Semiotic Review 2: Monsters - article published June 2015

Licking the Ceiling: Semantic Staining and Monstrous Diversity

Michael Dylan Foster

Abstract:This essay proposes a concept of “semantic staining,” whereby images and meanings bleed into other images and other meanings, indelibly altering them within the popular imagination. By exploring this process as it is expressed in the naming and description of monsters—a cultural category particularly open to interpretation—we gain broad insight into the way knowledge is ordered, transmitted, and altered through time and space. We can also examine specific historical moments in which homogenizing forces and different media change taxonomies, categorization, as well as language and culture more generally. Based on a mixture of archival and field research, the essay focuses specifically on several examples of major and minor Japanese yōkai (monsters and supernatural phenomena) and how shifts in nomenclature change the way they are imagined.

Keywords: taxonomy, yōkai, naming, folkbiology, Japan

No one can articulate a syllable which is not filled with tenderness and fear, which is not, in one of these languages, the powerful name of a god. To speak is to fall into tautology.

--Jorge Luis Borges, “The Library of Babel”

Japan has a plethora of yōkai, a word variously translated as monsters, spirits, goblins, fantastic beings, or supernatural phenomena. Such things have long played a role in local traditions, belief systems, and folklore, as well as in visual and literary arts, and more recently in commercial and popular culture, including film, manga, anime and games. In fact, yōkai leap with uncanny facility from one platform to another, constantly remediated, and constantly taking on new meanings. They are equally at home in folk and commercial culture, and operate simultaneously within personal, local, regional, and national contexts. It is impossible, and probably fruitless, to make broad generalizations about yōkai, except to say that they are characterized by variety, abundance, and constant change.

Given the multi-platform, media mixed, heteroglot exuberance of yōkai then, acts of naming and describing provide order and a rationale of sorts: a semblance of semiotic consistency. A desire for order inspires (or is manifest in) what I call an “encyclopedic mode” of discourse, a tradition of encapsulating each individual monster into a concise and (relatively) self-contained unit in which it is labeled, described, and sometimes illustrated.1 This is of course the kind of natural-history format that informs modern zoological and botanical systems of taxonomy.

In the case of yōkai this format is particularly interesting because, like monsters everywhere, uncanny creatures and phenomena challenge taxonomic systems. Almost by definition, they are the anomalous and out-of-the-ordinary examples that, by refusing to fit into existing parameters, force systems to restructure, regroup and redefine. Alternatively, if we try to cram monsters into a bestiary of their own, we are also confronted with questions that lay bare the very scaffolding of any system of classification, revealing the invisible criteria that determine categorization schemata. It is exactly because yōkai challenge taxonomy—both by disrupting established systems and also by representing a boldly heterogeneous group of phenomena that somehow needs to be organized—that they work as powerful defamiliarizing agents in the often taken-for-granted order(s) of nature.

In the current essay, I focus on a few cases involving the labeling of yōkai in which the transference and transformation of names, images, and meanings through different media—the movement of signs through time and space—reveal how the chaos of local variation is marshaled together into a centralized taxonomic system. These examples demonstrate a complex play of signifiers and signifieds whereby images and meanings bleed into other images and other meanings in a process I call, tentatively at least, semantic staining. Like a stained shirt or carpet, an existing “thing” is indelibly colored by contact with another. As with the staining of wood, the object itself is forever altered even as the stain accents the grain and contours of the “original.” Thinking through the metaphor of semantic staining provides insight into the complex ways in which meanings change through time and place and from person to person.

While this essay and these ideas are still very exploratory and inconclusive, I hope the metaphor of staining is at least a little useful for thinking through questions of taxonomy and categorization, as well as language and culture more generally. My examples are specifically based on the Japanese case and I argue in fact that they are ultimately historically contingent, but at the same time the broader ideas here are transferable to other discourses of monsters and the questions they raise. Ultimately, the optic of semantic staining allows us to explore the processes through which ideas (broadly defined) travel through and between cultures. We can see these movements occurring on several different levels, including language and terminology, description, imagination and belief. I have not explicitly distinguished these different levels from each other in the pages that follow, in part because they are analogous to each other and also mutually constitutive, but I present the material here with the hope that others might more carefully parse out these layers. All of this will make more sense with some concrete examples, the first of which, coincidentally, concerns a yōkai that causes stains.

The Ceiling Licker

In the latter half of the twentieth century, manga artist and yōkai researcher Mizuki Shigeru (b. 1922) particularly inspired a resurgence of yōkai in the popular imaginary, and his images are ubiquitous throughout Japan today. Mizuki is known not only for his original narratives and newly invented yōkai, but also for documenting and illustrating yōkai from all parts of the country. In one of his many monster catalogs, he presents a picture of an emaciated, lanky creature with frilly hair. Seemingly suspended in mid-air, it is licking a wooden ceiling with a long tongue. Mizuki explains:

There is a yōkai called “tenjōname” [ceiling-licker]. You would think that it would be a great help for neatly licking clean the ceiling, which normally does not get cleaned—but this is not the case. It is fine that this “ceiling-licker” licks the ceiling without being asked, but it actually causes dirty stains to adhere. When there is nobody around in an old house, temple, or shrine, it comes out and licks with its long tongue...It seems that if they found stains on the ceiling, people in the old days thought it was the handiwork of the tenjōname.

Mizuki goes on to describe his own experience with this creature through a personal, nostalgically framed, remembrance: “When I was a child, there was an old woman in the neighborhood who was particularly knowledgeable about yōkai. On occasion, she used to stay at our place, and she looked at the stains on the ceiling of our house and said, ‘Look! The tenjōname comes out at night and makes those stains’” (Mizuki 1984:152-153).

Figure 1. Bronze figuring of <em>tenjōname</em> on display in Sakaiminato (Tottori Prefecture), hometown of Mizuki Shigeru. Photograph by author, 2007.

Figure 1. Bronze figurine of tenjōname on display in Sakaiminato (Tottori Prefecture),
hometown of Mizuki Shigeru. Photograph by author, 2007.

In another catalog, Mizuki discusses this yōkai in a similar fashion, placing it with others in a section labeled “Yōkai I have encountered” (Mizuki 1994:46-47). And in a memoir about his childhood growing up in the rural port town of Sakaiminato, he makes explicit the way he learned about the tenjōname from Nonnonbā, an “old woman in the neighborhood” who taught him local folklore and instructed him in the way of the yōkai. He explains that when she told him “with a serious face” that the stains on the ceiling were made by a creature called the tenjōname, “there was no room for doubt” (Mizuki 1997:20).2

Mizuki’s description certainly seems plausible. Throughout Japanese history, different kinds of yōkai were offered as explanations for otherwise unexplainable things, from strange sounds in the forest to glimmering lights in a graveyard. It would not be surprising if the long-tongued tenjōname was invoked wherever water or mold stained the ceiling of an old house. And it turns out, in fact, that others have documented this same creature. In a 1929 collection of yōkai-related material, for example, folklorist Fujisawa Morihiko (1885-1967) notes that “in old houses and old halls, stains on the ceiling are the result of this yōkai” (Fujisawa 1929:10). Iwai Hiromi (b. 1932) goes into more detail, proffering a description similar to Mizuki’s and commenting that “nobody has ever seen the figure of this yōkai. But people have imagined it as having a very long tongue and being tall enough to reach the ceiling” (Iwai 1986:139; also 2004: 77-78). According to writer and folklorist Yamada Norio (1922-2012), there is even a legend of a tenjōname captured in the house of a retainer for the Tatebayashi clan, who proceeded to bring the creature to Tatebayashi Castle (in present day Gunma Prefecture) where it was employed to lick up otherwise hard-to-reach spider webs (Yamada 1974:22).

It would seem then that the tenjōname has all the qualifications of a legitimate folkloric monster found in both belief and legend. In fact, however, if we hunt for more signs of the creature, we run into questions about memory and documentation, both on a personal and cultural level. It turns out, for example, that the origins of Yamada’s story of the retainer are obscure; Yamada himself provides no citation or documentation. And Iwai’s description of a yōkai that “nobody has ever seen” is most likely based not on word of mouth, but ironically on a picture that he has seen. Folklorist Komatsu Kazuhiko points out that certainly both Iwai and Mizuki have seen an image of the tenjōname drawn in 1784 by Toriyama Sekien (1712-1788), an Edo-period illustrator and literatus whose series of yōkai catalogs are still the go-to guides for images of the Japanese otherworld.3

Sekien documented indigenous folk knowledge about yōkai and also translated monsters from various Chinese sources into pictures. But in addition he created new yōkai. Although it is impossible to confirm whether he actually fabricated the tenjōname from scratch or was illustrating something already circulating in oral/local tradition at the time, his is the first known image of the creature and Sekien’s work had the effect of transforming the tenjōname into a visualized and named representation (Komatsu 1994:63).4 It was undoubtedly this image that inspired Mizuki’s own image of the creature as well as Iwai’s comments.

Figure 2. Toriyama Sekien’s <em>tenjōname</em> from <em>Hyakki tsurezure bukuro</em> of 1784.

Figure 2. Toriyama Sekien’s tenjōname from Hyakki tsurezure bukuro of 1784.

Sekien’s entry on the tenjōname includes an explanation: “It is said that if the ceiling is high, [the room] will be dark and in winter it will be cold; but this is not because of the way the house is made. Rather, it is entirely due to the machinations of this yōkai [kai]. Just thinking about it will give you the creeps” (Inada and Tanaka 1992:273). This rather confusing explanation is complex with literary allusions. Specifically, it refers to the famous essay collection, Tsurezure-gusa (translated as Essays in Idleness; 1333), by the monk Yoshida Kenkō (c.1283 – c.1350).5 In entry number 55 of this text, Kenkō writes that “a room with a high ceiling is cold in winter and dark by lamplight” (Keene 1967:50-51). Sekien’s tenjōname description plays with this assertion: it is not the architecture that makes it dark and cold, he suggests, but the haunting of the tenjōname.

I mention here only the most obvious reference in Sekien’s image. Scholars have carefully studied Sekien’s work and suggested other allusions that would have been accessible to Edo-period readers. For example, yōkai expert Tada Katsumi has suggested that the creature itself may be a visual pun on the Edo-period fire companies, with its frilly clothing, mane, and fingers resembling a matoi, the flag-like banner carried by the firemen (Tada 2001:320-321).

What I would like to stress here, however, is not Sekien’s creativity and literary-imagistic playfulness but that his own entry contains absolutely no mention of stains on the ceiling. Despite the fact that Mizuki’s image some two centuries later is unabashedly derived from Sekien’s, it elides the allusion to Kenkō and offers instead a nostalgic folksy explanation, complete with a wise old woman to convey the folkloric knowledge of the past. Where did Mizuki get this? It is of course possible he simply made up this part of the story, rhetorically transforming an individual artistic creation into a folk product of collective imagination. Certainly the old-woman-taught-it-to-me narrative is much more resonant and memorable for a post-war manga readership than Sekien’s allusion to a literary classic.

On the other hand, perhaps Mizuki is to be believed absolutely, in which case we can surmise that shorn of its association with Yoshida Kenkō, the image of the tenjōname entered the cultural imaginary of rural Japan, eloquently morphing over the next two hundred years into an explanation for why ceilings become stained over time—an explanation that Mizuki then heard as a child growing up in rural Japan. That is, it is possible Sekien did create the tenjōname and that his invention migrated into oral tradition and village folklore where it acquired new meanings. Whatever the case, the work of these two artists demonstrates the dynamic and non-linear relationship between the oral and literate (and visual), as well as the lively interplay of tradition and individual creativity necessary for the making of monsters.

But my point here is not to seek the true origins of this fairly minor yōkai; rather I trace its murky provenance as a single example of the way in which meanings are always unstable and mutable. Even as the image and name of this creature remain essentially unchanged, the explanation for its behavior shifts significantly. My guess is that in present-day Japan, most people familiar with the tenjōname know it only from Mizuki’s description, and likely buy into his explanation of it as something from the prewar rural imagination. Its genesis (presumably) in the sophisticated urban milieu of Toriyama Sekien has been semantically stained, as it were, by this modern nostalgic explanation, a meaning as arbitrary as the first perhaps but more resonant with the postwar historical imaginary. And we see how this new explanation indelibly colors the old; the contours of the original—lanky body, long tongue, frilly hair—remain, but in the popular imagination the tenjōname is now associated with wizened old women, rural farmhouses and stains on the ceiling.

Kappa, also called Kawatarō

The tenjōname is only a very limited case, and really a pretty obscure one. In the next few pages, I provide another example—this time, a much more complex one concerning a yōkai known throughout Japan as the kappa. The kappa is famous as a water sprite that lives in rivers and ponds. It is notorious for nasty habits such as drowning horses, cows and small children. In the contemporary popular imagination it resembles a giant frog or turtle: the size of a young child but disproportionately strong, it is commonly portrayed as a scaly-slimy creature, greenish in color, with webbed feet and hands, and a carapace on its back. A concave indentation (or saucer) on top of its head contains water; if this water is spilled, the kappa loses its super-human strength.

Despite its murderous tendencies, the kappa can be both playful and exceedingly honest; it particularly enjoys sumo wrestling and is known to challenge passers-by to a match. One tactic for defeating a kappa in sumo is simply to bow beforehand; when the kappa bows in response, it spills its strength-giving liquid from the saucer on its head. Kappa are said to be fond of cucumbers, a fact celebrated in Japan and much of the rest of the world by the fact that roll sushi made with cucumbers is called kappa maki. The kappa is at once a dangerous, demonic monster—a reason to warn children about swimming alone—and also an amusing, if disgusting, water sprite.

Along with this folkloric image of the kappa, there is also a simultaneous image of the kappa as a kind of cutesy “character.” In the guise of this sanitized version, the kappa is ubiquitous in Japanese popular culture today, appearing on children’s television shows, as advertising icons, and even as local mascots for “village revitalization” in communities throughout the country. To be sure, the creature is still associated with folklore, but it is also very much part of the popular and commercial imagination, more often than not portrayed as a manga-esque or anime character: green, cute, loveable, harmless.

Figure 3. Kappa figurine atop a postbox in Satsuma-Sendai (Kagoshima Prefecture). Photograph by author, 2010.

Figure 3. Kappa figurine atop a postbox in Satsuma-Sendai (Kagoshima Prefecture). Photograph by author, 2010.

The description above—both the folkloric and popular culture image—would be recognized by people throughout Japan, and has even made it onto the international stage with mention in the Harry Potter series (Rowling 1999; 2001). This image of the kappa is indeed somewhat monolithic. When folklorist Itō Ryōhei taught a class on Japanese culture at a university in Taiwan, for example, he asked his students to draw pictures of famous yōkai, including the kappa. More than any of the other yōkai, images they drew of the kappa were similar to each other and all very similar to the way a Japanese college student might illustrate one: “a small cute creature, a head with hair surrounding a saucer, a beak and webbed hands and feet, and a shell on its back” (Itō 2011:220-221). In other words, a fairly unified/generic image of the kappa exists within Japan and, through mass media and other means, has been transmitted to other cultures as well.

While all this seems straightforward at first glance, the situation is actually more complicated. Historically speaking, the word kappa is just one dialectal variant; according to some estimates, in fact, there are over one hundred regional variations on the name and characteristics of this water creature (see Ishikawa 1985:41; also Wada 2005). I am oversimplifying significantly here, but generally we can divide the country into east and west.6

Until around the middle of the Edo period (c. 1600-1868), the shell-bearing, amphibious kappa described above was distributed in the eastern part of Japan, from the Kantō region (centered in Edo/current-day Tokyo) to the northeast of the country. And although there were numerous different names for this creature, kappa was one of the more common labels associated with it.

In contrast, in western Japan (from the Osaka-Kyoto area to Shikoku and Kyushu) there was a creature called a kawatarō, or some variation of this name, that was hairy and walked upright, more like a monkey than a frog or turtle. An early Japanese encyclopedia known as the Wakan-sansaizue (c. 1715) presents a line drawing of the creature and explains:

About the size of a ten-year-old child, the kawatarō stands and walks naked and speaks in a human voice. Its hair is short and sparse. The top of it head is concave, and can hold a scoop of water. Kawatarō usually live in the water but in the light of the late afternoon, many emerge into the area near the river and steal melons, eggplants and things from the fields. By nature the kawatarō likes sumō; when it sees a person, it will invite him [to wrestle]…If there is water on its head, the kawatarō has several times the strength of a warrior…The kawatarō has a tendency to pull cattle and horses into the water and suck blood out of their rumps. People crossing rivers must be very careful…(Terajima 1994:159).

Figure 4. Kawatarō from the <em>Wakan-sansaizue</em> (c. 1715) by Terajima Ryōan.

Figure 4. Kawatarō from the Wakan-sansaizue (c. 1715) by Terajima Ryōan.

This is the first-known extended description and illustration of a kappa-like creature, and clearly contains many of the elements (indentation on head, love of sumō, pulling of livestock into the water) that would come to be associated with the water sprite nationwide. However, the word kappa is nowhere to be found in the Wakan-sansaizue and in many ways the creature described here looks quite different from the shell-bearing kappa image common today. It is no coincidence that Terajima Ryōan, author of the Wakan-sainsaizue, was from Osaka in the west of the country. Similarly, a 1754 text called the Nihon sankai meibutsu zue shows a group of kawatarō from Bungo (modern day Oita Prefecture) in Kyushu walking upright and playing near a river; the text describes them as “the size of five-year old children and having hair all over their bodies” (see Kagawa 2012:6).7 In short, the western image of the kawatarō was that of a hairy monkey-like creature; while it shared some characteristics of the eastern kappa, it was also clearly distinct.

But somewhere along the way, the image of the kawatarō got stained by the image of the kappa. Most likely this was a gradual process driven by the publishing industry, the center of which shifted to Edo with the emergence of inexpensive popular books, generally known as kusazōshi, that were often rife with illustrations. It was in this milieu that Toriyama Sekien produced his influential yōkai catalogs. His first series of illustrated monster books, published in 1776, included an image of a web-footed creature with a round indentation on its head, peering out from a clump of flowering lotus plants. The picture is labeled “Kappa, also called kawatarō.” With this simple caption, then, Sekien subsumes the name and image of the western kawatarō under the name and image of the eastern kappa. Whether intentional or not, the phrase, bolstered by the accompanying image and the power of a popular publishing industry, signifies the moment of an indelible staining, the instant one “species” of water spirit becomes a “sub-species” of another. This is not to say that Sekien himself caused the change; more than likely, he was simply articulating the results of a process that had been going on for years. But whatever the case, in the years after Sekien’s work, and certainly by the late nineteenth and early twentieth century, the word kappa and this amphibious, slimy version of the creature, would become the common image of Japanese water sprites throughout the country (Kagawa 2012:6-7).

Figure 5. Kappa, from <em>Gazu hyakkiyagyō</em> (1776) by Toriyama Sekien.

Figure 5. Kappa, from Gazu hyakkiyagyō (1776) by Toriyama Sekien.

In contrast to the overwhelming ubiquity of the kappa, it is fair to say that most residents of Japan have probably never heard of a kawatarō. Or if they have heard the word itself, it is through a local legend where the creature is explained as being “like a kappa” or “a kind of kappa” or, perhaps, “a local name for kappa.” In short, kappa is the default and somewhat generic signifier—both name and image—for kappa, kawatarō, and dozens of other localized water creatures found throughout the country.

But so what? Isn’t this process of naming and renaming, of influence and assimilation and diffusion, a natural one? In discourses of folkbiology, ethnobiology, taxonomy and classification more generally, scholars are often able to locate “prototypes” within particular sets of “natural categories.” Such sets are “defined upon prototypes that are definitive of the focal area and that more or less cover objects at a distance from the focal area—that is, fuzz-off toward the edges of the semantic field” (Laughlin 1993:20).8 Clearly, in this case, the kappa is the most representative or typical member of the family of mischievous Japanese water creatures. The kawatarō and many other local variants show a “family resemblance” (Rosch and Mervis 1981) that allows them all to be part of the same general category even as they are not quite central or representative of it.

The question remains, of course, as to how one “thing” becomes central to an entire category, or inversely, how other “things” become less central. In one sense, the process by which the kawatarō is subsumed under the sign of the kappa is metonymic for all processes of synthesis and influence. But by examining the case of a fantastic creature—as opposed to biologically or ontologically “real” animals—the contours of these processes are defamiliarized and more striking. Ultimately, of course, “kappa” is not a so-called natural category, but rather a cultural category; its representative or prototypical status is contingent on human relations, mutable interpretations, and historical events. The fact that the kappa and not the kawatarō image became the exemplar for Japanese water creatures is a historical artifact.

Figure 6. Sign for Kappa yokochō (Kappa alley) an entertainment neighborhood in Osaka, where the local water creature used to be called kawatarō.

Figure 6. Sign for Kappa yokochō (Kappa alley), an entertainment neighborhood in Osaka,
where the local water creature used to be called kawatarō. Photograph by author, 2013.

Monstrous Taxonomies9

The case of the kappa raises several points of interest in terms of the way objects are classified and organized. First, as mentioned earlier, the taxonomy of yōkai, while seeming to follow patterns similar to those for animals and plants, is different simply because the “object” being classified is not stable. It is illusive, elusive, and therefore also allusive—the nature and form of a yōkai in one community can be stained with the characteristics of a yōkai from another community so that the people in the first community adopt the descriptive language of the other community to describe their own local yōkai. And not just to describe it: they start to envision it and understand it through this “other” language. Over time, then, this staining alters the yōkai’s hue and appearance so that eventually it becomes a different yōkai. Such a transformation does not generally happen with a “real” creature such as, for example, a squirrel or an elephant; to be sure, names and descriptive language may differ over time and place, but presumably the morphology of such animals remains generally the same. Inversely, because yōkai are culturally and historically contingent, individual distinctions between them are relatively unstable and therefore they have limited natural resistance to being reclassified, renamed, and even reshaped.

Second, not only does the monstrosity and elusiveness of yōkai destabilize any formal classification system imposed on them, but the variety of names associated with them also problematizes the possibility of type-token distinctions. Some yōkai names refer to what we might consider, in modern scientific taxonomic terminology, genus or family or species, but others are more akin to proper names, associated only with a single creature known in a single community.10 Any list of yōkai will necessarily and somewhat randomly cut across diverse levels of taxa. Despite this difficulty, one of the most common ways of documenting yōkai—at least since Toriyama Sekien’s time—is through encyclopedic classification. Such classificatory attempts eloquently reveal the impossibility of a stable taxonomy: they include arrangement by Japanese syllabary order (e.g. Murakami 2000), by geographical distribution such as prefecture (e.g., Chiba 1995) or by places of encounter such as “Yōkai of the rivers and bogs,” “Yōkai of the ocean,” or “House yōkai” (Itō 2011). And some, such as a few of Mizuki Shigeru’s compendia, are refreshingly subjective, with sections entitled, “Yōkai I have met” and “Celebrity yōkai” (Mizuki 1994). Although writing of European attempts to classify the monstrous, Harriet Ritvo’s words also neatly apply to the yōkai situation in Japan: “Capacious, motley, and irredeemably vernacular, the category ‘monster’ proved invulnerable to expert analysis” (Ritvo 1997:137; see also Foster 2015:105-114).

Finally, I want to emphasize that this impulse towards systematization—and all the paradoxes it engenders—emerges out of a modern context. In the study of the supernatural, modernity is often portrayed as a culprit, the killjoy intent on disenchanting the party of the fantastic. In Japan, this was explicit: most famously the dynamic Buddhist philosopher Inoue Enryō (1858-1919) set out on a mission to banish superstitions and eradicate beliefs in yōkai, replacing them with modern scientific knowledge so that Japan could join the fraternity of modern nation-states.11 While this disenchanting ideology contributed to a fading of yōkai from the cultural imaginary, I would argue that just as influential as the rationalistic rhetoric that debunked specific supernatural phenomena is a rationalizing logic of modernity intent on amalgamation and standardization. This reflects ideology, of course, but also develops from concomitant processes of urbanization, migration, technological innovation, and commercial market forces.

In Japan, these dynamics start emerging during the early-modern Edo period (c.1600-1868). Without going into great detail here, suffice it to say that during this time cities expanded immensely, making Edo (current-day Tokyo) larger than Paris or London at the time (urbanization) (see Hanley 1987:1). As people from the countryside moved to cities (migration), they brought with them local beliefs that clashed and meshed with others. Meanwhile vibrant new forms of popular media were evolving—from kabuki and bunraku drama, to inexpensive books and woodblock prints (technological innovation). These were sold by a thriving and entrepreneurial merchant class to a newly urbanized public (market forces). Within this context, yōkai (and many other traditional beliefs and practices) changed form, becoming hybridized and less locally specific. This is the milieu in which Toriyama Sekien, for example, produced his creative volumes.

All these processes continued in one fashion or another through the twentieth century, and indeed continue to today. This kind of consolidation of disparate traditions under the rubric of a single standardized identity is part and parcel of the creation of a national identity. Even as the late-nineteenth and early-twentieth centuries in Japan witnessed the linking of distant locations and tiny villages through a massive network of railroads, so too in more subtle ways, systems of beliefs were linked together—including those relating to monsters. And just as the railroad and other transportation systems radiated out from large metropolises such as Tokyo, so too the influence on the yōkai world spread out from the urban centers, staining the peripheries.

Even on the level of the supernatural then, a nationalist agenda made less distinctive the particular belief systems of particular locales. Perhaps it is no coincidence that it was also during the early twentieth century that Japan embarked on its colonial incursions into countries on its own periphery; whether we can talk about a colonization of local monsters or not, certainly the standardization of many common yōkai reflected a broader mindset of unification that bled into all aspects of social-political-cultural life at this time. In some cases, such semantic staining might be intentional, but more likely it was also a byproduct of capitalism and the commercial advance of popular literature and images (in the Edo period) and transportation infrastructure, advertising, and other media (in the twentieth century). Although they did not entirely disappear, local versions of water creatures all around the archipelago were stained by the dominant image of the kappa and the meanings it brought with it. The process can be thought of as a kind of syncretism, in which the local gods are translated into a version of the central god—a way in which disparate communities, through a tradition of monsters, are tied together as a “nation.”

I am speaking in superlatives here to be provocative. But in fact this informal rationalization of water monsters (and other yōkai) was occurring at the same time as very formal and intentional policies of the Meiji government (1868-1912), particularly the so-called shrine mergers (jinja gōshi) in which local spaces of worship were forcibly merged with others and brought together into a hierarchical system. Although the merging of water-spirit beliefs was by no means as institutional or potentially nefarious, the processes are similar. Memories are overwritten with the contingencies of the moment; old beliefs are stained by new ideologies, retaining only the contours and grain of the earlier iteration.

In a sense, the standardization of yōkai reflects a loss of local hegemony. If yōkai by their very existence represent a taxonomic challenge, then the heterogeneity of vernacular yōkai continued to pose a threat to emerging systems for reigning in and ordering “superstitious” beliefs and narratives. There is a tyranny of the ascendant monster here, a centrifugal force that spins outward and stains regional variants with a dominant name and set of characteristics. Perhaps there is also an attendant desire on the part of the periphery to identify its local yōkai with those in the metropolis. It is part of the process of nation-building itself, a unification of disparate local beliefs into a single image-name that all “Japanese” can share. Just as cities are built over the ghosts of former cities, so the shiny, generic image of the kappa overrides older, more localized kawatarō and other water creatures. Suppressed by contemporary standardizations, these earlier iterations are all but invisible now: quaint ghosts of a past diversity.


My discussion of the kappa-kawatarō issue here is broad, and admittedly based on two quite generalized images; I should reiterate that I have simplified quite brazenly for the sake of clarity. In order to explore these dynamics more specifically, I would like to present an intensely local example that demonstrates the extent to which the kappa image/name stained even the most localized manifestation. For my own fieldwork, I have spent a great deal of time in the village of Teuchi on the island of Shimo-Koshikijima off the west coast of Kagoshima Prefecture. This is a very small area: the island has a population of fewer than 3000 people and the village of Teuchi only has about 730 residents. The only access to the island is by ferry—about an hour from mainland Kagoshima Prefecture, which itself is about as distant as you can get from Tokyo and still be on one of the four major islands of Japan. All this to say that Teuchi is on the periphery of the nation even now, and in the early twentieth century it was certainly a long distance geographically, technologically, and psychically from centers of power and culture.

Figure 7. Village of Teuchi, on the island of Shimo-Koshikijima, Kagoshima Prefecture. Photograph by author, 2010.

Figure 7. Village of Teuchi, on the island of Shimo-Koshikijima, Kagoshima Prefecture. Photograph by author, 2010.

There on the island, I came across a yōkai I had never heard of before—something called a gamishiro.12 I first learned about the creature from an islander in his late fifties named Hironiwa Yoshitatsu who told me that as a child he was warned not to swim in the ocean after eating cucumbers because he would be attacked by a gamishiro. Looking back, he suggests this may have been a way for adults to prevent children from stealing cucumbers, a local summer crop. His parents, Hironiwa Mamoru and Hironiwa Eiko, both in their 80s, confirmed that they were told the same thing when they were children, so the gamishiro has been around for at least eighty years. Another fiftysomething islander recalled being told that she should not go swimming alone in the ocean because a gamishiro would grab her leg and pull her under, but she does not recall an association with cucumbers.

Most people could not give me a specific physical description of the gamishiro: they just explained that it was a frightening and dangerous creature lurking in the ocean. However, Teuchi resident Hashiguchi Yoshitami, now in his late eighties, described it as having a cavity on its head and claw-like limbs. As a child he was warned that the gamishiro lurked particularly in two deep pools of water in the bay, and that it would pull you down by grabbing your Achilles tendon. A young relative of his almost drowned in one of those pools, he said. When the child was eventually rescued, everybody checked his ankles for marks from the gamishiro’s claws, but there were none.13

Some of these details, such as the association with cucumbers and the cavity on its head, are clearly cognate with predominant kappa beliefs from the Kantō region (and elsewhere). At the same time, some of the characteristics—such as the fact that it pulls people down by the ankle or Achilles tendon—are not necessarily associated with the kappa image today, though nor are they difficult to connect with it. On the other hand, the fact that the gamishiro is associated with the ocean (only natural, of course, for an island) distinguishes it from most kappa, which generally live in freshwater ponds or rivers.14

It is evident then that even in descriptions of the creature collected in the early twenty-first century, the gamishiro is distinct in some ways from the kappa. To be sure, it may function in similar ways (a warning against swimming, for example) and it definitely shares some physical and behavioral characteristics with the kappa, but it is ultimately a monster very much associated with a specific location (two deep pools in the bay) and the particular concerns of a small island village. But what struck me most vividly was that everybody who told me of the gamishiro—at least seven different island residents—immediately noted that it was like a kappa. Hironiwa Mamoru, for example, explained to me simply that a gamishiro was “what we would now call a kappa” [ima de iu kappa]. And then he went on to say that “nobody has actually ever seen one” [jissai mita hito wa inai].

The comment here, of course, is reminiscent of Iwai’s point earlier that nobody had ever seen a tenjōname, and while the context is completely different, they gesture to a similar process. Hironiwa-san had never actually seen a gamishiro, so he used the generic, national image of the kappa as a comparative device. And with such a move, the specific image of a creature locally known to kids growing up eighty years ago in Teuchi becomes indelibly stained with the image of the kappa. It is impossible to know, of course, how much of this generic kappa image actually informs the present understanding of the gamishiro. Did the “original” gamishiro enjoy cucumbers too? Did it have a cavity on its head? And when did the association with the kappa become the standard way of describing it? Nobody I spoke with could answer these questions. But that is just the point: once something is stained, it is permanently altered. There is no seam between that which came first and that which came after. Unlike paint, the color permeates the substance it touches and cannot be scraped off. Hironiwa-san may never have seen a real gamishiro or even a picture of one—I have not been able to find any illustrations of the creature—but there are plenty of images of the kappa in popular culture. And these images become the gamishiro.

Description often works through analogy, through metaphor or simile, whereby one thing is explained in terms of another. Dictionaries cannot avoid tautology. Semantic staining too begins as a vibrant analogic process but implies an even more pervasive overlapping of thing and description of that thing; the comparison becomes a blending whereby the object (or creature) in question becomes indelibly synced with its analogy. The gamishiro is like a kappa, and therefore the gamishiro is a kappa.

Of course, it is important to note the context of my discussions with island residents here. I have frequently visited (and occasionally lived on) Shimo-Koshikijima for some fifteen years and the people I spoke with are friends. They also know that I am particularly interested in the study of yōkai. Nevertheless, while on the island I am still considered a non-islander, a non-native speaker of Japanese and ultimately identified as a “foreigner” (specifically as an “American”). When they explain to me that the gamishiro is “like a kappa” or “what we now call a kappa,” my interlocutors are participating in a pragmatics of semantic staining, choosing a term they know I will understand—that is, they select appropriate language for the given discursive context. But such language usage is not limited to speaking with foreign visitors or non-islanders. I have witnessed the same dynamics when older islanders describe gamishiro to island children who know about kappa from television and manga but never realized the local version had a local name. Hironiwa-san’s phrase, ima de iu kappa, might also be (liberally) translated “what young people would call a kappa.” In a sense, then, this pragmatics may reflect islander recognition that the word gamishiro is specific to this place and to a time now past. It needs to be translated into the standard and contemporary vernacular because both the local and the past have become, as it were, foreign countries.

The semantics of semantic staining

In the preceding pages, I have suggested the notion of semantic staining as a way to think about how yōkai change over time, both in terms of nomenclature and meaning. To reiterate more systematically, semantic staining is a process with certain stages or characteristics. First, two different things are perceived as analogous, connected by apparent similarities with regard to qualities (shape, size, color, behavior) or name or even sound (see e.g., Berlin 2006 on phonaesthesia). The relationship is Iconic. As Peirce puts it, an Icon is a “Sign that represents its Object in resembling it” (Peirce 1966:368). Or in Susan Gal’s words, “Icons are signs that represent by being taken as in some way similar to their objects” (Gal 2013:34; emphasis added). As Gal’s phrase “being taken” implies, similarity is neither natural nor inevitable: ideology, politics, history, and other socio-cultural factors are responsible for the “attribution of necessity to a connection” (Gal and Irvine 1995:973).

Second, in the process of semantic staining, the relationship between the two terms is asymmetrical: one must become dominant. This is also, of course, a factor explicitly determined by ideological, cultural and historical circumstances. Third, the dominant thing becomes the default descriptor for the other thing, gradually staining it so that the original meaning is no longer viable or visible as a separate entity, though its contours may vaguely remain. Fourth, similarity is subsumed by sameness: words of comparison, such as “like” or “as” or their equivalents gradually disappear and the descriptor becomes more than a way of describing—it becomes the thing itself. And finally, although we may be aware that such a process has occurred, there is no reversing it; the staining is permanent.

I propose this process as a very inchoate way of understanding how yōkai change over time and place. My exploration is still inconclusive, but hopefully will stimulate others to work on these problems from a more semiotic or taxonomically informed perspective. So, I end here openendedly with a few propositions that might be pursued in the future.

Not prototypical prototypes.

In terms of monstrous taxonomy, prototypicality and categorization are based on cultural forms and not natural forms. Family resemblances are in the eye of the beholder. To a certain extent, prototypicality is a negotiation between perspectives. In the case of the gamishiro, for example, the image of the local creature was stained by the imported (and overwhelming) concept of the kappa. While scholars of taxonomy and folkbiology may seek variables that determine prototypes and categories for animals, such as particular types of fish (de Oliveira, Barreto, Begossi 2012; Bang, Medin, Atran 2007), the same sort of close attention to local taxonomies of monsters might reveal critical cultural variables based not only on perceptions of nature, but on perceptions of more elusive and numinous fields of belief, imagination, creativity and dreams.


The process of semantic staining leads to an ambiguation of species, as it were, so that they blend together, or so that the prototype becomes the hegemonic example cognitively “colonizing” all the others. Presumably this also leads to a decrease in diversity. Of course, this can happen with all sorts of classification systems, but with ontologically real (capturable) animals, for example, merging of nomenclature may mean less diversity of nomenclature but does not translate into fewer actual animals on the ground (or in the water or air). On the other hand, with yōkai and other imaginary zoologies, a reduction of signifiers can indeed translate into fewer signifieds. When a local water-dwelling creature called a gamishiro is called a kappa, the gamishiro all but disappears from the taxonomy. Or rather, it fades from the lexicon—because in distinction to natural histories, the ordering of yōkai (and monsters elsewhere) is less a taxonomy than it is a lexicon: in many cases words precede things.


Of course, local names are analogous to local dialects. National taxonomies/nomenclatures are akin to hegemonic languages that semantically stain the subtleties of local and regional dialects. By explaining to me that the gamishiro was “like a kappa,” my friends were “translating” the local dialect into a language that I, and other non-islanders, could understand. But it was not only non-islanders like myself who required translation; younger islanders too do not understand what a gamishiro was, just as they too have trouble following the dialect of their elders. Through this process of translation, old words may not be completely erased, but they are indelibly stained: the local gamishiro, like the subtleties of any dialect, are colored by and made comprehensible only through the lens of the “outside” or “standard” language.


But even as they are subject to the centralizing, standardizing forces of modernity, local monsters also embody the potential for resistance. Japan’s postwar recovery and rapid economic growth caused massive and rapid urbanization and accompanying depopulation of rural areas. In response, great efforts were made to revalue (and repopulate) the peripheries, often through nostalgically infused processes of “village revitalization” (mura okoshi), in which a local community would develop a particular cultural resource for purposes of identity and tourism. In many cases, the resident yōkai was given new life as an icon of the region.15

In this context, difference was celebrated: in the city of Tōno in Iwate Prefecture, for example, the traditional water creature may be called a kappa like so many others, but much is made of the fact that Tōno’s native kappa are red in color as opposed to the more common green. The implications of making such distinctions are, of course, ambiguous. Even as revitalized local creatures evoke a unique identity in contrast and resistance to the standard form, so too their difference—and the need to accentuate it—serves to highlight the hegemony of the norm. Such narcissism of minor differences, as Freud might put it, creates distinctions between local communities/yōkai/dialects even as it keeps them all within the broader rubric of an established nation/taxonomy/language.


Because the examples of semantic staining I have outlined here for both the tenjōname and the kappa took place in the past, we can observe the dynamics of change in step-by-step slow motion. I would contend however that a similar process is occurring all the time with yōkai and other phenomena. It can be seen, for example, in the transmission of Internet memes across sites, languages, and cultural groups. In particular, it is worth noting the current phenomenon of “character” creation, through which kappa and all sorts of other yōkai become crystallized with particular characteristics, and are then used as manga characters, local mascots, toys, etc. Just as the process of naming standardization was a modern practice, the process of “characterization” is, at least in part, an effect of postmodern commodification and technological practices. Indeed, while usually not yōkai per se, so-called yuru-kyara, loose or wobbly characters used for promotion of tourism and local products, are created and disseminated through processes very similar to those that occur with yōkai.16 The steps visible in the examples of historical yōkai provide some sense of what is happening at light speed through the contemporary media mix.

Figure 8.  “Bari-san” yuru-kyara stuffed “animals” for sale. Bari-san is associated with Imabari City in Ehime Prefecture, famous for its chicken products. Photograph by author, 2013.

Figure 8. “Bari-san” yuru-kyara stuffed “animals” for sale. Bari-san is associated with Imabari City in
Ehime Prefecture, famous for its chicken products. Photograph by author, 2013.


In the final analysis, then, semantic staining is a process through which diversity of cultural expression is attenuated. If biodiversity is critical to the health of the natural environment, what of monstrous diversity or, more broadly, cultural diversity? What does it mean when an abundant and subtly distinct taxonomy of monstrous creatures is organized into large fuzzy sets, in which certain exemplars semantically overpower less central figures? When diversity among monsters is reduced in this way, how is human imagination affected?

As the yuru-kyara example suggests, concomitant to the process of semantic staining may be a process of semantic mutation, whereby previously unimagined forms and combinations infuse the lexicon. Somewhere, right now, a new tenjōname is being invented. If the rationalizing processes of modernity are in part responsible for attenuating the diversity of monsters in Japan, then inversely, perhaps the rapidly shifting multi-platform mixed media of the contemporary moment can lead to a reinvigoration and infusion of new creatures in the cultural imaginary. Perhaps the diverse ultra-specialized niches of media and social networks—to say nothing of commercial products such as the massively popular Yōkai Watch franchise—will engender radically new “local” yōkai of all sorts. Or perhaps, regardless of technology, this process is always already occurring: somewhere right now, just as in the past, children are imagining a water creature emerging from their local river or pond. It may look a bit like a kappa but it will also have particular features, and perhaps a specific name, that only they will understand.17

Ultimately, whether yōkai appear as cyberfolkloric monsters or commercial “characters” or new versions of local creatures, they continue to challenge familiar categories and taken-for-granted structures. They remain in motion: emergent, changing, staining and being stained. And monstrous taxonomies are never just about monsters.

I would like to express my deep gratitude to Paul Manning for encouraging me to explore this material and for keeping after me, always with good humor, to get the article written (and rewritten). I also thank Christopher Bolton and two anonymous reviewers, all of whom provided brilliantly insightful suggestions on how to improve and expand my thinking on these matters. I thank Michiko Suzuki for too many things to mention here. And my appreciation also goes to my friends on Shimo-Koshikijima for their hospitality, generosity, and for putting up with my silly questions about local monsters.


Unless otherwise specified, all works in Japanese are published in Tokyo.

Azuma, Hiroki. 2009. Otaku: Japan’s Database Animals. Translated by Jonathan Abel and Shion Kono. Minneapolis: University of Minnesota Press.

Bang, Megan, Douglas L. Medin and Scott Atran. 2007. “Cultural Mosaics and Mental Models of Nature.” Proceedings of the National Academy of Sciences of the United States of America 104, no. 35 (Aug 28, 2007): 13868-13874.

Berlin, Brent. 2006. “The First Congress of Ethnozoological Nomenclature.” The Journal of the Royal Anthropological Institute 12 (2006): S23-S44.

-----. 1992. Ethnobiological Classification: Principles of Categorization of Plants and Animals in Traditional Societies. Princeton: Princeton University Press, 1992.

Chiba Mikio. 1995. Zenkoku yōkai jiten. Shōgakukan.

Figal, Gerald. 1999. Civilization and Monsters: Spirits of Modernity in Meiji Japan. Durham: Duke University Press.

Foster, Michael Dylan. 2015. The Book of Yōkai: Mysterious Creatures of Japanese Folklore. Oakland: University of California Press.

-----. 2009a. Pandemonium and Parade: Japanese Monsters and the Culture of Yōkai. Berkeley: University of California Press.

-----. 2009b “Haunted Travelogue: Hometowns, Ghost Towns, and Memories of War.” Mechademia 4: War/Time (2009): 164-181.

-----. 2008. “The Otherworlds of Mizuki Shigeru.” Mechademia 3: Limits of the Human (2008): 8-28.

Fujisawa Morihiko. 1929. Yōkai gadan zenshū, Nihon hen jō. Chūō bijutsu sha.

Gaiman, Neil. 2002. American Gods. New York: HarperTorch.

Gal, Susan. 2013. “Tastes of Talk: Qualia and the moral flavor of signs.” Anthropological Theory 13 (1/2): 31-48.

Gal, Susan, and Judith T. Irvine. 1995. “The boundaries of languages and disciplines: How ideologies construct difference.” Social Research 62, No. 4 (Winter): 967-1001.

Georges, Robert A. and Michael Owen Jones. 1995. Folkloristics: An Introduction. Bloomington: Indiana University Press.

Hanley, Susan B. 1987. “Urban Sanitation in Preindustrial Japan.” Journal of Interdisciplinary History 18, No. 1: 1-26.

Haslam, Nick O. 1998. “Natural Kinds, Human Kinds, and Essentialism.” Social Research 65, No. 2 (Summer): 291-314.

Iguchi, Takashi. 2002. “Depopulation and Mura-Okoshi (Village Revival).” In Iwai Yoshiya ed., Forestry and the Forest Industry in Japan, 259-277. Vancouver: UBC Press, 2002.

Inada Atsunobu and Tanaka Naohi, ed. 1992. Toriyama Sekien gazu hyakkiyagyō. Kokusho kankōkai.

Ishikawa Jun’ichirō. 1985. Shinpan kappa no sekai. Jiji tsūshinsha.

Itō Ryōhei. 2011. “Yōkai no hakubutsushi.” In Komatsu Kazuhiko ed., Yōkaigaku no kiso chishiki, 77-107. Kadokawa gakugei shuppan.

Ivy, Marilyn. 1995. Discourses of the Vanishing: Modernity, Phantasm, Japan. Chicago: University of Chicago Press.

Iwai Hiromi. 1986. Kurashi no naka no yōkai. Bunka shuppankyoku.

-----. 2004. Yōkai to ema to shichi fukujin. Seishun shuppansha.

Kagawa Masanobu. 2012. “Kappa imeeji no hensen.” In Kokuritsu rekishi minzoku hakubutsukan ed., Kappa to wa nanika: Dai 84-kai rekihaku fōramu, 6-9. Sakura-shi, Chiba: Kokuritsu rekishi minzoku hakubutsukan.

Keene, Donald trans. 1967. Essays in Idleness: The Tsurezuregusa of Kenkō. New York: Columbia University Press.

Knight, John. 1994. “Rural Revitalization in Japan: Spirit of the Village and Taste of the Country,” Asian Survey, Vol. 34, No 7 (July): 634-646

Komatsu Kazuhiko. 1994. Yōkaigaku shinkō: Yōkai kara miru Nihonjin no kokoro. Shōgakukan.

Lakoff, George. 1987. Women, Fire and Dangerous Things. Chicago: University of Chicago Press.

Laughlin, Charles D. 1993. “Fuzziness and Phenomenology in Ethnological Research: Insights from Fuzzy Set Theory.” Journal of Anthropological Research 49, No 1: 17-37.

Medin, Douglas L., and Scott Atran, eds. 1999. Folkbiology. MIT Press.

Mizuki Shigeru. 1997. Nonnonbaa to ore. Chikuma bunkō.

-----. 1994. Yōkai gadan. Iwanami shoten.

-----. 1984. Zoku yōkai jiten. Tōkyōdō shuppan.

Murakami Kenji. 2000. Yōkai jiten. Mainichi shinbunsha.

Nakamura Teiri. 1996. Kappa no Nihonshi. Nihon Editāsukūru shuppanbu.

Nozawa, Shunsuke. 2013. “Characterization,” Semiotic Review Issue 3: Open Issue 2013; – November 2013

Occhi, Debra J. 2010. “Consuming Kyara ‘Characters’: Anthropomorphization and Marketing in Contemporary Japan.” Comparative Culture 15: 77-86.

-----. 2012. “Wobbly Aesthetics, Performance, and Message: Comparing Japanese Kyara with Their Anthropomorphic Forebears.” Asian Ethnology 71 (1): 109-32.

Peirce, Charles S. 1966. Selected Writings (Values in a Universe of Change). Edited with introduction by Philip P. Weiner. New York: Dover Publications.

Ritvo, Harriet. 1997. The Platypus and the Mermaid and Other Figments of the Classifying Imagination. Cambridge: Harvard University Press.

Robertson, Jennifer. 1991. Native and Newcomer: Making and Remaking a Japanese City. Berkeley: University of California Press.

Rosch, Eleanor. 1975. “Cognitive Reference Points.” Cognitive Psychology 7: 532-547.

Rosch, Eleanor and Carolyn B. Mervis. 1981. “Categorization of Natural Objects.” Annual Review of Psychology 32: 89-113.

Rowling, J.K. (as Newt Scamander), Fantastic Beasts & Where to Find Them. Scholastic Press, Arthur A. Levine Books, 2001.

-----. 1999. Harry Potter and the Prisoner of Azkaban. Scholastic Press.

Saitō Tsugio. 1996. Yōkai toshi keikaku ron: Obake kara no machi zukuri. Sairyūsha.

Sillitoe, Paul. 2002. “Contested Knowledge, Contingent Classification: Animals in the Highlands of Papua New Guinea." American Anthropologist, New Series 104, No 4 (December): 1162-1171.

Tada Katsumi, 2001. “Etoki gazu hyakkiyagyō no yōkai,” Kai 12 (December): 320-321.

Takakuwa Fumiko, “Shimo-Koshiki no minzoku: Shinkō densetsu o chūshin ni,” Shimo-Koshiki-son kyōdo-shi hensan iinkai ed, 985-1007. Shimo-Koshiki-son kyōdo-shi. Shimo-Koshi-son: Shimo-Koshiki-son kyōdo-shi hensan iinkai, 2004.

Terajima Ryōan. Wakan-sansaizue 6. Heibonsha, 1994.

von Sydow, Carl Wilhelm. 1948. “Geography and Folk-Tale Oicotypes.” In Laurits Bødker ed., Selected Papers on Folklore, 44-59. Copenhagen.

Wada Hiroshi. 2005. Kappa denshō daijiten. Iwata shoin.

Yamada Norio. 1974. “Nihon no yōkai, henge, majintachi.” In Yamamuro Shizuka ed., Yōkai, majin, shinsei no sekai. Jiyū kokuminsha.

Yamada Shōji. 2010. “‘Yurui’ to ‘katai’ no aida: Nihon no ‘yurukyara’ masukotto o kangaeru.” In Understanding Contemporary Japan, International Symposium in Indonesia, 157–166. Kyoto: International Research Center for Japanese Studies.

Yasui Manami. 1997. “Machi zukuri, mura okoshi to furusato monogatari.” In Komatsu Kazuhiko ed., Matsuri to ibento, 201-226. Shōgakukan.


1. For more on the encyclopedic mode, see Foster 2009a: 31-48.

2. For more on Mizuki Shigeru, his yōkai, and his construction of nostalgia, see Foster 2008; 2009b.

3. The tenjōname appears in Sekien’s Hyakki tsurezure bukuro of 1784.

4. Murakami Kenji (2000:235) suggests that the Tenjōname is a Sekien invention that plays with the idea that a ceiling is a boundary between two worlds.

5. The name of Sekien’s catalog itself, Hyakki tsurezure bukuro, also references this same work with the word tsurezure. For Sekien’s works, see Inada and Tanaka 1992.

6. The history of the kappa and related water creatures in Japan is tremendously complex and I stress again that my analysis is highly simplified for the purposes of this essay. For a comprehensive examination of the kappa throughout Japanese history, see Nakamura 1996. For a presentation of the overwhelming variety of kappa-related creatures and beliefs, see the 763 pages of Wada 2005, a dictionary of “kappa traditions.”

7. For an image of the Nihon sankai meibutsu zue, see Nihon sankai meib­utsu zue, Hirose bunkobon 3, Kyushu University Library website, n.d., (accessed September 10, 2014).

8. On prototypes, fuzzy sets, family resemblances and other issues in classification and taxonomy see, for example, Rosch 1975, Lakoff 1987, Haslam 1998. On folk biology and ethnobiology, see Berlin 1992, Medin and Atran 1999, Sillitoe 2002.

9. I hesitate to use “teratology” here to indicate the ordering of yōkai because of the roots of teratology for describing birth defects and other so-called abnormalities. In the case of yōkai at least, the creatures in question may be different from humans and other animals, but are not necessarily defective or abnormal; they rarely fit within the rubric of so-called monstrous births.

10. The corpus of yōkai is rife with oicotypes (or oikotypes), a term adopted appropriately “from botany to identify subsets of types that are delimited geographically, culturally or linguistically” (Georges and Jones 1995:150). The word was introduced to folkloristics by Swedish folklorist Carl Wilhelm von Sydow (see von Sydow 1948).

11. For more on Inoue Enryō and his relationship with modernity, see Figal 1999; Foster 2009.

12. Although I have been unable to find a “gamishiro” anywhere else, the similar-sounding “gameshirō” is identified as one of many variant names for kappa-like creatures in the Kagoshima area (Wada 2005:696). In fact Takakuwa (2004:1007) notes gameshirō as the name of the creature in question on the island, but does not mention gamishiro. The game part of gameshirō is a name for a kappa-like yōkai found in various different regions and may imply an association with turtles, or kame. For game, see Murakami 2000:117.

13. Interviews with author, February and March 2012.

14. It is possible that the gami in gamishiro is related to the word kami, the general Japanese word for deity.

15. Such mura okoshi and machi zukuri (town-making) movements began during the 1960s in response to the depopulation of rural communities throughout Japan and were often, not surprisingly, fraught with political and economic complexities. There is a vast literature on these issues. See for example, Robertson 1991, Ivy 1995; Knight 1994, Iguchi 2002, Yasui 1997. Saitō 1996 focuses specifically on the use of yōkai in these processes.

16. Yuru-kyara mascot figures, often developed to represent a particular town or region, are exceedingly popular in Japan currently. See Occhi 2010; 2012; Yamada 2010; Nozawa 2013. Similar to yōkai, they feature characteristics unique to the location with which they are associated. The most successful (i.e., popularly known and commercially profitable) yuru-kyara are those with relatively loose copyright restrictions so that they can be reproduced or modified for different purposes with relative freedom.

17. A number of popular fictional works, such as Takahata Isao’s anime Heisei tanuki gassen pompoko (1994) and even Neil Gaiman’s 2001 novel American Gods (Gaiman 2002), reflect a sense that there is an increasingly limited space for monsters and gods in the human imagination. Inversely, Azuma Hiroki’s theory of “database consumption” suggests that the very process of classification—of extracting and organizing “settings” and motifs—provides a database for the creative assembling and proliferation of new and heterogeneous monsters, gods, and other characters (see Azuma 2009). I thank Christopher Bolton for suggesting the connections here.