SOURCE: McCullough, Helen Craig. Introduction to Tales of Ise: Lyrical Episodes from Tenth-Century Japan, translated by Helen Craig McCullough, pp. 1-65. Stanford, Calif.: Stanford University Press, 1968.

[In the following excerpt, McCullough explores the importance of the figure of Narihira to the tales and comments on the difficulties of classifying Ise by genre and establishing its authorship.]

JAPANESE COURT POETRY IN THE NINTH AND TENTH CENTURIES

Tales of Ise is anonymous and of uncertain date, and so are a majority of its poems, but it is probably safe to assume that few, if any, of the poems are more recent than 950, and that most of them were written during the ninth century.¹ The poems coincide roughly in period, therefore, with those in the first imperial anthology of Japanese poetry, Kokinshū, or Collection of Ancient and Modern Times, in whose title “ancient times” means essentially the early decades of the ninth century. There are, moreover, basic similarities of theme, technique, and tone between the two collections, as well as a partial duplication of content. The kinds of poems considered worthy of an imperial anthology were also the kinds admired by the unknown author or authors of Ise monogatari, and they represent the practice and taste of contemporary court society …

Especially important to the content and tone of court poetry was the rule of taste as it applied to the natural world. The lives and interests of the nobility centered on the imperial city of Heian. To be forced to reside in the provinces was to be exiled; long journeys were undertaken with extreme reluctance, attended by timorous (if justified) misgivings about, the dangers of the road, and lamented in floods of plaintively nostalgic poems. But the incomparable beauty of Japan's mountains and streams persuaded these city dwellers to build suburban villas and hunting lodges, and to venture on pilgrimages to nearby shrines and temples. No pastime was considered more elegant, and none was more typical of the age, than a brief excursion into the countryside, where, with a few companions, a gentleman sipped wine and composed verses inspired by the ever-changing face of nature. In the capital itself, every aspect of nature could be observed in the landscape gardens of the great mansions, whose lakes, streams, hills, rocks, and plants were designed with infinite pains to suggest well-known scenic places and to appear to advantage at different seasons of the year. There one might watch the nocturnal voyaging of the harvest moon, select a single perfect chrysanthemum to accompany a gift, or, seated alone of a wintry morning, indulge in elegant musings on the resemblance between plum blossoms and snow.

No well-bred person remained unmoved in the presence of nature, and few aristocrats allowed their reactions to pass unrecorded in verse. A nature poem was expected to demonstrate a capacity for close observation, a keen appreciation of beauty, and, most important, a sensitivity to the poignancy inherent in the relationship between beauty, ephemerality, and the human condition. On viewing a cherry tree in full bloom, someone like Prince Genji, the idealized hero of The Tale of Genji, was likely to write of the evanescence of the blossoms, or to reflect that although the tree would bloom again, man's own springtime comes but once. In the cicada's papery husk and the withering plumes of summer grasses, Heian poets recognized the tyranny of time; and nature's more permanent aspects inspired them to melancholy reflections on the brevity of human existence and the uncertainties besetting man's troubled passage through the world.

A courtier responded, then, to the first green of spring, to snowfall, mist, and rain, to the scattering of blossoms. And the tone of the response was the tone of Heian society. Its basis was a love of beauty so sophisticated, so highly refined, that it permitted no careless outpouring of spontaneous feeling. Emotion was conveyed through understatement and veiled allusion, through the subtle, half-spoken evocation of a mood; it was contained within channels of formality and propriety and expressed with decorous elegance.

The cultivated man was expected to display equal taste and sensitivity in his relations with others. He conducted affairs of the heart with delicate solicitude for the lady's feelings, as when Prince Genji, blundering into a relationship with an awkward, red-nosed princess, made himself permanently responsible for her happiness because it was clear that she would never attract another suitor. In other social relationships as well, the requisites were tact, a ready wit, and discernment—the ability to grasp the nuances of a situation and respond appropriately, preferably in verse.

Poetry concerned with human relations reflects these social requirements, and so does most nature poetry, because natural phenomena are almost invariably presented in terms of their effect on man in a specific situation with social connotations. Heian poems are occasional pieces, instruments of social intercourse, and much of their interest derives from the circumstances of their composition. How well has the poet met the implicit challenge? How skillfully has he shown his mastery of the rule of taste?

Kokinshū and Tales of Ise both contain the following poem by Ariwara Narihira:

Akanaku ni
Madaki mo tsuki no
Kakururu ka
Yama no ha nigete
Irezu mo aranan.
Must the moon disappear
In such haste,
Leaving us still unsatisfied?
Would that the mountain rim
          might flee
And refuse to receive her.

It is apparent, of course, that Narihira and others have been watching the moon, and that they regret its setting. The ritual of gazing at the moon has the prestige of Chinese precedent, familiar from innumerable poems and paintings, and is thus immediately classifiable as a gentleman's pursuit. The poet also demonstrates the requisite feeling for the impermanence of beauty. But it is the apt response to specific circumstances that makes the poem truly elegant. To appreciate the response, we need the background provided by Tales of Ise's prose context, which reveals that the imperial prince Koretaka, off on a hunting trip with a group of attendants, has spent most of a long spring day under blossoming cherry trees, sipping wine and chanting verses. At the prince's villa, the company has sat far into the night, gazing at the moon. When at last the moon sinks toward the hills and the prince prepares to retire, Narihira detains him with the poem Akanaku ni, a graceful, compressed allegorical expression of sentiments that in another society might be spelled out at tedious length: “Today has been a rare experience. You have allowed us to accompany you to the fields; you have entertained us most graciously at dinner. We shall never forget the cherry blossoms at Nagisa, the poems on the return journey, or the stories this evening as the moon climbed the skies. It is hard indeed that such a day in such company must end. Will you not linger with us a while? Life with its precious moments passes as swiftly as the moon crosses the heavens; let us enjoy it while we may.”

In the eyes of Japanese critics, Akanaku ni succeeds because it combines grace, sensitivity, and wit in a moving expression of the dominant aesthetic ideal of the age, the blend of elegance and pathos known as mono no aware. The same is thought to be true, in varying degree, of the other poems in Ise monogatari—all are model responses to social situations, evoking mono no aware, the pathos of life, by means of apt metaphors, elegant diction, and images rich in literary connotations.

Ariwara Narihira (825-80) was the grandson of two emperors. His father, Prince Abo, a son of ex-Emperor Heizei, was exiled to Kyūshū in 810 at the age of 18, after participating in an attempt to restore Heizei to the throne; he returned to the capital around 824 and went on to a moderately successful official career. Narihira's mother was Princess Ito, a daughter of Emperor Kammu. According to Sandai jitsuroku (True Records of Three Reigns), the official history for the period, the poet was his father's fifth son. He and his brothers were made commoners, with the surname Ariwara, in the year following his birth.

Aside from scattered hints in Kokinshū poems and their headnotes, the chief source of information about Narihira is a terse obituary notice in Sandai jitsuroku, which, like all the histories, concentrates primarily on the official careers of its subjects. The editors' list of Narihira's offices and ranks shows that he advanced very slowly, and that, unlike his older brother Yukihira, he never rose to the top level of the bureaucracy. After a series of minor appointments, he was promoted to first one and then the other of the two offices with which he is identified in Tales of Ise: he became Commander of the Right Horse Bureau in 865 and Middle Captain in the Imperial Guards of the Right in 875. In 879, shortly before his death, he was named Director of the Archivists' Bureau. His highest court rank was Junior Fourth Upper.²

The chief reason for this poor showing was that during Narihira's lifetime the Fujiwara family was gradually excluding outsiders from important public offices. Fujiwara pressure was irresistible unless a man happened to combine exalted lineage with exceptional ability or seniority, as, for example, Yukihira and some of the Minamoto ex-princes did. Narihira pursued the course usually adopted by a courtier of high birth who was unwilling either to enter the Buddhist priesthood or to seek his fortune in the provinces. He stayed in the capital, performed a few routine duties, and devoted most of his time to social activities. His life as a fashionable gentleman of leisure inspired his poetry and was indirectly responsible for the development of Tales of Ise in the form in which we know it.

The passage that describes Narihira in Sandai jitsuroku, though famous, has been made tantalizingly cryptic by the editors' determination to write in balanced Chinese prose. Each of their four comments consists of exactly four characters.

1. Taibō kanrei: “In appearance he was elegant and handsome.” This probably means that he was more than ordinarily good-looking, since official histories do not as a rule comment on a man's appearance.

2. Hōjū kakawarazu: Hōjū usually means “self-indulgent,” kakawarazu, “regardless of” or “in spite of.” The phrase implies that Narihira did as he pleased regardless of established conventions—but when and how? As an official? In his private life? One possible interpretation, that he was promiscuous, has become the central element in a proliferating legend that has made this poet one of the great romantic figures of Japanese history, comparable in some ways to the twelfth-century warrior Minamoto Yoshitsune. He has been regarded as a real-life Prince Genji, irresistibly attractive to women and highly sensitive in his relationships with them, and indeed with men as well—in short, a model Heian courtier. This is his role in Tales of Ise, which both stems from the legend and reinforces it. But attempts to find an independent, factual basis for the legend do not get very far, since headnotes to Narihira's Kokinshū poems (the other chief source of information) show only that he was involved with three or more ladies. It is possible that the stories about his amatory prowess have been greatly exaggerated, and that the Sandai jitsuroku editors were being less specific than has traditionally been supposed. … It has also been suggested, on the basis of the third comment in the obituary, that the phrase means that he refused to bother with Chinese studies because he preferred to write Japanese poetry.

3. Hotondo saigaku nashi: “Almost no saigaku.” In Sandai jitsuroku, sai (“talent,” “intelligence,” “aptitude”) usually seems to mean scholarly attainments or capability as a Chinese scholar, and gaku (“education”) Chinese learning. Puzzled Tokugawa scholars, unable to understand how someone with “almost no Chinese education” could have had an official career, decided that the negative nashi had somehow slipped into the text instead of the positive ari. The present tendency, however, is to hold that Narihira's duties could not have been exacting, and that the text probably means what it says.

4. Yoku yamatouta o tsukuru: “He excelled in the composition of Japanese poems [as opposed to Chinese].” This is a rather unusual remark for the editors of a standard history, who typically took the position that only a man's official career and his achievements as a Chinese scholar were worth recording. (The biography of the great Man'yōshū poet Ōtomo Yakamochi, for example, says nothing about his verse.) It probably signifies both that the status of the waka was rising and that Narihira's genius was recognized by his contemporaries.

This is as far as the official history takes us. For more information about Narihira, his associates, and his way of life, one can only rely upon his Kokinshū poems, poems written to him, and the relevant headnotes. … Narihira wrote a number of poems relating to prominent figures of the day: the Fujiwara regent Mototsune; Fujiwara Kōshi (sometimes called Takaiko), a consort of Emperor Seiwa; and an imperial prince, Koretaka.

Prince Koretaka (844-97) was a son of Emperor Montoku (r. 850-58) who had appeared to be destined for the throne until his infant half-brother was named crown prince instead, in 850. Koretaka was the son of a lady from the Ki family; his successful rival, the future Emperor Seiwa, was the grandson of a Fujiwara. Narihira married a daughter of Ki no Aritsune, Prince Koretaka's uncle. During his years as Commander of the Right Horse Bureau, he was apparently in the prince's service (perhaps because of the family connection), and he seems to have been a valued and congenial member of the prince's literary coterie. A poem in Kokinshū bears touching witness to his grief when the prince suddenly took religious vows at the age of 28 and retired to a hermitage. Tokugawa scholars were wrong, one feels, when they explained the relationship between these two men as merely a political alliance against the Fujiwara, though it can be assumed that disappointment was among the ties that bound them together.

During the Tokugawa period, a political explanation was also zealously sought for the most famous of Narihira's legendary love affairs. The lady was Fujiwara Kōshi (842-910), “the Empress from the Second Ward,” niece and adopted daughter of the Fujiwara chieftain Yoshifusa. Kōshi became a junior consort (nyōgo) of Emperor Seiwa in 866, when she was 24 and the emperor was 16. She bore the future Emperor Yōzei two years later, and received the title of ex-empress (kōtaigō) after Yōzei's accession in 876. For seven years, from the time her son was named crown prince until Emperor Seiwa abdicated, she was the most important of the imperial ladies, and she seems to have led a gay life, surrounded by luxuries, and to have maintained a literary salon frequented by Narihira and other poet-courtiers. Later her fortunes waned. Yōzei, who proved to be criminally insane, was deposed in 884 by Kōshi's brother, Mototsune, and in 896 Kōshi was deprived of the title of ex-empress because of a sensational liaison with a Buddhist monk.

Kōshi's affair with Narihira supposedly took place while her adoptive father, Yoshifusa, was waiting for the future Emperor Seiwa to reach puberty so that she could be made an imperial consort. According to legend, Narihira visited her in the palace of her aunt, Fujiwara Junshi, “the Empress from the Fifth Ward,” became intimate with her, and finally abducted her.

A generally reliable eleventh-century chronicle, Ōkagami (The Great Mirror), says of Kōshi:

It is not at all clear how the lady happened to become an imperial consort, because she was the girl whom Middle Captain Narihira carried off and hid while she was still living a sheltered life at home. Her elder brothers … went to fetch her back. [Since her elopement ruled out the possibility of a formal presentation], there would have been no opportunity for the emperor to meet and fall in love with her if she had been kept as rigidly secluded as most girls. But she paid visits to the Empress from Somedono [Meishi, Seiwa's mother], with whom she was on intimate terms, and no doubt he noticed her on one such occasion.

As a result of the elopement (the legend continues), Narihira was obliged to go to eastern Japan for a time, and after his return his official career languished because of Fujiwara hostility. According to some Tokugawa scholars, the seduction was a political ruse devised by Narihira because, as a partisan of Prince Koretaka and the Ki family, he wished to prevent the match between Seiwa and Kōshi.

Kokinshū and Tales of Ise are the sources for this romance, which has always been regarded as one of the great scandals of early Heian court history. …

Many scholars have questioned the historicity of the trip to the east, usually on the grounds that it ought to have produced more and better poems, and there is in any case no evidence to support the old argument that the trip was made because of the disclosure of an affair with Kōshi, or possibly to try to recruit anti-Fujiwara support. About the affair itself, one can merely speculate. Perhaps the dashing young man did indeed snatch the susceptible maiden from under the arrogant noses of her Fujiwara guardians; perhaps not. Eleven hundred years later, it can be said with certainty only that the legend was firmly established well before the end of the Heian period.

Kokinshū shows Narihira writing poems for Kōshi, as for Prince Koretaka, on purely public and social occasions. We learn from the same source that he composed a graceful poem for Mototsune's fiftieth birthday celebration, and there are other indications that he was outwardly on good terms with the Fujiwara, regardless of his private sentiments. A lament for Yoshifusa is attributed to him in the second imperial waka anthology, and the main figure in Ise monogatari is consistently courteous and complimentary in his dealings with the family.

From Kokinshū headnotes (corroborated by episodes in Ise monogatari), one can deduce that Narihira's closest associates were men of his own sort, cultivated, interested in poetry, high-born but of little consequence politically: e.g., Ki no Aritsune, his father-in-law; Fujiwara Toshiyuki, an important poet who was Aritsune's nephew; and Ki no Toshisada, a minor official whose poetry was good enough to be included in Kokinshū.³

Narihira's nine poems in the Kokinshū “Love” category disclose the following information about his celebrated amours. One poem, with the companion verse by the lady and that verse's headnote, constitutes the evidence for his marriage to Aritsune's daughter; one was sent to a stranger; two seem to hint at the relationship with Kōshi; one is a complaint against a lady who had resisted his advances; three others are addressed to a lady or ladies whom he had presumably wooed successfully, perhaps Kōshi, perhaps Aritsune's daughter, perhaps one, two, or three others. And one is said to be a reply to the poem below:

When Narihira went to Ise Province, he paid a most secret visit to the Virgin. The next morning, as he was thinking of her longingly (he could not very well send her a message), someone brought him this.

Kimi ya koshi
Ware ya yukiken
Omōezu
Yume ka utsutsu ka
Nete ka samete ka.
Did you, I wonder, come here,
Or might I have gone there?
I scarcely know …
Was it dream or reality—
Did I sleep or wake?

Section 69 of Tales of Ise says, “The Virgin was the one who served during the reign of Emperor Seiwa,” but Kokinshū is not specific, and its phrase saigū narikeru hito might even be rendered “a lady in the Virgin's service.” Within a few decades, however, the combined influence of the two books had made this particular Virgin part of the Narihira legend, and by the early eleventh century she was being accepted as the mother of one of the poet's sons. Modern scholars have been more skeptical. They point out that if such a love affair took place it must have been a well-kept secret, since otherwise the Virgin would have been removed from office (she served her full term without incident), and that whereas hints of the Kōshi liaison recur repeatedly in Tales of Ise, suggesting that there was some foundation for the persistent rumors, the same is not true of the supposed association with the Virgin. … The only certainty is that of the three specific ladies to whom Kokinshū alludes, or appears to allude, Aritsune's daughter, a suitable wife for a man in Narihira's position, has received little attention in the legend, while Kōshi and the Virgin, the glamorously inaccessible imperial consort and high priestess, have become its main pillars.

There are only a few other bits of information with which to touch up this sketchy picture of Narihira and his life: intimations of periods of depression, caused no doubt by dissatisfaction with his offices and ranks, or by unhappy loves; additional exhibitions of sensibility; and indications of a warmly affectionate, generous nature. The total is not an impressive foundation for a major legend, and yet it evokes a curiously vivid image. Like Prince Genji, Narihira seems to personify the supreme Heian ideals of elegance and sensitivity. And it is fairly certain that this is why he became the model for the character, identified only as “the man of old,” who lends unity to Tales of Ise's vignettes.

Since Narihira has always been regarded as a great poet, his name has become attached to many poems from other hands. In order to appraise his work, one must first discard these accretions. A mere 150 lines remain, but they are enough to prove that his reputation rests on solid achievement, to illuminate his place in the classical tradition, and to clarify the nature of his indebtedness to Chinese example. Here is one of his poems, composed “On seeing the cherry trees at the Nagisa House”:

Yo no naka ni
Taete sakura no
Nakariseba
Haru no kokoro wa
Nodokekaramashi.
If this were but a world
To which cherry blossoms
Were quite foreign,
Then perhaps in spring
Our hearts would know peace.

It can scarcely be denied that this poem stands squarely in the Six Dynasties tradition of intellectualism. Furthermore, it is not an isolated aberration. Narihira constantly hypothesizes, explains, wonders, poses rhetorical questions, and devises ingenious metaphors. Whether he was a Chinese scholar or not, there can be no doubt that he used the techniques of Chinese poetry and that his themes show Chinese influence. The “message” of Yo no naka ni is a familiar one: “We could enjoy the spring if we were not obliged to worry constantly lest wind or rain destroy the blossoms.” Similarly, the basic idea in the famous Tsuki ya aranu—the contrast between the ephemerality of human experience and the eternal sameness of spring—can be traced through numerous Chinese and Japanese variations to an early T'ang original. Still other examples could be cited. But in poetry the crucial thing is the use made of materials, not the materials themselves. It was natural that Narihira and other early Heian poets should have sought to adapt Chinese conventions to the waka. The important question is: how successful as poetry was the result?

Few poets have treated their subjects with more technical skill or greater lyric beauty than Narihira. Brower and Miner's analysis of Yo no naka ni sheds useful light on his methods.

Although appropriate diction is but one of Narihira's many accomplishments, it is the language in cadences of Virgilian resonance more than anything else which makes his art so appealing. He and Komachi often fill the third and fifth lines with liquid, inflected adjectives or verbs that develop the sounds of preceding words with a perfected lyricism. [The poem Yo no naka ni] by Narihira, for example, is one that will never have a wide appreciation among foreign readers, but its pure diction and lovely rhythms will always appeal to the Japanese. The major pause at the end of the third line anticipates the strong conclusion, in which the “o” and “k” sounds of the preceding line are given a new direction; the first and fourth lines have slightly longer pauses and their grammatical structure is similar, except that each of the two nouns in the fourth line has one more syllable than the corresponding noun in the first line.

[Robert H. Brower and Earl Miner, Japanese Court Poetry. Stanford: Stanford University Press, 1961]

One of Narihira's special qualities is that, like Komachi, he remains very much a lyric poet in the ancient native tradition, even though he accepts the conventions of a new era, and indeed does much to shape them. He stands at a happy point in time; his poetry, like Heian civilization itself in the ninth century, combines the glorious Nara vitality, sense of wonder, and emotional intensity with a sophisticated Chinese elegance too new to have felt the paralyzing hand of conventionality. In his day, repetition of hackneyed themes and techniques had not yet threatened to reduce the waka to a mechanical exercise, nor was the function of Japanese poetry quite as public as it soon became. None of Narihira's Kokinshū poems was written for a contest, and, so far as can be determined from the headnotes, very few were composed on topics set by other people. Most of them instead represent his response to personal experience, affected in a general way by knowledge of the Chinese classical tradition, but shaped specifically by a warm, easily moved nature and a rare poetic sensibility.

Of course Narihira's lyricism is not the lyricism of the Man'yōshū poets. The following love poem, linked in the legend to the affair with Kōshi, illustrates his characteristic mode of expression:

Once, quite without premeditation, Narihira began to make love to a lady who lived in the western wing of a palace belonging to the Empress from the Fifth Ward. Shortly after the Tenth of the First Month the lady moved away with no word to him, and though he learned where she had gone, it was impossible to communicate with her. In the spring of the following year, when the plum blossoms were at their height, poignant memories of the preceding year drew him back to the western wing on a beautiful moonlit night. He lay on the floor of the bare room until the moon sank low in the sky.

Tsuki ya aranu
Haru ya mukashi no
Haru naranu
Wa ga mi hitotsu wa
Moto no mi ni shite.
Is not the moon the same?
The spring
The spring of old?
Only this body of mine
Is the same body …

Whereas the emotion in the typical Man'yōshū poem is directly expressed and easily apprehensible, Narihira presents a psychological state so complex that the poem can be interpreted in at least four different ways. (1) “Is it possible that the moon is actually different this year? The spring too? They seem so to me. Everything has changed; I alone remain the same.” (2) “How could the moon and the spring be different this year from last? Clearly, they are not. Physically, I too am still the same, but I feel entirely different. I was happy before; now I am miserable.” (3) “Everything is as it was before. The moon, the spring, I myself—all are the same. The only difference is the absence of the woman I love; the change is there, not in me.” (4) “I don't know whether or not the moon and the spring are different. I suppose not, although they seem very different to me. But I am painfully aware that I am the same person who once was happy here. Of that there can be no doubt at all.”

Tsuki ya aranu is typical of a group of poems in which Narihira has used polished craftsmanship and lyric intensity to explore the nature of illusion and reality, and, in so doing, to create a dreamlike, otherworldly atmosphere. The poem has exceptional formal distinction. Stops at the ends of lines 1 and 3 strengthen the emotional impact, and great aural beauty is achieved by patterns of repetition: Tsuki ya … Haru ya; Tsuki ya aranu … Haru naranu; Haru ya … Haru naranu; Wa ga mi … Moto no mi. The early negatives (aranu, naranu) balance the later affirmation. And by leaving the final line unfinished, Narihira transcends the limitations of his brief medium, leaving in the air a variety of provocative suggestions. This poem and others like it establish Narihira as a major Japanese poet.

The Kokinshū compilers, however, preferred the witty conceits of Yo no naka ni and Akanaku ni. The Chinese preface says, in obvious reference to Tsuki ya aranu and poems of its kind, “Ariwara Narihira's poems suffer from an excess of emotion too tersely expressed. They resemble drooping flowers, somewhat lacking in color but still fragrant.” In other words, Tsurayuki and his colleagues felt that there was an imbalance between form and content in Narihira's poetry: the “lingering fragrance” was the portion of the content denied expression by the inadequacies of the form. The compilers had no objection to overtones per se. … But whereas Narihira, in his best poems, used linguistic resources to evoke complex human emotions, and even to raise metaphysical questions, the more typical Kokinshū author was concerned with creating an impression of ethereal, elegant beauty. …

Different ages establish different concepts of beauty and of the function of art. Ours is broader than that of Heian Japan—less refined, perhaps. … Tsurayuki, while recognizing Narihira's greatness, believed that there was something not quite well-bred, something disturbing, in the intensity of his feeling. Tsurayuki and his colleagues recognized that there could be no poetry without emotion. “Poetry has its roots in the human heart,” Tsurayuki wrote at the beginning of his preface. But the Kokinshū compilers rejected excessive displays of passion for the same reason that they rejected distressing themes and unorthodox images—such as those of Yamanoue Okura's “Dialogue on Poverty” in Man'yōshū,⁴ with its description of whining children and tattered bedclothes, or of the country lass's poems in Tales of Ise, with their homely chatter of roosters, cisterns, and silkworms. The ideal of “decorous elegance,” they felt, could best be served by giving preference to conservative diction, seasonal themes, and gently melancholy reflections on the ephemerality of worldly things.

It can be argued that the Kokinshū compilers and their followers sacrificed too much by limiting themselves in this way, stayed too much under Chinese influence, lost the freedom to experiment and grow, and made the eventual decline of their art inevitable. But it was a deliberate choice, made for reasons that command respect. Their ideal was beauty, taste, and sensitivity; and they achieved it with notable success. The remarkable thing about the Japanese classical tradition is indeed not its sterility but its vitality—the fact that for generations its poets produced works of great distinction, which, after centuries, are still fresh and moving. The survival of this poetry best answers the question of Chinese influence, for mere imitation seldom confers immortality. The good classical poets used Chinese materials creatively. To call them imitators or borrowers is to disregard their finest achievements.

TALES OF ISE

The ideals of the Kokinshū compilers set the general literary tone of the period during which Tales of Ise came into being. Tales of Ise itself, however, stems primarily from the indigenous tradition of lyric love poetry, as modified through contact with the new Sino-Japanese culture. Symbolically, it contains all 30 of Narihira's Kokinshū poems but only one poem attributed to Tsurayuki.

Tales of Ise resists capsule description. Early modern scholars, seeking to relate it to recognized literary genres, found themselves obliged to assign it to a new category, the uta monogatari (“stories about poems”), created expressly for it and for two slightly later works written under its influence. As the name suggests, it is a hybrid form, combining the traditional emphasis on poetry with a new interest in the potentialities of the prose medium. It represents an early stage in what was perhaps the most significant literary phenomenon of the Heian period—the flowering of true prose literature, which culminated in The Tale of Genji.

Tales of Ise can be described as a collection of short short stories, but many of its episodes consist of nothing more than a poem with a one-line introduction. In such sections, and in numerous others with only slightly more extended prose contexts, it strongly resembles a poetic anthology with headnotes. We cannot however call it one, for its prose elements are, on the whole, too prominent. It is a transitional form, halfway between the poetry collections and the later prose works, such as The Tale of Genji, in which the role of poetry, though important, is unmistakably subordinate.

In most Ise monogatari texts there are 209 poems distributed among 125 sections. In a poetic anthology such as Kokinshū, the poems are arranged with great care, first by general categories (Spring, Summer, Autumn, Winter, Felicitations, Parting, Travel, Love, etc.), and then by subjects within categories. The structure of Tales of Ise is much less tidy. There has been an attempt at rational organization, but at least two mutually contradictory principles have been observed. One is biographical. In the first section of most texts, a young man, recently come of age, flirts with two beautiful sisters at the Nara capital. That section's initial phrase, mukashi otoko (“in olden times a man,” “once a man”), recurs in episode after episode. The anonymous protagonist woos a lady whom the reader is invited to identify with Fujiwara Kōshi, the future consort of Emperor Seiwa, and then journeys to eastern Japan “because of certain problems that had made life in the capital uncomfortable for him.” In the east he exchanges love poems with several ladies. Before long he is back in the city, writing poems on public occasions, pursuing casual amours, and traveling in neighboring provinces. He is dispatched to Ise Province as an Imperial Huntsman, and at the Grand Shrine, in a secret nocturnal meeting, he wins the heart of the Ise Virgin. He serves a son of Emperor Montoku, Prince Koretaka, until the prince takes religious vows and retires to a hermitage. Meanwhile he grows older. He appears to be in demand as a poet on social occasions, although he is described as a humble old fellow who is not good at writing poetry. His poetic exchanges with ladies continue. Near the end of the book he complains—not for the first time—that life has been unkind to him, and in the final episode he writes a death poem.

The poems composed by the “man of old” include all of the 30 attributed to Ariwara Narihira in Kokinshū and four of the 11 attributed to him in Gosenshū; the court offices he holds are offices held by Narihira; and many of the stories about him parallel Kokinshū headnotes to poems by Narihira. But it is clear that there has been no serious, consistent attempt to present a historically accurate biography. Poems listed as anonymous or attributed to other writers in Kokinshū are presented as compositions of the “man of old.” The two Kokinshū poems by Ki no Aritsune's daughter and Narihira are given a fictional setting. Two unconnected Kokinshū poems by [Ono no] Komachi and Narihira are joined in a romantic episode. Narihira's poem about autumn leaves floating in the Tatsuta River, described in Kokinshū as having been written for a picture on a folding screen, is said by Ise monogatari to have been composed at the river. Kokinshū poems by Narihira and by others are consistently endowed with contexts that change their meaning. In many episodes, moreover, the chief male character is someone very different from Narihira—a man living in a remote country district, the young son of an ordinary couple, a minor official traveling in Kyūshū, or the son of an itinerant peddler.

The placement, and indeed the inclusion, of some episodes seems to have been dictated by a determination to group stories dealing with the same locality (eastern Japan, Kyūshū, the home provinces), the same person (Kōshi, Prince Koretaka, the Ise Virgin), or the same subject (married life, court service, unhappy love). Many of the tales in such groups have no connection with the Narihira theme, and some of them … badly disrupt whatever unity the chronological principle confers. Either Tales of Ise was written by an extraordinarily muddleheaded author or it owes its present form to two or more men who sometimes worked at cross purposes.

A book of this kind, made up of short, loosely linked episodes, was peculiarly susceptible to revision, especially in view of the circumstances under which it circulated during its crucial early history. Mentions of Ise monogatari in The Tale of Genji show that it was regarded as a classic by the beginning of the eleventh century, and that it must therefore have been widely known. It was transmitted in manuscript versions, which were copied by readers who saw nothing wrong with making revisions, deletions, and, especially, additions that seemed to them to improve the text. Almost all Heian works have undergone a similar process, but in the case of Ise monogatari an exceptionally complex network of textual lines attests to repeated and extensive changes.

Every extant full-length version of Ise monogatari contains a core of some 110 episodes that appear to constitute the oldest part of the book. Some families of texts have five additional sections, some 15, some 20 or more. The process of accretion can be detected. For instance, every textual line includes Section 69, the story of the encounter with the Ise Virgin …, but Sections 70-75 in our text [Tales of Ise: Lyrical Episodes from Tenth-Century Japan], which are short episodes echoing the theme of Section 69 (and which contain no Kokinshū poems), appear in other texts in different sequences or not at all—as though they had been added late and were not carefully integrated. Episodes that seem little more than variations or elaborations of earlier ones may also have been added later. Section 83, containing a Kokinshū poem by Narihira in a context similar to the Kokinshū description of a visit to Prince Koretaka at Ono, is followed by a section presenting a poem of unknown provenance in an almost identical setting. Similarly, the abduction story in Section 6, which follows two sections based on Kokinshū poems by Narihira, both hinting at the liaison with Kōshi, is a fanciful elaboration of the Kōshi theme, with, again, a poem of uncertain origin. In a number of texts, there is another abduction story, which sometimes follows Section 6 and sometimes appears elsewhere.

Many other examples could be cited to support the thesis that extant texts of Ise monogatari are the result of a long and complex development. The above will be enough, however, to indicate the hazards involved in attempting to discern in Tales of Ise the hand of a single author with a single theme, philosophy, or purpose. It has been said that the book was intended to illustrate different kinds of love, as experienced by men and women of widely varying ages, temperaments, and social classes, and extending to the attachments between parent and child, friend and friend, and superior and inferior; or that it has the didactic aim of describing the behavior appropriate to a cultivated man; or that it tries to demonstrate the kinds of poems suitable for various occasions, particularly romantic ones; or that the original author wanted to explain the circumstances surrounding the composition of certain poems. Some or all of those motives were no doubt operative at one time or another, but it is probably best to think of the work as a response to contemporary interest in the waka [a Japanese poetic form consisting of 31 syllables]—the poems themselves, the circumstances of their composition, their authors—and in the emerging possibilities of the prose medium.

It can be assumed that the form of the poetic anthology, with its introductory comments preceding the verses, served the original author as a point of departure, and that there is a close relationship between Ise monogatari and one particular anthology, Kokinshū. Of 62 Kokinshū poems in Ise monogatari, the prose contexts of 25, all written either by or to Narihira, correspond so closely to the Kokinshū headnotes that there can be no question of coincidence. Either Ise monogatari has borrowed from Kokinshū, Kokinshū has borrowed from Ise monogatari, or both have drawn on the same source or sources. Examination of other poems common to the two texts suggests fairly conclusively that Kokinshū has not borrowed from Tales of Ise, and that in some cases, at least, Tales of Ise has relied upon Kokinshū. We have already noticed that two poems by Narihira and Komachi, which merely adjoin one another in Kokinshū are in Tales of Ise made the focal points of a romantic episode involving the “man of old” and an unidentified lady. In some versions of Ise monogatari, a Komachi poem from Kokinshū is presented as a composition of the “man of old” (Section 143), anonymous love poems that in Kokinshū have brief introductory remarks or none are provided with extended settings, and other Kokinshū poems are revised to fit new contexts.

It is tempting to suppose that an early tenth-century courtier, wishing to compile a collection of exemplary love poems, used Kokinshū as his principal source, drawing heavily on the early anonymous love poems and including most or all of the poems attributed to Narihira, who was already becoming legendary; and that the resultant work, containing few poems by other historical figures, came to be dominated by Narihira's presence. One could argue that gradual accretions tended to increase the book's resemblance to a biography, and that by the eleventh century it was accepted as such, as demonstrated by its alternative titles, Tales of Narihira and The Journal of Narihira.

The long trend among Japanese specialists, however, has been to favor the third possibility listed above, namely, that Kokinshū and Ise monogatari drew on a major common source. Kokinshū, like other imperial anthologies, is based largely on private poetry collections, and it is assumed that Kokinshū poems by Narihira, together with their uncommonly long headnotes, derive from a vanished “Narihira Collection.” The collection, it is thought, also formed the basis for the original version of Ise monogatari, even though it is recognized that at some point there was borrowing from Kokinshū.

The reasoning behind the “Narihira Collection” position is closely related to the question of Ise monogatari's enigmatic title. As mentioned above, the text probably once had an alternative title or titles, but the name Ise monogatari seems to have been well established by Murasaki Shikibu's day. It is the one attached to all known manuscripts, and the one used by all commentators from the Heian period to the present. Monogatari means “tale” or “tales,” but by the late Heian period, or perhaps earlier, the significance of Ise was no longer understood, and attempts at explanations had produced a number of theories, classifiable under three broad categories.

1. Theories ascribing a special meaning to the word. One of these suggests that i stands for “female” and se for “male,” in which case the title would mean “Tales of Women and Men”; another holds that ise is a variant of yose, from yoseru, “to sew, piece together, collect,” and that the title means “Collected Tales.”

2. Theories postulating a special connection with Ise Province or a person named Ise. One of the most persistent of these, apparently dating back to the mid-Heian period, links the work to Lady Ise, the poetess (fl. ca. 877-940), who, according to a highly unlikely story, married Narihira at the age of 12 and revised his draft of the book after his death.

3. A theory that the title derives from Section 69 and related episodes dealing with the Ise Virgin. Its proponents argue either that the Ise episodes are the most important in the book or that the original work must have begun with Section 69.

The etymologies are unconvincing, the proposed connection with Lady Ise suffers from lack of supporting evidence, and it cannot be argued successfully that the Ise Virgin episodes are the most important in the book as we know it. But late Heian scholars reported, and recent research appears to confirm, that there was indeed once a version of Tales of Ise that began with Section 69 of the present text, although it has not survived intact.⁵ The present tendency is to regard it as an early form that was much less sophisticated structurally than extant chronologically oriented versions. It is the effort to account for this line of texts, called the “Imperial Huntsman” line, that has led Japanese scholars from the Tokugawa period on to postulate a hypothetical “Narihira Collection,” rather than Kokinshū, as the most important source for Ise monogatari. Proponents of the “Narihira Collection” theory have usually assumed that the Ise Virgin episode came first in the collection and, accordingly, first in the original uta monogatari, which was therefore called Ise monogatari, in keeping with a not uncommon practice; and that the old name was retained after the original version was supplanted by the present chronological one. A recent refinement by a leading young scholar, Fukui Teisuke, presupposes not one “Narihira Collection” but two. The “Imperial Huntsman” texts would derive, according to Fukui's theory, from Collection A, a random assortment of poems by and to Narihira, with the so-called exchange with the Ise Virgin at the beginning. Collection B would represent a rearrangement of Collection A into conventional categories; and all extant complete Ise monogatari versions would be developments from Collection B. Fukui believes that extant full-length versions, when divested of presumed accretions, progress conventionally through the seasons and on to love poems, as he supposes Collection B to have done. Since Ise monogatari has clearly borrowed from Kokinshū, Collection B would, he says, have been later than Kokinshū, which presumably would have borrowed from Collection A.

The great merit of Fukui's theory is that it disposes of the “Imperial Huntsman” line and proposes a simple, logical development for the present full-length versions. The theory's weakness is that, like the older “Narihira Collection” hypothesis, it rests on scanty and inconclusive evidence. Until new facts come to light, the “Narihira Collection” must remain mere plausible conjecture.

It was noted earlier, in connection with the title, that an ancient tradition named Narihira as the author of Ise monogatari. By the end of the Heian period, scholars recognized that Narihira could not have written the book in its final form, since the text included his own death poem and other materials not available during his lifetime. Many medieval commentators therefore took the compromise position that Narihira had begun the work and that someone else, perhaps a relative or Lady Ise, had completed it. Narihira's modesty, they said, accounted for the description of the “man of old” as a humble old fellow who knew nothing about poetry. Other pre-modern critics gradually produced other candidates—Ki no Tsurayuki among them—supporting elaborate hypotheses with bits of circumstantial evidence, and ignoring inconvenient conflicts. But scholars now admit frankly that there are simply no facts with which to work, and that the problem is unlikely to be solved. Even if a long-buried diary were suddenly to yield the original author's name, it still would reveal nothing about the many other people who must have contributed in major respects to the book.

The problem of dating is equally recalcitrant. There must obviously have been an important period of activity at some point after the compilation of Kokinshū (ca. 905). Whether or not one supposes that to have been the initial period will depend upon one's opinion concerning the source used by the original author. If, as most scholars assume, the principal source was a “Narihira Collection” identical in content, if not in form, with the one used by Kokinshū, Tales of Ise may have originated at any time after Narihira's death in 880 (not before, since Kokinshū contains the death poem). In textual studies designed to establish a cut-off date for the process of development, modern scholars have unearthed no evidence of important accretions after the middle of the tenth century. The Tale of Genji, written around 1000, describes Ise monogatari as old. It can be concluded, therefore, that the work had assumed a fairly stable form by the beginning of the eleventh century, and probably several decades earlier.⁶

In summary, then, Tales of Ise is an anonymous work, the product of a gradual development that took place around the first half of the tenth century. In form it resembles a poetic anthology with greatly expanded and fictionalized headnotes. Its 209 poems deal with all the chief topics of Japanese court poetry, most importantly with love. A majority of the verses are anonymous. Because a famous ninth-century poet, Ariwara Narihira, is the one historical person who figures in it prominently, both as an author of poems and as a protagonist, it has sometimes been regarded as a biography of Narihira, but to view it in that light is to give it a unity and historicity that it lacks. It can best be approached not in terms of preconceived notions about its theme or purpose, but as an interesting combination of poetry and prose, and as a source of insights into the psychology, values, and behavior of Heian society.

Notes

1. See Waley, The Tale of Genji; Bonneau; Nippon Gakujutsu Shinkōkai, The Manyōshū: One Thousand Poems; and B/M. Vos's study of IM includes a literal translation of the best text, but it is not intended for the general reader. Keene's anthology, pp. 67–75, contains a selection from IM translated by Vos and Richard Lane.

2. There are 209 poems in most versions of IM. About a dozen of them appear in MYS, and thus belong to an earlier period. Sixty-two others, of approximately 30 percent of the total, appear in KKS (completed ca. 905); 11 others in the second imperial waka anthology, GSS (commissioned in 951); and 23 others in a large private collection, KWR, which probably dates from around the third quarter of the tenth century. (IM and KWR share 68 poems, including five in MYS, 33 in KKS, and six in GSS.) Of 96 poems shared by IM with KKS, GSS, or KWR, one-third or more were written by Ariwara Narihira, about 15 by people associated with Narihira or roughly contemporary with him, and virtually all of the remainder by anonymous authors. Most of the anonymous poems probably date from the early ninth century. This leaves in question about a hundred additional poems, many of which were attributed to Narihira by medieval commentators. Some may actually be his, others were very likely commissioned by one or another in IM's putative succession of authors, and others were probably old poems either preserved in now vanished collections or otherwise known to an IM author. Of the group as a whole, it can be said only that they are at least as old as IM itself, i.e. that they are for the most part probably no more recent than the mid-tenth century. For further discussion of IM's date and authorship, see pp. 64–65.

3. B/M, p. 171.

4. Sansom, A History of Japan to 1334, Chap. IX.

5. Seidensticker, pp. 33–34.

6. Ikeda et al., p. 507.