FILM AS TEXT: RADICAL CONSTRUCTIVISM AND THE PROBLEM OF NARRATIVE IN CINEMA.
The present paper is intended to broaden the debate on constructivism and film theory. It is divided into two sections: first, I outline the main features of Radical Constructivism as proposed by Ernst von Glasersfeld, and I argue that objections to such a radical approach arise from a (common) misunderstanding of the Radical Constructivist position; and second, I discuss David Bordwell’s theory of narration in the cinema from such a position. I argue that as Bordwell assumes that the spectator’s constructive activity is compatible with a realist epistemology the relationship between syuzhet and fabula he describes is flawed as he cannot account for the communication of narrative information from screen to spectator. I describe the nature of communication in the cinema and propose an alternative, hierarchical relationship between the syuzhet and the fabula.
David Bordwell (1996a) raises a number of philosophical, theoretical, and methodological issues that are relevant to all working within the field of film studies. In this paper I wish to focus on a single detail of his critique: the casual dismissal of a radical approach to constructivism in film theory. Bordwell states two objections to such a perspective that I quote here in full. He writes that,
Having rejected the strong/radical version of constructivism, Bordwell (1996b) proposes a ‘moderate constructivism’ that seeks to unite the biological predispositions of the cognising subject (Naturalism) and the pre-existing materials out which the subject constructs meaning in the world (Conventionalism).
Bordwell’s objections demand a response – if only to introduce a genuinely radical approach to constructivism to film studies. The development of cognitive approaches to the cinema over the past twenty years has brought constructivism to the fore in film studies (e.g., Bordwell 1985, 1989a, 1989b; Branigan 1992), though there has been little debate as to what the various forms of constructivism (e.g., Constructive Realism, Social Constructionism, Radical Constructivism) have to contribute to our understanding of the cinema. This paper is intended to broaden that debate. It is divided into two sections: first, I outline the main features of Radical Constructivism as proposed by Ernst von Glasersfeld, and I argue that Bordwell’s objections arise from a (common) misunderstanding of the Radical Constructivist position; and second, I discuss Bordwell’s theory of narration in the cinema from such a position. I argue that as Bordwell assumes that the spectator’s constructive activity is compatible with a realist epistemology the relationship between syuzhet and fabula he describes is flawed as he cannot account for the communication of narrative information from screen to spectator. I describe the nature of communication in the cinema and propose an alternative, hierarchical relationship between the syuzhet and the fabula. On the ‘historical’ definitions of these terms within the original major debate of their ‘inventor’ (i.e. Russian Formalism), see the principal texts by Viktor Shklovskii (1990); along with some general accounts with regard to the meaning of ‘syuzhet’ and ‘fabula’ in Formalist theory of aesthetics (Striedter 1989; Erlich 1965/1981; Jackson and Rudy 1985; Steiner 1984).
Radical Constructivism is part of a wider ‘constructivist’ movement in the philosophy of science (Schwandt 1994). It is a highly disciplinary school of thought that draws on psychological, biological, linguistic, and sociological evidence, and is closely linked to Cybernetics (Foerster 1973, 2003), Autopoiesis (Maturana and Varela 1987), and Personal Construct Psychology (Kelly 1955a, 1955b). (To differentiate Radical Constructivism from the ‘historical’ (Russian) aesthetic movement ‘Constructivism’ see the classic monograph by Christina Lodder , and most recently Maria Gough  and Barrett Watten . See also the important collections of Richard Andrews and Milena Kalinovska [eds.]  and Selim Khan-Magomedov . With respect to the early ‘Constructivist’ cinema and Sergei Eisenstein see, for instance, a general survey of Albera , along with Petric  and Bergan ).
Radical Constructivism is not concerned with knowledge per se, but with the mechanisms of knowledge construction. As such, Radical Constructivism has been described as ‘post-epistemology’ (Noddings 1990), and the school’s founder, Ernst von Glasersfeld (1974, 1984, 1990, 1995), has described it as a ‘theory of knowing’ rather than a ‘theory of knowledge.’
Radical Constructivism is an attempt to avoid the epistemological trap of realism.
Radical Constructivists argue that all knowledge is constructed. However, Radical Constructivism is not solipsistic: it does not deny ontological reality. Nor does it assert the existence of a mind-independent reality, but states that our knowledge is not knowledge of an ontological reality. As we cannot transcend the limits of our experience it is impossible to tell (and therefore unnecessary to know) to what degree our knowledge reflects an ontological reality. From the perspective of Radical Constructivism both ontology and epistemology are redundant: Riegler (2001: 1) describes Radical Constructivism as ‘agnostic.’ It is this statement of the function of cognition – ‘Cognition serves the subject’s organisation of the experiential world, not the discovery of ontological reality’ (Glasersfeld 1988: 83) – that distinguishes Radical Constructivism from trivial modes of constructivist thought: ‘those who merely speak of the construction of knowledge, but do not explicitly give up the notion that our conceptual constructions can or should in some way represent an independent, “objective” reality, are still caught up in the traditional theory of knowledge’ (Glasersfeld 1991: 16).
What do Radical Constructivists mean when they say that ‘the function of cognition is adaptive?’ The principle of adaptation is derived from the work of Jean Piaget (1937, 1967; see also Glasersfeld 1997), who approached the construction of knowledge as a biologist, and from the theory of evolution he imported the concept of adaptation into the study of cognition. Adaptation involves two complimentary and simultaneous processes: a cognising organism primarily seeks to organise experience in terms of the psychological structures (schemes) it already possesses, i.e., it seeks to assimilate experience; if the result of this process creates a perturbation the organism attempts to accommodate the error either by modifying an existing scheme or creating a new one. It is this balance between assimilation and accommodation that Piaget describes as adaptation. Knowledge is actively constructed, and is adapted to fit the environmental constraints that act on an organism in order to avoid internal contradictions and achieve equilibrium. Glasersfeld (2001: 39) describes the principle of adaptation in Radical Constructivist thought:
The key to evaluating competing knowledge claims, therefore, is not to seek to compare them to a mind-independent reality that cannot be known, but to assess their cognitive viability or functional fitness.
Bordwell’s objections to a radical approach to the construction of knowledge derive from a misunderstanding of the Radical Constructivist position. First, Radical Constructivism demands that knowledge is judged in terms of its cognitive viability, and it demands this of itself. Radical Constructivism is not self-refuting as it includes its own operations within its description of the production of knowledge. Glasersfeld (1991: 13) writes:
Radical Constructivism is ‘second order knowledge;’ it is knowledge about itself. A constructivism that does not account for its own operations is not radical; that is, it is not complete and remains within a traditional theory of knowledge. Bordwell’s first objection thus applies not to Radical Constructivism but to trivial modes of constructivist thought that continue to offer the (forlorn) hope of knowing an ontological reality. Bordwell’s first objection is thus applicable to his own ‘moderate constructivism.’ For Bordwell, the term ‘constructivism’ is not intended to be used ‘in the epistemological sense that is commonly opposed to realism;’ rather, it is intended ‘to signal the importance of constructive inference, or inference-like procedures, in our mental activities’ (Bordwell 1989b: 34). His model for making meaning takes a constructivist approach that ‘assumes that it is possible to arrive at inferences which are at least approximately true; it is thus compatible with a critical realist epistemology’ (Bordwell 1989a: 277n9). Confronted with such a model the following question remains: How do we know that our inferences are ‘at least approximately true?’ An answer to this question has eluded Western philosophy since Socrates, and, without a solution to the ontological problem, constructivism remains trapped within the epistemological trap of realism. As it assumes a ‘critical realist epistemology’ Bordwell’s ‘moderate constructivism’ is, in the precise way described by Glasersfeld, trivial.
In his second objection Bordwell states that a radical approach to constructivism ignores ‘biological’ explanations, and, therefore, has little to contribute to our understanding of ‘cross-cultural regularities.’ This is simply inaccurate. Radical Constructivism is influenced by, and influential upon, research in biology, neurobiology, ecology, and physiology (see Maturana 1978; Maturana and Varela 1980; Maturana and Varela 1987; Peschl and Riegler 1999; Boden 2000; Grossing 2001; Llinas 2001; Stewart 2001). Radical Constructivism can offer insights into the development and distribution of cognitive invariance: for example, Diettrich (2001) argues that perceived regularities are invariants of cognitive operators, and as such are human specific constructions and not ontologically independent. There is no reason why Radical Constructivism, which encourages us to reflect on our constructions, should be empirically limiting: ‘To the living creature, then, the universe is real, but it is not inexorable unless he chooses to construe it that way’ (Kelly 1955a: 8).
A key problem in Bordwell’s rejection of a radical version of constructivism is his use of terminology. The terms ‘radical’ and ‘constructivism’ are deployed by Bordwell to describe contemporary film theory of the Screen type with no apparent awareness of Radical Constructivism as Glasersfeld has proposed it as a theory of knowing. I would find it surprising if contemporary film theorists were to accept the Radical Constructivist description of cognition. A similar confusion is evident in Carroll’s (1996) use of the term ‘social constructivists.’ Carroll does not specify to whom this term is to be applied, though as it is used in his worthy critique of contemporary film theory I assume that he, like Bordwell, is referring to contemporary film theorists. Carroll’s use of this term is confusing, as it shows no apparent awareness of or respect for the distinction that is to be made between Radical Constructivists such as Glasersfeld and Social Constructionists (Gergen and Gergen 1991).
Radical Constructivism and Narrative Cinema
In Narration in the Fiction Film (1985), Bordwell sets forth a poetics of narration in the cinema to account for the dynamic processes of the spectator’s narrative activity and the formal operations of cinematic narration. In this section I consider Bordwell’s theory of cinematic narration from a Radical Constructivist perspective, and I argue that as he assumes that the spectator’s constructive activity is compatible with a realist epistemology the relationship between syuzhet and fabula he describes is flawed. I go on to propose an alternative, hierarchical relationship between the syuzhet and the fabula.
Narration in the Fiction Film
Bordwell bases his theory of cinematic narration on the work of the Russian Formalists. It is a theory that assumes a distinction between ‘the story that is represented and the actual representation of it’ (1985: 49); a distinction between the narrative as it is constructed by the spectator (the fabula) and the formal systems of representation employed in a film (syuzhet and style). Bordwell describes the spectator’s activity in constructing a narrative in the following terms:
Presented with two narrative events, we look for causal or spatial or temporal links. The imaginary construct we create, progressively and retrospectively, was termed by the Formalists the fabula (sometimes translated as ‘story’). More specifically, the fabula embodies the action as a chronological, cause-and-effect chain of events occurring within a given duration and spatial field. … The fabula is thus a pattern which perceivers of narratives create through assumptions and inferences. It is the developing result of picking up narrative cues, applying schema, framing and testing hypotheses. … It would be an error to take the fabula, or story, as the profilmic event. A film’s fabula is never materially present on the screen or soundtrack. … What is given? … The syuzhet (usually translated as ‘plot’) is the actual arrangement and presentation of the fabula in the film. It is not in the text in toto. It is a more abstract construct, the patterning of a story as a blow-by-blow recounting of the film could render it (Bordwell 1985: 49-50. Original emphasis).
Bordwell frames his account of narrative comprehension within the Constructivist school of cognitive psychology. He describes the spectator as an active mobiliser of inferential and problem-solving processes in the testing and fixing of perceptual hypotheses:
The organism is tuned to pick up data from the environment. Perception tends to be anticipatory, framing more or less likely expectations about what is out there. … The organism interrogates the environment for information which is then checked against the perceptual hypothesis. The hypothesis is thus either confirmed or disconfirmed; in the latter case, a fresh hypothesis tends to appear (Bordwell 1985: 31).
Bordwell argues that viewers construct a hypothesis and then match elements of a film (e.g., narrative events, characterisation, film style) to this hypothesis in order to construct a film’s meaning.
In his adherence to a realist theory of film spectatorship, Bordwell stresses that a film, in some form, already exists prior to the mobilisation of the spectator’s psychological processes: ‘Construction is not ex nihilo creation; there must be prior materials which undergo transformation’ (1989: 3). These prior materials are understood to be ‘cues’ in a film, and it is these that form the basis for the spectator’s constructions. Bordwell’s Constructivism is a theory of how the spectator actively engages with a film in the production of meaning.
Sweeney (1994) argues that as Bordwell makes no rigid ontological distinction between the pre-interpreted film and the interpreted film, and given the realist epistemology in which Bordwell locates his theory, ‘one must look to some supplemented or freighted object as the ontological construction of his theory.’ There must be a ‘realistic’ given in a film – in narrative cinema, the syuzhet – and the spectator transforms the text ‘by supplementing it, making the text take on meaning.’
Bordwell’s theory of narration in the fiction film is problematic in two ways. First, Bordwell assumes a model in which the inferences of the spectator as assumed to be ‘at least approximately true,’ but offers no argument as to how a cognising subject could objectively confirm his realist approach. His theory of the spectator’s production of meaning cannot be validated in its most essential particular. Second, Bordwell’s poetics of narrative lacks a coherent theory of communication. He rejects ‘communication theories of narration’ in which a message is passed from sender to receiver, but identifies the communicativeness of a narrative as being ‘how willingly the narration shares the information to which its degree of knowledge entitles it’ (Bordwell 1985: 59). As the fabula is constructed by reordering the information made available by the syuzhet this implies the communication of meaningful narrative information from the screen to the spectator, but Bordwell does not define the nature of communication or information in the cinema and their relationship to meaning.
Communication and Interpretation
The Radical Constructivist approach has a significant impact on the understanding of communication. In stating that the cognising subject, according to its needs, actively builds up all knowledge it is necessary to accept that meaning is constructed and cannot be communicated from one point to another. Meaning does not exist prior to its construction by a cognising subject; it does not reside in language, images, behaviour, or objects. Meaning cannot be communicated, and Shannon (1948) states that semantics is irrelevant to any understanding of communication. Glasersfeld argues that the belief that the inferences of a cognising organism derived from the environment in which it is embedded accurately reflect that environment is based on an ‘untenable conception of “communication.”’
If the constructivist movement has done anything at all, it has dismantled the image of language as a means of transferring thoughts, meanings, knowledge, or ‘information’ from one speaker to another. The interpretation of a piece of language is always in terms of concepts and conceptual structures which the interpreter has formed out of elements from his or her subjective field of experience. Of course, these concepts and conceptual structures had to be modified and adapted throughout the interactions with other speakers of the language. But adaptation merely eliminates those discrepancies that create difficulties in actual interactive situations – adaptation ceases when there seems to be a fit. And fit in any given situation is no indication of match. To find a fit, simply means not to notice any discrepancies (Glasersfeld 1991: 23. Original emphasis).
Bordwell (1989a: 64) sets the following question: ‘Any interpretative practice seeks to show that texts mean more than they seem to say. But one might ask, why does a text not say what it means?’ From a Radical Constructivist perspective not only can a spectator not know to what extent a film says what it means, the argument that a film communicates meaning to a spectator cannot be sustained. From this perspective communication as the transfer of meaning from source to destination is an illusion, and as Glasersfeld states: ‘There seems to be a blatant contradiction between the claim of “communication” and the apparently irrefutable subjectivity of meaning’ (1983: 211). This contradiction may be overcome if we consider what takes place in communication in the cinema, and subsequently examine the relationship of communication to information and meaning.
Communication may be defined as the successful transfer of information from one point in space-time to another. Communication may involve the transfer of information in one direction only or may be bi-directional, and the subjects involved in communication may be human, machine, or a combination of the two (Read 1998). Stonier (1997) argues that like energy, information should be considered a basic property of the universe, and should be defined operationally as the capacity to generate organisation. Any system that exhibits organisation contains information, and this is as true for the arrangement of molecules in a crystal as it is for the pattern of letters printed on a page. Information is the raw material that may, when processed, yield a message.
How is information communicated in the cinema? The first step is to produce a recording of the subject before the camera. Reflected light from the various elements of the scene passes through the lens of the camera and falls on the film stock where it affects the light sensitive coating to produce a latent image. The film is then processed to first develop the latent image and then to fix it as a permanent pattern of silver on the film base. This process is repeated in producing a positive print by passing the negative and unexposed stock simultaneously past a lens and a light source. In the projector, light is beamed through the moving film and is magnified by a lens for projection onto a screen. Reflected light from the screen enters the spectator’s eye, triggering a wide range of sensory and neurological processes in the spectator’s assimilation and analysis of new information. Sound waves emerging from a source are picked up by a microphone and converted into electrical signals. These signals are amplified and fed to a recording head where they give rise to a magnetic field that affects a tape coated with a metallic oxide. Sound is recorded as a pattern in the structure of the magnetic coating. On playback this pattern of magnetic oxide gives rise to an electronic signal, which is then amplified and causes mechanical vibrations in the diaphragm of a loudspeaker. These vibrations are mechanically coded as sound waves that give rise to a further set of neurological activities when heard by the spectator (Redfern 2004).
In the cinema then, a succession of images are produced, printed, and projected. The information about each image that is transmitted is the intensity of light and colour with its associated position in the frame. Sound is also transmitted. In the cinema three types of information are transmitted: light, position, and sound (Read 1998: 1). It is evident from this description of the transformation and transfer of information in the cinema that meaning has no role to play.
The process of subjectively constructing meaning is typically taken to involve a process of interpretation. Since the 1960s, when film studies entered the academy, the production of interpretations has become the goal of analysing cinema, and has become so routine and has reached such a scale that it has been described as an ‘industry’ (Bordwell 1989a). Radical Constructivism adopts a different approach and focuses on interpreting as one activity performed by a subject rather than interpretation as a product. Interpreting is a sense-making activity that involves the building up of conceptual structures. Those structures are constructed by putting ‘facts’ into a context. From a Radical Constructivist point of view, ‘facts’ are not a part of reality, but are elements of the observer’s experience: ‘Empirical facts, from the constructivist perspective, are constructs based on regularities in a subject’s experience’ (Glasersfeld 1989b: 447. Original emphasis). Facts are the result of an active process of observing. These facts are then put into a context by a spectator (i.e., they are interpreted). It is at this point that the interpretation of films typically ceases, but cognitive film studies takes the process further to engage in an activity of theorising that involves ‘the production of generalisations or general explanations or general taxonomies and concepts about film practice’ (Carroll 1996: 39). Taken as a whole this process involves a hierarchy of constructions: the construction of facts (observing), the construction of conceptual structures (interpreting), and the construction of theoretical statements (theorising). A final activity in the subject’s construction of knowledge involves testing a theory in new experiential contexts. A theory is viable if it maintains its usefulness and serves the goals of the subject in subsequent observations, interpretations, and theorisations.
If we cannot know to what extent our constructions reflect ontological reality then our constructions must remain hypothetical; and this is the case where we attempt to construct the mental and conceptual operations of others and of texts. If the cognising subject is a cinematic spectator then the interpretation of cinematic texts must be understood in terms of the ‘concepts and conceptual structures’ formed by the spectator out of the elements of ‘his or her subjective field of experience.’ The interpretation of films must therefore be understood not in terms of the meaning(s) a film may possess or that may be determined by a film’s formal elements, but in terms of the conceptual structures and cognitive operations of the spectator (Redfern 2004). The constructions of a cinematic spectator must remain hypothetical: a spectator cannot know to what extent his or her construction of a film matches that film. It is clear from this that the spectator cannot match the elements of a film to the hypothesis he or she constructs in order to produce a film’s meaning.
Radical Constructivism and Narrative Cinema
From a Radical Constructivist perspective it is necessary to redefine some concepts in our understanding the spectator’s construction of narrative in the cinema.
1. [The fabula] ‘is the developing result of picking up narrative cues…’
As Vico (1710/1988) noted, the word ‘fact’ is derived from facere – the Latin ‘to make’ – and the construction of facts involves work on the part of the observer. Like empirical facts, which from a Radical Constructivist point of view are ‘elements of the observer’s experience,’ narrative cues in the cinema are not objective data of information waiting to the ‘picked up’ by the spectator. ‘Facts’ in the cinema are not the pre-existing cues in a film described by Bordwell, but are the result of an active process of observing on the part of the spectator in which information is selected from his or her experience of a film. Facts, as Gough (1993) observes, are ‘testimonies to experience,’ and the subsequent constructive operations of the spectator make sense of this experience by constructing links between these facts.
2.‘Presented with two narrative events, we look for causal or spatial or temporal links.’
Bordwell’s model assumes that a narrative exists, albeit only partially, prior to the active processes of construction that the spectator brings to a film. However, this does not tell us how the spectator knows what is (or is not) a narrative event, and raises the question that if the spectator were already aware that he/she is being presented with two narrative events, why would he/she look for causal or spatial or temporal links? If the narrative were to exist prior to the spectator’s experience of a film, the activity of the spectator is trivially constructive, and is limited to joining up the dots between pre-existing narrative events rather than constructing those events and the links between them. Rewriting Bordwell’s statement from a Radical Constructivist perspective places the emphasis on the constructive activities of the spectator:
When the spectator experiences two filmic events he/she looks for causal or spatial or temporal links; if the spectator is able to construct such links between two filmic events, then he/she may infer that those events are narrative events.
Where a spectator is unable to make such inferences he/she may regard them as being non-narrative (e.g., associational, poetic) events.
3. ‘The syuzhet (usually translated as ‘plot’) is the actual arrangement and presentation of the fabula in the film.’
In constructing links between narrative facts the spectator constructs the syuzhet, which cannot be taken to be ‘the actual arrangement and presentation of the fabula in the film.’ If the spectator is unable to transcend his or her experience of a film and does not have access to that which is objectively given in a film, then the syuzhet that is in the film remains beyond the reach of the spectator. The definition of the syuzhet as ‘the patterning of a story as a blow-by-blow recounting of the film could render it’ remains, but it must be understood as a recounting of his or her subjective experience of the film by the spectator and not as something that is objectively given. It may be a tautology to state that knowing requires a knower, but as Ludwig Fleck states: ‘Of what should absolute reality be independent? If you want it independent of humans, you should consider that it would then be useless for humans’ (quoted in Glasersfeld 1989b: 435). A syuzhet that exists independently of the spectator is of no consequence in the construction of narrative in the cinema.
4. ‘The fabula embodies the action as a chronological, cause-and-effect chain of events occurring within a given duration and spatial field.’
The spectator constructs the fabula as a meaningful narrative by reordering the syuzhet into such a form, but the semantic content of this construct is solely the product of the operations of the spectator and has not been transmitted across a communication channel. Therefore it is necessary to describe a new relationship between the syuzhet and the fabula that recognises the radical nature of the spectator’s operations in constructing a narrative. Tom Stonier’s analysis of the production of meaning in information systems bears obvious similarities to Bordwell’s account of the viewer’s activity and to Piaget’s description of cognitive adaptation, and suggests a means by which syuzhet and fabula can be related within a coherent framework. Stonier describes meaning as the product of a two-step process of information work: the recipient selects a specific piece of information from an environment rich in information, and then analyses the selected piece of information in order to produce meaning. Information may be described as meaningful when the recipient is able to successfully link it to pre-existing information structures. Stonier gives the following definition of meaning:
Meaning is a state which is achieved when the incoming information becomes integrated into the information structures already present in the host. This message-context complex (a semantic complex) may, in turn, be further information-processed. An advanced information system such as the human brain may treat the semantic complex as if it were a message even though it has been internally generated. This would allow for the possibility of a second-order message-context complex (the original semantic complex plus its secondary context). By repeating this process, a hierarchy of contexts may be achieved which, if sufficiently elaborate, creates understanding (Stonier 1997: 187. Original emphasis).
The relationship between narrative facts, the syuzhet, and the fabula is a hierarchical relationship of this type. The spectator’s construction of the syuzhet is a first-order semantic complex. The syuzhet is then further information-processed in order to produce a second-order semantic complex – the fabula. Taken as a whole the spectator’s construction of narrative involves a hierarchy of constructions: the construction of narrative facts, the construction of causal or temporal or spatial links between those facts (syuzhet), and the construction of a chronological cause-and-effect chain of events (fabula). A final activity in the spectator’s construction of narrative involves testing the fabula in the light of subsequently constructed facts. A fabula is viable if it maintains its usefulness and serves the understanding of the spectator in subsequent experiences. These four stages are represented in Figure 1, which refers only to the experience of the spectator and does not refer to (and does not need) a reality that exists independently of the experience of the spectator. As the model contains a feedback loop, the knowledge that the spectator constructs is self-referential. In this model each stage in the construction of a narrative involves an input of work; and at each stage this input leads to an increase in organisation, thereby producing a hierarchy of organisation.
Bordwell defines narration in the fiction film as ‘the process whereby the film’s syuzhet and style interact in the course of cueing and channeling the spectator’s construction of the fabula’ (1985: 53), but in the absence of a theory of communication in the cinema, this definition cannot account for how these cinematic processes structure the spectator’s activity. Radical Constructivism offers a solution to this problem by accepting the unavoidable subjectivity of the spectator, and that it is the spectator who segments the diversity of experience into narrative facts, combines these to form a syuzhet, and further processes this construct to abstract a fabula. The ongoing process of refining the fabula in the light of subsequent experiences allows the spectator to construct a stable narrative. Bordwell is correct to argue that in the cinema construction is not ex nihilo, but the prior materials that undergo transformation are not pre-existing cues that are transmitted unproblematically from screen to spectator. Those prior materials are the pre-existing information structures of the spectator, and these structures are organised hierarchically.
Figure 1: Stages in the Spectator’s Construction of Narrative in the Cinema
The development of constructivist approaches to film studies has contributed much to our understanding of the cinema. They have suggested new models for how we experience film biologically/ecologically (Anderson 1996; Anderson and Anderson 1996; Hochberg and Brooks 1996), emotionally (Platinga 1995; Smith 1995; Tan 1996; Grodal 1997), and cognitively (Bordwell 1985; Branigan 1992; Currie 1995). They have investigated the interactions between audiences and films (Staiger 1992, 2000); and have focussed our attention on the philosophical, theoretical, and methodological assumptions scholars make in their research on cinema (Carroll 1988, 1996; Bordwell 1989a). This research has been carried out within a traditional realist epistemology, but as the above discussion of Bordwell’s model of narration in the fiction film indicates the case for the realists, even those of a constructivist bent, has yet to be conclusively made.
Winston (1996) describes the cinema as the product of a ‘lens culture,’ the linking of a realist ideology, scientific innovation, and aesthetics, that emerged in Renaissance Europe and was defined by an ‘addiction’ to realism. The apogee of this culture came in the nineteenth century with the development of photography and the cinema, technologies that have been described by Bazin (1967: 12) as ‘discoveries that satisfy, once and for all and in its very essence, our obsession with realism.’ Given these origins, it is unsurprising to find that the key concept to have shaped our understanding of the cinema – as a technology, as an art form, and as the site of academic discourse – is realism: ‘In relation to the cinema, the issue of “realism” has always been present’ (Stam 2000: 141). However, as John Hill states:
There is probably no critical term with a more unruly and confusing lineage than that of realism. … its continuing use-value as either a descriptive or explanatory concept would often seem to be in question. Amidst this plurality of uses, one consistent implication does appear to survive: that the distinctive characteristic of realism resides in the ambition to, in some way or other, approximate reality, to show ‘things as they really are’ (Hill 1986: 57).
It is the ambition of realism that is most problematic, and which remains unsolved. A Radical Constructivist approach to the cinema makes it possible for film scholars to do something other than discuss the cinema in terms realism without lapsing into solipsism or the ‘anything goes’ relativism that Bordwell rightly challenges. If film studies remains locked within a traditional realist epistemology, then cognitive, constructivist, and historical poetic approaches to the cinema will prove to be of limited use, unable to establish the validity of their most basic claims in their most essential particular. If we are to take the metaphor of construction seriously then we must adopt a radical position.
Albera, F. (1990) Eisenstein et le constructivisme russe, Stuttgart; Lausanne (Suisse): L’Age d’homme. (Series : Collection Histoire et theorie du cinema).