Introduction

In attending to the diverse history of cinema, one is tempted to wonder what forces might have been at work in its shaping. Why exactly is it that a typical Hollywood movie is the most popular form of cinema, and could it have turned out to be otherwise? And surely the history of film must be regarded as a progression towards a more effective mode of narration? Such questions ultimately fall back on the more general puzzlement regarding whether film history is best explained in terms of a natural or cultural development. That is, do we find certain forms of film more accessible because we are biologically predisposed to do so or because they belong to a cultural tradition with which we are familiar? The study of early cinema will inevitably run into questions such as these, especially when it is bent towards explaining the evolution of its formal properties. Care is necessary all along because the assumed viewpoint will shape, perhaps to a greater extent than is often realized, our consequent conceptions of the communicative, expressive, and meaningful dimensions of film, and art in general.

In much literature (not only structuralist), there is a healthy suspicion towards the inclination too readily to view what are actually customs and conventions as natural rather than cultural. It might seem natural that the earth should be represented with the Northern hemisphere upwards, or that higher tones should be called 'higher' rather than 'lower'. But of course, on reflection these matters are no longer as naturally 'given'. These instances might not seem ideologically important, and thus not always very pressing. But such a 'natural' line of thinking might on other occasions prove to be more obviously reactionary, as when men were said to be naturally superior to women, or whites to blacks, etc. Thus, when American historians argue that the American style of movies is more popular than others (existing or not) because it is more naturally accessible, there is reason to suspect that this might be a consequence of too limited a viewpoint. Conversely, however, stating that all conventions are cultural through and through leads to nothing more than the loss of a useful distinction. What would be the point of insisting that a custom is ‘merely’ cultural, if we agreed that it cannot be otherwise? Would we not still say that some choices are less arbitrary than others? Surely, one might argue, choosing people's faces rather than, say, feet as the object of portrayal is not mere convention, but rather has its quite natural explanation. And what is more, the fact that most film historians are westerners - and thus liable to be 'biased' - does not automatically make whatever they say incorrect.

The main purpose of this paper, therefore, will be to discuss the attempt to decide whether the force behind the evolution of early cinema was natural or cultural. It is to be my contention that such an attempt cannot be definitely successful, because it rests on either metaphysical speculation, or a rather restricted view of (film) art. The frequent use of words such as 'form' and 'function' in contemporary historicizing of cinema bears witness of a very persuasive conception of film which, in ways which will become clear as we go along, has certain strong ramifications for how culture and nature in cinema history are conceived. This in turn will lead me to discuss the perceptual/philosophical underpinnings of these more or less (usually less) explicit 'formalisms' (Actually, 'formalisms' here also include anti-formalisms, to the extent that they basically share the same view of how art can be divided into form and content, or rather, text and context.). Thus, of equal concern will be to investigate what we are in fact saying when we argue for one solution or the other. My task, then, will not be to offer a solution to the puzzle, but rather to work towards its dissolution.

The Nature of the Art

RECEIVED history tells us that film in its earliest days was concerned mainly with the fascination with represented movement, a "harnessing of visibility” (From Tom Gunning's seminal essay on the earliest cinema, "The Cinema of Attractions: Early Film, Its Spectator and the Avant-Garde", reprinted in Thomas Elsaesser (ed.) Early Cinema; Space, Frame Narrative (BFI, 1990)). After this, along with the technological development, it gradually came to the service of storytelling. This shift is usually regarded as a change of focus from showing to telling. Initially the documentation of reality was, as it were, self-sufficient. Put bluntly, the employees leaving the factory were simply employees leaving a factory (I refer, of course, to the Lumière brothers’ first film, Workers Leaving the Factory, of 1895), whereas in the later classical 'continuity' mode such a scene would have been fraught with meaning, seeing as it would then be symbolic, part of a language of film narration. Or as Thomas Elsaesser puts it: "the early cinema situates itself in a tension between presentation and representation, performance and narrative, showing and telling - with the proviso that one should not see such binary oppositions as mutually exclusive, but rather as polarities that imply each other or 'values' that regulate each other." (Thomas Elsaesser, "Afterword", in Early Cinema; Space, Frame Narrative (BFI, 1990), p. 405.) Following Noël Burch, this change is often referred to as a change of modes, in which different sets of logic - and so norms - apply. ( See e.g. Noël Burch, Life to those Shadows (BFI, 1990), where he specifically speaks about "modes of representation", implying a range of norms not limited to cinema, but to representational conventions more generally. David Bordwell and his fellow historical poeticians (e.g. Kristin Thompson and Janet Staiger in The Classical Hollywood Cinema has made further use of the notions of norms and modes, although also of practice and production.)

Burch identified the change as something which could largely be explained by reference to the hegemony of the bourgeoisie. The new mode was imposed because it fitted with the respected mode of representation, complete with a way of 'positing' the spectator in a voyeuristic¬ /contemplative - rather than critically distracted - relation with the movie (the alternative would subsequently become the project of the Avant-garde). In other words, Burch held that various formal properties, such as certain practices of editing and mise-en-scene, were constituents of a bourgeois way of narrating, and subsequently that the increasing presence of them proved his theory right. On closer scrutiny, however, much of Burch's film history, although still highly regarded, proved to be of an a priori variety. Along with the renewed interest in cinema's early days came people like André Gaudreault, Tom Gunning and David Bordwell to point out Burch's shortcomings, and that he tended to draw somewhat hasty conclusions and fall back on speculation rather easily. (Bordwell and Thompson present a thorough critique of Burch's film history for sometimes disregarding or even altering the facts to suit his overall arguments: "Burch's 'materialist' history rests upon assertions which, however intriguing as speculation, remain unsupported, and often unsupportable, by concrete historical evidence" (p. 10), in "Linearity, Materialism and the Study of Early American Cinema", Wide Angle 5:3 (1983), pp. 4-15. Elsewhere, they have forwarded similar arguments, i.e. concerning the lack of factual accuracy and questionable methodology, against Barry Salt's historical account.) An accurate understanding of film history, it was now argued, would not be reached in an arm chair, but rather in the archives (This attitude, moreover, seems surrounded by an air of optimism (and even positivism). On many occasions, one can read about the 1978 FIAF Conference at Brighton, after which the study of early cinema apparently took its first steps on 'solid ground', so to speak, towards a 'clearer understanding'. Cf Tom Gunning, ”Non-Continuity, Continuity, Discontinuity: A Theory of Genres in Early Film”, Iris, 2:1 (1984) , pp. 101-112: ”There is no question that the understanding of this history ultimately must include [...] a close and comparative viewing of all existing films with the tools of analysis that structuralism and semiology has provided to film study”, ” the time has now come for a diachronic comparison of filmic systems within history.” (p.102).).

A closer and more detailed study meant an analysis focused on specifics, i.e. in the case of cinema such measurable objects as shot lengths, camera positioning and movement, editing, lighting, etc. In this light, the prime interest in the development of cinema lay in the discovery of the changes of 'formal' practices. Changing editing practices, for instance, could thus be interpreted as changing modes of the construction of space - in a poetic and aesthetic sense - which in turn - under headings such as 'continuity' ('Continuity' must be one of the most varyingly used words in the history of film, having distinct, though related, senses in a number of contexts, such as continuity of narrative time and space, of pictorial composition, understood varyingly as a poetic ideal or as an aesthetic experience, sometimes supposed to be value-laden and sometimes neutral, and other times again referring to a specific linear conception of the development of cinema, and so on.) - would give rise to questions concerning causal explanations. That is, why did the 180 degree rule, point-of-view, or 'shot/reverse shot' editing, etc., come to receive such dominant roles in the 'Classical Hollywood Paradigm'? Apparently, such a material poetics would have the advantage of dealing with easily observed and organized historical 'facts'.

The proper object of historical study, then, would be form and its functions. Although it might not at first glance appear that way, movies made today share quite a few formal principles with the earliest ones. There are, however, some such principles - ways of narrating - which for one reason or another have not prevailed (or which have become increasingly uncommon). Seeing such 'extinct' formal strategies, one is often struck by how clumsy they seem. To take an example from an early Swedish movie, Thomas Graals bästa film (Mauritz Stiller, 1917), there is a scene with a married couple in a room, where the husband leisurely attempts a trick consisting in kicking a cigarette from the floor and next trying to catch it with his mouth. In performing the trick, he accidentally not only succeeds in getting the cigarette airborn, but simultaneously the slipper used to kick with, letting it go so it disappears to the left of the frame. Next there is a shot of his wife, seated and preoccupied with something of her own. A second later we see the slipper land right next to her, presumably giving her quite a scare. Now, what in this context is interesting about the scene is how it is narrated, i.e. how Stiller chooses to tell and show us this event. The slipper, as noted, left the frame on the left. In the subsequent shot, of it landing on the wife, however, it enters again from the left, quite the contrary to our expectations. Normally, that is, a shot depicting something having gone off to the left of the frame would be followed by a shot of it entering from the right. 'Normally', that is, at least to us experienced cinema goers of the nineties. Although we may be a little disrupted by the unusual representation of the route of the slipper, we do, however, understand the sequence (without much effort at that). Here the question arises, though, if we were thus disrupted because we have become more used to the alternative way of depicting similar events, or if the alternative somehow is a naturally less disruptive way of narration (which in turn could explain why it has gained dominance). In other words: there being several forms which can perform the same function, what makes one of them historically preferable?

Another example is given by David Bordwell, in "Convention, Construction, and Cinematic Vision" (in David Bordwell & Noël Carroll (eds): Post-Theory: Reconstructing Film Studies (Wisconsin, Madison: The University of Wisconsin Press, 1996), pp. 87-107.), where he discusses the emergence of 'shot/reverse shot' editing as we know it today. He notes how the same event, of two people talking to each other face-to-face, might be (and has been) represented in several different ways. In doing this, he discusses problems with both the "naturalist" and "conventionalist" assumption of the development of cinematic language. Disregarding the most 'primitive' way of filming two people talking, without cutting or movement of camera (presumably because he does not regard this 'tableau style' as in the relevant sense a "device"), Bordwell argues that the so-called "naturalist" position can't be correct, since the shot/reverse shot is not sufficiently faithful to "spontaneous perceptual activity" (Ibid., p. 88.) to thus explain its dominance. "The best equivalent to a viewer moving her or his glance from one character to another would seem to be obtained by simply swiveling ("panning") the camera from speaker to speaker." (Ibid., pp. 88-9) Instead, the most common way to represent two people talking is not by panning, but by cutting from one character to the other, while simultaneously changing camera positions. But, as Bordwell recognizes: "this deviation from the natural-equivalent premise opens the door to quite a different theoretical position. Once shot/reverse-shot cuts are admitted to be dependent upon purely "artistic" considerations, we can ask if they are not simply conventions." (Ibid., p. 89) He then goes on to question what is to be understood by the "conventionalist" argument. He dismisses the suggestion that shot/reverse shot is an arbitrary technique, and proposes we "bypass the nature/culture couplet" (Ibid., p. 91), and understand cinematic conventions as something in-between, as it were, neither fully natural, nor totally arbitrary, determined by what he calls "contingent universals".

Bordwell's account echoes Noël Carroll's insistence on conceiving of cinematic forms as "cultural inventions", rather than being wholly cultural (i.e. arbitrary) or natural. Discussing the emergence of the plow as analogous to that of certain movie techniques he writes: "It was adopted because it worked, because it met a cultural need by accommodating features of nature and biology." (Noël Carroll, Mystifying Movies: Fads and Fallacies in Contemporary Film Theory (New York: Columbia University Press, 1988), p. 143.) He repeats this view in a recent article, where he stresses the equivalence of point-of-view editing to ordinary perception (Noël Carroll, ”Toward a Theory of Point-of-View Editing: Communication, Emotion, and the Movies”, reprinted in Theorizing The Moving Image (Cambridge University Press, 1996), pp. 125-138.). In the piece he differs from Bordwell, however, by more explicitly defending the idea that perceptual proximity is an ideal in the shaping of a popular film language. Whereas Bordwell carefully states that "It seems very unlikely that our abilities to recognize humans and objects in images owes nothing to our biological heritage" (Bordwell, ”Convention, Construction, and Cinematic Vision”, p. 91. Original emphasis.), Carroll on the other hand takes this as a matter of course, arguing that this is the prime reason why some movies ("Hollywood International") are enjoyed by large audiences all around the world. However, as Carroll is eager to point out, such natural movie narration is culturally motivated, i.e., "supposing that the aim of mass movie entertainment is to engage (numerically) mass, untutored audiences, point-of-view editing is a ready source of communication because of the way in which it taps into or exploits biologically rooted, perceptual behaviors." (Carroll, ”Toward a Theory of Point-of-View Editing”, p. 134. Emphasis added.) Nowhere does he argue that movies were destined to be shaped by what is naturally accessible, he only states that such was their fate in our capitalist culture.

That point-of-view editing is keyed to our biological makeup undoubtedly enhanced the reception of the point-of-view structure. This is not to say that every successful communicative structure deployed by the mass arts will be biologically rooted, but only that biologically rooted structures, like point-of-view editing, will be attractive devices [...] in promoting visual comprehension on a mass scale. (Ibid., p. 135.)

Two lines of objection might be raised at this point. First, there is the concern that such a focus on the perceptual accessibility of cinematic forms risks neglecting other shaping factors. It is, for instance, imaginable that part of the reason why point-of-view and shot/reverse shot editing won precedence was that these devices allowed filmmakers more freedom than their alternatives. One such advantage with the shot/reverse shot technique is that it is possible to thus represent two people talking to each other without both of them having to be in front of the camera at the same time, etc. And there are of course other similar material, technological and practical aspects which might be overshadowed by that kind of emphasis on formal aesthetics (As is the case with any focus, however purely descriptive its aim, formalism has its normative aspects, since it can be taken to imply a relation between modes, whereby there are norms and deviations from them, rather than independent sets. This is presumably what Alison Butler has in mind in her review of Bordwell’s project, ”New Film Histories and the Politics of Location”, Screen 33:4 (1992), pp. 413-426.). A second and perhaps even more serious problem with these accounts is that it is not certain whether what seems natural today also seemed natural to the cinema audience of a hundred years ago. To return to the example of the slipper in Thomas Graals bästa film, the viewer of today finds the unusual representation of its flight somewhat disrupting. But it seems that we cannot decide for certain if this is due to our being accustomed to another way of depicting such an event, or if that is naturally the case. In the history of the philosophy of visual art, there is a similar dispute regarding the 'naturalness' of perspective drawing. On the one view (e.g. Erwin Panofsky’s & Nelson Goodman's) perspective is, like everything else, simply a convention ('made, not taken'), whereas on the other (e.g. Ernst Gombrich's), it in fact allows us more closely than its alternatives to capture how things really look.

Senses of Sensibility

IF one thing has thus far become clear, it is that talk of nature and culture in relation to film form is not always very clear. If it is stated that a cinematic technique is either natural or cultural, we still stand in need to ask a series of questions: in what respect is this so, how do we determine whether or not this is true, and why is this being said? Following Ludwig Wittgenstein, we might want to investigate the 'grammar' of the words, in this case concerning the nature/culture couplet, in order to see what assumptions underlie their use. Remembering that the words in question are just concepts with which we express ourselves (e.g. rather than names for something objectively observable), we can ask how it makes sense to say whether conventions are natural or not.

In the arguments presented so far there is an insistence on seeing cinematic conventions as neither wholly natural, since they are representational devices ('constructed and signifying'), nor wholly cultural, since they are not totally arbitrary. This would entail that it could not make sense to speak of perspective drawing or point-of-view editing being more or less natural conventions. But this is not the case. We do describe intelligibly the cut of the slipper entering from the 'wrong' direction as a less natural - or immediately graspable - device than its alternative. We are perhaps careful in doing this if we suspect that at the time of its release the chosen cut might in fact have been regarded as the more natural. If somebody were to include such a cut in a movie today it is likely that we should view it as, say, a mistake (or pastiche, or avant-garde, or what not) since today, at least, such a cut would not be perceived as natural. (Note that it is the choice of form, not the form ‘on its own’, that is more or less in accord with what we regard as natural, and hence judged thereafter.) In a trivial - but not unimportant - sense, all cinematic devices are cultural, conventional. That is, films are man-made artefacts, and not natural phenomena. Nobody claims otherwise, of course. What the naturalist position seems to be saying, rather, is that some devices are better than others at representing reality. They are better at "promoting visual comprehension", to borrow Carroll's words (quoted above), presumably since they are "biologically rooted". The conventionalist view also says that some forms seem better than others at representing reality, but that this is simply due to our being 'conditioned' to them.

When we regard a style as 'realistic' or 'natural', we do this because we are compelled to feel that it is somehow unstylized, or untampered with. Of course, at the same time we acknowledge that it is a style all the same. 'Realism' is thus properly conceived of as a style like any other, or rather as a cluster of styles. What makes up a realist style is heavily context dependent, and in no way 'untouched reality'. Rather, realisms are better treated as aesthetic movements which emphasize ‘aspects of reality’ which have been ignored in earlier styles. Thus, what one generation or culture holds to be realist art might seem heavily stylized to another. There cannot, in this sense, be a perfect realism ('reality itself, yet represented'...), nor a perfectly natural form. Still, we cannot help but experience perspective drawing as the most natural device of representing space, just like we cannot help perceiving the slipper-cut as less natural than its counterpart. That is, we cannot disregard our perception of reality in order to compare it to others, or independently of our own perspective set up criteria by which to evaluate it. In other words, we are necessarily unable to show how the slipper-cut is ‘merely’ a convention, even though we might argue that that is likely to be the case. Consider, for instance, historical movies made in the past, say in the 70's, which have a distinctly ‘seventies’ look and feel (the cast having dated haircuts, etc.), which people at the time presumably were not aware of. Similarily, a movie like Scorsese's The Age of Innocence, although perhaps seeming timeless to a contemporary audience, will necessarily be seen as a distinctly nineties-product to future audiences. Just which aspects of the film will be regarded as particulary dated is not something we can know beforehand, however. Indeed, filmmakers often strive to give movies a timeless look, e.g. by avoiding obviously passing trends, but the finished product will always contain aspects specific to the time of its making. (There is a specific kind of exception to this observation, viz., a movie might mimic a style of a past era so as to possibly deceive the viewer about its origin, e.g., a movie made in the 1990’s might so successfully capture the look and feel of a 1920’s feature that it does not bear any marks of the time of its actual making. But, importantly, such a movie is also confined to the specificity of style, i.e. it does not transcend stylistics.) To some extent, then, although we cannot show how, the forms that we perceive as natural are culturally specific.

At the other end of the spectrum, the view that an artistic convention might be totally arbitrary is a curious one. David Bordwell argues that an action "counts as arbitrary if the same goal could have been achieved by an alternative means, with no additional costs or difficulties." (Bordwell, ”Convention, Construction, and Cinematic Vision”, p. 93.) In this respect, for instance, deciding whether to, at a given moment, rotate the camera clockwise or counter¬ clockwise by tossing a coin, makes the outcome arbitrary. But there are problems with this conception. In another respect, the choices to rotate the camera at all - and to let the tossing decide which way - are not. Also, the reasons for choosing one method rather than another would often more suitably be characterized as aesthetical than as rationally calculated (just like we say that the choice not to eat people is an ethical one, rather than simply a convention based on calculation. (Of course, one could object that such aesthetical considerations are a luxury, reserved for filmmakers who do not necessarily have to think about maximizing profit. Although this is true in one sense, i.e. one's aesthetic freedom can be compromised by such 'exterior' motives, it does not undermine the argument, since what one is confined to under such circumstances is just another aesthetics, viz. a profit maximizing aesthetics.)) Still, this does not stop us from coherently calling some artistic choices arbitrary. But usually when we do so, it is an aesthetic response, judging a choice to be more or less arbitrary, e.g. implying that for some (aesthetical) reason, we do not feel that, say, rotating the camera in the opposite direction would make any difference, or that it was a very imaginative choice. That is, intentions are aesthetically interesting. Thus, when many of the devices used by Dreyer in Prästänkan (1920) are experienced as disruptive, we might very well be inclined to take it that such was his intention, i.e. at least he probably did not attempt to copy the continuity model of his time, but worked according to an altogether different logic. He might not have succeeded at reaching one goal or another, but in so far as we do not take ‘American continuity’ to be one of those, we have no reason to treat the disruptive moments as mistakes. If, on the other hand, we do suspect Stiller’s slipper-cut to be a mistake, this is because we imagine he might have been trying to follow the American model of movie narration

In regard to the nature/culture dispute Bordwell proposes 'the golden mean', so to speak, by arguing that certain forms are better than others since they rely on 'contingent universals' - part nature, part nurture. As he recognizes himself, this alternative at least has the advantage of 'getting on with it', without having to decide once and for all which view is the correct one. What he shares with the 'dogmatists', however, is the view that we need certain skills, be they biologically or culturally rooted, in order to understand what film forms communicate, and that this in turn would explain the shape of cinema history.

A Misconception of Perception (and its Perception of Cinema)

BUT how are we to decide whether a certain device really is natural, or if it purely seems that way? As already suggested, it is a mistake to treat the matter as if it were a question of the device ever being one way or the other. Rather, the chosen technique seems more or less natural: to our eyes, perspective drawing seems to be the most natural way of depicting space, be that as it may. Hence, it is not a very fruitful way to go about asking whether or not the slipper-cut in some metaphysical way is less natural than its alternatives. To us, westerners of the nineties, the cut seems 'unnatural', whereas it is imaginable that to the Swedish audience of its time it seemed 'natural', i.e. not more disruptive and unusual than its alternatives would. A related issue is the dispute about whether or not to regard the history of cinema as a progress toward more and more convincing representation, and whether the forms of today make up the last step in that evolution. In view of what has been said so far regarding the dangers of equating our realism and naturalness with the realistic and natural way of depicting the world, it will perhaps be suggested that conventionalism in the shape of relativism is the correct way out. But just like it is a mistake to assume that our way is the way, so there is nothing to support the claim that our way is only as good as the next, since this (as argued above) cannot be shown. Such an investigation would require criteria irrespective of either viewpoint, which of course are impossible to obtain. Hence, the only conclusion to be drawn (which is really a description) is that it certainly seems like ‘our’ naturalness is the ultimate naturalness, but that any such claim is likely to have to be revised in the future.

What, then, makes us inclined to describe a given narrational structure as being either natural or cultural, if, as argued, such a distinction cannot ultimately be upheld? Or more correctly: in what respect does it make sense to speak of the matter in such terms? Here it is instructive to look more closely at the context in which this conception figures. Regard the following claim by Carroll:

point-of-view editing serves the purposes of movie narration so well because to the degree to which it is keyed to biologically rooted and transculturally distributed features of perception, it guarantees fast pickup and a high degree of accessibility to mass untutored audiences, crucial desiderata for the persistence of any device in the economy of the movies. (Carroll, ”Toward a Theory of Point-of-View Editing”, p. 134 Emphasis added.)

In an important respect this is similar to Bordwell’s claim that ”some choices are weighted because human proclivities favor certain options”, and ”many artistic conventions are more appropriate to certain ends than others” (Bordwell, ”Convention, Construction, and Cinematic Vision”, p. 92. Emphasis added.) I.e. there is an underlying functionalist assumption of cinematic forms, according to which one form can be regarded as more or less natural than another, in achieving an end. In other words, to understand a certain form is to recognize the function it performs. This, apparently, is what enables us to say that ”the same goal could have been achieved by an alternative means” (Ibid., p. 93. Emphasis added). Returning to the scene with the flying slipper, we might imagine the same thing, i.e. the flight of the slipper, being communicated in a number of ways. Thus, a natural form would be one which performs this function well, without the viewer having to be familiar with a culturally specific set of symbols, just like telling somebody something in their own language ‘gets the job done’ more elegantly than would telling them the same thing in a foreign language or in morse code. Such an account would also suggest that we could substitute our talk of nature and culture with that of degrees of ‘familiarity’, in the manner that Bordwell proposes we dispense with the couplet and focus on some ”contingent universals of human life”.

But there are problems with this conception of movie narration. For one thing, it encourages the view that the form of film can be studied in isolation from the content, and that form performs the function of ‘transporting’ the content. This division corresponds to the view that art can be grasped in mentalistic terms, whereby the meaning of an artwork is defined as an entity originating in the mind of the artist. The statement that the same situation, e.g. of two people talking, can be depicted in a number of more or less efficient ways, amounting to the same result, is characteristic of this assumption. On reflection, however, such a view shows itself to be absurd. Form and content are aspects of art in much the same way that the mind and the body are aspects of a person: neither is conceivable in isolation from the other, i.e. both concepts rest on dualistic reductionism. Thus, a discussion might be depicted by shot/reverse shots, or altogether without cuts, and still communicate the same information. However, the scenes will not in all relevant respects be experienced as identical, since communication of information is not the sole function of art (in which case a film arguably would lose nothing if substituted, say, with a speech). That is, the chosen form is relevant to the experience (and appreciation) of art in a way that it is not to systems of communication (such as the morse code).

It furthermore seems likely that such formalistic conceptions of film are related to a certain theory of perception, according to which it makes sense speak of words as sounds that we hear and then interpret, or sights as ‘sense data’ being likewise interpreted. According to this theory, then, we see or hear things because we have - ‘biologically as well as culturally rooted’ - abilities to interpret the content of the perceived form, as it were. It is peculiar, however, that we should think of perception as an act of interpretation, resting on an ability, since these are words we usually have reason to use when a lack of them is conceivable (to say that somebody is able to do something, for instance, usually entails that they have a talent that not everybody has, or that this was not always the case, etc.) Normally, that is, e.g. watching Dreyer’s La passion de Jeanne d’Arc, we would not say that we ‘first’ receive a bundle of light rays, which we ‘then’ interpret as an image of a woman’s face, which we ‘then’ interpret as sorrowful. Rather, we would simply say that we - as long as we are not blind - see a woman’s face, which we perceive (simultaneously interpret) as sorrowful. It makes sense to speak of interpreting in this latter connection because it would not be surpising to hear that other people saw the face as scared, or spiteful or what not - only if we learned that many others did not see a woman’s face but, say, as a transsexual’s would we have reason to say that we saw (interpreted) the face as that of a woman. Similarly, it would be awkward to make a distinction between, on the one hand, ‘sensing’, and on the other ‘interpreting’ a given filmic device (as representing, for example, a character’s point-of-view), except in cases when we attend to other possible perceptions of the same instance. In this light the idea that the perception of the form - e.g. of the flight of the slipper - is distinct from the perception of the content - the flight ‘itself’ - shows itself to be an illusion.

Conclusion

What I hope to have accomplished by these somewhat scattered remarks, I want to emphasize, is not some kind of demonstration of the ‘difficulties’ involved in trying to provide an answer to our original question, but the realization that the question itself is ill-conceived. Already in posing the question, whether some forms are more natural than others, we are making certain reductive assumptions concerning the function and perception of film, e.g. that we are able to isolate the form of film, and that perception rests on innate or acquired abilities, etc. Futhermore, what is being asked for are causes to explain the development of cinema history.

Film history, then, is sometimes viewed as an aspect of nature which can be studied in a way similar to other natural phenomena. Why did she catch a cold, we might ask, and get the answer that she did because she sat next to somebody with the flu in the cinema theater. Although this answer might still our curiosity, other causes (read: theories) might aslo have been given: she did because at the time she was exhausted (had lowered immune defences), because she did not wash her hands when she got home, etc. (a Freudian might be interested in a Freudian explanation, an anthropologist in an anthropologist, a priest in a theological one, and so forth.) Usually we are satisfied with just the one answer, since (normally) we are aware of the different circumstances which need to be present in order for something to happen, but most of them are trivial - asking why the glass fell to the floor, we are usually not interested in hearing about Newton’s law of gravity, although gravity is ‘one of the causes’. Similarly, regarding cinema history, there are indeed a number of more or less scientific causes to be offered (industrial, economic, sociological, political, Darwinian, Hegelian, astrological, etc, etc.), and we can either emphasize the importance of one or two, or of the plurality. However, there is a crucial difference in respect to film history, viz. that it is a cultural practice. In a sense, films are actions rather than objects, we understand them like we understand people in general. The girl catching the cold has no reason for doing so, but somebody making a film behaves, and can give reasons for this behaviour. Hence, to understand why somebody has caught a cold is different from understanding why in a certain culture movies of one kind or another were made.

Thus, problematically, the study of the ‘shaping powers’ of cinema shows itself to be interested in a history of causes for what we cannot understand without recourse to histories of reasons. At the outset of this essay I asked if we find certain forms of film more accessible because we are biologically predisposed to do so or because they belong to a cultural tradition with which we are familiar and why exactly it is that a typical Hollywood movie is the most popular form of cinema. In conclusion, then, these are questions which cannot be answered since once investigated they cease to make sense.