Classical philology, the dominant discipline in higher education until the early twentieth century, has an ambiguous status; a definition of that status could explain why the field has never produced a theory of meaning, although it has produced numerous methodological works on which the other literary disciplines are based. On the one hand, it evolved into a formidable and illustrious science during the nineteenth century, on the same footing in principle as all the others: thus it relinquished its prerogatives while still defending them by invoking the sacrosanct and prestigious nature of its subject matter. On the other hand, the means at its disposal within academies and universities, even if these were used only to organize subdivisions and auxiliary branches, were granted by virtue of its role within the pedagogical systems of the various countries involved. Yet scientific results were not taken into consideration at this level. Even in the past, the texts for which classical philology was responsible had long borne the imprint of the purposes they were intended to serve, purposes always different from those of the original texts, precisely where those texts had had the most profound influence on systems of thought. This ambiguous status—science and guardianship—prevented philology from questioning the values that it invoked and that were the basis for its influence; this alone explains why it instinctively declined to develop a theory of works or a technique of reading, or to define the nature of the texts under its aegis. Research of that sort would have been incompatible with the stability it represented. And yet, some reflection on the conditions of their practice might have led philologists to conduct a historical review of the uses made of the texts, {15|16} uses that created the subject matter of the field. Having failed to engage in such reflection, they have deprived themselves of the epistemological grounding that would have justified the field’s ambitions and scaled down its pretentions.

The philologists contrast one interpretation with another; they accumulate points of contention ad infinitum through marginal notations and corrections, as if controversy—which they mock, as Montaigne did, but still engage in—were integral to the nature of commentary. But these debates, while necessary to the survival of the practice, would be supplanted if the origins of the proposed translations and of the causes of error were studied; such study would show that, if scholars have not reached a conclusion, it is because, for reasons linked to the way their culture is organized, the very framing of the questions has made it impossible to draw conclusions.

An analytic, rather than descriptive, history of interpretation, which might reconnect opinions with the factors on which they were based, is particularly valuable if it analyzes the process of constituting an illusory science; and, in the specific case of Homeric criticism, it can show how the various approaches, originally intended to clear up difficulties, in fact transformed those difficulties into means appropriate to the organization of a discipline that has prevailed for over a century and still sustains debate. This situation arose because the approach in question had managed to design a virtually autonomous academic and academically profitable system of questions and answers that do not actually refer to the subject; Homer is the case in point. This doxography is necessary to the practice because it isolates the critical points that Homeric analysis, the philological method par excellence, exploits; it views them as flaws in the work, whereas the process of reading, in its very flow, blends and assimilates them. Above all, however, only a comprehensive view of the whole series of interpretations and the historical factors that underlie them, even if we are simply looking at errors, can allow us to move beyond doxography. The best way of avoiding the professional game of controversy is to categorize all opinions and to reconstitute, through them, the representational system that is still in use, except when the text is called upon to support an ideology or to implement a theory, aesthetic or otherwise. In the latter case, the material in question is marked by the tradition developed by earlier interpretations, so much so that it is the tradition that makes use of the users. The philologist critic has as his material, on the one hand, the text or the potentiality of a meaning, and, on the other, all the ways the text has been used—that is to say, the history and culture of each interpreter; thus it is necessary to determine whether or not one is speaking in the interpreter’s name when one means to speak in the name of the text.

The scholars’ disdain for theories—which they teach even though they view them as mere devices or distractions compared to facts, such as archaeological {16|17} remains or “established” meanings—explains why philologists have never included the history of errors in a systematic deliberation on which their science could have been based. The absence of a theory of their own corresponds to their scorn for the theories of others. The principle for analyzing the Homeric poems, which constituted one of the achievements of the philological method, was rarely discussed by the authors who applied it, and no doubt it would not have been imposed as a practice if it had not produced facts that could be reinterpreted indefinitely. The errors that these facts implied were never discussed, because they lacked believers and advocates. Now, facts established by error cannot be acknowledged in the system in which they lurk without a re-­examination of the principles that constituted them; in the same way, one cannot go beyond the accepted meanings without analyzing the tradition those meanings have produced.

Philologists have been even more preoccupied with the genesis of Greek epic than with the birth of tragedy; for more than a century, the question of genesis has been at the center of a debate in which what was at issue was certainly not Homer himself, but rather the material attributed to him. The successive positions adopted by philologists are more evident here than elsewhere; they control and encompass hypotheses and theoretical formulations down to the smallest detail, so that readers are authorized to equate a particular hypothesis with a particular theoretical position, thus bypassing the laborious and thankless task of contending with the numerous variants of a given hypothesis. Thus there will never be any question, in the present essay, of criticizing errors or even revealing them as such.

The nine Wanderings of Odysseus, presented in groups of three, are found in Books 9, 10 and 12. They are arranged in such a way that two brief narratives precede a long one:

• Book 9: Cicones, Lotus-Eaters, Cyclops
• Book 10: Aeolus, Laestrygonians, Circe
• Book 12: Sirens, Charybdis and Scylla, Cattle of the Sun {18|19}

The descent into the Underworld, which is closely linked to the Circe episode, is recounted in Book 11.

Kirchhoff thought he had two proofs of this division of the poem into layers; according to him, the wrath of Poseidon, which struck after the blinding of the Cyclops, formed the kernel of the epic of the Return, since the god is the hero’s implacable enemy. The other “wrath,” that of Helios in response to the sacrifice of his cattle, is superfluous. Secondly, Kirchhoff detected traces of a shift from the third to the first person of the verb in the simple fact that Odysseus speaks of events he has not witnessed himself. These indications affected the second cycle, Books 10 and 12, which the organizer had introduced afterwards.

Some years later, Wilamowitz proposed a different structure, which retained the division of the Adventures but shifted the moment of their linkage, and thus the origin of the poem, further back in time; the writer of our Odyssey was modifying a text that had already been modified. In fact, Wilamowitz imagined an “earlier Odyssey,” comprising a certain number of constitutive elements such as the epic of Calypso and a descent into the Underworld; it featured a longer poem that included Books 10 and 12, that is, the second cycle of Adventures that Kirchhoff had isolated. The “editor” who had already compiled the “Ancient Odyssey” gathered these scattered parts into a whole. The split motifs and the doublets, considered imperfections, could be explained by the fact that he had to use the same theme more than once. In the Calypso epic, the wreck of the raft led Odysseus to Ithaca; in the poem used by Wilamowitz, Odysseus succeeds in reaching the Phaeacians after the shipwreck. By situating the Calypso episode between the shipwreck and the arrival on the island of the Phaeacians, the “editor” had created confusions that interfered with the reading of the Odyssey, as the scholarly readers of the time saw it.

Among the elements that Wilamowitz attributed to the earliest version of the descent into the Underworld, he retained the wrath of Poseidon, the cause of Odysseus’ tribulations. He concluded that the blinding of the Cyclops, the origin of the curse, formed part of the same textual unit. He endorsed Kirchhoff’s opinion that the beginning of the Wanderings (Book 9) had been composed in the first person, while the adaptation of the other stories (Books 10 and 12) had led to a transposition to direct discourse. Thus the obvious links between the first and second parts of the Wanderings, as well as those between the stories of Calypso and Circe, were not ascribed to authorial intent but to a perfunctory work of adaptation. In fact, they were interpreted as borrowings. From then on, every connection confirmed the hypothesis of a secondary fusion, where traces of the operation had remained visible.

The writings of the Analysts proved so difficult that the average reader found them inaccessible, and this acted as a selective principle. It is arduous enough to follow an argument that is based on textual fragmentation. Add to that a critical apparatus indicating attributions, allusions, footnotes, implicit references, and it all showed contempt for literary and non-specialist writing. The theories of earlier scholars and colleagues in the field were presumed to be well known, invoked as self-evident, and relentlessly corrected. Reading the work of an Analyst is a test for the initiated.

In setting themselves apart from ordinary readers, the Analysts abandoned exegesis. Indeed, their operations could not be based on an obscure text needing clarification. Critical explication of a text of any kind was relegated to an inferior level, the realm of teachers. In contrast to Aeschylus, Homer was considered an author without affectation, since he had always been proclaimed a naïve {21|22} genius. It was easier to construct vast hypotheses than to acknowledge the real difficulties posed by exegesis. But above all, if philology was to retain its stature, it had to be differentiated from an activity that was considered unoriginal, appropriate only for students. Analysis was a science: not a science of the text, but a science that used the text to constitute itself as science. Basing their power on the place society accorded to the humanities in the training of elites, and on the importance assigned to the preparation of schoolteachers, the philologists were all the more powerful because they built their scientific prestige on a subject matter that was never discussed in response to the pedagogical needs of their audience, nor in terms of the usefulness of their knowledge. Science lay elsewhere; it had different goals.

All this led to a paradoxical situation: the leading scholars, conscious of this dualism and of the pedagogical inefficacy of Analysis, belittled the naïve understanding of the classics that had nevertheless been the prerequisite of their own academic power and the basis for their work. Left to the Unitarians and the public at large, aesthetic appreciation characterized what was not science.

The scientific method, cold and relentless, is a necessary evil for “modern man.” But the aesthetic sense remains a first cause, which must be used with skill since it offers the wherewithal to support the cult. A certain linguistic violence—the use of a curt anti-rhetorical language—characterizes university society at the highest levels, especially in Germany, and sometimes in England: it is a language of the concrete, of efficacy, through which this group believes it has access to the reality of the affairs and actions from which the University as such is cut off. (In France, the style has remained more pompous and hollow, less aristocratic, because the differentiation between scientific and aesthetic discourse was not produced in the same way.) Like a mask over a brutal and destructive reality, the sense of beauty has the sole function of compensating for a necessary evil and of making the inevitable acceptable—for Homer, division and disparagement. This form of discourse—heir to the rhetoric of the humanists, carefully preserved in France—and even Romantic glorification are occasionally used by a scholarly enterprise that rejects them. Academism is not taken seriously enough for it to get drawn into the game, always remaining prudently allusive, failing to assert itself, and protecting itself by referring to facts.

Thus science, which is addressed to universal reason, occupies the position of an esoteric activity, reserved for the rare initiates. In contrast, meaning, which is hidden and speaks only to visionaries, is in the public domain, with the result that the in-depth exegesis that would have gone beyond pedagogical practice, and that continues to thrive in certain circles, is dismissed as sheer fantasy or theological raving. {22|23}

Analytic criticism could be augmented by other sciences because its work with texts led to the discovery of vanished civilizations in their successive stages and their mutual relationships, and because it sought to rely on historical material. It had already taken over the methods of textual criticism by making even accidental defects into something rich in historical meanings. These accidents served the Analysts’ aims so well that the repetition of such a fortuitous event might well cause readers to doubt that it had actually occurred. Their criticism became confused with comparisons of documents; as a result, literary history was flooded with factitious works. This historical type of exercise, which produced facts that were then related to other external facts, provided material for much university research, for many published books and endless controversies. Interpretation, confirmed by the text, would unquestionably be less rich, less expansive, and finally less “profitable.”

Odysseus is presented with two routes to follow: Circe’s description of them will help him choose, but he must make up his own mind (lines 57–58). Once past the Sirens, he will find “overhanging rocks” that the gods, in their language, call “Planctae.” On one side, among the reefs, there is a smooth rock beyond which no bird flies, where at times even one of the doves bringing ambrosia to Zeus may disappear. Each time, Zeus replaces the dove. On the other side, no boat skirting the rock can avoid shipwreck, whether it is swallowed up by the waves or engulfed in flames. Only the ship Argo has managed to pass safely on her return from Colchis, thanks to Hera, who protected Jason (lines 59–72).

Detail, which is of interest only to philologists, the experts, enters the construction built up by the lines of descent and everything that depends on them; here, the Planctae retain only their philological function; they are no longer a sign, as they were when, in the text, they maintained links with other parts of the narrative. The example demonstrates that the enormous work of two centuries of philology has often only minimally improved our understanding of Homer’s text. From antiquity on, the Planctae have been Rovers. The error predates the analysis. So Homeric science has deprived readers even of what they could have understood on their own, by subjecting these passages to a scholarly transformation. The text was difficult, but once problems were erased they could never be resolved. From this interpretive void have arisen both constructions that tended to increase the interest of the text by external means and errors that obscured and deformed it.

Despite scholarly differences of opinion, this other route is solidly anchored in Homer’s text today, an absurd stump, a vestige of a lost epic, or something like an incidental fact; and what has disappeared is the remarkable transfer of the action into a different action.

Unitarianism was the natural position for an educational system that was based on the authority of the text (and not on that of the interpreter). It asserted its objection to Analysis in order to prove that Homer could be read without “science.” It was not as if anyone, in establishing the text as it stood, had tried to examine it in depth. An initial line of defense consisted of turning to tradition and to the usual practices of scholastic explication. Later on, other scholars tried to demonstrate that Analysis could be considered a heresy.

It is true that Analysis unfairly adopts the forceful speech of the natural sciences, but the resistance mounted by those defending the rhetorical use of texts reveals their motives: by invoking the rights of the imagination and of emotion, they challenge all criticism, be it aesthetic or historical.

The lack of interest in the text itself was compensated for by the wealth of scholarly work produced. Since Analysis invited other analyses, the reconstitution of voyages based on elements in the poem, corrected ad infinitum, provided more fluid and malleable material for discussion than the text itself, and expanded the field of philology toward the historical and geographical sciences.

Stripped of its proper purpose, the narrative can nevertheless be salvaged and used to establish a geography. Concrete information is culled from the myth—the configuration of a promontory or a port, a place name with its subsequent positioning; this is linked to authorial intent, while the rest is viewed as material the poet inherited. Bérard goes even further: he believes that the whole tale transposes details that are geographically accurate, so the role of interpretation is to eliminate the myth and to translate the elements into descriptive terms. He considers the difficulties arising from this approach as mere imprecisions, and ascribes them to Homer’s lack of knowledge.

Other critics, like Wilamowitz, believed that the descriptions revealed traces of a primitive localization, earlier than the one preferred by the singer, and no longer associated with the narrative but rather with a fixed literary tradition. Thus Thrinacia, the island of the sacred cattle, would be placed in the Peloponnese region, and Scylla and Charybdis would refer to the strait between Cythera and Cape Malea. On this basis, starting from the original localization (which would be compatible with the source of an Oriental poem), once the western land of the Cyclops had been separated from the rest, in the interplay of re-readings the sites could be grouped between East and West, so as to be aligned with three phases of ancient history: the concentration in the Aegean {34|35} Sea (where the Oriental poem originated), the expansion to the shores of the Black Sea (the poem itself), and the colonization of the west (the land of the Cyclops in Book 9).

Bérard had laid down a firm rule: “verisimilitude always retains its rightful place in an Odyssean tale.”

It has often been pointed out that this geographical research is fragile and inconsistent, but production has continued nevertheless. The work was fruitful, and that alone was sufficient reason to continue. It is useless to point out the erroneous details, whether Circe lives in Malaga or Campania, or even to criticize the legitimacy of this research, if error itself is not elevated to the status of a subject to be studied, as a means to uncovering its deeper interest and to understanding that meaningless results have not meant that geographical research itself is meaningless.

It was a matter of attaching facts drawn from the text to events, so as to give these events a historical dimension. But by bringing in an external principle of explication based on fortuitous and changing data, the historians in question lost any access to a historical perspective, and failed to consider the conditions under which they themselves were exploiting the text. Using Homer for their own purposes, and treating its meaning as already known in advance, they were {39|40} not in a position to equip themselves with a theory of the text, nor to secure the conditions under which understanding can emerge from the very difficulties of reading. Fixing the historical location of a work forces the reader to go through the subjectivity of the author, and thus to accentuate the anomalies, as so many means for perceiving the distinctive feature of a gaze. The idea of a narrative as the reproduction of a ready-made history, transmissible and translatable, or as a means of describing something external to the text, prevents readers from seeing that the narration includes signs that make it enigmatic and that are, in a way, detached from it. The Wanderings of Odysseus, like other ancient texts, have a depth that has been proven by a tradition of exegesis. Their esoteric nature, the fact that the story cannot be read without demanding reflection on its meaning, precludes any arbitrariness in the interpretation of signs. The elements of the narration maintain a rigorously determined relationship with the interpretive elements within it, so that the meaning remains open only as long as the totality of the relationships, in every episode, has not been explained.

The existence of an exegetic network, of which repetitions, for example, are a part, allowed a blinkered criticism to perceive signs that were incomprehensible on the narrative level as anomalies. Indirectly, the possibility of Analysis vouches for the existence of an accompanying exegesis that replicates the text with a form of allegory that does not replace one text with another but reads the same text differently. Accepting this principle would lead to the definition of a practice that, by carefully distinguishing between the content and the way it is used, as revealed by the text, would retain the specific nature of the reading. We must persuade ourselves that the text disappoints the expectation that the naïve narration encourages. Error consists in producing a meaning that the text deflects. The project is not linked to tradition by a form of dependency, but by the fact that it obliges readers to posit a state of affairs that is different from the one it describes. The object of expectations thus cannot be that of modern critics, who reduce the tale to a primitive form, even as they declare it to be full of urbanity and style: we must reconstitute the expectation that arose from the disappointment of the initial expectation. This means accepting the perceived difficulties while at the same time learning to see the hidden problems.

The insertion of a legend such as the tale of the Argonauts can be considered an intellectual phenomenon, like an element of language. If the Planctae that the ship Argo passes are Scylla and Charybdis, the question arises why, and for what purpose, the solitary passage of Odysseus clinging to the wreck is absorbed into one accomplished by the heroes of all of Greece. The problem of the unity of the poem can only be posed in these terms: unity originates in the necessary relationship among the individual episodes; it is established as the result of the successive interpretations suggested by the text. {40|41}

The research on what is called “oral poetry” begun by Milman Parry and Albert Lord has made no contributions to the topic of the Odyssey’s composition beyond those already made by Analysis. Yet the two men consider, in a new and certainly more accurate way than before, not the genesis of Homer’s poem—which they do not examine, even though they think they have done so—but the genesis of the monumental compositions favored, long before Homer, by the practice of oral poetry.

They have demonstrated the importance of metrical schemas and formulaic models and actions (“formulaic patterns”). Albin Lesky, in his article “Homeros” for the great encyclopedia of the sciences of antiquity, wrote in 1968:

Instead of going along with the Analysts and representing the poet as an ancient genius whose poem was damaged by succeeding versions—as if primitive poetry, symbolized by Homer, had been lost in his work and needed to be recovered—Kirk’s “oral poetry” rehabilitated, in the final stage, the editor and master craftsman. Tradition would lead to a Homer, and this Homer would never stray from tradition. But in the absence of a search for the distinctive traits of this character, by multiplying Homer ad infinitum in his past, Kirk places him back in an unknowable earlier time, since the constitutive act of the Homeric poem is not grasped.