A Semantic/Syntactic Approach to Film Genre

Abstract

Most comfortable in the seemingly uncomplicated world of Hollywood classics, genre critics have felt little need to reflect openly on the assumptions underlying their work. Everything seems so clear. Why bother to theorize, American pragmatism asks, when there are no problems to solve? We all know a genre when we see one. Scratch only where it itches. According to this view, genre theory would be called for only in the unlikely event that knowledgeable genre critics disagreed on basic issues. The task of the theorist is then to adjudicate among conflicting approaches, not so much by dismissing unsatisfactory positions, but by constructing a model which reveals the relationship between differing critical claims and their function within a broader cultural context. Whereas the French clearly view theory as a first principle, we Americans tend to see it as a last resort, something to turn to when all else fails. Even in this limited, pragmatic view, whereby theory is to be avoided at all costs, the time for theory is nevertheless upon us. The clock has struck thirteen; we had best call in the theoreticians. The more genre criticism I read, the more uncertainty I note in the choice or extent of essential critical terms. Often, what appears as hesitation in the terminology of a single critic will turn into a clear contradiction when studies by two or more critics are compared. Now, it would be one thing if these contradictions were simply a matter of fact. On the contrary, however, I suggest that these are not temporary problems, bound to disappear as soon as we have more information or better analysts. Instead, these uncertainties reflect constitutive weaknesses of current notions of genre. Three contradictions in particular seem worthy of a good scratch. When we establish the corpus of a genre we generally tend to do two things at once, and thus establish two alternate groups of texts, each corre-

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 309 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.