Structural/ Materialist film attempts to be non-illusionist. The process of the film's making deals with devices that result in demystification or attempted demystification of the film process. But by 'deals with' I do not mean 'represents'. In other words, such films do not document various film procedures, which would place them in the same category as films which transparently document a narrative, a set of actions, etc. Documentation, through usage of the film medium as transparent, invisible, is exactly the same when the object being documented is some 'real event', some film procedure', some 'story', etc. An avant-garde film defined by its development towards increased materialism and materialist function does not represent, or document, anything. The film produces certain relations between segments, between what the camera is aimed at and the way that 'image' is presented. The dialectic of the film is established in that space of tension between materialist flatness, grain, light, movement, and the supposed reality that is represented. Consequently a continual attempt to destroy the illusion is necessary. In Structural/ Materialist film, the in/film (not in/frame) and film/viewer material relations, and the relations of the film's structure, are primary to any representational content. The structuring aspects and the attempt to decipher the structure and anticipate/ recorrect it, to clarify and analyse the production-process of the specific image at any specific moment, are the root concern of Structural/ Materialist film. The specific construct of each specific film is not the relevant point; one must beware not to let the construct, the shape, take the place of the 'story' in narrative film. Then one would merely be substituting one hierarchy for another within the same system, a formalism for what is traditionally called content. This is an absolutely crucial point.'
Through usage of specific filmic devices such as repetition within duration one is forced to attempt to decipher both the film's material and the film's construct, and to decipher the precise transformations that each co/incide/nce of cinematic techniques produces. The attempt is primary to any specific shape, otherwise the discovery of shape (fetishising shape or system) may become the theme, in fact, the narrative of the film. This is a crucial distinction for a (dialectically) materialist definition of structural film. That is why Structural/ Materialist film in fact demands an orientation of definition completely in opposition to the generally used vague notions concerning 'Structural Film'.
Each film is a record (not a representation, not a reproduction) of its own making. Production of relations (shot to shot, shot to image, grain to image, image dissolution to grain, etc.) is a basic function which is in direct opposition to reproduction of relations. Elsewhere in this essay I shall try to elucidate further this problematic of production versus reproduction. Suffice it to say here that it is the core of meaning which differentiates illusionist from anti-illusionist film. When one states that each film is a record of its own making, this refers to shooting, editing, printing stages, or separations of these, dealt with specifically. Such film mitigates against dominant (narrative) cinema. Thus viewing such a film is at once viewing a film and viewing the 'coming into presence' of the film, i.e. the system of consciousness that produces the work, that is produced by and in it.
There is this representational 'reality' one is aiming the camera at. This remains true even if for example the representational content is pared down to the filmstrip itself being pulled through the printer. In fact this isn't necessarily a paring down at all. The Structural/ Materialist film must minimise the content in its overpowering, imagistically seductive sense, in an attempt to get through this miasmic area of 'experience' and proceed with film as film. Devices such as loops or seeming loops, as well as a whole series of technical possibilities, can, carefully constructed to operate in the correct manner, serve to veer the point of contact with the film past internal content. The content thus serves as a function upon which, time and time again, a film-maker works to bring forth the filmic event. 
The usage of the word content so far has been within the common usage, i.e. representational content. In fact, the real content is the form, form become content. Form is meant as formal operation, not as composition. Also, form must be distinguished from style, otherwise it serves merely in its reactionary sense to mean formalism, such as: this formal usage (e.g. Welles) versus that (e.g. Sternberg).
Film as material
The assertion of film as material is, in fact, predicated upon representation, in as much as 'pure' empty acetate running through the projector gate without image (for example) merely sets off another level of abstract (or non-abstract) associations. Those associations, when instigated by such a device, are no more materialist or nonillusionist than any other associations. Thus the film event is by no means, through such a usage, necessarily demystified. 'Empty screen' is no less significatory than 'carefree happy smile'.  There are myriad possibilities for co/optation and integration of filmic procedures into the repertoire of meaning.
The mental activation of the viewer is necessary for the procedure of the film's existence. Each film is not only structural but also structuring. This is extremely important as each moment of film reality is not an atomistic, separate entity but rather a moment in a relativistic generative system in which one can't simply break down the experience into elements. The viewer is forming an equal and possibly more or less opposite 'film' in her/his head, constantly anticipating, correcting, re-correcting - constantly intervening in the arena of confrontation with the given reality, i.e. the isolated chosen area of each film's work, of each film's production.
In dominant cinema, a film sets up characters (however superficially deep their melodramas) and through identification and various reversals, climaxes, complications (usually in the same order) one aligns oneself unconsciously with one or more characters. These internal connections between viewer and viewed are based on systems of identification which demand primarily a passive audience, a passive viewer, one who is involved in the meaning that word has taken on within film-journalese, i.e. to be not involved, to get swept along through persuasive emotive devices employed by the film director. This system of cinematic functioning categorically rules out any dialectic. It is a cinematic functioning, it should be added, analogous on the part of the film director to that of the viewer, not to mention the producer, who is not a producer, who has no little investment in the staking out of the economics of such repression. What some of the more self-defined 'left-wing' directors would rationalise in terms of dialectic are merely cover-ups for identification, selling the same old wares, vi: Antonioni and the much less talented Bertolucci, Pasolini, Losey, not to mention committed right-wing directors. Thus, if a character is somewhat more complex, or if the acting is of a higher order, or if the lighting cameraman does most of the work, then the director rationalises the work which would seem to imply that he is as taken in by the phantasy as the viewer. Whether he is or not (there are few she's in such a position) is in fact irrelevant. The ideological position is the same.
There is a distinct difference between what can be termed the ambiguousness of an identification process  and a dialectic functioning. Ambiguousness posits each individual viewer (or reader, listener, etc.) as subject: the subject, that is, who forms the interpretation. One becomes posited, formed, constituted, in fact, as the subject of the self-expression and self-representation through the mediation of a repressive ideological structure. That ideological structure is in this case narrative cinema, part of which is the process of identification. Ambiguousness aligns itself as a concept (and therefore as a reality) with the concept of freedom and individualism. The two latter concepts are extremely rigidified in late capitalism. The individual also thus becomes posited as static, as essence, as ideal (or referring to the possibility of such). The individual becomes posited as unitary, 'free' view, centred in deep perspective space away from the screen, and invisibly solidified, ever-present. Our whole formation towards, and in, filmic enterprises, is dominated by such ideological strangleholds.
The commercial cinema could not do without the mechanism of identification.' It is the cinema of consumption, in which the viewer is of necessity not a producer,  of ideas, of knowledge. Capitalist consumption reifies not only the structures of the economic base but also the constructs of abstraction. Concepts, then, do not produce concepts; they become, instead, ensconced as static 'ideas' which function to maintain the ideological class war and its invisibility, the state apparatus in all its fields.
The mechanism of identification demands a passive audience, a passive mental posture in the face of a life unlived, a series of representations, a phantasy identified with for the sake of 90 minutes' illusion. And that 'phantasy' is often not even the (insipid) utopian romance of what should be' (Marcuse's justification for Goethe's poems) nor the so-called 'intervention' in bourgeois morality that at moments may be approached in de Sade, Lautréamont, Sacher-Masoch (never without intensely counterproductive repressions and paranoiac violence stimulating and appeasing the bourgeois' tastes and tolerances).
Identification is inseparable from the procedures of narrative, though not totally covered by it. The problematic centres on the question as to whether narrative is inherently authoritarian, manipulatory and mystificatory, or not. The fact that it requires identificatory procedures and a lack of distanciation to function, and the fact that its only possible functioning is at an illusionistic level, indicates that the problematic has a clear resolution. In that sense, it is more of a problem than a problematic. The ramifications of the crucial question are very limited. Narrative is an illusionistic procedure, manipulatory, mystificatory, repressive. The repression is that of space, the distance between the viewer and the object, a repression of real space in favour of the illusionist space. The repression is, equally importantly, of the in-film spaces, those perfectly constructed continuities. The repression is also that of time. The implied lengths of time suffer compressions formed by certain technical devices which operate in a codified manner, under specific laws, to repress (material) film time.
Narrative and deconstruction
A further point on narrative: while the deconstruction of narrative as an academic exercise is not of vital import, it would be in any case a useful function towards expropriating the ownership of the codes of narrativity. Which means that the meanings formed by certain filmic operations could be analysed and no more be the privileged possession of the owners of the means of production; in this case, the means of production of meaning in film. Thus deconstruction exercises, in their limited way, are not irrelevant as sociological insight into certain filmic operations. Deconstruction exercises, maintained filmically (i.e. on film, in film) are direct translations from the written into film, and are thus filmically reactionary, though illustrative of certain ideas about film. The re-translation back into language (words) would seem to negate the necessity of narrative-deconstruction being undertaken on, or in, film, rather than in writing. This has now dawned, perhaps, on the overzealous graduates who wish to make statements about certain usages of narrative.
Apart from work in deconstruction, there is also that filmwork which is interpreted as deconstruction, works which have as their basic project an overhauling (not a criticising and not a smashing) of narrative, such as the pseudo-narratives of Robbe-Grillet's appalling films, or Straub's post- (and sometimes pre-) Brechtian exercises in distanciation and reflection. (Even here the Brecht of the theatre is mistaken for the Brechtian theoriser.)  Other examples are Dreyer's purist set pieces of dramatics, straightforward identificatory narratives, the identification merely shifted from the psychological/emotional to the psychological/rationalistic. The identification into the narrative is through the thoughts, the ideas about the actions, the decisions, the ratio, instead of the melodramatic unthought motivations of characters propelled by unthought 'fear', 'desire', etc. as in most other films. A study is urgently needed on the theme of narrative versus non-narrative form and on the inadequacy of the mechanistic deconstruction approach which ends up illustrating rather than being, which ends up static, time denying, posited as exemplary rather than relative, contradictory, motored into filmic, durational transformation through dialectic procedures.
Two art movements had their special effects on the current avant garde, Structural/Materialist film, and on those structural films which are working in that direction. The art movements were: the aesthetics of Abstract- Expressionism (though not necessarily the imagist results) and Minimalism (to include such work as Stella's).  A major problem erupts here: that of making visible the procedure, presenting such as opposed to using it. Throughout this essay, virtually every problem centres on the opposition between usage and presentation, incorporating versus foregrounding, etc. There exists also the problem of the 'sensitive' artist, ever-present in the final object, which can be one end the means to which is an art which may record its own making. But the other end, and the division must be carefully analysed and researched with each case in question, is that of an art which is not an imagist creation, a decorative object (narrative or otherwise) separated from its means of production without a trace left. If the final work magically represses the procedures which in fact are there in the making, then that work is not a materialist work. This is a crucial point as to usage versus presentation. And in each work many factors are operating which produce either an over-determination of the usage (i.e. repression) of the procedures, or an over-determination of the presentedness of the procedures.
Jacques Derrida has clarified what in fact is at stake in a work, in the procedure of constituting a work. His definition of differance (with an 'a') is useful precisely because it clarifies an aspect of work which previously was latent but not brought to speech, not adequately theorised, and which therefore always fell back into the ideology of illusionism and unseen subject (the artist).
We shall designate by the term differance the movement by which language or any code, any system of reference in general, becomes historically constituted as a fabric of differences ... Differance is what makes the movement of signification possible only if each element that is said to be 'present', appearing on the stage of presence, is related to something other than itself but retains the mark of a past element and already lets itself be hollowed out by the mark of its relation to a future element. This trace relates no less to what is called the future than to what is called the past, and it constitutes what is called the present by this very relation to what it is not, to what it absolutely is not; that is, not even to a past or a future considered as a modified present... We ordinarily say that a sign is put in place of the thing itself, the present thing - 'thing' holding here for the sense as well as the referent. Signs represent the present in its absence; they take the place of the present. When we cannot take hold of or show the thing, let us say the present, the being present, when the present does not present itself, then we signify, we go through the detour of signs. (J. Derrida, in Speech and Phenomena, 'Differance')
The aesthetics of Abstract-Expressionism in fact could produce an imagist object which never separated itself from individualist psychological origins, whereas the 'same' aesthetic base could function in certain works as production itself presented, distanced. Such presentation of production functions in certain drawings of targets by Jasper Johns (for example), distancing the object as object, as created text, towards which the various marks added to each other, negating, erasing, produce further elaborations towards an as yet unfulfilled total surface.  (Total is used in the sense of at some point coming to a stop.) The essential locus is again the question of psychological orientation, that is, identification, whether into the 'fantastic' or the 'real' or the 'surreal', in opposition to stated notions of distancing. But it must be clarified that the distancing is not from some wholly elaborated fantastic, real or surreal, from which a distance is created. Rather, the text itself is elaborated and constituted in such a way that the whole work process of reading the marks necessitates a reading of differences and a dialecticisation of the material procedures which produce the marking one is confronted with. The subject of the work is not the invisible artist symbolically inferred through the work's presence, but rather the whole foregrounded fabric of the complex system of markings itself.
What Frank Stella may have verbalised correctly (see footnote 8) did not prevent his work from becoming exactly the Abstract-Expressionist problem, the whole conglomeration of feelings, associations, seductions, repre- sentations which an imagist work demands no matter how 'process'-oriented the production process itself was. Similarly the process of making a Welles or Fassbinder film is not in an adequate way the product. This is the root of the whole problem I am trying to get at. Some of Stella's early works could escape this Abstract-Expressionist route, just as many of Johns' and Giacometti's works fail to avoid or solve that problem. Process as general definition is in fact vacuous. This vacuous definition is nevertheless filled, ideologically rigidified, in such a way that few works escape through the gap left, and those works are a conjuncture (happenstance or not) of a whole range of incidents and factors, co/incide/nces which enable this escape from the co/opting 'process' definition (and concreteness). This 'escape' is not a displacement (which would therefore create a misunderstanding, or a theoretical gap, elsewhere) or a suppression, but an adequate solution of questions correctly posed in terms of materialist practice and theoretical embodiment.