Darko Suvin                                                                                                        (1996, 9,340 words)

It should be clear that our study of the nô must include the resources of comparative literature. Indeed …we must study the nô drama from every possible direction, using every useful approach.
Konishi Jin’ichi, 1960

0. How May Intercultural Theatre Studies Approach Nô?

0.1. To begin with, two notes about the object and approach of this essay. First, any student of Nô is acutely aware that its verbal signs are in constant interaction with a rich array of musical and other acoustic signs as well as with  dance and other optical signs; the final dance of the shite¹  takes by itself often one third of performance time. Though it is misleading to call the verbal text of Nô a libretto, encompassing meaning resides only in the performance as a whole. Nonetheless, my approach is predicated on the hypothesis that the integral “stage story” is the backbone of Nô’s “performance text,” and that some crucial parameters of this theatre story-telling can be inferred from natural language and must be fruitfully isolated for an initial discussion.

Second, I take it that it is by now widely accepted that there is no stable object “out there” which would be analogous to the Moon seen imperfectly or anamorphically in our telescopes or spectrometers. The first thing that strikes one in Nô studies is (at least in my experience) how inevitable it is to foreground this realization in them. Instead of approaching a stable ontology, wiser and sadder today, we have to deal with the epistemology of competing societal stances towards, and indeed competing definitions of, an entry in the imaginary cultural encyclopedia of a given spacetime (Eco). What such an entry refers to, its referentiality, is perhaps more arbitrary  in the case of the Nô play than in most other fields of theatre studies. This will be exemplified by the quite basic or constitutive decision of what are the significant units of Nô. For the purposes of this essay I shall attempt to skirt the minefield of referentiality (cf. at least Whiteside) by refusing both the contemporary ontological absolutists  who believe they can travel with Plato and Saul of Tarsus from shadowy contours to the deep Truth, and the deconstructionist absolutists who believe no legs are lost in minefields and no delights incurred or emotions  hurt in sexuality since all these are–since reality in general is–simply matters of discourse. To my mind the golden in-between position is to be found in a dialectical refusal to segregate to opposed poles the seeing subject and the object seen, since the act of seeing presupposes and proves that they share the same structure. As seminally introduced by Hegel, “It becomes clear that behind the so-called curtain which is supposed to veil the inner, there is nothing to see unless we step behind it, not just in order that there be someone who can see, but equally in order to have something there to be seen” (129; tr. from Taylor 147). An analogous denial of enclosed interiority or depth for both observer and cultural text is suggestively formulated in Paul Ricoeur’s two-tier discussion of sense and reference in texts:

This is perhaps most precisely phrased in Brecht’s discussion of criteria for the art of acting:

The positions (enunciations) were thus not brought near to the spectator but distanced, the spectator was not led but left to his discoveries. (17: 984)

For such a stance, it is impossible to believe any longer that  cultural constructs are finally anchored by referring back to a pre-existent eternal reality or inside to an equally fixed “objective” essence of Nôness. Rather, the pragmatically useful concept and category of “Nô,” its world and its values, would become present or unfold in front of the (culturally constructed) text, as the observer’s discovery, as a methodological necessity and mandatory presupposition for speaking at all about this congerie of phenomena. As Fredric Jameson concluded: “The study of the referent, however, is the study, not of the meaning of the text, but of the limits of its meanings and of their historical preconditions…” (Ideologies 108); meaning is a given text’s “‘historically operative’ significance or function…, the meaningfulness of a gesture that we read back from the situation to which it is precisely a response” (Ideologies 145-46). The quest for meaning is unavoidable because to my mind inseparable from the human condition, the interpretive mysteries subsist, yet not in any monadic unit that bestows meaning –be that subject or object. If we are still forced to use these terms, let us at least have meaning arising between the two societal, historical and situation-bound,  constructs of knower and known. The essence is historical and dynamic, the founding epistemological unit is not a monad but a relational dyad (this too can be found in Hegel).

In that sense–but only in that one–I believe the term “Nô” does refer, albeit to a subject-object constellation rather than to an objective reality; so that it still makes sense to talk about Nô (or Renaissance theatre or Brecht).

0.2. The shifting and evidently “socially constructed” nature of a unit of Nô studies–a good instance of what Saussure identified (if misnamed) as the arbitrariness of linguistic and other signs- -can be followed on at least three levels. First, when we speak about any Nô title, which text and/or variant are we speaking about? Texts and scores were until this century transmitted exclusively by the performing “extended families” (ryû), which had their own criteria of selection. Konishi has noted how “many plays popular today were not well liked during the Muromachi period, ” including some of Zeami’s masterpieces; the staging of Kinuta, possibly Zeami’s favourite, lapsed for about a quarter millennium so that no performance traditions survived. Obversely, evidence  from  various sources, centrally from  Zeami’s  own writings, establishes that “a number of important works performed at that time have not been retained in the [modern] repertory” (Rimer xxvii). Other aspects were simply changed: Konishi goes on to discuss how (e.g.) the dramaturgic agent played by the waki was in a well-known play degraded from daimyô (feudal lord) to yeoman.² In fact, from the ca. 250 plays today considered canonical, only about 100 are performed in the original, pre-Edo verbal text. “Over half a millennium has passed since Zeami himself performed his plays, a period in which changes have inevitably taken place,” remarks Hare, so that (a  striking and accurate parallel) a discussion of Zeami in terms of today’s plays “would be tantamount to discussing Shakespeare’s style using Polanski’s Macbeth” (54); he follows this up with a careful discussion (55-61 and 267) of what we can assume has changed from Zeami’s to present-day performances. A possible conclusion might be that the central red thread of plot outline has in most cases changed little, but that there are numerous, sometimes startling, transformations in many segments. In some cases the play variants may diverge widely,  amounting to rewrites, and the modern printed versions may be “often simplified, even bowdlerized” (Rimer xxvii). Furthermore–although we seem to know very little about the changes in dance and instrumental music–all proportions must have been radically altered by the approximate doubling of performance time in the quarter millennium after the latter part of 16th Century, when the performance of one Nô took no more than 45 minutes (see O’Neill’s painstaking proof, Early Nô 88-90)! Other major shifts included the introduction of the tsuyogin and yowagin –roughly, “major” and “minor”– scales ca. 1680,  the increased stylization of masks and the canonization of their kinds ca. 1600 (cf. for both Konishi, “Approaches” 13 and 15-17) as well as the freezing of fixed dimensions for the Nô stage (cf. Rimer xxiv).  Finally, political monopolization of all Nô groups by shogunal patronage led to a strong ceremonializing trend of shortening texts, cutting down on secondary and parallel –especially female –agents, and slowing the dances (see Yokota 261-66, 268-70, and passim, also Brazell, “Nature” 214ff.). In sum, there is general agreement that the original Nô performances were more realistic and much less solemn than after the Edo period (cf. O’Neill, “Background” 20), so that, in a diplomatic formulation, “the kind of stately experience usually offered today seems at some variance with the rough-and-tumble world described in [Zeami’s] treatises” (Rimer xxiv).

However, all such deep changes in play texts shall not be pursued further here (I attempt to discuss in somewhat greater detail the case of the Nô-play Tanikô, where the shite-role probably shifted to a different agent in Edo time,  in Suvin, “Use-Value”). Instead, I shall briefly face two other levels. For, second, if we take as our unit one single play, how many plays are we speaking about? And third, if we wish to approach any significant generalization about Nô, how do we categorize them into higher-order units or groups?

On the second question of this subsection, in the Nô’s heyday, the Muromachi period (14th to 16th Century), perhaps more than 1,000 plays were written, of which about 600 survive. To them another 1,000 or more plays were added later–mainly in the Edo period, 300 years on–but these were not at all or very rarely performed, so that O’Neill suggests they should be considered as (a separate sub-genre of?) closet plays.³ However, only about 250 are in the canon performed by the five Nô schools, the first trace of whose establishment–a list of Nô plays no longer in the active repertoire–stems from 1686 (Matisoff 254). In this canon, there are perhaps less than 100 plays performed  and discussed with any frequency (according to a forthcoming investigation of mine, there are over 150 Nô plays translated into one of the main European languages in reasonably accurate versions, but only 84 of them more than once). Yet frequently there exists “a misapprehension that the existing repertory or canon of 200 plays suffices to study the nô….Most of the plays in the acting canon are excellent dramas, superior to the rest in many respects, and we seldom find a masterpiece among the uncanonical plays….Is it reasonable, [however,] to discuss the style–much less the history–of nô drama on the basis of the present-day repertory…?” (Konishi, “Approaches” 3-4)

 My subsequent argument demands a reminder that in this canon Zeami is the central “playwright,” in the integral sense including text, music, and probably dance (he was also protagonist and organizer-“director,” of course). Attributions in this genre– where plays are often anonymous or rewrites–are always difficult and most frequently a matter of guesswork subject to oscillations in what one might call constructive and deconstructive waves, so that Zeami’s opus is estimated at somewhere between 50 and 110 titles. It has been increasingly recognized as not only splendid but also very personal or idiosyncratic, rather different both from his predecessors (such as his father Kannami) and from his successors, and as having set up some central parameters in terms of which Nô is being discussed and perceived– first by the players themselves, and consequently by the often intensely clannish Japanese critics,  on whom necessarily much of the foreign discussion builds. (To this should be added in our century the impact of the newly published theoretical writings by Zeami;  they are highly interesting but also obscure and will be only fugitively mentioned here, within a methodological choice that favours implicit over explicit poetics.) A first conclusion is, therefore, that a contemporary critic must be very cautious when generalizing from the plays seen in performance and known from translations or critical discussions to Muromachi Nô in general. Our view today is a combination of Edo-period and 20th-Century filtering and refraction. The resolute shogunal moulding of Nô into an official institution cut it off from the commoners and congealed it into rigid upper-class forms; and the post-Meiji refraction of the last 100 years largely retained that tradition within its own, nostalgic–and sometimes indeed chauvinist–agenda of a quintessential “Japaneseness.” Thus, a selection of plays and aspects to be foregrounded, quite analogous to the procedures of condensation and relocation (Verdichtung und Verschiebung) Freud discovered in “dreamwork,” is inevitable and sometimes very illuminating, but only on condition that it be clearly seen as what it is: a particular slant (or series of slants) rather than “as it really was” (the historian Ranke’s illusory wie es eigentlich gewesen). Nô theoreticians in particular should bear this in mind when attempting speculations such as those in my following sections.

 On the third question, the contingent nature of determining significant units or macro-texts (cf. Barthes 155ff.) holds in spades for the various Nô categorizations. The commonest one, used in actually composing a Nô-performance program (though decreasingly so in practice)  is a system of division into five categories (goban date), variously called First to Fifth Piece; or Deity, Warrior, Woman, Miscellaneous, and Final Piece (kami, shura, kazura, zatsu with various subtitles, and kiri-Nô); and other names. It reposes squarely on the nature or type of shite-role in each group, though there is strong evidence for “an older form of Nô before Zeami established the Shite as the single central figure” (Shimazaki 42), and at some point after Zeami the waki-role became a second focus in some prominent plays. The first category consists of auspicious plays about the blessings by deities, the second is about dead warriors recounting their memorable downfall, the third consists of graceful plays about beautiful women, living or dead,  and about the spirits of plants or non-sentient beings. It is quite remarkable that only the first two categories are clear-cut. The third begins to grow eclectic; the fourth  or miscellaneous may have for shite-role mad people, living warriors, vengeful spirits, female deities, etc., while the fifth-category shite-role is a beast, demon, non-warrior spirit, etc. However, though extremely important from the Edo period to the present, this system was arrived at precisely in the conservative systematization of the Edo period and it cannot be extrapolated backward except as an open theoretical imposition on a period that did not know it (cf. Rimer xxvii). I shall use it as a convenient shorthand to get to more important distinctions. Some other distinctions, such as the division into of geki or “dramatic” Nô vs. furyû or “spectacular” Nô (cf. Hoff-Flindt 214 and Terasaki passim) may be getting at worthwhile oppositions but seem to me too impressed by a quasi-Aristotelian, in fact European 19th-Century drama-model, and therefore not very useful today. More useful may be the distinction between plays where the shite remains in the same role throughout, called one-part Nô, and those where he plays two roles, changing mask and costume, called two-part Nô. Still, a theoretically minimally satisfactory grouping of Nô plays on the basis of clear, compatible, and encompassing parameters is not yet present.

 0.3. My initial denial of objectivism does not at all mean (as I concluded in 0.1 against philosophical idealists and agnostics) that there is  nothing out there interacting with our critical eyes and constricting what they may see. As Einstein once remarked, the belief into the existence of an external world, independent of any single observer, is the basis of all knowledge (though we would have to add today that the knowledge is constituted by the interaction of observers with observed). In Shakespearean studies, e.g., Terence Hawkes’s famous  “sense of a text as…an area of conflicting and often contradictory potential interpretations, no one or group of which can claim ‘intrinsic’ primacy or ‘inherent’ authority…” (117) manifestly has to rely on a provisionally stable text and also on a pragmatic authority, however socially contingent, for (its) given purposes. I quite agree with him that “Our ‘Shakespeare’ is our invention” (124); but I have argued at some length elsewhere that it is not only our invention (Suvin, “Modest”). Thus, on the one hand, all interpretations are negotiations between a set of presuppositions and an object partly constructed by the obscure wish to articulate those same presuppositions. But on the other hand, the object and the presuppositions are partly constricted by material limits, resistances, and vectors coming from  “out there.” The more one knows about the practice and history of interpretation, the more one realizes that what the interpreter finds is not the fixed meanings of a text (or of the Book of Nature) but only other interpretations. One could apply to the Edo-period categorizations Harry Berger’s discussion of authoritative closure by “citation” of past categories (though I believe you can never “prevent” interpretation, not even in societies with a myth-hegemony):

 Myth and ritual, for example, are citational acts that…prevent against interpretation by closing down on unbound meaning. Such performances can succeed only because they presuppose and continually reinvent an iterable corpus of detextualized ‘texts’ that privilege a certain interpretation of cosmology, values, and norms, i.e., an interpretation of what there is in the world, what one ought to think about it…. (155)

 Therefore, a second conclusion from the foregoing discussion, complementary to the first one, might be that, while not disregarding the significant continuities within the nonetheless changeable traditions of the Nô acting schools, both Edo and post-Meiji Nô can be illuminated by Hobsbawm’s definition of an “invented tradition.” This is:

 …a set of practices … which seek to inculcate certain values and norms of behaviour by repetition, […and] normally attempt to establish continuity with a suitable historic past….The peculiarity of “invented” traditions is that the continuity with it is largely factitious. In short, they are responses to novel situations which take the form of reference to old situations, and which establish their own past by quasi-obligatory repetition. (1-2)

 “Seek to inculcate” is a bit one-sided, and Hobsbawm speaks later  of other types of “invented traditions” that establish, symbolize or legitimize real or artificial communities and institutions (9). Nô is clearly not instituted in the ad hoc way of various European and other nationalisms, “mass produced” since the 19th Century, but more like Scottish “Highland Traditions” or English Christmas carols, i.e. with a significant but unacknowledged break in continuity (Hobsbawm 6-7); with these caveats, much of the rest in the quote and argument seems applicable.

 Thus, there is no cognitive reason why a contemporary critic cannot, with due humility, attempt to emulate the Edo-period shites of Nô or the earlier 20th-Century Japanese critics who have in fact excogitated the present canon and categories of Nô, i.e. “invented” the present “tradition,” and why she or he cannot find for new purposes some new units and categories–or indeed new functions for carefully reformulated old, not quite invented categories. Since any macro-text (e.g. “the Nô play”) is always established as a cognitive and pragmatic unit/y by specifiable agents from a specifiable point of view or stance, it stands to reason that for a different purpose all texts current in (say) Zeami’s lifetime might constitute a macro-text. This is in fact what Nô studies routinely do when elucidating any play by Zeami by means of its intertextuality with tankas, Kannami, The Tale of Genji, etc. My present purpose, as suggested by the references to Barthes, Eco, Jameson or Ricoeur, is “not an ultimate appraisal but some answers to the practical questions: what particular attractions and values have the Nô plays for [people] of the twentieth century?” (Wells 154). Now, for various interests these attractions and values will be legitimately different, ranging from interpretations of particular segments or aspects in plays that increase the spectator’s delight to very general questions that may be posed by these plays to a comparative dramaturgy and theory of theatre. While I believe that ultimately any theory is justified mainly by feeding back into an increased understanding of historical texts and attitudes, it may itself at times be a both necessary and enjoyable detour, to what one hopes would be a new vantage point for mapping. This is the road I propose to take here.

 1. A Move Toward Recategorization: Zeami and the Anamorphosis of Warrior Nô From Deity Nô

 1.1. Proposing to practice what I preach, I shall now attempt to build on the crucial  distinction between genzai nô (where the role played by the shite or protagonist is a dramaturgic agent alive in the time of the story) and mugen nô (the “visionary” plays where this agent comes from outside of that quasi-empirical time, i.e. is a deity, spirit, and similar). This division was developed in the 20th Century, the term mugen nô being introduced by Sanari Kentarô in 1930 and genzai nô by Yokomichi Mario in the 1950s (Terasaki 14-17). Thus, it is equally anachronistic in relation to the Muromachi period, and it is not absolute or all-encompassing, but it does seem to repose on some dominant structural features of the plays such as central agent and time-horizon, so that it is better suited for a theoretical discussion than the empirical bricolage of the fivefold categorization. Today the genzai nô–the “everyday” plays in which, to repeat, the shite plays a human living at the time of the stage story–are to be found mainly in the fourth category and partly in the third and fifth categories (Miscellaneous, Woman, and Final Nô). Zeami himself wrote both genzai and mugen nô,⁴ but his major innovation seems to lie in the series of masterpieces which codified the mugen nô.  Thus, for reasons both of fascination with his talent, comparable to very few people in the history of world theatre such as Aeschylus, Shakespeare, Molière or Brecht, as well as of ideological preference by the commentators, the main impression about Nô plays (certainly outside Japan) is based on mugen nô. This is a Meiji-period invention, sometimes traced back to the influential position formulated by Haga Yaichi in 1899: “the essence of Nô is the ghosts” (cited by Terasaki 27); in Europe, this was furthered by the elitist fantasies of Pound and particularly Yeats.

 Yet historically speaking, the defining of Nô in terms of mugen nô only is an optical illusion. Even prescinding from a possible different ontology in medieval Japan, for which numinous beings were perhaps differentiated from empirical ones in other ways than in secularized thought, it seems that at the time of Nô’s inception and early development the  more “realistic” plays–or better, the plays developing exclusively within supposedly empirical time-horizons–constituted its mainstream. Other templates clearly existed,  and in particular the auspicious visit of a godhead in folk-rite-derived plays carried forward by priestly patronage: “some kinds of liturgical dramas,  whose direct posterity is today still found in the kagura plays…of most Shinto shrines. Which, of course, does not mean that Nô itself is religious art….” (Sieffert 16; cf. also, e.g., Honda, O’Neill, Early Nô 85-93 and passim, Raz, Audience 83-85 and passim, as well as Wang for early Chinese parallels) Nonetheless, as has been argued by Kobayashi Shizuo, there is evidence that such old shushi sarugaku (sarugaku plays performed by priests in temples) were not an exclusive or possibly even not a prevailing ancestor of Nô, as compared to the plebeian, empirically oriented and often satirical, senmin sarugaku (Terasaki 22-26; cf. O’Neill, Early Nô 4-9, 87, and passim, Raz, Audience 80 and 98-101). At least one prominent history  of Japanese literature has no doubts at all: “[W]e  must conclude that living-figure nô…–and, moreover, other plays that had a dramatic plot–were the mainstream of nô in its early period” (Konishi, History 524). Finally, if one takes into account all Nô plays written, genzai nô seem to account for almost half of the 2,500-3,000 pieces, while in the canonic 250 repertory plays they may account for half or more (Terasaki 76-77 and 197; other estimates place genzai nô at one third of the repertory, further evidence of the unclear categorizations).

 1.2. I shall begin by presenting an initial argument in the form of a table of parameters which, starting from the dominant categorization into five groups, proceeds to doubt it by means of the division into genzai and mugen nô and proposes to refine this in favour of a trisection. Leaving other matters aside, I shall then advance to focus on the first two categories–the Deity and Warrior plays–which are obviously, prima facie correctly distinguished sets, more or less monolithic because mainly created for the same purposes and same audiences by Zeami.  I am banking on these clear groupings being apt for an initial theoretical discussion which could lead us to what I take to be the central theoretical problem in a discussion of Nô dramaturgy, and of immense consequence for any theory of dramaturgy and theatre, namely: the puzzle of whether conflict is necessarily present in Nô.

What I propose is, it can be seen, a first step (and only first!) toward a new categorization. As was mentioned, beginning with category 3 (which is characterized as having elegant and soft protagonists) the subdivisions of the fivefold categorization grow hopelessly miscellaneous. The very useful division into genzai vs. mugen Nô gives us a good pointer by differentiating in theshite-role between the presentation of living humans and of non-human entities (deities or specters), which entails significantly different time-flows (the arrow of empirically sequential time in genzai nô vs. multiple qualitative times in mugen nô) and time-horizons. We could call this parameter “shite-role reference and temporality.” I believe it may turn out to be necessary for a sensible definition, but it is not sufficient. This seems easily proved by the fact that each of these two categories again contains admittedly heterogeneous groups of plays, with little similarity to each other beyond the shite-role’s reference-cum-temporality. One division of genzai nô, e.g., breaks it down into plays: 1/ dealing with events contemporary to the author; 2/ drawn from previous literature (such as the Komachi plays); 3/ realistic, non-masked portrayals of historical warriors– e.g. Ataka; 4/ episodes from the earlier life of spirit-Nô heroes or heroines, marked by a title beginning in “Genzai,” such as Genzai Matsukaze (Terasaki 65-73). The mugen nô subdivisions are still more heterogeneous, as I hope partly to show in the next section. It is rather as if Kafka, Tolkien, Steven King, Isaac Asimov, and William Morris were all to be put into a single category called Fantasy Fiction (a grouping that alas is being propounded, but has at least the excuse that itdeals with matters no older than the 19th Century and usually not deemed of central generic or genological significance). Other parameters have to be introduced for a basic sense-making categorization. To begin this, in a combination of structural and historical arguments, I propose the above trisection into Deity Nô, Warrior Nô, and Others.

As indicated by the little overlap at the bottom of Table 1 between mugen nô and my “other categories,” the boundary between my second and third categories is deficient. In particular there are in Woman Nô obviously some plays (e.g. Seigan-ji, Eguchi or the plant-spirit plays such as Bashô) which are homologous to the laudatory Deity Nô, and some (e.g. Yuya, Sôshi-arai Komachi, Senju, Yoshino Shizuka) which tend toward a conflict, albeit backgrounded, muted, and/or finally amicably resolved or superseded. In between, some plays are rather similar, possibly homologous, to the Warrior Plays.5 Were this not so inelegant, my second category ought more precisely to be called “Warrior Nô and some Woman Nô and possibly some plays from the goban date Fourth and Fifth Pieces.” It is therefore adopted here as an intermediary step, provisional and clearly needing supersession, for the sole purposes of the present initial essay; it should be considered further at a later stage of the approach broached here.

In the meantime I may, to begin with, claim that I am starting from what most previous categorizations also started from, the quite central attitude-cum-emotional-aura transmitted by the shite-role, who may be largely “the incarnation of some powerful emotion” (Keene, Nô 24).
However, the shite is clearly at the antipodes of an individualistic protagonist, since the emotion this role is transmitting to the audience results from an interaction between her or his desires and the all-encompassing stance toward the world that this agent shares with the audience. Thus, my categorization introduces as parameters what seem undeniably matters of overriding importance for these plays: the shite-role’s–and the audience’s–central desire(s) and its (their) relationship with the play’s cosmological framework (permanence vs. impermanence), which is consubstantial to the type and ranking of the values espoused by the play.

1.3. I shall narrow my focus now to the relationships between the Deity Nô and Warrior Nô categories. It is generally acknowledged that the Warrior Nô were practically Zeami’s singlehanded invention, in 13 plays of which at least half are masterpieces: he certainly or probably composed ab ovo or reworked into its distinctive shape Atsumori, Kanehira, Kiyotsune, Sanemori, Tadanori, Tamura, Tomonaga, Tsunemasa, Yashima, Yorimasa, and possibly also Ebira, Tomoakira, and Tomoe. All of them are even today in the repertory of all five Nô schools (ryû), except for Tomoakira which is in the repertory of four schools. The three dubious titles and three more non-Zeami Warrior Nô in the canon are centrally imitations of his model (93 more “inactive” plays seem to have been written in this mould, see Terasaki 76). How did Zeami come to compose them thus, whence did he take their common distinctive elements and aspects?/6

Now, it would be very possible and I think useful to pose the question what do the Nô deities and warriors stand for or signify. There is little doubt that one of their central intertexts is the power relations and value horizons of Japanese politics (in the widest sense of long-duration orientations) in the country community as a whole and on the shogunal court in particular. Our distinctions between religion and politics do not obtain in the Middle Ages, when (e.g.) the deities were guarantors and indeed personifications of “an ordered country,” as monotonously repeated in one after another Deity Nô. However, before proceeding to this, I believe much more work is due on the signifier level. Therefore, I shall discuss here some central implications, modalities, and consequences of this trajectory of Zeami’s. The generally accepted thesis that the Warrior Nô were created by Zeami taking as their template the Deity Nô seems to me not only correct but inescapable. This argument may be strengthened by the fact that this innovator yet also continuator within a living theatre tradition explicitly championed such a procedure, e.g. when transforming his troupe’s traditional totally fierce demon type and Nô-play into “a variant derived from the form of the warrior,” where the usual “demon form” was transformed by adding to it the “heart/mind of a human.”7 Zeami proceeded then from “the prototype first established in kami nô” (Konishi, History 526 and passim) to Nô plays such as the Warrior Nô–as well as to some Woman Nô–in which the shite plays a human, not a deity. My argument will again be proposed first in the form of a synoptic table.

The originating parameters here are clearly the first two, the play’s overall action and central agent (shite-role), which amount to transplanting the same central understanding of action (or story) as revelation from divine to human dramaturgic agents. However, this shift is simultaneously feasible and problematic. It is possible because the humans are dead, it is problematic because their passions are still human: “Violently my heart yearns for the earthly world,” exclaims quite typically, if in a somewhat simplified translation, the eponymous hero of Yorimasa (Shimazaki 118). There is a common theological or ontological trait between godheads and dead warriors in relation to the time-horizon of other stage roles (primarily the waki-role’s horizon, imaginatively identified by the audience with their everyday horizon). As Professor Brazell put it, “[t]here is a connection between the nature of the [shite-role] and the way in which the time and space are manipulated” (“Nature” 206): both divinities and fallen warriors come from outside, and thus deny the exclusivity of, the quasi-empirical or human time-horizons on the stage, and yet both can manifest themselves within them. Furthermore, both these types of agents appear in two shapes–in a kind of disguise and then in propria persona— which require, and correspond to, the two parts of these Nô plays.

Yet Zeami’s bold shift from Deity to Warrior Nô is also fraught with problems, for the theologico-ontological status of godheads and dead humans is after all very different. Beside coming from outside the waki-role’s human time, they have little in common. In Deity Nô, the numinous fullness of the Shinto divinities makes of the shite-role of the play’s first part (the maejite) a human incarnation of the same deity that appears in the second part (as nochijite) in one of its divine aspects. In Warrior Nô, that joy-bringing succession and clear uplifting movement necessarily gives way to the troubled succession from maejite as misleading spectre– presented by Zeami, in a direct copy of Deity Nô, in seemingly human form–to nochijite as vision of the warlike nobleman from the past (cf. Konishi, History 524-25). True, the nochijite warrior visions are more vivid than the shadowy European stage-ghosts in the tradition of Euripides and Seneca, which as a rule do not attain protagonist status (e.g. in Hamlet or Macbeth). While I doubt that Zeami deals in anything resembling individualist identity or Self (cf. Suvin, “Soul” passim), I concur that “an inability to escape the ties of a past life, love, hate, longing, and pride…[are] characteristics [that] persecute Zeami’s shite” (Hare 242). The vividness can be attributed to various factors of the Japanese cultural tradition, which in Nô come to a head precisely in Zeami’s eduction of the warrior ghosts out of the resplendent Shinto divinities that bring the blessings of fertility in the latter part of Deity Nô. This is why I have substituted the more specific “passion” for the “desire” of Table 1. Yet this same change radically transforms the logic of the story. It enforces and sets into train a set of complex anamorphic manoeuvres so that the stage-action may make sense to the Nô audience.

The resulting play-model for Warrior Nô/8 is by necessity significantly modified. I mentioned that at the beginning its maejite appears on the stage in the same way as in Deity Nô–as a lay person, usually an old man, of the locality the waki-role is visiting. But instead of a felicific (medetai, happiness-bringing–cf. Gundert 225ff.) divine revelation or theophany, the Warrior Nô proceeds then to a tale of woe and regret. This is epistemologically speaking also a revelation, and I have found it convenient to call it the revelation of a final permanence, an absolute cosmic closing chord, as supreme value. But if the dramaturgic syntax is the same, the semantics diverge. If both Nô groupings reveal a permanence, it is in Warrior Nô not a positive but either a frankly negative permanence (the warriors are often in the special realm of torment  for them9); or, at best, it may become, in a second and final revelation, the “positive negativity” or zero of Buddhist nirvana attained at the end of the play (the warriors’ pre-history is usually recounted in verse, music, and dance to the waki-role listener so that he, a priest, may pray for their salvation through purgation of their martial passion as well as of any other earthly attachment–e.g. to poetic fame in the case of Tadanori).

True, in the “non-duality” (Loy) of Mahayana Buddhism a proper passage through desires may be finally conducive to enlightenment; but once enlightenment is achieved, desires are definitel left behind. This is clearly foregrounded in Warrior Nô. For all these reasons, Zeami’s adaptation of the Deity-Nô model has to modify it much more strongly in the second part of the Warrior plays, where the real situation will out. The founding divergence between Deity and Warrior Nô is that the fallen warriors’ final permanence not only contrasts with the impermanence of their earthly life, it also differs radically from the “positive” permanence of the Shinto deities. Reaffirming the evergreen continuity of time’s ongoing cycle (e.g. in Takasago’s marriage of two great historical ages of poetry as magical creativity), the deities ensure and bring forth the future to an audience that is ideally supposed to represent the community. Their fulness of value and time–the potential simultaneity of all times–is in sharp contrast to the dead warriors bringing up only their personal past (albeit typical of a hegemonic social class) for anamnesis and redemption out of time. True, in both Deity and Warrior Nô the audience in the present also experiences delight and probably (more so before the Edo period) a magico-religious awe at seeing in the shite-role’s apparition (cf. Gundert 221) respectively the promise of communal felicity or the representative passion and comforting of the great transgressor (cf. Ruch, “Medieval” 306); this should qualify my oversimplified isolation of positive vs. negative permanence, necessary for a tabular overview. Still, in Warrior Nô the only connection that I could see to the community’s future is in the nature of an awful warning about what is to be avoided.

The Warrior Nô therefore induces an affect quite distinct from Deity Nô (though these affect  are symmetrically obverse and quite compatible). Or it might be better to see the Warrior-Nô revelation as working through a number of passionate affects, such as the affection for another person–kin, lover, master–or the striving for fame, but finally leaving the spectator with two main, complementary affects: first, of the sweetly sad beauty of that special impermanence of passions mentioned above, well-known in the Japanese poetic tradition and retrospectively named mono no aware, the sense of the fugacity or tears of things (cf. Morris 208-09); and second, of the final peace of nirvana.10 The latter is not always present (e.g. not in Kanehira or Yashima), but when it is present it is usually achieved in the last verses and seconds on stage, after the main dance. Furthermore, while the stage agent may have freed him/herself of passion, this achievement can only be transmitted to the audience by way of a sympathy affect, complementary to yet also sustained by the clash of desire and fugacity. The shite-role may (in some plays, not always) by his confession and the waki-role’s sympathetic intercession have been purged of “wrongful clinging” to earthly ambitions and passions (môshû), but the audience must affectively cling to the shite’s story.

For both these weighty reasons–the brief, indeed sometimes omitted, instant of mostly verbal affirmation that the shite-role’s passions have been extinguished, and the affect-laden transmission of all stage propositions–, the shite-role’s supposed release into nirvana does not seem to me so important for the spectator as verbally inclined commentators looking for official religious doctrine would have us believe. Zeami’s affect-laden refunctioning of nirvana seems incompatible with the Buddhist a-ranâ, dwelling in Peace, which combines conflictlessness with full absence of passion (cf. Diamond Sutra 9 in Conze ed., 44-45). As Professor Ruch convincingly establishes, the popular imagination and art of medieval Japan freely chose among elements of official Buddhist doctrine (embracing, e.g., karmic causality) and then even more freely or fancifully alloyed them with aspects of other “available traditions and doctrines” so as to make them “most emotionally and aesthetically satisfying to the need of the moment” (“Coping” 100 and passim). In particular, “[the Nô] plays…are not religious rites or treatises, but art; and their goal is not primarily to set forth ideas….An accurate account of Nô should convey its lack of [doctrinal] rigor.” (Tyler, “Path” 170, and see Gundert 5). This is why there can be an inherent and, as far as I can see, unresolvable contradiction between the peace of nirvana as value-horizon striven for and explicitly invoked, and nirvana as affect to be induced in a spectator who balances it with the aching beauty of the warrior’s doomed passions. The dead warriors move fluidly across the supposedly strictly separated realms of transmigration, in every play they are found “clinging to the [earthly] place of their greatest love, deepest hate, or greatest pain” (Ruch, “Coping” 102), usually near their grave. This is why I have in Table 2 called nirvanic yearning a “subordinate” affect. The Nô plays mingle “accuracy of canonical detail” with a “disregard for overall paradigm integrity [concerning the Buddhist afterlife, DS]” in a doctrinally incoherent “illogical eclecticism of major scope” (Ruch, “Coping” 103 and 129): which was yet, I would maintain for the superior instances at least, affectively coherent.

1.4. Finally, and perhaps for a comparative dramaturgy most interestingly, the last two parameters of Table 2 deal with opposition and conflict (cf. my “Revelation vs. Conflict”). In this case, Zeami’s twofold adaptations in the Warrior Nô template are not complementary but divergent. Deity Nô had no trace of conflict of individual wills. It simply told, by stage means, a story (katari, the principle of classical Japanese theatre –cf. Raz, “Japanese”) enacting the wakirole’s– and the audience’s– advance from ignorance to knowledge, or from blindness to enlightenment: a trajectory characteristic for epochs of practically absolute, and thus theoretically–or discursively–unchallengeable, norms such as partly the European and even more so the Japanese Middle Ages (in sua voluntade è nostra pace, “in His will is our peace,” is the way Dante formulated it). In Warrior Nô it was impossible to fully evacuate individual subjects and their oppositions or confrontations. First, there is the brute fact of death in war, as a rule recapitulated in the nochijite’s final dance; often, however, the opponent is either not important, simply an agent of destiny (as Rokuyata in Tadanori), or non-existent in any individualist sense, as in the suicide of the eponymous protagonist of Yorimasa, so that this as it were background conflict seems to me quite subordinate. Second, there may be a “conflict” (or better opposition) within the protagonist, which is in some cases also double: the tension between two passions (poetry and martial prowess in Tadanori), and the overall or umbrella opposition between any earthly desire and the fleetingness or impermanence (lack of stay, hakanasa), which is in Buddhism foregrounded as unavoidable suffering and in this Nô category evidently demonstrated. This is a full inversion of the happy enlightenment from the Shintoist Deity Nô into the mono-no-aware enlightenment, further tending toward a more complex, bittersweet but equally blessed, nirvana.

Notes

*/ My thanks go to the Saison Foundation for support in Spring 1992; to the Japan Foundation for an earlier Short-term Fellowship; and to the SSHRC of Canada for a two-year Research Grant–all for work on Japanese theatre. For introduction to viewing Nô I am grateful to Günter Zobel, Yamada Kazuko, and Elsa Hatsumi Tsuzuki, and for useful indications of critical material to Frank Hoff, Fredric Jameson, Karatani Kôjin, Shelley Fenno Quinn, and Jacob Raz. George Szanto, Rich Gardner, and David Loy raised potentially fruitful critical objections to earlier drafts. None of them is even remotely responsible for my views. Translations from texts identified as not being in English are mine. Japanese names are adduced with family name first.

1/ I assume that the terms of shite (= doer, central and dominant player, protagonist) and waki (= sideman, introductory player) are by now so domesticated that I do not need to put them into italics. However, analytic and indeed ideological ambiguities may appear when the shite, e.g. Kanze Hideo, is confused with his role, e.g. the Takasago deity (this shite-role I sometimes faute de mieux call “protagonist,” though there is no antagonist and strictly speaking no agon in many Nô); the same holds for “waki,” used in criticism both in its proper meaning and as ellipse for “waki-role.” I have therefore reluctantly decided that the clumsy addition of “-role” may in all ambiguous cases be unavoidable semantic hygiene. Other terms are italicized and explained in the text, but I might repeat the main ones: maejite, nochijite = in two-part Nô, where the shiteplays different roles, name for shite in the first and second part; genzai nô = Nô where the shite plays a living human in the (supposedly) empirical time of the stage story; mugen nô = Nô where the shite plays “non-empirical” entities (deities, specters), sometimes appearing in the wakirole’s dream.
It might be added that the original implication when the term of mugen nô was coined, namely that it always presents an empirically non-existing dream or vision, is not acceptable any more. It was always semiotically tricky to enforce different epistemological statuses on actors equally present on the stage. But even without invoking semiotics, it would be absurd to exclude from mugen nô one of its mainstays, Deity Nô, where the divinities appear in real spacetime (Kanai Kiyomitsu observed this apropos of Takasago in 1969, as cited in Gardner 33; and in fact Konishi seems to follow that line, cf. History 525).

2/ Konishi, “Approaches” 4 (based on Nose Asaji) and 14-15 (based on costume changes for the waki of Yoroboshi) and cf. 8ff.; had his splendid overview from 1960 had the impact it deserved on criticism in European languages, my Section 0 could have been different and briefer.

3/ O’Neill, Early Nô 101 and 94, also Konishi, “Approaches” 4-5; and see Goff 3, Keene, Nô 41 and passim. Sieffert follows Tanaka Makoto’s “Yôkyoku no haikyoku” (“Uncanonical Nô Texts,” in Nogami Toyoichirô ed., Nôgaku Zensho, Tôkyô 1942, 4: 133-35) by speaking–as do Konishi and O’Neill–of “nearly 3,000 Nô libretti,” 12; the discrepancies are largely due to the yet imperfect bibliographic control of a list of some 3,000 titles, which comprises in some cases same title for different plays and in other cases same play under different titles.
For quicker orientation, I recall the standard historical periods referred to: Muromachi–14th to 16th Century; Edo–17th to 19th Century; Meiji–1866-1912.

4/ Sieffert allots Zeami 48 plays in the Deity, Warrior, and Final categories and 59 plays in the other two categories. O’Neill, another impenitent “attributor,” gives him from among plays “thought to be directly connected with one in the modern repertoires” 49 in the Deity, Warrior, and Final categories and 76 in the other two categories (Early Nô 161 and 165-71). On the otherpole, Hare (42-47; he follows Omote Akira) confines himself to 46 plays in the present canon  plus 15 “of uncertain attribution,” which leads him to allot four fifths of Zeami’s plays to mugen nô (236)…

5/ The complexities of the Woman-Nô category as constituted in the Edo-period, which also contains some plays that are possibly derivative from the revelation model but do not precisely follow either the conflictual or the revelatory pattern (e.g. Ohara Gokô, Ohmu Komachi or Yôkihi), are such that a separate investigation is needed.

6/ In the vexed and shifting matter of attributions I follow Dômoto Masaki; cf. also Shimazaki 73. In view of my comparatist purpose, I am in this paper not touching upon Zeami’s own stage “micro-syntax” of division into dan (sections) and its build-up from blocks or subsections (called shôdan by Yokomichi), nor upon Zeami’s jo-ha-kyû theory of rhythm; cf. Hare for detailed treatment of both. Zeami’s distinctions based on what I would call social type (woman, oldster, warrior) would have to be considered carefully in a more detailed treatment.

7/ I have used both the Rimer/Yamazaki (157) and the Quinn (79) translations with minor
modifications. I am much indebted to their useful comments.

8/ A somewhat different sub-type is presented by Zeami’s three kachi-shura (Warriors’ Victory) Nô, Ebira, Tamura, and Yashima, whose shite-role is a victorious warrior. This may modify the second four parameters in Table 2 (most clearly in the case of Tamura). Whatever these plays’ dates of composition, such Nô may therefore be typologically somewhat nearer to, or a transition from, Deity Nô to the “pure” make-shura (Warriors’ Defeat) Nô. At any rate, it seems most significant that the consensually most memorable plays and protagonists are the “defeat” ones, and that even the “Victory Asura” plays leave the victors in the Warriors’ Hell (ashuradô, see next note). Obversely, and I think characteristically for Zeami’s heterodox position, these “Victory Asura” plays were the most frequently performed ones for the Tokugawa shoguns (see Yokota 366). Though the kachi-shura are not analyzed here, it may therefore be seen how they could be accommodated by the above Table. For other possible divisions of Warrior Nô based on position between the poles of dance-cum-song and narration, on the social rank of shite-role, etc., see Shimazaki 185-87.

9/ Doctrinally, in Japanese Buddhism this ashuradô is not a hell but one of the six “realms of transmigration” or states of sentient existence (rokudô) through which people are cyclically reborn according to their karmic deserts until released to a paradise and ultimately into nirvana (see Ruch, “Coping” 93ff.). However, it is a place and/or state of the dead warriors’ perpetual, bloody torture in frenzied battle, befitting their passion, and usually translated as “hell” in Nô renditions (see e.g. Shimazaki 64, 84, 92, and 136). Professor Ruch stresses a significant difference between officially “commissioned” and “noncommissioned” works, e.g. “religious/ didactic writ” vs. poetry and most prose in Japanese Middle Ages (“Coping” 106 and passim).
Nô seems in this usage too to show its hybrid nature, arising out of a contradictory fusion of “low” with “high” cultural and ideological elements (cf. Ruch, “Medieval”).

10/ This matter of affects is crucial for any discussion of Nô as plays composed to have effects for given audiences. It is also the major unresolved crux of present-day theatre theory. Cf. some interesting hints on the “affect pattern” of mono no aware in Hijiya-Kirschnereit 214.

Works cited

Barthes, Roland. Image–Music–Text. Sel. and tr. S. Heath. New York: Hill & Wang, 1977.

Berger, Harry, Jr. “Bodies and Texts.” Representations no. 17 (1987): 144-66.

Brazell, Karen. “Atsumori: The Ghost of a Warrior on Stage.” par rapport no. 5-6 (1982-83): 13- 23.

—. “On the Nature of Noh,” in eadem ed., Twelve Plays of the Noh and Kyôgen Theaters.

Cornell East Asia Papers 50. Ithaca: Cornell U East Asia Program, 1988, 205-31.

Brecht, Bertolt. Gesammelte Werke. Frankfurt: Suhrkamp, 1973.

Conze, Edward, ed. Buddhist Wisdom Books. [London: Arnold, 1958].

Dômoto, Masaki. “Bangai-kyoku kenkyû josetsu” [Introduction to the Study of the Noncanonical Pieces of Nô]. Bull. of Geinô-shi kenkyû-kai no. 94 (1986): 1-14.

—. Interview with D. Suvin, Kamakura 2/6/1993.

Eco, Umberto. “Dizionario versus enciclopedia,” in his Semiotica e filosofia del linguaggio. Torino: Einaudi, 1984, 55-140.

Freud, Sigmund. Die Traumdeutung. Leipzig: [n.p.], 1909.

Gardner, Richard. “Takasago: The Symbolism of the Pine.” Monumenta Nipponica 47.2 (1992): 203-40.

Goff, Janet. Noh Drama and “The Tale of Genji.” Princeton: Princeton UP, 1991.

Gundert, Wilhelm. Der Schintoismus im japanischen Nô-Drama. Mitteilungen DOAG XIX. New York: Johnson Reprint, [1965]; orig. 1925.

Hare, Thomas Blenman. Zeami’s Style. Stanford: Stanford UP, 1986.

Hawkes, Terence. That Shakespeherian Rag. London: Methuen, 1986.

Hegel, Georg Wilhelm Friedrich. Phänomenologie des Geistes. Sämtliche Werke, Vol. 2. Ed. J. Hoffmeister. Leipzig: Meiner, 1949.

Hijiya-Kirschnereit, Irmela. “The Concepts of Tradition in Modern Japanese Literature,” in P.G. O’Neill ed., Tradition and Modern Japan. Tenterden: Norbury Publ., 1981, 206-16 and 301-02.

Hobsbawm, Eric. “Introduction: Inventing Traditions,” in idem and Terence Ranger eds., The Invention of Tradition. Cambridge: Cambridge UP, 1983, 1-14.

Hoff, Frank, and Willi Flindt. “The Life Structure of Noh.” Concerned Theatre Japan 2.3-4 (1973): 209-56.

Honda, Yasuji. “Yamabushi kagura and bangaku.” Tr. F. Hoff. Educational Theatre J. 26.2 (1974): 192-208.

Jameson, Fredric R. The Ideologies of Theory, Vol. 1. Minneapolis: U of Minnesota P, 1988.

Keene, Donald. Dawn to the West, Vol. 2. New York: Holt, 1987.

—. Nô and Bunraku. New York: Columbia UP, 1990.

Konishi, Jin’ichi. A History of Japanese Literature, Vol. 3. Tr. A. Gatten and M. Harbison, ed. E. Miner. Princeton: Princeton UP, 1991.

—. “New Approaches to the Study of the Nô Drama.” Tôkyô Kyôiku Daigaku Bungaku-bu Kiyô 27, Kokubungaku Kambungaku ronsô 5 (1960): 1-31.

Loy, David. Nonduality. New Haven: Yale UP, 1988.

Matisoff, Susan. “Holy Horrors,” in James H. Sanford et al. eds., Flowing Traces. Princeton: Princeton UP, 1992, 234-61.

Morris, Ivan. The World of the Shining Prince. Harmondsworth: Penguin, 1985.

O’Neill, P.G. Early Nô Drama. Westport CT: Greenwood P, 1974.

—. “The Social and Economic Background of Nô.” Maske und Kothurn 27 (1981): 19-29.

Quinn, Shelley Fenno. “How to Write a Noh Play: Zeami’s Sandô.” Monumenta Nipponica 48.1 (1993): 53-88

Raz, Jacob. Audience and Actors. Leiden: Brill, 1983.

—. “The Japanese Itinerant Storyteller,” in his Aspects of Otherness in Japanese Culture. Tôkyô: ILCAA, 1992, 35-62.

Ricoeur, Paul. Hermeneutics and the Human Sciences. Ed. John B. Thompson. Cambridge: Cambridge UP, 1981.

Rimer, J. Thomas. “The Background of Zeami’s Treatises,” in [Zeami,] On the Art of the Nô Drama. Tr. idem and Yamazaki M. Princeton: Princeton UP, 1984, xvii-xxviii

Ruch, Barbara. “Coping With Death,” in James H. Sanford et al., eds., Flowing Traces.
Princeton: Princeton UP, 1992, 93-130.

—. “Medieval Jongleurs and the Making of a National Literature,” in John W. Hall and Toyoda Takeshi eds., Japan in the Muromachi Age. Berkeley: U of California P, 1977, 279-309.

Shimazaki, Chifumi. Warrior Ghost Plays from the Japanese Noh Theater. Ithaca: East Asia Program, Cornell U, 1993.

Sieffert, René. “Introduction” to idem ed., Nô et kyôgen: Printemps–Été. Paris: POF, 1979.

Suvin, Darko. “A Modest Proposal for the Semi-Demi Deconstruction of (Shakespeare as) Cultural Construction,” in Loretta Innocenti et al. eds., Semeia: Itinerari per Marcello Pagnini. Bologna: Il Mulino, 1994, 67-76.

—. “Revelation vs. Conflict: A Lesson from Nô Plays for a Comparative Dramaturgy.” Theatre J. 46.4 (1994): 523-38.

—. “The Soul and the Sense: Meditations on Roland Barthes on Japan.” Canadian R. of
Comparative Lit. 18.4 (1991): 499-531.

—. “The Use-Value of Dying: Magical vs. Cognitive Utopian Desire in the ‘Learning Plays’ of Pseudo-Zenchiku, Waley, and Brecht.” Brock R. 3.2 (1994): 95-126 [somewhat different Japanese version in Hihyô kûkan (Critical Space) no. 10 (1993): 110-33].

Taylor, Charles. Hegel. New York: Cambridge UP, 1977.

Terasaki, Etsuko. “A Study of Genzai Plays in Nô Drama.” Diss. Columbia U, 1969. Ann Arbor: University Microfilms, 1970.

Tyler, Royall. “‘The Path of My Mountain’: Buddhism in Nô,” in James H. Sanford et al., eds., Flowing Traces. Princeton: Princeton UP, 1992, 149-79.

Wang, C.H. “The Lord Impersonator: Kung-shih and the First Stage of Chinese Drama,” in Ying-hsiung Chou ed., The Chinese Text. Hong Kong: The Chinese UP, 1986, 1-14.

Wells, Henry W. The Classical Drama of the Orient. London: Asia Publ. House, 1965.

Whiteside, Anna. “Theories of Reference,” in eadem and Michael Issacharoff eds., On Referring in Literature. Bloomington: Indiana UP, 1987, 175-204.