Review: Narrative Structure in Comics by Barbara Postema

Barbara Postema's new book, Narrative Structure in Comics, joins a nice list of books on comic theory to emerge this year, and it provides some interesting analyses and insights. The book is clearly written and should be readable for most audiences.

Upfront, I found it unclear who the book is aimed at. If this is an introductory book or for a popular audience (perhaps suggested by the appendix of basic terms at the end?), some of the terminology and ideas should be more carefully laid out and should reflect or acknowledge the consensus of the field. If this is a scholarly tome, some of the background literature should be more carefully acknowledged and debates discussed in detail.

For example, quite pertinent to my own interests, she quickly dismisses analogies between the structure of comics and the structure of language, stating that “…images communicate largely without rules… the smallest elements of images have no set meanings, and the way these elements are combined or even repeated are not governed by rules like grammar” (p. xvi). Viewpoints like this are repeated throughout the text, stated as accepted knowledge or as obvious, despite the absence of a scholarly consensus on the issue or any evidence to support it. (Compare this to Miodrag's recent Comics and Language—review coming soon—which spends 200+ pages discussing this argument!)

In fact, comparisons have been made along the lines she disparages since the 1970s within "semiotic" approaches to comics—the viewpoint she takes throughout (see my review "Comics, Linguistics, and Visual Language")—which go unreferenced. I assume her argument is directed against these researchers' theories, but it's hard to know without references or augmentation. My own work highlighting parallels between language and comics with a contemporary linguistics/cognitive viewpoint also go undiscussed and unreferenced throughout. It's hard to know whether such omissions are by choice (then: why?) or out of ignorance (which I doubt).

This again makes it unclear who the audience is: are readers supposed to take such flippant claims as reflecting the consensus of a field? Are they supposed to know that they reflect an opinion of a particular group of people (or at least, this author), but that there is an ongoing debate about this issue?

Clarity in this regard would aid the book on the whole, even if made as a note in the introduction, and would certainly aid me in reviewing it. I'd be more harsh on the book if it's meant as a scholarly treatise than if it's meant as an introductory book. I'll strike a middle ground here.

Theoretical contribution

This issue of audience permeates my main problems with the book, which is that I did not see much novelty in the theoretical contribution. Her broadest argument is that every level of structure in comics involves "filling in gaps" of information that go unprovided in their representations. This applies to the inferences about objects that are made by viewing only the simplistic forms of cartoony images as well as to the inferences created about a sequence by viewing only what is depicted in panels.

Certainly, no one denies that inference occurs in the understanding of sequences of images (or individual images), and indeed many people have said this, including McCloud, Saraceni, myself, and a host of others writing about comics (and other types of narratives!). Just how her theory of inference goes beyond others in any substantive way remains largely undiscussed. Is this a reflection of the book being aimed to an introductory audience? It's hard to tell.

A comparison may be useful: Saraceni's dissertation was extremely detailed, technical, and precise, only to be overly-simplified in the "textbook" version that was widely published as The Language of Comics (his articles fared better, striking a nice balance in tone). I haven't read Postema's dissertation, but such a simplification may be the case here as well.

In some ways, this book feels like a meshing of McCloud's focus on inference with a desire to simplify, clarify, and in some cases go beyond Groensteen's ideas, but aimed at a wider, less academic English-speaking audience (thankfully with less opacity than the original). For example, her contribution to discussing page layouts is merely to categorize the ways in which panels might be physically juxtaposed: do they have big gutters, no gutters, inset panels, etc? This focus on the physical is comparable to Groensteen's taxonomic treatment of how word balloons relate to panel borders (do they touch, do they overlap, etc). Such analyses at least provide a way to cut up the taxonomic space of comics, but they don't tell us much about their understanding. They're just a list of things in comics—and not even an all that interesting or insightful one!

Also, I certainly don't believe that all of these variations carry the intrinsic aspects of "meaning" that Postema seems to argue that juxtapositions in layout carry with them. For example, she makes much of the difference between when a single line separates panels versus having a larger gutter. However, when people first start drawing comics (for example, as kids and often adults), they usually just use single lines to separate panels. Do the "meaningful" "significations" of a gutter versus a line maintain in these, non-published works as well?

Semiotics vs. cognitivism

This example highlights the nature of a significant portion of my critique of the theory, no matter the audience. That is, the dissatisfaction with the book's orientation within a "semiotic" paradigm—a trend I'm finding in many recent works on comics and which I do not consider to be useful for looking at the formalist elements of this medium. These ideas draw from linguistics concepts from the 1930s until 1950s, principles of which were borrowed by the social sciences and humanities for looking at many other non-language domains. However, the actual study of language (i.e., linguistics and psychology) largely abandoned this approach in the 1960s in favor of ideas that attempt to describe principles of cognition, for which those "semiotic" ideas were not useful and only described a surface structure. Since that time, cognitive science research has developed enormously in ways that are insightful for describing language and beyond.

Though Postema appears to argue against a structuralist idea of language, she still maintains the broader framework of semiotics. There are many places where notions from contemporary linguistics or cognitive science would provide more convincing frames for the analyses of comics she makes (which, admittedly, are often insightful and interesting despite their non-cognitive framework). For example, Postema's discussion of Jacks Luck Runs Out by Jason Little nicely highlights how the story draws inspiration and structure from the rules and aesthetics of playing cards. However, her discussion remains at a surface level, stated in terms of connotation and denotation as if such principles carry with them explanatory power of "understanding."

In terms of cognitive science, this comparison would be covered by the theory of "conceptual blending," which illustrates how one domain of knowledge can creatively draw from another domain of knowledge. Blending theory would actually be beneficial for many of the examples discussed throughout. Granted, Postema's insights would likely remain much the same if framed in terms of blending, but framing them as such would allow her to probe deeper issues at work in this analysis, as well as to connect such a discussion of comics to how we understand the workings of the mind and brain more generally (and, dare I say it, language!).

As it stands, the analysis remains at surface level, without casting much light on what might be going on "under the hood" in the actual understanding of this material. This is especially true of the broader claim that readers "fill in gaps" when reading comics: Just what is creating this inference if not the mind/brain? And if it is the mind/brain (which I doubt Postema would deny), then why not actually target those principles guiding such processes?

Here is where the semiotic approach is most clearly lacking from cognitive approaches. It is important to remember that the "comic medium" itself does not actually contain these structures, as purported by the semiotic approach. Rather, the human brain is what instills and interprets such structures within the graphic form, so that is where our research should be focused in such matters.

Corpus analyses of comics

I stumbled across this interesting project called the eBDtheque database which has started coding various comic pages to create a searchable corpus. The full project is online here, while their paper describing the project is found in this pdf.

This rather large research group looks like it's actually been doing quite a lot of computational approaches to analyzing comics, which is very cool.

Besides recording background information about the authors, publishers, etc. of the comics, this database in particular seems to take comic pages and code them across three main dimensions. 1) How many panels per page, 2) how many balloons per page, 3) how many lines of text within balloons.

While this coding scheme is fairly limited, they have said that they'll be expanding it to look at other dimensions, such as the angle-of-viewpoint and filmic shot type. What makes the project fairly impressive though, is that the data in this corpus is not just human coded, but also involves extraction from the computers themselves. All that makes this a project worth watching.

Naturally, I'd love to see this type of project be developed further. My two cross-cultural studies so far comparing panels coded from Japanese and American comics are hints of what having a corpus will allow.  I have a research library so far of roughly 4,000 comics and graphic novels that is ripe for research to be done on them. The compelling idea here though is not just doing isolated studies. Rather, the idea would be to create a massive database that could be searched in many different ways for various inquiries across many dimensions. In fact, I had talked with programming savvy friends about starting such a project years ago.

Once my own actual lab is up and running, I'd like to start on a project like this by recruiting students to help with the coding. However, another way to build a large database is to "crowdsource" the work. This would involve having people across the world complete "training" sessions to become proficient at the theories and coding schemes involved, and then they could login and code comics they have around them.

This would help with creating a large, internationally diverse corpus available to many scholars to use, and would be created by people across the world. Imagine being able to do large scale searches across millions of comic panels from across the world in order to do comparisons about the structure of visual language!

Methodological advice for comics scholarship

In my previous post, I drew a distinction between the type of scholarship often done by people who are creators of comics (i.e., fluent in the visual language they study) and scholarship by those who lack this fluency. I argued that there are certain things that non-fluent scholars are unable to recognize, and that such impairments can significantly limit their viewpoint on studying the structure of sequential images.

However, there are ways in which these limitations can be overcome and significant contributions can still be made. Much of this comes down to methodology (important also for those who are fluent creators as well), which is a more important critique I often have about scholarship on comics:

1) Be specific — Sweeping generalizations about "how comics work"—especially those based on specific exceptional examples—are almost of little contribution in understanding the structure of the visual language. Papers in linguistics or psychology almost never try to generalize to "all of language" or "all of cognition," but rather pinpoint a specific structure and then rigorously detail it. Scholars of the structure used in comics will be far better served by finding a very specific instance of something of either 1) theoretical or 2) empirical interest and then rigorously investigating it.

1.1 - Things of theoretical interest might be a particular idea or theory about the structure of sequential images, such as "how panel transitions (don't) work" or "how panels structure space" which can then be rigorously examined across many comics—both "normal"and "artistically exceptional." It always amazes me that so many scholars subscribe to McCloud's theory of panel transitions, yet practically none of them follow his methodological lead in trying to apply those transitions in quantitatively in the way that he tabulated transitions in various comics. Every time I did this, it directly poked holes in the theory of transitions.

1.2 - Things of empirical interest would be seeing a phenomenon in a comic, and then rigorously analyzing how it operates in other books (if it does) and exploring the theory behind why it might occur. A good example of this method is Abbott and Forceville's examination of why some character's hands turn into stumps in a manga. They noticed a curious phenomenon, then detailed why they thought it—specifically—happened.

2) Manipulate and test — Following the need for specificity, you should then manipulate structures and test them on people. If you aren't fluent in the visual language and can't just manipulate it in your mind (or even if you can...), manipulate things based on assumptions of your theory, and then give those creations to people who are fluent.

For example: I've often seen people claim that "sequences in comics can't just be strictly linear" or "sequences/narratives can't be governed by a rule system like a grammar because nothing constrains the sequences in such a way." Yet, not one of the papers I've read that claims this ever provides actual evidence supporting such critiques (compared to, say, my psychology experiments which directly test this and say otherwise).

Here's what they should do with the most simple of manipulations: take a sequence of images that they claim "isn't constrained" and then scramble the images into random orders. If every possible order makes sense, then they might be right that there are no constraints. If even one order doesn't make sense, then 1) there are constraints, 2) their critique/hypothesis is wrong, and 3) it then behooves them to follow up and figure out why that one order doesn't make sense (i.e., figure out the constraints!).

Even if you don't have the intuitions to create the sequences by drawing them, you can still manipulate existing sequences based on a thesis (again, of a specific phenomenon), and should have the intuitions to notice if what you did has an effect. Similarly, you can show many manipulated sequences to other people and start to derive data that tells you which manipulations might be behaving in different ways.

...and congratulations, at this point you're doing science!

Comic creation expertise and theory building

When I first started entering into discussions about visual language research, the fact that I actively created comics seemed like an important point that I would often stress. While I am a little more forgiving about non-creators doing theoretical work these days, I think there are some interesting trends that fall out of whether a person actively has created comics or not.

People who create comics, especially at a professional level—like Eisner, McCloud, and myself—tend to focus on the fairly normal, banal, and ubiquitous elements of the medium. This research usually emphasizes regularities, showing how things are systematic in unexpected ways, and/or providing taxonomies to better understand the pervasive and everyday elements of this visual language. 

On the other hand, people who do not actively create comics tend to theorize or discuss about the exceptional, unusual, and "rule-breaking" aspects of comic creation. They discuss the most artistic and poetic creators, and their examples usually are the most "interesting" ones out there. They are also the most likely to disparage the idea that regularities do exist in visual sequences.

This distinction makes sense to me. To people who are not actively immersed in the visual language, the things they find most interesting are the things that break the rules. Normal, ubiquitous elements of the system are just that: normal. They are less interesting because they are commonplace and less noticeable. These people also have less intuitions for the nature of the regularities, being unable to manipulate such structures in their own minds (thereby giving the illusion that no such structures are there).

For actual creators, the banal is highly interesting, because it allows us to understand and articulate the basic process and the intuitions that we have while creating. Our scholarship can be informed by how we actually think, as opposed to just what we see on a page far removed from the creative process. The medium is not a static, received thing, but rather a dynamic, created process.

I feel that this distinction is fairly important. As I've said before in my "advice to aspiring theorists," there is a danger in placing too much focus on the exceptional examples. Exceptional examples can be useful for revealing what makes them exceptional—and thereby highlighting how they contrast from the normal. However, building a theory entirely around the exceptional will leave out the normal, and it will ultimately be left unable to deal with the most basic aspects of the medium.