hiyuka in Book 11 is a comment that states the kind of object to which feelings and thoughts in that poem are compared and entrusted (kibutsu yushi). It is extremely unclear how the hiyuka in this book can be distinguished from the poems under kibutsu chinshi. The single hiyuka in Book 13 is listed as one of the five categories in that book, namely the categories of miscellany, love poems, dialogue poems, hiyuka, and elegies. The fourteen hiyuka in Book 14 appear to be part of the Azuma uta (Poems of the East) collection, and each of them is identified as a regional hiyuka, as indicated by comments such as “a hiyuka of the province of Suruga.” All in all, this medley of subcategories, overlappings, and regional identification shows that the nature of hiyuka as a category is rather unformed and ill-defined.

Hiyu is a figure of speech which may variously involve comparison, juxtaposition, analogy, and fusion of images to express an idea or emotion. While other categories are named by content (elegies, love exchanges, miscellany) or form (tanka, chôka, sedôka), hiyuka is so named presumably for the poetic technique of hiyu predominant in the poems. The closest English equivalent to the term hiyu, I think, is “metaphor.” I render hiyuka as “metaphorical poems” to mark the category's emphasis on poetic technique, as well as for the various aspects of metaphor it exhibits, as I shall elaborate in due course. The immediate question which arises with the naming of a category for the poetic technique it exhibits is whether the use of such a technique is exclusive to the category, and if not, what justifies the creation of the category. The fact is, the use of metaphors is so common in poems of all categories in the Man'yôshû that to name a category of poems as metaphorical poems compels one to look for special characteristics of that category that distinguish it from other categories. The compilers also make remarks which lead us to believe that such characteristics exist, as indicated clearly in the comment on poem 7:1375, “The above poem is not a hiyuka.” The appended comment adds that the poem was placed under the category of hiyuka because it was composed by the same poet who wrote the poem that precedes it. But it remains debatable whether the poems distinguished as hiyuka exhibit consistent and distinguishable characteristics that set them apart from the other categories.

Hiyuka as a category was established by Ôtomo no Yakamochi who belongs to the fourth and latest phase of the Man'yôshû poets.³ The fourth phase began in 734 during the reign of Emperor Shômu and lasted until New Year's Day of 759, on which the last and latest recorded poem of the Man'yôshû was composed. It is a phase in which poetry was widely used in social functions.⁴ This phase features the poems of Ôtomo Yakamochi and members of his family. In fact, Books 17 to 20 resemble chronicles of the Ôtomo family recorded in a mixture of prose and poetry. This characteristic is one of the strongest pieces of evidence to show that Yakamochi was one of the major compilers of the Man'yôshû, the compilation of which extended over eighty years and involved a number of compilers whose identities cannot be conclusively established.

In his composition and compilation, Yakamochi shows a sensitivity towards the principles of poetic forms and theories that were available during his time. For instance, he classified the chôka he wrote about landscape (sansui: mountain and water) as fu (exposition). On another occasion, in 753, he wrote three poems on “spring sorrow” (19:4290-4292) and subtitled the second poem as “a poem written according to the stimulus (J. kyô, C. xing),” revealing his awareness of the poetic technique of xing.⁵ Yakamochi's use of these poetic terms in his compilation shows his interest in formulating a language for critical discourse by integrating and interpreting existing poetic theories. It is in this kind of intellectual environment that the category of hiyuka is believed to have been engendered.

A number of critics, including Nakanishi Susumu, Nakajima Mitsukaze, and Edwin Cranston, share the opinion that hiyuka as a category corresponds to sômon (love poems) and replaces it at certain points.⁶ Since Book 1 consists only of zôka, Book 2 of sômon and banka, and Book 3 of zôka, hiyuka, and banka, it is understandable that Book 3 is often regarded as Book 1 and 2 combined. In an essay titled “Man'yôshû no rizumu to imeeji” (Rhythm and Imagery in the Man'yôshû),⁷ Nakanishi argues that the use of metaphor (hiyu) is the special characteristic of love poems, and the transformation of the “image in the heart” (shinshô) into an image of a thing, i.e., a concrete image (busshô), is the fundamental function of imagery in the love poems of Man'yôshû. But regarding hiyuka as a substitute for sômon amounts to confusing a category based on poetic technique with one based on content. Moreover, in Book 13, sômon appears side by side with hiyuka, an indication that hiyuka does not necessarily always replace sômon. In Book 10, the two hiyuka (1889, 1978) are listed under zôka, and this challenges the view that hiyuka correspond to sômon. The critic Morimoto Masayoshi also argues that the last forty-nine zôka in Book 7, from poem 1247 onwards, should be considered as sômon by nature of their content, so Book 7 can be seen as being made up of four categories—zôka, sômon, hiyuka, and banka—instead of three.⁸ Clearly, despite the fact that major and minor categories and even encyclopedic sub-categories were devised, there was still considerable ambivalence on the compilers' part in deciding what sort of poems should come under a certain category. This kind of ambivalence is an indication of the uncertainty implicit in the gestation of possible new forms or genres, and to say that hiyuka is the replacement of sômon is to disregard that complicated process. Furthermore, since hiyuka and sômon exist side by side in certain books, and since their names imply that they are categorically different, one is compelled to look for characteristics of hiyuka which might legitimize its existence as an independent category.

The fact that the term hiyu corresponds to the broad notion of “comparison” leads critics to construct more sub-categories within that given category in an attempt to define hiyuka. Yamazaki Kaoru, for instance, points out that there are “complete hiyuka” and “incomplete hiyuka.”⁹ In a “complete hiyuka,” the metaphor dominates the entire poem, and the intent or feeling (hon'i) is hidden behind the metaphor. In contrast, in an “incomplete hiyuka” the hon'i is exposed directly in the poem. Yamazaki illustrates with the following pair of poems:

7:1390 (COMPLETE HIYUKA)

Ômi no umi
nami kashikomi to
kazemamori
toshi wa ya henamu
kogu to wa nashi ni(10)
In the sea of Ômi
since the waves are fearsome
I wait for a favorable wind.
Will the years pass by
before I ever row out?(11)

7:1400 (INCOMPLETE HIYUKA)

Shimazutau
ashibaya no obune
kazemamori
toshi wa ya henamu
au to wa nashi ni
Like the swift-footed small boat
that sends news to the island
awaiting a favorable wind,
Will the years pass by
before I meet her?

What makes poem 1390 a “complete hiyuka” is that the circumstance of “not being able to meet” is hidden in the metaphor, while in poem 1400, it is spelled out in the last line, “au to wa nashi ni.” Using this standard, Yamazaki divides all the hiyuka into “complete” and “incomplete” hiyuka. In fact, Yamazaki is talking about different aspects of the use of metaphor, and it is clear that the idea of “complete hiyuka” derives from the rhetoric of “hidden comparison” (in'yu), and the idea of “incomplete hiyuka” comes from the rhetoric of “direct comparison” (chokuyu). Chokuyu corresponds to the “prosaic” metaphor which frequently operates like a simile or a predication,¹² and in'yu corresponds to a more complex metaphorical relation in which the tension of comparison exists but is not explicitly expressed. Yamazaki's analysis of the hiyuka to some extent clarifies the kinds of metaphors used in the hiyuka, but runs the risk of forcing the poems into either one of two subdivisions.

Nakajima Mitsukaze uses the same approach but creates even more subdivisions within the category.¹³ He places the hiyuka in Book 3 under the respective headings of chokuyu (direct comparison), in'yu (indirect comparison), fûyu, (comparison that involves indirect criticism), and joka (poem whose opening lines prepare for the introduction of a later idea). However, he is unable to substantiate his analysis with appropriate examples, except for citing the much quoted example for chokuyu, in which the use of the word goto (like) clearly establishes the prosaic metaphorical relation in the form of a simile.

7:1367 (HIYUKA; ENTRUSTING [THOUGHTS] TO ANIMALS)

Mikuniyama
konure ni sumau
musasabi no
tori matsu ga goto
ware machi ya semu
Like the giant flying squirrel
that lives on the treetop
waiting for a bird,
shall I too pine away
from waiting?

In my analysis of the hiyuka, I find it helpful to look at the various uses of metaphors in the poems in an attempt to trace the general characteristics of hiyuka, and although I think some metaphorical relations in the hiyuka far exceed in sophistication those in poems from other categories, I am inclined to share the position of critics like Morimoto Masayoshi and Ogibata Tadao that hiyuka as a category overlaps with a number of other categories as far as its characteristics are concerned.¹⁴ This leads me to speculate that there must be other factors, most likely historical, that contribute to the formation of the category of hiyuka. Before examining historical factors let us look at a few hiyuka.

The tying of a marker (shimeyui) is a frequently used image which originally involved the use of a piece of grass or twig to tie a knot or a rope to mark off an area.¹⁵ An image much used in hiyuka as well as in poems in other categories, it came to stand for exclusive possession, either by the gods or humans, of a sacred place (such as a shrine), a special area (such as a field or a mountain), or a loved one. For instance,

3:394 (HIYUKA) POEM BY YO NO MYôGUN

Shimeyuite
waga sadameteshi
suminoe no
hama no komatsu wa
nochi mo waga matsu
The little pine
on the beach at Suminoe
that I have decided on
and tied my marker to—
later it will still be mine.(16)

The metaphorical relation in this poem is fairly straightforward: the little pine stands for the loved one as well as the object of possession, and the marker is the symbol of possession.¹⁷ The metaphor and the object or thought it enhances or indirectly conveys are logically related or, by convention, share certain similar qualities. Thus the transference of meaning from the metaphor to the object is readily acceptable and comprehensible. This kind of metaphor, where the comparison is direct and applicable, is frequently used in the hiyuka. It is also seen in poem 1336 in which the image of the burning of fields in spring evokes a “burning heart.”

7:1336 (HIYUKA; ENTRUSTING [THOUGHTS] TO GRASS)

Fuyu gomori
haru no ôno o
yaku hito wa
yakitarane ka mo
aga kokoro yaku(18)
He who burns the weeds
on the open plains in early spring—
Is it not enough?
Must he burn my heart too?

There is another kind of metaphor in some hiyuka that is more subtle and complex. This kind of metaphor relies heavily on poetic contextualization. This is not the same as the historical contextualization in Book 3, which provides information about the circumstances or people involved in the composition in order to suggest to the reader a possible interpretation of the metaphor. Instead, when a number of poems constantly use a set of metaphors to refer to a corresponding set of thoughts or ideas, a poetic context is established for understanding or deciphering an individual metaphor within that set. The clothing metaphor functions basically on this level. Poem 1315, for instance, requires interpretation, but the hidden meaning is not readily available unless one finds out what the clothing metaphor conventionally evokes.

7:1315 (HIYUKA; ENTRUSTING [THOUGHTS] TO CLOTHING)

Tachibana no
shima ni shi oreba
kawa tômi
sarasazu nuishi
aga shitagoromo
Since I live on the island
of Tachibana,
the river is far away;
I sewed my underrobes
without laundering the material.

One intuitively links the image of the unlaundered cloth and underrobes to something impure or intimate, but the correspondence is certainly not as explicit as linking, for instance, the process of flowering and fruition to the maturation and consummation of love. It is here that one relies on other poems employing similar metaphors to validate the association of the metaphor with the thought, feeling, object, or event that is indirectly expressed. The association in poem 1315 lacks the straightforward clarity that is present in the set of poems on the tying of a marker, and in many ways that makes the metaphor in poem 1315 more complex and interesting, and urges the reader to search for its meaning in the larger poetic context. Clothing, especially intimate apparel, is a ready metaphor for love or sexual relationships, and in the context of the Man'yôshû, it is often used to stand for clandestine love affairs. A waving of the sleeve by the Prince to Princess Nukada in the murasaki field implies an illicit love affair,¹⁹ and the coarse fabric of wisteria cloth “to which one eventually gets accustomed” traditionally stands for one's acquaintance with another person, as expressed in the poem below:

3:413 (HIYUKA) POEM RECITED AT A BANQUET BY ÔAMI HITONUSHI

Suma no ama no
shioyaki kinu no
fuji koromo
mato ni shi areba
imada kinarezu
Too wide, the gaps between the threads
of the coarse wisteria cloth
worn by the fishers of Suma
when they burn the salt;
I have yet to wear it to my comfort.(20)

Moreover, the dyeing of garments, particularly in one shade of red (kurenai), suggests sexual involvement that should not be exposed, as in poem 1313:

7:1313 (HIYUKA; ENTRUSTING [THOUGHTS] TO CLOTHING)

Kurenai no
fukasome no kinu
shita ni kite
ue ni torikiba
kotonasamu ka mo
I wear the garment,
dyed a deep red,
underneath.
If I wear it outside
will there not be rumors?

Needless to say, the wetting of sleeves (12:2849), undressing (12:2846), and untying of sash (12:2851) are all potent images of love and sexual union, and the recurring clothing metaphor in these poems creates a rich and well established poetic context to elucidate most of the obscurities in less-than-direct metaphorical relationships.²¹

There are a number of hiyuka which exhibit metaphors of even greater complexity. In these poems, the relation between the metaphor and the object, event, or thought of comparison can only be surmised but never firmly established, by virtue of the fact that these metaphors defy one's expectations and that the poetic context for reading these metaphors is scanty. If the poems with direct comparisons, such as the juxtaposition of a flower and one's lover, are considered as simple, prosaic metaphors whose metaphorical relation is based on resemblance, then the metaphors in the poems which I am about to discuss are “essential metaphors” which deal in “a more complex, instantaneous, and even non-logical relation.”²² This idea resembles the concept of in'yu (hidden comparison), and satisfactory application of this kind of metaphor conveys Aristotle's idea of a good metaphor, which is essentially a metaphor that evokes “an intuitive perception of the similarity in dissimilars.”²³ One must add that a complex metaphor challenges and stimulates the reader's imagination in ways that a prosaic metaphor cannot, as in the following poem:

7:1332 (HIYUKA; ENTRUSTING [THOUGHTS] TO MOUNTAINS)

Iwa ga ne no
kogoshiki yama ni
irisomete
yama natsukashimi
idekatenu kamo
Into the craggy mountains
I entered for the first time but
finding the mountains so lovely, ah
how hard it is to get out again.

The general reading habit of interpreting every indirect expression as a metaphor for romantic love renders the mountains into erotic images. But the metaphor in the poem above is sufficiently fluid to invite other interpretations. Entering the mountains can be symbolic of spiritual rapture, as in the case of sudden enlightenment. Even if one regards the mountain as an erotic image, much remains open to speculation as to what entering the mountains for the first time really means. When it comes to interpretation, the mountain metaphor is more elusive than the clothing metaphor in the sense that the latter, given the poetic context, tells us explicitly that it refers to sexual involvement, as in poem 1313. Of the metaphors that I have examined, the mountain metaphor is by far the most subtle and complex vessel of hidden meaning. One is tempted to say that metaphors like this express the quintessential characteristics of the hiyuka, but similar poems with images inviting interpretation as indirect expressions of hidden meanings are also present in other categories; for instance, poem 1092.

7:1092 (MISCELLANY; POEM ON A MOUNTAIN)

Naru kami no
oto nomi kikishi
makimuku no
hibara no yama o
kyô mitsuru ka mo
I have only heard the name—
Like the rumblings of the thunder—
of the Hibara Mountain in Makimuku.
But today, ah, I have seen it!

This brings us back to the original question concerning the creation of the category of hiyuka. If there are poems in other categories that closely resemble hiyuka and share many of its characteristics, then what warrants the establishment of an independent category as such? What are the sources for setting up such a category, and what does the formation of this category say about the process of compilation?²⁴

THE ORIGINS OF THE TERM “HIYU”

The Chinese critical tradition, which emphasizes the pragmatic and political functions of poetry, fits rather poorly with the expressive orientation of Japanese poetry.²⁵ Yet compilers of early imperial anthologies constantly referred to Confucian hermeneutics and adopted the Chinese critical vocabulary to lend prestige and legitimacy to the rising poetic form of waka. This practice of paying homage to the Chinese emphasis on the political ends of poetry is already evident in the formation of the category of hiyuka. The term piyu (J. hiyu) is used extensively in the explication of the terms feng, bi, and xing, part of the earliest Chinese exegetical vocabulary that forms the view of poetry as political criticism. In borrowing the term hiyu to name a category, the compilers of the Man'yôshû showed an awareness of the political dimensions of Chinese poetics, but ultimately the political aspect of the term hiyu was abandoned and only its rhetorical functions were retained in Japanese poetry. In this section, I shall discuss briefly the use of piyu in the context of feng, bi, and xing and its adaptation in the Japanese context.

A. “HIYU” IN RELATION TO “FENG”

The character pi refers, in its broadest sense, to a comparison.²⁶ In the Analects, the role of a ruler is compared to the North Star: “He who exercises government by means of his virtue may be compared to the north star, which keeps its place and all the stars turn toward it” (2:1).²⁷ It is used in the same way in the line, “to be able to take for comparison something that is close at hand” (6:29).²⁸Pi also refers to the act of understanding: “Although the one who speaks is sincere, the one who listens has not understood yet” (from “The biography of Bao Yong,” in the Hou Han shu).²⁹ Finally, pi also means to admonish by way of a comparison or a circumlocution. “Pi means yu (to use a comparison [in order to persuade or criticize])” (Shuo wen).³⁰ “Pi is like a matching object. To match and criticize indirectly” (Attached Commentary to the Yi jing).³¹ The compound piyu (J. hiyu) refers to a comparison, to comparing, or admonishing indirectly by using a metaphor.³²

The term piyu began to occur as part of the critical vocabulary in exegetical texts in Zheng Xuan's (127-200 a.d.) time, and was frequently linked with the idea of feng.³³ In his appended commentaries to Mao's annotation of the Shi jing, concerning the term feng, one of the six principles of poetry in the Great Preface of the Shi jing, Zheng Xuan comments that “to transform by suasion and to criticize indirectly both refer to the use of comparisons (piyu) and not direct castigation”.³⁴ Another exegetical text of the Six Dynasties adds that, “feng means comparison or indirect admonition by way of comparison” (Yu pian).³⁵ The Tang commentator, Kong Yingda (574-648), in his notes to the Shi jing, confirms Zheng's interpretation by adding that “feng can be likened to the wind causing objects to sway. Thus it was called piyu, which means not using direct castigation”.³⁶ Finally, the annotator of the Wen Xuan (Selections of Refined Literature), Li Shan (?-689), comments that the word feng in the phrase “present Sweet Fountain to the throne in order to use suasion” in Yang Xiong's rhymeprose of the same title means “not daring to speak directly [on the subject]”.³⁷

The use of the term piyu in the definition of feng leads to the speculation that the category of hiyuka in some ways corresponds to the principle of feng as described in the Chinese exegetical literature in terms of meaning and function. The late Edo kokugakusha (scholar of national studies), Kishimoto Yuzuru (1788-1846), author of the Man'yôshû kôshô,³⁸ thought that such was the case.

Hiyu[ka] means poem of comparison [tatoeuta], in the sense that it discloses one's thoughts by entrusting them to object[s]. The following [section] consists of love poems, but since they use objects as comparisons, they are different from ordinary love poems. … To understand them as love poems is not sufficiently accurate. As far as hiyuka is concerned, in Zixia's Great Preface of the Mao edition of the Shi jing,³⁹ it is said that there are six principles of poetry—feng, fu, bi, xing, ya, and song. “Those in high position, by means of feng transform those below them, and those below, by means of feng seek to redirect those above them. When the main intent is set to music and the admonition is indirect, then those who speak do not commit any offense, while it is enough for those who listen to take warning”; thus it is called feng.⁴⁰ Li Shan's annotation says, “Transformation by suasion and indirect criticism both refer to piyu [indirect expression] and not direct castigation.” In the [kana] preface to the Kokinshû, one of the six kinds of poetry is soeuta, and this refers to poems that entrust [thoughts to an object] (yosoeuta), which in fact is the principle of feng.

In defining the relation between feng and hiyuka, Kishimoto was mindful of the common grounds of the rhetorical device of comparison in both Chinese and Japanese literature, but appeared not to be able to render his extended quote from the Great Preface on the political function of feng relevant to his explanation of hiyuka. It is dangerous to assume that because the term piyu is used in the definition of feng, hiyuka must therefore be derived from the principle of feng. Furthermore, based on the fact that hiyuka is further divided into encyclopedic sub-categories of entrusting [thoughts] to things (mono ni yosu), Kishimoto made the connection between hiyuka and soeuta as described in the kana preface of the Kokinshû. Actually, the category soeuta itself is a forced parallel to the principle of feng in the Great Preface. Out of deference for Chinese civilization and in emulation of the Chinese poetic tradition, the presumed writer of the kana preface of the Kokinshû, Ki no Tsurayuki, like many compilers before and after him, incorporated some form of Chinese poetics in his discussion of Japanese poetry to enhance the prestige of his preface, but the relation between soeuta and feng is at best tenuous.⁴¹ The forced correspondence of soeuta and feng in turn invites skepticism toward the supposed relation between hiyuka and feng.

One cannot determine the relation between feng and hiyuka without first delving into the complexities of feng. In an article titled “Notes on the Wind: The Term ‘Feng’ in Chinese Literary Criticism,” Donald Gibbs traces the etymology of the character feng as well as the use of the term in Han and Six Dynasties exegetical texts, and the following discussion of the term feng is partly based on his analysis.⁴² The invention of the graph feng itself, with the constituent elements of a sail and a serpent, suggests the idea of motion. The earliest, traditional account of the term in Shuo wen says that “When the wind stirs (causes movement) small creatures are born; the creatures live eight days and then are transformed.” Thus the early definition of feng consists of the idea of the wind as a mover and a transformer of things.⁴³

The term feng is used in naming the first section of the Shi jing, the Kuo feng, “Winds of the states.” These are songs collected from fifteen different states to reflect the voice of the people. “In olden days there were officials in charge of collecting poems, so that the ruler could observe the customs, learn about successes and failures, and reflect upon what was correct” (Hanshu yiwen zhi).⁴⁴ In other words, these poems were gathered by officials for the purpose of political critique.

This idea of the wind as a power that sways and, by extension, transforms and influences things, is further elaborated in the Great and Small Prefaces appended to the Shi jing in Han times. The Great Preface says:

Those in high position, by means of suasive force (feng) transform (hua) those below them, and those below, by means of critical persuasion (feng) seek to redirect those above them.⁴⁵ … Thus, (as for the first principle of poetry), it is called Persuasion (feng).

The Small Preface says:

The word “wind” means “power to change,” it means “teaching”; just as wind moves things, teaching transforms things.⁴⁶

It is clear that feng as a literary term is laden with political connotations, and naturally the art of comparison (piyu) used in the definition of feng is emphasized not simply for its appeal to the imagination and its enhancement of the rhetoric of poetic expression, but also for its effectiveness in delivering a message about ethical propriety in a roundabout and possibly more subtle and persuasive manner, so that it can be used in situations that require diplomatic exchanges, notably in meetings among the heads of states or between a retainer and his lord. In the Analects, it is said that “though a man may be able to recite the three hundred odes, yet if, when entrusted with a governmental charge, he knows not how to act, or if, when sent to any quarter for a mission, he cannot give his replies unassisted, of what practical use is it?” (13:5)⁴⁷ Confucian hermeneutics does not promote the notion of poetry as a form of aesthetic expression without purpose, mission, or ethical concern, but emphasizes the role of poetry in moral enrichment and the improvement of the state. The following poem from the Shi jing, quoted in part, illustrates the political function of feng:

POEM 113

Big rat, big rat,
Do not eat our millet.
Three years we have served you,
But you have not been willing to heed us;
It has gone so far that we will leave you;
We go to that happy land.
Oh, happy land, happy land!
Then we shall find our place.(48)

The Small Preface to this poem says, “‘Big Rat’ indirectly criticizes heavy taxation. The people criticize the ruler obliquely for levying heavy taxes, feeding on the people, as well as for being negligent in administrative duties, avaricious, and intimidating, and liken [him] to a big rat.”⁴⁹ This kind of comparison (piyu), which is thought to be the essential feature of feng, is always connected to political criticism and related to ethics.⁵⁰ It is thus inaccurate to consider hiyuka as a derivative from the concept of piyu that is linked with feng, because hiyuka are not read as “indirect criticisms [of the government] by way of comparisons” in the Japanese exegetical tradition; nor is hiyuka as a category established for a morally edifying purpose.⁵¹ Since there is no evidence that hiyuka were gathered or composed for a political purpose, and since the rhetoric of indirect expression is employed mainly to enhance the linguistic sophistication and enrich the meaning of the poem, it is clear that hiyuka as a category does not express the principle of feng as presented in Han exegetical texts.⁵²

B. “HIYU” IN RELATION TO “BI” AND “XING”

The term piyu had also been used to define the rhetorical device xing. The Huangqing jingjie (The Imperial Qing Explication of the Classics) notes that the term xing in the phrase “poetry can stimulate (xing) [the mind]” in the Analects “refers to making a comparison (piyu)”.⁵³ It is likely that this kind of explication leads critics to refer to the rhetorical devices bi (comparison) and xing (stimulus) as possible sources for the concept of hiyuka.⁵⁴

Bi and xing are aspects of metaphorical usage. Bi refers to comparison based on resemblance, and it corresponds to a simple, straightforward metaphor, while xing refers to a more complex metaphor which is not necessarily based on analogy. In the Zhou li (The Rites of Zhou), Zheng Zhong (fl. ca. 58-76) says,

A comparison makes a comparison to an object, and a stimulus entrusts a situation to an object.⁵⁵

Zhong Hong (469-518) in the Shi ping (Gradings of Poets) writes that,

When meaning lingers on, though writing has come to an end, this is “xing.” When an object is used to express a sentiment, this is “bi.”⁵⁶

Zheng Xuan defines bi and xing on the basis of their function in an exegetical tradition that reads poetry as political commentary,

With a comparison one sees a present failing, does not dare to castigate directly, and selects a categorical correspondence to speak of it. With a stimulus one sees a present excellence, disdains flattery, and selects a good situation to encourage it by comparison.

(Yu, 58)⁵⁷

Kong Yingda reiterates the comments of Zheng Zhong and Zheng Xuan and explains the literary aspects of the terms,

“A comparison makes a comparison to an object”; anything that says “like/as” (ru) is a word that is a comparison. … ” A stimulus entrusts a situation to an object,” thus xing means to arouse (qi): to select a comparison which draws forth the categorical correspondence and stimulates one's heart/mind. In the text of the Classic of Poetry, all examples of using plants, trees, birds, and beasts to manifest meaning are words that stimulate.

(Yu, 59)⁵⁸

Liu Xie (465?-523?) links bi with xing in the chapter entitled “On Comparison and Stimulus” in the Wenxin diaolong (Elaborations on the Heart of Literature),

Therefore, bi means to match, and xing means to arouse. What matches a meaning uses close categorical correspondence in order to indicate a situation. What arouses emotions relies on the subtle to formulate conceptions. Arouse emotions, and forms of the stimulus will be established. Match a meaning, and examples of comparisons will be produced. A comparison stores up indignation to castigate with words; a stimulus links analogies to satirize.

(Yu, 165)⁵⁹

There are two veins that run through the definitions of bi and xing. One of them is concerned with the political function of the rhetorical devices. Bi and xing are part of the larger tradition which sees poetry as a vehicle for self-cultivation and social betterment. It is a tradition that perpetuates a kind of moral and political didacticism that can sometimes be deadly unimaginative and authoritarian; nonetheless, it forms the backbone of the Chinese view of poetry.

The other vein underlines the literary functions of bi and xing. What defines bi—to match, to use close categorical correspondences—points to a simple metaphorical relation as in a simile. What defines xing—to arouse, to rely on the subtle—refers to a complex metaphorical relation that is less direct in terms of comparison and as a result more evocative, potent, and fluid. Actually, since both bi and xing are metaphors, the line between them is often blurred; sometimes they occur simultaneously in a poem, and sometimes it is difficult to determine whether a certain metaphorical usage is strictly bi or xing.⁶⁰

The burgeoning of critical works that comment on both the rhetorical and political aspects of bi and xing was fostered in part by the critical atmosphere of the Six Dynasties. It was a time in which sophisticated discourse on rhetoric, prosody, and style flourished, a time in which literature was not just studied for its didactic use. Yet the Confucian view of literature, far from receding into the background, still held a significant place in the literary milieu of the time.⁶¹ This mixture of concerns for the aesthetic and pragmatic criteria of literature constituted the continental tradition of poetics and exegesis with which the Japanese came in contact. The Confucian critical stance was no doubt the more historical and revered one, and is often quoted in Japanese critical texts for its prestige.⁶² Yet in examining hiyuka in relation to the various definitions of bi and xing, one notices that the rhetorical aspect alone is well endorsed in the hiyuka themselves, while there is hardly any trace of the political aspect of bi and xing.

The metaphor (piyu) that is used to define feng, bi and xing has a political dimension which the same term in hiyuka does not share. Hiyuka speak indirectly of many things: of longing, loneliness, regret, embarrassment, frustration, but they do not speak with the voice of political criticism. The term hiyu, as derived from Chinese usage, was so laden with political meaning that, from the start, it is a misnomer for a group of poems that are primarily love poems characterized by linguistic sophistication. The term was borrowed for its prestige, but in the end it had to be abandoned, like an ill-fitting cloak. Just as the echoes of Confucian political concern in the early Japanese critical works eventually give way to the more indigenous poetic views of yûgen (mystery and profundity) and en (beauty) elaborated in Fujiwara Shunzei's Korai fûteishô (“Poetic styles past and present,” 1197), so does the name hiyuka fade away after the Man'yôshû in favor of categorical headings which better define the range of themes and concerns of Japanese poetry.

Yet despite the fact that the term hiyuka was eventually abandoned, metaphorical poems (hiyuka) are in fact among the most mature and sophisticated works in the Man'yôshû. In that sense, they provided a source of inspiration as well as a linguistic model for subsequent poetic creations. Of the twelve poems in the Kokinshû that are believed to have originated from the Man'yôshû,⁶³ one is from the category of hiyuka (MYS 7:1351, KKS 4:247), and the other one is modeled closely on a hiyuka (MYS 7:1379, KKS 14:720):

MYS 7:1351 (KKS 4:247) (HIYUKA; ENTRUSTING [THOUGHTS] TO GRASS)

Tsukikusa ni
koromo wa suramu
asatsuyu ni
nurete no nochi wa
utsuroinu tomo
I shall dye my robe
with the moon grass
even though after it gets wet
by the morning dew,
the color will fade.

MYS 7:1379 (HIYUKA; ENTRUSTING [THOUGHTS] TO A RIVER)

Taezu yuku
asuka no kawa no
yodomeraba
yue shi mo arugoto
hito no mimaku ni
If the ever-flowing
River of Asuka
should stop,
someone would surely go and look
to try to find out why.

KKS 14:720 (IT IS SAID THAT THIS IS BY NAKATOME NO AZUMABITO)

Taezu yuku
asuka no kawa no
yodominaba
kokoro aru to ya
hito no omowamu
If the ever-flowing
River of Asuka
should stop,
wouldn't someone think
it had deep feelings?

Since the one hundred and sixty-four hiyuka only make up a small fraction of the 4,516 poems in the Man'yôshû, the fact that two of the twelve Man'yôshû poems that appear in the Kokinshû are hiyuka or hiyuka-like suggests that they formed part of the fountainhead of a long and increasingly sophisticated tradition of metaphorical usage.

My examination of the hiyuka as a category provides grounds for some observations about the process of compilation at the dawn of the anthologization of Japanese poetry. First of all, the Man'yôshû's classification is a tentative attempt to classify poems by form, technique, region, and content, and this results in much overlapping and ambiguity among the categories, as I have illustrated with poems in my discussion of the characteristics of the hiyuka. A direct problem that arises from this system of organization is the creation of categories, like the hiyuka, which lack defining characteristics that distinguish the individual categories from one another.

Secondly, in the discussion of the origins of the hiyuka, I grappled with the phenomenon of continental influence. The overwhelming eminence of the Confucian view of poetry was keenly felt by Japanese compilers and critics, but the increasing interest in technical discussion in the Six Dynasties poetic treatises also appealed to their poetic consciousness. One aspect of the Japanese response to Chinese poetics involved a kind of selective incorporation and application. In the case of the hiyuka, the compilers incorporated the critical term piyu into their categorization and arranged as well as composed poems under that category. In doing so, they toyed with the idea of piyu and came up with their own interpretation of the term. As I have shown, they concentrated on the linguistic cleverness of piyu and abandoned its political connotations. As a result, hiyuka turned out to be a far cry from the origins of the term applied to the poems so categorized.

Emptied of its political connotation, the rhetoric of hiyu really does not have a distinguishing edge and shares too many similarities with the metaphors used in poems from other categories in the Man'yôshû, particularly those in the category of love poems. Perhaps this is why hiyuka as a category was absorbed into other categories and disappeared altogether after the Man'yôshû. The two other categories based on poetic technique alone, kibutsu chinshi and seijutsu shincho, also disappeared in subsequent anthologies, while sômon survived as the category of love in the Kokinshû as well as in other collections, and zôka and the seasonal categories retained their indomitable importance in the process of anthologization. Categories based on technique alone could not stand as independent categories because so many poems in other categories shared the same techniques of composition. Thus, the identification of poetic techniques so useful in criticism and analysis was not viable in the process of classification.

Notes

1. The word “entrust” is used here in the translation of the character ki (yosu) in kibutsu because ki does not simply mean “to send [thoughts to objects],” but involves the meaning of “relying on objects to express thoughts” or “trusting to objects.”

2. Tanka are short poems in the syllabic pattern of 5-7-5-7-7. Chôka are longer poems in the syllabic pattern of 5-7-5-7 … 7-7, and the poem length is not regulated. Sedôka are poems in the syllabic pattern of 5-7-7-5-7-7.

3. Since Kamo Mabuchi, critical views of the phases of Man'yôshû composition have varied. For a succinct discussion of the four phases of the Man'yôshû, see the article by Gomi Tomohide and Ono Hiroshi, entitled “Shuyô na kajin to kafû no tankai” (The major poets and the development of poetic styles) under the entry for “Man'yôshû,” Nihon koten bungaku daijiten 5 (Iwanami Shoten, 1985), pp. 559-62.

4. An example of poetry used in social functions is the utagaki organized by Emperor Shômu in 734, a court ceremony in which young men and women met at a gathering and exchanged poems of proposals and acceptance and rejection, possibly inspired by the ancient sex picnics. See Waka bungaku daijiten, ed. Hisamatsu Sen'ichi et al. (Meiji Shoin, 1962), pp. 78-9. Also see Earl Miner, Princeton Companion to Classical Japanese Literature (Princeton University Press, 1985), p. 302.

5. The term xing first appeared in the Zhou li (Rites of Zhou) in reference to music, along with feng, ya, song, fu, and bi. These later resurfaced in the Great Preface of the Shi jing (Classic of Poetry) as the six principles (C. liu yi, J. rikugi). There is a profusion of definitions and interpretations of these six principles throughout the ages, as well as numerous English renderings. For instance, Pauline Yu rendered them as “airs” (feng), “elegances” (ya), “hymns” (song), “exposition” (fu), “comparison” (bi), “stimulus” (xing). See Pauline Yu, “Metaphor and Chinese Poetry,” Chinese Literature: Essays, Articles, Reviews 3.2 (July 1981), pp. 205-24, esp. 213-15. Donald Gibbs rendered them as “folk airs/suasive force”(feng), “Elegantia” (ya), “Elegies” (song), “narration” (fu), “analogy” (bi), and “evocative image” (xing). See Donald Gibbs, “Notes on the Wind: The Term ‘Feng’ in Chinese Literary Criticism” in Transition and Permanence: Chinese History and Culture, A Festschrift in Honor of Dr. Hsiao Kung-ch'uan, David C. Buxbaum and Frederick W. Mote, eds. (Hong Kong: Cathay Press, 1972), p. 292, note 4. For a list of informed Western-language studies of these terms, see John T. Wixted, “The Kokinshû Prefaces: Another Perspective,” Harvard Journal of Asiatic Studies 43 (June, 1983), p. 228, note 32.

6. Nakanishi Susumu, “Man'yôshû no rizumu to imeeji” (Rhythm and imagery in the Man'yôshû), Man'yôshû kenkyû, Gomi Tomohide and Kojima Noriyuki, eds., vol. 2 (Hanawa Shobô, 1975), pp. 101-15; Nakajima Mitsukaze, “Hiyuka mondôka ron” (On metaphorical poems and dialogue poems), Man'yôshû kôza 6 (Shun'yôdô), pp. 91-126; Edwin Cranston, article under the entry for “Man'yôshû,” Kôdansha Encyclopedia of Japan, vol. 5 (Kôdansha, 1983), pp. 102-11.

7. See note 6, under Nakanishi Susumu.

8. See Morimoto Masayoshi, “Man'yôshû kan shichi kô” (A study of Book 7 of the Man'yôshû), Kokugo to kokubungaku 4 (8), pp. 1230-54.

9. See Yamazaki Kaoru, “Man'yôshû no hiyuka” (The hiyuka in the Man'yôshû), Man'yôshû taisei 7, pp. 129-44.

10. Unless otherwise stated, all quotes from the Man'yôshû are taken from Man'yôshû, Kojima Noriyuki, Kinoshita Masatoshi, Satake Akihiro, eds., vols. 2-5, Nihon koten bungaku zenshû (Shôgakkan, 1971).

11. In my translation of the Man'yôshû poems, I consulted and made adaptations based on the following Japanese edition and English translations: Man'yôshû, Kojima et al., 1971; The Man'yôshû, J. L. Pierson, trans., 20 vols. (London: E. J. Brill Ltd., 1933); The Man'yôshû, Nippon Gakujutsu Shinkôkai, trans. (Columbia University Press, 1965); The Ten Thousand Leaves, Ian Levy, trans. (Princeton University Press, 1981).

12. See the definition of metaphor in the Princeton Encyclopedia of Poetry and Poetics (Enlarged Edition), Alex Preminger, ed. (Princeton University Press, 1974).

13. See Nakajima Mitsukaze, “Hiyuka mondôka ron” (On metaphorical poems and dialogue poems), in Man'yôshû kôza 6 (Shun'yôdô), pp. 91-126.

14. Ôgibata Tadao, “Hiyuka no seikaku” (The characteristics of hiyuka), Ôsaka Man'yô Gakkai (July 1953), pp. 1-8.

15. This definition is taken from the headnote to Poem 115, Man'yôshû, 1 (Shôgakkan, 1971), p. 126.

16. Translations of poems from Book 3 are based on Levy 1981, with some modifications.

17. The image of the marker also dominates the following set of poems: 7:1348, 3:400, 8:1510. The metaphorical relations expressed in these poems are again very straightforward. Poems 400 and 1510 are remarkably similar in terms of content, structure, and language, and both express the idea that the speakers' possessions shall defy the natural course of withering and continue to thrive. Yet one is placed under the category of hiyuka and the other under sômon. This kind of overlap recurs again and again. See Poems 7:1364 (hiyuka), 8:1463 (Spring; love poem), 10:2286 (Autumn; love poem), and 11:2759 (kibutsu chinshi).

18. I did not translate the makurakotoba (pillow word) for spring, fuyu gomori.

19. MYS 1:20

    Akanesasu
    murasakino yuki
    shimeno yuki
    nomori wa mizu ya
    kimi ga sodefuru

    Going this way on the crimson-
    gleaming fields of murasaki grass,
    going that way on the fields
    of imperial domain—
    won't the guardian of the fields
    see you wave your sleeves at me?

    (Translated by Levy, Ten Thousand Leaves, p. 48)

20. A similar use of the metaphor of the wisteria cloth is also seen in poem 12:2971 (hiyuka).

21. The clothing metaphor is not limited to the category of hiyuka. It is found in the categories of seijutsu shincho (12:2846, 49) and kibutsu chinshi (12:2851). For instance:

    12:2846 (SEIJUTSU SHINCHO)

    Yoru mo nezu
    yasuku mo arazu
    shirotae no
    koromo wa nukaji
    tada ni au made ni

    Unable to sleep at night,
    unable to feel at ease,
    I shall not remove my robes of white cloth
    until we directly meet again.

    12:2851 (KIBUTSU CHINSHI)

    Hito no miru
    ue wa musubite
    hito no minu
    shitabimo akete
    kouru hi so ôki

    The outer robe that people can see
    is fastened,
    the underrobe that evades people's eyes
    is untied—
    Days of longing are many.

22. See the discussion of metaphor in Preminger, p. 491.

23. Poetics 1458b; cf. Rhetoric 1405a, quoted in Preminger, p. 491.

24. Morimoto Masayoshi thinks that the following “poems on things” (eibutsu) can be read as hiyuka: Book 7:1078, 85, 91, 1109, 21, 29. See Morimoto, p. 1230.

25. For a discussion of the pragmatic theories of Chinese literature, see the chapter titled “Pragmatic Theories” in James Liu, Chinese Theories of Literature (University of Chicago Press, 1975), pp. 106-16. For a discussion of the delineation of the pragmatic ends of poetry in the Kokinshû prefaces, see Wixted, 1983, pp. 215-38, esp. 234-38. Also see the discussion of the same topic in the chapter titled “Kokinshû and Its Prefaces: Genesis and Aesthetic” in Helen McCullough, Brocade by Night: ‘Kokin wakashû’ and the Court Style in Japanese Classical Poetry (Stanford University Press, 1985), pp. 293-367, esp. 302-8.

26. A comparison is a kind of prosaic metaphor. In quoting Chinese texts that involve the term piyu, I shall render it as “comparison” when it clearly refers to a case of comparing, but I shall render it as “metaphor” when it clearly refers to a broader range of metaphorical relations.

27. This translation is based on James Legge, The Four Books (Shanghai: The Commercial Press, 1962), p. 12.

28. Cf. Legge's translation, “to be able to judge others by what is nigh in ourselves,” ibid., p. 77.

29. Fan Hua, Hou Han shu, vol. 2 (Beijing: Zhonghua Shuju, 1965), p. 1921.

30. Shuowen qiezi gulin, vol. 3 (Shanghai: Shangwu yinshu guan, 1959), p. 973.

31. Ibid.

32. For the definition of pi and piyu, I consulted Morohashi Tetsuji, Daikanwa jiten, vol. 10 (Taishûkan Shoten, 1955), p. 600.

33. For the discussion on feng in this paragraph, I consulted Daikanwa jiten 10, p. 534, and Ôta Hyôzaburô, “Nihon kagaku ni okeru Shina shiron no eikyô” (The influence of Chinese poetics on Japanese poetics), Kokugo to kokubungaku 15 (4), pp. 121-42. Revised and reprinted in his book, Nihon kagaku to Chûgoku shigaku (Shimizu Kôbun Shobô, 1968), pp. 42-53.

34. Mao shih Zheng jian, vol. I.1 (Shanghai: Zhonghua Shuju, 1930), p. 11.

35. Yu pian, 30 fascicles. An enlarged version based on the Shuo wen. Compiled by Gu Yewang (Liang dynasty), enlarged by Sun Qiang (Tang dynasty), and revised by Zheng Pangnian et al. (Song dynasty).

36. Kong Yingda, Mao shih zhushu, 4 vols. Ashikaga gakkô hiseki sôkan 2 (Kyûko Shoin, 1973), 1, p. 53.

37. Wen xuan Li zhu yishu, vol. II.7 (Taipei: Zhonghua Congshu, 1967), p. 5.

38. Kishimoto Yuzuru, Man'yôshû kôshô (Evidential studies on the Man'yôshû) (Kokin Shoin, 1925), 3 (2), pp. 41-42.

39. The Great Preface (Da xu) was formerly attributed to Bu Shang (507-400 b.c.), but in more likelihood written by Wei Hong. See Wixted 1983, p. 217. The annotators of this edition are traditionally identified as Mao Heng and Mao Chang.

40. This translation is based on Gibbs 1972, p. 288; Liu 1975, p. 112.

41. Concerning the six principles (C. liu yi, J. rikugi), Wixted writes, “An attempt at the application of these critical terms was made by Yoshimochi and Tsurayuki [in the Kokinshû prefaces]. Variously interpreted and inconsistently applied by Chinese commentators to the Shi jing, the terms had become in China a sacrosanct formula invoked for the purposes of legitimizing one's critical stand. Yoshimochi and Tsurayuki employed them in much the same way.” See Wixted, p. 230. For a translation of the kana preface, see McCullough, pp. 318-19.

42. Gibbs 1972, pp. 285-93. Also see Gibbs' discussion of feng in “Literary Theory in the Wen-hsin Tiao-lung, Sixth Century Chinese Treatise on the Genesis of Literature and Conscious Artistry,” (Ph.D. dissertation, University of Washington, 1970), pp. 98-107.

43. Gibbs, “Notes on the Wind,” p. 287.

44. Quoted in Wang Jingchi, Shi jing tongshi (Taipei: Furen Daxue, 1980), p. 2.

45. Gibbs also traces the usage of the term feng in the sense of “influence” in the Analects. When asked about his view on capital punishment, Confucius said, “The virtue of the Superior Man is wind; the virtue of the common people is grass. Wind on the grass is bound to bend it” (Analects 12:19, Legge, Classics 1, pp. 258-59). The metaphor was drawn from the Book of History, “You are the wind; the inferior people are the grass,” Legge 3, p. 539. For a discussion of the metaphor, see Gibbs, “Notes on the Wind,” pp. 288 and 293, note 8.

46. The excerpts from the Great and Small Prefaces are taken from Gibbs' translation in “Notes on the Wind,” p. 288, with slight modifications.

47. This translation is taken from Legge 1962, p. 178. Cf. The Analects, D.C. Lau, trans. (Penguin Books, 1979), p. 145. Donald Holzman comments that, “[To Confucius,] the Shi jing was important not as a work of literature, but as a tool of diplomacy, an extra-literary guide to morality and an aid to social living.” See “Confucius and Ancient Chinese Literary Criticism” in Chinese Approaches to Literature from Confucius to Liang Ch'i-Ch'ao, Adele A. Rickett, ed. (Princeton University Press, 1978), pp. 21-42. For a list of informed Western-language works on Confucius' view of literature, see Wixted 1983, p. 225, note 23.

48. This translation is based on Bernhard Karlgren, The Book of Odes, Chinese Text, Transcription and Translation (Stockholm: The Museum of Far Eastern Antiquities, 1950), pp. 72-73, with slight modification. Cf. Legge's translation in The Book of Poetry, Chinese Text with English Translations (Paragon Book Reprint Corporation, 1967), pp. 125-26. The Chinese text is taken from Wang Jingzhi, ed., Shi jing tongshi (Taipei: Furen Daxue, 1980), p. 236.

49. Wang, p. 236. Cf. Legge's own preface to Poem 113 which reads as follows: “The ‘shih shu’: metaphorical. Against the oppression and extortion of the government of Wei.” The Book of Poetry, ibid., p. 125.

50. The Chinese tradition of political or historical contextualization is used to a certain extent in the reading of some of the eleven “foretelling songs” (waza uta) in the Nihon shoki (720), but Japanese scholars think that it is highly debatable that the songs originally carry any political messages. See Sey Nishimura, “Retrospective Comprehension: Japanese Foretelling Songs,” Asian Folklore Studies 45 (1986), pp. 45-66, esp. 49-51.

51. It is not impossible to read some hiyuka as political critiques, but for the most part they are conventionally treated as love poems. Poem 391 in particular strikes me as one that is potentially laden with political meaning:

    3:391 (hiyuka) A poem by the priest Manzei, superintendent for the construction of the Buddhist temple Kanzeon-ji in Tsukushi.

    Tobusatate
    ashigarayama ni
    funagi kiri
    ki ni kiriyukitsu
    atara funagi o

    Standing thick ends of branches in the stump
    as offering to the mountain god,
    they have cut the ship-timber on Ashigara mountain,
    Cut them as ordinary timber—
    Alas, that ship-timber!

    The speaker in this poem bemoans the fact that good timber is not put to its best use, and traditionally this poem is read as a love poem in which the speaker regrets that someone else has claimed the woman he likes. (See the interpretations of Kojima et al., Pierson, and Levy.) This kind of reading is prevalent because the imagery in hiyuka has traditionally not been treated as political metaphor but rather as a metaphor for love in accordance with the emphasis on the aesthetic function of poetry in Japanese conventions of reading. If it were read in the Chinese exegetical tradition, the speaker of the poem could well be expressing the frustration of someone whose ability or virtue is not recognized. This reading would cast this poem into the long tradition of poems about frustrated scholar/courtiers featured prominently in the Shi jing and Chu ci (Songs of the South).

52. In the Six Dynasties, the term feng was infused with a more expressive shade of meaning by Liu Xie, the author of Wenxin diaolong, who regards feng as a literary element capable of “transforming the emotions.” He writes that “when you are saddened and would set forth your feelings, you must begin with feng.” See Gibbs, “Notes on the Wind,” p. 289. The use of the metaphor in the hiyuka is closer to Liu Xie's relatively expressive mode of interpretation, but there is no tangible evidence to show that the idea of hiyuka is connected to the term feng as used in Liu Xie's text.

53. Huangquing jingjie, Ruan Yuan, ed., 360 fascicles. Fascicle number 286 is Mao shi bushu (Additional notes to the Mao shi), 5 vols, Jiao Xun, ed.

54. The author[s] of Man'yôshû kôgi more or less suggest[s] this possibility in the definition of hiyuka: “Hiyuka should be read as tatoeuta. In the preface to the Kokinshû, the fourth kind of poetry is tatoeuta, all of which are poems that express thoughts indirectly by entrusting them to objects.” See Man'yôshû kôgi 2, Yamaka Yasue et al., eds. (Kokusho Kankôkai, 1912), p. 274.

55. From Zheng Xuan's Zhou li zhushu (Annotations to the Rites of Zhou), in Shisan jing zhushu, I.7/46a/p. 684. Quoted and translated in Pauline Yu, The Reading of Imagery in the Chinese Poetic Tradition (Princeton University Press, 1987), p. 58. Subsequent page references in Yu will be given in the text.

56. See Shi pin zhu (Gradings of Poets Annotated) Chen Yenqie, ed. (1927; rpt. Taipei, 1958 and 1960), p. 4. This translation is based on John T. Wixted, “The Nature of Evaluation in the Shih-p'in (Grading of Poets) by Chung Hung (a.d. 469-518)” in Theories of the Arts in China, Susan Bush and Christian Murck, ed. (Princeton University Press, 1983), p. 238, with slight modification.

57. Zhou li zhushu, in Shisan jing zhushu, I.23/158a/p. 796.

58. Mao shi zhengyi, 1A/6b, p. 44.

59. Wenxin diaolong shiju, Zhao Zhongyi, ed. (Guangxi: Lijiang Press, 1982), p. 307. Yu's translation, modified.

60. This leads the Song commentator, Zhu Xi, to comment that many poems consist of both “bi and xing.” See Zhu Xi, Shi jizhuan 1 (Shanghai: Wenxue Kuji Kanxingshe, 1955), p. 12.

61. The following is a brief outline of the development of Chinese poetics from Han to Six Dynasties: (a) In the last years of the Han dynasty, Cao Pi's (187-226) “Lunwen” (Essay on Literature) paved the way for “the declaration of independence of literature” and “a study of literature in and of itself,” despite the fact that Cao Pi still felt obliged to say that “literary writing is a major occupation in the managing of a state.” For discussions on “Lunwen,” see Holzman 1978, pp. 40-41; and Holzman, “Literary Criticism in China in the Early Third Century a.d.,” Asiatische Studien/Etudes Asiatiques 28.2 (1974), pp. 113-36, pp. 121-36. (b) By the Six Dynasties, there was a significant number of critical works that emphasized discourse on tones and style, many of which are referred to in Kûkai's Bunkyô hifuron (819), such as Shen Yue's Sisheng pu, Liu Shanjing's Sishen chigui, and Wenbi shi (author unknown). For a study and English translation (of three of the six chapters) of Bunkyô hifuron, see Richard Wainwright Bodman, “Poetics and Prosody in Early Medieval China: A Study and Translation of Kûkai's Bunkyô hifuron” (Ph.D. dissertation, Cornell University, 1978). (c) Donald Gibbs points out that there is strong evidence suggesting an expressive emphasis in the thinking of Liu Xie (465?-523?), the author of one of the greatest works of Chinese criticism, Wenxin diaolung (Elaborations on the heart of literature). But Liu Xie is also concerned with the political function of literature. For a discussion of the tension between the political and aesthetic concerns in literature from Confucian to Liu Xie, see Donald A. Gibbs, “Literary Theory” (Ph.D. dissertation, University of Washington, 1970), pp. 24-39. For a discussion of Liu Xie's emphasis on the expressive, see p. 38. (d) For a succinct discussion on the literary view of Xiao Tong (501-531), the compiler of Wen xuan (Selections of refined literature), as well as the views of his contemporaries, see David R. Knechtges, “The Literary Milieu of the Liang and Xiao Tong's View of Literature” in Wen xuan, or Selections of Refined Literature 1 (Princeton University Press, 1982), pp. 11-21. Knechtges points out that while the Qi-liang era (479-556) was a time of great attention to literary craft, some scholars were disturbed by the neglect of the classical norm. In an essay titled “Treatise on Carving Insects” (Diaochong lun), Pei Ziye (460-530) stressed the use of literature in moral instructions. On the treatise, see John Marney, “P'ei Tzu-yeh: a minor Literary Critic of the Liang Dynasty,” in Selected Papers in Asian Studies 1 (Albuquerque: Western Conference of the Association for Asian Studies, 1976), pp. 161-71.

62. For instance, both Ki no Yoshimochi, the presumed writer of the Chinese preface of the Kokinshû, and Ki no Tsurayuki, the writer of the Chinese preface of the Shinsen wakashû, paid homage to the Chinese view of literature by drawing from the following passage in the Great Preface of the Shi jing in a practically formulaic manner: “Therefore, nothing approaches the Book of Poetry in maintaining correct standards for success or failure [in government], in moving Heaven and Earth, and in appealing to spirits and gods. The Former Kings used it to make permanent [the tie between] husband and wife, to perfect filial reverence, to deepen human relationships, to beautify moral instruction, and to improve social customs.” This translation is taken from Liu 1975, pp. 111-12. For the text of the Chinese preface to the Kokinshû, see Nihon kagaku taikei 1, Sasaki Nobutsuna, ed. (Fûma Shobo, 1963), pp. 41-42; for the Chinese preface to the Shinsen wakashû, see p. 44.

63. The Man'yôshû poems which appear in the Kokinshû are not taken from the text of the Man'yôshû but are believed to be orally transmitted. In fact, the Man'yôshû was not regarded highly by the Kokinshû compilers and the man'yôgana (script) had become obscure, so they made a point not to include Man'yôshû poems in their anthology. See headnote to poem 1351, Man'yôshû, Kojima Noriyuki et al., eds., Nihon koten bungaku zenshû 2 (Shôgakkan, 1971), p. 267.