RETURN

TO

H I S T O RY*

D

iscussions about the relations betwcen structuralism and his­ tory have been numerous, highly involved, and often confused - not only in France but in Europe, America, and perhaps Japan as well, I don't know. This has been the case for several reasons that are simple to enumerate. The first is that no one agrees with anyone else on what structural­ ism is. Second, in France the word "history" means two things: what historians talk about and what historians do in their practice. The third and most important reason is that many political themes or con­ cerns have run through this discussion about the relations between history and structuralism. It should be said, moreover, that I don't intend to dissociate today's discussion from the political context in which it is situated. On the contrary: in the first part I would like to lay out the general strategy, the battle plan of this debate between the structuralists and their adversaries concerning history. The first thing to note is that structuralism, at least in its initial form, was an undertaking that aimed to give historical investigations a more precise and rigorous method. Structuralism did not turn away from history, at least not in the beginning; it set out to construct a history, one that was more rigorous and systematic. I will simply take three examples. Franz Boas, an American, can be considered the founder of the structural method in ethnology.l Now, what was that *Rekishi heno kaiki," Paideia II (I February 1972), pp. 45 - 60. This text is based on a French transcription reviewed by Foucault. Robert Hurley's translation.

4 20

Aesthetics, Method, and Epistemology

method as he conceived of it? It was essentially a way of criticizing a certain form of ethnological history that was done in his era. Edward Tylor had provided the model for it.2 This history assumed that hu­ man societies all follow the same evolutionary curve, going from the simplest forms to the most complex. The evolution did not vary from one society to another except in the speed oftransformations. Further, the great social forms such as marriage rules or agricultural tech­ niques were seen basically as kinds of biological species, and their extension, their growth, their development, and their distribution were thought to obey the same laws and patterns as the growth and spread of biological species. In any case, the model that Tylor used to analyze the development and history of societies was the biological one. Tylor referred to Darwin, and more generally to evolutionism, in order to tell the story of societies. Boas's problem was to free ethnological method from that old bio­ logical model and to show how human societies, whether simple or complex, obeyed certain internal relations that defined them in their specific organization. That process within each society is what Boas called "social structure," and he thought analysis of the structure would enable him to do a history of human societies which was no longer biological but truly historical. So, for Boas, it was not at all a matter of discarding the historical point of view in favor of, say, an antihistorical or ahistorical point of view. I took the example of Boas, and in the same way I could have taken the example of linguistics, and phonology in particular. Before Nikolai Trubekskoi, historical phonetics traced the evolution of a phoneme or a sound across a language.3 It tended not to account for the transfor­ mation of an entire state of a language at a given moment. What Tru­ betskoi wanted to do with phonology was to convert it into the tool that would enable him to go from the individual history of a sound, as it were, to the much more general history of the phonetic system of an entire language. I could take a third example that I will recall briefly. It's the applica­ tion of structuralism to literature. When, a few years back, Roland Barthes defined what he called the "level of writing" as against the level of style or the level of language, what was he trying to accom­ plish?4 Well, that becomes clear when one looks at how literary his­ tory was studied in France from about 1950 to about 1955. During that period, either one did the individual, psychological, and perhaps psy-

Return to History

4 21

choanalytic history of the writer, or one did a general, overall history of an epoch, of a whole cultural ensemble, a collective consciousness, if you will. In the first case, one never encountered anything beyond the indi­ vidual and his personal problems; in the second, one only reached very general levels. What Barthes wanted to do by introducing the notion of writing [ecriture] was to reveal a certain specific level on the basis of which a history of literature as literature might be under­ taken, recognizing that it has a particular specificity, that it goes be­ yond individuals, who reside within its space- and further, that in the midst of all the other cultural productions, it is a perfectly specific element with its own laws of conditioning and transformation. By introducing this notion of writing, Barthes wished to establish a new possibility of literary history. So I do think we need to bear in mind that, in their initial projects, the different structuralist ventures (whether they were ethnological, linguistic, or literary, and the same could be said regarding mythology and the history of the sciences) at the outset were always attempts to fashion the instrument for a precise historical analysis. Now, one has to acknowledge that this undertaking did not fail-I don't mean to say that at all -but was not recognized for what it was; most of the adver­ saries of the structuralists agree on this point at least, that structural­ ism would have missed the very dimension of history and would be in effect antihistorical. This criticism comes from two different horizons. First of all, there is a theoretical critique whose inspiration is phenomenological or ex­ istentialist. It is argued that structuralism was obliged to abandon whatever good intentions it may have had, that in fact it gave an abso­ lute privilege to the study of simultaneous or synchronic relations over the study of developmental relations. When, for example, the phonologists study phonological laws, they study language states, without taking their temporal development into consideration. How can history be done if one fails to take time into account? But that is not all. How could it be said that structural analysis is historical, since it privileges not only the simultaneous over the successive but also the logical over the causal? For example, when Claude Levi-Strauss analyzes a myth, what he tries to determine is not where the myth comes from, why it came into being, how it was transmitted, why a

422

Aesthetics, Method, and Epistemology

particular population has recourse to the myth, or what led another population to transform it. He is content, at least in a first phase, with establishing logical relations between the different elements of the myth; temporal and causal relations can be established within the space of that logic. There is another objection, finally: Structuralism does not consider freedom or individual initiative as a factor. Against the linguists, Jean-Paul Sartre objects that language is never anything but the outcome, the crest, the crystallization of a basic, primordial human activity. If there was no speaking subject to continually take up language, inhabit it from within, shape it, deform it, utilize it, if there was not this element of human activity, if there was not speech at the very heart of the language system, how could language evolve? So, as soon as one leaves human practice aside, considering only structure and rules of constraint, it is obvious that one is again miss­ ing history. The objections that have been raised by phenomenologists or exis­ tentialists are generally adopted by a certain number of Marxists whom I shall call "summary Marxists," that is, Marxists whose theo­ retical reference is not Marxism itself but, in fact, contemporary bour­ geois ideologies. On the other hand, objections have come from a more serious Marxism, that is, from a truly revolutionary Marxism; these objections are based on the fact that the revolutionary move­ ments that have occurred, that are still occurring among students and intellectuals, owe next to nothing to the structuralist movement. There is perhaps a single exception to this rule; it's the case of Althus­ ser in France. Althusser is a Marxist who has applied certain methods that can be regarded as structuralist to the reading and analysis of Marx, and Althusser's analysis has been very important in the recent history of European Marxism.5 This importance is tied to fact that Althusser freed the traditional Marxist interpretation from all the hu­ manism, from all the Hegelianism, and from all the phenomenology that burdened it, and thus made possible once again a reading of Marx that was no longer an academic reading but a truly political one. But as important as these Althusserian readings were at the start, they were quickly outstripped by a revolutionary movement that, although developing among students and intellectuals, is, as you know, an es­ sentially antitheoretical movement. One might add that most of the

Return to History

revolutionary movements that have developed recently in the world have been closer to Rosa Luxemburg than to Lenin. They place more trust in the spontaneity of the masses than in theoretical analysis. It seems to me that until the nineteenth century the primary aim of historical analysis was to reconstruct the past of the great national ensembles by which industrial capitalist society was divided up or tied together. From the seventeenth and eighteenth centuries onward, industrial capitalist society established itself in Europe and the world according to the schema of the great nationalities. History had the function, within bourgeois ideology, of showing how these great na­ tional units, which capitalism needed, came from far back in time and had asserted and maintained their unity through various revolu­ tions. History was a discipline by means of which the bourgeoisie showed, first, that its reign was only the result, the product, the fruit, of a slow maturation, and that this reign was thus perfectly justified, since it came from the mists of time; next, the bourgeoisie showed that, since this reign came from the dawn of time, it was not possible to threaten it with a new revolution. The bourgeoisie both established its right to hold power and warded off the threats of a rising revolu­ tion, and history was indeed what Jules Michelet called the "resurrec­ tion of the past." History assigned itself the task of bringing the whole national past back to life. This calling and role of history now must be reconsidered if history is to be detached from the ideological system in which it originated and developed. It is to be understood, rather, as the analysis of the transformations societies are actually capable of. The two fundamental notions of history as it is practiced today are no longer time and the past but change and the event. I will cite two examples, the first borrowed from structuralist methods, the second from properly historical methods. The purpose of the first one is to show you how structuralism has given or, at any rate, tries to give a rigorous form to the analysis of changes; and the second aims to show how certain methods of the new history are attempts to give a new status and meaning to the old notion of event. As the first example, I shall take the analysis that Georges Dumezil did of the Roman legend of Horace.6 It is, I believe, the first structural analysis of an Indo-European legend. Dumezil found isomorphic ver­ sions of this well-known story in several countries, Ireland in particu-

Aesthetics, Method, and Epistemology

lar. There is an Irish narrative in which one sees a character, a hero named Cuchulain, and this Cuchulain is a child who has rece.�ved from the gods a magical power that gives him an extraordinary strength. One day when the kingdom in which he lived finds itself threatened, Cuchulain leaves on an expedition against the enemies. At the gates of the palace of the opposing leader, he meets a first adversary whom he slays. Then he continues to advance. He meets a second adversary and kills him, then a third, whom he also kills. After this triple victory he can go home, but the combat has brought him to such a pitch of excitement, or rather, the magical power he received from the gods has inflamed him to the point that he becomes red and glowing, so if he returned to his town he would be a danger to every­ one there. To quell this burning and seething force, his fellow coun­ trymen, on the way back, decide to send him a woman. But it so happens that this woman is the wife of his uncle. The incest laws prohibit such a sexual relation, so he cannot extinguish his ardor in this way, and they are obliged to plunge him into a bath of cold water- but he is so hot that he makes the bath water boil and they have to soak him in seven different baths before he cools down to a normal temperature and can return home without being a danger to others. Dumezil's analysis differs from the analyses of comparative my­ thologies that had been done before him. In the nineteenth century there was a whole school of comparative mythology in which one merely showed the resemblances between one myth and another, and in this way certain historians of religions had managed to find the same solar myth in almost all the world's religions. Dumezil, on the other hand - and this is what makes his analysis structural - compares these two narratives only in order to establish the differences between the first and the second. He identifies these differences in a very pre­ cise way. In the case of the Irish Cuchulain, the hero is a child; sec­ ond, he is charged with a magical power; finally, he is alone. Consider the Roman myth: the hero, Horace, is an adult, is old enough to bear arms, has no magical power-he is simply more clever than the oth­ ers, since he invents the ruse of pretending to flee and then returning, merely a slight distinction within the strategy, but he has no magical power. There is another set of differences in the case of the Irish leg­ end. The hero has such a strong magical power, and this magical

Return to History

power is so intensified in battle, that his return bears a danger to his own town. In the case of the Roman narrative, the hero returns as a victor, and among those he meets he sees someone who has betrayed her own country in her heart- his sister, who took the side of Rome's enemies. The danger was thus shifted from outside the city to the inside. It is no longer the hero who is the bearer of danger, it is some­ one different from him, though belonging to the same family. Finally, there is a third set of differences. In the Irish narrative only the magi­ cal bath in the seven tanks of cold water can calm the hero, whereas in the Roman narrative a juridical ritual, not a magical or religious one, is required, that is, a trial, then an appeals procedure, then an acquittal, before the hero can regain his place among his contempo­ raries. Dumezil's analysis- and this is the first of its features-is therefore the analysis not of a resemblance but of a difference and an interplay of differences. In addition, Dumezil's analysis is not content with drawing up a table of differences; it establishes the system of differ­ ences, with their hierarchies and their subordination. For example, he shows that in the Roman narrative, from the moment that the hero is no longer that young child endowed with a magical power, but a soldier like the others, it is clear that he can no longer be alone in the face of his adversaries, because a normal man confronting three nor­ mal adversaries would necessarily lose. Consequently, around the hero, Horace, the Roman narrative has added two partners, the two brothers who even things up for the Roman hero facing the three Curiattii. If the hero were charged with a magical power, it would be easy for him to defeat these three adversaries; but once he is a man like the others, a soldier like the others, it is necessary to frame him with two other soldiers, and his victory will be obtained only by a kind of tactical trick. The Roman narrative made the Irish hero's ex­ ploit a natural one. When the Romans introduced the difference that consists in putting an adult hero in the place of a child hero, when they presented a normal hero, and no longer a character charged with magical power, there had to be three and instead of one against the three. So one has not just the table of differences, but the connec­ tion of differences with one another. Finally, Dumezil's structuralist analysis consists in showing what the conditions of such a transfor­ mation are.

Aesthetics, Method, and Epistemology

Through the Irish narrative one makes out the profile of a society in which military organization rests essentially on individuals who have received their power and their strength from their birth: their military strength is tied to a certain magical and religious power. In contrast, what appears in the Roman narrative is a society in which military power is a collective power. There are three Horace heroes, and these three Horace heroes are only functionaries so to speak, since they have been delegated by those in power, whereas the Irish hero had himself taken the initiative of his expedition. The combat unfolds within a common strategy: in other words, the Roman trans­ formation of the old Indo-European myth is the result of the transfor­ mation of a society essentially made up, at least as concerns its military stratum, of aristocratic individualities, into a society whose military organization is collective and to a certain extent democratic. And while structural analysis may not solve the problems of the his­ tory of Rome, you can see how it ties in directly to the actual history of the Roman world. Dumezil shows that it would be pointless to look in the narrative of the Horaces and the Curiattii for something like the transposition of a real event that would have occurred in the first years of Roman history; but, by showing the schema of transforma­ tion of the Irish legend into a Roman narrative, he also reveals the principle of historical transformation of the old Roman society into a state-controlled society. You see that a structural analysis like that of Dumezil can be linked to a historical analysis. On the basis of this example, it could be said that an analysis is structural when it studies a transformable system and the conditions under which its transfor­ mations are carried out. Taking a very different example, I would now like to show how certain methods currently employed by historians make it possible to give a new meaning to the notion of event. People are in the habit of saying that contemporary history concerns itself less and less with events and more and more with certain broad, general phenomena that would extend through time, as it were, and would remain immo­ bile through time. But for several decades historians have been prac­ ticing a so-called serial history, in which events and sets of events constitute the central theme. Serial history does not focus on general objects that have been con­ stituted beforehand, such as feudalism or industrial development; se­ rial history defines its obj ect on the basis of an ensemble of

Return to History

documents at its disposal. Thus about ten years ago a study was done of the commercial archive of the port of Seville during the sixteenth century: everything having to do with the entry and exit of ships, their number, their cargoes, the selling price of their goods, their national­ ity, the places they came from, the places they were sailing to. It was all these data, but only these data, that constituted the object of study. In other words, the object of history is no longer given by a kind of prior categorization into periods, epochs, nations, continents, forms of culture . . . One no longer studies Spain and America during the Re­ naissance; one studies- and that is the sole object-all the documents relating to the life of the port of Seville at such-and-such a date. The consequence- and this is the second trait of this serial history- is that this history doesn't use these documents to immediately decipher the economic development of Spain; the object of historical research is to establish, on the basis of these documents, a certain number of rela­ tions. In this way it was possible to establish - I'm referring again to Huguette and Pierre Chaunu's study on Seville-year-by-year statisti­ cal estimates of the entries and exits of ships, classifications according to countries, and distributions in terms of goods? Based on the rela­ tions they able to establish, the Chaunus were also able to plot the curves of development, the fluctuations, the increases, the stoppages, the decreases; they could describe cycles and establish relations, fi­ nally, between this group of documents concerning the port of Seville and other documents of the same type concerning the ports of South America, the Antilles, England, and the Mediterranean ports. The his­ torian, you see, does not interpret the document in order to reach behind it and grasp a kind of hidden social or spiritual reality. His work consists in manipulating and processing a series of homoge­ neous documents relating to a particular object and a particular ep­ och, and the internal or external relations of this corpus of documents are what constitute the outcome of the historian's work. Using this method- and this is the third feature of serial history-the historian can reveal events that would not have appeared in any other way. In traditional history it was thought that events were what was known, what was visible, what was directly or indirectly identifiable, and the work of the historian was to search for their cause or their meaning. The cause or meaning was essentially hidden. The event, on the other hand, was essentially visible, even if one sometimes lacked the docu­ ments to establish it with certainty. Serial history makes it possible to

Aesthetics, Method, and Epistemology bring out different layers of events as it were, some being visible, even immediately knowable by the contemporaries, and then, beneath these events that form the froth of history, so to speak, there are other events that are invisible, imperceptible for the contemporaries, and are of completely different form. Let's take up the example of the Chaunus' work again. In a sense, the entry or exit of a ship from the port of Seville is an event with which the contemporaries inhabiting Seville were perfectly familiar, and which we can reconstruct without too many problems. Beneath this layer of events, there exists another type of events that are a bit more diffuse- events that are not per­ ceived exactly in the same way by the contemporaries, but which they have a certain awareness of all the same, for example, a lowering or an increase in prices which will change their economic behavior. And then, beneath these events as well, you have others that are hard to locate, that are often barely perceptible for the contemporaries but nonetheless constitute decisive breaks. Thus the reversal of a trend, the point at which an economic curve that had been increasing levels off or begins to decline, such a point is a very important event in the history of a town, a country, or possibly a civilization, but the people who are its contemporaries are not aware of it. In our own case, de­ spite a relatively precise national accountancy, we don't exactly know that the reversal of an economic trend has occurred. The economists themselves don't know whether a stop in an economic curve signals a great general economic reversal of the trend or simply a stop, or a little intercycle within a more general cycle. It is history's task to un­ cover this hidden layer of diffuse, "atmospheric," polycephalic events that determine, finally and profoundly, the history of the world. For it is quite clear to us now that the reversal of an economic trend is much more important than the death of a king. Population increases, for example, are studied in the same way. The fact that Europe's demographic curve, which was pretty much stationary in the course of the eighteenth century, rose abruptly at the end of the eighteenth century and continued to rise in the nineteenth is, in part, what made possible the industrial development of Europe in the nineteenth century, but no one experienced this event in the way that one might have lived through the revolutions of 1848. An inquiry has begun concerning the modes of alimentation of European populations in the nineteenth century. It was noticed that at a certain moment the quantity of proteins consumed by Europeans started to

Return to History rise sharply. This is an extremely important event for the history of consumption, for the history of health, for the history oflongevity. The abrupt increase in quantities of proteins consumed by a population is, in a sense, much more significant than a change of constitutions or the transition from a monarchy to a republic, for example. It is an event, but an event that cannot be grasped by the classic and tradi­ tional methods. It can only be dealt with by an analysis of series of documents that are often neglected, series that are as continuous as possible. So with serial history we don't at all see the event dissolving in favor of a causal analysis or a continuous analysis but, rather, lay­ ers of events multiplying. 1\vo major consequences follow from this, and they are intercon­ nected. The first is that history's discontinuities will multiply. Tradi­ tionally, historians dwelled on the discontinuities in events such as the discovery of America or the fall of Constantinople. It's true that such events may involve discontinuities, but the great reversal, for example, of the economic pattern -which was characterized by growth in Europe during the sixteenth century, which stabilized and became regressive in the course of the seventeenth century- marks another discontinuity that is not exactly contemporaneous with the first one. History appears then not as a great continuity underneath an apparent discontinuity, but as a tangle of superimposed discontinui­ ties. The other consequence is that one is led in this way to discover different types of time spans in history. Take prices, for example. There are so-called short cycles: prices rise a little, then, reaching a certain ceiling, they come up against the threshold of consumption and at that moment they go back down a little, then climb again. These are brief cycles that one can isolate without any difficulty. Be­ neath this short time span, this oscillatory span, as it were, you have more important cycles that last twenty-five to fifty years, and then, farther down, there is what are called, in English, secular "trends" (the word is passing into the French language), which is to say great cycles of expansion or recession that, in general, wherever they have been observed, cover a period of twenty-five to one hundred and twenty years. Then, beneath even these cycles, there is what French historians call "inerties," that is, large-scale phenomena operative over centuries and centuries: for example, agricultural technology in Europe, the ways of living of European farmers that remained largely unchanged from the end of the sixteenth century to the beginning

45 0

Aesthetics, Method, and Epistemology

and, in some places, up to the middle of the nineteenth century - an inertia of the peasantry and of the agricultural economy above which one had the great economic cycles and, within the great cycles, smaller cycles, and finally, at the top, the little price and market fluc­ tuations that can be observed. History, then, is not a single time span [durce] : it is a multiplicity of time spans that entangle and envelop one another. So the old notion of time should be replaced by the notion of multiple time spans, and when the structuralists' adversaries tell them, "You're neglecting time," these adversaries do not seem to real­ ize that it's been a long time, if I may say so, since history got rid of time, that is, since historians stopped recognizing that great unitary time span which would sweep up all human phenomena in a single movement. At the root of historical time, there is not something like a biological evolution that would carry away all phenomena and all events. In reality there are multiple time spans, and each one of these spans is the bearer of a certain type of events. The types of events must be multiplied just as the types of time span are multiplied. That is the mutation that is occurring at present in the disciplines of his­ tory. And now I will finally arrive at my conclusion, with apologies for reaching it so late. I believe that between the structural analyses of change or transformation and the historical analyses of types of events and types of duration, there is, I won't say exactly an identity nor even a convergence, but a certain number of important points of contact. I will indicate them by way of ending this talk. When they deal with documents, historians do not treat them as something to be interpreted, that is, they don't look behind or beyond them for a hid­ den meaning. They treat the document with a view to the system of its internal and external relations. In the same way, the structuralist, when he studies myths or literature, doesn't ask those myths or that literature what they may translate or express of the mentality of a civilization or the history of an individual. He makes every effort to bring out the relations and the system of relations characteristic of that text or that myth. Rejection of interpretation of the exegetical approach, which looks behind texts or documents for what they sig­ nify, is an element that one encounters both among structuralists and among today's historians. The second point is, I believe, that structuralists and historians alike are led in the course of their work to abandon the grand old

Return to History

45 1

biological metaphor of life and evolution. Starting in the nineteenth century, people have made a lot of use of the idea of evolution and adjacent concepts to retrace or analyze the different changes in hu­ man societies or in man's practices and activities. This biological metaphor that enabled one to think about history offered an ideologi­ cal advantage and an epistemological advantage. The epistemological advantage was that, with biology, one had an explanatory model that had only to be transposed term by term to history. Thus it was hoped that this history, becoming evolutive, would finally be as scientific as biology. As to the ideological advantage-very easy to identify - if his­ tory is indeed caught up in a time frame analogous to that of life forms, if the same evolutionary processes are at work in life and in history, then human societies have no particular specificity, and they have no other lawfulness, no other determination or regularity than life itself. And just as there is no violent revolution in life, but simply a slow accumulation of tiny mutations, in the same way human history cannot really have the potential for a violent revolution; it can never harbor within itself anything more than imperceptible changes. By metaphorizing history on the analogy of life, one thus guaranteed that human societies would be incapable of a revolution. I think that struc­ turalism and history make it possible to abandon this great biological mythology of history and duration. Structuralism, by defining trans­ formations, and history, by describing types of events and different types of duration [dune], make possible both the appearance of dis­ continuities in history and the appearance of regular, coherent trans­ formations. Structuralism and contemporary history are theoretical instruments by means of which one can - contrary to the old idea of continuity - really grasp both the discontinuity of events and the transformation of societies. NOTES I

2

3

Franz Boas, The Mind oj Primitive Man (New York: Macmillan, 1911); Race, Language and Culture (New York: Macmillan, 1940). Edward Tylor, Researches into the Early History oj Mankind and the Development oj Civiliza­ tion (London: Murray, 1865); Primitive Culture: Researches into the Development oj Mythology, Philosophy, Religion, Art and Custom (London: Murray, 1871), 2 vols.; Anthropology: An Intro­ duction to the Study ojMan and Civilization (London: Macmillan, 1881). N. Troubetskoi, Zur aligemeinen Theorie der phonologischen Vokalsysteme, Travaux du Cercle linguistique de Prague (Prague, 1929) vol. I, pp. 39 67; Grundz-uge der Phonologie, Travaux du -

432

Aesthetics, Method, and Epistemology

Cercle Iinguistique de Prague (Prague, 1959), vol. 7. [Principles of Phonology, trans. Christiane A. M. Baltaxte (Berkeley: University of California Press, IgB9)]. 4

Roland Barthes, Le Degre zero de l'ecriture (Paris: Seuil, 1955), [Writing Degree Zero, trans. Annette Lavers and Colin Smith (New York: Hill and Wang, 1968)].

5

Louis Althusser, Pour Marx (Paris: Maspero, 1965); "Ou 'Capital' a la philosophie de Marx," in Althusser, Pierre Macherey, and Jacques Ranciere, Lire "Le Capital" (Paris: Maspero, 1965); vol. I. pp. 9 - 8g; "L'Objet du 'Capital,' " in Althusser and Etienne Balibar, ibid., vol 2, pp. 7 - 185.

6

Georges Dumezil, Horace et les curiaces (Paris: Gillimard, 1942).

7

Huguette and Pierre Chaunu, Seville et l'At/antique (Paris: Sevpen, 1955- lgBo), 12 vols.