Yoshikai Naoto
Translated by Vyjayanthi Selinger
Introduction
No one would dispute the importance of the topos of kaimami (lit. “seeing through a gap in the hedge/ fence”), or stolen glimpse, when studying Japanese literature from the Heian period (794–1191), particularly The Tale of Genji (Genji monogatari, ca. 1008).
After all, these stories usually require the male protagonist’s visual “encounter” with a sequestered aristocratic lady as a pivotal plot point to get the story going. Though most marriages for upper-class families were likely arranged by the parents, authors of the Heian tales often incorporated chance encounters, perhaps because people assumed arranged marriages would not set the couple’s heart aflutter (as in Lady Aoi’s sparkless marriage to the eponymous hero of the Genji).
Given the significance of kaimami, I conceived this trope as a scholarly term and theorized it through my readings of the Genji and published it as Kaimamiru Genji monogatari (Peering into The Tale of Genji).[1] In this book, I argue that if kaimami were easily achievable, the reader would not have found it riveting. In fact, the Genji purposely creates hurdles for a man to catch a furtive glimpse of a woman. His efforts to overcome these hurdles function as the beginning of courtship and a chance “encounter” is the sine qua non of courtly tales.
Whereas these aspects of kaimami are common knowledge for most of the Heian literature specialists, this does not hold true for others, including those who work in ancient or modern literature. I have noticed that not only are non-Heian specialists unfamiliar with how kaimami works in Heian tales, but they also tend to conflate “peeking” with illicit voyeurism. This was also the case for the foreign researchers and international students I have met in Japan; I feel that we are in need of forging a common understanding of kaimami across borders.
1. Towards an International Theory of Kaimami
One cannot speak of the theory of kaimami without mentioning the groundbreaking study by Imai Gen’e, “Kodai shōsetsu sōsaku-jō no ichi shuhō: Kaimami ni tsuite” (A method in constructing ancient novels: On the subject of kaimami), which reframed the mundane act of “peeking” as kaimami and sublimated it into a literary device for advancing a love story.[2] Given how highly Imai’s theory was regarded and how it shaped the reigning consensus in my field, I used to simply apply it to my own literary analyses without much reflection. However, as I continued my research, it became apparent that Imai’s theory was far from perfect and non-Heian specialists did not readily accept it.
First of all, Imai’s romanticization of kaimami is proven to be a hard sell, as I noticed the unusual aversion to this concept among contemporary readers of Heian tales, especially foreign students in Japan. Even though kaimami connotates elegance and allure in the cultural milieu of the Heian court, in the eyes of foreigners today, it often appears negatively with a criminal thrust. As a result, my foreign students tell me that kaimami is wrong and no one should be doing it.
The tendency for Westerners to stigmatize the act of peering seems to have influenced how kaimami scenes are translated. I hear that the word “peeping” is used to describe kaimami in some English translations of the Genji. Such rendering automatically invalidates the argument that kaimami is an important literary device. Though one might brush this off as a matter of cultural difference, I cannot bring myself to do so. But I don’t know how to convey to my international students that kaimami is a legitimate and elegant motif and it was devised to propel a love story forward. This is a critical hurdle we must overcome in the globalization of the Genji.
Nevertheless, foreigners are not the only people who would contest the idealization of unconsented peering. Even in Japan, the act of looking was at times unfavorably perceived. As I took a broad historical view, it became clear to me that the trope of “serendipitous glimpse” was a strictly Heian literary convention. In pre-Heian texts, seeing can be associated with taboos and folk customs that prohibit looking: by looking at a prohibited person or object, one could bring about misfortune. Similar ideas are found in foreign works such as Charles Perrault’s fairy tale Blue Beard, in which a wife enters a forbidden room and is confronted with the horrifying sight of the corpses of the previous wives.
Furthermore, in setsuwa (anecdotal literature) of medieval times, the act of looking is often endowed with the power to unmask the true identity of another. This motif, which is particular to the setsuwa genre, indicates that people in medieval Japan believed in the “magical force of the eye,” sharing similarity with the “forbidden sight taboo” of ancient texts. I have therefore come to understand kaimami in Heian texts as a rather unique development, with connotations quite divergent from the act of seeing in different historical or cultural context.
Of course, I am not declaring that the Heian construction of kaimami is a distinct thing unto itself with no connection to anything prior. Heian authors could have taken the sight taboo and turned it on its head, using it to drive the love story instead of inviting misfortune. If so, it is possible for the remnants of the taboo to hover in the background. I therefore do not deny the possibility that the prohibited and the forbidden could inform the act of kaimami. At the same time, it is also true that kaimami adopts a viewpoint distinct from a sight taboo, a difference that is central to its use as literary motif.
2. Kaimami as a Lexical Term, Kaimami as a Concept
A search for the textual mentions of the term kaimami revealed that, despite its well-worn use in academic circles, it appears fairly few times. As mentioned before, this word does not appear in pre-Heian literature, and is used for the first time in the The Tale of the Bamboo Cutter (Taketori monogatari, ca. ninth century). Even in the Genji, it only appears six times. Nevertheless, theories of kaimami are made possible by the fact that scholars do not limit themselves to the cases in which the narrator literally describes as kaimami. Instead, they fold in actions that are referred to as nozoku (peer into), ukagau (quietly observe), and miru (gaze) to construct (and fantasize) the concept of kaimami. Imai is no exception.
Put another way, scholars have conceived kaimami beyond its literal meaning, since they cannot limit themselves to the rare situations where characters peer through a gap in the hedge. Consequently, kaimami as a concept has come to include many shapes of looking, including ones that have little to do with a romance. Further, I have discovered that, unlike other literary devices, there is no locus classicus or prototypical example of kaimami. All of these findings suggest that the kaimami theory, as we know it, is built on a faulty premise; it casts a wide net to survey various forms of peeking, takes up only the examples that turn into courtship, and represents them as archetypes of kaimami. It seems that the theory emerged as the scholars’ “wishful fantasy” on what kaimami should be like, while any “noisy data” were excluded from the analysis.
Furthermore, it should be noted that beginnings of kaimami theory are traceable not to the Imai article, but rather to an older one written by Akeda Yonesaku, “Genji monogatari no kaimami” (Stolen glimpses in the Tale of Genji”).[3] From more than a dozen scenes of kaimami in the Genji, Akeda introduces, as representative examples, Yūgiri’s furtive peering at Tamakazura and Genji as well at Lady Murasaki in the Nowaki chapter, Kaoru’s stolen glimpses of Ōigimi and Naka no Kimi in the Hashihime chapter, Kashiwagi’s and Yūgiri’s covert peering at the Third Princess in the Wakana chapter, Prince Niou’s glimpses of Ukifune in the Ukifune chapter, Genji’s furtive observation of Utsusemi and Nokiba no Ogi in the Utsusemi chapter, Kaoru’s surreptitious observation of the First Princess in the Kagerō chapter, and Naidaijin’s secretive viewing of Ōmi no Kimi in the Tokonatsu chapter. The major scenes of kaimami in the Genji are almost exhaustively covered in this article.
Because Akeda’s article is closer to an overview of kaimami in the Genji than a full-fledged research study, its contribution is not comparable to that of Imai’s piece, which puts forth a methodology. Nonetheless, it is worth remembering that Imai’s theory is not the first. After Imai, Shinohara Yoshihiko wrote a piece called “Genji monogatari ni itaru nozokimi no keifu” (The genealogy of peering before The Tale of Genji).[4] By explicitly expanding the scope of analysis to the generic act of nozokimi (peering), Shinohara unfurls a wider story world than one that is limited to kaimami.
In this respect, Shinohara could have critiqued Imai’s kaimami theory for his misleading method of selecting examples. Yet, for some reason, Shinohara never mentions Imai’s theory, and that omission is regrettable. Had Shinohara taken up Imai’s theory and critiqued it, I am certain we would have re-evaluated our understanding of kaimami at an earlier date. Because Shinohara’s study did not engage Imai’s theory, however, Shinohara himself is normally overlooked by other scholars.
3. Work on Kaimami Going Forward
It goes without saying that my research on kaimami benefited greatly from Imai’s seminal study. To develop his ideas further, I offered the following suggestions in my work.
- Kaimami as a conceptual term has both narrow and broad definitions. The narrow definition is the literal meaning of the term, so the scope is too limited to be useful. In its broader sense, the term itself is superfluous. The notion encompasses a variety of acts of seeing and it does not need to drive a love story, allowing us to revisit the concept and further modify our understanding of kaimami. In the future, scholars may hone this conceptual apparatus by bringing in perspective theory, narrative theory, and image (picture) theory. We also need to consider and learn from foreign theories that are used for analyzing non-Japanese literature.
- I clarified that kaimami is not a one-way act between the male observer and the female observed, and thus necessitates a careful reconsideration of scenes in which female characters steal a furtive look at men. In doing so, we may discover that what appears to be a typically gendered kaimami is actually a case of mutual peeking, the observer transitioning into the observed and vice versa, the observed becoming the evaluator, and the observed putting on an act for the observer. Going further, we need to consider how the kaimami scene creates a theatrical effect for the text.
- I reaffirmed the importance of hearing in these scenes. Traditionally, scholars made a distinction between kaimami and eavesdropping. However, the two senses have parallel and mutually reinforcing effects. I proposed that auditory information is critical for the topos of kaimami to function effectively.
- I also reaffirmed the significance of the sense of smell. This is because olfactory references in scenes of kaimami emplace the gazer in the scene, proving his/her existence.
By considering these principles as we analyze and probe the texts, it becomes possible to critique the existing consensus around Imai’s approach to kaimami and update his theory. It also allows us to reflect on our own shortcoming of uncritically adopting Imai’s framework thus far. With more thoroughgoing analysis and re-evaluation, we are bound to see greater utility in kaimami as a framework for textual analysis.
In particular, we need to consider, with some urgency, whether or not the implications of taboo and immorality associated with nozoki (peeking) are applicable to the Heian topos of kaimami. By doing so, we can appropriately acknowledge our debt to Imai and put forth a newly refined theory of kaimami.
In any case, recall that I mentioned “image theory” above. Since kaimami is a method of describing the story world through the perspective of the protagonist (observer), such a scene can be easily rendered into a painting. As a result, Genji pictures (Genji-e) and Genji scrolls (Genji monogatari emaki) frequently include scenes of kaimami. Along these lines, I shall note that the kaimami theory can be easily combined with and enhanced by image theory and other methods. It is my hope to see comprehensive and interdisciplinary approaches to kaimami flourish for many years to come.
[1] Yoshikai Naoto, Kaimamiru Genji monogatari (Tokyo: Kasama Shoin, 2008).
[2] Imai Gen’e, “Kodai shōsetsu sōsakujō no ichi shuhō: Kaimami ni tsuite,” Kokugo to kokubungaku 25, no. 3 (March 1948).
[3] Akeda Yonesaku, “Genji monogatari no kaimami,” Nihon oyobi Nihonjin, October 1927.
[4] Shinohara Yoshihiko, “Genji monogatari ni itaru nozokimi no keifu,” Bungaku Gogaku vol. 68 (August 1973). This piece was later included in Genji monogatari no sekai (Tokyo: Kindai Bungeisha, 1993).