The bibliographical knowledge of Athenaeus and his characters is of a cartographical nature; it organizes a space, subdivides it, and proposes different points of view, which lead from the literary genre to the text and to the quotation, or vice versa. This shift in the hierarchy between container and contained corresponds to a change of scale and to specific forms of articulation: the articulation of sections of the library, literary genres, works within the corpus of an author, books, the rolls that compose them, and the words that fill them.

Callimachus’ Pinakes are, in practice, the map of a library aimed to cover the totality of paideia. That orientation-table traces the horizon shared by the deipnosophists, even if, after Callimachus, the strata of erudition, Hellenistic first, then imperial, were deposited on the Alexandrian base, to which they brought the additional depth of learned monographs, philological conjectures, lexica, and commentaries. However, the structure of the map itself was not subject to change, since it was based on the inventory and the ordering of the literary and intellectual heritage of classical Greece. Even though the map was enriched by inevitable corrections and additions of detail, it was not modified by them. Rather, it acquired relief, thanks to the development of a metaliterature which is a center of activity, in commentary, edition, collection of words, facts, and quotations, all according to the creation of new objects of knowledge. The Deipnosophists are but a stage in a history which is still ongoing today.

The members of Larensius’ circle master a certain number of literary techniques which allow them to situate themselves on that map in its two dimensions, horizontal and vertical. They have at their disposal an intellectual arsenal to take advantage of that library and “work” the quotations they extract from it, with the aim of producing collectively, in the course of their convivial conversations, a form of knowledge. Those literary practices take the form of a group of gestures, that range from the most concrete (handling books) to the most abstract (interpreting, criticizing, commenting, correlating). The literary form chosen by Athenaeus, a polyphony of speeches that follow the development of the table conversations, stresses the construction of knowledge, and not only its objective content. Unlike a lexicon or an encyclopedia, which, if anything, use objective information without clarifying its origin, Athenaeus privileges a reflexive and dialectical dimension within which his characters make explicit their praxis, their intellectual tools, and their criteria of judgment. The Deipnosophists thus offers a privileged testimony on the expertise of scholars, and introduces us into the heart of the practices of a circle of grammarians, rhetoricians, physicians, and musicians who share, beyond their different specialties, the same basic techniques.

To present those techniques would involve taking into account all the components of imperial literary culture. Knowledge distinguished ordinary reading from specialized reading, capable of resolving the difficulties that one comes against in reading ancient texts. That knowledge is mentioned in general terms, as an indispensible prerequisite to understanding, for example, an epigram by Simonides (10.456e). However, it also manifests itself in the course of the text, both within the conversations of the deipnosophists and within the Deipnosophists which relates them, since Athenaeus remains the director of all those literary games. This grammatical culture manifests itself, for example, in the procedures necessary to solve a problem of interpretation relating to a word used by Anacreon: to collect the corpus of previously suggested solutions, discuss them, and look for complementary information in specialized treatises (14.634b–636c). It also comes through in the isolated explanations brought to a text in the very course of its recitation, the oral equivalent of the “marginal notes” or “footnotes” around a written text: the comment, “by ‘leaves’ he means not those of the fig tree, but those of the poppy,” is specified in a parenthesis during the quotation of the list of flowers used for crowns contained in Nicander’s Georgica (15.684a). These scholars are conscious of the necessity of verifying the texts they read, comparing several copies if necessary, and noting their significant variants: this is what happens with the work of Nicander himself (15.684c). The deipnosophists do not hesitate to discuss the conjectures of the Alexandrine editors, conscious of the importance of the establishment of the text for the interpretation of the meaning of the works that they read. It could simply be a question of modifying the punctuation of a verse of Homer (1.12 a–b), of comparing three verses of Eubulus with Alexis (1.25f–26a), of endorsing Aristarchus’ suggestion concerning the suppression of a verse of the Iliad (2.39d), or of noting the different orthography of a word in Hesiod and in various copies of Antiphanes’ Minos (2.58d). [1]

Reading has an exploratory and heuristic function: it provides Athenaeus’ characters with the material for their conversations, where the rules seem to be, like in Callimachus’ learned poetry, that nothing should be said that is not based on the evidence of a source. [2] Reading is thus collecting interesting facts, words, and quotations—in sum, everything that is worth mentioning and remembering at the same time. Although the deipnosophists share the same library, they are nevertheless very different with respect to what they can gather from it and with respect to what attracts their interest as they are reading. “While reading the twenty-eighth book of Poseidonius’ Histories, I found, my friends, an absolutely delicious passage on perfumes, a passage that would not be out of place in our sumposion” (15.692c). The talent of that reader lies in offering a paradoxical point of view on the work of Poseidonius, taking the Stoic to task on a question of minor importance if to the point, while his treatise without doubt included more serious content. Reading means taking note and memorizing, so as to reactivate the memory of a read passage during a conversation, or, possibly, in the writing of a new text. Thus, one may have read Phylarchus’ entire History, as in the case of Myrtilus, without having noted the passage of Book 23 on the cities of Ceos, where there are neither prostitutes nor flute-players; a passage that on the contrary, did not escape Cynulcus (13.610d).

The characters of Athenaeus thus devote themselves to an activity characteristic of the scholars of Hellenistic and imperial times: they take reading notes so as to collect and order raw material destined to be reused in new erudite writings or during conversations. [3] Besides, the banquet conversation lent itself particularly well to that sort of exhibition and one could prepare a repertoire of quotations and lines precisely in view of such meetings. Thus, at the beginning of his work, in a section that is unfortunately epitomized, Athenaeus mentions the figure of Charmus of Syracuse who had verses and proverbs ready for each course of the banquet. As for Calliphanes, he prepared by copying the three or four first verses of a certain number of poems (1.4a–c). Next to those practices of dilettanti who wished to give themselves a patina of erudition, the deipnosophists appear in the guise of authentic scholars who have read their texts from beginning to end, or at least who boast they did, and have compiled them in a comprehensive enough manner to be able to use the words and quotations suitable for the various moments of the conversation.

Being capable of quoting texts from memory is a characteristic of the educated man (pepaideumenos), even if he is a slave (3.108d: Myrtilus’ slave) or a cook (3.102b; 9.381f–382a). Indeed, learning poetry or extracts of theatrical plays by heart was a common school exercise, and reciting those texts was among the pastimes of cultivated Greeks (4.164a; 8.335e; 11.482d; 12.537d; 14.620b; 15.693f–694a). However, the deipnosophists were not satisfied only with quoting texts learnt by heart. They had special techniques that allowed them to navigate within the corpus of memorized texts. Indeed, being able to find the verse, phrase, or word required by a particular moment in the conversation presupposes the capacity to move in a non-linear manner either among texts learned by heart or in collections of fragments and quotations that were previously isolated: while the recitation of a poem follows the order of the verses from beginning to end, Athenaeus’ characters are capable of immediately recalling to memory a particular verse, and of putting it in relation with a verse of another composition. That non-linear memory suggests a form of mental indexing that makes it possible to produce series of quotations ordered according to the same keyword. Mnemotechnics make up for the limits of the ancient book, the papyrus roll, where the columns of text succeeded each other linearly without pagination or numbering of the lines of text, without an index nominum or an index verborum. Finding a specific passage in a roll was not impossible: a sign put in the margin of the column could serve as a point of reference, as can be seen from the practice of the Alexandrian philologists, who equipped the margins of the text’s columns with diacritical signs. When, however, it was a matter of mobilizing scores of quotations on a single theme or a keyword, as Athenaeus’ characters are doing, looking for occurrences directly in books turned out to be an absolutely impossible operation. Without doubt, the lexica and glossaries had the function of providing a great number of quotations, and specialized treatises on the most disparate subjects also collected material extracted from the literary sources. The rolls of reading notes, where the scholars organized their extracts, supplied an instrument of personal archiving; Athenaeus and his characters make great use of it. However, in their conversations, the deipnosophists consult neither lexica nor reading notes, but resort to their memory. Is that a literary fiction that hides the work of compilation done by Athenaeus? Perhaps. Yet the practice of quotation represented in those terms also corresponds to an effective form of “gymnastics of the mind,” [4] which consists in moving within memorized texts as one would move within a concrete book-roll, whose reading could be interrupted and resumed at any line of the text. [5]

The deipnosophists extend this game to the entire library. Their challenges and their exchanges consist in putting together when they speak the highest number of quotations extracted from various texts. If, as in the case of King Cassander mentioned by Athenaeus (14.620b), a cultured Greek, educated at the school of the grammarians, could know by heart large extracts of the Homeric poems (when he did not know them integrally), [6] it is difficult to believe that the deipnosophists had memorized in their entirety the works they cite. The meaning of the word “memorization”, however, needs to be clarified: to quote a passage literally is not the same as to paraphrase its content, or to know that this piece of information can be found “somewhere” in a given book. Thus a vague assertion based on an inaccurate memory (“Hegesander spoke of the citrus somewhere, but I don’t remember where.”) does not resist the certainties of someone who has just finished reading the entire text and certifies that, in that text, there is no mention of citrus whatsoever (3.83a–c). That notwithstanding, the fact remains that the quotations rely on memorized texts, and that the listeners are able to control and verify the quotations of others by confronting them with their memory of the texts. Inviting Ulpian to start his catalogue of flower crowns, Myrtilus takes care to ask him not to cite the treatise On Crowns by Aelius Asclepiades, because it is known to everyone (15.676e–f).

Athenaeus’ particularity resides in presenting that process of recollection and that itinerary within the library in the form of a collective oral performance, in the mutual exchange and polyphony of a conversation. The mnemotechnical virtuosity of the deipnosophists is at the level of their excellence in the field of paideia. It is difficult to escape the temptation to associate their performances with what Eunapius tells us of the sophist Longinus, one of Porphyry’s masters: Longinus was a sort of “living library” (bibliothēkē tis … empsuchos) and a “wandering museum” (peripatoun mouseion). [7] Under the guidance of Longinus, Porphyry indeed reached the culminating point of paideia, at the summit of grammar and rhetoric. The author of many works, Longinus distinguished himself especially by his critique of the ancients, and in literary circles his judgment reigned supreme among his contemporaries.

Why this strange status for Longinus? A library and a museum at the same time, he is, so to speak, an entire Alexandria by himself. He embodies Alexandria’s erudition and critical authority, and his works fill the libraries of others. He also, however, bases his judgments on his knowledge of the works of the ancients, and that living library is in effect an incorporated, memorized library. Eunapius’ description could easily be applied to Athenaeus’ characters: to Ulpian, Masurius, Larensius (whose profound critical sense is highlighted from the very beginning of the work: 1.2b), or to Charmadas, that Greek mentioned by Pliny the Elder who, when shown a book in a library, was able to recite it by heart as if he were reading it. [8] The specific techniques that could allow one to forge a memory of that sort are attested by Seneca, who mentions his contemporary Calvisius Sabinus, who had bought at high price slaves trained to be living books: each one had learned one classical author by heart—Homer, Hesiod, or the Lyrics—and had the suitable quotations ready at the disposal of their forgetful master during the banquet conversations. [9] The wealth of Sabinus’ library depended on the number of living books that he had succeeded in putting together: some Greek classics, it would seem, were sufficient. Unfortunately, Sabinus also had a very weak short-term memory, and he was unable to repeat the quotations his living books whispered to his ears.

Reading in one’s memory as if one held a book. Pliny’s Charmadas could illustrate one of the rules of ancient mnemotechnics: things are written in memory as they are written on tablets or papyrus rolls. To remember is to read. [10] Material books and libraries are the model of the immaterial books and libraries of memory. [11]

Keeping in mind the deipnosophists’ words so as to report them to Timocrates, Athenaeus reproduces on a macro-structural level the performance of his own characters (7.277b; 14.643d; 15.665a–b). Like them, he has to search his memory (anapempazesthai), not for what he read but for what they said (10.459b–c). However, contrary to his characters, most of the time he remembers entire, continuous speeches, and has less frequent recourse to that form of combinatorial and non-linear memory we have analyzed. [12] In the narrative frame of the Deipnosophists, Athenaeus reconstructs from memory the long thread of the banquet’s conversation, the succession of speakers, and all the texts and authors each of them quoted. This recollection process is duplicated: first, in the oral account of Athenaeus to Timocrates, and second, in the transcription of this account in fifteen book-rolls. Athenaeus was too erudite to ignore the tradition about the invention of the art of memory. The poet Simonides of Cos was able to identify the corpses of Scopas and his guests, after the roof of the banquet room fell and crushed them all, since he remembered the position of each of them on the couches around the table. [13] This is the founding myth of ancient mnemotechnics, relying on a system of organized places where one can store information, either facts and ideas or literal quotations. In order to retrieve this content, one has to go through these mental places and to activate the data stored in each of them. Athenaeus relies on two different systems of mnemonic places: the first one is the organization of the banquet room, where the guests are lying on couches around central tables; the second one is the sequence of events and dishes, from the very beginning of the banquet to the end of the symposium. The mix of these two systems allows Athenaeus to reconstruct the entire thread of the conversation, linking together the speakers and the topics they discussed.

In the Deipnosophists memory is performance, and as such it is inseparable from the interaction of the guests. What sets it in motion is the game of questions and challenges that takes place within Larensius’ circle; what sets off recollection is the zētēsis, a technique of investigation and framing of questions that invites one to explore the libraries of the mind. Zētēsis is inseparable from a mode of experiencing culture and knowledge: this is attested by Ulpian, the investigator par excellence, symposiarch and promoter of the game of the dialogue (1.1d–e). Zētēsis is, in its proper sense, “the search for solutions, explanations or responses to a given question”: this intellectual activity can mobilize the most profound critical capacities, and Larensius’ eulogy mentions his skill in that art (1.2b–3b). Finding the answer requires secure erudition, and sometimes time, many interlocutors, and the luck of coming across a book that gives the solution (15.676e–f). It is also a social activity, which presupposes a search for interlocutors, an audience, the pleasure of searching together, and of being the first to find the goal. The banquet, the place and time of distraction and paideia, provides a privileged framework for this party game, [14] but Ulpian (as we have seen) also practiced it in the streets, the baths, and the shops of booksellers. So is it a matter of finding the correct answer to the questions? Or of finding answers whatever the cost? The deipnosophists provide answers that complete each other, that add nuances, and sometimes that contradict each other. The zētēsis is the driving force behind the conversation and the exchanges and it modulates the rhythm of the interventions, facilitating some to suggest solutions while others search in silence (3.83a–c). That way of proceeding is strongly criticized by Cynulcus (3.97c), but in fact he is himself a virtuoso of it (3.106e), although in less obsessive a manner than Ulpian.

Zētēsis is an essential operator in Athenaeus’ text: it allows the passage from things to words, from food to discourse on food, from table to library, from one guest to the other one. The passing of dishes and the ritualized unfolding of the meeting automatically impose the subjects of research, so that the deipnosophists do not need to use their imagination or to negotiate the appropriate themes like the guests of Plutarch. [15] Formulating a question, however, stimulates constant and reflexive attention to what is being said, done and seen in the course of the meeting.

Asking questions about whatever exists in the world, seizing on words and attributing a problematic strangeness to them: that is the sport to which Ulpian devotes himself, often knitting his brows (9.385b); that way he fills an essential function in Athenaeus’ text, since he is the one who initiates changes of subject in the conversation, both in the progressive specification of the topics and through explicit digressions (3.115b). Others less gifted than he could find in books lists of problems ready to be used in the conversation, together with their solutions: the written text thus came to compensate for blanks in the memory and erudition of his readers. [16]

This intellectual exercise appears in the form of a real party game that is not devoid of didactic implications, a game to which the members of Larensius’ circle devote themselves with pleasure, with a sense of humor yet at times grudgingly: “We feed on questions,” remembers the house master himself (9.398b). This collective procedure for the production of knowledge clearly perpetuates techniques that were in use in the Athenian philosophical schools and in the Museum of Alexandria itself. [17] Athenaeus privileges a model of inquiry that requires simple and immediate answers, but does not ignore the fact that zētēsis could also take on more complex questions, real riddles on which generations of scholars had practiced their skills, and zētēsis had given rise to a specialized literature that expounded the problems and sometimes the solutions as well (see e.g. 3.85e; 15.673d–674b). Those banquets devoted to all sorts of philological, literary, and lexical research suggest a further analogy between Larensius’ circle and the Museum of Alexandria. According to Porphyry, the Museum’s scholars were in the habit of suggesting zētēmata, and of putting down in writing the various solutions offered. [18] One can admit, following William J. Slater’s invitation, that an important dimension of “academic life” at the Museum of Alexandria consisted in banquets, where philological problems, real or facetious, were discussed while drinking, and the solutions competed with one another in their ingeniousness. Some of Aristophanes of Byzantium’s textual conjectures may have drawn their inspiration from those traditional questions. [19]

Plutarch and Athenaeus thus testify to the happy fortune that the Alexandrine model knew among private circles, where the symposium was the setting for erudite discussions, ordered around the questions offered to the sagaciousness of the audience. [20] Contrary to Plutarch’s guests—to the most eclectic curiosity and to certain aspects that are closer to the Aristotelian tradition of “problems”—the deipnosophists remain within the field of literary and lexicographical erudition. Besides, it is significant that Athenaeus chose to organize his work on the basis of that operational principle. The zētēsis introduces us to the heart of his project and to the implications inherent in his enterprise; I shall underline three of them in order to close our journey: the relation to language; the particular regime of textuality specific to the Deipnosophists; and finally, the construction, within the work, of a reflexive knowledge, through which a culture tries to decipher itself.