Before we go further, a few basic questions need to be asked:
1. Is language needed in meaningful discourse interaction?
No, language is not needed. Discourse can function without language, and it can be meaningful too.
2. Can narrative be conveyed without language?
The answer is yes, narrative can be conveyed without using language. This is evident in some still pictures and some silent movies.
3. Is language essential to cinema?
Language can often be found in cinema, but this does not mean that it is essential, as there are some movies that do not use language.
4. If language is not essential to cinema, how is the ‘linguistic’ analysis of cinema then possible?
If language is found in a cinematic text, linguistics can of course be used to analyse the language in the text. The use of linguistics becomes an issue if linguistics is imposed on the non-linguistic aspects of the cinematic text, which may be resistant to a ‘linguistic’ analysis. The problem however, is that this has been done rather frequently in the history of cinematic scholarship.
Informal comparisons with the grammar of language have been made early in the history of film, during the silent era, when it was believed that the universal language is found. In the words of Christian Metz (in his Film Language: A Semiotics of the Cinema, p. 51), it was asserted during this time that ‘[t]he image is like a word, the sequence like a sentence, for a sequence is made up of images like a sentence of words, etc.’ Such comparisons have also been made by the director Raymond Spottiswoode, as clearly indicated in the title of his book, A Grammar of the Film.
More formal usage of linguistics for the analysis of cinema have been made in formalism, structuralism and semiotics. For example, in the structuralist approach of Metz, cinema is regarded as a compound of many languages. Among the cinematic constituents chosen in a linguistic approach to cinema is the shot. The problem however, which is seen in some of the works of Metz, is the linguistic equivalence of the shot: is the shot equivalent to the word, the utterance or the sentence?
The attempt to look for linguistic equivalents does not stop with the shot. There have been earlier attempts, for example, by the Russian formalists and by the Czech structuralist Jan Muka?ovský (also spelt Mukaovský), to use the terms syntax and semantics for film. The use of these terms for cinematic analysis is still with us. Further use of linguistic concepts for cinema have been made by the Italian film director and writer, Pier Paolo Pasolini, who looked for the phonemic and morphemic equivalents for cinematic discourse.
The use of linguistics for the analysis of cinema can be seen in various semiotic approaches (for a discussion, see the next page of notes). One approach that could be mentioned here is that of the Moscow-Tartu Semiotic School, which has a more considered approach to the relationship of film with language than some of the earlier approaches. According to an idea which they further developed from the Russian Formalists, and which can be seen in the work of Yuri Lotman – a prominent member of this school – there is a difference between what they call the primary and secondary modelling systems. Language is a primary modelling system which shapes our perception of reality and the world at large, whereas artistic ‘languages’ are secondary modelling systems, which may depend on the primary modelling system of language, but are quite distinct from it. These artistic ‘languages’ work at a higher level of abstraction than language; in other words, they may be abstracted from language, but they are not identical with it.
The underlying assumption in many of the ‘linguistic’ approaches to cinema can be related to what is known as the Sapir-Whorf hypothesis: that language shapes our vision of reality. This hypothesis is controversial. To say that language ‘shapes’ other semiotic systems, which hence have similar structural patterns to language, is debatable. To say that those semiotic systems that have been shaped by language are themselves ‘linguistic’ is to make an even more contentious assertion.
Among scholars who have related cinema with linguistics, Metz is perhaps the most prominent. His early scholarship placed a strong relationship between language and cinema. Metz later became more critical of the relationship between language and cinema. For example, he criticised the relationship between
To Metz, the difficulty of associating the linguistic word with the cinematic shot is based on the fact that shots, unlike words, are infinite. Another difference is that shots are created for the film, unlike words, which pre-exist their use in linguistic discourse. Shots also have more possibilities of meaning than words, but in spite of this, their references are more localised than the references of words. Furthermore, shots, unlike words, do not really work on the paradigmatic plane. Although Metz earlier suggests that the shot can be associated with the sentence (the notion of sentence in the French original may be closer to the notion of utterance), he questions this association in his later work.
Despite his strong views of the association of language with cinema, Metz further qualifies his view about the association of the cinema with language in his earlier work by stating that although cinema is a ‘language’, it does not have a language system (pp 44-9) – ‘langage sans langue’.
In spite of his criticisms, Metz continued to associate language and linguistics with cinema in his later work. Among the assertions and claims that can be criticised in his later work are:
Metz’s notion of the five cinematic tracks is useful, but we notice here that only two of the tracks are fully linguistic (2 and 5 below):
Recorded moving image
Recorded linguistic sounds
Recorded non-linguistic sounds
Recorded musical sounds
Saussure has been influential in cinema studies and in semiotics in general (more discussion on Saussure's influence on semiotics in the next document). But the Saussurean model is not the only linguistic model used in film studies. Other models apart from Saussure's have been used. One prominent linguistic model (or set of models) is that associated with Noam Chomsky and other generative grammarians (refer to the print versions of John M.Carroll's “A Program for Cinema Theory” (1977) and Warren Buckland's The Cognitive Semiotics of Film (Cambridge, 2000) for some approaches that have used the generative model for film analysis).
Among the concepts used for cinematic analysis are surface and deep structures. The problem here is that quite a lot of other phenomena work on the basis that what appears on the surface may not be exactly the same as what can be found beneath the surface, in spite of the fact that the former may depend on the latter. Also, it is not the case that if surface and deep entities occur in other phenomena, they can be related to similar processes in linguistics, or that these processes are actually dependent on linguistic processes or on our understanding of these processes. In fact, generative grammar rules can be applied only at the level of the sentence, and there are difficulties if we try to apply them to discourse as a whole, even if the discourse under consideration is made up entirely of language.
Another problem with the generative model is its division of sentences or phrases into grammatical or ungrammatical ones, sometimes described as sentences that are well-formed or ill-formed. The question of grammaticality in this model is a yes–no judgment with no judgments lying in between. Thus all and only grammatical phrases and sentences are included in the generative grammar of a language, and all fuzzy and semi-grammatical examples are excluded. The problem here is that judgments of grammaticality are not always universal, even within a particular speech community which speaks a common dialect of the language. Also, even with reference to the judgments of individual users of a language or dialect, judgments of grammaticality are not stable, and are not always a question of yes or no. Many sentences or phrases in a language, can be graded on the basis of what is more or less acceptable, and there is a cline of grammaticality, which becomes even more evident when we consider language change. Thus the concept of grammaticality or well-formedness is problematic even within linguistics, as not all linguists agree that clear-cut notions of grammaticality should be used as the invariable basis for the comprehensive grammar of a language, and trying to apply it or related concepts to cinema will certainly create further problems.
Another generative concept that that has been applied to cinema studies is that of competence, which is contrasted to performance: competence is the idealised understanding of the grammar of a language, whereas performance refers to actual language use. Like the ideas of ‘surface’ and ‘deep’, the concepts of competence and performance have general applicability, and one need not think of them as being strictly based on the corresponding concepts in linguistics. In fact, the concepts falter if one thinks of them strictly in terms of linguistic competence and performance. In this regard, it has been noted by Ehrat (2005 p. 176) that ‘There is no cinematic competence… corresponding to language competence in cinema and a corresponding grammaticality; there is only performance.’
Finally, the concept of the modularity of mind, which is associated with the Chomskyan approach to generative grammar, may actually militate against the use of generative linguistic concepts in non-linguistic domains. The concept of the modularity of mind hypothesises that the human mind consists of autonomous innate modules that perform specialised functions. Thus the modules involved in language learning and in linguistic production and comprehension, for example, are not transferrable, and cannot be related to the non-linguistic modules present in the mind. Thus generative linguistic concepts should not be applied to non-linguistic domains, even if they can be unproblematically used for language.
Ultimately, in the discussion on the use of linguistics for cinematic analysis, we should ask if the linkage is merely metaphorical or analogical (Coëgnarts 2017), and if there is an underlying reason for its usage. Silent movies may be helpful here in establishing or interrogating this link: except for its appearance in inter-titles, language is obviously less important than in sound movies. The visual element continues to be important after the advent of sound. As this is the case, why should linguistic concepts and models continue to be used for the overall analysis of cinema? In relation to this question, it is also useful to look at how a written original is converted to a film, and how the importance of language is lessened as a consequence, and concomitantly, what is retained or removed in the conversion. Language, and hence linguistics, become less important in the movie when compared to the written original.
© 2010-2018. Last revised: 11 August 2018.