Soyons sans illusions. L'homme ne se souvient pas du passé; il le reconstruit toujours.
L. FEBVRE

1There is no doubt that history and translation are bound together. Translation represents not only a central process in historical work, but is, in itself, a historical practice. However, so far these ties have not forged connections across the two disciplines. It must be acknowledged that the difference between the status of translation and history in the research community is such that the use of translation by historians has long been considered "normal" and "natural," while translators studying the history of their profession (so far of little interest to those who are historians by trade) are in general careful not to identify themselves as historians. Based on this notion, this article will focus on two points: first, general issues concerning the role of history in the history of translation; second, the question of periodization, to show how it is fundamentally subjective and marked by its time as well as by the institution (in a greater sense) that produced it.

1. Role of history

2First let us delve into questions which are important to scholars, who, without being historians by trade, study the history of translation with varying and sometimes uncertain outlooks on the proportions of history. Consider on this point the reflections of Pym regarding the role of history in the history of translation,1 an issue which is not clearly addressed, however, in his book devoted to the examination of methods used in constructing the history of translation.

3What about the status of these two disciplines, history with a capital H (a large hatchet for the wink to Perec!) and translation? History — a noble discipline among the social sciences, a veritable institution in itself. History with its diverse fields of research, its periods (long and short), its current fields (economic, for a time), culture, social practices, beliefs, portrayals, ideas, women and youth; but also the history of war, of ideology. Furthermore, the duty of memory is now inseparable from historical practice and straddles the discipline between science and memory. One cannot forget also the entire issue of discourse of and on history, its writing, the plotting of the story, essential to the historian, the "how we write history" to reference Paul Veyne (1971). Lastly come into play the issue of historiography, the sometimes uncertain realm where knowledge and social demand strike an awkward balance, the equally important question of practices and methods, the many issues uncovered by this question, and even more importantly, the (contemporary) epistemological debate which relates to the role of historians as well as the objectivity of historical knowledge. From the positivist movement inherited from the nineteenth century to Marxist or structuralist scientism that marked the middle of the twentieth century — the claim of a historical truth, modeled on the idealized image of the natural sciences, driven by the scientific philosophy — the historical discipline has experienced a very thorough questioning of itself. Some of the questions are linked to the status of figures in history, others to its ability or inability to dissect and portray the event, not to mention the often difficult relationships the discipline has with sociology. Hence, in the end, the position, relativist par excellence, by which the realities of the past authorize, by their essence, by nature, each group, each community (think of American Black History or of the most radical movements of feminism) to conceive of its own norms of truth, the historical discourse now finally able to be considered a fiction among others. To cite only two examples, think of Hayden White in the United States or of Paul Veyne in France, for whom history is nothing more than a narrative in which the historian constructs a plot around certain ordered events according to editing chosen in light of the research subject, the theme under study. We return later to precisely this issue, in discussing periodization in the history of translation. There is no doubt that there was need to, and that there is still a need to, otherwise define the very concept of historical objectivity and broach the question of knowledge, historical or otherwise, as relevant to an ensemble of social practices, to avoid the stalemate of scientific dogmatism.

a. Historia a Debate website

4These questionings on the status of history and of those who create it are particularly well addressed on a very interesting Internet site, built out of a series of contacts, reflections and debates. It is the Historia a Debate2 site, under the direction of Carlos Barros, a history professor from Santiago de Compostela (Spain). The initiative grew from a desire to put forward a "critical position," inviting dialogue among various historiographical movements, mainly the continuism of the sixties and seventies, postmodernism, and the return to the old history, amusingly described by those in charge of the "last historiographical breakthrough" site. This site is an excellent example, unique in its genre and rare, of practical and completely relevant application of the Internet to propose and sustain debates on such vast topics as What is history or What are its main components or What are the most important tasks for the historian nowadays. The interesting part of this work actually lies in the fact that the format of the Web itself enables change and updates, as well as the interaction of numerous participants. Recently this site presented the results of an extensive international survey3 conducted between the months of March 1999 and the end of the year 2001, a survey aiming to give an overview of the state of the art, to offer a diagnostic on the state of the discipline, aiming to define it, to explain its methods, writing the present and the past. This survey is too vast to discuss in much detail here, but it is of great interest, particularly for "non-specialists," as are many translatologists. The survey addresses many points relating to possible definitions of the discipline (for example, "history is the science of society," "history is the science of humanity through time," or "history is not a science because it does not allow one to know the objective truth of past facts," or "all history is contemporary"). It raises questions such as "How does reflection on methodology interest historians? Historiography? The theory of history? The theory of society?" It explores the ways in which philosophers and thinkers such as Marx, Freud, Weber, Durkheim, Lévi-Strauss, Foucault, Habermas, Bourdieu, Derrida, and Ricoeur have influenced, or continue to influence, the work of historians. These issues that arise in history create history, and build histories connected to the sometimes moving and incomprehensible subject of translation; translators are not often considered.

b. History in the history of translation

5Pym is not wrong when he points out the relative absence of history in the history of translation. Could this be due to a "chasse gardée" phenomenon, because translation historians consider themselves, as specialists in the field, authorized by this fact to trace the contours and determine the issues, their issues throughout centuries, in a position of reclamation, of defence (normal, incidentally, if the "ancillary status," — to use the words of Berman — of the translator and of his or her practice throughout the centuries is to be considered)? But how do we know that the translator lies at the heart of the history of translation, as the approach adopted by a number of recent research projects (for instance Delisle's Translators in History and Portraits of Translators) on the history of translation seems to indicate. In other words, aren't we trying to make it happen? It is precisely the question of SUBJECT that is treated in the first proposal of the History Under Debate Manifesto that has around three hundred and ninety subscribers, mainly historians. The manifesto is presented in eight languages4 and includes sixteen methodological, historiographical and epistemological proposals. It touches upon issues that vary from the reclamation of "SCIENCE WITH A SUBJECT," which avoids both the outdated, naive objectivism of positivism and absolute postmodern subjectivism, to the defence of the ETHICAL FUNCTION of history by virtue of the role it plays in society and more particularly in its formation, the awakening of consciousness and the education of citizens. It does so by discussing the new scholarship which engenders not only the enlargement of the source concept but also the resorting to "non-sources" that embody the silences, the errors and the gaps in history [see Santoyo in this volume]. In other words, the "history built from ideas, hypotheses, interpretations," which enables it to discover, to construct, even to deconstruct sources, combines itself with the celebrated "history written with documents."

6Thus the issue arises of knowing to what extent to pursue the hypothetical. To cite one example (among many), particularly interesting because it is related to translation history, let us mention the review of Pym's work Negotiating the Frontier: Translators and Intercultures in Hispanic History (2000) by Julio-César Santoyo.5 In this book Pym presents a great number of hypotheses, a great deal of interpretation, considered as mere speculation by Santoyo: "The reader will find a good sample of such speculation on the final pages (30–33) of the first chapter, with its fifteen would's, five perhaps's, six probably's, four might's, two imagine's, and two hypothetical's, besides a heavy dose of such other qualifying terms as apparently, possibly, likely, apparent, suppose, may and so forth." Clearly two visions of how to write history, two postures towards historical facts and objects, and two perspectives on how to present or represent the past. Debilitating relativism on one part? Manipulation of evidence? Too much confidence in the reality of the referent on the other? Too much emphasis on raw data? The reader will be able to draw his own conclusions; at the risk of sounding too ecumenical, can it be said that the two approaches do not need to be considered as exclusive, one from the other?

2. Periodization

7The second point will address the topic of periodization, which we know to be indispensable, though at times it can also be problematic because it is so systematic. Thus, in the words of Antoine Prost (1996, 116):

History cannot do without periodization. But periods have a bad reputation in our profession. From Lord Acton, who declared a century ago, "Study problems, not periods," to radical critiques from Paul Veyne or François Furet, periods are a problem. Truly, the problem lies in the preconceived period, served-again period, the one inherited by historians, not the living one. Periodizing in itself is a generally accepted, legitimate practice, and no historian could do without it. It is the result that appears suspicious: the period as an arbitrary frame, a constraint that deforms reality. Because once built/constructed, the period as a historic object inevitably works autonomously.6 [our translation]

8This issue can be quite delicate. If periodization constitutes an indispensable division for defining and delimiting historical evolution, as well as the historical object itself, it often appears to be arbitrary, or at times even inaccurate. The concept of century, base for the "constraints of periods," has been denounced by a number of historians. Other periods have been revisited: with respect to the limits of classical periodization which defines the Middle Ages as the period between the end of Antiquity and the beginning of the sixteenth century, Jacques Le Goff, conscious of the difficulty that such a periodization of short intervals would represent for general understanding, proposed a division of time into longer periods, with shorter sub-periods. Therefore, his long Middle Ages extend from the third to the nineteenth centuries, comprising a first period from the third to the tenth centuries, a second period from the tenth to the fifteenth centuries, and lastly a third period from the sixteenth to the nineteenth centuries. The unity of each period lies largely in the fact that technological phases do not evolve in marked fashion. As Le Goff states, the "era of the mill" will be replaced by the machine of the nineteenth century. The same occurs in the economic, social and "cultural" fields, as well as the eras of famine and of Christianity. To give another example, referring to the same period, let us mention Huizinga (1996) and his famous Autumn of the Middle Ages, a work originally published in 1919 in which he establishes three distinct periods, among others: a twelfth century described as "bubbly," a thirteenth century described as "central," and a fourteenth century christened the Autumn of the Middle Ages. It is evident that the temporal divisions will undergo modifications or changes relative to the issue under study. Let us cite just two examples from the medieval context. First, in reference to the subject of the early university institutions, consider the distinctions generally made between the twelfth and thirteenth centuries. At first, the school followed the instructor, and then later the instructors followed the school. In other words, before institutions appeared, the professor was central, attracting students from different places; later on, professors traveled to the large educational centres, schools and early universities. Clearly, the interest in this type of periodization lies in the fact that the structures and the organization of instruction would be close to the "object" of study. To continue within the realm of knowledge and teaching, let us mention the concept of The Renaissance of the Twelfth Century, a concept put forward by Charles Haskins (1970). It is perhaps Marc Bloch (1952) who best brought to light the artificial character of all periodization:

The Middle Ages, in fact, exist because of their limited pedagogical usefulness: it allows us to structure programs and provides a single name for learned techniques whose extension is far from being clearly determined within traditional dates. The Medievalist knows how to read old texts, can read charts and knows old French. It is something, of course. But it is insufficient in order to establish, through exact divisions, a science based on reality.7

9The "translation object" has itself given rise to a number of periodizations that, if they served well to structure the topic of analysis by applying a certain number of categories or by proposing divisions deemed essential to the organization of knowledge in some way, have also fallen into the "most obvious teleological traps," to adopt (adapting it)8 the expression used by Jean-Frédéric Schaub (2003) in the work that he has recently dedicated to the relations between France and Spain, particularly in the seventeenth century. This is, in fact, my second point: are the temporal classifications not necessarily founded on an evolution which leads to an outcome that is an end in itself? In other words, do we not reduce these classifications to essentialism, to historical continuity and totality in which we know they do not constitute any pertinent hypotheses? Since it is periodized, the "translation object," which would refer to the product or the agent (the translator), appears to be outside the limits or points of contact and articulation of literatures and of the ensemble of intercultural and interlinguistic exchanges that exist within a vast movement. Furthermore, the various periodizations applied to this phenomenon over the course of the last fifty years demonstrate a variety of differing points of view and approaches. While some of these focus on practices, others prefer to reflect upon the surrounding theories. These periodizations or "divisions" allow us to focus on the actual facts of translation or on the reflections surrounding them and thus become better able to define them, explain them, and even analyze them within their own context, as well as in relation to other tendencies rather than in isolation. The periodizations also present the (translation) facts within a continuity, which sometimes confines them to idealization, assimilating, in a sense, the phenomena into a linear movement, a succession or natural chain of events which does not allow the periods of rupture or silences to show through. Neither does it demonstrate the relationships of domination, rather than negotiation, between languages, cultures and literatures.

10a. In 1969, Ljudskanov, for reasons associated with the political and intellectual context of that era, probably found the road of periodization to be the least bumpy or chaotic. He proposed structuring the practice of translation into the following four stages: Word for word translation during High Antiquity and Antiquity, Sense translation from the Early first century AD to the fifteenth century, Free translation from the sixteenth to the eighteenth centuries, and finally Adequate translation from the nineteenth century to what he calls the present day, that is to say, the end of the 1960s.

11Ljudskanov's text is marked by the state of discourse of its era, in which the first studies in machine translation were conducted. From those studies arose an argumentation based on the illusion of a translational invariant.

The practice of adequate translation received a new impetus and rose to an even greater level of perfection thanks to the development of ideas within the Soviet school of translation, which, among others, gives the name to this historical type of translation (Ljudskanov 1969, 25).

12In this article, the basic principles of the fourth and final category to which the Soviet School of Translation is associated are said to be founded on "the most progressive ideas that have been expressed in this field in the past" (ibid., 25, note 21), as well as on "the point of view of the Marxist-Leninists on the language of translation." In other words a "tradition," with all that this concept entails: two sides, part old and part new, but above all a construction based on the necessity to represent translation and its practices within a historical movement marked by progress and result.

13b. In his 1975 work After Babel, which has since become famous, George Steiner proposed a periodization also comprised of four phases, but situating all translational practices from Cicero to the end of the eighteenth century within the same category. Steiner referred to this phase as empiricism, as he envisioned it to be "the epoch of primary statement and technical notation" (Steiner 1975, 236). A second phase corresponds for George Steiner to the hermeneutic approach, a stage where the theory and the reflections on what it means to understand a text predominate. This second stage, for George Steiner, gives to the subject of translation a clear philosophical aspect, and extends from Friedrich Schleiermacher (1813) to Valéry Larbaud (1946). A third stage, marked by automatic (machine) translation and formalism, begins, according to Steiner, in the forties and ends in the seventies. The practice and reflection on translation during this time is influenced by the research on automatic translation enhanced by structuralism and transformational generative grammars. Overlapping this period is a fourth, originating at the beginning of the sixties and marked by interdisciplinarity. Anthropology, sociology, philology, comparative literature, lexical statistics, sociolinguistics, and poetics are called upon to help in the effort to discover more about translation and to analyze its products, to explore its different facets.

14c. Another periodization in four stages was proposed by Julio-César Santoyo in 1987. A first phase of oral translation extends from prehistory to 3000 years BC; a second of written translation is determined from 2400 BC to Cicero. The period when reflections on translation arise stretches from Cicero to the end of the eighteenth century. "Real" theorization begins with Tytler (1791) and Schleiermacher and lasts into the eighties. The periodization here proposed presents a double decoupage between modalities, methods (opposing oral to written for example), reflections, and thoughts on translation. But are these distinctions valid? Can they be used as a base, a foundation? In other words, can they be considered true categories that allow analysis of structure? To provide another example from medieval times, one can say that in the case of the work carried out in Spain during the twelfth century, the use of an oral pivot language (Romance) was particularly important, and as a matter of fact central, to the transmission of knowledge from Arabic into Latin, in practical as well as in symbolic terms. As far as the distinction of practice from reflection or theory is concerned, although it has permeated the various discourses on translation for many years, it can hardly be taken for granted. It may certainly be said that the impact of information technologies on the profession has contributed to the discovery of more and more common ground and forums for discussion between translation professionals and scholars.

15d. Let us turn now to one last periodization, namely the one proposed by Michel Ballard (1992) under the form of five general categories. The first, dubbed Sources, extends from Prehistory to the fourth century. During the second, from the fifth century to the fourteenth century, translation appears as a relay, one could say a go-between. The third covers the fifteenth and the sixteenth centuries, when translation goes hand in hand with discovery and the opening of new horizons. The fourth period is marked by the Belles infidèles in seventeenth-century France; this is for Michel Ballard the moment that gives birth to translation studies per se. The fifth and final category is called Divergences and Reactions (Divergences et Réactions): it begins in the seventeenth century and ends with Walter Benjamin (1892–1944). One cannot but see that this last category is in fact a vast reservoir that contains several centuries of practice and reflection on translation. In this temporal classification more than in any other one, the teleological movement is clear: translation practice and the reflections that surround it are structured around the predictive notions of sources, mediations that tend towards an end, and as a matter of fact, here, give way to the birth of translation studies. In this case, as in any other temporal classification, there is no mention of the presence or absence of translation nor of the gaps between and within periods: the object of translation is considered as natural and given, when it is in fact constructed and predetermined.

16These different ways of (re)presenting translation history and of analyzing its objects in the wider sense (though we are far from the project of a universal history of translation put forward by the committee on translation history of the International Federation of Translators in the sixties) have no doubt been useful to translators and translation scholars. They might even have interested people from other disciplines, have provided a certain structure to our field, have applied a certain number of categories, and have established temporal limits and periods that have allowed us to give a shape to knowledge. But one cannot be blind to the fact that, as with any temporal classifications, they appear as preconstructions which accommodate the object of translation and present translation (process, products, and actors) as part of a teleological movement, as a practice that moves towards a determined and essential end. Through Michel Foucault and others, we have learned that practice opens the way to the object, not the other way around. This explains why it is now common to see the object of translation (practice or reflection on it) as moving and changing, influenced by different paradigms (linguistic, cultural, or deconstructionist). This new approach to translation and translation studies offers a perspective that presents the advantage of allowing us to look at translation objects not as givens but as constructions, as representations, structured by translation scholars into categories which themselves have a history and are based on different interests and power relations. We do not doubt that the time has come to critically re-evaluate such categories.