Comprehension of Film Narrative

Jacek Ostaszewski

 

Cognitive film theory assumes that the process of understanding a film narrative consists in solving a cognitive problem, whereby a viewer tries to construct a coherent story, grasp its direct meaning and discover the cause-and-effect relations that govern a particular film. As David Bordwell states in Narration in the Fiction Film, “to understand a film’s story is to grasp what happens and where, when, and why it happens.”[1]

My empirical research on comprehension of film narrative is based on the analysis of papers written by the candidates taking their entrance exam to Film Studies Institute at the Jagiellonian University in Cracow, and papers written by the students of the mentioned Institute. Having seen Krzysztof Kieslowski’s Three Colours: White (1993) and Elia Kazan’s The Arrangement (1969), the candidates were required to write their on-the-spot analysis of a given film (96 and 86 papers respectively). On the other hand, the students were allowed a week’s time for writing an analysis of Otto Preminger’s Laura (1944) (87 papers).

The papers delivered by both candidates and students are mainly of descriptive character, which results from a single viewing of a given film. For this reason I have found the papers extremely useful as an illustration of the process of film comprehension. Furthermore, I assume that the ability to provide a linguistic description and to conceptualise a film’s story is actually the ability to comprehend film narrative.

I shall begin with presenting the results of my studies on these papers with reference to common mistakes and various biases in the film’s description, as I agree with Elke van der Meer’s statement that “errors and mistakes in judgements and thinking processes are often far more interesting than correct solutions, because they provide the right key to understand the nature of human thinking.”[2]

Taking into account the limited amount of time I am given, I will not discuss all types of bias I have come upon. Instead, I shall concentrate on those examples, which are the most surprising and at the same time most typical of the process of constructing a cognitive representation of a fiction film.

On the example of the papers on Laura I would like to show how the knowledge of typical film representations is revealed in constructing a situational model; the papers on The Arrangement will serve as an illustration of typical mistakes in the construction-integration process; while Kieslowski’s film Three Colours: White will be considered in the context of overinterpretation that may accompany the process of comprehending a film, or sometimes even replace it.

Thanks to its changing narrative perspective and the ambivalent combination of realistic and oneiric elements in the second part of the film, Laura cannot serve as an example of a typical Hollywood film; as it was convincingly justified by Kristin Thompson in Breaking the Glass Armor (1988). However, not many viewers can notice and understand this fact. Most viewers perceive Laura as illustration of their own stereotypic images and expectations pertaining to a film noir, even though they can notice a defamiliarization device, which consists in concealing the identity of a victim and providing a surprising solution. Preminger’s Laura is a “simple” story not because of its form as such, but because of the viewer’s expectations that film cannot overcome.

The most surprising bias the analyses of the film pointed to was an erroneous assumption that an ancient clock and Laura’s portrait are located both in Laura Hunt and Waldo Lydecker’s apartments. The authors of analyses emphasised that in a crime story certain objects become cues or evidence in police investigation. Although the mistake appeared in four papers only (which accounts for 6.6% of all papers), it illustrates the work of several mechanisms in cognitive processing and understanding a fiction film.

Firstly, the mistake results from misunderstanding the shift from the film’s credits to the first scene, which takes place in Waldo’s apartment. This must be the only explanation for the mistake. Laura’s portrait makes the background for the credits. The credits end with a fade-out, but before a fade-in, we can hear Lydecker’s voice over. After the fade-in, the camera smoothly travels into a pan showing Waldo Lydecker’s living room full of antiques, works of art and craft, and concentrates on the figure of detective McPherson who takes a brief look at a portrait of a woman (though it can hardly be seen, it is still evident that it does not show Laura).

The mistake results from the students’ lack of knowledge of typical conventions governing classic Hollywood cinema, namely, the rule of “hard editing” (straight cut) within a single scene contrasted with “soft editing” which was used for linking scenes and sequences, and as such served as a punctuation mark. Fade-out/fade-in, which is used in this particular case, signals a far more distant leap in time-space, because it links the sequences. The viewer of classical Hollywood films knew that after fade-out/fade-in it is necessary to define “when” and “where” anew. The authors of the four papers I have mentioned did not realise how important the function of a long and clear fade-out and fade-in was, and so they identified Laura’s portrait from the credits with a portrait in the first scene. This erroneous assumption is further elaborated on when in the third scene of the film the viewers’ and the characters’ attention is drawn to the portrait itself, and the authors jump to the conclusion that although Laura is absent from the diegetic world (as she was murdered), we still are presented with two portraits of hers.

This mistake confirms a procedural status of knowledge on filmic forms of representation. None of the authors mention the style of the film. It seems irrelevant to a situational model they construct for the story told in Laura. With respect to the status of procedural knowledge, however, the elements of style are essential for the constructing of cognitive representation, even though they do not make a part of this representation. The case of the four papers including the mistake I have discussed here, seems to confirm this statement.[3]

Another conclusion we may draw refers to the students’ knowledge of filmic forms of representation, or rather, we should say, to the lack of such knowledge.

It seems that the mistake committed in the beginning has a particularly strong impact on the subsequent opinions of the viewers of a film. Although the viewers learn the portrait was painted by Jacoby (who was in love with Laura, which excludes even the slightest chance for another copy of her portrait to exist), they seem to ignore it. Therefore, it seems that the viewer is ready to verify the “strong” assumptions made at the preliminary stage of viewing only if a film itself denies them directly and ultimately, or if a general coherence is at stake.

When trying to understand Laura, the viewers construct the plot at the level of referential meanings based upon a model detective story. In the case of The Arrangement, however, such a simple model cannot be employed, as the film makes the viewer construct the story and seek its explicit meanings simultaneously. As one of the reviewers pointed out: “The film is dense with ponderous allegories, symbols and things which ‘imply something’ or ‘create the atmosphere’.”

The candidates watching The Arrangement were high school graduates who had finished a course in the history of literature, so they had no difficulty in discovering allusive constructs based on simple clichés. Sometimes they even revealed an exaggerated urge to do so.

The descriptions of the second scene of the film included in many papers provide a clearly erroneous construction, which, nonetheless, hilariously draws upon allusive and allegoric rhetoric employed by Kazan. In one paper the author claims that the drivers of the two trucks between which Eddie (Kirk Douglas) was driving in the scene of his attempted suicide, ostentatiously smoked Zephyrs; while in fact, neither of them smoked a cigarette; and it was a passenger in one of the trucks who indeed smoked ...a cigar. In two other papers the authors come to an amazing conclusion that it was Eddie himself who was driving a truck he crashed into.

However, the most significant “bias” in understanding The Arrangement is the erroneous recognition of Uncle Joe’s role (E. J. Andre). We meet him in the scene at a hospital when Dr Weeks informs Eddie of his father’s condition. Uncle Joe, who accompanies Eddie’s mother, awkwardly introduces himself to Dr Weeks as the brother of the deceased, the mistake tactfully corrected by Gloria explaining who he actually is. A moment later Eddie talks to his father, who shows the traces of senile dementia or paranoia aimed at his 71-year-old wife. Eddie’s father accuses her of having an affair with an Irish lover who chases her like a dog whenever he (i.e. father) takes his eyes off her. Strange as it may seem, in as many as 13 papers (15.1% of all) the authors identify an “Irish Lover with ... Uncle Joe!”

This false construction may be the result of incorrect assumptions made in deductive thinking. In this context the essence of a deductive conclusion boils down to the idea that on the basis of the assumptions regarded as true and correct, one draws the conclusion, which explicitly reveals the information included in the assumptions. In this particular scene, the viewer selects and verifies the elements, which are indispensable for the construction of a mental model of a film (in Johnson-Laird’s understanding). The information of Mother’s “love affair” is regarded as a relevant premise for the conclusion. Then, retrospectively, one assumes that another relevant premise may be a figure accompanying Mother in a waiting room; the figure whose identity seems doubtful for the viewer. This reasoning is not challenged by other elements of the scene, so the viewer simply concludes that the man, passing himself off as the brother of the deceased, must be Mother’s lover. This conclusion derives from two assumptions: 1. Mother has a “love affair”; 2. There is a man accompanying her in the waiting room.

What is more, the solution appears plausible, because it helps achieve a generally coherent vision of the film. In the final scene of the funeral Gwen accompanies Eddie/Evangelos, Arthur stands next to Florence, while Mother is accompanied by Uncle Joe (“Irish Lover”) – in the opinion expressed by many students coupling the protagonists in this way conveys a clear message – although Eddie managed to abandon the old, hated arrangement, it is instantly replaced with another one. In one paper the author writes: “Eddie’s Mother comes with her lover to her husband’s funeral. For me, this implies rejecting the slightest chance for the disentangling of the arrangement.”

Thus, it seems that miss-recognition of the supporting character of Uncle Joe and his role in a situational model, is ultimately confirmed in the final scene. In the students’ view, a new “arrangement” which is presented at Sam Arness’s funeral, univocally identifies Uncle Joe with “Irish Lover.”

Interesting as it seems, this false conclusion, which is crucial for the comprehension of the whole story, is not verified in the off-line mode, whereby making it difficult to introduce and embed other important pieces of information into a situational model. Let us take the example of Uncle Joe’s flashback memory, which takes place when he’s playing cards with Eddie/Evangelos in the hospital ward. In a conversation with Dr. Weeks, Eddie suggests removing his father from the hospital. This is accompanied by Joe’s memory - a brief sequence of several frames shot in sepia, which show the immigrants’ arrival in New York by ship. Every flashback has its justification. If the viewers have problems with identifying whose flashback it is, no wonder they cannot establish local coherence. Of course, the viewers solve the problem by assuming in the integration process that the flashback was conceived by someone who is well-grounded in parent space of a situational model. Some authors point to Eddie and claim the flashback is his memory from childhood, others suggest that this is his father’s recollection. Similar misconceptions accompany the students’ interpretation of the scene when Eddie/Evangelos visits his dying father in the Old People’s Home. Uncle Joe searches Sam’s jacket and tries on his shoes unknowingly confirming Sam’s senile delusions (at the funeral Uncle Joe wears the same shoes). In one paper the author writes: “At times Eddie compares his life with that of his father’s, for example when he puts his shoes next to those belonging to his dad.” As we can see, the incongruities that occur in the phase of the integration process are solved by the viewers who try at all cost to create a coherent cognitive representation of The Arrangement.

While lowering the redundancy threshold (at least in comparison with Laura), the film form leaves the percepts of gestalts without clear verbal “labels” that might facilitate their conceptualisation, which however, does not hamper unconstraint play of imagination allowing for numerous configurations of events and characters to be tested. In consequence, during these cognitive operations of constructive and integrative character, certain “biases” occur. Constructive mistakes were spotted in as many as 62 out of 86 papers on The Arrangement (which accounts for 72% of total). In most cases the mistakes are of minor importance. However, even the analyses based on totally absurd premises, which definitely change the plot of the film, surprisingly manage to achieve the level of global coherence of situational model. These papers, which indeed prove their authors’ lack of understanding of The Arrangement as a film narrative, at the same time show that, despite their idiosyncrasies, they follow the principle of cognitive rationality and attach importance to coherence and logic of their argument.

In the case of the third film – Kieslowski’s Three Colours: White (1993), one of the most characteristic and at the same time, most frequent biases was overinterpretation (in Umberto Eco’s understanding).[4] According to Eco, overinterpretation is seeking in a film the elements of minimal meaning and assigning them with maximal consequences so that they match a prior premise (or premises). The viewer’s initiative in the top-down processes is aimed at seeking confirmations for his/her hypothesis. Overinterpretation is, therefore, not the excess of interpretation, but actually, “underinterpreatation.” Thus, we might suggest that overinterpretation is reading a film in a paranoid and obsessive pursuit that the viewer takes up.

What is symptomatic of the examined papers is the viewers’ relatively poor interest in the actual contents of the film. On the other hand, their attention is focused on the figure of Krzysztof Kieslowski as the representative of a quality art cinema. Although the viewers understand the story itself, they quickly abandon analysis in favour of judgment (valuation). As a result, they often fail to see the interesting aspects of the story itself. Few of them have noticed that the film deals with such issues as struggle for love and understanding, or that it is extremely packed with action in comparison with other films made by Kieslowski.

Nevertheless, most students focus on “symbols” and “hidden meanings” which – as they assume – must be conveyed by the film. What is more, they use the analysis as a tool, which will help them find those “concealed messages.” That is why they approach the story of Karol and Dominique with a touch of suspicion. Even the title itself is supposed to have a symbolic meaning, as it refers to the slogans of the French Revolution. A three-coloured flag – blue-white-red one – represents the ideas of freedom, equality and brotherhood. Thus, White means equality. The viewers interpret white in various ways, referring to numerous symbolic meaning of the colour, for instance, innocence, chastity, madness, death, capitulation, emptiness, truth, or chill. Having established symbolic potential of white, the students laboriously search for white elements and motives and provide us with an endless list including: white pigeons; white tiles at the Paris underground station; a white car Dominique drives; white snow, Marianne’s white bust; Karol’s white shirt and T-shirt; white opera-glasses, Dominique’s white wedding dress and veil; a white telephone Karol uses to phone Dominique; finally, birds’ white excrements on Karol’s coat. I should make it clear that among these catalogues one may also find some sensible remarks which neglect the excessive symbolic of the colour itself, focusing more on the colour interplay (e.g. the use of dirty white) or on white in contrast with other colours, such as black or red. Nevertheless, even in this area one may come upon various idiosyncrasies. In one paper we read that the other main character of the film is the colour itself: “white as a background for red and blue. The characters drive white, blue and cherry-red cars. Their sheets are red and blue; so are their toothbrushes. While the posters at the Underground are mostly red, the light reflected in the asphalt of the street changes it into a blue river. The walls of the Underground are white, while Mikołaj’s scarf he never takes off is red. In the final scene Karol stands in the prison yard, carrying a blue bag, which contains cherry kompot and a cake wrapped in the white paper.”

The procedure of cataloguing is carried out in deep conviction that there must some mysterious truth or Kabalist sense behind the interplay of colours.

If we analyse the conclusions that the students draw, we discover that their imagination knows no constraints. When elaborating on the suggested idea of equality they match quite different motives. The first scene in which Karol entering the court building “gets caught by surprise” by a pigeon, as one of the authors euphemistically puts it, makes him or her conclude: “Could equality in this world be nothing but an illusion? A pigeon flying over us in the sky – just a bird, you might say – but still has an advantage over us. It can do to us something we cannot.” Another student comments on the same scene: “In France even birds do not take Karol seriously.”

A cognitive problem that we face in the case of overinterpreting Three Colours: White lies in the fact that the viewers do not integrate film “cues” in a coherent entity on the basis of data derived from the film, as we might normally expect. On the other hand, having made a prior assumption they begin with a seemingly “coherent” entity and use a film form to confirm their already stated vision (with no restraints at all). The construction-integration processes that take place during the off-line mode of perception are here “filtered” by the clichés, which are both suggested by the context of the film (the reviews and the filmmaker’s declarations) and by the ambiguous form of the film itself. On the one hand, the symbolic meaning of the colours and the idea of equality are placed on such a level of abstraction that they do not help conceptualise the story and its message; on the other however, the urge to read the film in terms of such a symbolic results in overinterpretation.

When considering the reasons for overinterpretation (which obviously brings no sensible solutions – what can we learn about reality using the symbolic of colours?), it seems that its sources should be sought in the social reception and context of Kieslowski’s films. The phenomenon of overinterpretation I have discussed here seems to stem from the pressure of communicative context. The “late” works by Kieslowski were received by the reviewers with admiration and respect, and this attitude seems to have influenced a general tone of criticism aimed at his films, and replaced fair argumentation concerning such issues as essential truth and metaphysical illuminations the viewer is supposed to discover in Kieslowski’s films. Indeed, the necessity of seeking profound truths may be partly blamed for overinterpreting simple matters, which are actually the subject of the film.

On the basis of the empirical research I have come to the following conclusions:

[1.]  Film comprehension depends on three general assumptions:

1.     The goal assumption; the viewer watches a film and intends to concentrate on the actions taken up by fictitious characters in order to construct a story.

2.     The coherence assumption; the viewer aims at constructing a coherent and consistent semantic representation both at local and overall levels.

3.     The explanation assumption; the viewer tries to explain the situation, events and actions presented in a film, their interrelations and justification.

 

Developing those assumptions, the viewer constructs a complex cognitive representation of a fiction film one watches, and at the same time, comprehends. Representation is the result of both a construction process (concluding), and of an integration activity, whereby some pieces of information are selected, while others are neglected. This is, of course, consistent with a constructionist theory, according to which the principle of search after meaning is a fundamental principle of film perception; while “discovery” of the meaning is the consequence of understanding a film.

[2.] The examples of possible “readings” of films show that the lack of coherence in micro-scale (problems with categorising the events, misunderstanding the relations among the characters) does not disturb the coherence in macro-scale (at the level of understanding a global meaning of a story).

Problems at the level of local coherence make the viewer impose their own structure of coherence, leading to a paradox, as the reading of a film, however logical and consistent it may seem, turns out to be partially a misreading, because it goes beyond the actual message of a film. However, even in such cases, we cannot speak of film comprehension in terms of a binary opposition of understanding and misunderstanding. Instead, we should speak of an extent or a degree of a film comprehension.

[3.] Complexity of mental representation of a film stems primarily from an elliptic character of a film story-telling, which means that some pieces of information are skipped or concealed, while others are manifested in such an excessive way that the viewer perceives the events in form of gestalts, namely, complex entities whose features cannot be reduced to the properties of their elements; whereas other missing pieces of information are discovered during the top-down process activating prior knowledge. These pieces of information may also take the form of gestalts – complex and unreducible structures of meaning. This hypothesis explains why the same film may be received and understood in different ways by various viewers. According to this hypothesis, the process of a film comprehension or the process of establishing and negotiating its coherence might consist in correlating the percepts[5] of gestalts and emphasising these aspects which match one another.

In every act of recalling, discussing, summarising or paraphrasing a film narrative, which is a manifestation of a film comprehension, the viewer “flattens” and functionalises a complex “architecture” of gestalts that make up a cognitive representation of a film and its situational model. This is achieved by updating and emphasising these aspects of gestalts which might guarantee the coherence of the whole structure and which are consistent with epistemological context assumed by the viewer in order to cope with pragmatic challenges of a given situation. The most common and, at the same time, the simplest strategy of film comprehension is the strategy of causal logic. It is also the strategy most film narratives refer to, thus making it even better internalised. The viewer considers and links the events and actions of a film in terms of cause and effect, while the clarity of this cause-and-effect relation promotes conceptual processing and understanding of a film narrative.


Notes

[1] David Bordwell, Narration in the Fiction Film, Madison 1985, p. 34.

[2] Elke van der Meer, Gedächtnis und Inferenzen [in:] Dietrich Dörner, Elke van der Meer (eds.), Das Gedächtnis. Probleme – Trends – Perspektiven, Göttingen 1995, p. 343.

[3] See Peter Ohler’s works referring to John R. Anderson’s concepts and Robert N. Kraft’s experiments.

[4] See Umberto Eco with Richard Rorthy, Jonathan Culler, Christine Brooke-Rose, Interpretation and overinterpretation, Cambridge 1992.

[5] A percept is understood here as mental representation of a single sensual data, or of a set of data in form of a gestalt. This is a viewer’s “private” experience of film information.