Why Should We Praise Depth in Animation? — Animated Works as Models

Aspires to be someone who gets other people off.

Editor’s note: Austoria has done it.

Abstract

Depth is an indispensable critical category across the various forms of animation criticism. We often describe animation as “deeply meaningful” or “thought-provoking,” or say, more specifically, that a work “lays bare and interrogates human nature,” “is full of humanist concern,” or “reflects on modernity,” and so on. Although depth is a concept with indistinct boundaries, it is clearly always introduced into our criticism of animated works as something to affirm and praise. In this essay, I examine a question that is easily overlooked: why is depth a critical category worthy of praise in animation? This involves considering the multiple meanings of depth in animation criticism and comparing them with depth in scholarly research, which is often regarded as the exemplary expression of depth in human thought. Having clarified the question, I propose an answer: animated works can be understood as a distinctive kind of model, allowing the content we regard as deep to offer unique insights in neglected areas and from overlooked perspectives. This account brings out the ontological similarities between animated works and scientific models. I go on to discuss the advantages of this approach.

Keywords: depth, animation criticism, models, ontology, interpretation.

1. Introduction

Whether in everyday assessments of animation on online platforms or in more serious critical essays, depth is an indispensable critical category. What distinguishes it is that it usually concerns the most abstract and conceptual aspects of animation, making this loosely bounded category difficult to define precisely. Yet we routinely discuss an animated work’s “depth,” consider what it “really wants to say” through its themes, and even use theoretical frameworks to explore its responses to, and resonances with, grand themes.

Given how evidently widespread appeals to depth are in animation criticism, examples of what this category might encompass are easy to find. Interpretation and depth are conceptual twins: when we think a work possesses certain properties that fall under the category of depth, we are actually constructing what is deep in it through interpretive practice. Depth has as many particular forms as there are particular acts of interpretation in the world. But a harder question than what depth means is this: what gives depth a legitimate place in animation criticism? Or, put another way, why should we praise depth in animation criticism?

The question itself may seem puzzling. Why should depth, as a category of animation criticism, be a problem? Almost all of us would at least agree that “excellence in the animation” is an obviously positive assessment of an animated work. I will therefore begin, in section 2, by explaining in detail why depth is problematic as a positive critical category. To clarify the problem, I will compare our typical understanding of depth with a paradigm of intellectual depth: that found in scholarly writing. This comparison will expose contradictions in our unthinking praise of depth, and these are precisely where the problem lies. In section 3, I will briefly introduce the concept of a model, widely used across academic disciplines, and explain why an ontological similarity makes it reasonable to bring this concept to bear on our problem. In section 4, I will discuss how treating animated works as models allows us to understand interpretive practice and the content we regard as deep in animation. Finally, in section 5, I will explain how this model-based account addresses the central contradictions in the puzzle of depth, and briefly show why treating animated works as models of real-world systems offers a better response to that puzzle than certain alternative views.

2. The Puzzle of Depth: Difficulties Arising from Everyday Intuitions

Although examples of depth as a critical category are easy to give, this does not mean we are always referring to the same aspect of a work when we say that one animation “has real depth” or another “makes you think.” In actual usage, the category has many faces. Sometimes we mean the multiple layers of meaning disclosed by symbols or metaphors; sometimes, the overt and covert structures of storytelling. Sometimes we mean merely that certain kinds of imagery appear in an animation, opening up possibilities for interpretation. Different aspects may also combine or overlap, and so on. Broadly speaking, what these various forms of depth share conceptually is a demand to interpret what is hidden. When we think an animated work possesses properties that fall under the category of depth, we do not mean that features directly describable in observational language display those properties.¹ Rather, we mean that elements of the animation can be endowed with those properties through our interpretive practice.

In this sense, the most typical form of depth is engagement with grand themes, since criticism concerning such themes most clearly exemplifies systematic, orderly interpretive practice. The meaning of life and death, the tension between war and peace, the divide between justice and evil, authenticity and alienation amid technological progress: themes of this scale usually require a relatively serious approach, and may even require us to introduce theoretical frameworks into our discussion. Whether or not we agree with a particular interpretation, we would generally agree that if, for most critics, an animated work as a whole offers interpretable elements relating to grand themes, it can unquestionably be regarded as deep. I will therefore limit the discussion in this essay to depth in its most typical sense: engagement with grand themes.

For most critics, at least, an animated work’s successful presentation of grand themes is undoubtedly a strength worthy of praise. To see why unthinkingly praising depth in animation criticism is problematic, we can consider why we generally praise depth outside animation. This usually rests on an everyday intuition about epistemic superiority. To call a claim deep often means that it supplies “better” knowledge—for instance, a better explanation or prediction of a phenomenon.² For example, “Suicide is a social phenomenon that takes various forms and results from various social factors” is deeper than “Suicide is an individual act caused by psychological problems,” because the former better explains how suicide comes about and the mechanisms behind it. In more everyday language, we might say that the former offers “deeper insight.” It is precisely this epistemic superiority that makes us prize depth so highly.

This intuition about knowledge is also reflected in much of the more serious criticism of animation. Such essays draw on scholarly resources to give a better account of how a work appropriately engages with the grand themes in question. Deleuze, Lacan, Agamben, Badiou: these names from contemporary intellectual history keep appearing in animation criticism today. Structuralism, feminism, existentialism, postmodernism: these overlapping currents of thought likewise permeate current interpretive practices, not to mention the enduring popularity of psychoanalysis. After all, scholarly writing stands in our culture as a paradigm of depth. If we want our criticism to seem more insightful, drawing on the wisdom of our predecessors is surely a reasonable choice.

But can the everyday intuition behind our praise of depth directly explain why we praise it in animated works? If we accept this explanation, then what we admire as depth in animation is actually epistemic superiority. When we say that the 1995 Ghost in the Shell is a deep animated work, we are really saying that it provides reliable knowledge—about technology, mind, soul, and so forth—and that, to some extent, this knowledge is available only through this work. Accepting this explanation, however, inevitably brings difficulties. A comparison between depth in animation and depth in scholarly writing will make these clear.

Once we accept this explanation based on everyday intuition, depth becomes reducible to epistemic superiority. Most people would also agree that scholarly discussions of grand themes tend to have greater depth than their presentation in animation. Scholarly works, for instance, formulate questions on the basis of more accurate and detailed information, develop their arguments more rigorously and systematically, and produce findings that must undergo assessment by a scholarly community. We should therefore be entitled to argue that, by the criterion of depth, scholarly works generally deserve more praise than animated works, since they provide “better” knowledge. But this conflicts with familiar facts. At least so far, almost no one discussing the presentation of a grand theme in animation—the mind in the Ghost in the Shell series, for example—compares it with scholarly discussions of that theme and uses the comparison to criticize the animation’s insufficient depth. This shows that we are well aware that the vast majority of animated works offer, at best, “profound insights in a zoo,” and so we do not press them too hard to provide superior knowledge of grand themes. If we really believed depth deserved praise because it embodied epistemic superiority, we ought to demand that animation’s treatment of grand themes move ever closer to rigorous scholarly research. Yet this is not how we actually criticize animation.

A more important problem is that we can scarcely compare animated works to determine which offers deeper insights. The standards for judging one insight “better” as knowledge, and therefore deeper, may differ from context to context. But we would all agree that depth in scholarly writing derives not only from the views it advances but, more importantly, from their justification. In principle, a reasonable assessment of a view requires a thorough examination of how the relevant evidence and arguments support it. This is almost impossible in animation. However aptly we interpret Cannon Fodder, the segment of Memories, as an indictment of totalitarianism and technological alienation, the work offers only a presentation, not a justification of the view in any sense.

These two difficulties rule out using our everyday intuition about epistemic superiority to explain why depth is praised in animated works. This is the puzzle posed by depth as a critical category: animation contains things we greatly admire, things we can interpret and describe using expressions that fall under the category of depth. Yet our everyday reasons for praising depth cannot readily explain why we praise these things in animation.

Of course, we could reject the idea that this everyday intuition about depth is a legitimate resource for answering the puzzle. In section 5, however, I will argue that this strategy not only comes at too high a price but may still have unavoidable defects. An indirect response grounded in everyday intuition therefore remains the more promising approach.

3. The Ontological Similarity between Animation and Models

In the previous section, I argued that our everyday intuition about depth cannot directly explain why depth in animated works deserves praise. If we do not want to abandon our existing understanding and use of depth as an evaluative category, we need to construct an indirect explanation through some intermediary.

Across academic disciplines, models are widely understood to serve as intermediaries. In both the natural and social sciences, the phenomena we study are invariably intricate and bewilderingly diverse, while the theories we hope to arrive at must capture regularities within a class of phenomena. By describing phenomena in idealized terms, models become indispensable intermediaries between theory and phenomena. In microeconomics, for example, discussions of agents’ economic behavior usually model those agents on the assumption of a “rational agent.” Whatever meaning we give to rationality in different contexts, this model plainly differs from an actual person in the real world. Yet such a partially distorted model is indispensable to our search for regularities in economic behavior at the micro level.

Whatever account we give of the nature of models and how they represent their target systems, the epistemic role of this conceptual tool is clear. Although the vast majority of models are human constructions, the knowledge we use to construct them does not encompass all their properties. By studying models, then, we can learn about properties not covered by our existing knowledge and, in turn, learn something about those properties in the target systems the models represent.

Can we introduce the concept of a model as an intermediary in answering the puzzle of depth? In other words, can we treat animated works as models when addressing this puzzle? Answering this question requires us to focus on ontological similarities. We need to clarify which ontological features animated works share, to some extent, with models, and on that basis justify treating them as a kind of model.

The usual ontological question is: what exactly is a model? Models of all kinds are proliferating across disciplines. Even among familiar examples, the concept refers to very different objects: physical entities, fictional concepts, mathematical formulas, descriptive statements, and so forth. We can accordingly divide models into numerous overlapping categories, without exhausting the possibilities: material models, conceptual models, mathematical models, computational models, descriptive models, model organisms, and so on. Precisely identifying the ontological features shared by these different concepts of a model is therefore no easy task. Fortunately, this difficulty also means that we need not make an overly strong claim about the nature of models. Different models may, in practice, bear family resemblances to one another, and we often bring new conceptual constructs into the category as contexts change. Thus, when considering whether animated works can be regarded as models, we can try to identify shared features among some typical models and use these as criteria for treating something as a model.

On this ontological question of what models have in common, I will draw in part on an answer offered by Michael Weisberg. Seeking to account for models and modeling practices in contemporary science, Weisberg identifies three principal kinds of model: concrete, mathematical, and computational.³ Their common features can be summarized in two parts: a model’s structure and an interpretation of that structure.⁴

It is obvious that every model has an internal structure. Consider concrete, physical models. The structure of a plastic DNA model is the double helix made of plastic; the structure of an aircraft model for wind-tunnel testing is a geometric form that can be placed in the tunnel for the relevant tests; the structure of a model of a building is a small-scale object made from materials such as reinforced concrete. If we regard an animated work as a model, it plainly has a structure too: the temporal ordering of its images, sound effects, dialogue, and so on.

More decisive than the structure itself in making it a model of something is the interpretation of that structure. The most important part of this interpretation specifies a representational relation between the structure and its target. Weisberg acknowledges that, in principle, any structure can be regarded as a model representing some target, but not every structure can appropriately represent every target.⁵ Appropriate representation must meet a number of specific conditions, which we can broadly summarize in terms of isomorphism. A plastic object stands in a representational relation to a DNA molecule precisely because its spatial structure is isomorphic to that of the molecule. Similarly, a set of mathematical formulas can be regarded as a model of a target phenomenon because the relations among its variables exhibit a certain formal isomorphism with that phenomenon. This does not, of course, require a model to represent something that actually exists: Ptolemy’s geocentric model, for example, represents nonexistent objects such as “epicycles.” Recognizing this is crucial to treating animated works as models. On the one hand, to count as a model, something need not in principle represent an actual target system or phenomenon. On the other, the best scientific models are often not those most faithful to reality: models are constructed with all kinds of fictional elements. In particular, when we do not require a model to pursue the epistemic goal of truth, we can even construct one that, as a whole, represents a fictional object. As long as we use it as a model on the basis of isomorphism in some sense, it still possesses the structure required to be regarded as a model.

An objection may remain: what exactly does an animated work “represent”? Here, too, we can draw an analogy with science. Even with the most outlandish scientific model, we can readily explain the representational relation between model and target. A fairly natural account of this relation between scientific models and fictional objects runs roughly as follows. Although these fictional objects do not actually exist, if they did exist, they would occupy particular positions in space and time, have particular physical features, and belong to a network of causal relations. Model-based scientific practice investigates precisely the circumstances described by such conditionals.⁶ There have, for instance, been models of the ether in the history of science. Although contemporary physics no longer has a place for this substance, we can still say that those models, taken as a whole, represent a nonexistent object called “ether.” Suppose a conventionally trained physicist today, knowing perfectly well that ether does not exist, nevertheless uses sophisticated mathematics to construct an elaborate model of it out of personal preference. We could say that the model describes what would be the case if ether really existed: the substance could then be accurately described by this model.

The situation in animated works is slightly more complicated. R. M. Sainsbury points out that authors of fiction regularly describe things that genuinely exist in order to enhance a story’s sense of reality. An author may describe fictional events involving a fictional character, for instance, while setting those events in a real location whose geographical features correspond exactly to reality.⁷ As a product of moving-image technology, animation is especially adept at conveying this sense of reality. In Fate/stay night, the mechanics of the Holy Grail War are fictional; a King Arthur who remains forever a teenage girl is fictional; the history of feuds between families of mages is fictional. Yet the story’s locations display many realistic features. Both the entirely red Fuyuki Bridge and the Western-style Tohsaka residence have geographical features almost identical to those found in reality. Here we can see an important difference between an ether model and an animated work. The former simply represents a fictional object. The latter represents a collection of events: those that would occur if things that exist in reality, things that could exist, and things that could not exist all actually existed together in a particular spatiotemporal setting. Alternatively, we can treat what the animation represents as an entire “world,” and regard that world as a “candidate” for physical existence. The world of the animation does not actually exist, but if it did, it would be a concrete physical entity. [Note 7] Whichever account we adopt, representation in animated works can be explained through the conditional “if it really existed…”

An animated work plainly has a structure. By interpreting that structure, we can identify a representational relation between the work and a collection of events that is fictional as a whole. In this sense, animated works possess the general features of models. We can therefore say that animation and models are ontologically similar, and that this similarity makes it reasonable to regard animated works as a kind of model.

4. Model-Based Interpretation and Depth

In the previous section, I discussed the ontological similarities that allow us to treat animated works as models. The question now is how this model-based account can offer a satisfactory response to the puzzle of depth in animation.

Let us first return to the puzzle itself. Its substance is that our everyday intuition about depth—that depth implies a kind of epistemic superiority—cannot adequately explain why we prize this category so highly in animation criticism. More specifically, our everyday intuition requires animated works to carry some clearly identifiable knowledge-bearing content, whereas, on the usual understanding, animated works seem to have no necessary connection to such content. This produces the two principal grounds for the puzzle discussed in section 2: the low standards applied to depth in animation, and the inability to compare the depth of animated works effectively.

Since we are reluctant to abandon our everyday intuition about depth, an appropriate strategy is to use the model-based account to rethink what we regard as deep in animation. Certain epistemic features of models can connect particular content in animation with its capacity to carry knowledge. At the same time, as I have already noted, depth in animated works often emerges through interpretive practice; only through such practice does particular content come to be regarded as deep. To rethink the nature of this content, we must therefore first clarify what our interpretive practice ordinarily amounts to in this sense.

In approaching interpretation through the model-based account, we should first consider an epistemic feature of models that I call a surplus of properties. A model may contain more properties than those introduced in its construction. Or, put another way, when we model a target object, our model never exhausts all the properties of that object. [Note 8] This feature is a prerequisite for any model’s capacity to offer new insights. The surplus allows us to investigate a model further, learn about properties we did not understand when constructing it, and extend them to the target itself. In ordinary scientific practice, this feature is directly connected to the modeling process. All modeling idealizes or even deliberately distorts its target: a model entirely faithful to its target is neither possible nor necessary. Yet this partially distorted representational relation is often sufficient for the properties we wish to understand.

Once we regard an animated work as a model, we can see the same surplus of properties at work. Scientists investigate scientific models through various systematic research procedures to learn about their “surplus” properties; when watching animation, we likewise use various written and unwritten interpretive practices to uncover its “surplus” properties. In animation, this surplus takes the following form: the ideas, thoughts, and intentions a group of creators brings to the making of a work do not exhaust the ideas, thoughts, and intentions that the work itself may present to its audience. Animation is a structure open to interpretation, and the temporal sequence of its images, sound effects, dialogue, and so forth is, of course, deliberately arranged by its creators. But this does not mean that the original ideas, thoughts, and intentions behind the arrangement are the only legitimate claims about the animation’s content. Even if the creators explicitly explained in writing the metaphors and symbols behind every image, they would still only be explaining the representational relation established between the animation’s internal structure and fictional objects during the modeling process. The audience’s interpretive practice is another matter altogether. Unless we are diehard proponents of author-centered interpretation, we must acknowledge that the creators [Note 9] neither can nor need to strive, while making an animation, to set firm limits on the audience’s interpretations.

From the model perspective, then, much of our interpretive practice consists in uncovering an animated work’s surplus properties—properties its creators, as model-builders, did not usually introduce consciously into the creative process. Moreover, a model’s surplus of properties is not merely a feature of its structure; it is also manifested in the representational relation between the model and its target. We always regard the surplus properties we learn about through a model as properties of the target itself. In this sense, we can reasonably understand interpretation as an investigation of the surplus properties of animation as a model. At the same time, it makes claims about properties of the represented collection of events that were not known to us during the modeling or creative process.

On this model-based account, interpretation never speaks solely about the animated work itself. Through model-based reasoning, it also makes claims, justifiable to some extent, about what the work represents. The purpose of these claims generally has to do with one of a model’s epistemic functions: explanation.

Clearly, all models have an explanatory function. This usually arises from an isomorphic relation between model and target, on the basis of which we extend properties discovered in the model to the target itself. If we want to know why most sports cars have streamlined bodies, for instance, relevant data from a wind-tunnel model, the graph of a function in a mathematical model describing the circumstances, or even the output of a computer model based on empirical formulas can all provide an explanation. A model’s explanatory function does not, however, require every explanation to involve calculation. If reasoning about a model yields properties that can affect our claims about its target, this too counts as a realization of that explanatory function. Indeed, this way of realizing it is more common in everyday life. To explain why the arch is so widely used in buildings, for example, elementary-school general-knowledge lessons regularly use paper models of different shapes to explain its greater reliability compared with other geometrical structures.

As a kind of model, animated works have precisely this explanatory function that does not depend on calculation. In animation, a claim is usually explained through the presentation of events in the plot. An appropriate presentation inclines us to agree or disagree with that claim. A particularly clear example comes from episode 24 of Fate/Zero. Inside the Holy Grail, Kiritsugu Emiya sees that if he adheres to his creed of “sacrificing the few to save the many,” the catastrophic outcome will defeat his purpose: he will ultimately save only a very small number of people. (Many critics, however, regard this sequence as merely a variant of the trolley problem, and thus as very shallow and clichéd.) What this part of the work represents is quite clear: a certain class of events in which we make ethical choices by counting people. Like models in general, this part of episode 24 plainly idealizes what it represents. In real situations, the smaller group might include those closest to us, and the cost to those sacrificed need not always be as extreme as death. But just as every idealized model has a range of application, this part of Fate/Zero, considered as a model, is relatively successful—at least regarding the claim that “sacrificing the few to save the many is justified”—in explaining why that claim is unacceptable.

In most animation, of course, the explanatory function is not so clear. Just as with ordinary scientific models, interpreting animation as a model usually requires sufficient insight and the relevant background knowledge. Consider the widely acclaimed 1995 Ghost in the Shell. What we see is less an explanation of a particular claim than a presentation of near-future ways of life and forms of society. Yet even with such indeterminate content, animation can still be regarded as a model with an explanatory function—one that depends heavily on our interpretive practice. For example, we can treat Motoko’s experiences in the story, together with their technological and social context, as a model open to interpretation. What this model represents is the class of events constituted by our ways of treating the mind in a technologically advanced society, if such a society actually existed. Interpretation may allow the model formed by this material to serve a variety of explanatory functions. We might take this particular part of Ghost in the Shell, as a model, to explain an externalist view of the mind, or the dissolution of the mind–self under technological domination, or even the necessity of a deus ex machina, and so on. A model’s explanatory function does not limit us to explaining a single claim. As the frequent disputes over the interpretation of data in scientific models also show, interpreting animation as a model is invariably complex and plural.

Of course, what an animated work represents as a model usually includes things that are essentially fictional. We are unlikely, for instance, to accept that the Unicorn Gundam’s various mysterious powers could ever occur in reality. The collections of events represented by animated works therefore often make sense as subjects of discussion only within the context of the animation itself. In such cases, the explanatory function of these models operates chiefly at the level of principle. As with the models involved in thought experiments that posit extreme situations, interpreting animation as a model more often explains why a claim is universally unacceptable in principle than why it is acceptable in any sense.

Having clarified interpretation and animation’s explanatory function through the model-based account, we can now naturally arrive at a new understanding of what constitutes depth in animation without abandoning our epistemic intuition about depth. When we use this category to discuss particular content in an animated work, what we really mean is that this content, as a model of the collection of events it represents, serves an explanatory function to some degree through the critic’s interpretive practice.

5. The Advantages of the Model-Based Account and Views Contrary to Common Sense

In this final part of the argument, I will explain the advantages of reconstructing depth as a category of animation criticism through the model-based account. As I have emphasized throughout, one reason for proposing this strategy is our reluctance to give up the commonsense understanding of depth. Conversely, understandings of depth that depart from common sense may suggest other responses to the puzzle. I will briefly discuss and assess these views.

The first ground for the puzzle in section 2 was that, compared with the depth found in scholarly writing, we are satisfied with a low level of depth in animated works. If depth in animation criticism really does mean epistemic superiority, then, when judging animated works for their depth, we should not be satisfied with the low-level insights they provide. We can respond by considering the structure of animation as a kind of model. An animated work’s structure usually brings together real and fictional things in a complex combination. Taken as a whole, this combination is original, which means that every animated work—and even each particular part of it—can be regarded as quite a novel model.

In the 1995 Ghost in the Shell, for instance, there is a distinctive structure linking characters such as Motoko, Batou, and the Puppet Master with the work’s setting and plot and the images through which the story takes shape. Models or discussions with comparable structures do not appear in any scholarly work, and no scholarly work has therefore elaborated on them. In this sense, we can say that particular content regarded as deep in an animated work does indeed offer a unique insight through critics’ interpretive practice, however trivial that insight may seem.

This does not mean, of course, that depth in animated works can always be understood in this way. At least in some contexts, the supposed depth of certain content instead dissolves in the face of scholarly depth. Episode 24 of Fate/Zero is a typical example. As many critics point out, Kiritsugu’s experience inside the Holy Grail is really no more than a reworking of the trolley problem, and looks quite shallow beside the serious discussions found in intellectual history. From the model perspective, this criticism amounts to saying that this part of Fate/Zero, as a model, is too similar to the many models and discussions that fill existing scholarly works, and therefore offers no unique insight.

The second ground for the puzzle was that the depth of animated works cannot be compared. The main reason was that, even if particular content genuinely offers deep insights, we cannot see how those insights are justified, and therefore cannot find grounds for supporting or opposing a claim. Clearly, the explanatory function of models has already resolved this problem. As a model, animation can, in principle, explain through showing and thereby effectively incline us to support or oppose a claim. It is therefore also possible in principle to compare the depth of animated works. This depends on whether the respective classes of events represented by two works are sufficiently similar to allow comparison.

In explaining the model-based account’s advantages, we must also ask whether abandoning the intuition that depth rests on epistemic superiority might resolve the puzzle more simply. I do not think this strategy can give us a satisfactory answer.

Specifically, this strategy holds that our praise for depth in animated works has nothing to do with their particular presentations of grand themes or the insights they display. What we call depth in animation is merely a summary description of certain aesthetic experiences. This is closest to a popular view of animation: what we appreciate is the pleasure of watching the work itself, not something beyond it. However deeply a scholarly work discusses a grand theme, it is not as “good to watch” as a shallow animated depiction of that theme. Animation’s engagement with grand themes is inconsequential in itself. What really moves us is what belongs to the animation itself, such as excellent music and animation. We could even say that, provided these things were good enough, it would make no difference to our high opinion of the Ghost in the Shell series whether it depicted the dissolution of the boundaries of the self in a cyberpunk society or merely told a B-movie story set in the near future.

The problem with this strategy is that it leads us to a crude formalist aesthetics. Criticisms of the various formalist positions exemplified by Clive Bell therefore apply to it as well. It departs from our ordinary critical practice: the vast majority of critics discuss an animated work’s story, and even sakuga fans—viewers most devoted to the craft of animation—do not dismiss the value of the particular story being told for the sake of the images themselves. More importantly, this strategy cannot account for differences in viewing brought about by cultural context. Chinese and Japanese audiences may find different aspects of the same work pleasurable, while the works cherished by seasoned animation fans may differ from those popular with general audiences. These examples all show that, when watching animation, we are not merely watching “animation itself.” We inevitably interpret a work, in writing or otherwise, within a particular context, and it is through this practice that we experience what the critical category of depth describes.

6. Conclusion

In this essay, I first explained why depth as a critical category is itself a problem. Our everyday intuition that depth means epistemic superiority cannot explain why depth deserves praise in animation criticism. I then drew on Weisberg’s account of the shared features of familiar models to argue for an ontological similarity between animated works and models, establishing the legitimacy of treating animation as a kind of model. On that basis, I offered a new understanding of interpretive practice in animation criticism and of the content we regard as deep. This understanding can adequately answer the two principal grounds for the puzzle of depth. Finally, I briefly argued that abandoning our everyday intuition about depth and responding to the puzzle on the basis that “animation is just animation itself” leads to an unacceptable formalism.

Overall, treating animated works as models is not a strained analogy. On the contrary, it has a reasonably secure ontological basis. The model-based account also offers a fairly satisfactory response to the puzzle posed by depth as a critical category while preserving as much common sense as possible. Yet we must acknowledge that this strategy still has limitations, the most important being its neglect of the nature of fiction. This essay has dealt with the various problems of fiction only through a fairly natural, commonsense view. In fact, even on the relationship between models and fiction alone, several different positions exist. [Note 10]

References

1. Bell, Clive. Art. Translated into Chinese by Xue Hua. Nanjing: Jiangsu Education Publishing House, 2004.

2. Sainsbury, R. M. Fiction and Fictionalism. Translated into Chinese by Wan Meiwen. Beijing: Huaxia Publishing House, 2015.

3. Weisberg, Michael. Simulation and Similarity: Using Models to Understand the World. OUP USA, 2013.

4. Frigg, Roman. “Models and Fiction.” Synthese 172.2 (2010): 251.

5. Frigg, Roman, and Stephan Hartmann. “Models in Science.” Stanford Encyclopedia of Philosophy (2006). Online: seop.illc.uva.nl↗.

6. Giere, Ronald N. “Why Scientific Models Should Not Be Regarded as Works of Fiction.” In Fictions in Science: Philosophical Essays on Modeling and Idealization, edited by M. Suárez (2009): 248–258.

7. Godfrey-Smith, Peter. “Models and Fictions in Science.” Philosophical Studies 143.1 (2009): 101–116.