For many reasons, Eric Rohmer’s Perceval le Gallois (1978) is one of the most remarkable films of all time. We can think, of course, of Rohmer’s translation of Chrétien de Troyes’ 12th-century verse into an idiosyncratic version of modern French, in which the grammar, though versified, resembles modern syntactical forms, but which frequently retains archaisms such as “moult” (“many”, “very”), “oncques” (“never”) and, more predictably, “pucelle” (“maid”), “sire” and “valet”. This even, on occasion, takes on a humorous tinge, such as the frequent use of the verb “baiser” in its original, chaster sense of “to kiss”, whereas it has come, in the present-day world, to exclusively mean “to fuck”.

More boldly, Rohmer retains the octosyllabic metre from Chrétien’s epic poem, and has the actors recite it in a unique diction which required a year’s worth of rehearsals to master. Verse is both spoken by the characters and sung by a chorus which intervenes into the action at regular intervals, accompanied by Guy Robert’s “music based on tunes from the 12th and 13th centuries”, as the end credits put it. Startlingly, Rohmer’s fidelity to the text is such that characters in the film will often narrate their own actions in the third person. Rohmer characterised his method in the following terms: “Here, no indirect speech at all, but a lot of direct speech, linked by a narration whose flavour can not be replaced by any image or cinematic montage. I will keep it such as it is, and it will be spoken not only by the narrators, but also by the protagonists – who will speak of themselves in the third person, and will employ, when necessary, this ‘he said’, of which all the cinema (much like the theatre) has to do is make use.” (1) Once again, comic effect is achieved, when, for instance, Fabrice Luchini as Perceval utters the words: “He dared not utter a word”.

The care given to diction is mirrored by the attention accorded to gesture in the film: the characters are frequently positioned in the scene to deliberately evoke mediaeval art, and adopt specific gestural poses, which Rohmer described in the following terms:

I was inspired by miniatures in which one sees people with raised hands, elbows often close to the body, body a bit inclined, head lowered. I showed these images to my actors and asked them to try and incorporate them while rehearsing. In the beginning it was artificial, but it became less so later on. (2)

In speaking of the film’s remarkable nature, we can also think of its narrative structure, in which Rohmer sticks rigidly to the original text, making no concession to the dual fact that Chrétien de Troyes’ poem was left unfinished – only 9000 lines were written out of a likely 12-15,000, and Rohmer ignores the “Four Continuations” later added to the text by other authors – and that it was written before most of the narrative conventions of the bourgeois novel had even been formulated. Most elements of “literary realism” such as we have come to know them are absent from the film: the characters lack psychological motivation and development; the narration takes on a more circular rather than linear form, and includes major digressions such as the episode with Gauvain. Indeed, Rohmer seemingly takes delight in ending the film with a 10-minute sequence recreating the Passion Play – with Fabrice Luchini also in the role of Jesus Christ. If spectators had trouble fully comprehending the antiquated French verse of earlier in the film, they are left stranded in this sequence, commented by a Gospel poem sung in unsubtitled Latin.

And let us not forget the never-ending procession of comely maidens who illuminate the film with their presence, including early turns for Arielle Dombasle as Blanchefleur, and Marie Rivière and Pascale Ogier in multiple minor roles.

Yes, Perceval le Gallois is truly one of the most remarkable films of all time: at once both starkly different to the type of film that Rohmer has become known for – with its studio filming, (relatively) large budget and period setting, it could hardly be further away from the quasi-documentary shooting conditions of the ‘Six Moral Tales’ or the ‘Comedies and Proverbs’ – and yet, at the same time, distinctly Rohmerian – both for the fact that it is inscribed within a lineage of period films by Rohmer which have escaped categorisation (Die Marquise von O… [The Marquise of O, 1976], L’anglaise et le duc [The Lady and the Duke, 2001], and, with the most obvious affinity, Les amours d’Astrée et de Céladon [Romance of Astrea and Celadon, 2007]), as well as the fact that it continues the thematic preoccupation with matters of love, morality and fidelity developed in all his films. And then, of course, there are the comely maidens…

If Rohmer can best be conceived of as a filmmaker who dialectises the classical and the modern, then there is no greater example of this synthesis than his Perceval. And yet there is one aspect to the film which, above all, demonstrates this dialectical synthesis, and which, for me, makes the work so remarkable. But in order to adequately appreciate the true wonder of what Rohmer has achieved with the film, you, my loyal reader, will have to bear with me, as I take you on a perilous journey through the murky recesses of post-68 French Althusserian film theory.

Oh, I can hear your scoffing already. Am I not aware, you splutter, of the gulf that exists between the idealist musing of the Bazino-Catholic Rohmer and the ideological theorising of the Althussero-Tel Quelo-Marxists of Cinéthique and the post-68 Cahiers du cinéma? Why, I only need look at the contretemps between Rohmer and the Cahiers writers in April 1970 (handily reprinted in this issue of Senses) to perceive just how at odds their respective visions of film, politics and the universe were. How could I possibly hope to demonstrate the majesty of Rohmer’s film using the theoretical perspectives of his adversaries? Surely this is an undertaking that both parties would vigorously repudiate, notwithstanding the fact that the film, upon its French release in March 1979, was enthusiastically supported by Cahiers – by this time a very different beast to what it was earlier in the decade. (3)

But no. In one regard there is a striking correlation between Rohmer’s aesthetic ambitions in making Perceval and the developments in film theory instigated by the Cahiers writers. And so, to demonstrate this affinity, I will take you back to an interview given to Cinéthique by the Tel Quel contributors Marcelin Pleynet and Jean Thibaudeau in mid-1969. In this interview Pleynet made the following influential comments – and I entreat you for the permission to quote extensively from this text:

Have you noticed how all the discourses that can be held on a film, or on the cinema (and large quantities of them have been held), all start off from the a priori non-signifying existence of an apparatus which produces images, images which can be used indifferently for this or that, on the right or on the left? Does it not seem to you that before interrogating themselves on their ‘militant function’, filmmakers ought to interrogate themselves on the ideology that is produced by the apparatus (the camera) which determines the cinema? The cinematic apparatus is a properly ideological apparatus, it’s an apparatus which diffuses bourgeois ideology, even before diffusing anything else. Even before a film is produced, the technical construction of the camera produces bourgeois ideology.


In terms of the ideological production of the film, as well as the material that the film uses, the problem remains the same, whether in 16 or 35, to wit: a camera producing a directly inherited perspective code, constructed on the model of the scientific perspective of the quattrocento. It would be necessary to speak at greater length about this… to show how the camera is minutely constructed to ‘rectify’ all perspectival anomalies, to reproduce in its authority the code of specular vision such as it is defined by Renaissance humanism. (4)

The interview as a whole, and the passage from which this quote is extracted in particular, served to galvanise the writers at Cahiers. In October 1969, their famous editorial (well, famous in film theory circles at least), “Cinéma/Idéologie/Critique”, heralded the onset of the journal’s Marxist phase, in which films were prized to the extent to which they contained “a critical de-construction of the system of representation”. (5) Then, in a period beginning in early 1971 and continuing for roughly a year, a flurry of articles appeared in the journal concerning themselves, at least in part, with the role of the Renaissance perspectival code – as first elucidated by the Italian theorist Leon Battista Alberti in his work On Painting in 1435 – in the ideological composition of cinematic representation. Three writers, in particular, each embarked on a series of pioneering articles. The Lacan-influenced Jean-Pierre Oudart began with “L’effet de réel” in issue #228, from March-April 1971, which was continued with his “Notes pour une théorie de la représentation” in the two following issues. (6) Pascal Bonitzer carried the torch in issue #229 (May 1971), with a series of articles under the banner “Réalité de la dénotation”, which set out the work of Godard/Gorin, Straub/Huillet and Marguerite Duras as virtually the only contemporary films to be supported, owing to their shattering into fragments of the cinema’s “imaginary cube”, which was based on the “figurative system begun by the Renaissance”. (7)

But the role of Renaissance perspective in the cinema’s aesthetic code is most thoroughly delineated in the series “Technique et idéologie”, started by Jean-Louis Comolli in issue #229 as a polemical response to Jean-Patrick Lebel’s book Cinéma et idéologie, and continuing for six installments until issue #241 (September-October 1972). (8) The series was broken off, uncompleted, with a tantalising “to be continued” closing the sixth installment. Happily, this promise was belatedly fulfilled last year with the reprinting of the entire series together with a contemporary update, “Cinéma contre spectacle”. (9)

In the original articles, Comolli specifically tackled Pleynet’s above-quoted comments in order to oppose Lebel’s view that the camera, rather than being an ideological instrument, retains a level of “scientificity”. Comolli agrees in essence with Pleynet that perspectiva artificialis – the imitation of three-dimensional space on a two-dimensional plane through the gradation in size of the objects represented and their positioning in an imaginary representational “cube” receding to a horizon point – “has acted as a repressive system”, but for the Cahiers writer this is more due to the fact that “the inscription of doubt and lack has been systematically covered over by the inscription of normality and the centrality of the eye.” (10) In fact, Comolli tempers Pleynet’s view that “the image produced by the camera can do nothing but confirm and redouble ‘the code of specular vision such as it is defined by Renaissance humanism’.” (11)

For Pleynet’s view leads the cinema to the practical impasse arrived at by both Cinéthique – which advocated a “mute” cinema deprived of its ideological functioning – and Guy Debord, whose film Society of the Spectacle (1967) called for “Not the negation of style, but the style of negation”. If the camera is by its very nature an ideological instrument, then nothing it could produce could avoid bolstering the dominant (bourgeois) ideology, save for a quickly stale self-reflexivity or aesthetic negationism.

While Comolli admits that the cinema is “thought out, fabricated and purchased” within ideology and as an ideological instrument, he agrees with Lebel “in refusing to brand it with a ‘natural ideological blemish’”. (12) In fact, Comolli offers a way out for cinematic practice, and once again I will have to provide you, my patient reader, with a prolonged citation:

Due to the single, centralising eye of the camera, the deep-focus image is still organised on the basis of an axis which is perpendicular to the surface of the screen, much like Alberti’s ‘central ray’ which at the same time we know to be the assignation to the spectator of a rigorously fixed point of view, the real centre of the spectacle. In this image the laws of the perspectival system (with its ‘normality’ and its censures, the logocentrism that it installs) are thus renewed, and one can even say that this is the only occurrence in the cinema (or photography) of such a renewal, since a ‘flat’ image, without depth of field, (that of a telephoto lens, for example), participates in a different representational encodement, produces a different type of space, film having the particularity (that photography evidently does not have, unless composed in collage) of being able to depict, through the use of montage, different figurative codes in succession, with varying degrees of proximity to the Quattrocento ‘model’. This possibility, always offered, more rarely exploited scientifically or systematically by filmmakers and camera operators, of varying the types of lenses – and therefore the figurative codes – leads us to relativise the remarks made by Marcelin Pleynet cited at the beginning of this study: only two or three types of lenses – those which best ‘imitate’ normal vision, and which are therefore certainly used most frequently, let’s say most ‘naturally’ – are ‘minutely constructed to “rectify” all perspectival anomalies’. (13)

Do not let yourself be intimidated by Comolli’s labyrinthine sentences, the lesson is simple: through a use of overly flat, two-dimensional images, or a systematic application of differing focal lengths, filmmakers have the opportunity both to subvert the dominant use of Renaissance perspective in the cinema, and to provide the viewer with alternative pictorial schemas. We can see this in certain of Godard’s films at the time. La chinoise (1967) and Le gai savoir (1969), for instance, both contain numerous shots of characters shown frontally against a flat background, in which the image’s depth of field is completely annihilated, and this practice led Brian Henderson, in an article with some unfortunate major flaws, to see Godard as heading “Towards a Non-Bourgeois Camera Style”. (14)

And so, after this long and (hopefully) fascinating detour, we return to Rohmer. Allow me to summarise my argument in advance: Perceval le Gallois is one of the most remarkable films of all time because it offers an alternative visual schema to the Renaissance perspectival code dominant in the cinema since the Lumières.

But Rohmer’s technique is very different to Godard’s flattening of the image, as the man himself acknowledges:

To flatten the image with a telephoto lens, or to put people against the wall without any space behind them, would be too simple, too naive, because photography could never imitate the paintings; I don’t want to imitate painting. … Contrary to what one might expect, I want to give great importance to the third dimension and had thought of using the wide-angle lens, which would correspond more to the point of view of people in the Middle Ages. It’s difficult to explain all this. Space will be presented through enormous sets with a sky as a background which should make everything even bigger.” (15)

So it is that Rohmer’s refusal of Albertian perspective comes not through a manipulation of the camera lens, but through his exceptional use of set design in the film. Indeed, it is the artificiality of Perceval’s studio setting which will immediately strike the viewer as the most unusual aspect of the film. As befits the picaresque nature of Chrétien’s tale, Perceval’s myriad adventures take place within a single sound stage, flanked on each side by gilded castles, between which Perceval, atop his steed, sempiternally trots. In between: a sprinkling of spherical trees, conspicuously made out of steel, and a backdrop consisting of cardboard cut-outs of mountains, islands and castles.

Rohmer was forthright about his aesthetic program with Perceval:

Look, I am not trying to show the Middle Ages as we would see it if we could go back in a time machine and photograph it; I am searching to rediscover the vision of the Mediaeval period as it saw itself. This, it seems to me, one can attempt to accomplish, while we will never know the Middle Ages as they really were, and besides, it would not interest me to try. This attempt to create a film according to the period’s own conceptions of itself – I find this exciting. (16)

And it is this desire which leads to Perceval’s most notable breakthrough: Rohmer bases the film on pre-Renaissance aesthetic codes and conventions, taken from mediaeval painting, tapestry, manuscript illuminations and stained-glass windows. This noticeably applies to the film’s colour scheme: “realistic” colours are relinquished and replaced with a profusion of gold and brilliant hues, while subtle shading and contours are eschewed in favour of bold outlines and contrasts.

But the most frankly disorienting aspect of the film is its renunciation of the use of perspectiva artificialis. By shooting in a completely artificial location, Rohmer has used his control over the spatial relationships in the film’s mise en scène to recreate those found in mediaeval art. As Naomi Wise writes:

The proportion of his castle is, amusingly, precisely that of mediaeval art, in which a man was often as tall as a house, as in the 14th-century drawings for the Terence des Ducs manuscript. Rohmer’s overtly artificial castles are necessitated by the mediaeval view of nature, mediaeval visual arts, and most of all, mediaeval sense of proportion. It would appear that the people of that era had little sense of priorities: all sense impressions appear to have been vivid, overwhelming, and equal, and a man in front of a castle was, of course, as tall as the castle. (17)

Sooner than the Terence des Ducs manuscript, shots such as this one in Perceval remind me of one of the most famous works of art of the early Middle Ages, the Bayeux Tapestry, made roughly a century before Chrétien’s poem.

Moreover, as François Géré notes, the dissonance of Rohmer’s use of space and proportion is only heightened by the fact that his aesthetic world is, unlike mediaeval works of art, populated by living, moving creatures:

Our habits of verisimilitude are all the more impeded by the movement, in this space, of very real people and horses whose volume and dimensions conform to the principles we are accustomed to, and which make the proportional incongruence of the elements of the set even more marked. Questions surge from our perceptual panic: what is this universe? What are its organisational principles? (18)

Perhaps the only comparable attempt in the history of the cinema to disrupt the spectator’s perceptual habits in such an all-encompassing manner is Das Cabinet des Doktor Caligari (The Cabinet of Dr. Caligari, Robert Wiene, 1919). Indeed, Linda Williams declares that “the film’s dazzlingly innovative narrative style and total re-thinking of filmic space are as radical for 1979 as Caligari was in 1919.” (19) But she later mitigates this analogy, writing: “Rohmer does not attempt the intentional disorientation of the mixture of human figure and trompe l’œil backdrop of a film like Caligari.” (20)

Distancing the film from Caligari would seem to match Rohmer’s own views, as is evident from the following answer given to an interviewer:

On the other hand, I don’t want either German expressionism or the mise-en-scène of Laurence Olivier’s Henry V, where he tried to reconstruct the world of miniatures with false perspectives. In Romanesque art there are no false perspectives, there are no perspectives at all. There is a scale for objects, but not for space. Space does not exist. So it would be false to do Caligari sets. We have to find something else. (21)

Nonetheless – and I do not wish to dispute Rohmer’s views of his own film too strenuously – it is hard to overlook the striking affinities between a shot like this in Perceval:

And a shot like this in Caligari:

So, weary reader, I have come to the end of my voyage through the innermost reaches of French Althusserian film theory, as it pertains to Rohmer’s Perceval le Gallois. As a parting note, I wish to take you back to that original quote by Marcelin Pleynet. You may well have noticed that I brazenly inserted an ellipsis into the text cited, which dissembled a portion of the interview that I now take all too much glee in revealing to you.

After Pleynet asserted that “Even before producing a film, the technical construction of the camera produces bourgeois ideology,” his Cinéthique interviewer, Gérard Leblanc, interjected with: “16mm can produce less of it than 35 or 70mm.” To this Pleynet responded with a brusque: “Absolutely not, whether in 16 or 35mm, a Rohmer film remains a reactionary film, as much in terms of the ideological production of the film as the material that the film uses.” (22)

What profound historical irony, then, that the filmmaker fleetingly derided by Pleynet as “reactionary” was to make a film which, more than any other, fulfilled his aesthetic program by offering a systematic, radical alternative to the “code of specular vision such as it is defined by Renaissance humanism”!


  1. Eric Rohmer, Le goût de la beauté (Paris: Cahiers du cinéma, 1984), p. 142 (own translation).
  2. Nadja Tesich-Savage, “Rehearsing the Middles Ages” (interview with Eric Rohmer), in: Film Comment Vol. 14 No. 5 (May 1978), p. 52.
  3. See Danièle Dubroux, “Le rêve pedagogique”, in: Cahiers du cinéma No. 299 (April 1979), pp. 42-43; François Géré, “Poor and Lonesome” in Idem., pp. 44-46; and a later interview with Rohmer: Pascal Bonitzer and Serge Daney, “Entretien avec Eric Rohmer” in: Cahiers du cinéma No. 323-324 (May 1981), pp. 28-39.
  4. Gérard Leblanc, “Économique, idéologique, formel…” (interview with Marcelin Pleynet and Jean Thibaudeau), in: Cinéthique No. 3 (1969), p. 10 (own translation).
  5. Jean-Louis Comolli and Jean Narboni, “Cinéma/Idéologie/Critique” in: Cahiers du cinéma No. 216 (October 1969), p. 13 (own translation).
  6. Oudart published the following articles under the series title “Cinéma et réalité”: “L’effet de réel”, in: Cahiers du cinéma, No. 228 (March-April 1971), pp. 19-26; “Notes pour une théorie de la représentation”, in: Cahiers du cinéma, No. 229 (May 1971), pp. 43-45; and “Notes pour une théorie de la représentation (suite)” in: Cahiers du cinéma, No. 230 (July 1971), pp. 43-45.
  7. Bonitzer’s four articles were: “Réalité de la dénotation” in: Cahiers du cinéma, No. 229 (May 1971), pp. 39-41; “Réalité de la dénotation (2): Le Gros Orteil”, in: Cahiers du cinéma, No. 232 (October 1971), pp. 14-23; “Fétischisme de la technique: la notion de plan”, in: Cahiers du cinéma No. 233 (November 1971), pp. 4-10; and “Hors champ (un espace en défaut)”, in: Cahiers du cinéma No. 234-235 (December 1971-January/February 1972).
  8. The Comolli articles are the following: Part I: “Caméra, perspective, profondeur de champ” in: Cahiers du cinéma No. 229 (May 1971), pp. 4-21; Part II: “Pour une histoire matérialiste du cinéma” in: Cahiers du cinéma No. 230 (July 1971), pp. 51-57; Part III: “Pour la première fois” in: Cahiers du cinéma No. 231 (August-September 1971), pp. 42-50; Part IV: “La profondeur du champ ‘primitive’” in: Cahiers du cinéma No. 233 (November 1971), pp. 39-45; Part V: “Effacement de la profondeur/avènement de la parole” in: Cahiers du cinéma No. 234-235 (December 1971-February 1972), pp. 94-100; Part VI: “Quelle parole?” in: Cahiers du cinéma No. 241 (September-October 1972), pp. 20-24. The first two installments were translated by Diana Matias in a commonly available, but deeply flawed, English version of the article, and so I have provided fresh translations of the passages used here.
  9. Jean-Louis Comolli, Cinéma contre spectacle (Paris: Éditions Verdier, 2008). The following page references refer to the republished version of the article, which nonetheless remains unchanged from the original.
  10. Cinéma contre spectacle, p. 149.
  11. Ibid., p. 135.
  12. Ibid., p. 154.
  13. Ibid., p. 156.
  14. See Brian Henderson, “Towards a Non-Bourgeois Camera Style” (1970) in: Bill Nichols (ed.). Movies and Methods: An Anthology. (Berkeley: University of California Press, 1976).
  15. “Rehearsing the Middles Ages”, p. 52.
  16. Ibid., pp. 51-52.
  17. Naomi Wise, “Perceval”, in: Film Quarterly, Vol. 33 No. 2 (winter 1979-80), pp. 51-52.
  18. “Poor and Lonesome”, p. 44 (own translation).
  19. Linda Williams, “Eric Rohmer and the Holy Grail”, Literature and Film Quarterly, Vol. 11 No. 2 (summer 1983), p. 72.
  20. Ibid., p. 78.
  21. “Rehearsing the Middle Ages”, p. 52.
  22. “Économique, idéologique, formel”, p. 10 (own translation).