SEMIOTICS OF ART Prague School Contributions Edited by Ladislav Matějka and Irwin R. Titunik The MIT Press Cambridge, Massachusetts, and London, England '-»-'**»■■ -1.....-■"■*•"-&*- ... -^ .^^;..^^..^mi|,-- ,, ■■arfliaALiJK^AÉifaa^..,»>., ^ > ^uHyaA^JÉ^cwáÉ^ Structure, Sign, and Function Selected Essays by Jan Mukařovský translated and edited by John Burbank and Peter Steiner New Haven and London Yale University Press teý.^...^.,, ^■•■■> .1 «±^Um**áb^u,luáĚki»~***auáamái±* „—aAafe, . . <^.**ste--^<~**i******i>»>~ 13 Time in Film Film is an art with many facets. There are threads connecting it to literature (both the narrative and the lyric), drama, painting, and music. Film has specific formal devices in common with every one of these arts, and every one of them has influenced film in the course of its development. The strongest ties, however, link film to the narrative and the drama; this is patently clear from the number of filmed novels and plays. We can say even more: the episte-mological conditions provided by its material place film between the narrative and the drama so that it has some basic characteristics in common with each of these arts. All three arts are related by the fact that they are arts of plot, and their theme is a series-of facts connected by temporal succession and a causal bond (in the broadest sense of the word). This has its significance both for the practice of these arts and for their theory. In practice this close kinship facilitates the easy transposition of theme from one to another as well as the heightened possibility of mutual influence. At the beginning of its development film was under the influence of the narrative and the drama; now it is beginning to reciprocate by means of a reverse influence (for example, the influence of the filmic techniques of the shot and the panorama on the presentation of space in modern narrative prose). For theory the closeness of film to the narrative and the drama makes possible their comparison. To this closely bound trinity of arts we can appropriately apply the general methodological rule that the comparison of materials which have many common features is theoretically interesting because, on the one hand, latent differences come sharply to the fore against a background of many similarities and because, on the other hand, we can arrive at reliable general conclusions without the danger of precipitous generalizations. In this outline we wish to attempt a comparison of filmic time with "Čas ve filmu," an article for an unrealized anthology on film, written in the second half of the thirties; published in Studie z estetiky (Prague, 1966). 191 .|?í-fliffl MSÉH HUk ,^J%é*áSS&,. "■*"» ■"■"•* idfeE.^. ~**áa.~~j3,*áUĚk± m* IfrTi'fflHtfiiEII 192 STRUCTURE, SIGN, AND FUNCTION dramatic and narrative time, both in order to elucidate film itself through a comparison with arts which are theoretically better studied and in order to obtain through the help of film more precise characteristics of time in the plot arts in general than has been possible up to now. We have already said that the most basic common feature of the film, the narrative, and the drama is the plot character of their themes. The plot can be defined in the most elementary way as a scries o! facts connected by temporal succession; it is thus inevitably connected to time. Time is therefore an important structural component in all three of these arts, though each of them has different temporal possibilities and requirements. In the drama, for example, the possibility of the presentation of simultaneous plots or even the displacement of segments of temporal series (the performance of what happened earlier after what happened later) is very limited, whereas the exploitation of simultaneity and temporal shifts are normal in the narrative. In this respect, as we shall see, film stands between the drama and the narrative as regards temporal possibilities. If we wish to understand the differences among the temporal structures of the three contiguous arts, we must realize that there are two temporal levels in each of them: one provided by the plot sequence, the other by the time which the perceiving subject (the viewer, the reader) experiences. In the drama these two times elapse parallel to one another. When the curtain is up, the flow of time is the same on stage as in the audience (if we disregard subtle discrepancies which do not disturb the subjective impression of sameness; for example, activities whose course does not have significance for the action, such as writing a letter, are abbreviated on stage; the flow of the audience's real time can also be projected into a much larger scope if the parallelism of temporal proportions is preserved). The time of the perceiving subject and that of the plot thus elapse side by side in the drama; therefore the plot of a drama takes place in the viewer's present, even if the theme of the drama is temporally located in the past (a historical drama). Hence the feature of dramatic time that Zich designates as its transitori-ness, which consists in the fact that only that section of the plot immediately before our eyes appears to us as present, whereas what has preceded is at the given mom.ent already swallowed up by the past; the present then is in constant movement toward the future. Ü^MMto wĚmim .^^-uJbi^ia+Amteiiimvi**,. mm. *^tMWatAMl>MtoA TIME IN FILM 193 Let us now juxtapose the narrative to the drama. Here, of course, the plot is also presented as a temporal sequence. The relation of this plot time to the temporal flow which the perceiving subject (the reader) experiences is, however, quite different, or more precisely, there is no relation between them. Whereas the flow of plot time in the drama is connected to the elapsing of the viewer's time to the extent that even the duration of the drama is limited by the normal ability of the viewer's concentrated attention, it docs not matter how much time we spend in reading; we can read a novel continuously or intermittently, in a week or in two hours. The time in which the narrative plot takes place is completely detached from the real time in which the reader lives. In the narrative the perceiving subject's temporal localization is felt as an indefinite present without temporal flow reflecting itself against the background of the elapsing past in which the plot takes place. Through the separation of plot time from the reader's real time there is the possibility—theoretically infinite—for the" condensation of plot in the narrative. A plot covering many years, which would require an entire evening in a dramatic performance, even with great temporal omissions between single acts, can be summed up in one sentence in a narrative: "A certain rich man married a beautiful young girl who, however, soon died and left him a little daughter, Helen."1 If we now put film beside these two types of temporal structure, that is, the drama and the narrative, we see that here again it is a matter of a different exploitation of time. At first glance it might seem that film is temporally so close to the drama that their temporal structures are the same. A more painstaking examination, however, will show that filmic time also has many characteristics which distinguish film from the drama and bring it closer to the narrative. In particular, film has an ability to condense plot quite similar to narrative condensation. Here are a few examples. Consider a long journey by train, which has no significance for the plot since it elapses "without any event." The narrative writer would sum this up in one sentence. The film director shows us a railroad station before the departure of the train, the train going through the countryside, a person sitting in a compartment, and perhaps the arrival of the train at the place of destination; thus in a few meters of film and in a few short minutes he "depicts" 1. J. and K. Čapek, "Mezi dvěma polibky," Zářivé hlubiny (Prague, 1924), p. 46. f 194 STRUCTURE, SIGN, AND FUNCTION synecdochically the action of many hours or even days. An even more illustrative example is Šklovskij's film Zapiski iz mertvogo doma [Notes from the house of the dead] where the march of a column of convicts from Petersburg to Siberia is presented in the following way: We see the legs of the convicts and their guards tramping over the frozen snow, and at the same time we hear the song that they sing; the song continues, and the shots change; we catch sight of a winter landscape, then the procession itself, again close-ups of legs, and so on; suddenly we realize that the landscape through which the procession is passing is not wintry but springlike; summer and autumn landscapes flash by in the same manner; the song goes on uninterruptedly, and when it is finished, we see the convicts already at their destination. In this way a journey of many months was summed up in a few minutes. The departure of plot time from the viewer's real time is obvious in these cases. In the same way that a narrative writer could abbreviate the long period of the journey by omitting all the detailed events in a short span of several sentences, the scenarist condenses it into a few shots. Another characteristic which filmic time shares with narrative time is the possibility of transition from one temporal plane to another, that is, the possibility of the successive presentation of simultaneous actions, on the one hand, and the capability of temporal return, on the other. Here, however, the analogy of film with the narrative is not so unconditional as in the previous case. Jakobson has recently pointed out that simultaneous actions are applicable only to film with captions, in fact to that kind of film with a narrative component (the verbal presentation of action), since a caption of the type "And meanwhile," conjoining simultaneous actions, is a narrative device.2 Flashback also has more limited possibilities in film than in the narrative, though it is not as impossible as in the drama. As an example let us cite an excerpt from Delluc's screenplay Le Silence: 52—Pierre's drawn face; he is remembering. 53—Pierre seen from afar in the middle of the apartment. He is searching his memory. Slowly. But for us the shots follow one another very quickly. 2. Editors' note. Mukařovský is referring to Jakobson's article, "Úpadek filmu?" [The decline of film?], Listy pro umenia kritiku 1 (1933): 45-49; reprinted in Roman Jakob-son, Studies in Verbal Art (Ann Arbor, 1971), pp. 150-56. UUMÉÉlitriMÉiÉi ääü. -■*' ^^^^^^^^^^ '*MAtl*ltJ aaüCw^KA ^ .^.tfflifc. -^^■*-^- TIM E IN FILM 195 54—Aimée, in an evening gown, falls forward in the middle of the drawing room. 55—Smoke. 56—A revolver. 57-Pierre's hand (the same rings) is holding the revolver. 58—Aimée stretched out on the rug. 59 ľicrrc standing in front of her. He throws the revolver down. 60-Pierre is bending over and is going to lift Aimée. 61-The servants arc coming. Pierre instinctively steps back. 62—Pierre's face after the murder. 63—Pierre's face now remembering the scene. 64-A shot of Pierre in the past, at his desk. He is writing. Aimée sits on the arm of his easy chair and tenderly kisses him. A visitor enters. It is Jean, an elegant young man. Aimée leaves, annoyed. Jean follows her intently with his eyes. Pierre notices this and becomes worried. 65-A dinner. Suzie, next to Pierre, is speaking to him with as much emotion as circumstances permit. Jean, next to Aimée, is passionately courting her. Aiméc's embarrassment; she is compelled to remain courteous. Pierre watches them anxiously. 66-The same evening. A corner of the drawing room. Suzie is bothering Pierre (who is no longer thinking about his jealousy). But Pierre is either cautious or faithful. He gracefully slips away from her. 67-Anothcr corner of the drawing room. Jean is harassing Aimée with his amorous insinuations; she doesn't know how to get rid of him. 68-Pierre notices them and again becomes furious. Suzie comes up to him, all smiles, but he coolly spurns her. 69-Suzie's face. She is insulted, and her pride has been terribly wounded. 70-Pierre in his smoking room. One morning. He is opening his mail. 71-An anonymous letter: "If you do not want to be intentionally blind, you will defend your honor. Keep a close eye on your wife." 72-Pierre, nervous and grim. He goes out. He hides behind a door opening onto the street. «Hilu; mIMéhMmhééi ^ ^»OUsiL*******-^*. i^aaa*"^ ■'-iI-ftMlIíf-ŤIlll "^"afa&- 196 STRUCTURE, SIGN, AND FUNCTION 73—Jean, very elegant, dressed for a visit, in the street. He goes into Pierre's house. Pierre goes in after him. 74—Jean in the drawing room. Aimée comes in. She reproaches him, begs him to leave her in peace, etc. ... He laughs, doesn't want to know anything, cries out that he is in love, etc. . . . etc. 75—Pierre behind the door. 76—Jean grabs Aimée. She defends herself. He kisses her against her will. Smoke. Aimée falls. Jean flees. 77—Aimée stretched out on the rug. 78—Pierre standing in front of her, the revolver in his hand.3 Here we have an obvious flashback: a murder and only afterwards a depiction of how it took place. The flashback is presented here, of course, in a loose temporal sequence, that is, it is motivated by the free association of a recollecting individual. In the drama such a displacement of plot segments would necessarily be understood as a miracle (the resurrection of a dead person) or as a surrealistic destruction of the unity of theme but never as a return to the past. This is because dramatic time is strictly irreversible due to the close bond between plot time and the perceiving subject's time. In sound-track film as well, we could hardly imagine such a transition from a closer temporal plane to a more remote one, even though it is motivated by recollection, for sound (in the above case, a shot and the characters' conversations), added to the optical impression, would make the break between temporal planes impossible. It would not be very plausible, for example, for the person whom we see as dead to appear and even speak in the following scene. In the progression from silent film with captions to silent film without captions to sound-track film the possibility of a temporal shift thus decreases. Nevertheless, the possibility of such a shift is not totally suppressed even in sound-track film. For instance, a flashback motivated by a recollection can be presented in such a way that the recalled scene is rendered only acoustically (a reproduction of a past conversation which the viewer has already heard) while the recollecting person is shown on the screen. How can we explain these characteristics of temporal structure in film? Let us first note the relationship between the time of the perceiving subject and the time of the picture projected on the 3. L. Delluc, Drames de cinema (Paris, 1923), pp. 24-27. «Mri miUĚM *■*•-* -^^^ifc^. - ■^■■■te^t^ilMi^iaiirriTTTi TIME IN FILM 197 screen. It is obvious that the temporal flow which the viewer experiences is deautomatized in film similar to the way that it is in the drama: the "pictorial" time flows parallel to the viewer's time. This is the resemblance of film to the drama. It is precisely this similarity that explains why film stood so close to the drama in its beginnings and again upon the introduction of sound-track. We must, however, consider another question. Is what we sec in front of us on the screen actually the plot itself? Can we identify the time of the filmic picture with the time of the filmic plot? The above examples provide the answer. If a march of many months from Petersburg to Siberia can be presented in a film without interruption and without any obvious temporal jumps in a few minutes, it is apparent that the presupposed plot (which, of course, does not actually have to be performed continuously) elapses in a different time than the picture. Its temporal localization is also different. We are aware that the action itself belongs to the past, whereas what we see in front of us on the screen we interpret as an optical (in some cases, an optical-acoustic) message about this past action. Only this message takes place in our present. Thus filmic time is a more complex structure than narrative and dramatic time. In narrative time we must take into account only one temporal flow (the elapsing of the plot), and in dramatic time a dual flow (the plot sequence and the elapsing of the viewer's time, the two lines necessarily parallel); in film there is a triple temporal flow: the plot elapsing in the past, "pictorial" time flowing in the present, and the perceiving subject's time parallel to the preceding temporal flow. Film gains ample possibilities of temporal differentiation through this complex structure. The exploitation of the viewer's own experience of temporal flow provides film with a versimilitude similar to the versimilitude of dramatic plot (rendering it present); but at the same time the sequence of "pictorial" time inserted between plot and viewer prevents the automatic linking of the plot flow with the real time in which the viewer lives. This makes possible the free play of plot time in a way similar to that in the narrative. We have already cited examples. Now we shall add one more concerning the interruption of plot flow in film. It is well known that there are static clustered groups of motifs in the narrative as well as dynamically ordered motifs (those bound by temporal succession), that is, that the narrative has the possibility of temporally static description as esa Üi ■na mM «uéMé iüüi •attü lifílŕiiií-riÉÉinriÉrf-'friiiMÉiTii 198 STRUCTURE, SIGN, AND FUNCTION well as that of temporally progressive narration. Tynjanov has shown that descriptions cut off from the temporal sequence of the plot also occur in film. He attributes descriptive power to detail, citing a scene in which the robbers, who are leaving a burglarized house, are described. The description is achieved by means of details concerning their weapons, and so forth. At this moment time has stopped. Tynjanov extends this discovery to detail in general and declares that it is excluded from the temporal flow.4 We could, of course, cite examples of details strikingly incorporated into the temporal sequence. The fallacy of this generalization does not, however, mean that Tynjanov's observation about the case in point is unimportant. Here it is indeed a matter of the suspension of the temporal flow of filmic description made possible only by the fact that the flow of pictorial time mediates between the viewer's time and that of the plot. Plot time can stop because even at the moment of its suspension "pictorial" time flows parallel to the viewer's time (which here, in contrast to the case of the narrative, is deautomatized). There are other possibilities lor playing with time in film at the boundary between "pictorial" time, which corresponds in its course to the viewer's time, and plot time, which is separate. These arc slow-motion and fast-motion film, as well as "reversed" film. In fast- or slow-motion film the ratio of the speed of plot time to "pictorial" time is deformed. A much larger (or much smaller) segment of plot time than we are accustomed to is appropriated to a specific segment of pictorial time. In reversed film the plot sequence elapses regrcssively, whereas the flow of "pictorial" time bound to the viewer's real time is naturally felt as progressive. In conclusion let us return to the problem of time in plot arts in general in order to attempt a more precise solution than we could suggest at the beginning of this article. In analyzing film, we have detected three kinds of temporal sequences: the first created by the flow of the plot, tlie second by the movement of pictures (objectively we could say: by the movement of the film strip in the projector), and the third based on the deautomatization of the real time experienced by the viewer. However, traces of this triple temporal stratification can also be detected in the narrative 4. J. Tynjanov, "Ob osnovax kino" [On the principles of film], in Poetika kino [The poetics of film], ed. B. Ejxcnbaum (Moscow, 1927), p. 66. Ä^IŽI i ÉMftÉtfMÉÉ ■Háhfei arttft^anü»^ «trn***-****,-****-,. lili '\Ti\](tíÍĚĚkĚabůJ'milUMmA^' TIME IN FILM 199 and the drama. As far as drama is concerned, the existence of two extreme temporal streams—plot time and the perceiving subject's time—is not in doubt. As for the narrative, there is, to be sure, only one distinctive temporal flow, that of the plot, but, as we have already remarked, the viewer's time occurs here at least as a static present. In both cases the existence of two temporal strata is thus discernible. What is seemingly missing is the third stratum, which mediates between these two extremes in film; it is what we have called pictorial time with respect to the material of film. What, in fact, constitutes this time? It is the temporal extent of the very work of art as a sign, whereas the other two times are defined with respect to things which are outside of the work itself. Plot time is related to the flow of a "real" event which is the content (plot) of the work; the perceiving subject's time is, as wc have remarked, merely a projection of the viewer's (or reader's) real time into the temporal structure of the work. If, however, "pictorial" time, which we could perhaps designate more generally as "semiotic" time, corresponds to the temporal extent of the work, it is obvious that its preconditions are also present in the narrative and the drama whose works also unfold in time. If we now look at the narrative and the drama, we discover that even here the duration of the work itself is reflected in its temporal structure through the so-called tempo, a term meaning the rhythm of the narration in the individual parts of narrative prose and the overall pace of the stage work (determined by the director). In both cases the tempo appears to us much more as a quality than as a measurable temporal quantity; however, in film, where the temporal extent of the work is based on the mechanically regular motion of film machinery, a quantity also manifests itself in semiotic time, and this time comes to the fore as a distinct component of the temporal structure. If we thus accept as a necessary epistemological precondition three kinds of temporal strata in all plot arts, wc can say that film is the art where all three strata obtain equally, whereas in the narrative the stratum of plot time comes to the fore, and in the drama it is the stratum of the perceiving subject's time (while the stratum of plot time is passively bound to this). If we ask—and not only for the sake of symmetry—whether there is an art in which semiotic time prevails, we must turn to the lyric where we can see a complete suppression of the perceiving ÉMIHUIÍIMÉIMÍÍIÉí Jěm ,^;. .^.jsjaí^^tíb^^a^s^ s» • •» ^^riifiÉÉf«ntr m íinír t 200 STRUCTURE, SIGN, AND FUNCTION subject's time (the present without signs of temporal flow) and of plot time (motifs are not connected through temporal succession). Proof of the full significance of semiotic time in the lyric is the importance for it of rhythm, a phenomenon linked to semiotic time, which, with the help of rhythm, becomes a measurable quantity. .a*^ -,-,^a^. ^i^JiMy........yjL'i.^^imi^^AA^. Emm