from Rick Altman, “A Semantic/Syntactic Approach to Film Genre”

The difference between semantic and syntactic definitions is perhaps at its most apparent in familiar approaches to the western. Jean Mitry provides us with a clear example of the most common definition. The western, Mitry proposes, is a "film whose action, situated in the American West, is consistent with the atmosphere, the values, and the conditions of existence in the Far West between 1840 and 1900." Based on the presence or absence of easily identifiable elements, Mitry's nearly tautological definition implies a broad, undifferentiated generic corpus. Marc Vernet's more detailed list is more sensitive to cinematic concerns, yet overall it follows the same semantic model. Vernet outlines general atmosphere ("emphasis on basic elements, such as earth, dust, water, and leather"), stock characters ("the tough/soft cowboy, the lonely sheriff, the faithful or treacherous Indian, and the strong but tender woman"), as well as technical elements ("use of fast tracking and crane shots"). An entirely different solution is suggested by Jim Kitses, who emphasizes not the vocabulary of the western but the relationships linking lexical elements. For Kitses the western grows out of a dialectic between the West as Garden and as Desert (between culture and nature, community and individual, future and past). The western's vocabulary is thus generated by this syntactic relationship, and not vice-versa. John Cawelti attempts to systematize the western in a similar fashion: the western is always set on or near a frontier, where man encounters his uncivilized double. The western thus takes place on the border between two lands, between two eras, and with a hero who remains divided between two value systems (for he combines the town's morals with the outlaws' skills).

Now, in passing we might well note the divergent qualities associated with these two approaches. While the semantic approach has little explanatory power, it is applicable to a larger number of films. Conversely, the syntactic approach surrenders broad applicability in return for the ability to isolate a genre's specific meaning-bearing structures. This alternative seemingly leaves the genre analyst in a quandary: choose the semantic view and you give up explanatory power, choose the syntactic approach and you do without broad applicability. (10-11)

We need to recognize that not all genre films relate to their genre in the same way or to the same extent. (12)

Far from postulating a uniquely internal, formal progression, I would propose that the relationship between the semantic and the syntactic constitutes the very site of negotiation between Hollywood and its audience, and thus between ritual and ideological uses of genre. (13-14)