Since the beginning of film history, film-makers and film-theoreticians have tried to compare film with verbal language in a serious attempt to justify the film medium. The main difficulty lies in the use of the term ‘language’ to describe words and images. It is very surprising that the concept of ‘film language’ has been around since the time of the earliest films (Thomas 1995; 1).
To define film and television as a language like all other literary languages might be wrong. But to suggest that film and television is some kind of a language and that they both make use of several codes and signs, in order to constitute a structure or a grammar, might be a step in the right direction. What are the actual differences and similarities? That is one of the concepts a number of semioticians have discussed and have tried to illustrate ever since the start of the work of Levi-Strauss. In this essay I’ll look into some of their theories and try to establish an understanding of what is actually meant by ‘language’ in the sense of visual images.
We know that if we want to read a literary text in any language we have to learn the signs and codes – the letters and the words – of that particular language. Therefore, to be able to ‘read’ the visual images appearing in film and television, we have to learn to understand the signs and codes of cinematic media. In my essay I’ll outline some of these codes in order to establish an understanding of the sense of grammar in cinema. Moreover, I’ll try to outline the differences and similarities between the visual and the spoken language, in order to investigate the semiotics of cinema.
In order to explore the contention that cinema constitutes a language, it’s necessary to establish what different semioticians have argued in this matter. Semiotics deals with many specific approaches to the study of art as a language.
The term semiotics…means theory of signs. Linguistics is one such theory, since it is the theory of language-as-a-system-of-signs. The semiotics of cinema is, similarly, the theory of film-as-a-systeme-of – signs (Braudy 1998; 90)
The theory of cinema semiotics can be traced back to the linguistic theories of Ferdinand de Saussure and it uses language as a method for analysis (Monaco 1981; 417). Saussure argues that language material creates meaning but that has no meaning in itself. To transfer the literary semiotic theories to the visual cinema, has proven to beprblematic, because according to Saussure, the cinema is not a conventional sign like words are. Saussure argues that in cinema the signifier is almost identical with the signified, which suggests that a visual image of a car is much closer to the actual object than the word “car” is (Thomas 1995; 2). We have been taught that the word “car” is ‘identical’ with the image of a car, but without the knowledge of the English language the word “car” is only a collection of letters, without any meaning.
Already before Saussure’s semiotics, an American philosopher C.S. Peirce developed a semiology different from the one of Saussure. Peirce identified three types of signs: The iconic sign, the indexical sign and the symbolic sign.
Let’s start off by looking at some of the theories and statements of Christian Metz, whose approach to cinema is probably one of the most thorough and interesting of cinema theories. Metz believes in the linguistic approach to cinema, but in order to justify the study of cinema as a language, the perception of language had to be redefined. Any form of communication is a language, but Danish, English and Spanish is a ‘language system’, Metz states (Monaco 1981; 157). Therefore cinema may be a language of some kind but it is not a language system. As Metz puts it: “It is not because the cinema is a language that it can tell us such fine stories, but rather it has become a language because it has told such fine stories” (Metz 1991; 47). Metz argues that film is to cinema, what a novel is to literature. Moreover, Metz suggests that denotation is to be studied before connotation. According to him, the denotation is the basic form of cinematic material, because it presents, it doesn’t interpret. Denotation is the images that make up a story. Connotation has to come second, he says, because what the images connote is not directly presented by the basic material of the film and connotation is only partly indicated by the denotation (Braudy 1998; 91).
Metz argues that there’s no unit in film that equals the word in language. The image, which he believes is the smallest unit in cinema, is already at the same level of a sentence or a paragraph. This fact leads him on to compare the shot and the word, which illustrates his strong relations to the linguistic semiotics (See appendix 1). Metz’s point is that in literature you can imagine, you can create your own visual images, whereas in cinema you can’t, because the images have already been chosen for you. For instance, not many readers of the trilogy The Lord of the Rings would have created the same visual image of Frodo, as the director of the film invented him on screen. In this context, film doesn’t suggest: it states. It puts the visual images in front of us. Metz explains that all speakers of the English language at a certain age have learnt the code of English – they can create sentences. However, he suggests that cinema doesn’t present a language that is available as a code, because the ability to create cinematic utterances relies on talent and training (Stam 2000; 111). Metz wrote: “…when a ‘language’ does not already exist, one must be something of an artist to speak it, however poorly. For to speak it is partly to invent it, whereas to speak the language of everyday is simply to use it” (Monaco 1981; 163).
One of Metz’s theories is that film is made of a series of minimal sequences which he calls syntagmas. An example: “…a series of different views of a house… might constitute what Metz calls a descriptive syntagma’– showing what the house is like and not presenting an event unfolding in time” (Braudy 1998; 92). In this way syntagmas can be identified on the basis of montage and how the montage is put together.
Another difference between the cinematic and verbal language is that the cinematic one can suddenly be shifted in a new direction, for instance, by new technology or by aesthetic decisions. Those introduced by Dogme95, for example, which created a renewal within the cinematic language by cutting away all cosmetics of cinema and present the audience with the ‘truth’.
Even as Metz concluded that cinema isn’t a ‘language system’, because he believes that it lacks minimal units and double articulation, it still suggests a systematicity that resembles that of language. In the same way as literary language expresses itself through written material, cinema expresses itself through five tracks: moving photographic image, recorded phonetic sound, recorded musical sound and writing. “Cinema is a language…an artistic language, a discourse or signifying practice characterised by specific codifications and ordering procedures” (Stam 2000; 112). Such a statement indicates that when we learn to understand cinema, we automatically learn to decode the visual images, as well.
Metz vs Wollen
Another semiotician , Peter Wollen identifies one of the problems with Metz’s theory: he puts all his effort into the denotation instead of regarding the connotation as a great part of the cinematic language, as well. Most aspects of the plot of a film – who the characters are, what they are doing – are indicated by other clues, of the plot. According to Metz such aspects are a part of the connotation, whereas I personally understand them as a part of the denotation. This is also my reason for believing that it’s necessary to look at both denotation and connotation in order to understand the film, television, paintings… any kind of art. Denotation and connotation can’t exist without one another.
Metz was criticised for being to programmatic: “A theoretical program that leads to so little in practice is clearly suspect” (Braudy 1998; 92-93). Wollen, believes in the semiotics of cinema, as well. However, instead of relying on Metz’s syntagmas, he leans towards the theory of signs by C.S. Peirce. According to Wollen cinema is primarily iconic and indexical. He says that the important codes in cinema have to do with the poetic rather than the linguistic, which leads him away from Metz’s idea of finding any unit in cinema, which resembles the linguistic language. Wollen rejects Metz’s suggestion that the purpose of cinema semiotics is “to study the ordering and functions of the main signifying units used in the filmic message” (Braudy 1998; 95). He states that cinema has both meaning and significance but it doesn’t carry messages, whereas language is often used to communicate a message. Even as Metz and Wollen disagree about the actual nature of semiotics, they still both believe that the study of cinema is a part of such semiotics. Metz appeals to the linguistic approach: the semiotics of language become the semiotics of cinema. Wollen sees the filmic strengths as different from the linguistic strengths.
Another theorist, Gregory Currie, argues that “cinematic representation… is not linguistic, quasi-linguistic or even remotely linguistic” (Currie 1995; 114). On the background of this statement he disagrees with both Metz and Wollen. He rejects all attempts to analyse the cinematic language on the background of literary language. He supports this argument with the statement that cinematic language only uses one medium: the sight, whereas literary language can be seen, heard and touched (Freeland (nd); 1). Totally contradicting Metz’s theory that cinematic language consists of five tracks, which I’ve already mentioned. He also overlooks the fact that film and television can both be seen and heard… the soundtrack is a great part of a film. However, even as he rejects cinema as a language he agrees that it is a way of communication: “There is no set of conventions that function to confer meaning on cinematic images in anything like conventions confer (literal) meaning in language” (Currie 1995; 130). Currie states that there is no smallest meaningful unit of film. To him, there’s no interaction between two shots and therefore the shots don’t constitute a language. Currie doesn’t believe that images can be put into sentence-like structures because they don’t have a grammar (Freeland (nd); 3).
Coming back to the question of double articulation… In the early stage of the debate of film as a language a question of minimal units and articulation in the sense of Andre Martine’s notion of double articulation of minimal units of sound (phonemes) and minimal units of sense (morphemes) was raised (Stam 2000; 112).
As previously mentioned, Metz disagrees with the assumption that cinema has double articulation. The cinema works with shots of a constructed reality, and doesn’t have anything similar to the linguistic phoneme. However, he suggests that montage could be seen as a kind of articulation of the depicted reality. But not an articulation that corresponds to the linguistic sense of articulation. Again he states strong relation to the linguistic approach and refers to the difference between the filmic shot and the literary word (See Appendix 1).
Through his analysis Metz doesn’t find any unit in cinema that resembles the linguistic phoneme, which means that cinema can’t have double articulation.
In stead of trying to identify the shot as the smallest unit, Pier Paolo Passolini sees the minimal unit being formed by the signifying objects in a shot, and because of this identification of the smallest unit, cinema can have double articulation. Passolini’s cinematic double articulation consists of cinemes (instead of phonemes) and im-signs (instead of morphemes). The fact that the smallest unit exists within a shot, means that the cinemes are different to the phonemes, because there is to be found an infinite number of cinemes in a single shot.
Semioticians like Eco and Garroni disagree with Passolini and they criticise him for confusing the meaning of double articulation. Eco states that objects can’t have a second articulation because they already constitute meaningful elements (Stam 2000; 113), and it is impossible to separate the meanings of an element. According to Eco there are no phoneme-like or morpheme-like units in the cinematic images, and therefore no double articulation.
Metz states that the justification of analysing the similarities between film and language consists in their common syntagmatic nature. In language phonemes and morphemes are combined to create sentences, in cinema image and sound are combined to create syntagmas (Stam 2000; 115). By moving from one image to another film ‘speaks’ to the viewers and it communicates. Metz states that no image resembles another image but most narrative films resemble one another in their structures. However, if we look at Alfred Hitchcock’s The Birds (1963) – as Messaris has done in his book Visual Literacy: Image, Mind and Reality- such a theory breaks down. For instance, in the scene where the female protagonist sits in a schoolyard, we see shots of birds gathering behind her intercut with her sitting, smoking a cigarette. Finally when she turns around, behind her is full of birds. The editing of this sequence breaks with a ‘rule’ that says that if a character is seen looking out of the frame of a shot, the next shot is to show at what the character is looking (Messaris 1994; 75).
This brings us to the question of the grammar of film and television. Are there general ‘rules’ in filmmaking in order to make the audience ‘read’ images in a certain way? In Messaris’ book he refers to Monaco’s statement in How to Read a Film, where he says that film is only “like” a language and that it’s impossible to be ungrammatical in cinema. Messaris argues that if this statement was actually true, the viewers would always be able to identify the filmmakers intended meaning and that the interpretation of the images would be independent of the structure of the film. In order to break with such statement he illustrates an example of the New Wave, which introduced jump cuts, for instance, with no great consequences for the viewers’ understanding. To explain this Messaris states:
… the viewer’s interpretation of edited sequences is largely a matter of cross-referencing possible interpretations against a broader context… rather than a matter of decoding formal devices. To put this more compactly: Interpretation is driven by the narrative context, not the code (Messaris 1994; 79).
It’s necessary to understand what the images connote in order to understand the plot. We have learnt to ‘read’ and to understand films even as they may not follow the ‘grammatical’ structure of filmmaking. Therefore, as I mentioned in my introduction, learning to decode the signs of cinema is almost the same as learning to read.
When comparing literary language and cinema we are bound to find some differences and similarities. Differences: The cinema doesn’t consist of signs that are similar to words. In cinema signifier and signified are almost identical, in contrast to verbal language. Cinema is not a language like Danish and English is and it’s practically impossible to be grammatically incorrect in cinema. It’s not necessary to learn a vocabulary. In cinema it’s difficult to identify a small unit like a word. A small number of words can produce an infinite number of interpretations. Cinema and language doesn’t have the same method of expression. Cinema does not suggest: it states. (Thomas1995; 2). Cinema can shift in different directions. In language we can choose our own images, in cinema we are presented with chosen images. We learn to ‘read’ cinema before we have any knowledge of the verbal language. Similarities: It’s – in some way – structured like literary language. It presents the receivers with a number of possible meanings. You can learn to ‘read’ cinema. Both cinema and language is a way of communicating (Thomas 1995; 3). They both exist on several levels, which rely on age, gender, cultural-, social-, and academic- background.
So, does cinema (film and television) constitute a language? In my opinion, it does constitute a form of communication, a one-way communication that differs from literary language in ways that have already been mentioned. Throughout my essay I’ve tried to establish a notion of what different semioticians mean when they refer to cinema as a language, with the main focus on the theories of Christian Metz. I’ve explored how they have all taken different approaches to this notion and I’ve discussed how they both agree and disagree on several points concerning cinematic language, which also indicates the difficulty in such a notion. It’s interesting to notice that almost every one of them goes to extremes trying to find some kind of justification of their own work. They seek to present cinema as a language so it can be seen as a respected medium that is at the same level as literary language. They are desperate to put cinema into some sort of structure and give it a grammar that can be understood on several levels, in order to justify it as a medium for study and research. However, as I’ve explored throughout my essay, this becomes difficult because the language of cinema can suddenly shift into new directions, exemplified by Dogme95 and The New Wave.
Metz’s comparison between the literary word and the cinematic shot.
November 2002