Torben Grodal: The experience of realism in audiovisual representations The aim of the article is to analyze some of the ways in which realism is experienced. Thus, it does not aim to provide a definition of what is real, but to describe and characterize some of the different ways and elements that cause viewers of audiovisual representations to have an experience of `realism', based on the premise that `realism' can be described as an evaluative feeling based on perception, cognition and habituation. The starting point is a discussion of how the experience of `realism' is linked to perceptual specificity, but also to some mental schemas that provides typicality and familiarity, `recognizability'. Then the article discusses how some types of `realism' are based on the idea of the AV-screen as a transparent `objective' window, whereas other types of realism emphasise the role of filmmakers, reporters or protagonists, for instance by making subjective deconstructions of transparency in order to indicate the presence of the communicator. The article discusses how some types of realism focus on a specific external reference, whereas other forms, `categorical realism' wants to portray the general `essence' of things, and yet other forms, lyrical `realism' activate subjective-associative references. Reality can also be constituted by assertions: some human agencies assert that something is real, or constituted by interaction with the represented. The article further discusses how and why realism is often connected with seriousness and even with negative and painful experiences. The article finally discusses how a postmodern `reflexive' scepticism towards realism in audiovisual media is an emotional stance. Some of the central terms in describing media representations are `reality' and `realism'. Because they are central it is no wonder that they are used in many different ways, causing ambiguity or inconsistency in their use. Furthermore, the concept `reality' (and derivatives like `real') is not only central for describing media representations but for understanding `real' life and also here the concept is used in many different ways. Our concepts of what is real are based on many different elements, and in a given representation the different elements may each have their own reality status. A fiction film shot on location in New York may provide a `realistic' visual picture of New York and realistic acting but have a wildly improbable story. Animators of fantasy films may put a lot of energy into making the movements of fantasy creatures look `realistic', that is, made in accordance with the viewers' experience-based `movement schemas'. A non-fiction film may tell a true story but with pictures that only provide a weak experience of reality. That our concepts of `reality' and `realism' are based on many different parameters (schemas) in a given audiovisual representation means that the viewer performs a series of different mental operations in order to assess the reality status of each parameter. The viewer may also evaluate the reality status of the film or program as a whole. Realism is mainly a word used about representations in order to describe their relation to a physical and social `reality' exterior to such representations. Realism may be applied for fictitious as well as non-fictitious representations, because realism does not imply that what have been represented is true and `real' in all its aspects, only that the representation is experienced as being a concrete representation that is or might be true. The evaluation of realism in a given representation is based on three elements: 1. A characterization of the representation. 2. Some concept of how the `real' world looks like. 3. The relation between representation and the concepts of how the `real' world looks like. The degree to which there is a positive match between representation and concepts of `reality determine the degree to which the representation will be evaluated as `realistic'. Thus, the question of realism does not only imply a characterization of representations, but also a characterization of what different individuals, groups, cultures or epochs understand as `real', and that is a rather difficult and complex task. Some parameters on which such evaluations of what is `real' are based, are relatively universal. Thus, for instance, the basic perception of the physical world is largely based on innate dispositions and on certain anthropocentric presuppositions (like `earth is down', `heaven is up' (cf. Lakoff 1987, Johnson 1987)). There are therefore universal norms by which the degree of realism in the perceptual dimensions of representations may be evaluated. Such an evaluation does, however, not necessarily express how a given culture experiences a given representation. Even if people in the middle ages perceived the basic features of the world in a way very similar to us they were not able to represent their perception of exterior reality to the same degree as we can. They may therefore have perceived their medieval paintings as more realistic representations than we do because our representational context is different. But other parameters, for instance those norms of human behavior that anchor evaluations of realism, may vary over time, and from group to group. Even if nonverbal communication is based on innate and universal dispositions it is well known that different groups enhance or suppress certain features within the nonverbal communication (cf. Ekman, Friesen 1975).. Moral norms that are `unrealistic' or deviant in one society may be considered `normal' in another culture. Knowledge of what exists in the world differs from person to person and from epoch to epoch. Thus, in evaluating whether a given representation may be labelled realism we need to perform a historical or cultural analysis that consider the knowledge of modes of representation and knowledge of the world among representation- makers and among intended or actual recipients. The dominant trend within film-and-media studies have for several decades been aimed at describing the historical relativism of the experience of reality, but a description of the pertinent and universal features have often been neglected, thus over-emphasising the relativism of `realism'. A further complication in the evaluation of `realism' is derived from the fact that in a central use of the word `realism', the representation shall not only provide a good match to reality, but also a good match to those aspects of reality that are considered to be typical aspects of reality. Thus, many aspects of our concepts of `realism' and `reality' are not only descriptive but also imply normative evaluations of what is typical of real life. Representations of the private life of film stars or millionaires might be just as `real' as the `Kitchen Sink' representations of the life of ordinary people, but the latter will often be considered to be more centrally realistic than the former. The concept of realism is defined not only with reference to `reality' but also in contrast to two different terms, the abstract and the fantastic. This complicates things, because the abstract is not necessarily `unreal', models within science like abstract models of atoms have very real references to reality, and similarly you might argue that many (although certainly not all) abstract representations within the arts and in media are representations of some general schematic features of objects and thus representations of some very real phenomena. Abstractions are not only something that we find in models and `exterior representations, even our basic mental representations use abstractions', in the form of simplified schemas. It is therefore evident that realism as opposed to abstraction provides different meanings than realism as opposed to the fantastic. Thus the contrast between what is `abstractly' realistic and what is `concretely' realistic point to an experiential bias in the dominant use of the term realism. What is realistic must not only or not necessarily be real, for many adressees it must also be experienced as real. In order to understand the experience of the real in media representation we must look into those basic mechanisms that constitute our experience of what is centrally real. Perceptual specificity A basic component in our experience of physical and social reality is that what we perceive, what we see and hear, is perceptually determinate, particular, complex and unique. Few objects are exactly identical (except when factory-made), no situations will repeat themselves in exactly the same way in real life. But even if we perceive a given situation or object as something unique, our minds will process these particular and unique phenomena into something more simple and more general. We perceive the particular uniqueness of a given tree but on the other hand our minds recognizes that it possesses some non-specific features (trunk, branches, leaves). When we store the perceived in memory we will tend to store it by a rather compressed, general description, although we may recall the specificities when we see the perceived again. Our slotting of what we perceive into schemas provide that superior categorization that provides the background for our comprehension of reality. In a developmental and historical perspective representations take their point of departure in the abstract, schematic descriptions. Small children and our prehistoric ancestors draw or have drawn humans by making circles for head and eyes, lines for limbs etc. Verbal descriptions start on a very general categorical level, man, woman, river, and the ability to provide very specific and detailed `realistic' verbal descriptions has, as shown e.g. by Erich Auerbach(1974) in his book Mimesis only been developed through a long historical process. Thus, in a historical perspective, the particular and unique objects have been the hallmark of direct perception because only that could provide the necessary complex-specific salience typical of the experience of the `real', whereas mental or physical representations might be `schematic'. All other things being equal, perceptual uniqueness and complexity enhance the feeling of realism, because the representation is directly simulated in our brains as if we were confronted with reality. Stick men or computer graphics with poor texture are less realistic than good photographs. The perceptual realism is often dealt with in relation to `indexicality': A photograph or a film is supposed to be the true imprint of reality via the photographic process. However, seen from an experiential point of view it is not only our knowledge of the photographic process that provides the feeling of `realism' but the very salience of the experience, and the `realistic' salience may be created independently of any indexical relation, because it might be `staged' or produced in a computer. Seen in a developmental perspective, knowledge about `indexicality' is acquired, small children take film and television for granted and their fascination of audiovisual representations and eventual crude distinctions between more or less `realist' programs take place even if they have not the faintest ideas of `profilmic events' or `photographic processes'. The impression of reality caused by a complex-particular film sequence is an immediate reaction, based on innate dispositions in combination with universal basic experiences of reality. Many people might find it difficult to understand why `perceptual existence' experientially precedes a full understanding of reference and its reality status, but the simple explanation to this is that, except for memories, dreams and thoughts, the innate functioning of our brains presupposes that `seeing' and `existence' is the same thing, because our brains were not constructed in a media environment. Thus, for small children, realist films may not be perceived as representations of some profilmic events, but as a display of some `humans' and `scenes' with a special reality status (they cannot be touched or interacted with). Even for adults it demands a special effort to imagine the `profilmic event' as distinct from the immediate experience. Our increasing knowledge of `reference' is added to and enriches the phenomenal experience, in the same way as our knowledge about the rotation of the earth etc. may be added to and enriching our basic experience of the sun as rising in the morning, crossing heaven and setting in the evening, without, however, to replace the basic phenomenological experience. I know when seeing the evening news that they portray a profilmic event in some remote TV studio, but my immediate experience does not reflect that fact. They are somehow present in my TV set and in my consciousness, and only a special effort enable me concretely to imagine the profilmic circumstances. Our knowledge modifies our experience of realism. We know that a science fiction film is staged or animated, we know that even if the famous Rodney King video tape has a poor quality, it shows something that is real. There is thus a potential conflict between `realism by perception' and `realism by knowledge'. It has been argued by Wolterstorff (cf. Plantinga 1997) that fiction films do not assert anything about the truth of a given story, but just invites us to consider that state of affairs, in contrast to nonfiction films that are assertive because they assert that the films' states of affairs actually have occurred in actual worlds as portrayed. However, perceptual salience is an assertion of existence, even when it comes to fantasy creatures. But those assertions may to a smaller or larger degree be contradicted by our cognitive assessment of `realism', that is, our evaluation of whether the representation has a possible concrete existence as portrayed. Thus all audiovisual representations are assertions of existence. The specification of which type of existence must be decided by what Branigan (1992) calls `decisions about assigning reference', based on our knowledge about the world, including on our trust in the addressing agencies (filmmakers, TV stations etc.) and their assertion of reality status (true, invented etc.). Schematic typicality and familiarity The experience of realism by perceptual salience, including uniqueness and complexity (mostly) takes place on the background of basic familiarity and typicality, because our understanding of what reality looks like is based on previous experiences that are stored as `familiarity' and as `schematic descriptions' of movements, objects, persons, or situations. We have produced some mental schemas of how people walk or how working class people behave that we use in order to evaluate whether a given representation is `realistic' or not. If a representation is very much different from our schemas we may think that it is not very realistic. But if it looks just like our schemas without any specificity it is not felt as being very realistic either. The purely schematic representation may be felt as being a stereotypical representation that lacks the flavor of the particular and unique time/place/object. As pointed out by the Russian formalists, salience decay over time if we become `over- familiar' with and habituated to some experiences, and thus the feeling of `realism' may also decay. Realism could therefore be considered as a balancing act between the unique that provides the `salience of the real', and the typical that provides the cognitive credibility and familiarity of the real. As our imagination and representative capacity often rely on schemas and typicality, the confrontation with the nonfiction reality is often important for creating specificity. The bias for typicality provides a certain normative dimension in the concept of realism, because there are many real phenomena and representations of real phenomena that are not typical or familiar. Realism has, for instance, as mentioned, an anthropocentric bias. A high angle shot is experienced as less `realistic' than a shot from the level of the eyes of a standing human eyes, except when motivated as the vision of a bird or a person flying. But even if a deviating angle is motivated, the feeling of `deviation from the typical' persists in the experience. Furthermore, our norms of `realism' are influenced by considerations of `canonical view' (cf. Grodal 1997, Humphreys, Bruce 1989). The recognition of an object is not equally easy from all angles. A horse seen from behind is more difficult to recognize than a horse seen from a sideward position. Most people will draw or record objects from a canonical angle in order to provide easy recognition. Canonical angles may also be felt as being more `realistic', less expressive, even if the object in principle is also familiar from non-canonical angles. Historically, `realist representations' has often preferred the daily uneventful life, especially that of the middle and the lower classes. Thus, de Sica's Bicycle Thief has for many film critics been regarded as a prototypical realist film because it shows the problems of everyday life for ordinary working class people. But such a narrow definition of `realism' does not seem to correspond to what viewers or even critics perceive as `realism'. Even representations of intense actions such as emergency room events or strikes -- as in Biberman's Salt of the Earth - may be perceived as `realistic'. The preference for `uneventfulness' and `daily life' in the concept of realism of some critics may be seen as a polemic against those fiction films that stage `improbable actions' that has a low relevance for ordinary life, and not necessarily a polemic against those representations that portray more extraordinary situations (like sudden illness or social conflict) that may have a high degree of relevance for `ordinary people'. A common use of the word `realism' even points to an emotional impact as a valid criterion for `realism'. Scenes that portray all the gory details of mutilations may be seen as strongly realistic. Kracauer (1960) pointed out that garbage, dirt or sewer grates were important motives for film's effort of establishing `physical existence' because they ordinarily were unnoticed. But I might add that also the emotional impact of those phenomena that Kracauser called `the refuse' is important for the experience of `realism'. Transparent access or addresser-and-recording-mediated realism In several respects the typical fiction films are better able to provide a basic `transparent' perceptual-experiential realism than typical nonfiction films, although its capability for realism is only one option among several (I shall return to the strengths of nonfiction representations later). The perceptual realism of a `realist' fiction film provides the experience of a concrete, unique reference, although its realism is general and hypothetical. Ford's The Grapes of Wrath or August's Pelle the Conqueror portray many scenes that may have been true on a general level, although their specific concrete reference is staged. But during the viewing viewers will often experience the images as if they have concrete references, although they know that it is only a general and hypothetical one. The typical fiction film is based on narrative forms that follow actions and events that develop in time in a way that is similar to that of everyday experience. Fiction films are often focalized by one living being, and the existence of living focalizers provide an experience of online, concrete interaction with the what is shown. This is similar to our basic relation to the world by our personal focalization, including our emotions and concerns in relation to that world. Furthermore, the `classical' `seamless' way of presentation that is especially developed in Hollywood tries to follow the basic perceptual and cognitive rules for experiencing real life events (cf. Bordwell 1986, Bordwell, Staiger, Thompson 1988). But it is just as important to note that fiction films can stage many events that are inaccessible or only accessible with difficulty for nonfiction films. Nonfiction films cannot provide a firsthand view of a murder or the secret negotiations in government or business. Even those events that are in principle accessible (plane crash, street violence etc.) very seldom have a camera present at the proper moment. Even if a `professional' camera is present the circumstances will not allow such optimal solutions that could be made if a similar scene was shot in a fiction film. The widespread private use of video cameras may provide some footage, but then from a point of view or of a lower quality than that of a fiction film. The sheer presence of a camera may change the behavior of those filmed, whereas a good actor should be able to perform as if there was no camera present. Some nonfiction films or programs compensate for the above-mentioned shortcomings in several ways. The simplest way is to emulate the narrative form, although often with some kind of narrating agency in the role as focalizer, whether in the form of voice over or intertitles/subtitles, or by making the reporter or the documentarist into a visually present focalizer. It is also possible to take a more radical step, like the one taken by `docudramas', namely to try to restage a real event. Here we get first hand access to how the murder happened, to the secret negotiations, or to the small everyday dramas linked to accidents and illness as in docudramas about emergency situations. Thus the docudrama can provide basic perceptual realism to the portrayed events. It may even use the real life protagonists as actors in the docudrama and thus possess a basic personal specificity that allows the viewer to have a full real life involvement, because the drama and the protagonists are concretely and uniquely real. The eventually unprofessional acting may have a salience and specificity of its own, compared with professional acting. There are however some problems with the docudrama procedure. The docudrama has to be told after the fact, in contrast to the way in which most fiction films, including realist fiction films, tell the events in a hypothetical presence. Docudramas may therefore provide an experience similar to that of the `passive melodrama' that is often told from a temporal position later than that of the events. Because the viewers are aware of the fact that it is a `re-enactment' they may speculate as to its correctness, in contrast to the typical realist fiction film in which the addressers possess the full insight in the hypothetical phenomena. Thus, docudramas may succeed in making some events visually salient, but are in danger of not being able to provide the strong activation that are caused by the feeling that we are witnessing unique, real events. A more radical or paradoxical `compensation' for lack of perceptual realism consists in emphasising its shortcomings by making imperfect perceptual realism into a sign of `reality'. This implicitly links the perfect perceptual realism of many fiction films with something `unreal' and `staged', thus undermining perceptual realism as an indication of `realism'. Many documentaries and non-fiction film do not try to achieve perceptual realism, on the contrary. The grainy pictures, the imperfect focus and framing, the erratic camera movements and bad lighting all indicate that this is shot under non-staged nonfiction conditions, although such images do not emulate the way in which the human eye would perceive such situations. We may follow `real' policemen or firemen at work, and although we may have difficulties in seeing what is happening in the murky alleys of Los Angeles, we are supposed to get a feeling of unstaged `reality'. Fiction makers -- for instance the Dogma film group -- have used similar techniques with the intention of making realist fiction films. The method may be exemplified with the fake documentary (the fake home movie), like Myrick's and Sanchez's The Blair Witch Project. In this film one of the typical effects of `imperfect perceptual realism', namely the subjectivizing effect, is obvious. The erratic, handheld camera and the often bad resolution enhance the feeling of stress, isolation and weirdness of the experiences in the wood, and the amateurish quality of the representation corresponds to the helplessness of its pretended makers, the young persons in the wood. That an important aspect of `imperfect perceptual realism' (an `anti-seamless' presentation) is the production of a subjectivizing effect provides an interesting insight into the relation between the innate and the acquired aspects of `realism'. Viewers obviously learn to connect certain features in audiovisual representation with the circumstances of its recording, so that for instance they learn to connect a degraded picture coming from a transmission from some satellite in space with the realistic difficulties of transmitting over vast distances or learn to connect imperfect pictures with `live reporting'. So just as we learn how zebras look like we learn about media recording and media transmission. But on the other hand, the fact that such imperfect transmissions are often experienced as creating subjective experiences points to some innate specifications in the way human perception take place. The `imperfections' block the easy recognition of the objects. Our innate way of categorizing deviating perceptions is to think and feel that some subjective factors are blocking the basic `fusion' of mind and object world. Deviations from perfect vision and hearing destroy our normal lack of awareness of the subjective aspects (the proximal aspects) of our perception. Deviations furthermore create `stress' and arousal (cf. Grodal 2000). Given the proper context (the horrible experiences in The Blair Witch Project or the stressful work as a policeman in the Los Angeles streets), you might argue that the lack of perceptual realism is transformed into an `emotional realism' by inducing the viewer to experience the nervous, stressful situations. In the more trivial uses of `imperfect perceptual realism' the aesthetics may either serve as a rhetorical device (the lack of salience in the object is compensated for by camera motion and odd framings and angles that provide a rhetorical salience) or as a `self-conscious' emphasis on the `addresser' of the program (`we were here, we did the recording'). Thus the central question is whether the imperfections are motivated by the recording situation or not. The distinction between `perceptual realism' and `emotional realism' is linked to the problematic of the `aspectualization' of the word `realism'. The makers of many different modes of representations claim they are `realists', but only portray special aspects of reality. Terms like `magic realism', `extreme realism', `poetic realism', `psychological realism', or `social realism' shows how different modes of representation may claim that they are portraying special aspects of reality, and later in this paper I will argue for the relevance of the term `categorical realism. Even standard films may use elements derived from e.g. expressionism in order to provide `realist' descriptions of deviating mental states. These different `aspectualizations' of the word `realism' are a logical consequence of the fact that reality is a multi-faceted and an open-ended concept because new knowledge adds new elements to the understanding of what is regarded as real. We can only deal with this problem of `aspectualization' pragmatically: Some aspects of reality are typically evaluated as being more `normal' and `intersubjective' than others. There are some representations that can be called `realism' without any specifications, like Bicycle Thief or High School, whereas other representations need specifications that mark them out as special types of realism, like the `psychological realism' of Polanski's Repulsion. There are not only different types of realism corresponding to representations of different aspects of reality, such types are also evaluated and graded in such a way that some types are evaluated as more prototypical or central than others. The aesthetics of imperfect perceptual realism is the very opposite of a Bazinian aesthetics of realism. According to Bazin (1967) the essence of film art is its ability to perform an objective `mechanical' recording of the world that leaves no trace of an intervening human subject. Films and photographs should be the perfect replication of the world. We thus have two different ideas of realism: one centred in the subjective vision of the experiencer, the addresser, the other centred in the object world, based on the presupposition that such a neutral representation is possible. Some scholars and filmmakers have disputed that a neutral representation is possible. For instance Bill Nichol's book about documentaries, Representing Reality (Nichols 1991), is based on the presupposition that neutral representations are ways in which ideological representations pose as neutral and natural, and that nonfiction programs therefore must be self-reflexive and admit that they are specific representations. Imperfect perceptual realism is one possible way of marking the presence of an addresser or experiencer. But it might also be argued -- following Bazin's ideas -- that we are sometimes able to minimize our awareness of our subjective situation of observation and therefore have a feeling of such an objective `disinterested' observation of the world. That the experience of disinterestedness may be characterized as a feeling does not necessarily make it subjective, because (as argued in Grodal 1997), certain feelings and emotions are central tools for marking the reality status of given experiences. Thus, addresser-based and object-based realism are different and equally valuable ways of representation. Abstract categorical realism and lyrical representations Perceptual realism is, however, not the only way in which we experience reality, just as the narrative form is not the only form that we use in order to assemble data (cf. Branigan 1992). Many audiovisual representations use a categorical form, whose aim is not to provide a vivid portrayal of some concrete situations and experiences, but on the contrary, to extract some more abstract understanding out of concrete examples. Even if the famous documentary Night Mail uses some narrative elements (a train ride from London to Glasgow) and some illustrative visual examples of some of the activities, the purpose is not only to provide a concrete experience of a night on the night mail, the purpose is also to provide a more general and abstract knowledge of some aspects of the British mail system. Even if an important segment of nonfiction programs try to portray some very specific, concrete-unique situations, another important segment try to portray general, and thus `non-perceptible', non-concrete and non- unique situations. Thus categorizations and mental models serve a quite different function than being a `silent background' or tacit knowledge for our understanding of concrete situations. In the categorical film the abstract schematic level is in the focus of the representation. Photographs in a dictionary, say of a tiger, do not denote the concrete and specific tiger, but tigers at a schematic level (cf. Branigan 1992). Monty Python derives much fun out of conflating the concrete and the schematic use of images of trees and larks. The experience of reality that might be provided by such abstract portrayals of phenomena in the world I will call `cognitive realism' or `categorical realism' as opposed to `perceptual realism'. The difference between `perceptual-narrative' representations and categorical representations may be illustrated with the nonfiction montage sequence at the beginning of Beverly Hill Cops. We are presented with different shots that obviously have no narrative connection with each other. But they seem to have a 'theme' in common: all the shots show some aspects of the poor life in a black ghetto in Detroit. The different persons and locations have a very concrete and unique 'realistic' specificity. Had the scenes been linked by narrative elements they might just have been perceived as concrete perceptual realism. But as the viewer cannot find a narrative link, by default they search for a 'thematic', a categorical connection, and may find `poverty' and `black ghetto life'. This process does not fully deprive the individual pictures of their concrete reference, but nevertheless they get double status as concrete reference and as concrete illustration of a more abstract theme, poverty. The narrative part of the film mostly takes place in the rich Beverly Hills, and although it is very possible to see the scenes and events as focally describing the opposite theme `richness' and `white upper- class life', the categorical understanding is now only felt as a background for the concrete events, because the narrative form evoke a concrete focus. What is the source of that feeling of `realism' that is connected to categorical presentations? Plato and other thinkers within the `idealist' tradition thought that `ideas' were more real than appearances (or `perceptual realism' in the language of this paper), they were the essence of reality. But this does of course not provide a psychologically satisfying explanation, especially because perceptual realism is such an important elicitor of `feelings of reality'. However, as mentioned, when small children draw, they often draw the `formal essence', legs as lines, stomach and head as circles. The schemas are the mentally pertinent features in the experience of the ever-changing phenomenal world. I will therefore propose that there is a kind of `schematic salience' that provides a feeling of reality that is abstract and atemporal and gets its power by being the mental essence of many different experiences, in contrast to the feeling of perceptual salience that is connected to the temporal, specific and unique. In fiction films they are often dealt with as the redundant themes in the concrete narrative progression. To activate this `schematic salience' the presentation need to make concrete illustrations and exemplifications so that the schematic salience is supported by perceptual salience. But the schematic realism presupposes that the final reference is not only to the concrete and unique but also to the general and pertinent. Night Mail is not a documentary about a specific night and a concrete train, the concrete illustrations refer to the general and atemporal-repetitive. The illustrated and focused categorical presentation provides a double feeling of realism and reality. Its exemplifications provide the unique perceptual salience, and its categorical presentation provides the categorical essentialist salience. The reason I emphasise that the presentation must be `focused' is that if not, the presentation will lead into yet another type of experience, the lyrical mental-associative experience that to a much lesser degree points to the exterior, intersubjective reality. It is often pointed out that many documentaries provides a lyrical experience, and Plantinga has described three typical `voices' in nonfiction films, the formal, the open, and the poetic voice. However, no explanations have been provided for the seemingly paradoxical phenomenon that a type of presentation that should be `facts-oriented' often intentionally or unintentionally provides a lyrical experience. Let us take a look on one of the famous lyrical documentaries, Ruttmann's Berlin, Symphony of the City. The film provides a long series of shots from the life in Berlin, supposedly taken on one specific day, because it starts in early morning and ends late in the evening. Sometimes the film shows several shots of similar events sometimes it contrasts different events. But except for the fact that that all shots are taken in Berlin and shows a diurnal cycle, there is no clear focus, no arguments are put forward, and no clear categorical analysis provided. Therefore the viewers `by default' build up a web of associations between the different objects and events that only have a diffuse center in the idea that this shows `life' in a modern city, Berlin. In `real life' everything is concrete and is fixed in a continuous forward-directed time and a contiguous space. Associations between widely different times, spaces and objects, as well as associations based on similarity and other formal features typically only exist in minds (we may associate New York and Tokyo as important capitals, link different historical periods, find a similarity between eyes and sun), or in representations. An experience based on associations is typical of a lyrical experience, which contrary to most narrative fictions and contrary to the concrete documentary does not primarily refer outwards to a concrete reality, but refer `inwards' to the existence of the experiences in the mind. In the focused categorical audiovisual representation categorical `assertion' provides the link between exterior world and mental representations, but in the unfocused presentation this assertion of exterior existence is lacking. This provides some curious effects: Many lyrical films are in principle straightforward or even extreme `documentaries', for instance some films of Marguerite Duras (cf. Grodal 2000), but are nevertheless experienced as being strongly lyrical. The reason for this is that the viewer cannot find some focused categorical assertions and therefore `by default' searches the mind for possible associations, and thus establishes a subjective associative web, cued by the film. Even a typical documentary like Night Mail lingers between a concrete, focused reference and a lyrical `mental' reference, helped by the lyrics of Auden on the sound track. The film's apotheosis of `speed' and `efficency' sometimes evoke a purely `idealistic-internal' reference. Assertions of existence and realism Many scholars have been interested in assertions from an epistemological point of view. Thus Wolterstorff and Plantinga (1997) point to the fact that nonfiction films perform pragmatic acts by which they assert that what is shown and told is a true fact in the world. Some theoreticians, like Nichols (1991) is very sceptical in relation to the overt assertive aspect of nonfiction films as it is found in what Nichols calls the expository mode, for instance in the form of a `voice of God', an all-knowing voice over. The problem of assertion in establishing reality and fiction could, however, also be approached from a psychological point of view. As discussed earlier, our basic experience of reality is not only linked to the perceptual processes but is also linked to the enactive processes: Those things, those perceptions are real that can guide our (re)actions. Thus, the question of `reality' and `realism' poses the problem of agency. In the typical narrative fiction film, the question of `agency' is solved through the existence of a protagonist for whom the perceived serves to orient some actions. In some types of `observational films' the nonfictive protagonists have a salience that provides the viewer with some kind of agency that provides the necessary `enactive relevance'. However, many types of film, like for instance a categorical documentary, do not have any `onscreen' agency that can provide relevance. But the explicit or implicit addresser of such films may for the viewer serve as an agency that performs some non-concrete actions by asserting that this or that is the case. To utter performatives, symbolic actions that claims that something exists, is (as shown by Austin and Searle) a central feature of human communication. The assertive-performative activity is a central way of transmitting a feeling of reality and factuality. Assertions need not be verbal, any framing or presentation is in principle an assertion, although it may not be felt if the viewer cannot reconstruct the principles on which the assertions are made. As discussed in relation to Night Mail, the assertions made during the film are central for providing the concrete focus, the concrete reference of the representation that might otherwise be perceived in a subjective-lyrical mode. Nichols's distrust of overt assertions as an ideological mode is a rather severe exclusion of a central way of communicating about reality. Abuse of addresser-agency is of course possible, but we cannot abolish addresser-agency in general just because it might be abused. It might also be argued that there is a certain `honesty' about overt assertions that makes it possible to have a clear impression of the asserting agency. This may be seen when we compare an `expository' film and its overt assertion with a so-called `observational' film that pretends to be recorded by `a fly on the wall', and thus seemingly is an objective recording without any asserting agency. However, most viewers seeing Wiseman's brilliant `observational' documentary High School will expect that the film provide some general assertions about the American High School system, although on a basis of concrete observation. But the film hides the addresser-agency so that the film is seemingly asserted `by reality' because Wiseman hides his asserting agency that have made the concrete shooting and editing. This `problem' with the `observational' mode has been noticed (cf. for an overview Carroll 1996, 224-52), and scholars like Nichols are therefore fond of the `reflexive' film that overtly thematises that it is a specific representation. However, it might be argued that a normal, non- naïve viewer of an overtly `assertive' expository film is very well aware of the fact that this is a film made by some human agency and that therefore a reflexive representation is not the only option for representing the asserting agency. Realism and virtual or actual interactivity Assertions are a special case of the way in which reality is experienced on the basis of actions and interactions. The addresser `interacts' with the represented by making assertions. The simplest form of interactivity is based on direct presence that allows a person to react on what she hears and sees, either verbally or physically. Media representations are in principle non- interactive (except for telephone and some computer applications). Films and programs may use different procedures in order to veil or overcome this problem. The simplest form of interactivity is `passive' interactivity or interactivity by proxy. Passive interactivity consists in speaking to or looking `towards' the viewer-listener. In this way the viewer may have a feeling of being a passive participant, a concrete listener-viewer and may have a very concrete feeling of sharing a real experiential space with the persons that are present in the media and have a feeling of situational uniqueness. Interactivity by proxy means having an audience that serves as a mediated proxy for the viewer-listener or an asserting addresser. The viewer-proxy may even use an interactive media, like the telephone or the net. A strong feeling of reality by interactivity is however produced by certain computer applications, and most prominently some computer games. Despite an often rather poor perceptual definition, such games can produce a strong feeling of reality because players can physically interact with the represented world. This provides us with an insight in those strong elements in the experience of reality that is linked to our (inter)actions, our output (cf. Grodal 2000b). Reality and realism as genuineness, seriousness, and pain Our experience of reality is often linked to evaluations of seriousness of intentions and consequences, and to evaluations of `genuineness'. Even the most realistic fiction film is not fully real, because the actors do not intend to kill each other they only pretend to have that intention. This is a specific consequence of a more general principle in our experience of reality and `realism' that it demands that some vital human (or animal) concerns are at stake. When playing or pretending we are supposed to be free to discontinue that behavior, as opposed to those acts that are the consequences of our `genuine' self and its vital concerns. A transmission of a football match might rarely be described as `realistic' whereas a transmission of a fight between police and demonstrators might be seen as a piece of `realism' because vital concerns are supposed to be at stake in the latter case. But if a documentary peeps into the locker room and shows how real concerns are at stake for the players, and show their `true backstage' identity it might easily be termed `realism'. In a television genre often called `reality TV' (Survivors, Big Brother etc.) which is based on staged performances by ordinary people, `amateur actors', it is often emphasised by programmers that the participants act in accordance with `serious' and therefore `real' motives. It is difficult to draw a strong line between what is just pretension, a `role', and what is `genuine'. Motives that make people act and pretend in real life (their `front stage' appearance as it is called by Goffmann cf. also Meyrowitz 1985) may be controlled by stronger motives that their `relaxed' private and intimate (back stage) behavior, although probably a `behind the scene' documentary about actors, politicians or football players would be evaluated as a more `realist' representation than a presentation of `front stage' performance. To sing a song, play an act or perform a sketch might be seen as being just as `real' as to work on an assembly line. However, it seems that the concept of `seriousness' and genuineness serve some vital function in the evaluation of what is real or realistic. It might be an implicit reflection on that problem that `realist' representations often prefer situations in which the agents are supposed to have a limited capacity of `acting'. Even if the life of the hero of Bicycle Thief from one perspective is trivial and uneventful, the loss of his bicycle and therefore possibly his job is an `existential crisis' that guarantees that his behavior is serious and genuine. The reality TV concept Big Brother emphasises its ability to evoke genuine behavior by pointing to elements of possible deprivation (Confinement for a long period of time in limited space that is under surveillance, nine out of ten participants have to be excluded by painful processes, the participants should be reduced to a basic or even deprived situation compared with present day life standards etc.). A consequence of the principle that a strong feeling of reality demands that vital human or animal concerns are at stake is that `realism' is more often attributed to those representations that portray negative emotions than those that portray positive emotions. This is perhaps based on the assumption that `pain' is more real than `pleasure' and thus evokes a more genuine behavior. A documentary about deprived and suffering people would probably more easy be labelled `strongly realistic' than a documentary about some happy jetsetters, even if it is not clear whether the first group is more typical than the other. Some of the classic realist films, like Ford's The Grapes of Wrath, or Biberman's Salt of the Earth, focus on deprivation and pain. Some philosophers and artists think that `death' is the ultimate reality. Freud even called those mechanisms that were linked to reflect on pain for the `reality principle' as opposed to the `pleasure principle'. Although Freud's terminology does not make sense from an evolutionary perspective, it might reflect a fundamental feature in the way in which humans experience reality. The psychologist Nico Frijda (1988, p.353) has claimed the existence of a `hedonic asymmetry': `Pleasure is always contingent upon change and disappears with continuous satisfaction. Pain may persist under persisting adverse conditions.' Thus, the bias in realist representations for describing suffering or deprived people may not only reflect a political wish for advocating empathy, but also a widespread feeling that pain and deprivation is more real than pleasure, although personally I understand but do not fully share that feeling, that evaluation.. Realism in an age of pervasive representations A series of scholars, for example Jean Baudrillard (cf. Grodal 1992), have argued that the fact that people in the age in which the `online' experiences plays a decreasing role in comparison to experiences linked to audiovisual representations leads to a dissolution of our experience of what is real, and this entails that the concept of realism becomes increasingly problematic. The salience of images does not warrant any real concrete and unique reference, and as even the perceptual world become increasingly man-made, there are no absolute and natural norms to anchor an authentic experience of reality. The present pervasiveness of representations even in every day life certainly raises some problems as to what is true or probable (realist), and what is only invented and improbable (fantastic or just lies and misrepresentations). However the problem of the reality of representations is at the heart of even the most rudimentary premodern verbal or iconic representations. From the time of the invention of language, sculpture and painting strange tales have been told about gods, magic causations or cyclops with only one eye, or visual representations of gods and other fantastic beings have been shown and asserted and believed to be true representations of actually existing beings. Even if premodern humans typically only had visual and acoustic access to a relatively narrow concrete environment, many of their beliefs about the world were molded by what they were told existed elsewhere. It is not obvious that modern viewers of fantasy and science fiction films have more trust in the realism of such representations than premodern humans had in fairies and gods, on the contrary. Even if `perceptual realism' activates innate dispositions, the massive exposure to audiovisual representations does probably not create more credulous viewers than verbal communication created credulous listeners. Some questionable representations are even more easily performed in a verbal than in an audiovisual media. Thus generalizations like `New Yorkers are happy and active' are easy to make verbally, but if they are illustrated visually even naïve viewers might have some doubts about the validity of the generalizations. It is therefore not obvious that the world moves into a direction of a crisis of representations of reality (except in the sense that the increase in knowledge increases the awareness of possible problematic relations between world and representations). Even heavy media users have an extensive experience of a non-media world, including an experience of media-platforms as physical objects, and their first-hand experience of other cultures will typically be greater than ever before. There is no reason to believe that the basic `perceptual' realist experience of the everyday intersubjective world have changed radically, it still provides the norm in relation to which special represented phenomena are evaluated and characterised. Furthermore, there are more complex public discussions about the validity of representations than ever before, so the proliferation of (audiovisual) representations is matched by an expansion of critical discourse. This does not mean that people cannot be seduced into believing that biased and stereotypical representations are real or realistic. All types of human communication are possible learning processes, but the physical co- existence of slaves and masters, or Nazi-officers and victims did not prevent strange concepts. Those high modernist and postmodernist discourses that have centred on a crisis of representation and `reality' may therefore better be explained from the point of view of the emotional dimensions of the cognitive appraisal of what is real and realistic. As briefly mentioned earlier, a vital component of our experience of reality is emotional, the feeling that something exists and thus can serve as a possible object of or scene for action. A central aspect of this feeling of reality is the feeling of a certain basic familiarity. Periods of rapid change, including expansions of representations may cause alienation, an emotional dissociation that make feelings of familiarity and thus the feeling of `reality' more difficult for many people. To some extent, therefore, some aspects of high modernism and post- modernism do not reflect `the essence' of the present period seen from within, but are on the contrary emotionally and experientially rooted in the conflict between past and present. Thus, those modern and `postmodern' critics and artists that express their scepticism of `realism' and their interest in `metafiction, `reflexivity¨ and `intertextuality' may be seen as symptoms of a cognitive and emotional period of transition that for emotional reasons need to foreground the representedness of representations. Even if postmodernism in principle should be against grand narratives and grand generalizations, their negative epistemological claim of `uncertainty' and scepticism is a grand historical narrative, that implicitly expresses a longing back to a Cartesian world in which some divine principle guaranteed the link between world and mind. However, the feeling of reality and realism is based on and serve our pragmatic interaction with the world. Whether we communicate with grunts, with words, or by means of audiovisual media we make and use concepts of what is actually and uniquely real or what is `typical', and we trust some communicators or representations as being better guides for our actions and concerns than others. I therefore think that concepts like `real', `realism' and `truth' are pertinent even for a postindustrial society with pervasive audiovisual representations. References Auerbach, E. (1974). Mimesis. The Representation of Reality in Western Literature. Bazin, A. (1967): What is Cinema? Berkeley: University of California Press. Bordwell, D. (1986): Narration in the Fiction Film. London: Methuen. Bordwell, D., Staiger, J., Thompson, K.(1988): The Classical Hollywood Cinema. Film Style & Mode of Production to 1960. London: Routledge. Branigan, E. (1992): Narrative Comprehension and Film. London: Routledge. Carroll, N.L Theorizing the Moving Image. Cambridge: Cambridge University Press. Ekman, P., Friesen, W. V.(1975):Unmasking the Face: A guide to Recognizing Emotions from Facial Clues. Englewood Cliffs, NJ: Prentice Hall. Frijda, N. (1988): `The Laws of Emotion', in American Psychologist Vol. 43, No. 5 349-358. Grodal, T. (1992): `Romanticism, Postmodernis and Irrationalism' in Sekvens 92 `Postmodernism and the Visual Media.