SOURCE: Carter, Steven D. “Rules, Rules, and More Rules: Shōhaku's Renga Rulebook of 1501.” Harvard Journal of Asiatic Studies 43, no. 2 (December 1983): 581-642.

[In the following excerpt, Carter analyzes the evolution of rules for linked verse.]

Every literary work is composed according to genre conventions of one sort or another. Usually these conventions operate at the unconscious level of artistic creation, guiding the writer in the choices he must make. But occasionally the “rules” of literary art achieve a more explicit formulation in the literary record, as in the Italian sonnet, or even in the modern gothic romance. Such is also the case with the rules (shikimoku) of Japanese linked verse (renga), which comprise what is perhaps the most detailed set of genre conventions in world literature.

The first important statement of principles for linked-verse composition is contained in Yakumo mishō (1221), a poetic treatise written by Emperor Juntoku (r. 1210-1221) at the height of the Shinkokin era, when court poets were beginning to show serious interest in the renga as an art form. Although most of the “rules” contained in Juntoku's work have little application to the genre in its maturity, even from this early articulation a few general ideas emerge with clarity. The first is that each verse in a sequence should be independent, both grammatically and semantically—that it should represent a total poetic conception and not merely a fragment. This meant that each verse had to stand on its own as a statement; no verse could be simply part of a thematic series. The second principle that becomes clear from a reading of Yakumo mishō is a similar one: that no one conventional theme (spring, love, travel, and so on) should continue over more than a few verses in a series. In other words, the ideal of linked verse, at least in Juntoku's eyes, was that all the categories of the court tradition be represented in each hundred-verse sequence (hyakuin), with no one category boasting anything like dominance. And the final principle that is apparent in the early Kamakura rules is that the repetition of categories or of links is not consistent with the ideals of the genre: “It is bad to introduce the same thing too many times in a hyakuin.” The vocabulary of linked verse was of course limited, and so were the categories of composition, but the ideal of variety remained foremost. “One composes linked verse,” says Juntoku, “by continually moving the sequence in unexpected directions.”¹

The history of renga rules after Yakumo mishō takes the form of a gradual accumulation of detail. In the late Kamakura and early Muromachi periods the simple principles gave way to great catalogues of prescriptions demanded by poets actually confronted with disputes in the renga session. On the surface these catalogues seem both incredibly complex and hopelessly trivial. But the earlier ideals of linked verse were never discarded by later rule makers. Indeed, the aim of the new rules was nearly always to realize the abstract ideals of early poets: to preserve balance in the presentation of ideas and images while at the same time creating a sense of constant change and variety in every sequence. Thus it became vital, over time, that certain images of the court world—blossoms and the moon, for example—be restricted in their usage. Tradition almost required that such images be given special treatment, while the ideal of thematic discontinuity demanded that such treatment be moderated and controlled within a larger pattern. This contest between the thematics of the court poetic heritage and the kinetic aims of the renga aesthetic of play informed the rules' gradual evolution from general principles to specific prescriptions. Finally, the rules made the genre into a form of poetry dominated by a peculiar aura—the aura of disordered order, the literary equivalent of the Zen landscape garden. So although the effect of the rules on any specific example of renga art may appear to be one of randomness, behind this randomness is concealed an ordering and organizing dynamic.

Although other thirteenth-century examples are unfortunately not extant, we know that soon after Yakumo mishō's appearance in 1221 such men as Fujiwara no Teika (1162-1241) and his son Tameie (1198-1275) were busy compiling rulebooks of their own for court linking sessions.² But the first truly significant rulebook to appear in the historical record after the Shinkokin age was a compendium of rules put together by Nijō Yoshimoto (1320-88) and his tutor, Gusai (d. 1376?), over a long period of labor that reached fruition in 1372. Entitled Ōan shinshiki (The New Rules of the Ōan Era), this work was no more original in its conception than earlier rulebooks seem to have been.³ In fact, Yoshimoto's purpose in writing the work was probably to standardize conventions in a world of competing renga masters.⁴ Still the 1372 rulebook is valuable for a number of reasons, not the least being because it introduces the most basic prescriptions of the tradition—restrictions on seriation, intermission, and occurrence—in terms that remain valid for the entire history of the genre. His restrictions on seriation limit the number of verses in which thematic or lexical categories may appear in sequence; his restrictions on intermission serve a complementary function by dictating how many verses must separate different appearances of the same word or category; and his restrictions on occurrence simply limit the number of times certain words may appear in a full sequence of one-hundred verses. In this respect Ōan shinshiki no doubt simply reflects the customs of its time, but it also sets the trend for later rulebooks, all of which adopt an identical stance toward their object, using the same basic framework and terminology. Otherwise it is only in terms of sheer bulk that Yoshimoto's work represents a new development in the history of the rules. Yakumo mishō contained only fifteen remarks on composition; Ōan shinshiki amounts to a categorized list of over a hundred restrictions and thus represents a substantial expansion of the renga domain.

If there is another way in which the 1372 rulebook set the trend for later work it was in its eclectic approach, for it was clearly the product of a long process of study and compilation and not one of original creation. Indeed, Yoshimoto himself voiced the attitude of most medieval poets towards the rules in a statement that may be taken as a credo for later generations: “As time passes, styles change, and there is no need to adhere to old precedents. But neither may one simply follow one's own preferences or incline too far toward one's own prejudices. Rather, one must look to the rules in use among masters of one's own time.”⁵ At first glance this declaration may appear rather open-ended in its implications; but since the masters of the medieval age were largely conservative by training and disposition, their predilection was for the standards of the past. As a result, the rules of the genre retained their neo-classical character in all periods. The final authority in matters of import resided in a community, but a community that defined itself in terms of its relation to tradition.

In their design and in their focus, then, Yoshimoto's rules of 1372 provided a pattern for all subsequent rulebooks. Their prescriptions on a variety of topics ranging from the use of base-poems (honka) to more esoteric information on the avoidance of “remote repetition” in linking became the model for the rules of later periods of renga history. One might even say that all later contributions to the rules read now as footnotes to Ōan shinshiki, although some of those footnotes turned out to be much longer than others.

For about seventy years after its appearance, Ōan shinshiki served, in court circles at least, as the standard rulebook. But as the renga grew in popularity so did the need for more detail in the rules. Yoshimoto himself added several appendices to his original work, and other poets too wrote their own handbooks for composition. While the authority of Yoshimoto's effort was seldom challenged, then, the adequacy of his work to the task of regulating a renga session was put into question regularly. This lead to a whole series of short additions to the rules—no one seen as complete or final, but all adding to the confusion of the renga world. And yet it was not until the 1450s, well after the death of even Yoshimoto's younger disciples, that there was a more concerted attempt to revise the rules in accordance with more contemporary practice. And then the man who put his hand to the task was none other than Yoshimoto's own grandson, Ichijō Kanera (1402-81).

Author of numerous studies of court lore and literature as well as of renga accouterments such as Renju gappeki shū (1476?), Kanera was one of the foremost scholars of his century and a prominent figure among the cultural elite of court society. He was heir to not only his grandfather's library but also to the latter's social position: both Chancellor and Regent in his time, he was the confidant of emperors, tutor to the shogun and his wife, and teacher to a whole generation of younger scholars and poets. It was perhaps less his erudition than his pride, however, that motivated him in his revision of Yoshimoto's rules, for he is known to have been intent on the idea of surpassing his grandfather's considerable achievements as a scholar.⁶ As might be expected, his status as Yoshimoto's literal and figurative heir won wide acceptance for his additions to the rules, called rather unpretentiously Shinshiki kin'an (1452), or Suggestions for a New Day. Incorporated into Yoshimoto's work as an appendix, Kanera's new rules became the basis for serious renga composition throughout the later part of the fifteenth century.⁷

As in the case of the rules of 1372, the Kin'an was not simply the product of one man's efforts, nor did it represent only one man's opinions. It was another statement of changes that had taken place in the rules over time. Before writing the work Kanera is known to have consulted with the greatest renga master of his day, Takayama Sōzei (d. 1455), and it can be assumed that he asked the opinions of other important poets too.⁸ Thus the contents of the Kin'an represent a consensus and reflect a growing need for more complete rules. Most of the comments in Kanera's work are in the nature of additions to Yoshimoto's lists. The words “bird,” “fire,” and “jewel,” for instance, are added to the number of things that may appear three times in a hyakuin; similarly, Kanera adds a few words to the list of things that must be separated by at least one verse—the category Plants and the words “garden thicket,” “autumn paddies,” and so on. Other entries emend slightly or clarify Yoshimoto's original statements in the light of current practice. The framework of the earlier rulebook remains intact, making Kanera's text entirely what its title suggests it to be—a group of new rules for a new age. But these additions were nonetheless important in their own time. Above all, they were authoritative, meaning that in some cases they actually replaced the rules of the past, or at least established a new interpretation of Yoshimoto's sometimes ambiguous intentions. While their influence on the essence of renga art may have been minimal, then, the effect of Kanera's suggestions on the day-to-day composition of renga sequences was probably considerable. They were not laws, of course, because senior poets were always free to accept or reject them according to their own practice. But Kanera's reputation guaranteed his work respect among even the highest practitioners of the art. At the very least the Kin'an increased the number of words to which the restrictions generally applied, thus bringing more vocabulary—usually from the imperial waka anthologies—into the process of conventionalization.

The next major rulebook to appear after Kanera's Kin'an was also a kind of footnote to Yoshimoto, albeit a more extensive one. Bearing the prolix title Renga shinshiki tsuika narabi ni Shinshiki kin'an tō (literally, The New Rules of Linked Verse, With Additions, Suggestions for a New Day, and Other Comments), it was compiled in 1501 by Shōhaku (1443-1527), a disciple of Sōgi (1421-1502), the greatest of renga masters. As with Kanera's work, Shōhaku's is a compendium of the rules current in his time. In his case, however, the time was the age of the hyakuin's maturity as an art form—the age of Sōgi, Kensai (1452-1510), Sōchō (1448-1532), Sōseki (1474-1533), and of course Shōhaku himself. Perhaps for this reason, the rules of 1501 have long been considered the definitive rules of the renga tradition.

Although the exact manner in which Shōhaku compiled his rulebook is not entirely documented, after a detailed comparison of texts and much tracing of the poet's activities during the last years of the fifteenth century, Kidō Saizō has concluded that the 1501 work was the result of a labor of editing and collation that had begun as early as 1482. In the spring of that year Sōgi and Sōi (1418-85), a prominent warrior-poet of the older generation, had met at Arima hotsprings in Settsu to compile a short appendix to Yoshimoto's rules. Shōhaku, acting as Sōogi's student and perhaps partly under his master's direction, continued the process of redaction, gathering rules from various sources written after Kanera's time with the idea of producing a definitive rulebook for his day. His work, which represented another great increase in the number of individual rule entries, reached completion nearly twenty years later, in 1501.⁹

One might expect of Shōhaku, after his many years of study and compilation, a thorough revision of the rules, but once again a different approach to the task of emendation was taken. Rather than an open revision, Shōhaku's work takes the form of an interlinear commentary—sometimes a critical commentary—on the rules of Yoshimoto and Kanera. To read the rulebook of 1501 is to read Yoshimoto, Kanera, and Shōhaku, along with some anonymous voices, in a kind of running discussion or argument. Preserving the rules in both their original and emended forms, Renga shinshiki tsuika narabi ni Shinshiki kin'an tō is thus a complex and sometimes confusing text—one that has contributed greatly to the reputation of linked verse itself as a rule-bound genre. But the work's greatest fault is in a way also its greatest virtue, for it allows the reader a chance to see exactly what kind of changes had taken place in the first century and a half of the rules' existence.

Shōhaku's comments are once more primarily in the way of additions and clarifications. But periodically he also indicates actual changes in practice and perception. In the following example, for instance, he takes Kanera's side in an “argument” over a matter about which Kanera himself seems to show some equivocation:

Yoshimoto: Once one has linked “wind” or “spring haze” to a blossom verse, one should not repeat the same link again in the same hyakuin. This rule holds even if the links are widely separated. And the same holds for all other cases of this sort. [Kanera: In modern times we do not go so far as to prohibit the repetition of links such as the one described above between “wind” or “spring haze” and a blossom verse. [Shōhaku: Current practice abides by Kanera's direction.]] It may well be, however, that we should follow The New Rules of the Ōan Era in this matter.¹⁰

This sort of “dialogue” between the three authors of the 1501 text makes it a frustrating experience in reading. Out of such a reading, however, emerges a fairly clear picture of the rules of Shōhaku's day. Shōhaku's comments often show that opinion was divided upon word usage and categorization: indeed, his work is more than anything an attempt to put at end to such disagreements. Numerous words have been added to each of Yoshimoto's original lexical and thematic categories; and the lists of words restricted as to seriation, intermission, and occurrence have grown apace. Yet the categories themselves are unaltered in substance, showing once again the conservative nature of the renga tradition in its most practical manifestation. Shōhaku's contribution to the growth of the rules was one of quantity primarily, and not of quality.

The new rules of 1501 seem to have been accepted very widely as the new standard for renga composition. This was partly because they represented the combined wisdom of all the major figures of the tradition, including Gusai, Yoshimoto, Sōzei, and Kanera, not to mention the contemporary masters Sōgi and Shōhaku. But the rapid acceptance of the new rulebook also says something about the state of the world of linked verse in the early sixteenth century. By this time linked verse had become a recognized institution with semi-official masters in most major castle towns, many of them disciples of Sōgi or his senior students. And the vehicle of composition, the hyakuin, was also by this time an increasingly well-defined genre—a fact to which the continued development of the rules itself attests. Since the environment of linked verse in Shōhaku's time was more rigidly controlled than it had been even in Kanera's time, uniformity in both composition and interpretation was more easily enforced.

Given the implications of this social context, one is not surprised to find that during the sixteenth-century renga masters elaborated considerably on Shōhaku's initial efforts. Beginning with Sōboku (d. 1545) the process of proliferation reached grand proportions, with many poets publishing their own commentaries on the rules, which usually included additions as well as interpretations. By 1600 there were perhaps five or six times as many words contained in the rules as there had been in 1501. Mugonshō, published in 1598 by a disciple of the great master Satomura Jōha (1524-1602) after an exhaustive survey lasting some twenty years, represents a culmination of this trend.¹¹ Essentially a dictionary of the rules that arranges the rules according to the iroha order, Mugonshō is encyclopedic in scope. And since it makes frequent clarifying references to Shōhaku's work it is invaluable as an aid to the interpretation of the rules in earlier periods as well. This is less true of Ubuginu, a mid-Edo period work of some two hundred printed pages compiled by an unknown poet. The latter is reliable as a source for rules of the Edo period but not as useful as a handbook for the reading of sequences from the golden age of renga art in the late years of the fifteenth century.¹²

The most important work for the consideration of hyakuin written during the days of the greatest renga classics is Renga shinshiki tsuika narabi ni Shinshiki kin'an tō, Shōhaku's rulebook of 1501. For this reason I have chosen to translate it in full. Esperanza Ramirez-Christensen has referred to the rules as a “grammar” of the genre, and the metaphor is an apt one.¹³ Originally Shōhaku wrote his rules as a help to poets and referees faced with the practical job of making complex decisions in the atmosphere of a renga linking session. But when inverted the rules can also be a guide to the classical reading of a sequence—to a reconstructing of the most basic interpretive strategies involved in any approach to the hyakuin as a genre. …

Notes

1. NKT, Supplement 3, p. 204.

2. The rulebooks of Teika and Tameie are among those listed in the catalogue of the Reizei House Library. Works by Fujiwara no Yukiie (1223-75) and Fujiwara no Nobuzane (d. 1265) are also on the list. See Kidō Saizō, Renga-shi ronkō, 1 (Meiji Shoin, 1971), pp. 185-86.

3. Ōan shinshiki is actually a later title for what was known at the time as Renga shinshiki (The New Rules of Renga). In order to avoid confusion with Shōhaku's Renga shinshiki tsuika narabi ni Shinshiki kin'an tō I have used the former title throughout this study. A text of the work is available in Okami Masao, ed., Yoshimoto rengaron shū, 1, Koten Bunko, No. 63 (1952), pp. 7-23. Ōan shinshiki represents the third of Yoshimoto's attempts to standardize the rules, the first being an appendix to Hekirenshō, ed. Ijichi Tetsuo (NKBZ, 51, pp. 46-61) and the second an appendix to Renri hishō, ed. Kidō Saizō (NKBT, 66, pp. 56-67). After the initial compilation of the 1372 work Yoshimoto also added a number of appendices (the tsuika of Shōhaku's title), all of which are included in the text of the 1501 rulebook.

4. For a good summary of the process by which Yoshimoto's rulebook came into being and the motives behind that process, see Shimazu Tadao, Renga no kenkyū (Kadokawa Shoten, 1973), pp. 175-201.

5. Renri hishō. NKBT, 66, p. 36.

6. Kanera is notorious for saying that he had surpassed even the famous Sugawara no Michizane (845-903) in a number of ways; Nagashima Fukutarō sees Kanera's work in court lore and his compilation of the Kin'an as particular evidence of a desire to best Yoshimoto. See Nagashima Fukutarō, Ichijō Kanera, Jimbutsu sōsho, 31 (Yoshikawa Kobunkan, 1959), pp. 97-98, 155.

7. Okami's text of Ōan shinshiki (see note 3 above) contains the Kin'an as an appendix.

8. Sōzei was official Renga Master of the Ashikaga government at the time. A note at the end of the Kin'an says that Kanera consulted with Sōzei before writing his rules, and it seems certain that he would also have consulted with other major figures such as Gyōjo (1405-69).

9. Kidō, Renga-shi ronkō, 2 (1973), pp. 495-516.

10. See translation, Section III.

11. It should be noted that since Mugonshō was written nearly one hundred years after Sōgi's death, it often interprets the rules in ways not entirely relevant to fifteenth-century practice. It is therefore as a cross reference that the work is valuable. For bibliographic information on Mugonshō, see note on abbreviations.

12. For bibliographic information on Ubuginu, see note on abbreviations.

13. Esperanza Ramirez-Christensen, “The Essential Parameters of Linked Poetry,” HJAS, 41 (1981), 555-95.

The following abbreviations are used in the notes:

MGS: Mugonshō in Nihon koten zenshū, Ser. 5 (1936).

NKBT Nihon koten bungaku taikei, 102 vols. (Iwanami Shoten, 1957-68).

NKBZ: Nihon koten bungaku zenshū, 51 vols. (Shōgaku-kan, 1970-76).

NKT: Nihon kagaku taikei, 10 vols., 3 supplements (Kazama Shobō, 1956-64).

RGPS: Renju gappeki shū, in Rengaron shū, Vol. 1, ed. Kidō Saizō and Shigematsu Hiromi, in Chūsei no bungaku, Ser. 1 (Miyagi Shoten, 1972).

RRS: Rengaron shū, ed. Ijichi Tetsuo, Iwanami Bunko, Nos. 5114-5119 (1953, 1956).

UG: Ubuginu, in Renga hōshiki kōyō, comp. Hoshika Sōichi and Yamada Yoshio (Iwanami Shoten, 1936).

ZGRJ: Zoku Gunsho ruijū, comp. Hanawa Hokiichi and Hanawa Tadatomi, 71 vols. (1923-30).