Colin Crisp. 'The Rediscovery of Editing in French Cinema, 1930-1945'. In T. O'Regan & B. Shoesmith eds. History on/and/in Film. Perth: History & Film Association of Australia, 1987. 57-67.


Colin Crisp

The material presented in this paper originates in a large-scale study of those technical practices which contribute to the structuring of the filmic narrative. Editing practices are clearly one of the most important of these technical practices, and a discussion of them has been central to several of the theories which deal with filmic style and structure - such as montage theory, realist theory, and Metz's syntagmatic analysis.

Metz produced a table of grammatical figures[1] which he claimed were an essential element in the intelligibility of filmic texts, yet anyone who has attempted to apply that table of figures to a particular film will be aware of its inadequacies and of the inconsistencies that result. It rapidly becomes apparent that these inconsistencies are not just incidental, due simply to an as yet imperfect classification system, but are on the contrary fundamental.

The editing practices which structure a filmic narrative owe their existence not to some linguistic necessity such as Metz posited, but rather to contingent historical pressures of an economic, technological, institutional and ideological nature. These structuring practices occur at all stages of a film's production - at the point of script-writing, of filming, of editing, and of music composition - and though not finally determined by any necessity, are nevertheless far from random or autonomous since a range of consistent historical constraints weigh on them all. The available technology may only permit certain possibilities and exclude others; economic factors may be decisive in determining what sort of technology is available, and to whom; the individuals involved will have a limited awareness of what is possible, and this in turn will have resulted from their formal or informal training. Again, they will be differently informed of, and trained in, the structuring practices in related media, such as the theatre, opera, symphonic music, variety programs. Moreover, economic pressures at different times and in different parts of the cinematic institution push filmmakers either towards conformity to a recognized norm, for a guaranteed return on investment, or towards innovation - say, to titillate the aesthetic sensibilities of the art-film market. Finally, the conditions in which films have come to be consumed - themselves contingent on economic and ideological factors - condition the way filmmakers have come to structure stories, the sort of information they have become accustomed to giving the viewer, the degree of clarity and orientation the viewer has come to expect.

The technical practices that result from such constraints, and which interact to structure the narrative chain of the filmic text, can therefore best be studied under a series of headings such as "institutional conditions", "economic constraints", "the individuals and their training", and "available technology". What emerges from such a study will not be any neat and precisely delimited table of grammatical figures such as Metz was working towards, but a complex of possibilities, in theory infinite, but in practice limited by the conditions of production. To explain how any one film is structured, and how it came to be so structured, would require a potentially endless investigation; but to recognise certain regular structuring practices that exist within a series of films produced in like socio-economic conditions, by filmmakers with similar class backgrounds and similar training, and using identical technology, is by no means so impossible a task.

Consistencies are to be expected. Equally interesting will be any marked inconsistencies and any systematic changes over a defined historical period, since these too ought to be explicable in the terms outlined above.

The following account summarises a series of investigations intended to identify such regularities and exceptions within the editing practices of the French cinema of the thirties and forties, as part of a larger project which will ultimately totalise all the technical practices contributing to the structuring of cinematic narrative.

The editing practices at work in any film are open to a number of quantitative analyses. The most basic of these is to note the number of shots in each film. Twenty-three target films were selected for analysis, chosen in such a way as to allow the individual influence of directors, scriptwriters and editors to be identified (see figure 1 below), and the number of shots in those films was recorded. Figure 2 details the results of this analysis. In itself, this information is of little interest, since the films are of different lengths. Two procedures for rendering it useful were undertaken. Firstly, the figures were standardised to the norm of a 100-minute film. If, for instance, the film was shorter than that, the figure was increased proportionately, and vice versa. Secondly, the average shot length (ASL) was calculated by dividing the number of shots into the duration of the image-track of the film. The results of these calculations are listed in columns 5 and 6 of figure 2.

What we can observe from them is first of all a general tendency for the editing rate to increase steadily between 1932 and 1946. In fact, it increases by 50%, from a slow 400 shots per 100 minutes (ASL of 15 seconds) to a figure of 600 shots per 100 minutes (ASL of 10 seconds) in 1946. With one exception this steady increase overshadows every factor that might be attributed to the presence of an individual filmmaker, making "creative" choices according to private aesthetic sensibilities. Whether the film is signed Carne, Gremillon[2] or Duvivier, whether edited by Yvonne Martin, Marthe Gottier or Henri Rust, whether scripted by Charles Spaak, Henri Jeanson or Jacques Prevert, the result will be an objective and relatively predictable increase in editing intensity as the years pass.

The extreme point in this progression was to be Jacques Becker's Antoine and Antoinette (1946), credited with 1250 shots and immediately notorious as an image-track which had been fragmented to an unacceptable degree.[3]

If this regularity in the data needs to be accounted for, so does the one marked exception to it: the films directed by Jean Renoir. With remarkable consistency, four of these films record an ASL of 18 seconds, well above that of any other film analysed, and the fifth of them (Le crime de Monsieur Lange) holds an easy record of 24 seconds. This begins to provide a tangible basis for auteurist accounts of Renoir's filmmaking, since it suggests that whoever he worked with, under whatever conditions, and despite a recognisable set of normal practices observed by all other filmmakers of the time, Renoir structured films in a way that was peculiar to him. Even the odd instance of Partie de campagne (ASL: 11.96 seconds) which seems to invalidate this conclusion, in fact tends to confirm it. The film was, after all, constructed from fragments of a projected feature film shot in 1936, but only edited a decade later by Marguerite Renoir, at a time when Renoir himself was still absent from France. What we now have, therefore, is at once a film edited long after it was shot, when editing practices (as we have begun to see) had changed considerably, but also the sole film edited by Mme Renoir in Jean Renoir's absence. Figure 2 records one consequence of this: it is edited far more intensely than his other films, and perhaps more like those which Mme Renoir was editing for Becker at the time. We can hypothesise that it would have looked very different had it been edited when filmed, and with Renoir present.

By his exceptional nature, Renoir emphasises the regularities in editing practices that were elsewhere being observed; but these are set off equally clearly by the comparative data accumulated by Barry Salt in relation to American editing practices.[4] The distinctive nature of the two national cinemas can be seen in the fact that although American editing rates were also in the vicinity of 15 seconds soon after the introduction of sound, they rapidly returned to an intensity approximating that of the last years of silent cinema, where they remained until the 1940s. Any explanation for the regularity, and the regular evolution of French editing rates must be of a sort to account for this discrepancy in national practices between the two cinemas.

Although the ASL thus proves unexpectedly productive in isolating certain characteristic features of normal French editing practices, any further progress is dependent on identifying quantifiable parameters of a more detailed kind. An analysis of the distribution of brief and long takes proved useful in achieving this aim. Intervals of 5 seconds were arbitrarily selected, and the number of shots within the target films which fell into each 5 seconds duration was calculated. These figures were recorded as a percentage (figure 5) and then standardised for a 100 minute film (figure 6). Thus, in figure 3, we can see that 30% of the shots in Le grand jeu are of 0-5 seconds in duration, 21.5% are of 6-10 seconds, and so on. From figure 4 we can observe that the editing practices in Le grand jeu, when standardised to a 100 minute film, will generate 113 shots of 0-5 seconds duration, 81.1 of 6-10 seconds, and so on.

The two tables condense a large amount of interesting information in a useful comparative format. For instance, they give the lie to Barry Salt's assertion that, on the basis of the ASL a normal distribution of shot lengths will allow one to calculate the number of shots of a given length in each film. In fact, very few films exhibit anything like a regular distribution of shot lengths, such as would generate a smooth curve. In Le crime de Monsieur Lange, for instance, a marked predelection is apparent for shots of 20-35 seconds duration, and 40-50 seconds duration, but a disaffection for shots of 15-20 and 35-40. In Nous les gosses shots of 0-5 seconds are predominant, though there is a relative favouring of shots of 25-35 seconds in duration; whereas shots of 10-25 seconds in duration are almost non-existent. Such irregularities in the distribution of shot lengths would clearly constitute material for a more detailed stylistic analysis.

One consequence of this is that any two films with the same ASL (such as Entree des artistes and L'etrange Monsieur Victor) may diverge considerably in their use of shots of different lengths. L'etrange Monsieur Victor contains a far higher proportion of shots 0-5 seconds long, but also a higher proportion of shots 30 seconds and above in length. Entree des artistes achieves the same average shot length by its concentration on shots of intermediate length. This kind of analysis can therefore help us to identify and quantify the stylistic features which characterise the editing practices of a specific film. We thus are beginning to develop means to identify both overall regularities within the cinema as a whole, and the specific practices which characterise a particular film within that cinema.

For our present purposes, one main observation is unavoidable: the overriding cause of the increase in editing intensity is a marked increase in the use of very brief takes, of a duration of less than 5 seconds. Where in the early thirties it was regular practice for only 30% of a film to be less than 5 seconds long, by the war's end, the norm was nearer 50%. In a given 100 minute film, brief takes of less than 5 seconds were being used about three times as often in 1946 as they had been in 1932-36. There is a corresponding move away from very long takes of a duration of 45 seconds or more, but since these were few at any time, the perceived effect is less.

Basically, this shift could be interpreted two ways: either as a general abbreviation of shots across the whole range, as stories came to be edited more intensely, or alternatively as the fragmentation of certain specific longer takes into a multiplicity of briefer takes. The former would imply no overall change in editing practices though the intensification itself might need to be explained, the latter would imply a radical alteration of editing practices.

The latter is the case, as a further analysis quickly proves. The increase in brief takes is not dispersed equally or randomly along the narrative chain, but occurs at local sites, where "clusters" of such brief takes occur more and more frequently as the years pass. Not only do such clusters occur more frequently, but the number of shots involved in the clusters increases. Such series of clustered takes, from occupying 2-5% of the image-track in the early thirties, came to occupy 7-10% in the late thirties and 15-20% of it in the mid forties. Only the films directed by Renoir consistently resist this trend, and this is the principal reason why his films retain a high ASL.

These sets of relatively objective statistical data provide a description of editing practices over this period which clearly called for an explanation. The first step in developing that explanation was to correlate the increasingly frequent clusters of brief takes with the diegesis. What was happening at such moments on the image-track? What sorts of events were so treated?

Of the three types of event commonly structured into series of brief takes, one type - the descriptive series which serves to establish or change the locality and/or mood - was never common, and did not increase significantly over the period examined. It can, therefore, safely be ignored.

The two other types of events, which do in fact account for the change in editing habits, are firstly dynamic external events (such as hunts, pursuits, shoot-outs, assault and battery) and secondly dramatic but basically static dialogues where the interior drama of developing psychological relationships is particularly intense. We might designate the practices that structure these two types of event as "expressive editing" and "psychological editing", respectively. Both will be well-known to theorists of film.

In expressive editing, the variation between short and long takes corresponds to the level of dramatic action. Short takes are supposed to generate the desired emotional intensity in the spectator, long takes to relax that tension. As the action builds towards a climax, this system will dictate that takes of shorter and shorter length shall be used.[5] The system is, of course, somewhat crude in its signifying power. Indiscriminately, a series of long takes may be called on to signify despair or tranquility, boredom or a transcendant oneness with nature; while a series of brief shots, equally indiscriminately, may be called on to signify the exciting or the confusing, the angry or the terror-stricken, the triumphant, the agitated or the violent.

What is interesting to note is that this rather banal editing practice, which had been elaborated and theorised in the twenties, is relatively uncommon in the early thirties. It becomes about five times more common in the mid-forties; or rather, clusters of expressively edited brief shots occupy about five times as much of the image-track then as they had fifteen years before. In this case, we can talk of a gradual return to (or rediscovery of) editing practices which had once been common but which had largely disappeared with the advent of sound cinema.

The same cannot be said in the case of psychological editing, which had not been at all common in French silent cinema, but which evolved gradually in the course of the years under study. It is a phenomenon related to the construction of plausible psychological characters with a highly developed inner life. While too complex to be adequately described here, it involves the adoption by the camera of the point of view of the fictional characters, whose statements and reactions are closely observed in a series of alternating shots and reverse shots. Additionally, whole scenes tend to be fragmented into a series of "glances" - hypothetically, those of an interested onlooker observing the action.[6]

Clusters of brief shots which thus fragment the narrative chain - typically on the occasion of dramatic dialogue - come to occupy four times as much of the image-track in the mid-forties as they had in the early thirties. Becker's Antoine et Antoinette, already mentioned, marks the extreme point of this intensification. The decoupage already proposed 850 shots, at an ASL of 7.5 seconds, but in the filming and editing this increased to 1250, or an ASL of 4.5 seconds, largely through the proliferation of psychological shot/reverse shot editing of dramatic dialogues. It rapidly became notorious as an example of unacceptable fragmentation of the narrative line, an extreme point in virtuoso editing beyond which a realist cinema could not go without risking foregrounding technique at the expense of audience involvement. It is worth noting, that the films signed Renoir, consistently resist this tendency towards ever more intense psychological editing, and this is the principal cause of their lack of clusters, and of their higher ASL.

Certain of the developments described above can be accounted for fairly readily by the transition to sound cinema, whereas others require quite an extensive investigation into the conditions of production of the day if we are to understand them.

The more direct effects of the introduction of sound need only be summarised rather briefly. One universal factor whether in France or elsewhere, was a radically reduced editing rate in the early thirties. A factor influencing this reduction was the enormously more cumbrous nature of the new technology. Microphones and shielded cameras tended to tie action down to static locations. Expressive editing became more difficult. The dominance of dialogue in the early sound film had a similar effect. Ignoring the appeals of montage directors, filmmakers of the thirties tended to respect the integrity of the spoken phrase. The move from visual towards verbal drama therefore at least initially hampered any inclination towards brief takes. This tendency was reinforced by the fact that, in those early years, sound and image was recorded simultaneously on the same filmstock. It was thus impossible to break up the image-track if the dialogue was to remain intelligible. It was, therefore, not only by adding a hitherto lacking element of reality that sound pushed the cinema towards realism: it actively hindered constructivist attempts to interfere with the recorded profilmic material. Moreover, the men and women of the theatre called in to act as midwife to the sound cinema[7] thought in terms of acts and scenes rather than of shots, and saw creativity as centred in character development through the clash and riposte of verbal play rather than in the clash and riposte of montaged shots.

It is obvious, then, that a new technology, itself the product of economic and institutional factors will affect in a variety of ways the style and structure of subsequent texts; that different individuals with different competences are going to be called upon; that filmmakers trained in another medium will produce differently structured texts; that notions of realism, themselves ideological will evolve in parallel and interact with these. A complete explanation would use these categories - training, parallel media, technology, institutional structure, and ultimately socio-economic and ideological factors - to account for the observed changes in editing practices (and by no means only in those) over the period 1932-1946. What follows outlines some of the areas such an explanation would cover.

It was inevitable that the ability of the French cinema to return to the intense editing of the twenties or to implement new forms of intense editing would be less than its American counterpart. The available technology was relatively inadequate, and new technology was introduced at a much slower rate. In America, large-scale investment in new technology in the early thirties had tangential effects in the editing sphere, where efficient editing tables were rapidly introduced, and a system (called rubber numbering) for manipulating multiple segments of film while retaining synchronisation was rapidly introduced.[8] In France, rubber numbering was apparently still not in operation after the war, and the job of synchronising segments of film was done by hand, using Indian ink.[9] Moreover, contemporary accounts frequently make mention of the problem of uncomfortable and non-standard tables. In fact, except in the area of camera technology, the French cinema was notorious for its technological inefficiency. Sound technology had itself taken years longer to win out than in America, colour was introduced only much later, arc lights replaced floods years after they were standard in America.

Partly the explanation lies in the relative insignificance of the studio as a market-force in France. The collapse of the two large production companies in 1934 meant that the industry lacked industrial giants of the sort that, in America, could plough money back into modernisation. Studios remained poorly equipped, and often sub-contracted work out to small workshops, with the consequence that specialist knowledge and competences became fragmented and isolated, and located in relatively under-capitalised firms. In addition, the 50% unemployment in the industry militated against industrial action, to improve these conditions. It also meant that French editors tended to work in short bursts, but not regularly. The relative continuity of work in the American industry, makes a striking contrast. French editors, working mostly if not entirely with one director, might not work on more than one film a year, and each in a different studio, with different if equally inadequate equipment.

This situation is clearly related to the overall state of the industry and of the economy in this period.[10] A detailed documentation of the crisis would be superfluous here, but the following statistics help to understand how such a situation could arise. Sound reproduction patents were in foreign hands, and the rights to use it cost the industry dearly. This cost came in addition to the raw cost of the sound technology, and served to treble the cost of production over five years. Yet this was at a time when the very introduction of sound was creating a language barrier which limited the market. Equally, as the depression began to affect France, towards 1933, salaries dropped, entry prices dropped, and receipts fell from 933 million francs in 1932, through 878 in 1933, and 832 in 1934, to 758 in 1935. Two, three, or even four films per screening further reduced the return per film. Cinemas go bankrupt (86 in 1933) but, more dramatically, production companies go into liquidation - 50 in 1932, 58 in 1933, 88 in 1934. Parallel to the collapse of large and middle-size companies, there is a proliferation of small production companies, forming and dissolving rapidly, yet working with a steadily decreasing overall capital base. In 1934, alongside the 88 that fail, 202 emerge, but with a total capital of one-eighth of the 88. This pattern, reproduced year after year, rapidly produces an incoherent industry, made up of small ad hoc groups organised to produce one or two films. Such groups hire studios for periods as brief as possible; they have no ability to pay for, and no need to press for, new technology.

Hence the technological backwardness of the French industry which explains the difficult conditions under which French editors were working; and this in turn goes some way to explain the relative tardiness with which they introduce or reintroduce those intensive editing practices which were already common in the USA in the early thirties.

This process was reinforced by another series of pressures within the personnel of the industry. The task of editing was downgraded to the point where no particular skill was expected of those filling the role of editor and no training for the task was thought necessary.

This is in stark contrast with the twenties, when no director would have thought editing unimportant. For many it was even the creative moment of the filmmaking process, far too crucial to leave to others. The focus, that is, was then on the end point of the filmmaking process rather than on the beginning - no very elaborate pre-planning was felt necessary. It became necessary, however, with the need to script dialogue, and the consequent introduction of theatrically-oriented scenarists.[11] Insofar as these prescripted exchanged tended to pre-determing the editing, it was now in the early decoupage phase of the filmmaking process that creative moment was thought to occur and that creative control was thought to be crucial. Editing came to be thought of as a time-consuming but routine and unimaginative task that could be done by anyone. It was one of the least skill-oriented and theoretical of the specialist roles. Hence the lack of training: recruitment was informal, through word of mouth of accidental meetings, the novice sat and watched for a few days, then had a go in turn.[12] No one would have thought, during this period, of interviewing editors to find out what were their ideas about their job, since their job was not seen as involving ideas. This is unfortunate, since the result is that no account exists of the theory of psychological editing, contemporary with its development. Retrospectively we can record its evolution, note its historical relationship to the introduction of sound and to the dominance of dialogued exchanges, recognise its importance in the development of realism in the construction of credible characters, and in audience identification; but only in the 1940s with the growth of the film school did any first-hand theorisation of the techniques take place.[13]

In a sense, such technical skills as editors were required to develop themselves contributed to this eclipse, since psychological editing, in developing a motivation and logic for the syntagmatic linking of shots, rendered that linkage "invisible". Far from striving to overtly structure the narrative chain, the editor was called on to "de-structure" it - to render the viewer's movement along it fluent, unhindered and unconscious. Hence the recurrent complaint of the editors throughout the 1940s and even later that they were doomed never to be appreciated, since the better their work, the less it would be noticed.

But it is undeniable that a contributing factor to this invisibility is the gender differentiation which early on marked the editing task as appropriate for females rather than males.[14] This differentiation is peculiar to the French cinema. In the US and in Italy it was normal for editors to be male.[15] In Britain it was considered so physically demanding as to be inappropriate for females.[16] In France, however, it was early assimilated to the role of seamstress, with the little woman snipping busily away in her backroom, tidying up after the men. "The editing table is no more complicated, after all, than a sewing machine."[17] Given that the directors were all male, a marital metaphor is frequently invoked, with its concomitant notions of film as child, of editorial "pregnancy" and of the moment of birth. The women editors were encouraged to see themselves as the obedient, intuitive and passive counterpart of the assertive, active and creative male. Understanding, flexibility, humility, economy and discretion were their allotted virtues, but not intellect or creativity. They could not be expected to understand, let alone demand more complex and costly technology. The bleak sweatshops lined with bare tables which dominate the rare illustrations of French editing rooms of the 1930s[18] were the inevitable result of, amongst other factors, this morale of female self-abnegnation. Only with the development of film schools, with their formal courses in editing and their manuals of technical procedures - only, that is, with the progressive professionalisation of the film industry, which itself was the long-term effect of the task specialisation produced by the move to sound, do we begin to hear demands for more efficient technology, better working conditions. And it is noticeable that in this newly professionalised industry both teachers and taught tend to be male. It is from these men that the subsequent theorisation of editing practices comes.

I would hope that two things would become clear as a result of the project outlined above. Firstly, and most immediately, the nature of the pressures which served to shape the particular editing practices of the first 15 years of the French sound cinema, and which, because of their occasional similarities with and frequent divergences from later French cinema and other national cinemas, serve to explain the peculiar characteristics of the film style and structure of that period.

But secondly, and more generally, the method by which any theorisation of film style and structure should proceed: not by reference to some a priori linguistic or grammatical table, but by isolating statistical regularities which must then in turn be explained by a detailed investigation into the historical conditions of production. And this investigation must take into account not only the individuals involved, their self-image and working conditions, but the institutional and socio-economic constraints which determine both the statistical norm noted when isolating regularities, and the degree of freedom available to individuals to vary from that statistical norm.


1. C. Metz "Problemes de denotation dans le film de fiction", in Essais sur la signification au cinema (Paris: Klinckseick, 1978).

2. An anomaly is evident in the ASL for Remorques. Briefly, the principal source of this anomaly is the storm sequences, which had to be edited to an unusual degree in order to conceal the gross inadequacy of the models used. Additionally, though filmed in 1939, the film was not edited till 1941.

3. Amengual, B., "Evolution actuelle du langage cinematographique", Image et Son, no. 5 (March), 1952.

4. Barry Salt, "Film Style and Technology in the Thirties", Film Quarterly, vol. 30, no. 1 (Fall, 1976). See also his recent book, Film Style and Technology: History and Analysis (London: Starwood, 1983).

5. For a discussion of expressive editing practices by French theorists, see the Bulletin no. 4 de l'IDHEC (September, 1946); M. Martin, Le langage cinematographique (Paris: Les editeurs francais reunis, 1977), p. 164 et seq; Dr. R. Bataille, Le savoir filmer (Lille et Paris: A. Taffin-Lefort, 1944); P. Mouchon, Cine-montage (Paris: Paul Montel, 1954); and Lo Duca, Technique du cinema (Paris: Presses Universitaires de France, 1948), p. 49, etc.

6. For early theorisations of these psychological editing practices in France, see the Bulletin no. 3 de l'IDHEC (July, 1946), and Bretoneiche's remarks in "Comment on fait un film: le rythme", Ecran francais, no. 94 (15 April, 1947). The fullest early account of them can be found in M. Martin, op cit, pp. 157-163.

7. Such as Marcel Pagnol, Sacha Guitry, and Marcel Achard.

8. B. Salt, op cit, pp. 28-29.

9. For details of this discussion see "Le montage", in D. Marion (ed.) Le cinema par ceux qui le font (Paris: Artheme Fayard, 1949), and the course on editing produced by Y. Noe for the Ecole de technique du cinema par correspondance. See also Le technicien du film, no. 58 (15 February, 1960).

10. For information concerning industrial conditions of production, this section draws on F. Courtade, Les maledictions du cinema francais (Paris: Alain Moreau, 1978); and on unpublished theses by J.Levy, Etude et presentation de la revue "La critique cinematographique" 1929-36, and E. Strebel, French Social Cinema in the 1930s: a cinematographic expression of popular front consciousness.

11. Moreover the escalating cost of production obliged producers to pre-plan extensively, in order to reduce studio-time to a minimum.

12. See the discussion by Suzanne de Troyes and Marguerite Renoir in Le technicien du film, no. 57 (15 January, 1960).

13. A film school had existed as early as 1923, with a section devoted to editing, but had reputedly had little effect on film production except in the area of camera-work.

14. Alongside the few men whose names appear - Jean Feyte, Rene La Hennaf, Leonide Azar, Henri Rust - there are dozens of women: Marguerite Renoir, Suzanne de Troyes, Louisette Hautecoeur, Yvonne Martin, Marthe Poncin, Madeleine Bonin, Marinette Cadix, Mme Huguet, Marguerite Beauge, Marthe Gottie, Madeleine Guy, Anne-Marie Cottret, Marity Cleris etc.

15. "Les femmes et la technique du cinema", La revue du cinema, no. 283 (April, 1974), Image et Son.

16. "Le montage", in La technique du film, ed. Stephen Watts, trans. Jean-Georges Auriol (Paris: Payot, 1939).

17. Appearing in Image et Son, no 283 (April, 1974), this phrase echoes similar remarks by Colpi, by Lo Duca, and by G-M. Coissac in Les coulisses du cinema (Paris: Les editions pittoresques, 1929).

18. The issue of La technique cinematographique July/August, 1934, devoted to an enumeration of the facilities available in French studios, fails in most cases to even mention an editing facility.

Year        Films                 Script/Dialogues      Director       Editor/Asst     

1932         Poil de Carotte          Duvivier          Duvivier      Marthe Ponein    

                                                                       (Jean Feyte)    

  1934        Le Grand Jeu         Spaak, Feyder     Feyder/ Carne       M. Simon      


 1934/5           Toni           Renoir, Einstein,       Renoir      M.Renoir (S. de   

                                      Le Vert                            Troyes)       

 1935/6   La Crime de M. Lange    Prevert, Renoir,       Renoir       M. Renoir (Mme   

                                     Castanier                           Huguet)       

  1936       La Belle Equipe      Spaak, Duvivier       Duvivier        M. Ponein      

 1936/7       Pepe Le Moko       Jeansen, Duvivier,     Duvivier        M. Beauge      


 1936/46      Une Partie de            Renoir            Renoir         M. Renoir      


  1937       Gueule D'Amour            Spaak           Gremillon        ---------      

  1937     La Grande Illusion      Spaak, Renoir        Renoir,        Mme Huguet      


  1938       La Bete Humaine           Renoir           Renoir,      M. Renoir (S. de  

                                                         Becker           Troyes       

  1938     L'Etrange M. Victor     Spaak, Achard,      Gremillon        ----------     


  1938        Hotel du Nord      Jeanson, Aurenche       Carne          M. Gottie      

  1938     Entree des Artistes    Jeanson, Cayatte    M. Allegret       Y. Martin      

  1939       Le Jour se Leve       Piot, Prevert         Carne         R. le Hennaf    

  1939       La Fin du Jour       Spaak, Duvivier       Duvivier        M. Ponein      

  1939       La Regle du Jeu       Renoir, Koch,         Renoir       M. Renoir (M.    

                                      Francois                           Huguet)       

 1939/41        Remorques        Prevert, Cayatte,     Gremillon,     Y. Martin (L.    

                                       Spaak             Daquin        Hautecoeur)     

  1941       Nous les Gosses    Modot, Hilero, Ayme      Daquin        S. de Troyes    

  1942      Les Visiteurs du      Prevert, Laroche       Carne           H. Rust       


 1942/3       Lumiere d'Ete       Prevert, Laroche     Gremillon      L. Hautecoeur    

 1943/4    Le Ciel est a Vous     Prevert, Laroche     Gremillon      L. Hautecoeur    

 1943/5      Les Enfants du           Prevert            Carne         H. Rust (M.     

                 Paradis                                                  Bonin)       

            Les Portes de la          Prevert            Carne         J. Feyte (M.    

                  Nuit                                                   Gottie)       

[More diagrams to come]