Linguistics and Film and Theatre Practices

Abstract

In this chapter, the concept of predication—borrowed from linguistics—is applied to film and theatre languages. The premise is that the rules that govern the establishment of predicative relationships in the sentence can be useful in identifying meaningful communicative segments of film and theatre discourses.

This section attempts to identify meaningful units of film language and the principles of their integration by looking at organisational paradigms of natural language and their corresponding cognitive realities in the deep structure. However, this endeavour is by no means premised on the idea of the equivalence in functioning between natural and film languages.

Christian Metz divided semiotic film studies into two stages. The first benefits from linguistic approaches, while the second stage “is properly semiological and translinguistic” (Metz 1974, p. 61). Metz views Saussure’s term le langage as suitable for film, while the term la langue is restricted to natural language. He believes that some linguistic concepts might be applied to film language but “with a great measure of caution” since film language does not possess the quality of double articulation: it lacks discrete units, and it has neither phonemes nor words. Nevertheless, Metz argues there are certain similarities between the two languages. He likens the film image to the sentence in the sense that the film image possesses the assertive status and quality of actualisation (p. 67). Metz attempted to discover the underlying codifications that govern the functioning of film language and the creation of meaning. He looked for the underlying codes and disregarded the external elements of the image that are infinite and constantly changing. His discovery of the underlying principles of the combinations of film images led to the grande syntagmatique theory. He identified eight film syntagmas, eight types of image ordering based on the criterion of time and space. Despite the criticism that his approach was too abstract, Metz situated the exploration of film within the film text itself, each text displaying a specific combination of codes. Metz’s shifting the focus of film explorations to the underlying rules that govern the workings of various codes gave an impetus to film studies that use linguistic methodologies to discover the film’s “universal grammar.”

Noam Chomsky’s Transformative Generative Grammar—the Standard Model—has been used in film studies to illuminate the processes of the transformation of the “universal grammar” from the deep level onto the surface level—the film screen. In the same manner in which the speaker selects grammatical combinations, due to language competence, a film director also selects and combines film units by resorting to the rules of the film language competence and performance. Spectators decode film discourse relying on their knowledge of filmic codes. Unlike the speaker of the natural language, the film director lacks a precise set of grammatical rules for the selection and combination of film language units. Rules, such as “always follow a long shot with a close–up,” “never cut more than 180 degrees,” or “always show an object moving continuously—don’t jump cut,” are not equivalent to linguistic grammatical rules but rather “proscriptions or prescriptions addressed to filmmakers” (Worth 1981, p. 56).

Normative grammar is only one aspect of linguistics and the least appropriate for film language analyses. Of greater advantage are the linguistic theories focusing on the cognitive and imaginative processes of the human mind and their association with symbolic language activity. The symbolic language activity is the surface manifestation of the hidden performance of the human mind. Film and theatre language, like any other creative human activity, is inseparable from these cognitive and imaginative activities. Any discourse, film, performance, or literary text is the outcome of these activities of the human mind. Hidden behind each discourse is cognitive structures of the deep structure. Explorations of different origins endeavour to illuminate the stages between thoughts and their surface expressions. To be transformed into a surface discourse, deep cognitive structures have to be organised and recognisable for the collective addressee. The addresser has to use the language accepted by the collective and to make the message communicable. It could be a symbolic linguistic system or a codified film or theatre language. In any case, cognitive structures at the deep level ought to be matched with their correlatives at the surface level. However, this transformation is not a straightforward process.

Chomsky’s Standard Model identifies the root node in the deep structure, which is then through various stages of the preterminal nodes transformed into the terminal nodes on the surface level. These transformations should be conceived as a dynamic process in the course of which thoughts are formed. It means that the creation and the expression of thoughts are a simultaneous process. Therefore, ideas, emotions, or thoughts are not completed at the deep level but are rather in an amorphous form that obtains its definitive shape only at the surface level.

Warren Buckland, in The Cognitive Semiotics of Film, argues that Chomsky’s transformational generative grammar “represents the ideal paradigm for cognitive film semioticians” (2000, p. 18). Cognitive film semioticians seek “to combine film semiotics and cognitive science with the aim of modeling film competence” (p. 25). Buckland also points out that “film theory has reached a relatively mature stage thanks in part to attempts to develop a film grammar based on Chomsky’s early theories of TGG” (p. 109). Buckland rejects the criticism of Chomsky’s Standard Model—that it lacks “psychological reality.” He emphasises that “the role of theory in semiotics is to make visible the underlying, non-perceptible system by constructing a model of it” (p. 7).

The methodology of lexical semantics proposed by Jurii Apresjan illuminates some important stages between thoughts and expression in natural language (Apresjan 1973, pp. 274–300). Apresjan assumes the existence of semantic or conceptual language, namely the existence of a general language of thoughts, translatable into natural language. Therefore, it is possible to express the same idea in different ways, for various expressions have identical content at the conceptual level: “The production of sensible sentences can be regarded as a process of translating from the semantic language into a natural language, and understanding sentences can be regarded as a process of translating from a natural language into the semantic language.” He concludes that “every thought has a single, standard representation in the semantic language” (pp. 275–6).

According to Apresjan, the meaning of the word does not correspond, as a rule, to the elementary semantic units of conceptual language. Word meanings can be divided into elementary senses which correspond to the elementary units of conceptual language. They are named distinctive semantic features. The words of natural language are frequently composed of two or more elementary senses. The elementary senses such as “self,” “someone,” “have,” “cause,” “cease,” “begin,” or “not” could each form one word, or one single word could be composed of more elementary senses. Also, one elementary sense unit can be part of different words. Thus, these elementary senses can be found in different Russian words: “vladet” “own” = have; “obladat” “posses” = have; “brat” “take” = cause oneself to have; “davat” “give” = cause someone to have (p. 276).

Apresjan singled out elementary situations which correspond to the elementary sentence types. In the stages of transformation, an elementary sentence might transform into a more complex one, but regardless of how complicated the sentence is on the surface level, it is reducible to one or several elementary sentence types corresponding to elementary situations. These situations might be: “disturbing a state of physical equilibrium (the wind rocks the house); causing the movement of an object toward the subject (iron is attracted to the magnet); situation of disturbing state of emotional equilibrium (the mother worries, is glad, about her son); situation causing negative emotions (he irritates me)” (pp. 295–6).

In Dependency Syntax Theory and Practice (1988) Igor Aleksandrovič Mel’čuk uses very similar concepts. He aims to uncover the deep-structure relationships underlying the combination of lexical units—wordforms. To demonstrate “which items are related to which other items and in what way,” he rejected the categorical syntactic symbols such as object, subject, noun phrase, and verb phrase, but instead he operates with single, concrete, not abstracted lexemes (p. 14). In the Meaning-Text Model, Mel’čuk established the relationships between the fundamental set of meanings and a set of texts. In this sense, any natural language is a logical device by virtue of which a “correspondence […] between the set of meanings and the set of texts” is created (p. 43). This model partly includes the reality-meaning model, which is at the same time the subject of various scientific disciplines, such as philosophy, psychology, logic, or cognitive science. A unified discipline which would explore the reality-meaning model has not yet been established. Its structure is extremely complicated and at this point is still premised on the presuppositions only (pp. 46–7). The model comprises “the discrete cognitive representation of observed continuous reality” (p. 47). According to Mel’čuk, there is “the interaction between the cognitive representation, the internal thesaurus of the speaker, the pragmatics of a given speech situation …” Therefore, the deep semantic structure that underlies each utterance consists of the “discrete cognitive representations,” which are matched with the corresponding linguistic symbolic units. The cognitive representations are embodied on the surface linguistic level (p. 47).

It is not difficult to set up parallelism with film communication. Film representation also emanates from the reality-meaning model or Apresjan’s conceptual language of thoughts. The difference is that in film communication, the director matches deep cognitive representations with the predominantly visual elements of film language on the surface level. The spectator, like the addressee of a linguistic message, endeavours to discern the deep structure hidden behind the one manifested on the film screen.

Heterogeneous reality might be reduced to a finite set of objects and elementary situations. The predominantly visual representation of film operates with infinite numbers of elements—objects from reality or generated reality—that might be reduced to a finite number of sets of objects. It means that objects from reality belong to a limited number of classes of objects. Natural language reduces the infinite number of objects from reality into a limited number of symbolic categories to organise human experience. The visual language operates in a similar manner. To organise visual experience the human mind classifies various objects into classes of objects. The mind at the same time perceives and classifies the elements of reality. It means that the visual human experience is, like symbolic language activity, also organised and systematised. To express ideas or emotions, the film director endeavours to find the appropriate visual correlatives for the deep-level cognitive structures.¹ Since communication presumes both encoding and decoding, the spectator decodes the film message in the opposite way, discerning the hidden deep-level cognitive structures from the surface level of the film image.

An adaptation of an episode from The Fortress for the screen might illustrate this correlation. The film director has to choose among the elements forming the visual lexicon to transform the cognitive representations of the deep level into a film image. Elementary situations of the deep level have to be matched with surface visual representations. Thus, an elementary situation of the following type—expressed in natural language as Spring in the town, merry children’s game, I don’t even notice all that, I am sad, obsessed with war memories, and the uncertain future—could be represented on the film screen in different ways. These could be conveyed by a combination of visual or auditory codes. Along with a choice of objects to be represented in the film image, the director also has to decide which film modalities² will be used to set up the relationships between the objects as well as to personalise film discourse. The choice of how to represent and actualise an elementary situation on the surface level—on the film screen—ought to be grounded in the principles reflecting the basic laws governing the functioning of film language. The assumption is that, on the surface level, the aesthetic message of the film ought to convey a maximal quantity of information most economically. The given structure should represent the most adequate organisation of the given content.

Buckland elaborates on another aesthetic criterion in relation to Christian Metz’s Grande Sintagmatique: the eight categories of GS “represent a standard grammar of film, one that generates well-formed sequences, and in relation to which deviant (or semi-) syntagmas can be comprehended” (p. 130). If the spectator can associate the deviant syntagmas “with one or more of the eight syntagmatic types of the GS,” the spectator can still “construct/comprehend a coherent narrative world.” This “additional processing effort” by the spectator “results in an aesthetic pay-off” as the spectator “must work harder to construct a coherent narrative world” (p. 131).

Film language, unlike natural language in ordinary use, is subjected to another set of restrictions which derive from the aesthetic code. In film and theatre languages the cognitive structures are directly transformed into the surface level, lacking the mediating stage of the symbolic linguistic system. However, it is not a completely straightforward process due to the existence of film and theatrical modalities, which function as filters in the transformation of the deep into the surface structure. Unlike the linguistic system, theatre and film modalities are less conventional and imposing. Such filters are more permeable, which in turn permits certain cognitive representations to be directly represented on the surface level. Conversely, natural language transforms thoughts, subjecting them to rigid syntactic rules. This is partly due to the conventionality of natural language as opposed to the iconicity of film and theatre languages. On the other hand, film and theatre diverge in the degree of iconicity of their signs. Since the objects on stage are “too real” in their material manifestation, a certain degree of abstraction is inevitable in theatre. The mobility of the theatrical sign is also relevant: one sign could easily replace another in theatre—an actor could become a prop or one part could metonymically replace the whole.

The syntactic analysis proposed by transformative-generative grammar reduces the unlimited number of sentences of one language into a limited number of sentence models. Actual lexemes are substituted by symbols. Symbols are the result of the process of abstraction, for they represent a class of words, the number of which is limited as well as the number of their mutual relations. It is possible to formalise potential syntactic relations between the lexemes of a given sentence. Elements of the sentence, or word forms, according to Apresjan, are in a hierarchical order.

It is important to discover which word forms are in direct syntactic relation and which are not. For instance, in the sentence: “in the windows glimmered the azure cold light of morning,” “glimmer” is the predicate which is in direct relation with the subject “light” and the object windows. “Azure,” “cold,” and “morning” are in an indirect relation with “light” and in a direct relation with “glimmer” (Mel’čuk 1988, p. 146).

Similar hierarchical relationships are detectable in the film image as well. Since film language operates with fragments of reality and not with sentences, the general notions of logic are more appropriate than the narrow grammatical categories. In this manner, Fillmore’s definitions of predication and argument are of more use than grammatical notions of subject and predicate. According to Fillmore, the predicate identifies “some property of an object or some relation between two or more objects,” while “the arguments of the predicate” are “objects concerning which predicate asserts something” (Fillmore 1968, p. 2).

Predications are conceived as the relationships between the arguments—objects—in the film image or on stage. Film image lacks a precise semantical determination—which a lexeme has—because an image can convey a wide scope of meanings transcending narrow denotative references. Because of the concrete nature of film image, the scope of meanings is inevitably narrowed, which in turn restricts the meaning of the image more than the lexeme does. Paradoxically, a wide range of choices of possible objects in film image narrows and determines its meaning. This paradox is easy to explain, for the image corresponds to the sentence rather than to the lexeme. An image establishes the context, the same as the sentence, while a lexeme is only an abstract element of linguistic vocabulary unless actualised in a sentence.

To be able to provide a semantic description of a verb and to identify its compulsory arguments and their roles, Fillmore selected a finite number of arguments and their roles. It is possible to discern similar relationships in film image. One-argument predication—such as Pa = TALL John or SLEEP John—has its equivalent in, for instance, a close-up shot of a human face.³ Each close-up of one object represents one-argument predication if it expresses a certain important characteristic of the depicted object.

If there is more than one human object in a film image, this could correspond to the two-or-more-place predication in natural language. For example, P a, b = HIT Harry, Mary, is a predication with two mandatory arguments, similar to the film image consisting of two objects involved in a certain kind of semantic relationship.

To define predications in more general terms, their meaning ought to be disregarded and only their function taken into consideration. Fillmore proposed a limited number of relationships between arguments and predications. They are of a general nature and demonstrate elementary judgements about the occurrences around us. These role types are: “who does something,” “who experiences something,” “who benefits from something,” “where something happens,” “what it is that changes,” “what it is that moves,” “where it starts out,” and “where it ends up” (p. 15).

Fillmore also singled out the categories: Place, Instrument, and Agents as possible functions performed by the arguments of predications. Such semantical and functional descriptions of a verb define its mandatory and facultative arguments and their functions. For instance, in “John hit the window,” the predication “hit” demands Place and Instrument while Agent is only facultative. However, if Agent is identified, Instrument is no longer mandatory. In film image, Instrument has to be seen because film image is concrete. If a window was hit, the image has to show who hit it and with what. However, an image might display the outcome of an action only. Thus, an image of a broken window corresponds to the passive construction in natural language.

On the other hand, certain predications in film image might be presented by the activation of the function of Instrument only: an image of a gun or a knife invokes the predication to kill.

Fillmore’s classification aims to create a lexicon in which each lexical item is defined by its most relevant syntactic, semantic, and phonologic characteristics, along with the instructions for how to interpret them. Using this model, the classification and structure of elementary film predications could be established. However, film lacks a finite set of lexemes. Instead, it operates with an infinite number of elements—objects—from reality, eligible to become the content of film image. However, possible objects of film image are easily reduced to the finite number of sets of objects and their mutual relationships. It means that each object in film image is a member of a certain class of objects. The number of objects is infinite while the number of classes of objects is limited. Natural language in the same manner classifies the heterogeneous reality into a limited number of classes of objects. Accordingly, the presented linguistic method could be applied to film language as well. Such analysis aims to discern in film language similar rules of a very general nature governing the process of establishing relationships between the objects of film image. It would, in turn, illuminate the functioning of film modalities.

The above-mentioned methodology of linguistic origin ought to be accommodated to the specific nature of film language. Linguistic material is static while that of film is dynamic. The logical relationships between predications and arguments have to achieve a dynamical dimension to be operative in film language. First of all, film language does not operate with sentences while the methodology proposed by Fillmore relates to sentences in natural language. There are some substantial differences between sentence and film image. There is only one predication in a clause and a certain number of arguments. The film shot is where predications should be sought. Shot is the basic unit of film language for it bears the dynamic dimension, which is the essence of film art. Shot is frequently composed of several predications and a large number of their arguments. In both sentence and shot there is a hierarchical relationship between predications and arguments. Unlike sentence, which is premised on linear ordering, film image is perceived simultaneously. However, this is relevant only from the point of view of perception. From the functional point of view, the space organisation of film image corresponds to the ordering of the sentence. Each possible position of objects on the screen has its value based on several parameters. In film image, similar rules govern the position of objects to the rules of the positions of subject or predicate in a sentence. The camera-distance modality (front, middle, or backplane), namely the position on the deep axis of the screen, or left or right from the screen centre, assigns a specific value to each of the variables. Since there are frequently several predications in one shot, the spectator scans the screen and focuses attention on certain spots only. The addresser of the film discourse, chiefly through the modality of camera movements, deliberately leads the addressee to establish predications between various objects. This process operates by virtue of film modalities. The modalities regulate the organisation of screen space. The position in the front, middle, or backplane; the size of an object in regard to the frame, camera lenses, angle, distance, and movement; special effects; and lighting—all these modalities perform the function of the semantisation of the screen space.

Frequently, there are several predications within one single shot. They are subject to the rules of hierarchical ordering, which means that all predications in one shot are not of equal importance. Film modalities regulate this process. The modalities act as grammatical elements of natural language. The same rules apply: “Word forms which occur one next to the other are not necessarily related to each other syntactically, and word forms which occur at a considerable distance from one another may still be related to each other syntactically” (Apresjan 1973, p. 147). The well-known shot composition in Citizen Kane illustrates the hierarchy of predications: in the deep focus we see Kane as a child playing in the snow, while the kitchen scene plays in the front plane.

The predications within one shot could be in active or passive relationships. In other words, one predication might or might not affect the other. It would be useful to set up precise criteria for the distinction between the two kinds of relationships. Since film image depicts reality—or computer-generated reality—there are certain objects within film image with only a representative function. This means that the predicative relation is not set up. The object stands for itself, failing to express some of its relevant attributes or to display the predicative relationship with other objects.

To illustrate this concept, let us imagine the elementary situation of the following type: the active interaction between two arguments of a predication, between two agents. This elementary situation could be represented on the surface level (in a shot) as a man and a woman in an active interaction—love or dispute. Since predications are defined by their function, the semantical relation would not be considered. Apart from the active participants of the image, passive objects are also its constitutive elements. For instance, a telephone, curtains, a chair, or a sofa performs a representative function, being props and semantically neutral. If conceived in this manner, they have only denotative meaning. However, one way to measure the aesthetic value is a lack of redundancy: none of the elements of a discourse should be redundant. In this view, objects within the film image are subject to rules of aesthetic choices. An object displays only denotative semantical value if it is not integrated into a predicative relationship. In other words, if it is not in some kind of meaningful relation with other objects or if it does not display some of its meaningful attributes which have an informative aesthetic value it is redundant. If the above-mentioned telephone starts to ring, the curtains move back and forth in the wind, the chair casts a shadow of strange forms by virtue of lighting or, if the woman embraces the back of the chair, hides behind the curtains, and the man drops the phone—the objects of the image achieve the predicative status. The process of establishing predications is premised on the mutual inclusion of the objects into meaningful relationships.

Conversely, if the director does not activate the semantic potential of the objects by their inclusion in a predicative relation within one shot or with the objects of other shots, they are, from the aesthetic point of view, redundant, which in turn impoverishes the image. In Hitchcock’s Notorious, a simple cup of coffee is embedded in several important predicative relationships within one shot and a sequence of shots: Alexander urging his wife Alicia to drink coffee; Alicia picking up the poisoned coffee and drinking it; the mother-in-law doing her embroidery; and Alicia holding her head in pain. Frequently, the camera is essential in activating the semantic potential of an object to create a predication. Objects themselves lack the capacity to set up mutual relationships. Similarly, in natural language lexemes (units of lexical meaning) are accompanied by grammatemas (units of grammatical meanings such as suffixes or prefixes) to convey complex meanings.

Accordingly, a film image rich in terms of its representative power, composed of all sorts of bizarre and amazing objects, is not necessarily semantically rich. If the predicative relationships between objects or within the object itself are not activated, such an image is aesthetically poor. The elephant pushing its trunk through the window into one Belgrade apartment and the monkey walking down the street, as in Emir Kusturica’s Underground (1995), are objects not included in meaningful relationships with other elements. They don’t have a strong aesthetic potential. Even if they connote certain meanings such as the Balkan menagerie, they are semantically feeble and in discordance with their representative potential.

Two-place predications presume the predicative relationships between two objects (Fillmore 1968). Their relationship could be both active and passive, symmetrical or asymmetrical. If one of the objects is passive while the other is active, the relationship is asymmetric. Frequently, this is the case with inanimate and animate objects. There are numerous possible interrelations between the animate and inanimate objects in film image. For instance, if inanimate objects are in the front plane while the animate protagonists are in the backplane, such a constellation bears a specific informative value: commonly, it implies the dominance of inanimate objects over people and their destinies. In regard to the hierarchical order of the predications of the image, the above-mentioned predication is frequently the most dominant among the simultaneous predications of film image. It might be called the primary predication. In Jasmila Žbanić’s film For Those Who Can Tell No Tales (2013) one primary predicative relation is established: that between the protagonist and the inanimate object—the bridge. The old bridge over the Drina River is the central “character” of the film. The bridge tells the story of the people who were killed and whose bodies were thrown into the river. The inanimate object becomes the film’s central predication. The camera keeps returning to the bridge as the plot of the film unfolds. The camera focuses on the details, the patina on the stone fence of the bridge, the time-inflicted wounds on the surface of the fence, and the protagonist, the Australian tourist, as she reads the message the bridge transmits. There is no difference between the inanimate and animate objects of the film image. Inanimate objects are frequently active constituents of predications.

If one object of a predication is active, while the other is passive, the relationship is asymmetrical. If they are both active, the relationship is symmetrical. For instance, a man and a woman talking or making love is a symmetrical predication.

A passive relationship has its validity only in cases where the passivity has a semantic potential. Otherwise, objects of such an image are redundant for the lack of a predicative relation. However, if the objects expected to be in a predicative relationship are both passive, the mere passivity could have an informative value. This is because of the created tension between the expected and the realised. For instance, two objects expected to be in an active interaction, let’s say two men, are depicted in the same time-space continuum as the only participants in the communicative situation, but any kind of communication between them—both verbal or visual—is missing. Since the objects of such a predication are not passive in the deep structure, for they are eventually in some kind of a meaningful relationship, the trace of which is deliberately hidden on the surface level, it cannot be considered a passive relationship.

In the practicum section, I will propose two simultaneous predications (stage configurations) in the adaptation of the novel The Island. The participants in the configurations, Ivan and Katarina, will often display passive predications on the surface level, but the intensity of the dramatic situation stemming from the deep structure will be even more emphasised by the seemingly passive predication on the surface level. Therefore, each predication in film image is essentially active. The entirely passive relation, both between inanimate and animate objects, lacks the semantic potential, being props and semantically redundant. In Chomsky’s trace theory, when an element of the deep structure is deleted on the surface level in the process of the transformation, its mental trace is present in the mind of the speaker. On account of the speaker’s language competence, the deleted element is restored (Buckland 2000, p. 126).

Frequently, a group of objects in film image is perceived as one object. If in a shot of a group of people, none of the constitutive objects are visually or acoustically emphasised, the whole group is one predication, since all objects express one, common characteristic unique for all of them. In Quo Vadis, Aida?, the soldiers forcibly usher a large group of men into a factory hall. We see slits on the wall revealing pointed machine guns, and we hear the sound of shooting but do not see the men being killed: the sound of the machine guns becomes the dominant modality in Žbanić’s aesthetic choice of avoiding graphic scenes. The men are treated as one predication, the sound of machine guns “informs” of the killing by establishing the predicative relationship between the group of soldiers and the group of victims.

Film image is commonly composed of several predications. The rule of a very general nature might be singled out: the more important predications are usually in the front plane, while those of lesser importance are in the backplane. However, by means of film modalities, the semantic potential of the objects in the backplane might be activated, for instance, by the size of the objects relative to the size of the frame, or by virtue of camera lenses or camera angle. In the film Dom za vešanje (Time of the Gypsies Emir Kusturica 1989) in the shot consisting of a large turkey in the backplane and of the protagonist Perhan and his grandmother in the front plane, the turkey is much larger than the other objects relative to the size of the frame. In this manner, the predication of a higher order is imposed and all objects of the image are brought into a predicative relationship with the turkey. The turkey is no longer a one-place predication—which expresses an important quality of an object. It becomes a constituent of the more complex three-place predication, along with Perhan and the grandmother. The size of the objects dictates the decoding of the meaning of the image. It brings mythology, one more level above the representational one: the fantastic realm of the film.

Film modalities as typical units of film language achieve new functions. This is especially prominent in the invention of new methods to activate the semantic potential of the objects not in the dominant position on the screen. It is a source of constant creativity and dynamism of film language as a whole.

The hierarchical relationships between predications in film have their counterparts in natural language. Syntactical relationships in natural language such as predicative, non-predicative, co-ordinate, and subordinate are to be found in film language as well (Apresjan 1973, p. 185). A typical relationship of subordination in film is, for instance, when landscape dominates the objects of film image. The predication, such as the quality of the landscape, influences the perception of other predications. Thus, the predication “gloomy, grey sky” dominates the other predications by virtue of film modalities. The size of the landscape could be emphasised through, for example, camera lenses, to activate the depth of the image. The other predications are subordinated: people in the front plane are dominated by the quality of the landscape.

The co-ordinate relationships between predications are those in which no single predication is dominant. For instance, a long shot of a restaurant, where people sit at separate tables, talk, eat, and drink quietly, does not establish a dominant predication. They are co-ordinated. However, since film is a dynamic medium, it might be expected that one of the predications is selected as dominant in the next shot, or the whole predication—the quality of the restaurant—might be drawn into a relationship with a predication of the other shots.

Fillmore’s analysis is concerned with the determination of the number of mandatory arguments of a verb. The comparison with film is possible to a certain degree. There are, both in film and in natural language, certain one-place predications which are, in fact, two-place. For instance, the verb “married” implies two arguments, but in a sentence is actualised as a one-place predication. It is considered, according to Fillmore, to be “suppression of, or failure to mention” (p. 11).

Film displays many possibilities to exploit the above-mentioned predicative relationships. The camera might show the predications where one of the mandatory arguments is missing. That argument could be revealed later in another shot. This is the source of dynamism between the arguments and their predications and between the predications themselves. It activates off-screen space, which is closely related to the concept of film predications.

The predication of the type: “a woman watches” is a two-place predication, but the second argument might be omitted. It has a special purpose: to focus the attention on the off-screen space to set up a dynamic relationship between the visible and the potential film space. It often serves to increase the dramatic effect. It functions in natural language in a similar way: “For certain predicates, silence (‘zero’) can replace one of the arguments—expressions just in case the speaker wishes to be indefinite or non-committal about the identity of the argument” (Fillmore 1968, p. 83).

In the film Kuduz (Ademir Kenović 1989), the murder scene is observed from the outside: the exterior shot of the camper and the diegetic sounds of the commotion inside. The two-place predication—the man murders the woman—is only partially represented. The deliberate omission of the arguments of predications in film—which might be both verbal and visual—is typical for the murder mystery genre. The murderer, the agent of a predication of the higher level, is often not revealed till the very end.

The source of the dynamic relationships between the objects is of two kinds: the first derives from the constantly changeable relationships within one shot, while the other is related to the creation of predications between two or more shots. In the former, the predicative relationships are set up by, for example, the position on the screen, lighting, by camera modalities—lenses, movement, direction, angle, etc. The constantly changing predicative relations within one shot can also establish higher-level predications within a single shot. Such complex predications are subject to the law of spatial and temporal continuity of the shot.

The relationships between objects of successive shots are subject to different rules from those operating within a single shot. Two consecutive shots might keep the same parameters of the communicative situation—participants, theme, space, time—or alter some or all of them. Metz’s second model of eight syntagmatic types, described in Film Language, is based on similar parameters. For example, the following criteria are relevant for the typology of the eight types of the film syntagmas: space or time continuity versus disjunction, simultaneity, parallelism, linearity, autonomy, ellipsis, chronology versus achronology, alternation, non-diegesis versus diegesis, etc. Metz uses the criterion of “the unity of action” for the sequence shot, and for inserted shots the criterion of “the cause of their interpolative nature,” which includes all kinds of logical relationships such as comparison, purpose, explanation, as well as shifts in fictitious levels—dreams, memories, fantasies, etc. (1974, pp. 121–133). Metz attempted to reduce an infinite number of film sequences to the basic underlying models. However, the most common criticism of GS is that it “leaves out some avant-garde films.”⁴

Various predications between different shots are set up on similar principles, but the concept of predication is more inclusive. First, it should be determined which of the parameters are the same for both shots and which are changed. What matters is the organic relatedness between the two shots. In other words, at least one predication from one shot ought to be somehow linked with one or more predications from the other shot. It could be the repetition of the theme,⁵ of one meaningful object, or the space-time locus, the continuation of the same parameters from the communicative act, etc. Even if the integration of at least two predications between two shots never occurs, this very act must have a certain meaning that spectators can decipher based on their film competence. Montage plays a chief role in the process of establishing predications between objects of various shots.

A systematic analysis of various shot sequences would probably reveal the laws which govern the creation of predicative relationships between various elements of each shot. It might also confirm the assumption that the degree of congruence between the form of film image and its content could be taken as the measure of its aesthetic value.

To discover the most general rules governing the formation of predications, their hierarchical organisation, and the setting up of the bonds between the predications of one shot with the others of different shots, it is necessary to redefine the very notions of both shot and sequence. The methods elaborated in distributional linguistics could be applied if adjusted to the unique nature of film language units.⁶ Metz also emphasised that “the work of film semiotics, in the vast majority of cases, is inspired more or less directly by models of structural or distributional linguistics …” Such an approach would “call […] attention to the regularities, co-occurrences, oppositions, recurrent ‘motifs’, progressions, etc. …” (qtd. in Buckland, p. 119).

First, it is important to discover the basic constant elements of the shot as well as its variable components. The objects of reality that compose the shot are variable. In the same manner, in which natural language units are defined by their relevant and constant features, film units could be so defined.⁷ Therefore, everything which is changeable and not mandatory for the definition of the shot should be left out. The shot ought to be defined in terms of its constant components. The shot is characterised by the three planes of composition: frame plane, depth plane, and geographical plane (Monaco 1977, p. 158). The position of an object in regard to these three planes has its value. Each of the positions represents a certain value in the film language system: front plane, backplane, left, right, in the centre, up, or down. The central position on the frame, unless neutralised by the performance of film modalities, has a higher value than the other positions (left, right, up, down). However, the value of the position is relative since one isolated element of film language obtains its value only when integrated with other elements. Therefore, the central position is not an a priori established value but always depends on the given realisation of the film image.

There are some constant laws—unless modified by the exercise of film modalities—which determine the value of objects in regard to their position on the film screen. They are defined as conventions of depth perspective: convergence—the image is getting narrower in-depth; density gradient—it is getting denser in-depth; overlapping—the objects overlap if they are behind each other in the depth plane (Monaco 1977, p. 158).

Unlike the frame, the shot is a dynamic category, for it is defined by one camera movement. The three planes of composition achieve dynamic dimension during the length of the shot. They constantly change. The objects change their position, new objects appear, and some disappear, moving along the depth axis or to the left, right, up, or down.

The boundaries of a shot are determined by the cut. The cut is a purely relational sign in film language for it carries the information that one shot ends and another commences. It has the same function of segmentation as the white space between the words in natural language.

The distinctive feature of the shot is the time-space continuum in which the camera makes one movement. At different points of this continuum, various predications and arguments might be established, but regardless of at which point they are included in the shot, they are subject to the laws of time and space configuration determined by the shot. For instance, if within one shot the camera follows a woman walking into a room and then, through the window of that room, it shows a man approaching the house, both predications are simultaneous. The camera might, in the same shot, depict two actions which follow each other, but both actions are subject to the same time-space continuity of the shot, which is the point of orientation in the time-space locus of the shot. The spectator is aware of the time-space homogeneity of the shot while at the same time perceiving two different shots as a separate time-space organisation. Between two shots there is no mandatory time-space continuum. In other words, two successive shots depict fragments of both continuous and noncontinuous reality. The mechanical laws of film discourse segmentation are not operative at this level. Therefore, to define film sequences, it is necessary to consider the categories of context, previously excluded from the definition of the shot. The segmentation of natural language discourse is conducted in a similar manner. Units larger than the sentence demand contextual analysis, while units of the lower levels such as sentences, lexemes, morphemes, and phonemes could be defined without considering the content. Since the sequence is a communicative category, similar laws govern both the formation of the speech act in verbal language and the film communicative act. Each speech act is defined by certain constant elements in regard to the context. These elements are the sender, the receiver, the message, a specific time span and location. The alteration of one of the parameters implies a new speech act (Fillmore, pp. 65–103).

Since in film language, the verbal code is not a necessary element for a communicative act to be formed, it is more appropriate to use the term communicative act instead of speech act. The predominantly visual communicative situation has its specific organisation different from a verbal one, yet similar laws underlie both communicative situations.

The terms locutionary source, the producer of the speech act, locutionary target or addressee, place of the locutionary source, place of the locutionary target, and time of the locutionary act (Fillmore, p. 67) are useful for the definition of the film communicative act. However, the different nature of film visual communication should be considered. Speech acts are predominantly defined by the verbal component. The verbal element of the film communicative act is only one among others and is sometimes missing. Therefore, if redefined for the specific film communication, any incitement for the formation of relationships between objects of the film image, both verbal and visual, is considered as the locutionary source. The object to which it is directed is then a locutionary target. For instance, the film image depicts a man looking through the window down the street. He is the locutionary source for he sets up the communicative relation with the object—the street—which is the locutionary target.

It is necessary to redefine the notion of the communicative act as it implies communication between people, excluding inanimate objects (Morris 1971, p. 92). In film, animate and inanimate objects could be in a predicative relation, which has a specific meaning and thus qualifies as a communicative act. In Jasmila Žbanić’s film For Those Who Can Tell No Tales (2013), the old bridge over the Drina River is frequently in the front plane, visually dominant, while the animate protagonists are often missing. The whole image is dominated by the predication of the bridge while the human participants in the predication are subordinated. In the film, the shots of the bridge are a one-place predication, conveying complex meanings. Thus, an inanimate object achieves a more important communicative status than an animate one. The question is whether a locutionary source can be an inanimate object.

The question is not whether the communication starts from a conscious source but rather—in accordance with the previously established definition of predication—whether an inanimate object is capable of setting up the relationship and initiating a predicative relation. Various inanimate objects are in predicative relations in film image and the director is the locutionary source. However, since film communication is double communication—communication within communication—one of the protagonists is the locutionary source of the inner communication. Accordingly, there are communicative acts in film that might lack both a locutionary source and a target. Still, the predicative relationships are set up, capable of communicating meanings, and at times very complex. This is possible because in film humans are not essential, mandatory elements. A film image could depict a seashore, a rustle of leaves, or a whistle of the wind. Colour or music, for example, could communicate anguish or joy, and thus can have meanings of their own. In film communicative acts, the camera takes over the mediating role of the locutionary source by setting up predications and by creating the connections between objects of the image, such as “the stormy sky and a man setting off on a journey.” Also, music can have a “generalizing function and trigger a ‘field of association’ likely to foster emotional identification” or “provide an emotional ‘beat’” as “music goes for the emotional jugular.” Stam also points out that “conventional film music has always worked to efface the instruments of production of the cinematic illusion” (Stam 2000, pp. 220–221). These roles that music can perform indicate that music is capable of establishing predications of a higher order both within a shot and in a series of shots.

The question of the delimitation between sequences is related to the boundaries of communicative acts. Both a series of shots, a sequence, or just one shot alone can create a single communicative act. If one of the mandatory parameters of the communicative act changes—participants, space-time locus—a new communicative act is created. In Tanović’s film No Man’s Land, the Bosnian and the Serbian soldiers are the only participants in most communicative acts. The space locus is identical—the trench between the battle lines—as well as the two participants. When the parameter of time changes, day to night, a new communicative act is established. The fictitious time of the film changes with the new communicative act, often manifested through the formal indications of the passing of time—the change from day to night or vice versa. The delimitation of the sequences is based on the alteration of one of the fixed parameters of the act of communication, in this case, time. The segmentation of the film discourse is thus related to the communicative components, constantly changeable.

A shot is, therefore, not an appropriate film language unit to define communicative acts since it is determined by the criterion of mechanical segmentation. It would be the same as if the speech act were matched with the sentence. What demarcates the shot is the continuity of a time-place locus. Within a single shot, the camera might depict one participant, turn to another, go back to the first participant, or include a new one. The camera could also, within one shot, “exit” a room, go to another, or introduce a new space. Shots, unlike sequences, possess a precise time-space orientation. Since the time-space locus is continuous within one shot, the spectator is constantly aware of the departure point in its time-space orientation. It is regulated by the performance of camera modalities, such as movements, angles, etc. Even when the spectator’s sense of place or time is on purpose disoriented within one shot—often through special effects—the cognitive associations the spectator establishes with the film’s “real” space and time enable the spectator to perceive the shot as grammatical or semi-grammatical.⁸

Montage is the most important film modality for the creation of communicative acts and sequences within the film discourse. There appears to be a similarity between the rules governing the editing process and the rules governing the formation of communicative acts. Kracauer demonstrates how continuity between two different shots can be achieved. The criteria that facilitate continuity between various shots are, in fact, the parameters of communicative acts noted above. Continuity is assured by the repetition of one of the fixed parameters of the communicative acts: to depict “phenomena in different places successively in time sequence,” a chase, for instance, is the repetition of the same object—predication—in two different shots. “To follow a chain of cause and effect” (Kracauer 1960, pp. 64–5) means that different shots represent one single communicative act since they repeat those objects/elements in the cause-and-effect relationship. The segmentation of film discourse into larger units of meaning or sequences helps to reveal the basic principles governing the functioning of film montage. It also could shed light on the creation of larger semantic units of a higher order that are integrated into the unique semantic whole of a film.

However, the application of the presented linguistic models has its limits. The language of film is an aesthetic system, unlike natural language. Communicative situations in ordinary language situations are not equivalent to communicative situations in film because the latter belongs to the secondary modelling system. Accordingly, the rules governing film montage—and other film modalities that are relevant to the concept of predication—are subject to another set of rules deriving from the aesthetic code. For instance, a combination of shots that is not in a direct logical connection—for example Godard’s Pierrot le fou, which Metz analyses in the context of the “potential” sequence—is an irregular combination.⁹ However, the language of film is subject to the rules of a higher hierarchy which regulates the creation of the aesthetic message by integrating separate fragments of reality into larger segments. These segments are not necessarily composed of shots in linear continuity or even in an apparent logical order. They represent communicative segments of a higher order, subject to the laws of an aesthetic code. Also, the film director imposes a specific rhythm on the order of the shots, thereby setting up rhythmical segments of a higher hierarchy. It is worth considering whether these rhythmical segments correspond with sequences, in other words, whether they represent a unique communicative act at the same time. This question relates to form and content in film language. One law of aesthetic creation demands the given form be the optimal organisation of the given content.

The application of linguistic models in investigations of the language of film cannot be exhausted solely by linguistic methods. Still, the proposed methodology for the exploration of film and theatre discourse that derives from linguistics might be enhanced in a variety of ways through the exploration of film discourses. Linguistic methodologies could be used, for instance, to discover distinct types of shots and sequences, and to reduce these diverse shots to a limited number of typical categories. By the methodology of morphologic algorithms, inaugurated by Zellig Harris, all shots of one film that, for instance, commence with a close-up, might be compared in order to isolate possible combinations. Different criteria could be combined to demonstrate whether or not some combinations are mutually exclusive. This approach would reveal if the predictability of film predications could be useful in determining what combinations are grammatical and what are not. Also, types of predications based on the character of arguments could be established. Various experiments of a linguistic origin could be conducted, such as the alteration of the “grammatical” elements of a shot while maintaining the same “lexical” elements. The idea is, in other words, to change the performance of film modalities while keeping the content of the image to detect significant changes in meaning.

In addition, it is possible to alter only the predicative relations of an image. This process might lead to some general conclusions. If certain predications demand specific modalities while excluding some others, one might conclude that, as in natural language, “syntax is more closely related to meaning than any other aspect of grammar. Words related to each other syntactically must be related semantically also” (Apresjan 1973, p. 175).

Segmentation of Theatre Language

The fundamental difference between theatre and film language is that drama lacks a mediating communicative system. In other words, a fictitious narrator is absent in drama.

Since drama lacks a mediating communicative system, it has developed techniques that compensate for the missing narrator. These techniques belong to a finite set of units performing a specific role. It is important to emphasise that the use of these techniques is conditioned by genre or by a specific historical period in theatre. Nevertheless, they are constant units in the inventory of theatre language and as such are either activated or not in certain periods or genres. These techniques could be used to integrate past and present events into the performance time.

Typical elements of theatre language are, unlike film modalities, varied in origin and nature. Despite their mutual heterogeneity, they perform the same narrative function, that of the role of the fictitious narrator, absent in drama. However, the nature of the functions of specific theatrical techniques has changed throughout theatre history. The basic dramatic concepts of action, subject-object relationship, and dramatic plot, for example, are notably changed in modern drama. These changes have also functionally altered the language units of theatre. The form of classical drama does not allow for a “separation between subject and object,” so the dramatic form is not adequate for the separation between subject and object in modern drama (Szondi 1987, p. 46).¹⁰ Characters, isolated in their destinies, lacking interpersonal relationships, are not appropriate for the dramatic form which requires action.

According to Honzl, dramatic action is the essential factor that integrates all other sign systems in theatre: speech, actors, costumes, the scene design, and music, “in the sense that we could then recognise them as different conductors of a single current that either passes from one to another or flows through several at one time” (Honzl 1976, p. 91).

On the other hand, the function of the creation of dramatic tension could be performed by various elements of film language. Montage and camera movements play the most prominent role in the creation of tension in film. The dramatic tension already present in the narrated story is reinforced, owing to the activity of film modalities. Since theatre language lacks these devices designed to reinforce dramatic tension, the very composition of dramatic action can perform this role. The dramatic action in classical drama follows the determined, conventional patterns “lusis—resolution, peripetia—dramatic change in a plot, anagnorisis—recognition, catastrophe […] the inversio and solutio, and the déenouements” (Pfister, p. 95). Drama possesses suspense potential produced by “tension between complete unawareness on the one hand and a certain level of anticipatory expectation” (p. 98). The concept of dramatic tension correlates to the double communicative process in drama: internal communication between dramatis personae and external communication with the audience. Dramas that culminate in a tragic or happy ending, with sudden peripeteia in the plot itself, eliminate the discrepant awareness between dramatic characters and between the characters and the audience. Dramas of this sort are said to have a closed form of ending, unlike dramas with an open ending, in which not “all the informational discrepancies are eliminated, [nor} all the conflicts resolved.” The plot structure of modern drama does not any more revolve around “crisis and conflict,” but could represent “a lasting condition” without resolution (Pfister, p. 96).

Szondi points out that in traditional drama “the decisions of the dramatis personae constantly modify the initial situation and move it toward a final point of resolution” (p. 55). Commonly, modern drama lacks such a concept of dramatic action since the source of dramatic tension does not derive from interpersonal relationships. Tension is anchored in the situation, which must provide “all necessary information.”¹¹ The very concept of dramatic tension in modern drama is re-examined. The action does not engender a change of situation but, rather, manifests a lasting condition. The dramatic techniques functioning to create dramatic tension are subject to constant functional changes.

In this regard, the distinction between the action and the event, elaborated by Pfister, is of considerable importance for understanding the process of the functional adjustment of dramatic techniques. Action is defined as “an intentionally chosen and not causally defined transition from one situation to the next” (Hübler, qtd. in Pfister, p. 199). This definition yields three components of dramatic action: “the existing situation, the attempt to change the situation, and the newly established resulting situation” (p. 199).

As outlined earlier, continuous reality is represented on the film screen in a fragmented manner, for each shot interrupts and fragments reality. Certain film modalities are specialised for the integration of the fragments of reality into an organic whole. A similar process occurs in drama, but the fragmentation of the dramatic world is considerably limited on account of the nature of dramatic form. The world of drama encompasses reality in its totality while film can represent fragments of reality complete in themselves.¹²

Consequently, dramatic form demands techniques that perform an integrative function to bond fragments of reality into a unified organic whole of the dramatic world. This gives rise to the question of the segmentation of dramatic discourse: What are the minimal segments of dramatic discourse, and what criteria determine their boundaries?

It is assumed that the specific nature of the medium dictates the segmentation of its units. Since theatre is not a “mechanical” medium as film is, it is not possible to segment the basic units of performance in the same way. Rather, it is necessary to include the criteria which consider dramatic content even at a basic level of segmentation of dramatic discourse. These criteria are the parameters of the communicative situation discussed earlier: participants, theme, space, and time. At the most basic level of dramatic segmentation, Pfister singled out configuration and defined it as the basic unit of dramatic language that represents “the section of the dramatis personae that is present on stage at any particular point in the course of the play” (p. 171).¹³ Configuration will alter along with the entry or exit of one figure. Thus, one of the parameters of the communicative situation, the participants, is the basis for the segmentation of dramatic discourse at the lowest level.

It is important to emphasise that, as in film, the communicative act in theatre is not the same as the speech act in natural language; communication might be activated without the verbal element, in both film and theatre. Yet, the nature of the communicative act in theatre is not identical to that of film. For example, the status of inanimate objects in film equals that of animate objects, which in turn makes it possible for a communicative act in film—one of the internal communicative processes—to function without an identifiable locutionary source. In theatre, the locutionary source of the internal communicative process is always an actor who, by virtue of dramatic action, either activates (or does not) the semantic potential of inanimate objects on stage. Inanimate objects come to life, that is, achieve the quality of predications, only if included in predicative relation by an actor’s action. Even in one-argument predications—predications that express the qualities of one object only—an object displays some of its properties important for the dramatic world, only if related to dramatis personae. Conversely, in film, the properties of inanimate objects—landscape, for instance—are capable of creating a predication that dominates others in the same shot.

The function of elements of theatre language is closely related to the very nature of the dramatic medium. The indispensable element of drama is dramatis personae: all other elements might be missing. The character generates dramatic action and integrates all other elements into an organic whole. Everything in drama is subordinated to a self-realising individuality.¹⁴ External conditions and inanimate objects are part of the dramatic world only if they are functionally related to dramatis personae.

As outlined earlier, configurations are defined by size—the number of figures—and duration. The configuration without any figures at all is an empty configuration (Pfister, p. 171). The other extreme is represented by the ensemble configuration, characterised by the presence of all figures. On the other hand, the duration of configurations plays an important role in the establishment of a specific tempo of performance. As a rule, short-lasting configurations speed up the tempo, while longer ones might indicate an unbearable slowness of the passage of time and the impossibility of anything happening (p. 172). It is obvious that the constellation of configurations is not merely a matter of form but is closely related to content.

Montage as well as the variation of the length of a shot perform similar functions in film. Through montage, the director imposes a specific rhythm on the fragments of reality in order to restructure that reality. Likewise, the rate at which configurations change is a powerful device for imposing a unique rhythm to the events represented on stage.

Pfister defines tempo on the surface level as “the speed of the movements and the frequency of the changes of speaker, configuration and locale” while at the deeper level tempo is defined by “the frequency with which the situation changes in the story itself” (p. 292).¹⁵ This distinction corresponds to that between plot and story. Through specific dramatic techniques, the deep level of a story is transformed into plot on the surface level. For instance, in Ibsen’s play Lady from the Sea (1960) a change of situation is rarely manifested on the surface level. The configurations remain the same, as well as the participants and locale, yet the situation has been changed at the deep level. The change of situation results from a decrease in informative discrepancies between the characters. In the conversation with Lyngstrand, Ellida finds out that her previous fiancé is still alive, which entirely alters the dramatic situation. When her husband announces that she is free to choose, the dramatic situation changes again. All these situative alterations are not manifested on the surface level.

On the other hand, in Alfred Jarry’s play Ubu Roi, the change of dramatic situations is often completely non-motivated in regard to dramatic dialogue. The motivation for the dramatic action is absurd and deliberately illogical. The constellation of dialogue in which interlocutors persuade each other by solid arguments to accept a certain position, thus altering the dramatic situation itself, is respected only on the formal plane of the discourse not on the content plane. The tragic dilemmas are thereby mocked and rendered absurd:

PA UBU:

No! nothing doing, I say! Do you want to ruin me just for these buffoons?

CAPTAIN MACNURE:

But look here, Old Ubu, don’t you see that your people are expecting gifts to celebrate your glorious coronation?

MA UBU:

If you don’t give them a great feast and plenty of gold, you will be overthrown in a couple of hours.

PA UBU:

A feast, yes, but money, never! Slaughter three old nags, that’s quite good enough for such scum.

MA UBU:

Scum yourself! How did such a crummy creature as you ever get slapped together?

PA UBU:

Do I have to repeat myself? I intend to get rich, I won’t fork out a penny.

MA UBU:

Don’t forget you hold in your hands all the treasure of Poland!

CAPTAIN MACNURE:

Yes, I know where there is a vast hoard hidden in the chapel; let’s distribute that.

PA UBU:

Just you try that on, you wretch.

CAPTAIN MACNURE:

Listen, old Ubu, if you don’t distribute some money, no one will want to pay their taxes.

PA UBU:

Is that really true?

MA UBU:

Yes, yes!

PA UBU:

Oh, in that case, I agree to everything. Bring up two or three million gold pieces, roast a hundred and fifty oxen and the same number of sheep, and see that there’s plenty left over for me. (1968, p. 35)

On the other hand, in the plays of Tennessee Williams, a change in a situation frequently corresponds with the boundaries of the configurations. Each new figure introduces a new conflict, and the mere entrance or exit of a figure indicates the alteration of a situation at the deep level (Rose Tattoo or Cat on Hot Tin Roof). The tempo at the deep level correlates with its surface manifestation, and that produces dramatic tension.

The smallest unit of theatrical discourse segmentation is “the partial change in configuration” (Pfister, p. 234). It is frequently defined as scene. Since the form and content of drama ought to be integrated, it is assumed that the entrances or exits of figures are motivated.

The next level of the segmentation is a complete change of configuration. It means that either all participants in the communicative act are changed or that an interruption of the space-time continuum has occurred. This level of segmentation could be named scene or act—depending on various traditions (p. 234). In any case, variations of the elementary parameters of communicative situations form the basis for segmentation at all levels of dramatic discourse. In different periods of theatre history, various techniques have been used to segment dramatic discourse. For instance, the chorus in classical Greek drama marks the division of the text into episodes (p. 235).

In addition to the devices fixed by the dramatic text, each performance can create its own techniques of segmentation. For instance, the curtains, the interval, and lighting are capable of the demarcating function (Pfister, p. 236).

The segmentation of dramatic discourse is closely related to dramatic action, which is the essence of drama. Pfister points out that it is possible to segment dramatic discourse at the deep level, which does not necessarily correspond with the segmentation at the surface level. At the first level of segmentation in the deep structure, Pfister distinguishes between action—“a single action by a particular figure”—and an action sequence—“the overall action of the whole text” (p. 199). Drama is commonly composed of several action sequences that represent a relatively closed system of chronological and causal relationships. Action sequences could be co-ordinated at the same fictional level of drama or superimposed at different fictitious levels, “the-play- within-the-play,” for example (p. 219).

Pfister, like Russian formalists, defines story as a chronological succession of events—what is narrated—while plot is the manner in which the story is narrated—the surface-level organisation of the deep-level story. It involves “temporal and spatial regrouping” and segmentation in accordance with different stages (p. 197). Story at the deep level consists of successive events that are on the surface level transformed into plot. The chronological and spatial continuity as well as the deep-level configurations are restructured on the surface plot level. Frequently, the surface-level segmentation of dramatic discourse does not correspond with the segmentation at the deep level.

The inventory of theatre language contains specific techniques for transforming story into plot. They are similar to the operations in natural language which transform deep structure into surface structure. The nature of the dramatic medium imposes certain restrictions on the transformation of story into plot. Since drama lacks a mediating communicative structure, the representation of story on the surface level ought to respect, to a certain degree, the time-space continuity. Because of these restrictions, theatre cannot resort to the film device of flashback,¹⁶ for example. In addition, scenic presentation of simultaneous actions is limited to those occurring at the same locale (Pfister p. 201).

To the extent that the principle of succession is disregarded, epic structures intrude and function as a mediating communicative system in theatre discourse (p. 202). Pfister selects two basic dramatic techniques capable of transforming story into plot: direct representation or open action, and narrative representation where the dramatis personae report hidden action, not presented on stage (p. 204).

At the next level of the segmentation of dramatic discourse higher level communicative units are created. The laws that govern the function of montage in film similarly operate at a higher level of the integration of dramatic discourse. Integration is also achieved through the repetition of at least one constant parameter of the communicative act: time, space, participants, or theme.

Pfister identifies the following techniques for integration of co-ordinated sequences in theatre: overlapping actions or events—one action or event occurs simultaneously in two different sequences; overlapping figure constellation—figure performs the same function in various scenes; and situative and thematic equivalencies, which “create a deeper, inner connection between the various strands of plot” (pp. 215–216). These techniques perform an integrative role, through the repetition of at least one of the parameters of the communicative situation.

In the inventory of theatre language, the number of functions of dramatis personae is limited. These functions are closely related to dramatic situations, and they are limited in number as well.

The distinction between figure and function, elaborated by Pfister, is based on the premise that one figure may not necessarily perform the same role and one function is not always realised by a single figure. The typology of the basic functions performed by figures leads to the typology of dramatic situations, and, accordingly, leads to the universalities of theatre language.

There is an evident correlation between the limited number of functions that the dramatic figure might perform and the elementary roles of the arguments of the predications proposed by Fillmore. The roles singled out by Fillmore are similar to the basic functions of the dramatic figure identified by the theorists of drama (Propp 1968; Souriau 1950; Ubersfeld 1999).

Two of Fillmore’s roles of the arguments—Agent and Counter-Agent—correspond to the notions of hero and opponent, common to all classifications of dramatic function.

To demonstrate this similarity, it is useful to compare Fillmore’s typology of arguments with Souriau’s typology of dramatic functions. Souriau identified six basic functions in drama. One figure might perform more than one function, and one function could be performed by more than one figure:

Le Lion—la force thèmatique—the most dominant force of a dramatic world that instigates all other actions in drama.; Le soleil or Le Representant de la Valeur—the value toward which la force thèmatique is directed; La Balance or L’Arbitre de la situation—the force that has the power of decision in a given situation.; L’Obtenteur or L’Astre Récepteur—the one who achieves the value sought; L’ Opposant or Rival—the one who confronts la force thèmatique; Complice or Le Miroir de Force or L’Adjuvant—the one who intensifies conflict, modifies equilibrium or the dynamic of a system, an adviser or accomplice. (Souriau 1950, pp. 83–113)

In the semantic description of verbs, Fillmore includes their compulsory arguments and their roles. The relations between the arguments and predications are elementary deep-structure relationships. Arguments play a limited number of roles, which are manifested on the surface level in different ways. One argument can play more than one role, which implies that arguments which play different roles might be identical. In the same way, dramatic figures perform a limited number of the basic functions—roles—very similar to those singled out by Fillmore:

Agent (A)—The instigator of the event.

Counter–Agent (C)—The force or resistance against which the action is carried out.

Object (O)—The entity that moves or changes or whose position or existence is under consideration.

Result (R)—The entity that comes into existence as a result of the action.

Instrument (I)—The stimulus or immediate physical cause of an event.

Source (S)—The place to which something is directed.

Experiencer (E)—The entity which receives or accepts or experiences or undergoes the effect of an action […]. (Fillmore 1968, p. 76)

Therefore, the deep structure of thoughts is a universal category transformed at the surface level in different languages according to the specific rules of each language. Dramatic language selects and transforms the relations within the deep structure in accordance with the demands of the dramatic conventions of theatre language. Classical drama establishes conflict situations leading to a resolution. Modern drama demonstrates situations in which the classical roles of hero and opponent are missing because this distinction at the deep level of thought is not always relevant in a world of modern drama. One might conclude that the cognitive representations or the deep structure of thoughts are historically and culturally conditioned categories. In the world of ancient drama, the deep structure of thoughts is conditioned by the mythological consciousness in which the will of the gods determines human fate. Dramatic form reflects such mythological consciousness. The essence of classical drama is the conflict between the passions and desires of the dramatis personae and the gods’ will. Dramatic action, through the stages of resolution, peripeteia, recognition, and catastrophe, leads towards the fulfilment of gods’ will.

In modern drama, conversely, seldom is there any action, but rather a dramatic situation that does not advance dramatic action. Often, only meaningless conversation justifies the existence of characters. Modern drama often reveals the impossibility and purposelessness of any deliberate action and, consequently, the functions of the figures are considerably altered. Since the essential structure of conflict is missing, the functional opposition between the hero and the opponent is missing.

Kott demonstrates how the tragic situation in modern drama becomes grotesque. In the classical drama, the hero is faced with a “choice by which one of the values must be annihilated” (1964, p. 194). The choice excludes any possibility of compromise, and the price is death. The tragic situation becomes grotesque when both alternatives are absurd and meaningless (p. 194).¹⁷

The functions of dramatis personae have also changed as with the other elements of theatre language. The surface elements of theatrical language and the deep structure of thoughts are interdependent categories. The functions deprived of correlates in the deep structure degenerate, while the new concepts of the deep structure initiate their new correlates on the surface level.

As mentioned earlier, dramatic figures might be characterised in both an implicit and an explicit manner. The explicit characterisation is premised on the self-characterisation of figures in both verbal and visual manner. The implicit characterisation is done from the perspective of other figures or an implicit author’s commentary—by means of setting,¹⁸ or by objects on stage that are in predicative relation to a character.

Film has recourse to similar techniques, but film also makes use of its specific modalities for figure characterisation. Theatre language resorts only to the above-mentioned techniques that are indispensable elements of its language. The distinctive characterisation of characters in film as opposed to theatre derives from the different nature of these media. Bettetini states that “on screen stands not an actor, but his image which is a great deal freer from the somatic and psychological characteristics of the character represented than would be the most perfect interpretation on stage” (Bettetini 1973, p. 82). The actor in film is an object used in the same manner as any inanimate object by a film director. Consequently, inanimate objects in film are capable of creating predications, while in theatre a character bestows a status of the argument of predications to an inanimate object.¹⁹

The dramatic structure as a whole, along with the theatrical representation of a given dramatic text, is considerably influenced by genre or specific historical periods in theatre. However, the theatrical techniques described here are constant elements of theatre language that are activated in certain genres and periods in various ways. The most important characteristic of these techniques is their functional dynamism. In other words, they constantly adjust to perform new functions as demanded by the deep structure of thought.