Sunday, February 08, 2009

The Pictorial and Linguistic Features of Comic Book Formulas

Neff, William Albert. 1977. The Pictorial and Linguistic Features of Comic Book Formulas. Doctoral Dissertation, University of Denver, Denver, CO.

Neff compares the patterns of formal properties for comics in different narrative genres of Adventure, Romance, Mystery, and Alien Beings or States. He analyzes the fields of panel shape (vertical, horizontal, square, circle), angle of view (lateral, high, low), and type of shot (close, wide) in comic panels. He also looks at pragmatic sentence types and parts of speech of stressed words from the text of comics.

His results show distinctions in the "formulas" of different genres, showing that different genres do use differing patterns for their distributions of these fields, and he then describes his interpretations for what those forumlas indicate about the genre (and vice versa).

While I don't doubt that such patterns for genres exist, this study had numerous problems. The categories for analysis were a little broad (only wide and close shots?) and often washed over in coding (diagonal panels were grouped as either horizontal or vertical). The interpretations of the genres' formulas also seemed a little like just so stories.

However, I take this entire study with a grain of salt because the sample size of his analysis is so small. For each genre, he uses only two comic books (pamphlets). While he does get statistically significant results using chi-squares, he pools frequencies across books, which eliminates any variation across books with no way to analyze it. Shouldn't he be using averages for this?

Two books per genre, and limited categories in the fields of analysis, are far too little to really get a sense of the patterns of an entire genre. His total number of panels in all was only about 530. In comparison, I consider my study comparing 300 panels in each of 12 Japanese and 12 American paperbacks (Cross-Cultural Space) to have been small in scope, and have just initiated a study of at least 200 books of varying genres and countries.

While I greatly appreciate the attempt at doing such corpus analyses (and am actively doing more), and especially like seeing it as a "hidden treasure" in the history of this type of study, this one unfortunately lacks the scope to be taken seriously.


Anonymous said...


It seems like you're highlighting the inadequacies of Neff's work (which I'm not familiar with), but it doesn't seem like you're acknowledging that it's relevance has most likely been diminished by time if anything. You're talking about a 32 year old dissertation which was most likely focusing on what was in production and available at the time. That most likely rules out Manga, and –I may be wrong on this– is going to be examining work that was less influenced by the now omnipresent ideology of post-modernism. I'm guessing Neff's work was very topical and thorough for it's time, but when you consider how much means of production / distribution / and discourse have changed and advanced in the past 32 years (longer than we've been alive) I could see how his work would need to be reexamined and updated in a way that is relevant to our current state of affairs.

It sounds like you're doing exciting work which I'd enjoy seeing, and when we're both retired and seeing each other at our 50th high school anniversary we can have a chuckle about the irrelevance of your dissertation. ;)


Neil Cohn said...

Thanks for the comment! For its time, the thought of doing this sort of analysis seems pretty good, but even then the execution should have been better.

I certainly didn't mean to imply that he should have been looking at manga (which is my own comparison of interest). Rather, for the comparison he wishes to make his sample is just way too small. He described going to a newstand to get the pamphlets, where — at the time — it should have been easy to get more than two comics of each genre.

Only two is not a large enough sample to really say anything, especially if you're not taking averages, which means that one book could be driving all of your numbers. One or both may not even be an accurate reflection of the genre, but you'd never know because of the way the study was carried out.