Managing the unmanageable: Emotional labour and creative hierarchy in the Japanese animation industry

The Japanese animation industry

The western fringe of metropolitan Tokyo is a hotbed of Japanese animation. Some 80–90 per cent of the domestic production capacity agglomerates here, with almost all major production companies in the industry clustering along the three parallel railway lines that run from Shinjuku and Ikebukuro stations westward (into cities of western Tokyo), and other cities of Saitama Prefecture. Between 2009 and 2010, I carried out 12 months of participant observation fieldwork in two animation studios. The first, UTT Studios, was located in the northern end of Suginami ward, where it is bordered by Nerima ward to the north and Musashino city to the west. Sprite Studios was located in Nerima ward, a 15-minute drive north of UTT. In both studios I interned as a production operator, an entry-level administrative worker, and worked with key creative staff such as directors and animators on a day-to-day basis on the shop floor. The ethnographic materials presented in this article are based on the work experiences of my fellow production operators as well as my own from fieldwork.

Although the first production of animation took place as early as 1917, the animation industry did not emerge as such until Toei Animation, Japan’s first large-scale commercial studio (then called Toei Douga), was founded in 1956. ² The televisual broadcasting of animation started in 1963, when the newly founded Mushi Productions, headed by manga artist Tezuka Osamu, produced Tetsuwan Atomu (Astro Boy) for Fuji Television. Originally a popular manga series penned by Tezuka himself, Tetsuwan Atomu centred on the adventures of a good-natured humanoid boy robot called Atomu (the astro boy).

While the show’s success in those early years of Japanese television animation was nothing short of phenomenal (Allison, 2006; Condry, 2013; Schodt, 2007), Tezuka proposed an unrealistically suppressed production budget to the broadcaster in an attempt to outbid his competitors, and this was to have far-reaching consequences for both Mushi Productions and the animation industry more generally. Mushi Productions’ low production budget became the standard with which other studios coming to television production had to work (Yamaguchi, 2004). According to Frederik Schodt (2007), Tezuka’s long-term English translator, the production of Tetsuwan Atomu was from day one driven by his passion to create a weekly serial television animation. For better or worse, sober economic calculations usually took the back seat in the operation of Mushi Productions and this, eventually, resulted in the studio’s bankruptcy in 1973.

The dissolution of Mushi Productions led to its creative body branching into several smaller production studios throughout the 1970s, many of which eventually grew into some of the finest production studios in the industry today (e.g. Sunrise, Production IG, Madhouse). Yet the lasting legacy of Tetsuwan Atomu and Mushi Productions are to be found in the way they set the perennial problem of net deficit between incurred cost and available budget in television production. It continues to paralyse the economic structure of commercial animation-making in the Japanese industry, sustained structurally by a substantially casualized workforce and an extensive subcontracting network.

On average, the production of a TV animation episode in the standard half-hour format mobilizes more than 100 workers over the production period, ranging from one to several months (but occasionally less than one month), depending on the production schedule. The yearly broadcasting cycle of television animation serials is divided into four quarters of roughly three months each. The broadcasting length of a given series varies from project to project, but in the past decade series running for either one quarter (i.e. 12–13 episodes) or two (24–26 episodes) have become more standardized. For such projects, the total number of workers involved in production will amount to several hundred. In order to deliver episodes to meet weekly television broadcasting schedules, a production company may simultaneously produce five, six or even 10 episodes, depending on its production capacity during an ongoing series.

The cost of production per episode varies from project to project, but generally ranges between ¥12 million (US$131,000) to ¥17 million (US$186,000) today, bringing the total production cost of a single-quarter series to ¥156 million (US$1.7 million) to ¥221 million (US$2.4 million), and twice that for a two-quarter series. ³ For the small and medium-sized enterprises that make up the majority of the Japanese animation industry, these are overwhelming investments to make two or four times every year.

Of the 300 to 400 production companies in the industry, no more than a dozen studios are independently capable of financial mobilization of this calibre. The rest operate as subcontractors on various levels, either as 1) original (sub)contractors (moto-uke) who receive production orders directly from financiers (i.e. a partnership of large corporations from TV, advertising, publishing, toy, and other outside industries); or, as 2) specialist studios (shita-uke) who receive subcontracting orders for particular production processes (e.g. Background Art, Key Animation). These specialist studios often themselves subcontract part of received production work to further subcontractors (mago-uke); the economic situation – and working conditions – usually exacerbate as one goes down the chain of subcontracting. For certain labour-intensive processes, most notably In-Betweening and Colouring (IB-TP), outsourcing to overseas subcontractors in China, South Korea, and Southeast Asia has become the de facto standard, accounting for approximately 70 per cent of such processes. The extensive networks and relations of subcontracting are of historical formation, linked with the development of commercial television animation production since the early 1960s. This is also true of overseas subcontracting, which began as far back as the early 1970s.

The common business model for financiers to recoup returns is through copyright merchandising and licensing. The vast number of production companies who operate as subcontractors are excluded from these channels of profit, and their relations are complicated and restricted further by the fact that the majority of commercial animations produced in Japan today are based on content originating in other platforms such as manga (comics), young adult novels, and video games. Studios which do not hold the copyright operate on a financial tightrope, balancing between the actual incurred cost of production and production orders paid in by either the financiers (in case of moto-uke), or by contractors one level above them (for shita-uke or mago-uke).

The extensive subcontracting relations between studios in the industry is paralleled by the extensive freelancing network comprising thousands of individual animators and directors, nearly all of whom work simultaneously on temporary project-specific contracts with multiple studios at various levels at any given time. As individual contractors in an extremely fluid labour market, they are without much job security, social security, or income stability in an industry with an estimated market size of ¥240 billion (US$2.6 billion) in 2007. ⁴

The basic unit of animation-making in the Japanese industry is called a ‘cut’ (katto). Like a ‘shot’ in filmmaking, an animation cut depicts a given duration of action, and a series of cuts makes up a scene. Take for example a scene showing a man walking into a room. The camera first follows him from behind. Second, it depicts his hand turning the doorknob. Finally, the camera shoots from the other side of the door as he enters into the room. These three actions will make up Cuts 1, 2 and 3. On average, the standard format of half-hour television animation in Japan is composed of approximately 300 cuts.

An animation cut, on the shop floor, refers also to the physical bundle of animation drawings. Using the above example, Cut 1 would consist of many paper drawings depicting the man walking toward the door. In the context of everyday work in an animation production studio, ‘cut’ is usually used in this second sense. These drawings are kept in numbered A3 sized ‘cut envelopes’ (e.g. from Cut 1 to Cut n). This dual existence – visual and material – gradually merges as the cuts proceed through production, most saliently when they are filmed and turned into an animated film.

The physical drawings of animation cuts are first created in the processes shown as ‘Layout’ and ‘Key Animation’ in Figure 1. Each box in the flowchart represents different production processes, and the overall progression of production is conceptualized as the vertical movement from top (‘Project Planning’) to bottom (‘Product Delivery’). The representation of the animation production process in this kind of flow diagram is standard practice in the industry, and one can find its iterations in a wide range of publications in Japan, from popular anime magazines to specialist textbooks. The extent to which it is the standardized representation of animation-making is also demonstrated by how one production operator actually drew his version of the flowchart for me during my fieldwork. OPEN IN VIEWER

Arguably, then, these flowcharts afford us a tentative glimpse of the animation production process in the Japanese industry: that it is paradigmatically conceptualized as an incremental process of visualization, whereby practitioners in different production divisions each bring their visual labour to the animation cuts. I say ‘paradigmatically visual’ because production operators, or any division representing administrative and managerial work, are characteristically absent from this representational logic even though on the actual shop floor and in the animation business managerial/administrative workers – such as production operators in my case – play no small part. In the rest of this article, I will examine just how this conceptual exclusion of non-visual, administrative labour from the representational diagram may be an implicit window into the micro-processes of workplace relations on the shop floor.

Production operators

As the shop floor administrator in charge of the production schedule, the work of production operators is, though critically important, also highly repetitive, physically and emotionally draining labour. In the everyday studio setting, the routine workload of the production operator revolves around two recurrent tasks: soliciting animation drawings from the animators, and having these drawings checked by the director. As subcontracting studios and freelancers work in a geographically disjointed network, production operators daily perform routine deliveries between studios and freelancers, quite literally driving the production process forward.

Production operators work under enormous pressure to mark important production milestones, but given their junior position in the organizational hierarchy, lack any actual power to give directions to the director and animators who create and manipulate the drawings. Because the operators are themselves unable to work visually on the production, they are not only rendered powerless, but also made dependent upon other creators for the effective management of the production schedule. As a result, the operators regularly find themselves in a situation where they have to adapt a strategically subordinate personality so as to beg and cajole the director and animators into getting their work done as quickly as possible. This is a constant source of deep frustration and, at times, of humiliation for operators. As the aforementioned Takamaki-san put it, this is not only a job ‘no one wants to do’ but also the one that feels ‘like some kind of messed up bullying (ijime)’.

In her study of the US airline industry, sociologist Arlie Hochschild (1983) developed the concept of ‘emotional labour’ as an important analytic to understand the subjective and experiential dimension of labour involved in the daily service work performed by flight attendants. In a celebrated passage, she (Hochschild, 1983: 7) defines emotional labour as follows:

This labor requires one to induce or suppress feeling in order to sustain the outward countenance that produces the proper state of mind in others. … This kind of labor calls for a coordination of mind and feeling, and it sometimes draws on a source of self that we honor as deep and integral to our individuality.

For the brief duration of such an exchange, which typically occurs in the main studio space, the operator who approaches the director stands while the director sits at a desk. The operator may leave the desk with a quick bow, albeit a casual one, with a customary and standardized phrase in the industry – yoroshiku onegai shimasu (thank you so much for your commitment). Likewise, when an operator is on the phone, one can usually tell if he or she is speaking with their animators (i.e. the ones participating in the operator’s episode) through similar cues: softened voice, a crafted thankful tone, sandwiched by social laughter. For the operators, communication with creators is part of everyday work, but that daily interaction is always carried out through these specifically crafted gestures, manners of speech, and a display of gratitude whilst asking creators to get their work done. Although creators, too, employ workplace honorifics and comportment in their verbal and physical interactions, neither directors nor animators seem to feel the same pressure to be complaisant.

The hierarchical relation between the operators and the creators is manifest not only in how the former conduct themselves around the latter, but also in how this power relation translates into differences in working hours and time management. While varying by project as well as across different production periods within a single project, the work hours of production operators average between 12 and 18 hours a day. While several hours each day are spent driving, the remaining eight to 12 hours do not necessarily all go into frantically juggling an avalanche of work (unless one is facing immediate production deadlines). During less busy times of the day – or the evening – operators spend their time at work waiting (machi), i.e. on standby.

This may sound rather benign, but waiting-as-work can become really taxing when it has to be done at half past two in the morning simply because the director happens to be a night owl and is still at work in the studio at that time. Even though the director does not usually need the operator to be present to do her own work, the operator is expected to be on standby in the studio, just in case he or she is needed for schedule- or technical design-related discussions; as the directors are freelancing with multiple projects, they do not necessarily memorize all the design specifics of the shows they work on. The particular workplace custom of the animation industry is also such that the operators must show by action how much they need the creators to do the visual work for them. This often boils down to the operators having to pull all-nighters in the studio during crunch time, just to show that they seriously require the directors’ work to be done as soon as possible, even though there is nothing the operators can do other than wait.

Potentially even more demanding terms apply to the operator-animator relation. Given the pressurized nature of the production schedule, it is often the case that the operator has to drive immediately to an animator’s workplace (be it his/her apartment or a different studio) once notified that their cuts (i.e. drawings) are ready. Animation cuts ready for the next production phase are called agari (‘up’, or ‘finished’ cuts). Although such notification is good news for the operator, a conflicting tension may arise when such calls come in past midnight when the operator was just about to wrap up for the day. Furthermore, what if two animators calls simultaneously on one night, and are in widely separated locations across western Tokyo?

The daily long drive(s) – ‘collection’ – by the operators to move animation cuts back and forth are amongst the most important and labour-intensive aspects of their job. An operator who is absent when a director needs him is excused if the reason is the ‘collection’. As one production operator (male, late 20s) succinctly put it, ‘an operator who cannot drive is completely useless’.

Driving to an animator immediately when the latter says the drawings are ready is not always necessary, but contingent factors often contribute to the situation being inevitable. The chief factor is usually scheduling: when the production is literally racing against time, collecting finished animation cuts as soon and fast as possible becomes an imperative for the operator, no matter the time of day. Another is prioritization. Unless hired at the top level of production (e.g. as general chief animator or character designer), animators usually work on multiple projects simultaneously. A situation arises when all the projects she works on are being produced under equally considerable temporal strains, whereby she needs to find ways to prioritize between them. A significant clue for the animator in this prioritization is the level of desperation evinced by the operators of the respective projects. An operator thus risks having her production side-tracked if she fails to come to collect the drawings as soon as they are done, since this could give an impression that her schedule is not as desperate as others, and the animator may decide to allocate her attention more fully to a different project. This would be acceptable if the production actually has some temporal leeway, but if not, losing the animator’s priority could have troublesome consequences. Finally, and quite irrespective of the matter of prioritization, an animator may simply demand that an operator come to collect his work as soon as he has finished. In such cases, the operator has little choice but to comply.

I suggest that a key to understanding the operators’ subordination lies in its correspondence to specific patterns of age and visual skill in the everyday business of animation-making. Like the Euro-American creative industries (e.g. Gill and Pratt, 2008: 14; McRobbie, 2002), the Japanese animation industry is a youthful industry: approximately 70 per cent of its total workforce is composed of men and women in their 20s (approx. 45%) and 30s (approx. 26%). ⁵ The demographic pyramid of the animation industry is therefore wide at the base, but narrows rapidly, and tilts increasingly towards the male side with age. ⁶ This generational structure creates a situation in which virtually everyone who holds major creative posts in the division of labour is senior, both in terms of age and career, to production operators, who do entry-level management work. The vertical and somewhat cultural relation of seniority thus mapped out across the operator-creator axis is further complicated by the organizational structure springing from the particular nature of the operator-creator axis in the context of creative visual production.

Thomas Malaby’s discussion of ‘the hierarchy of creation’ (2009: 124) in the context of the US video game industry is illustrative of this point. In his ethnography of the video game company Linden Lab, makers of a well-known online game Second Life, Malaby found a tension existing between what he calls ‘developers’ (programmers and others who directly create and manipulate the existing gaming platform) and ‘non-developers’ (who work in divisions such as administration and marketing) (2009: 107–24). The official management philosophy in Linden Lab discursively preferred an organizationally horizontal approach to decision-making and performance evaluation but, in the actual context of game development, employees’ outputs were evaluated on locally conceived scales of creativity within the studio that were hierarchical in character. As Malaby (2009: 122) observed, the ‘developers’ projects were in some sense paradigmatic of creativity’ against which the ‘nondevelopers often found themselves beholden to the interests and schedules of the developers’ (2009: 123). Employees’ outputs were, in reality, differentiated in terms of this locally enacted hierarchy of creation into classes of different creative significance. ⁷

Malaby’s concept of the hierarchy of creation applies well to the context of Japanese animation industry. As discussed above, the production processes of commercial animation are locally conceptualized as the incremental performance of visual labour from the upper-stream to the lower-stream divisions. In this, animation production was paradigmatically the visual production of animation cuts. Since operators’ output in the production process did not partake in the visualization process itself, this ‘exclusion’ enacted a second order of workplace hierarchy based on the embodied visual capacities, alongside seniority, between operators and creators on the shop floor. ⁸ Following Malaby, I address this combined hierarchical structure as a creative hierarchy in the workplace of the Japanese animation industry.

The strong imbalance of bargaining power between operators and other creators frequently made it difficult for the former to fulfil their responsibilities, as the management of the schedule on the shop floor boiled down to the effective management and orchestration of creative personnel. Yet the production operators lacked the most basic capacity of a manager, which is to give directions and assign short- and long-term goals to team members effectively with organizationally backed-up legitimacy. Drawing up a plausible production plan ahead of time was not in itself such a demanding task; problems emerged from difficulties in making others work according to the plan, and making sure everyone stuck to it. This latter challenge almost always faltered, throwing off the schedule. Devoid of power but having to manage nonetheless, operators resort to cajoling, urging, coaxing, begging and imploring their teams to meet their production milestones. This dynamic transformed everyday interactions and communications with creators into a nexus of considerable emotional strain and often overwhelming stress, compounded by physical exhaustion from lack of sleep, overwork and lengthy drives every day.

The operator-creator relation in the Japanese animation industry therefore stands in marked contrast to the editor-manga artist relation observed in the Japanese manga (comic) industry by Sharon Kinsella (2000), in two otherwise very proximate industries. Kinsella also observes a notable tension between editors and manga artists in the context of commercial publishing, but in her case it was usually the editors who held the upper hand in the communication and negotiation processes, ‘pushing manga artists to the periphery of the creative process’ (Kinsella, 2000: 10).

According to Kinsella (2000: 12), editors in Japanese publishing houses are ‘relatively wealthy, highly-educated and privileged individuals’ with secure long-term employment in what are some of the largest corporations in the country. To borrow from Bourdieu (1984), then, there is a notable disparity between the editors and the manga artists in terms of social, cultural, and economic capital, especially when the latter receive work from the former, who usually dominate the channels for commercial publication (unless of course the artists become best-selling ‘stars’, which may substantially redraw the power relation).

In the animation industry, the disparity in capital between operators and creators is not so clearly observed. While there are now more and more degree holders among 20-something operators and creators, those who hold power, being of an earlier generation, may or may not be university graduates. On the shop floor, any wider political economy of educational pedigree was implicitly dismissed on the basis that a bachelor’s degree had nothing to do with the visual production process per se and drawing well, especially among the creators, was seen to be of much greater importance. Here, the tacit formation of a creative hierarchy on the shop floor may inadvertently also serve as a counter-discourse to the wider social economy of educational credentials (gakureki) in Japan, as a kind of parallel meritocracy based on embodied skill. ⁹

The extent of emotional labour that the production operators perform in the solicitation processes of animation work from directors and animators may be glimpsed from the following outburst I heard from an UTT Studios operator I call Yokoyama-san, a man in his mid-20s, when we were seated in the security of a izakaya-style bar away from work in western Tokyo:

Man, it’s anything other than this idea that animators think [the operator is] lower (shita), directors think [the operator is] lower (shita) that pisses me off! The operator’s job, it’s hard but I can do it, but in the studio everyone seems to work with that idea [of hierarchy between operators and creators] as if that’s their entitlement. But then, other people often don’t see it as I do, and think it’s me who’s got an attitude. My boss once told me I should really swallow my pride, but then, what’s the point of working on like that?

Not coincidentally, this is where the importance of emotional labour as a sociological critique lies for Hochschild. It is not just problematic for the emotional sacrifice it demands from workers, but the ways in which such emotional sacrifice are institutionalized as routine work in a diverse array of workplaces exposes the extent to which selfhood is itself turned into a resource to enhance production and economic activities in today’s capitalist societies. In Hochschild’s words, ‘the worker can become estranged or alienated from an aspect of self … that is used to do the work’ (1983: 7, emphasis in the original).

Caught in the multiple middles

On one hand, the emotional labour of production operators can thus be analysed in terms of the generalizable work experience of being caught in the middle, especially when the ‘middle’ is produced by vertical – i.e. hierarchical – organizational dynamics. To quote Hochschild again, it is ‘when the emotional display that one person owes another reflects a certain inherent inequality’ in an institutionalized context that emotional labour becomes particularly taxing (1983: 19). To this extent, the everyday interactions that production operators have with creators participate in the kind of estrangement from the self, rather than its actualization, which can be characterized more generally as emotional labour.

On the other hand, I argue that ways in which workers such as ‘production operators’ become caught up in the ‘middle’ are specific to the organizational dynamics of the Japanese animation industry and, to a comparable extent, also in common with other creative industries. Organizationally, these workers operate at the intersection between management and creativity, and this makes them subject to performance criteria that are heterogeneous and often contradictory.

Insofar as this tension between management and creativity is concerned, the organizational dynamics of animation production companies may resemble a form of ‘heterarchy’. According to David Stark, heterarchical organizations have multiple and often competing evaluative principles that are in no simple way ordered hierarchically (2009: Location 585–95/6481). Organizations with multiple evaluative principles, moreover, ‘foster a generative friction that disrupts received categories of business as usual and makes possible an ongoing recombination of resources’ (Stark, 2009: Location 545, emphasis added). As I discuss below, the state of generative friction – or well-finessed creative chaos – is what studios generally aim to mobilize in the creative process of animation making. Yet, production companies are organized to achieve this state by imposing a rigid structure of hierarchy and, in this respect, they also differ substantially from typical heterarchical firms, which aspire to flatten traditional forms of hierarchy (Stark, 2009: Location 587, 701/6481). In what follows, I will expand my analysis of the creative hierarchy to elaborate with this observation ethnographically.

As Figure 2 shows, the division of labour for the studio-level management is simple compared to that for creative visual production (see Figure 1). Another important difference from the creative divisions is that the vertical juxtaposition of different management positions defines organizational hierarchy. Production operators (shown as ‘P.O.’ in Figure 2) work under the production desk, who in turn report to the producer. This is not the case for the creative divisions, where the vertical relation seen in Figure 1 primarily defines the flow of production process in temporal order and is not – at least in principle – an index of hierarchy. Although most creative divisions are, in reality, outsourced specialist subcontractors located elsewhere, the three ranks of management workers are hosted in-house. ¹⁰ OPEN IN VIEWER

Figure 3 shows how the creative and production management divisions are positioned vis-à-vis one another and upper management in the wider organizational context. Here, the lateral relation between the creative and production management divisions (Figure 1 and Figure 2) at the studio level is undercut by the vertical placement of ‘upper management’ (e.g. the board of executives, senior managers who oversee the sales performance of production projects). ¹¹ Since the upper management and the production management are, in reality, organizationally integrated, at this level of abstraction the Japanese animation industry concurs with the critical model of creative industries in which artists and creative individuals are largely confined within the institutional framework of business and corporate initiatives. OPEN IN VIEWER

As we have seen, however, this macro perspective does not adequately capture the actual work experience of production operators or account for the predicaments arising from everyday work. Production operators report to the production desk, the production desk to the producer, and the producer to upper management. Provided that this chain of reporting and command is complete, the existence of a creative hierarchy that makes operators subordinate to creators in the studio is not just taxing for the operators but, in a wider corporate context, organizationally subversive for the upper management and could be detrimental to the firm as a whole.

That animators and directors dominated operators was a taken-for-granted ‘fact’ of work at UTT Studios. Since production desks and producers had once worked as production operators themselves, this creative hierarchy is, however informal, a well-recognized aspect of studio work in commercial animation. During my fieldwork neither Sugita-san (production desk, male, mid-30s) nor Yamagami-san (producer, male, early 40s) seemed motivated to rectify the situation for UTT’s production operators. On the contrary, the structure of management as a whole was such that it was, if anything, complicit in maintaining a hierarchy between creators and operators in the workplace relation. Its very existence was seen as beneficial for commercial animation making.

To better understand this situation, we need to look at the workplace relations between production operators and higher ranked managers. The operators are tasked to manage the production schedules for individual episodes (approx. 24 minutes in length) of a serialized television production. Since studios produce multiple episodes simultaneously in order to meet the demand of the weekly broadcasting cycle, there is a need for an experienced mid-level production manager, namely the production desk, who can oversee several production operators and the progress of their individual episodes. Thus, while operators do whatever they can to solicit drawing work from animators and directors, they also prepare daily progress reports to the production desk, i.e. their actual boss.

These progress reports tie in with an aspect of schedule management work that production operators carry out daily vis-à-vis the management side of commercial animation making: the construction of the production database. As noted earlier, an episode of television animation is sequentially composed of approximately 300 animation cuts. In the studio, those 300 cuts exist as physically distinct individual bundles of animation drawings. These animation cuts enter the production process (think of Figure 1) simultaneously, but since some cuts are easier to animate than others, the pace at which they complete the visualization process varies greatly. What this means is that operators need to keep track of all 300-odd animation cuts individually as they traverse the production process, each at its own pace.

The production database is therefore a device and technique used by the operators to monitor the progress of these animation cuts. For this purpose a spreadsheet is drawn up, where vertical columns represent individual animation cuts (from Cut 1 to Cut n) and horizontal rows represent individual processes of visualization (e.g. key animation, colouring, filming) that each cut must go through. Two columns are allocated for each production process, wherein operators enter the dates on which the cut enters and leaves the process. Through this database operators obtain a synoptic vantage point from which they can see the whole production process numerically and prepare progress reports for the production desk.

Secondly, the production database records the number of drawings in individual animation cuts. Since more drawings directly lead to a higher cost of production, this information is used to monitor and contain production within the available budget.

Sprite Studios, another production company where I did fieldwork, was famous in the industry for having the strictest measures to contain the number of animation drawings used in a production, at a maximum of 3500 pieces. The number of drawings was simply, but universally, referred to as maisū – the ‘numbers’ – at Sprite, and was an intense locus of scrutiny in everyday work from the production desk and higher managers. One indication of this is the use of dedicated in-house software which connected production databases managed by individual operators to the company’s central servers, so that every manager in Sprite – operators and those above them – had panoptic access to the databases of ongoing productions.

Watching numbers carefully was central to the on-the-job training I received during my internship at Sprite Studios, to the extent it was concerned with the ability to manage production successfully: being able to deliver an animated episode with less than 3500 drawings was a significant aspect of performance evaluation for production operators. This was the case even though Sprite’s operators – or production operators in general – did not have the authority to intervene in the production process and reduce the number of drawings themselves; only directors had the power to alter the number of drawings prepared by the animators. Consequently, the drawing limit was another source of emotional labour where operators had to beg (this time the director) to be frugal with the number of animation drawings she allowed her team of animators.

In Sprite Studios, no phrase summarized the production control that the upper management strove to achieve better than sūji de miru – ‘seeing through numbers’. Animation is a visual craft; this defined how creative practitioners measured the progress of a given episode. In contrast, the expression ‘seeing through numbers’ points to a more abstract conceptualization, one paradigmatic of the business aspect of commercial animation making.

It is the production database that enables this numerical seeing. The visual characteristics of an animated scene, whether it depicts lovers embracing or a building exploding, are understood in terms of the number of drawings used to animate them. Instead of capturing the embrace as a narrative climax of a romance, as the director and animators intend to portray it, the production database may simply record it as 78 drawings. At Sprite the notion of ‘seeing through numbers’ was equated with the perspective of the upper management, where the ‘numbers’ tied in with other quantified aspects of commercial animation such as sales records and television ratings. Together with the strict limit of 3500 drawings per episode, the production database and the numerical accounting of animation were in place to keep production – and creators – in check within a viable and profitable business operation.

Although Sprite went further than most studios to enforce this management-led regulation of creativity, the basic idea that creators and the creative process had to be harnessed was shared with other studios. The phenomenal success and subsequent fall of Mushi Productions was a cautionary tale and something of a founding myth in the industry, and the consensus is that a studio cannot possibly survive financially if creators are given free rein without valid interventions from managers. Yet here lies the fundamental dilemma of the animation business: good animation is not necessarily made of good numbers. Indeed, anecdotal evidence often points to the contrary. As a recently retired executive (male, late-60s) at Sprite put it: ‘That a project finishes smoothly without major hiccups, according to the schedule and everything, is usually no guarantee that the result would be any good.’

A studio can produce a perfectly managed television project and end up with a perfect flop. The man behind these words, Furuhata-san, was an industry veteran who joined Sprite Studios as a production operator in the 1960s and eventually became a senior vice president. Drawing from the decades of experience at one of the largest production companies in Japan, Furuhata-san had his own theory about how good animation was to be created:

The studio must be full of vitality (katsuryoku). There is almost uncanny (makafushigi) correlation between the vitality of the studio and the vitality of animation it creates. If the studio begins to lose its vitality, then somehow, the animation it creates starts to lose its grip and appeal. That’s one fascinating thing about animation making. […] Often, the more delay we had in production, and the more chaotic the production process became, we came out the other side with a good and interesting result.

If these incidents appear straightforward manifestations of contradiction, this is because the animation business thrives on the strength of this contradiction. It is always in productive – or disruptive – tension between vitality and rationality, art and commerce, creativity and management.

At this point, the relationship between managers and creators in the animation business partakes of a more general problem recognized in organizational sociology between under-organization and over-organization. In her study of Burning Man, an annual weeklong event in which tens of thousands of people gather in the Nevada desert to erect a large effigy and set it alight, Katherine Chen learned the fundamental challenge of organizing efforts to be finding the right balance ‘that enabled creative chaos without debilitating chaos and totalitarianism’ (2009: Location 115/3136). If the Burning Man organizers failed to give it enough structure, they risked having the event gradually disintegrate and fall apart. If, on the other hand, they over-organized, with too much centralized control, then ‘the event’s creativity and vitality could suffocate’ (Chen, 2009: Location 78/3136).

Chen’s observation about finding the right balance for creative chaos drives home the fundamental dilemma that senior managers such as Furuhata-san grappled with at Sprite, and those like him at other studios across the Japanese animation industry. The indisputable fact was that the more time the director and animators spent on their drawings, the better the final quality of animation. The more drawings creators used in an episode, the closer the animation came to an illusion of life. But all these had to be accounted for in the context of viable business, and there was no easy answer to the problem.

Even though the dilemma applied to the whole structure of commercial animation making, production operators were among those worst affected by the paradox, for they had to bear the full blunt of the conflict between management and creativity in everyday work, facing the creators horizontally, while being supervised by other managers vertically. Indeed, the creative hierarchy works in favour of the upper management in this organizational context, because insofar as operators can still meet their schedule, it allows creators to exercise the maximum freedom possible within the limits of commercial production. To the extent that production operators need to work toward the contradictory goals of keeping the creative process in check while fuelling its vitality, they are forced to take on the role of an organizational buffer, absorbing the disruptive tension between management and creativity with intense emotional labour and excruciatingly long hours.

Managing creativity

In the Japanese animation industry, the creative hierarchy exists as an informal structure of organizational conceit to ensure and maximize creative freedom within the corporate framework. The ways in which this tension is managed are primarily hierarchical, rather than heterarchical, because the structure of organization relies on a particular group of actors, i.e. production operators, to resolve the disruption that could stem from it. Rather than being able to benefit from the heterogeneity, production operators are trapped in the multiple middles of commercial cultural production, sandwiched by both vertical (senior managers) and lateral (creators) forms of hierarchy.

The larger claim I make in this article, therefore, relates to the analytical purchase in paying situated ethnographic attention to manager-creator relations on different levels of organizational hierarchy. The capacity of directors and animators to work directly on the visual production of animation cuts on the shop floor gave them a privileged position of power in their everyday dealing with the production operators. From the organizational standpoint of production operators, who represent the management on the shop floor but lack its actual power, the only way to fulfil their responsibility promptly was to be complicit with this creative hierarchy.

While critical scholarship on the creative industries has rightly examined the relations between management and creativity in terms of the gradual overtaking of the latter by the former, in the micro-processes of workplace relations such a macro analysis does not always hold. Those structurally ‘tortured artists’ are themselves real and living human beings embedded in the micro-relations of workplace and the macro context of a volatile market. In their relative artistic subordination to the corporate structure of commercial production, they may also – consciously or otherwise – dominate others whose job is to assist and work for them in the actual production process. To the extent that senior managers continue to rely on the informal creative hierarchy on the shop floor to produce both financially and artistically successful animation, the management itself is dependent on the vitality of creative process it seeks to systematically harness and control. Thus structures of domination in actual creative workplaces are more complex and multi-dimensional than the one-sided triumph of management over creativity.