No one believes that film is literally a language or that there are literal film languages. So that is a red herring. But film philosophers since the 1990s have been attacking that idea that film is importantly like language. One of the most prominent articles in this areas is Gregory Currie's "The Long Goodbye: the Imaginary Language of Film" first published in 1993 and appear in an anthology I am currently using in a class: Philosophy of Film and Motion Pictures (Blackwell, 2006). Noel Carroll in his Introduction to Part II of the book endorses Currie's attack on semiotic theory, saying that his critique is "devastating." Although I think Currie makes some good points I hardly think that his critique is devastating. It has, however, led me to some serious rethinking about the nature of language itself.
To introduce my comments I think it is useful to refer to a passage in Carroll's discussion of Curry. One of the most telling points is that film, unlike a standardly accepted language as English, has no grammar. This leads Carroll to a claim, based on Currie's premises, which is so outrageously false that the entire enterprise of Currie and Carroll is thrown into question. The passage in question is in a paragraph about poetry. "English has a grammar, but English poetry qua poetry does not. What is right as poetry is not what abides by strict grammatical conventions, but that which, when executed, proceeds in cognizance of how things are usually done, but which nevertheless works with or against traditional procedures in order to move the reader in the way the poet intends. If an infinitive needs to be split, so be it. This is not poetically ungrammatical. For poetry has no grammar. Ditto film editing." (62). On this view, since language requires grammar and poetry has no grammar, then poetry is not language, or rather, not in language. Sure, he qualifies this by saying "poetry qua poetry" but it is not clear how the "qua" functions here. So neither film not poetry are language, nor are they in language. This is so deeply wrong, one has to question the foundations of the entire enterprise. And if contemporary Philosophy of Language entails this then there is something deeply wrong with it too!
For a more thorough analysis we need to turn to Currie himself. (I recognize that Carroll has written on this issue at length, but my focus here is on Currie.) Currie claims that film is not in any way interestingly language-related. But this is dependent on his understanding of natural languages like English. His view of English (which is, I grant, a standard one, and perhaps close to universally accepted by philosophers and maybe also by linguists) is that it is characterized by productivity, conventionality, recursiveness (meanings are assigned by convention to a finite stock of words and we combine meanings by rules of composition), and it must be molecular (which seems to me to mean the same thing as "recursive" except that it is a metaphor that understand the sentences as molecules and the words as atomic.) Then he says what I have found shocking (closely related to the implication by Carroll that poetry qua poetry is not in language).
"Since the atoms - words in English - are assigned meanings individually, and since the composition rules make the meaning of the whole a function of the meanings of the parts, we can say that meaning in our language is acontextual." I believe this to be wholly false. The meanings of words used depend on where and how they are used: meaning is contextual. (This is my extremely controversial claim.) Currie must be referring to some abstract thing, called "language," that we actually do not use in daily life. Currie's big point is that meaning in a film is contextual and, since acontextuality is is required for the presence of language then film cannot be like or involve anything like language. I grant that meaning in film is contextual, but so too is meaning in language! The meaning of an English sentence (not some ideal sentence floating outside of real use, but a sentence as used) is as contextual as a meaning of a sequence of shots in a film.
Then it turns out that Currie (and others like him...they are legion) holds that when we speak of utterance meaning then meaning is contextual. (I see this as damage control: a concession to the real world, although, as we shall see, a very weak one, since it the only context it seems to require is assuming the rationality of the speaker, i.e. a Davidsonian move.) The problem he thinks is that film theorists are confusing semantic (also called "literal") meaning and utterance meaning. My claim is that there is no so thing as semantic or literal meaning. It is a fiction. What is the literal meaning of Currie's sentence "There are several reasons for this, and I shall mention just one of them." (95) Is it a set of dictionary synonyms for each word in sequence? We cannot speak of anything like literal meaning here unless we think of the context in which it appears. The separation of literal or semantic meaning and contextual meaning is bogus.
The problem, I think, is a kind of Platonism. Plato believed that real reality was acontextual. The meaning of the Forms do not depend on context.
Consider Currie's discussion of the sentence "Harold is a snake." He correctly observes that it may mean in some context that a human named Harold is scheming. There is of course an ambiguity in the phrase "this same sentence is used." Sure, we can use "sentence" both for a set of words in sequence that can appear in different contexts and one instance in which that set of words appears. The second is the ur-meaning of "sentence." The other is just an abstraction and it is noteworthy that "sentences" in that sense have no meaning. Currie however says that the sentence "Harold is a snake" does not mean that Harold is scheming. He believes, it seems, that the meaning of the sentence is in someplace like the world of Forms. But that is not how language works. Language is a living, breathing thing. Language is how we actually communicate with each other. If I say that "Harold is a snake" in the context of intending that a human named Harold is a schemer this does not depend in any way on the claim that Harold is the name of an actual snake. The linguistic idealist like Currie (my label) seems to think that the true (utterance) meaning of the sentence (as it is used) depends in some way on the ideal meaning. It does not. Semantic meaning is a myth. It stands in place for the fact that we have certain models in our minds of how words can be used in certain contexts, models that are usually reflected fairly well by dictionaries (although dictionaries also have some normative force.)
So it is false that "the meaning of a particular utterance depends in part also on the meaning of the sentence uttered" if the meaning of the sentence uttered is taken to be the semantic meaning! I do not deny that there are conventions, dictionaries, and grammar books, and that these serve their important purposes mainly in helping us to communicate, conventions being the most important since we could certainly have language without the others. Semantic conventions do play a role in determining meaning.
The problem is in how these facts about conventions are used. Currie seems to think that utterance meaning is gained simply by adding to conventions of semantic meaning certain "non-conventional rules of rationality" ones that help us determine the intended meaning of the author. I do not want to go into all of the questions surrounding the notion of "intended meaning" which, itself, is something of a myth. But a quick look at the debates over that will show that determining the meaning of a sentence, for example in a literary work, or even in a philosophical writing, depends as much on the context of words that come before and after that sentence as on anything that can count as the "author's intention" which is notoriously nearly impossible to pin down independent of the immediate textual context itself. I say this without denying that interviews of the author or, in film, the director, can be immensely useful in shedding light on the meaning of the novel or film.
I would go further and bring up something mentioned by one of my students, that films should be seen as like novels, upon which they are often based. Neither films nor novels are languages as such, but language plays an important role in each, it is just that novels only use language. Films of course usually use a lot of language, literally English for example, in filmic dialogue. But in film we need to recognize that whatever the characters say is inextricably connected with the behavior of the actors who are playing those characters, including various gestures. When in Citizen Kane Orson Welles talks he is not just reading a script. The meaning of each thing he says is contextualized and modified not only by what came before and what comes after but also by the movements of his body, the intonation of his voice, and so forth.
Currie asks "does any of the story meaning that cinematic images convey possess the communicative features that we have attributed to the meanings of words and sentence?" This is a misleading question. They do communicate in the way sentences do, but this very often involves (and certainly should not exclude) communication by actors portraying characters using a natural language among other means of communication. It also includes the other surrounding contextual features which film semioticians often refer to as "the language of cinema." That is they include close-ups, tracking shots and so forth. This too is part of the overall way that directors and other key players in the film-making art communicate to us. Actors do part of the job but directors direct the whole enterprise. Doesn't it make sense that the whole package is language since the goal is communication?
Currie notes that George Wilson has taken what I consider to be the reasonable position that interpretation of a film needs to look at it holistically. (94) Currie then says "Wilson is right to say that what the cinematic images tell us about the story depends on the surrounding context of other images. But that is true also of words and sentences in a text, where there is no dispute about the presence of language. What kind of relation between described events is suggested by one bit of text depends upon the role that bit of text is seen to have in the context of all of the other bits...So the context-dependence of interpretation applies to literature as much as to film": which all seems completely right to me. But this undercuts the idea that language is irrelevant to film as it also undercuts the idea of the acontextual nature of language.
Currie imagines defenders of cinema language saying that we should not conclude that the "meaning intrinsic to the images themselves is contextually determined." I agree: this would be a bad move since it relies on the very same linguistic idealism that Currie unfortunately accepts. The problem is not, contra Currie, a matter of mixing up semantic and utterance meaning: the problem is with taking semantic meaning as something real. The defender of cinematic language should just insist that there is no acontextual meaning either in literature or in film, or anywhere else for that matter.