LITERARY UNIVERSALS WEBLOG: A series of informal observations and conjectures aimed at fostering more reflection on and discussion about cross-cultural patterns in literature.
Patrick Colm Hogan, University of Connecticut
At the end of May, I had the great pleasure of remotely attending Nigel Fabb’s valedictory address on the topic of endings, given on the occasion of his retirement from Strathclyde University. Nigel’s talk was characteristically erudite, witty, and insightful. Moreover, like so much of Nigel’s work, it introduced topics and approaches that are virtually certain to provoke further reflection, study, and theorization by Nigel himself and by others. Indeed, after listening to Nigel’s talk, I myself began thinking about just what features of endings might be candidates for universality.
Nigel stressed that isolating the “ending” is not simply a perceptual experience. It involves interpretation. In saying this, Nigel was extending the approach to endings that he developed in his book, Language and Literary Structure. A central idea of that book is that many kinds of form are our interpretation of what a text tells us, in context, rather than being observer-independent facts about the text. Endings are, by this account, one such kind.
A somewhat different, but complementary approach, would begin the interpretative isolation of endings by noting a key ambiguity in the term, “ending.” The basic reason why we need interpretation to isolate the ending is that, as narratologists would say, the story ending and the discourse ending are not necessarily the same. Put simply, a story is the sequence of events dramatized in a play or recounted in a narrative. Discourse is the manner in which those events are represented. For example, the discourse might represent story events in non-chronological order. Thus, a romantic comedy might begin at a wedding. As the bride and groom exchange rings and kiss, the unshaven best man turns to the somewhat rumpled maid of honor and whispers discreetly, “Well, that turned out rather better than we might have anticipated.” The film cuts to a clock, which stops its forward progress, then runs backward to eight hours earlier. It is the middle of the night, and the pajama-clad groom is having a ill-tempered exchange with the best man over why anyone would pack a wedding ring in a suitcase, when the item fits easily in one’s pocket—and especially when he knew that the ring was a precious memento of the bride’s grandmother and that she would probably call off the wedding if she thought the prospective groom had been so careless as to entrust that memento to such an untrustworthy guardian. The film cuts to the clock again, now running backward four hours and fifteen minutes, to when the best man is checking his luggage at the airport. We then follow the suitcase as it travels to some exotic destination that happens to have the same name as the regional airport in rural Wisconsin where the wedding was to take place. The rest of the film informs us of how the wedding ring was recovered, and the entire fiasco was concealed from the bride, due in large measure to the quick thinking of the maid of honor. The film ends with a montage of the newlyweds, driving off toward a sign that reads “Niagara Falls, Honeymoon Capital of the World, 836 Miles,” strings of clattering tin cans tied to the bumper, and the best man inviting the maid of honor for a cup of coffee, under a huge sign that says only “CHEESE.” (It’s a Wisconsin thing. Trust me.) In this fanciful scenario, the end of the story is presumably the wedding. But that end of the story occurred at the beginning of the narrative. The discourse ending comprises the departure for Niagara Falls and the possible beginning of a romance between the best man and the maid of honor.
Nigel empirically isolates some recurrent endings from which we might draw theoretical conclusions. Here, I will begin instead with a few theoretical conjectures to orient our look at some actual endings. The first thing we need to do is to characterize story endings more clearly. (A discourse ending is simply the last bit of a work—play, novel, poem, or whatever.) Broadly in keeping with Nigel’s view, I don’t believe that there are any absolute rules as to what constitutes a story ending. But one clear option relies on the common view that a story is preceded and followed by normalcy, which is to say, routines of everyday life. Something happens that disturbs this normalcy—for example, two people fall in love or an enemy invades one’s country. This sets in motion a series of further events until the disturbance is “resolved,” which is to say, things return to normalcy, to generally predictable routine. In a romantic story, the marriage of the lovers is a case of this sort. The subsequent honeymoon is to a degree ambiguous. It is not ordinary in the sense that it does not occur every day. But, within the (partially fictional) culture tacitly assumed in the preceding example, it is predictable. Moreover, it is what people in the culture would assume to follow the wedding (leaving aside some anachronisms, such as the tin cans, added in the hope of producing some comic effect). But, given the preceding problems, the wedding itself was uncertain. Thus, throughout the film, we presume that the usual sequences of events—including the honeymoon–will follow only because the (endangered) wedding has occurred. On the other hand, the hints of a budding romance between the best man and the maid of honor are not at all a matter of resolving the obstacles to the romantic union of the lovers in the main story. So, it is unequivocally a discourse ending.
Once we have distinguished story endings from discourse endings, we might reasonably expect that the two can either coincide or diverge. Coinciding is presumably the default case. Thus, one might anticipate that at least some narratives and dramas, cross-culturally, will conclude the discourse with the ending of the story. For example, we find the 18th-Century, Japanese playwright, Chikamatsu, ending his romantic tragedy, The Love Suicides at Sonezaki, in roughly this way. The lovers, weeping, first consider whether or not they will meet their parents “in the world of the dead” (56). The woman (Ohatsu) then demands that her lover (Tokubei) stab her. He does so, calling out, “Namu Amida” (56), thereby invoking the Pure Land Buddhist belief in the spiritual benefits of invoking the compassionate Buddha, Amitābha. When someone calls his name with sincerity, Pure Land Buddhists believe that Amitābha will aid that person’s enlightenment—thus, his or her freedom from worldly attachment—initially through rebirth in the “Pure Land” (see the entries for “pure land,” “Amitābha,” and “buddhakṣetra” in Buswell, et al.). Having stabbed Ohatsu, Tokubei quickly turns his dagger against himself. After they have both died, the narrator informs us only that “No one is there to tell the tale,” but even so people will “gather to pray for these lovers who beyond a doubt will . . . attain Buddhahood” (56).
This ending is almost entirely a simple representation of the suicide, thus of the final moment of the story. It does, however, contain two elements that suggest how divergent (i.e., non-coinciding) discourse endings might operate generally. In other words, they hint at the functions that might be served by separating the discourse ending from the story ending, or more simply why an author might do such a thing. I take it that the references to the bonds of the lovers with their parents serve to stress the centrality of attachment feelings (as distinct from mere lust) to the lovers’ emotional lives. This in turn makes their immanent attachment loss more salient and their condition all the more pitiable. (Most readers are, presumably, more sympathetic with people suffering attachment loss than with people who are merely sexually frustrated.) It also puts them in the position of children, in some degree heightening our sense of their vulnerability. In these ways, the references to parents intensify the audience members’ empathic response—though only short term. The reference to their eventual Buddhahood gives the viewer or reader hope—indeed, virtual certainty—that this deeply tragic moment will serve as the means to end suffering permanently. Like the reference to parents, this part of the discourse ending serves to modulate our emotion. But in this case, it limits the lovers’ suffering, thereby mitigating our empathic pain, providing us with a more distant, but also more enduring peace. We might in addition take these points to suggest that Chikamatsu is addressing Buddhist themes bearing on desire and suffering.
Thus, The Love Suicides at Sonezaki presents us with two possible functions of discourse endings: emotion modulation (either intensification or mitigation) and thematic assertion. Neither of these is unexpected. I have argued elsewhere (see, for example, Narrative Discourse 17-18 and Affective Narratology 103-104), following many previous writers, that literary narratives have two principal purposes: to produce emotion (in the first place, empathic emotion aimed at characters in the story) and to cultivate attitudes toward and ideas about the real world (thus, “themes” as I am using the term). As these are the two primary goals of literature, it is predictable that the primary goals of discourse endings would be the same.
Of course, narrative works pursue these functions through stories. Thus, we might also expect some discourse endings to bear on stories as such. Upon reflection, we may recognize something of this sort in The Love Suicides at Sonezaki. Specifically, the reference to Buddhahood focuses on the fate of the central characters, but it does so for events that occur subsequent to the story represented in this work, events that take place after the story ending. Other works may focus on the fate of “supporting,” or even peripheral characters. Of course, as already noted, some works give us the story ending early on, then conclude with something from the middle or the beginning of that story. Even so, all these story-based discourse endings (as we might call them) ultimately serve emotional or thematic purposes. For instance, in the fictional case presented earlier, the end of the story is actually the first event presented in the film. In cases such as this, the main emotion elicited by the narrative is often mirth, which might be inhibited if the viewer is worried about whether or not things will work out in the end. In Meir Sternberg’s terms, the change in discourse order, so that we know the story outcome from the start, shifts part of our emotional response from suspense to curiosity.
We find particularly illuminating cases of discourse endings in the two versions of Ji Junxiang’s The Zhao Orphan, which appeared in the Yuan and Ming dynasties. The Yuan version appeared during the Mongol rule of China and the play may be read as a call to rebel against this “foreign” domination (see “Paradoxes of Literary Emotion”). The play ends with the orphan’s rage against the would-be usurper, Tu Angu. The orphan contemplates how he will “slit open [Tu Angu’s] belly/Lop off his limbs” (67), and so on. This version is actually different from anything I have mentioned thus far. The discourse ending represents a point in the middle of the story, but the story ending is nowhere to be seen. Indeed, the story is unresolved. I take it that this has both an emotive and a thematic purpose, which become clear in the interpretation of the play as an indirect expression of anti-Yuan, Song loyalism. (The Song dynasty was the Han Chinese dynasty that preceded the Yuan dynasty.) Specifically, in that context, the discourse ending appears aimed at intensifying audience anger at the then-current rulers (i.e., the Yuan dynasty emperors). It implicitly characterize those rulers (thematically) as invaders and usurpers, thus as real-world targets for the sort of rage represented in the character of the Zhao orphan.
The ending of the Ming dynasty version is quite different. As Li points out, this version “tones down or deletes criticism of the . . . ruler and removes any suggestion that the Zhao Orphan . . . could have become a usurper” (19)—or, more properly, a rebel against the current, real-world ruler (who was tacitly characterized as a usurper in the Yuan period version of the play). In the Ming version, the orphan, Cheng Bo, receives the monarch’s permission to arrest Tu Angu secretly, after which the latter’s entire clan is killed and Tu Angu is tortured to death. There is clearly no thematic encouragement of rebellion in this version, unsurprisingly as the foreign, putatively usurping Yuan dynasty has ended. More precisely, the events of the play are presumably taking place in the pre-imperial period (to judge from the dates that Chinese writers have given to the historical sources; see Wai-yee Li’s “Introduction” to the play [17]). Nonetheless, they almost certainly serve to foster terror of the punishments for rebellion against the then-ruling Ming emperor, due to his implicit connection with the ruler in the play.
A possible problem here is that the brutal treatment of Tu Angu and the murder/suicide of the lovers will almost certainly lead some audience members to condemn these characters. Moreover, readers familiar with the Analects will recognize that terrorizing the public, as developed in the play, is far from the tiāndào or “way of heaven” envisioned by Confucius as the proper model for monarchs (see, for example, Analects 2.3 [available here; accessed 11 June 2024]). In part to modulate this very un-Confucian theme, the entire revenge is characterized as following the “way of heaven” (54). This technique of spiritual elevation applies directly to Cheng Bo, though indirectly to the ruler in the play and, thereby, to the Ming ruler also. (I refer to a feature of an ending as a technique if it is repeatedly taken up to specify and implement emotive and/or thematic functions, but is not a literary desideratum in and of itself.) We saw this technique previously in the Chikamatsu play, when the narrator asserts the lovers’ eventual Buddhahood. In both cases, this spiritual elevation serves in part to further secure the audience’s sympathy for the main characters—thus, Cheng Bo in the Ming Zhao Orphan and the lovers in The Love Suicides. In keeping with this function, the discourse ending stresses Cheng Bo’s filial obligations to his murdered parents and to his foster father. It then touches on the fate of the latter, before turning to an account of how various minor characters will be treated–how they will be memorialized if they died to save the Zhao orphan, and so on. This is a version of the discourse ending that recounts the (non-story) fates of the various characters.
Another, somewhat different instance of emotive and thematic spiritualization may be found in the discourse ending of Kalidasa’s 4th-century, Sanskrit drama, The Recognition of Shakúntala. This is a romantic comedy in which the lovers have been separated due to a curse and have remained separated due to a misunderstanding. Specifically, neither King Dushyánta nor Shakúntala knows about the curse, so neither understands why Dushyánta failed to recognize Shakúntala after their initial separation. The discourse ending includes the reunion of the couple—thus, the story ending. As part of this story-resolving conclusion, both lovers learn about the curse, which exonerates Dushyánta for his earlier rejection of Shakúntala, thereby reconciling the two. But this is not only a reunion of the lovers. It is a family reunion that highlights the affection Dushyánta feels for his young son, an affection that foregrounds the attachment component of the love that motivates the action of the play, much as the references to parents did in the discourse ending of Chikamatsu’s play. This is also connected with a general, spiritual elevation of this family, signaled most clearly by the fact that the concluding events take place in a heavenly hermitage, and by the final words of the play. Specifically, the concluding verses link the king—Dushyánta, and perhaps an actual monarch at the time of the play’s performance)–with heavenly beings. Interestingly, the king himself goes on to pray that he will be devoted to his subjects and will ultimately achieve spiritual enlightenment. This spiritualization has the usual thematic and emotive functions, with the addition of making Dushyánta more sympathetic through humility. (Thanks to Lalita Hogan for help with some points in the Sanskrit original.)
Two other points are worth making about this ending. First, the familial reunion brings the couple’s son, Bhárata, on stage and thereby facilitates the treatment of his eventual fate. Indeed, the play also recurs to Shakúntala’s adoptive father, allowing us to be reassured about his understanding of the situation and his continuing affection for Shakúntala. (It would be easy to worry that he would reject her for apparent promiscuity in becoming pregnant without any recognized marital bond.) Finally, the last act of the play is clearly parallel with the opening act and the invocation of Shiva at the end of the act parallels the invocation of Shiva that opened the play. Thus, the discourse ending closely recalls the discourse beginning of the work. This familiar technique of plot circularity often fosters a feeling of aesthetic pleasure through the reader’s recognition of an unanticipated pattern. (On aesthetic pleasure and unanticipated pattern recognition, see the first chapter of Beauty and Sublimity.) It may also intensify or otherwise modulate our emotional response by recalling for us the initial feelings of the lovers. In the case of this particular work, the circularity may be taken to have thematic implications as well, for the play’s treatment of forgetting suggests theories of the operation of memory traces in relation to spiritual practices (see Affective Narratology, 165-181).
The functions (thematic and emotive) and techniques (e.g., formal circularity) we have considered in the Japanese, Chinese, and Indian cases are, I believe, all familiar from European works as well. It is worth briefly mentioning a case from one further region—Nizami’s 12th-century Persian poem, The Story of Layla and Majnun. The poem has a total of 54 cantos. The story ending is spread over three of the last four. In canto 50, Layla dies. In cantos 51 and 52, Majnun grieves and dies. Additionally, in these two cantos, Nizami provides further indications that the unconsummated love between Layla and Majnun may be read as an allegory for the Sufi devotee’s love of God, which cannot be realized in a full, mystical union while the former lives in the mundane world. The 53rd canto sees the lovers buried together, their physical proximity suggesting their ultimate union after resurrection and judgment. This would seem to conclude the story proper. The final canto comprises a vision experienced by the old man who served as a go-between for the lovers. The vision is in part an instance of reporting what happens to the main characters after the story resolution. As in Chikamatsu’s play, however, this epilogue is also a spiritual elevation, for Layla and Majnun are transfigured in the divine Garden—itself a circular return, not to the beginning of the poem, but to the beginning of time itself or creation, though now intensified as “eternal” (thus, with no risk of a Fall). The poet goes on to explain that the moral virtues of the lovers, leading them to “right action” and “compassion,” are the outcome of love (176; I am grateful to Yass Alizadeh for her insights into the linguistic resonances of the original Persian). The intertwining of emotional and thematic purposes could hardly be clearer.
I hope Nigel will forgive me for taking up his fascinating analyses and moving them in a different direction. I say “different,” but that is not at all to say “contradictory.” First, there is certainly overlap between my treatment of endings and Nigel’s. More significantly, I hope Nigel will agree that the parts that do not overlap are complementary. Indeed, I hope he will consider a blog post here on his treatment of endings—which, I must stress, inspired my own, both the parts that overlap with his and the parts that do not.
As to future directions, I have sought to distinguish a range of functions and techniques of discourse endings. They may treat story materials by changing their order of presentation or by extending the tale to the post-story fate of the main characters or the non-story fate of ancillary characters. However, when they do so, the function of this added material is almost always one of emotion modulation or thematic development, for these are the usual desiderata of storytelling. There are different techniques an author may employ to achieve these ends. For example, he or she may associate the protagonists with attachment feelings (e.g., parents or children), thereby enhancing our empathy toward those characters. He or she might also seek to link the protagonists with spirituality, in effect disarming criticism of their (often unorthodox) behavior. Such elevation may be extended beyond the fiction to political figures in the author’s or audience’s world—whether powerful figures (as in the Ming Zhao Orphan) or unjustly disempowered figures (as in the Yuan Zhao Orphan). In some cases, the technique may be formal, as with circularity; then, its emotional effects are often a matter of aesthetic pleasure.
Though I have considered only a handful of works, it seems likely that the same broad principles will be found to recur in a range of cases in the written literatures of East Asia, South Asia, the Middle East, and North Africa, as well as Europe and the Europhone Americas. No doubt, the list of techniques could be expanded considerably. But they all probably serve the same broad emotive and thematic functions. What seems more potentially significant, however, is the study of pre-colonial (or indigenous) orature—African, Australian, North and South American, and so on. Indeed, orature is generally much underrepresented in the study of universals, so that it becomes difficult to tell if one has isolated an unrestricted literary pattern, or a pattern that appears only after the development of writing. In any case, Nigel Fabb’s recent work on endings—and, I hope, the preceding reflections on that topic—point toward yet another potentially fruitful area of research in literary universals.
[See also Nigel Fabb’s “Blog for July 2024: Endings are Created by Interpretations.”]
Works Cited
Buswell, Robert, et al. The Princeton Dictionary of Buddhism. Princeton UP, 2014.
Chikamatsu. Major Plays of Chikamatsu. Trans., Donald Keene. New York: Columbia UP, 1990.
Fabb, Nigel. Language and Literary Structure: The Linguistic Analysis of Form in Verse and Narrative. Cambridge: Cambridge UP, 2002.
Fabb, Nigel. “Why Endings Are Better Than Beginnings.” Strathclyde University, Glasgow, UK. 29 May 2024.
Hogan, Patrick Colm. Affective Narratology: The Emotional Structure of Stories. Lincoln, NE: U of Nebraska P, 2011.
Hogan, Patrick Colm. Beauty and Sublimity: A Cognitive Aesthetics of Literature and the Arts. Cambridge: Cambridge UP, 2016.
Hogan, Patrick Colm. Narrative Discourse: Authors and Narrators in Literature, Film, and Art. Columbus, OH: Ohio State UP, 2013.
Hogan, Patrick Colm. “Paradoxes of Literary Emotion: Simulation and The Zhào Orphan.” In The Routledge Companion to Literature and Emotion. Ed., Patrick Colm Hogan, Bradley J. Irish, and Lalita Pandit Hogan. New York: Routledge, 2022, 134-143.
Ji Junxiang. The Zhao Orphan [Ming version]. Trans., Pi-twan Huang and Wai-yee Li. In The Columbia Anthology of Yuan Drama. Ed., C. T. Hsia, Wai-yee Li, and George Kao. New York: Columbia UP, 2014, 17-54.
Ji Junxiang. The Zhao Orphan [Yuan version]. Tran., Wai-yee Li. In The Columbia Anthology of Yuan Drama. Ed., C. T. Hsia, Wai-yee Li, and George Kao. New York: Columbia UP, 2014, 55-72.
Kalidasa. The Recognition of Shakúntala. Ed. and trans., Somadeva Vasudeva. New York: New York UP, 2006.
Nizami. The Story of Layla and Majnun. Ed. and trans., Rudolf Gelpke. English version in collaboration with E. Mattin and G. Hill. Final chapter trans., Zia Inayat Khan and Omid Safi. New Lebanon, NY: Omega Publications, 1997.
Sternberg, Meir. Expositional Modes and Temporal Ordering in Fiction. Bloomington, IN: Indiana UP, 1993.