The heterogeneity of maid cafés: Exploring object-oriented fandom in Japan

Luke Sharp

University of New South Wales, Sydney, New South Wales, Australia

[0.1] Abstract—Maid cafés may be positioned as objects of (often oversimplified) fan communities in Japan; their heterogeneous qualities may be explored by presenting chronotopes with which they may be reconsidered. The problematics of viewing maid cafés in a homogeneous or reductive way (via iyashi-kei, "relaxation purposes," and moe-kei, "entertainment purposes") are considered, and the results of empirical research conducted at establishments across Honshū in Japan are presented. The vast differences in the services and interactivities provided in maid café settings are at odds with the homogenized fashion in which these venues have been treated in the popular press, in local promotional publications, and often in the academic literature. By presenting chronotopes and case studies with which maid cafés may be reconsidered, what it means to discuss maid cafés as a holistic phenomenon is challenged, and a heuristic for examining how spaces are constructed inside such establishments is provided.

[0.2] Keywords—Chronotope; Iyashi; Moe; Space

Sharp, Luke. 2014. "The Heterogeneity of Maid Cafés: Exploring Object-Oriented Fandom in Japan." In "Material Fan Culture," edited by Bob Rehak, special issue, Transformative Works and Cultures, no. 16.

1. Introduction

[1.1] The maid café may be treated as an object of fandom in Japan. In establishments throughout the country, patrons interact and converse with costumed waitresses, creating a vast array of experiences. Maid cafés are objects of affirmation and tribute to the Japanese maid persona; the variations of interactive services that they provide necessitate reconsideration of the term maid café holistically (note 1). After an initial discussion of the homogenizing of maid cafés and the methodology used for this study, I offer five chronotopes of maid cafés to highlight their variety and to demonstrate how they are not uniform. The chronotopes may be used as a heuristic for the consideration and evaluation of the communicative elements of maid cafés, particularly with how spaces are constructed inside them (note 2). This is useful for the consideration of, among many other potential space-related topics, how space is gendered inside maid cafés, how power relations between maids and customers are forged vis-à-vis space, and how public and private spaces converge in maid café settings.

2. The heterogeneity of maid cafés

[2.1] A maid café is a small establishment where waitresses dressed in some form of a maid costume engage customers (usually, but not exclusively, men) in conversation and provide entertainment, perhaps musically in the form of karaoke or through playing board or video games with patrons. Certainly this is how maid cafés are typically presented in the mass media. Central to these interactive elements is the role-play of the master/servant. Maids often greet customers as they enter an establishment with the phrase, "Welcome home, Master" (okaerinasaimase goshujin-sama), and on initial contact, they use formal language. Maid cafés have only been around since 2001, but they are best understood as part of a longer history of establishments in Japan's eroticism industry, to use Silverberg's (2007) term. Antecedents of maid cafés include the teahouses of the Tokugawa period (1603–1867) pleasure quarters, the cafés of the 1930s' "erotic grotesque nonsense" movement, and the hostess bars and clubs of the era after World War II. Sexual exchanges were not a primary part of hostesses' jobs in these early iterations, but they did happen. In contrast, in maid cafés, they uniformly do not. However, like these early establishments, they commodify elements of femininity, sexuality, and conversation (Galbraith 2011; Louis 1992).

[2.2] However, there remain large differences in the services and activities provided in each maid café in Japan, as well as in the forms of role-play that are present. To refer to maid cafés as if they were a single entity is thus inaccurate, despite the prevailing inclination to regard them in a homogenized fashion (note 3). It is important to reflect on this heterogeneity in relation to maid café fandom in Japan because it can potentially offer a new paradigm for rethinking the phenomenon holistically as well as offer new ways of analyzing how spaces are constructed.

3. Study design

[3.1] This study used ethnography as its methodology. It was structured to use the O'Reilly's (2009) "key concepts in ethnography." The fieldwork for this study involved several considerations of these key concepts, such as gaining access, establishing an insider role, building rapport, assessing both emic and etic perspectives, and avoiding "going native" (O'Reilly 2009, 3). Becoming a café regular was a risk, so the balance of "distance and empathy, insider and stranger" (O'Reilly 2009, 89) was imperative. The methods used were participating, asking questions, observing, collecting documents, and taking photographs. I spent May 2009 to August 2009 and September 2010 in the field in maid cafés in Japan. I was thus able to quantitatively add to the data by visiting cafés that had opened over that 12-month period.

[3.2] Although a comprehensive discussion of ethnography as a methodology is beyond the scope of this article, it is important to briefly address some of its important considerations vis-à-vis this study. One significant issue is how "proper ethnography" (Delamont 2004, 219) faithful to classic anthropological models differs from ethnography as an intermittent method of data collection (Aktinson and Hammersley 1998, 110). Wolcott highlights the distinction between the two as "ethnography as product" (that is, "doing ethnography" devoted to a set of principles within the discipline of anthropology) and "ethnography as process" (that is, "borrowing (some) ethnographic techniques" where researchers adapt fieldwork procedures for gathering data) (2008, 44). In this process, participant observation forms the crux of ethnography (Esterberg 2002). However, fieldwork has also come to incorporate the use of other research methods, including conversations and interviews, analysis of textual materials, and interpretation of visual sources such as photography, film, and video (Atkinson et al. 2001). I made use of all these modes in my methods.

[3.3] Given the conceptual contentions outlined above, and given that there are no prescriptive specifications for what ethnography is or how ethnographic research is conducted, it is not possible to neatly classify ethnography for this study as either product or process, to use Wolcott's (2008) terms. Although I used ethnographic techniques such as engaging in participant observation and analyzing photos, I also interacted with people in their everyday contexts and attempted to gain an insider's perspective of maid cafés and their customers. Hammersley and Atkinson (1995) acknowledge this fluidity of what ethnography is and can be, noting that positions exist between the polar extremes of ethnography for data collection and ethnography as paradigm.

[3.4] For this study, maid cafés in Honshū, the largest island in the Japanese archipelago, were sampled. Honshū was selected as the focal point of the study for reasons of accessibility and mobility, but more significantly, it offered the widest range of diverse locations in Japan. Although maid cafés exist on the other main islands of Japan—Kyūshū and Hokkaidō have several—these areas are limited in their size and population. Being situated on Honshū enabled me to access the largest urban centers and most highly populated cities in the whole of Japan (Tokyo, Yokohama, Osaka, and Nagoya), areas with medium-size populations (Sendai and Hiroshima), and much smaller prefectural capitals (Okayama, Nagano, and Wakayama). Honshū thus provided the broadest range of localities possible. I did not sample maid cafés; instead, I successfully visited every one in operation there from May 1, 2009, to July 31, 2009, and September 2010–73 in total.

4. Selecting maid cafés for study

[4.1] During my fieldwork, finding maid cafés was often problematic because no single protocol exists to describe them. Local publications and Internet listings subsume maid cafés with other maid-themed establishments, such as maid hair salons and maid reflexology clinics, as well as with other cosplay eateries. A number of venues market themselves as maid bars, maid izakaya (pubs), and maid kyabakura (hostess clubs) (note 4), yet all often fall under the umbrella term of "dining" or "food and beverage" in many local print media. The line between maid café and other establishments is thus often obscure as a result of this lack of explicit compartmentalization in local publications.

[4.2] I selected two factors to separate a maid bar and maid kyabakura from a maid café and a maid café from any other type of cosplay café: hours of operation (open from midmorning until late evening, unlike a maid bar or maid kyabakura, which open exclusively in early evening and often operate all night until early morning) and overall mode of dress of waitresses. For this study, a maid outfit must be the primary costume worn by the waitresses for the largest percentage of time an establishment is open; this was measured by the number of events days scheduled by a venue (note 5). These complexities in demarcating maid cafés indicate the wider problematics in their homogenization as a category or label.

5. Moe versus iyashi

[5.1] The propensity to view maid cafés in a homogenized fashion is found in both Japanese- and English-language media commentaries, in local promotional literature, and in the small body of academic literature dealing with the topic. On the occasions when they are referred to heterogeneously, however, they are placed into one of two categories: venues intended for relaxation (iyashi-kei, "healing type") and those intended to entertain (moe-kei, "moe type," or in English-language literature, simply "entertainment-kei") (note 6). Many establishments embrace these dichotomous and essentialized terms by self-labeling in their promotion.

[5.2] Although Galbraith believes "soothing" to be an aspect of leisure that otaku (often translated as "fans," "aficionados," or "enthusiasts" of manga and anime) particularly hold in high regard (2009, 108), the allure of repose as a marketing technique is not a phenomenon unique to maid cafés. As Allison points out in her discussion on toys for children in Japan, healing and soothing "are perpetual tropes in the marketplace of play goods these days" and are "said to be a relief from the stresses caused by consumer capitalism" (2006, 14). Maid cafés thus provide escapism, and they draw on this in their marketing as iyashi-kei. They both appeal to and capitalize on customers' emotional responses (Galbraith 2011). Allison refers to this as an encoded form of familiarity, consumerism, and technosocial interaction, which shape the fundamental construction of Japanese play (2006). In Western fan discourse, Jenkins (2006) describes something similar in affective economics, a marketing theory that seeks to understand emotions in the process of consumer decision making.

[5.3] The use of the alternative marker moe is restricted primarily to media commentaries and online reviews in fan blogs. For example, two free maps placed at the train station and local landmarks in Akihabara highlight the locations of cafés, listing these establishments under the heading moe-kanren (moe-related cafés). Some cafés themselves likewise suggest this association by incorporating the term moe into the name of the actual venue, such as Moekon@cafe (Akihabara) and Moe & Shandon (Osaka). The agenda for entertainment at the latter establishment is evident in its weekend "moe time" sessions, when live karaoke performances by the maids take place and the cover charge increases. Some establishments do not view moe and iyashi in absolute terms in their marketing strategy; for example, the flyer for Ichigo Miruku (Shibuya) states that it "offers to customers both moe and iyashi." By suggesting it is able to please everybody, regardless of their moe or iyashi orientations, this flyer effectively encapsulates the polar extremes of these two agendas and accommodates otaku more broadly.

[5.4] The terms iyashi-kei and moe-kei may be reasonable starting points to think of maid cafés in terms of functionality, but they are problematic. Because there is no fixed contextual definition of either iyashi or moe, the line between the two is subjective; the terms are arbitrarily applied by venues for marketing purposes. On the surface, the image portrayed by both the media and cafés themselves is that iyashi-kei venues strive for a tranquil ambience where interactive elements (often the catalysts for conversation) are minimal. Moe-kei venues, on the other hand, are depicted as highly communicative and interactive, loud, and colorful, with "a syrupy sweet atmosphere" (Galbraith 2009, 137). However, venues with an agenda of maid-customer interactivity often tout themselves as iyashi-kei. All branches of Pinafore (Akihabara), for example, have several interactive elements (based on conversing) available à la carte. They suggest an iyashi tendency on their menus by stating, "This establishment offers a space for healing" (iyashi no kūkan wo teikyōsuru kissaten). The perception of what is soothing and what heals the mind is not the tranquility of the café environment resulting from a lack of interactive service but rather the chance to converse and engage in communicative activities as if they were a form of therapy. These differences reflect a wide gap between the conceptions of what iyashi and moe processes entail among establishments.

6. Maid café chronotopes

[6.1] The difference between conceptions of healing and entertainment practices makes categorizing maid cafés as either iyashi-kei or moe-kei ineffective, especially in examining the spatialities of their settings. I thus suggest five chronotopes for maid cafés that illustrate the contrasts in major services and interactive elements. This categorization scheme challenges how the term maid café may be used holistically and permits reconsideration of the communicative spaces of fans and customers. I use the term interactive elements to refer to services that involve significant levels of interaction between maids and customers, such as cheki (Polaroid photographs taken with maids), rakugaki (art graffiti where maids draw images and words on served food with condiments such as ketchup), and notebook exchanges (where customers exchange messages and artwork in notebooks for the maids to read).

[6.2] In choosing the term chronotope, I have drawn on the work of Kamberelis and Dimitriadis (2005), who reconsider traditional qualitative research approaches by offering more fluid and variegated paradigms in which qualitative studies can be located. Their work developed from a need to rethink the fixed definitions associated with qualitative inquiry and its predetermined, concretized guidelines. Kamberelis and Dimitriadis challenge this type of epistemological demarcation with their development of the chronotope, a construct for reconsidering qualitative research. This includes assumptions about the world, knowledge, and the set of methods used for conducting research. Maid café chronotopes are thus constructs for the consideration of how spaces are created in their settings. In a similar vein to the caveat issued by Kamberelis and Dimitriadis, it is important to recognize that these maid café chronotopes ought not be viewed as fixed and static. Because they have been devised to correspond with the results vis-à-vis interactive elements and conversation, different chronotopes would be in order for the consideration of, for example, the interior dimensions of maid cafés or the uniforms that maids wear.

[6.3] It is inevitable that there will be some overlap with these chronotopes, so I recommend, as Kamberelis and Dimitriadis (2005) do with their constructs, that they be viewed as a heuristic for the consideration and evaluation of maid café interactivities and conversation—that is, what kinds of spaces are constructed—rather than as a list of prescriptive components that comprise maid cafés. This could entail elements such as how space is gendered inside maid cafés, how power relations between maids and customers are forged vis-à-vis space, and how public and private spaces converge in maid café settings. I identify the five chronotopes in table 1.

Table 1. Five chronotopes of maid cafés

Chronotope Description
A No or minimal interactive elements with little or no conversation between maid and customer.
B Low to medium level of interactive elements with moderate volume conversation between maid and customer.
C Venues where conversation is the main objective with a considerable level of interactive elements.
D Venues where conversation is the main objective with a strong emphasis on interactive elements.
E Theaterlike venues with an extreme level of interactive elements but with little or no conversation.

[6.4] With these chronotopes, I have disaggregated the terms interactive element and conversation because the relationship between services of an interactive nature in the mise-en-scène of the maid café and the conversations generated between maids and customers is multifaceted, and the two phenomena are not always linked. There are occasions when interactive elements and conversation exist independently of each other. For example, chronotope A cafés provide few interactive services (rakugaki, cheki, games) on their menus, and conversation between maids and customers, beyond ordering food and settling payment, is limited. In contrast, chronotope E cafés, which tend to be well known because they cater to tourists and receive media coverage, provide an overwhelming number of interactive services. Despite this, little conversation between maids and customers exists, as with chronotope A cafés (which generally have no interactive elements of this sort). This may be attributed to the type of clientele they receive (one-time customers) and the exceedingly high turnover of customers during opening hours (the maid to customer ratio is low during these times). In contrast, chronotope C cafés, which constitute the largest number of venues, specialize in the art of repartee while providing a considerable number of interactive elements. However, conversation between maids and customers is not always contingent on these elements; chat is generated as a standard service. This differs from chronotope D cafés, which use interactive elements as catalysts for conversation: these services form the foundation for chat, and the conversational content usually revolves around these services. This is reflected by the set menus (that is, fixed-price menus) that many of these cafés provide. Chronotope B cafés, in contrast, comprise venues that are on the quieter and less active end of the communicative spectrum; neither chat nor interactive elements are found in these establishments in any significant way.

[6.5] The manner in which conversation is generated, if indeed it is generated at all, is not uniform in maid cafés. The act of initiating conversation may or may not depend on interactive elements, and the degree to which these elements forge the content of the chat also varies. The differences in these communicative spectrums is illustrated in figure 1.

Two line graphs. The top one is titled 'Communication spectrum vis-a-vis services provided.' A line with 5 hatches dividing it is underneath. Atop the line read 'Marginal,' 'Considerate,' and 'Extreme,' spaced left, center, and right. Underneath the line, matching the 5 hatches, are labeled 'CHRONOTOPE A,' 'CHRONOTOPE B,' CHRONOTOPE C,' 'CHRONOTOPE D,' and 'CHRONOTOPE E.' The bottom line graph is titled 'Communicative spectrum vis-a-vis conversation.' A line with 3 hatches dividing it is underneath. Atop the line read 'Low Volume,' 'Moderate Volume,' and 'High Volume,' spaced left, center, and right. Underneath the line, matching the 3 hatches, are labeled 'CHRONOTOPES A&E,' 'CHRONOTOPE C,' and 'CHRONOTOPES B&C.'

Figure 1. Communicative spectrums of interactive elements and conversation. [View larger image.]

7. Chronotope A

[7.1] Although the term iyashi-kei implies that little interaction will occur, this label is only partially correct. It is true that participatory services are not available in some establishments, but in others, communicative interactivity may be present, even if at only marginal levels. A hybrid form of these two degrees constitutes the cafés of chronotope A, which accounts for the second smallest number of the five chronotopes identified, at 12.3 percent of all cafés. These venues provide no regular cheki or other photographic services; they sell a limited range of maid goods or no goods at all; and they do not offer games or notebook exchanges. Although food art created in front of diners is not unknown, predrawn rakugaki is the dominant feature. This practice appears to be the only element in chronotope A cafés that involves a perceptible communicative exchange beyond greetings and ordering, and with this exception, these establishments function like most noncosplay restaurants with table service.

[7.2] Correspondingly, because the food at these establishments is the main draw (unlike the appeal of chronotopes C, D, and E, where communicative experiences are the predominant attraction), the meals served in chronotope A cafés are of much higher quality than other venues, with diverse menus that may include fish and other seafood, deep-fried foods, gourmet pizzas, and various red-meat dishes. A vast array of alcoholic beverages are also available: Granvania (Akihabara) prides itself on having 30 European beers from more than twelve countries, Milkcafe (Osaka) proudly promotes its 300-plus types of spirits, and Hiyokoya (Akihabara) offers wines that cost more than ¥8,000 per bottle. This emphasis on the dining experience rather than the communicative experience can be seen in their marketing: the flyer for Granvania provides pictures and descriptions of their dishes, and some cafés do not mention the term moe and contain no images of maids.

[7.3] Despite the low levels of interactivity at chronotope A cafés, these venues sponsor event days. Chronotopes C, D, and E tend to advertise event days with great fanfare, but for chronotope A, the "event" may be little more than a day when an alternative menu is offered, with added bonuses such as an amnesty on expired reward cards or free-drink tickets. Event days at chronotope A cafés blur the lines between them and other venue types because as interactivity is heightened. Emaid (Osaka), for example, offers a cheki service only on event days, and M's Melody (Nagoya) holds a biannual satsueikai (photographic) event. Sometimes events occur frequently: Cure Maid Cafe (Akihabara) holds a concert each weekend where the maids perform on musical instruments. Some event days are also connected to the promotion of specific enterprises, which usually sponsor the event. This could be the launch of a new video game or computer software, where promotional products are given out to customers. Events may also simply be a change of usual costume to celebrate or correspond to an annual festival such as Valentine's Day, Halloween, or Christmas.

[7.4] There is a possibility that the focus of chronotope A cafés not being on communicative experiences is connected to another significant feature of these venues: the high number of female customers. Emaid (Osaka) and Wonder Parlour Cafe (Ikebukuro) in particular have many women patrons, with female customers actually outnumbering male customers. I observed women in chronotope A cafés in groups of two to four, though women dining alone were also present. In my discussions with fellow customers at other cafés, they suggested that the women who patronize chronotope A venues are avid cosplay fans motivated by creative inspiration: they visit to observe and admire the maids' costumes, seeking ideas for their next cosplay endeavor. My own observations and encounters with female customers support this interpretation, or at least its possibility. In Osaka, I witnessed young women (usually in pairs) dressed in some cosplay variant, and while visiting more than one café on the same day, I saw the same women at these different venues. Although some were in full costume, with wigs and makeup, others were only wearing cosplay accessories, such as mini top hats with veils. Similarly, at M's Melody in Nagoya, I was approached by a young woman who left her table to speak with me. She announced she was a "cosplay otaku" and proceeded to show me a collection of photographs of her in full costume. These observations suggest that for female customers, communicative exchanges are secondary reasons to visit cafés, and thus chronotope A cafés are a good choice for them. An esoteric appreciation of the surroundings is at the forefront of their motivation.

[7.5] Milkcafe, located in the Nipponbashi area of Osaka, is one example of a chronotope A maid café. From the outset, Milkcafe leaves no ambiguity about its intentions or function as an establishment. The first page of their menu contains an extensive list of expectations. The first three items state:

[7.6] 1) This establishment is a restaurant. Beyond offering food and drinks, we provide no other kinds of services.

2) Customers should be aware that the staff at this venue are not paid company—they are waitresses.

3) If your purpose is to find a date, we recommend that you spend your money at some other café that is that way inclined. (note 7)

[7.7] Although other cafés may not be so direct in their customer expectations, the explicitness of this text is indicative of the feel that chronotope A cafés aim to achieve: a setting where only minimal interaction exists between maids and customers, and where passivity in the experience of dining is the dominant objective. Pearson's (1998) reflection on performance relationships is apt here. The customer remains a constant passive spectator of the surroundings of the café; there is little opportunity for the role of actor to be taken. The declaration that its only concern is food and drink reinforces this, leaving aside the fact that the possibility of conversation with staff is remote. Similarly, branding the maids as waitresses, thus distinguishing them from paid company, circumscribes their role as servers; they are not generators of conversation. The third expectation elucidates the difference between Milkcafe and other establishments such as cabaret clubs.

[7.8] The focus on food, not interactivity, is further reinforced by the emphasis on the quality and range of products available to customers. Milkcafe's Web site clearly establishes it as an independent establishment not reliant on interactive elements. For example, during their Ai no Epuron (Love's Apron) ¥500 lunch promotion in September 2011, Milkcafe stated, "Milkcafe has no entry fee, table charge, service tariff or the like. We also have no tacky theatrical moe service or anything of the sort. Ordinary maid cafés are expensive and the food tastes awful, but Milkcafe is cheap and the food tastes great" (note 8). This differentiation between the ordinary maid café and Milkcafe, with an emphasis on the quality of food and the rejection of a moe-oriented service, indicates that the venue does not seek to provide interactivity between maids and customers. This typifies chronotope A cafés.

8. Chronotope B

[8.1] Chronotope B cafés provide a moderate volume of conversation on the communicative spectrum and provide between marginal and considerable volumes of interactive elements. Interactive elements and conversation in these establishments are best described as both fleeting and noncommittal: the communicative exchanges that do exist are not bound to any set length of time (unlike chronotopes C, D, and E, which have fixed time limits for certain services), and because the maids are not obligated to engage customers in such a manner, these exchanges are brief. In this sense, conversation initiated by maids between customers is arbitrary. If there are few patrons, she may spend more time chatting than if the venue were full (in which case she may not chat at all), and activities such as delivering food and greeting new customers take priority over conversation. Chronotope B cafés form less than a quarter of all establishments, at 19.2 percent (14 of 73 cafés). These venues typically provide no cheki or other type of photographic service (including satsueikai events), sell only a small range of maid goods, and do not provide games or notebook exchange services. Food art is perhaps the most interactive of the services provided. Consequently, rakugaki can act as the conduit for the limited conversation that is initiated. As with chronotope A cafés, chronotope B cafés rely on food art as the only element that initiates communicative exchanges between maids and customers, apart from exchanging greetings and ordering food.

[8.2] Although interactive elements are minimal and conversation between maids and customers is arbitrary at chronotope B cafés, event days are scheduled with a relative degree of frequency. Like chronotope A cafés, these events blur the line between them and cafés with an agenda for conversation. Event days seek to provide an occasion for more communicative elements than are usually present. Promotion of these events exists on both the Web sites of the cafés and at the physical venues via leaflets, posters, and announcements on bulletin boards.

[8.3] Fairy Tale, a café located in the Aoba-ku ward of Sendai in Miyagi prefecture, is an example of a chronotope B café. Chronotope B cafés characteristically have few interactive elements (many market themselves as having an iyashi agenda), but they simultaneously engage customers with a limited degree of conversation that is not time based. This is exemplified by the mission statement of Fairy Tale. Its Web site states that it aims to re-create a "classic image of a maid café" and offers "a place to unwind when feeling tired." However, highlighting the liminal space they occupy between relaxation and conversation, the statement continues with a description of what a fairy tale connotes in English: "The term fairy tale corresponds to dōwa or otogibanashi in Japanese, conveying a sense of nostalgia and amusement from within to those who hear it. In this manner, our café offers tranquility yet also friendliness" (note 9).

[8.4] Friendliness in this sense is best understood to be the act of conversing between maids and customers, while tranquility, as with chronotope A cafés, alludes to minimal interactivity. The initiative of either rests with the customer. Maids, if prompted, will converse with customers in the fashion of chronotope C and D cafés, although because it does not explicitly provide a time-based service, priority is given to delivering food and greeting new customers. However, if the customer chooses not to chat, either by failing to initiate conversation or failing to respond to maids' prompts, then communication between them advances no further; it is understood that they have taken the tranquil path. This arbitrary nature of conversation and liminal space that exists between it and interactive elements typify chronotope B cafés.

9. Chronotope C

[9.1] Chronotope C cafés constitute the largest number of venues, representing just under half of all establishments, 46.5 percent. These cafés, like chronotope D cafés, are high volume in terms of conversation on the communicative spectrum, with a considerable number of interactive elements. The main objective for these venues is to provide customers with an opportunity to converse with maids, for which interactive elements generally—though not always—act as a conduit. This is the differentiating characteristic between chronotope C and chronotope D cafés, where the interactive services provided consistently make the environment amenable to chat.

[9.2] Most chronotope C cafés provide a cheki service and regularly hold events. They place an emphasis on food art drawn in front of customers (as opposed to predrawn rakugaki) and sell a wide range of maid goods, both in the store and online. Although the moe incantation (chants that the maids say and customers repeat to magically make food more delicious) is an element found relatively widely, it is not common to all cafés in this category. The same is true of notebook exchanges. Games also play a part in the mise-en-scène of chronotope C cafés.

[9.3] Chronotope C cafés have a varied quality of food. Some establishments follow the gastronomical agenda of chronotope A cafés (and to a small extent chronotope B cafés), though generally food is more similar to the chronotope D and E cafés: of low quality and limited to a few main dishes of pasta, curries, and fried-rice omelette (omuraisu). Although blogs describing chronotope C cafés focus largely on the interactive elements of the venues, such as event days, there are also commentaries on the food, as café chronotopes A and B have.

[9.4] Ichigo Miruku, situated in the Shibuya ward of Tokyo, is an example of a chronotope C café. There is a cover charge of ¥400 per hour for male customers and ¥300 for female customers, with a reduced rate for students. This system is explained to customers as they enter, and any previous visits to maid cafés are confirmed at the same time. Ichigo Miruku promotes itself as a venue that that offers both moe and iyashi, and according to its Web site, it welcomes "grandpas and grandmas, mums and dads, boys and girls, couples, and foreigners." Indeed, this distinction between moe and iyashi seems reified by menus presented in two different folders to customers: one is the usual menu—the à la carte menu where individual items can be selected—and the other is the set menu. Although the presence of a set menu may seem more indicative of a chronotope D venue, where interactive elements have fixed time limits, in chronotope C, it is more representative of the level of engagement that customers can experience while conversing with the maids. Selecting a set menu usually guarantees the maximum time possible communicating with the maids. On top of the general chat she provides as part of the service at the café, customers who choose set menus are also afforded extra time spent playing games, taking cheki, and engaging in rakugaki activities. Conversely, customers who choose the usual menu items individually do not experience extended communication time, opting instead for the standard chat. The prerogative to maximize conversation thus lies with the customer. This typifies chronotope C cafés: conversation is not always contingent on interactive elements, and it is generated as part of the service, with food service being their main objective.

10. Chronotope D

[10.1] Chronotope D cafés form the third most common sort of maid cafés, at 13.7 percent (10 of 73 visited). They are venues that are high volume in terms of conversation on the communicative spectrum and between considerable and extreme in terms of interactive elements. The links between conservation and interactive elements in these establishments are strong. Unlike in chronotope C cafés, where conversation may not necessarily be contingent on the interactive services offered, in chronotope D cafés, conversation is founded on these interactive elements, the development of which is bound by a set length of time: 5-minute games, 10-minute à la carte photography, 30-second moe incantations. Chronotope D cafés almost always provide a cheki service and other forms of photographic opportunities; sell a wide range of maid goods; have a moe incantation; provide a variety of short- and long-duration games; and offer a notebook exchange service. Food art drawn in front of customers is also a common interactive element in chronotope D cafés, although predrawn rakugaki is not unknown. Because of the prevalence of these interactive elements, Pearson's (1998) performance relationships in chronotope D cafés are fluid and interchangeable. Through the construction of spaces enabled by the maid persona, the maids in the café initially begin as the active actors. However, customers' role as passive spectator alters once they begin interacting and become involved in the performance.

[10.2] Because the generation of conversation based on interactive elements is the dominant objective of chronotope D cafés, the food at many is not the point. Beyond drinks, some offer nothing more than small savory snacks such as potato chips and edamame. Those that do offer more are restricted to a few main dishes, such as curried rice, pasta, and omuraisu.

[10.3] Filles, located along Chūō dōri, in Akihabara, is an example of a chronotope D café. The café has a ¥1,800 cover charge known as a system, and the price includes one (nonalcoholic) drink and a short-duration game. The length of time for a stay is capped at 90 minutes. Food at Filles is an additional cost, and choice is extremely limited: a few savory snacks to accompany the drink, toast, sandwiches, cake. In short, the dining experience at Filles is supplemental to its communicative agenda. Customers do not come to eat but to interact and converse.

[10.4] Because the menu is so limited, food art does not exist at Filles. This is actually exceptional because all other cafés categorized as chronotope D have this interactive element. It does have a cheki service available for a fee, as in most chronotope C and E cafés. However, because the focus is on instantaneous communication, there is no notebook exchange service and, as expected, no collection of reading materials. The sale of in-store maid goods is also limited. Its shortfalls in terms of rakugaki and communication notes are made up with an interactive service rare among other cafés: the chance for female customers to dress up in full maid costume. At Filles, this is free of charge for 30 minutes and comes with the added bonus of the opportunity for the female customer to serve the food and drinks ordered by those she is dining with to them (unheard of at other cafés). This is marketed on Filles's flyers and Web site to male-female couples under the heading "Great news for couples!," suggesting to the women that it is an opportunity to know what it feels like to be a maid, body and soul (note 10).

[10.5] The maids immediately start chatting after explaining the system, and within moments after ordering, beverages arrive. Customers are given two poker chips, one white and one red, in a black folder containing the bill. When customers are ready to play a game with the maid (included in cover charge), they put the white chip on the table. For any games after that (at an additional cost), the red chip is put down. Additional games are capped at 15 minutes, but the initial game has no such time limit. Most other venues time all the games, and most do not usually last longer than 5 minutes. The maid informs customers that winners will receive a prize (usually some type of candy), but the consequence of losing is being made to do something embarrassing (hazukashii koto wo saseru). This usually entails the customer wearing a maid headband until his or her departure.

[10.6] The time spent conversing with maids at Filles is extensive, and by using the short games as a catalyst for chat, it typifies chronotope D cafés—that is, venues where the generation of conversation is always reliant on one or more interactive elements, which are usually part of a fixed system or set menu.

11. Chronotope E

[11.1] The communicative spectrum presented in figure 1 demonstrates that, despite there being links between interactive elements and the generation of conversation among maids and customers (café chronotopes C and D), the two can in fact be mutually exclusive. This is indicated by chronotope E cafés, where there is incongruence between the level of conversation (low volume) and the forms of participatory services available (extreme). Chronotope E cafés account for the smallest number of the five types identified, constituting 8.2 percent of all cafés visited. The draw for patrons visiting these venues is the emphasis on interactive elements and role-play in the fashion of theater restaurants. All provide cheki services (some also have regular satsueikai events), all sell a wide range of maid-related goods as souvenirs, and all have various moe incantations and a selection of games available. Because forging a regular stream of communication between maids and customers is not the primary objective of chronotope E cafés, the notebook exchange service is uncommon. Rakugaki is an especially noticeable (and marketable) feature of these venues, with food art images consistently appearing on their Web sites and other advertising material. Additionally, karaoke stage performances by maids are common—arguably what designates them as entertainment-kei venues. Because of this focus on musical performance and the corresponding cover charges, chronotope E cafés are among the most expensive to visit. At the opposite end of the continuum to chronotope A cafés in terms of interactive elements, in which the gastronomic experience forms the core purpose of the establishment, chronotope E cafés play on the novelty of what are popularly perceived as maid café fundamentals: the unfamiliar, unusual, and unique facets of stereotyped and exoticized maid cafés. Points of attraction include waitresses dressed in cosplay attire, intermittent musical performances, and spells cast on food and drink to make them taste better. These venues target and subsequently attract men, foreigners, male-female couples, mixed groups, women, and casual one-off visitors interested in gaining an ephemeral insight into maid cafés.

[11.2] Because chronotope E cafés exist to entertain, the food and beverages at these establishments are peripheral in terms of their variety and quality. This is clearly at opposite end of the spectrum as chronotope A cafés, which often go to great lengths to emphasize their high-good food. The culinary scenario that Galbraith (2009) describes of maid cafés serving a limited range of dishes that are staples of family restaurant children's menus—omuraisu, curried rice, and spaghetti—is truest for chronotope E cafés. It is possible that these dishes are the most amenable to the drawing of rakugaki, and that offering a more varied menu would diminish the opportunities for customers to partake in this particular interactive element. This is an especially marketable feature of chronotope E cafés, and the flyers for MaiDream, @Home Cafe, and Pinky Cafe all contain images of omuraisu, curried rice, and desserts covered in rakugaki. Similarly, all contain the word moe somewhere on the flyer, sometimes written in ketchup on a food item.

[11.3] Event days at chronotope E cafés occur with the highest frequency among all the café chronotopes—as often as two or three times a week, as is the case with @Home Cafe, compared to once a month with most chronotope A cafés. Many of the larger events, which occur around Halloween or Christmas, last for longer than a day and are usually drawn out over a 2- to 3-day period. The most common events at chronotope E cafés are satsueikai and costume changes.

[11.4] As with chronotope A cafés, chronotope E cafés attract considerable numbers of female customers. Female customers in chronotope A cafés were observed either alone or dining in groups of two to four together (without men). In contrast, the highest number of women in chronotope E cafés were accompanied by men, either as half of a couple or as part of a group of male-female couples. It is possible that for this demographic, the point of visiting a maid café is to listen to musical performances and engage in sensationalized interactive elements, and thus it acts as a social outing or date. Similarly, because of the focus on entertainment in chronotope E cafés, foreign tourists (in this case, those who appear white) were also present in noticeable numbers.

[11.5] The @Home Cafe in Tokyo's Akihabara is an example of a chronotope E café. Although it is often represented in the media as a singular entity, it is in fact fragmented into two branches (honten and donkihōte), with the former further divided into separate sites. The café prides itself on the large number of customers it has attracted; its Web site announces that it has had over one and a half million visitors. The @Home Cafe Donkiten is located on the west side of the Don Quixote department store's fifth floor, sharing the floor space with cosplay merchandise (with an entire section dedicated to different maid costumes) and an adult goods section that sells sex toys, pornographic DVDs, and erotica. In front of the chained-off entrance to the café is a small waiting area with chairs for customers to be seated on if the café is at full capacity. On the occasions I went, I was given an A4-size laminated card that stipulated the house rules of the establishment. This effectively functions like a consent form; acknowledgment of its being read permits entry into the café.

[11.6] The food menu at @Home Cafe Donkiten is one of the most limited of all cafés in Akihabara, if not Honshū. It is restricted to two types of pasta and four items in its rice category. The taste of the savory items is, at best, questionable and of low quality compared to that of chronotope A cafés. However, @Home Cafe is less about eating and more about the interactive events that accompany the food, such as moe incantations and food art. The limited menu and the food's poor quality are intended to be neutralized by the novelty and amusement of the interactive elements involved in their consumption. This counterbalance is indicated on the menus themselves, where colorful writing explains that for all the rice meals, "to finish the dish off, the maid will draw a picture on it with ketchup in front of all masters," as if to announce the dish is incomplete without a rakugaki component (note 11). Similarly, for spaghetti dishes, "the maid will mix in the sauce while she says a word like moe together with the master" (note 12). By explicitly announcing this as a part of the meal, customers are being made aware that what they are paying for is not the dish per se but rather the incorporation of these interactive elements into their dining experience—a novelty that does not exist at noncosplay cafés. Entertainment, not the food, is clearly at the forefront of their agenda. This typifies chronotope E cafés.

12. Conclusion

[12.1] Viewing maid cafés in a homogeneous or reductive way through the use of the terms iyashi-kei, "venues for relaxation purposes," and moe-kei, "venues with an agenda to entertain," is problematic. The chronotopes highlight their heterogeneity by permitting a more nuanced reading of the elements that comprise maid cafés. The need for a consideration of the heterogeneous qualities of maid cafés stems from the results of empirical research conducted at different establishments across Honshū, with vast differences in the services and interactive elements provided by the venues. The multiplicity of maid café settings is at odds with the homogenized fashion in which these establishments have been treated in the popular press, in local promotional publications, and in the academic literature.

[12.2] In response to this neglect of the diversity of the mise-en-scènes that exists among venues, the chronotopes may be used for maid cafés as a heuristic for viewing them heterogeneously; this will also permit maid cafés to be discussed as a holistic phenomenon in the future. The chronotopes are a heuristic for the consideration and evaluation of how spaces are constructed inside these establishments.

13. Notes

1. The origins of the maid persona in Japan are obscure and remain largely unexplored academically. Although Azuma (2009) claims that the first maidlike costume was present in the pornographic animation Cream Lemon: Black Cat Mansion, many Japanese fan Web sites insist it was neither a manga publication nor an anime series that vivified the maid but rather the highly popular video and gaming industry. Characters from so-called visual novels (interactive computer games with narratives) such as Sayori from Kindan no ketsuzoku (Forbidden kinship) and titles such as Kara no naka no kotori (Little bird inside the shell) and its sequel Hinadori no Saezuri (Song of the baby bird) are perhaps the first incarnations of the maid persona in Japan (; In these games, the attire of the maids was distinctly Victorian (a full-length black gown and white pinafore), with the latter two texts specifically set in Victorian Britain during the industrial revolution. The games were thus significant in positioning the origins of the meido (as the Japanese maid persona is known in Japan) and directly linked her to the female domestic servants of 19th-century England. The erotic story lines of these media consistently delineate the maid protagonist as a naive ingenue, vulnerable and at the total mercy of her master, the central male character of the narrative. However, over time, nonerotic incarnations of the maid persona developed, and according to Hayakawa (2008), this fascination with the maid identity and costume culminated in 1999 with the immense popularity of the manga series Mahoromatic. The eponymous heroine introduced fans of the maid genre to a different side of the maid persona, "a tough and noble young girl" (tsuyoku kedakai onna no ko). This persona was a precursor to numerous other maid-themed production, including Mori Kaoru's love story Emma, about a maid in Victorian England who falls in love with a wealthy upper-class man. The origins of the maid persona in Japan are thus connected to conceptions of dependence, servility, feminine passiveness, naïveté, (sexual) vulnerability, youthfulness, and domesticity. Although it is difficult to know whether these attributes accurately reflect the situations that domestic servants encountered in Victorian Britain, it aligns with how their experiences and histories have been represented and reconstructed by academics and numerous television dramas and movies in Britain.

2. By spaces, I am referring to the grounding work of scholars such as Lefebvre (1976), De Certeau (1984), and Soja (1996), who believed that space, although traditionally understood to have a purely geometrical existence, was best thought of as a production of social interactive elements. According to Lefebvre, because space is a social construction it is "not a thing but rather a set of relations between things" (1991, 81), in which power is effectively a medium for its production.

3. Locally produced promotional literature and mainstream variety shows in Japan frequently present maid cafés in a homogenizing way. Examples of such variety shows include YTV's Koeda no sugotoku and TBS's Shinchi Shikaikyū Kumagusu. In addition, documentaries or exposés about Japan have been produced abroad, including Japanorama (BBC, 2006), presented by Jonathan Ross, and In Search of Wabi Sabi with Marcel Theroux (BBC, 2009). The travel genre in particular has made several contributions in drawing maid cafés to the attention of foreigners, with blanket pieces on maid cafés. These include major international guidebooks on Tokyo and Japan, such as Lonely Planet (2008), Fodor's (2009), Michelin (2009), and Frommer's (2010), and travel television series, such as Channel Four's Globe Trekker in the United Kingdom, France 3's Faut Pas Rêver in France, and the Nine Network's Getaway in Australia.

4. Maid kyabakura might also be translated as a "maid cabaret club" or simply as a "maid club." As Faier explains (2009), contemporary hostessing establishments are fragmented into several specific venues—snack bars (sunakku), pubs (pabu), cabarets (kyabare), and clubs (kurabu)—all of which vary in price and services provided, though none explicitly offers sexual exchange, as establishments like pink salons (pinku saron) and soaplands (sōpurando) do.

5. There are no standard features that unite maid cafés when it comes to staff attire. Maid uniforms vary from black floor-length garments reminiscent of Victorian-era domestic servants to pink kawaii-type (cute) miniskirts, though one common element of all of these styles is a pinafore or an apron. Several texts have attempted to photographically document the difference in styles among maid cafés across Japan, including the Akihabara Housemaid-Cafe Costume Collection and Guidebook (2005) and the Parlour Maid Cafe Costume Collection (2006).

6. Some examples of productions and Web sites that have utilized this schema include NHK's Tokyo Eye, reviews on and, and listings on

7. In Japanese, these read as follows: (1) "Tōten wa inshokuten desu. Inshokubustu no teikyō igai no ikanaru sabisu mo itteorimasen"; (2) "Tōten sutaffu wa konpanion de wa naku, uetoresu desu"; and (3) "Deai mokuteki no okyakusama wa sōiita rui no omise ni sorenari no ryōkin wo haratte iku koto wo osusumeitashimasu."

8. The Japanese reads, "Milkcafe wa chajiryō, sekiryō, sabisuryō, nado issai itadakimasen. Wazatorashii gehinna 'moe' sabisu mo issai arimasen. Meido kissa wa takakute mazui, mirukukafe wa yasukute umai."

9. The Japanese reads, "Fairy Tale wa dōwa ya otogibanashi to itta imi no sōgo to narimasu. Kiku hito ni, dokoka natsukashisa ya tanoshisa wo ataete kureru dōwa. Sono dōwa no yōni, hajimete no kata ni mo ochitsukeru, yasashii omise ni shitai to omoimasu."

10. The Japanese reads, "Kokoro mo karada mo meido-san ni narikitte itadakima~su!"

11. The Japanese reads, "Saigo no shigae ni, goshujin-sama no mae de, meido ga kechappu de oe kaki itashimasu."

12. The Japanese reads, "Moe-na kotoba wo meido ga goshujin-sama to isshoni iinagara mazemaze itashimasu."

14. Works cited

Allison, Anne. 2006. Millennial Monsters: Japanese Toys and the Global Imagination. Berkeley: University of California Press.

Atkinson, Paul, Amanda Coffey, Sara Delamont, John Lofland, and Lyn Lofland. 2001. Handbook of Ethnography. London: Sage.

Azuma, Hiroki. 2009. Otaku: Japan's Database Animals. Introduction and translation by Jonathan E. Abel and Shion Kono. Minneapolis: University of Minnesota Press.

De Certeau, Michel. 1984. The Practice of Everyday Life. Berkeley: University of California Press.

Delamont, Sara. 2004. "Ethnography and Participant Observation." In Qualitative Research Practice, edited by Clive Seale, Giampietro Gobo, Jaber F. Gubrium, and David Silverman, 205–17. London: Sage.

Esterberg, Kristin. 2002. Qualitative Methods in Social Research. Boston: McGraw-Hill.

Faier, Lieba. 2009. Intimate Encounters: Filipina Women and the Remaking of Rural Japan. Berkeley: University of California Press.

Galbraith, Patrick. 2009. The Otaku Encyclopedia: An Insider's Guide to the Subculture of Cool Japan. Tokyo: Kodansha International.

Galbraith, Patrick. 2011. "Maid in Japan: An Ethnographic Account of Alternative Intimacy." Intersections: Gender, History and Culture in the Asian Context 25.

Hammersley, Martyn, and Paul Atkinson. 1995. Ethnography: Principles in Practice. London: Routledge.

Hayakawa, Kiyo. 2008. Meido Kissa de Aimashō. Tokyo: Āruzu Shuppan.

Jenkins, Henry. 2006. Convergence Culture: Where Old and New Media Collide. New York: NYU Press.

Kamberelis, George, and Greg Dimitriadis. 2005. On Qualitative Inquiry: Approaches to Language and Literacy Research. New York: Teachers College Press.

Lefebvre, Henri. 1976. The Survival of Capitalism. Translated by Frank Bryant. London: Allison & Busby.

Lefebvre, Henri. 1991. The Production of Space. Translated by Donald Nicholson-Smith. Malden, MA: Blackwell Publishing.

Louis, Lisa. 1992. Butterflies of the Night: Mama-sans, Geisha, Strippers, and the Japanese Men They Serve. Lincoln, NE: iUniverse.

O'Reilly, Karen. 2009. Key Concepts in Ethnography. London: Sage.

Pearson, Mike. 1998. "My Balls/Your Chin." Performance Research 3 (2): 35–41.

Silverberg, Miriam. 2007. Erotic Grotesque Nonsense: The Mass Culture of Japanese Modern Times. Berkeley: University of California Press.

Soja, Edward. 1996. Thirdspace: Journeys to Los Angeles and Other Real-and-Imagined Places. Oxford: Blackwell.

Wolcott, Harry. 2008. Ethnography: A Way of Seeing. 2nd ed. Plymouth: Altamira Press.

Transformative Works and Cultures (TWC), ISSN 1941-2258, is an online-only Gold Open Access publication of the nonprofit Organization for Transformative Works. TWC is a member of DOAJ. Contact the Editor with questions.