Evan Cameron's recourse to analytical terms derived from linguistic philosophy demonstrates some of the interesting things that can be said about film genre using this conceptual vocabulary. I have also been working on a related recourse to linguistic analysis in order to clarify some of the confusions that arise around genre names and their application, and this work perhaps has some use for talking about film noir. Philosophers since Russell have emphasized that there are two quite distinct ways of picking out things in the world. We can *describe* them, or we can *refer to* them, where a *description* gives specific characteristics, properties or qualities which may or may not be fulfilled by various things, and where *referring* picks things out by their spatiotemporal location. (This is Russell's famous Theory of Descriptions.) It can be shown that the two methods of picking out objects in the world are distinct, because, although it is quite possible to do both at once, it is also possible to do one without doing the other. I.e., we can refer to something without giving any description of it ("that thingamabob"), and w can also give a description which picks out *no* object in the world ("all those things which are both circles and squares"). A further distinction appears in the fact that the objects which fit under a given description are potentially infinite, whereas we cannot *refer to* an infinite number of things. These distinctions shed a certain light on the use of genre terms. The traditional distinction between *genres* and *cycles* corresponds in part to the distinction between *describing* and *referring*, as the latter category is spatiotemporally limited (a gangster *cycle*, e.g., was made during a certain period, at a certain studio, etc.). Descriptions such as "All those things which have the properties of being films with a certain kind of lighting, plot structure, characters, etc." pick out a *class* of films which is potentially infinite, whereas, when we talk about "certain films made during a certain period by certain directors, studios, scriptwriters, etc." we are *referring* or pointing to a particular group of objects which is finite. Thus a confusion arises when a "definition" includes both ways of picking out films, since the likelihood of there being any *new* films made in the late 1940's is quite slim indeed, whereas it is quite possible that there could be a new film next week which fit under a *description* meant to characterize film noir. These distinctions also helps us clarify the analytic project of studying genre, which involves starting from particulars and extracting from them a description. We can move from specific films (*Out of the Past*, *Double Indemnity*, etc.) to a description, but then the description becomes something quite different--not simply a description of specific films, but a potentially open-ended description of an infinite class of films. Further, multiple descriptions are possible for the same set of objects. Humans can be described as "featherless creatures" or "bipeds" or "animals" or any number of ways, although none of these pick out *only* humans. Similarly, genre names (such as "film noir") when understood as descriptions do not necessarily pick out *one* class of films distinct from all others (such as "the gangster film", "the crime melodrama," etc.) but may regroup films already understood as falling under different descriptions. Thus it is no surprise that Houston or Tourneur, e.g., had no idea of "the film noir" in their minds, as genres-as-descriptions do not need to be understood as being pre-given in advance, but can rather be constructed later by various critical projects. The implication of this analysis is that there are as many genres-as-descriptions as there are terms we can use to describe films, that genres-as-descriptions are created after the fact by a critical operation of description, and that the number of genres is potentially as infinite as the number of descriptions. Similarly, in this sense an individual film, once described, would be a genre which might only have one member. But there is no need for genre analysis to start from particular films and to abstract a description. We can perfectly well reverse the procedure and start from a description and look for films which fall under it. We can thus imagine genre-descriptions which have *no* members, thus bringing a certain poetry to the rather prosaic study of genre. We could, for instance, take Laura Mulvey's description of classical Hollywood cinema and invert the genders and arrive at a description which did not pick out *any* film (or if it did, it would be quite an interesting one!). The point of such a project would be that it would shed light on what kinds of films *are* made by showing what kinds of films are *not* made--an impossible film, a square circle. The project of genre studies would thus involve an examination of what meanings are possible and impossible during a certain period. Such "impossible" genres are not as fanciful as might be imagined: at one time the description "female buddy road movie" might have been thought of as having no members, but since *Thelma and Louise* (and some of its imitators), it becomes possible to retrospectively reconstruct a genre which might include *Gentlemen Prefer Blondes*. If this vocabulary indeed constitutes any kind of clarification, it is merely conceptual: that is, it does not speak to the appropriateness of this definition or that one, whether of film noir or of any other genre, but merely proposes criteria by which genres names can be understood as picking out films, thus underlining the difference amongst *kinds* of descriptions and their related critical projects. Edward R. O'Neill UCLA ---- To signoff SCREEN-L, e-mail [log in to unmask] and put SIGNOFF SCREEN-L in the message. Problems? Contact [log in to unmask]