At the time when heaven still embraced the earth, when Uranus still lay with full-hipped Gaia, an aeon before the Olympian gods, the Titans were born and with them, memory, or Mnemosyne. In the Hymns to Hermes, she is called the Mother of the Muses. She is the earliest of the goddesses, preceding even Apollo with his lyre. Hesiod mentions her as the goddess of the first hour of the world and describes her flowing hair as she stretches out beside Zeus on his couch, there to beget the rest of her nine daughters, the Muses. It is she who adopts the son of Maya, the "shamefaced"or"awful"nymph, and thus makes him the son of two mothers. She provides Hermes with two unique gifts: a lyre and a"soul."When the god Hermes plays to the song of the Muses, its sound leads both poets and gods to Mnemosyne's wellspring of remembrance. In her clear waters float the remains of past lives, the memories that Lethe has washed from the feet of the departed, turning dead men into mere shadows. A mortal who has been blessed by the gods can approach Mnemosyne and listen to the Muses sing in their several voices what is, what was, and what will be. Under the protection of Mnemosyne, he may recollect the residues that have sunk into her bosom by drinking from her waters. When he returns from his visit to the spring—from his dream or vision—he can tell what he has drawn from this source. Philo says that by taking the place of a shadow the poet recollects the deeds that a dead man has forgotten. In this way, the world of the living constantly makes contact with the world of the dead.
THE MODERN memory does not derive from the older Mnemosyne, but from another, later Latin word, memoria. Like words and text, memory is a child of the alphabet. Only after it had become possible to fix the flow of speech in phonetic transcription did the idea emerge that knowledge—information—could be held in the mind as in a store. Today, we take this idea so completely for granted that it is hard for us to reconstruct an age when recollection was not conceived as a trip into the cellar to pick up stores, or a look into a ledger to verify an entry. Since the fourth century B.C., memory has been conceived as such a deposit that can be opened, searched, and used. Philosophers have disputed where this deposit is located—in the heart, the brain, the community, or perhaps in God, but in all these discussions memory has remained a bin, a wax tablet, or a book.
For turning this idea topsy-turvy, Milman Parry ranks close to Einstein, although it took much longer before the implications of Parry's achievement were grasped, since humanists, as a rule, are much more conservative than physicists. Thanks to research done in the 1930s by this young Harvard classicist and his assistant Albert Lord, it is now clear that a purely oral tradition knows no division between recollecting and doing. The pre-alphabetic bard does not, like his medieval counterpart, draw on a storehouse of memories in order to compose a poem. Rather, he dips into a grab bag of phrases and adjectives and, driven by the rhythms of the lyre, spins the yarn of a tale.
Parry's thesis, submitted to the Sorbonne in 1928, argued that the Iliad could only have come into being through oral recitation and in the rhythm of spoken hexameters. According to Parry's hypothesis, there are two heterogenous processes by which social continuity is preserved: the flow of prehistoric epic tales that are never repeated word for word; and history that is built on the bedrock of words. In a purely oral tradition, songs, epics, and sayings do not hover above life. That life is a delicate, complex tissue steeped in epic recollections. As soon as the stream of recollections becomes even potentially visible as a narrative, this stream clots and turns into an authority, a point of reference, a socially disembedded rule, the excrement of lived wisdom that a new kind of wise man, called the scribe, can shape.
This epistemological heterogeneity between history and prehistory only gradually gained acceptance. It contradicts the assumption made by the sciences that categories exist to describe human experience tout court. Parry's hypothesis stood up only because the question whether a particular text represents the direct, firsthand transcription of a preliterate tradition can be answered according to strict rules.
The new field of research Parry marked out makes it possible today to determine with certainty whether a particular text is, in the strict sense, prehistoric—whether it is the faithful record of a preliterate improvisation, or the line of a speaker who uses language or memory to compose a text. During the last fifty years Parry's pupils have applied phonologically governed linguistics to the criticism of literary works and the study of oral tradition. In the course of their research, they observed that surviving elements of oral tradition often complemented the detailed study of the linguistic peculiarities of certain major Greek texts and subsequently of epics in other languages as well. They have developed, tested, and refined a number of criteria that make it possible to distinguish oral poetry from every kind of written composition with impressive consistency. Their criteria are the best way we know to evoke the elusive activity of preliterate recollecting in the time beforescriptaof information, originals, or copies emerged.
To begin with Parry's thesis about the Iliad: The Iliad reveals a mastery in self-limitation within given patterns that cannot be imitated self-conscious literacy. What Eric Havelock calls the "variation within the same"has never been approximated by any poet. Only texts that exhibit five forms of self-limitation simultaneously may be regarded as genuine, firsthand written records of oral improvisation: First, in Greek epics, the hexameters are composed of standard word groups. Second, those word groups are mutually attracted to one another during oral recitation. Purely statistically, there is an increasing probability of finding the same formulae in the same section of the epic. Third, the line usually coincides with a syntactic unit: Many lines could be ended with a full stop or a comma because at least the meaning comes to an end there. Fourth, a uniform—though complicated—pattern occurs at the level of the phoneme; combinations of sounds that fall outside the pattern inevitably point to written composition rather than oral improvisation. Finally, this quantitatively verifiable self-limitation relates even to the pattern of the story as a whole: The return of the hero, for example, is always, in oral improvisation, told in the same phrases within the same culture.
According to Parry, the question of the origin of Homeric epics had remained unsolved for so long because it had been wrongly framed. Even today much Homeric research is directed toward looking for an author. Who was the parent of those twenty-seven thousand hexameters? Was he an editor of songs that he had collected from people who knew them by heart? Was he a she? Or was he a godlike poet who composed them himself? Did he write them down, or did he get someone else to do it? Or did others learn them from him, memorizing them, so that much later, after the invention of the alphabet, like a Greek Samizdat, they could be written down?
For Parry, both hypotheses—that of the editor and that of the poet—were equally untenable. Neither learning by heart nor composing were possible in prehistoric times. Before writing, there was no text that could have been internalized and later reproduced like a film script or a part in a play. Not until there was a text could there be a recitation. In Plato's day, there were already people who knew the Homeric epic by heart—in the Ion, Plato describes Socrates' dialogue with such a mnemonist. Xenophon also tells of such a rhapsode who knew all of Homer's work by heart and was admired for it. But that very admiration is already Classical, providing proof that the rhapsode's act of memory was regarded as an extraordinary achievement. No oral society supplies accounts of an epic poet being admired for feats of recollection. They were neither prodigies nor super-Brahmanic mnemonists.
But neither was Homer a man of letters—for the simple reason that there were no letters. The lines of the Iliad do not consist of a series of words. Those who sang it were driven by the rhythm of the lyre. In the twenty-seven thousand hexameters, we can find twenty-nine thousand repetitions of phrases with two or more words. Homer sang as a prehistoric rhapsode—the Greek rhapsodein meaning to stitch together, a linguistic connection that is shared with the Sutras, stitched (sutured) together. Homer's art consisted of stitching together a series of stock words and phrases.
We are so used to drawing a distinction between speaking (and the language that we speak) and thinking (and the language in which it is clothed) that we are no longer capable of composing aloud by improvisation. This difficulty did not exist for the bard: He was composing and reciting simultaneously. As easily as he handles the Greek verb in the rhythm of speech, he finds the first stock word in the poetic vocabulary that leads him on to the next one that will fit in the hexameter. Choosing the one correct verbal inflection from the limited group of forms is as easy for him as selecting the phonetically and syntactically right formula from the vast, but after all finite, group of such formulae in the poetic vocabulary of his time.
In making his choice, the rhapsode was not so much concerned with the actual meaning of the particular adjective selected. It is therefore a mistake to judge these epics according to the aesthetic canon of the Classical Age. Homer, in contrast with Virgil, was not only word-less, but also languageless. The singer of the Iliad, carried along by the beat of the hexameters, was able to locate and use the wonderfully precise nuances of the Greek verb forms and to choose from the enormous store of"winged words."No object remains from this performance. The art of Homer consisted of fluent improvisation within strictly limited means: the art of Classicism gives poetic originality free rein. That originality consists of the deliberate recasting of a given text; that is to say it was based on improving imitation—the mimesis praised by Aristotle. For Virgil, the Aeneid was a work of art: It was an object that he continued improving by changing a word here and there—until, on his deathbed, he wanted to burn it in frustration. The Aeneid allows itself to be paraphrased. In contrast, Homer can only be rendered—the word cannot be pried from the meaning.
Parry's theory remained mere speculation until he managed to observe the singing of living traditional rhapsodes. In the 1930s, he and his pupil Albert Lord traveled to Serbia, where they made the acquaintance of a number of folksingers who still had their roots in the epic traditions of the Balkans. In Turkish coffeehouses and at peasant weddings they sang all night, telling stories to the rhythms of the gusla. Using the complicated equipment of pre-war days, Parry recorded their epics on metal discs in order to check his theory by observation.
No guslar ever repeated the same epic word for word. Every performance was, as Parry expected, a fresh attiring of the old story. For many years after Parry's death, Lord continued the research. He was able to observe the process whereby a youngster became a guslar. First, the young man spent years listening to the master singing. While tending his herds, he practiced using the stock formulae and so gradually became familiar with the poetic vocabulary. With growing assurance he was able, accompanied by the strum of the gusla, to fall back more and more upon those set pieces; but only a small number of skilled bards could draw, even in their maturity, upon the full repertoire of rhythmic fragments. The deeper his active mastery of the wealth of formulae, the clearer his understanding of the content of the songs he heard. Once this faculty was fully developed, he needed only one night's listening to a song he did not know in order to be able to reproduce that song himself a week later. No one could do it on the same day: The guslari say that a story needs time to ferment in the bard—at least a day and a night.
Parry's theory enables us to understand that so complex a structure as the Iliad was sung in a single draft—without the aid of written notes, plans, or drafts. According to Lord's observations in Serbia, it is entirely possible that a single bard assembled from formulae and sang tens of thousands of verses in one outpouring. The riddle of how such work is written down is also solved, according to Lord. In Serbia, he attempted, without tape recorders, to get an accurate written record of long epics. It emerged that collaboration between a clever town clerk and a mature guslar produced surprisingly good results.
At the start, the bard felt annoyed and uneasy about having to pause repeatedly in his singing and rely on plucking his gusla to keep him in time. Soon, however, the guslar began to enjoy this leisure and to use the additional time to utter the proper formula. And in the clerk he found a listener who allowed him to spin out his material at his own discretion until it was exhausted. The writing down of the Iliad could have taken place under similar circumstances, and Homer probably had the same attitude toward the text as the guslari: not one of them was the least bit interested in having so much as a line of the written record read back to him for checking.
The knowledge gained from this comparison of the Serbian guslarand Homer has proved helpful over the past fifty years in the study of cultures that have persisted beyond the reach of records. It has come to form one of the foundations of scholarly discussion of the epic in the Anglo-Saxon world and has led to entirely new insights in the study of the medieval epic.
Oral transmission of epics ceases with writing, and with it, at the dawn of history, fades the idea of memory as the goddess of immortal recollection. For the Classical poet of Greece no longer has need of recollections from a "beyond."No longer is each utterance like a piece of driftwood the speaker fished from a streamful of treasures, something cast off in the beyond that had just then washed up onto the beaches of the mind. No longer are thought and memory intertwined in every statement with no distinction between thought and speech.
When epic tradition becomes a recorded one and custom is transmogrified into written law, the poet's sources are frozen into the texts. He can follow the lines of a written text; the river that feeds its own source is remembered no more. Not one Greek city has preserved an altar dedicated to Mnemosyne. Her name became a technical term for "memory"now imagined as a page: the water of memory turns into the fluency of a writer and a reader. Fixed words on clay tablets acquire authority over the re-evocation of fluid speech.
Plato, in the early fourth century B.C., stands on the threshold between the oral and written cultures of Greece. The earliest epigraphic and iconographic indications of young boys being taught to write date from Plato's childhood. In his day, people had already been reciting Homer from the text for centuries, but the art of writing was still primarily a handicraft. From the seventh until well into the sixth century B.C., reading and writing were confined, in Greece, to very narrow circles. In the fifth century B.C., craftsmen began to acquire the art of carving or engraving letters of the alphabet. But writing was still not a part of recognized instruction: The most a person was expected to be able to write and spell was his own name. The taking of dictation and the fluent reading of written materials were not yet part of knowledge used for control and Education. Until the fifth century B.C., schooling in Athens was purely oral, musical, and gymnastic. Mousike stood at the core of the Greek curriculum: Poems were recited and improvised, rhythmic rhetoric was practiced, pupils learned stringed and wind instruments, singing and dancing. The few pictures in which a teacher is represented with a stylus in his hand show that the alphabet now made it possible for the teacher to read out to the pupils the poems to be learned. Thus a full century before the stylus was imposed on pupils, they were able to learn the texts by heart. That is to say, they gained an understanding of a fixed text that could be listened to, and a respect for the sound of its words, long before they were required to write or read fluently.
Plato's was the time of great change from instruction in elevated, rhythmic public speech to the predominance of prose speech. What formerly could only be recited or sung, can now be pinned down, penned down. The script can be copied, one copy serving as the source for another. The scroll can freeze"materials"for a teacher. It is not the speech but the language of the past that can be made present. Plato heard the Pythagoreans and Socrates. He does not claim to have dictation from them, but he does boast about his faculty of recollection. He is not a traitor like Hippias, who disclosed the orally transmitted secret teachings of Pythagoras. He is already a writer—however anachronistic that may sound. His dialogues are literary prose. He created the model—never surpassed—of the written dialogue that imitates speech. His literary oeuvre forms a counterpart to the record of Homeric song from prehistoric times.
Plato was not Greece's first author. But he was the first uneasy man of letters. He was the first to write with the conviction of the superiority of thought unrelated to writing. He was anguished by the effect the alphabet was exerting on his pupils. Their reliance on silent, passive texts could not but narrow the stream of their remembrance, making it shallow and dull. Earlier, this mistrust of the alphabet had been reflected in Aeschylus'Prometheus Bound* : Zeus punished Prometheus for bringing the alphabet— "the combining of letters, creative mother of the Muses' art, wherewith to hold all things in memory"—to mankind. Zeus had engendered his daughters in the pond of Mnemosyne so that they might bubble and flow, not be locked up in script.
Plato, who saw writing as a threat to the meditative search, kept coming back to the question of Mnemosyne: memory/ recollection. How do we bring the past into the present? He answers the question through Diotima in the Symposium, after he has been extolling Eros: "To what does the word meditation refer if not to knowledge that is past? When we forget, knowledge escapes us. Meditation then brings us to new knowledge and gives it the appearance of still being the same." Diotima describes the search for truth in terms that very closely parallel the process by which the Serbian guslar repeatedly retrieves the same material from oblivion and spins it into a new song. Plato's intellectual path, his access to truth and ideas, is an epic one. This becomes clearer when we read further in Diotima's speech: It forms part of her answer to Socrates, who wants her to teach him about the secrets of Eros. For Diotima, "meditating"is an expression or form of creative love, which in its search for the immortal is always giving itself anew and always withdrawing. Eros longs for what is permanent, and it takes shape when we meditate on the immortal truth, on eidos. Only this kind of loving meditation can lead to wisdom. Plato sees this search for the springs of truth as being threatened by a polymathy based on writing.
To give form to that threat, Plato "fabricates,"as Phaidros puts it, the story of Theuth, the inventor of letters. Theuth seeks to "sell"the letters to King Thamus of Thebes as a pharmakon, a medicine to strengthen the power of recollection and intellect of his subjects. The wordpharmakoncarries a suggestion of magic and the vegetable kingdom. It can be translated as"drug"—either a healing potion or a poison, depending on how it is used. Which of the two was meant was decided by the epithet: In some sayingspharmakonmeans"boon," in others "mischief."Theuth not only presents himself as the inventor of a new means, he also presents a new kind of end.
Thamus thanks him, but he refuses."O skillful Theuth,"he says,"being the inventor of an art is different from being the person who has to decide what advantages and disadvantages that art will bring to those who employ it. You stand before me as the father of letters. With a father's favor, you attribute to letters a fortune that they cannot possess. This facility will make souls forgetful because they will no longer school themselves to meditate. They will rely on letters. Things will be recollected from outside by means of alien symbols; they will not remember on their own. What you are offering me is a drug for recollection, not for memory.... Your instruction will give them only a semblance of truth, not the truth itself. You will train ignorant know-alls, nosey know-nothings, boring wiseacres."Thus in the Classical period memory became divided into two sorts: The natural—that which was born simultaneously with thought—and the artificial—that which could be improved, through precise techniques, or devices, and exercises. The Classical teacher of rhetoric still viewed recollection as the result of a journey, but not to the shore of a river to pick up a piece of driftwood that Plato called "similar"to another piece that had been lost beyond recall. The trip now led to a storage room, as Aristotle says, "to recover knowledge through previous sensationsheld in one's memory."Each of the three primary works of rhetoric (the anonymous Rhetorica Ad Herennium* \82--81 [B.C.], on which later Western traditions of memory training were based and which was attributed to Tullius; Cicero's De Oratore* \55 [B.C.]; and Quintillian's Instituto oratoria* \first century [A.D.]) describes essentially the same mnemonic technique. A person tries to imprint on his memory the interior of a building, preferably a spacious one, visualizing each location—stores, attics, stairs, fore-and antechambers—complete with accessories, such as furniture, paintings, and sculpture. The person then equates the ideas to be remembered with certain images (imagines agentes ); Quintillian uses the example of an anchor and weapon, perhaps to signify ships and war. These imagines agentes are mentally placed into various loci within the building. When the person wishes to "recollect"certain facts, he merely revisits these pre-designated places in the building, and gathers them up once again.
The construction of a memory palace met the needs of the rhetorical arts. To deliver a convincing speech, the speaker must remember it in a planned order; and to prepare for arguments, he must remember points that he has previously connected. (The idea of a planned order would have been, of course, alien to the epic poet, the story unfolding as inevitably as each note followed the next on his musical instrument.) The"palace"of memories provides not only the recollected facts, but also the shape, essential to a well-constructed rhetorical argument.
These architectonic images are suited to the shift from the aural to the visual emphasis that a script culture, like Greece by the end of the fifth century B.C., demands. In fact, Plutarch mentions that Simonides of Ceos, who was believed to have invented the"artificial"mnemonic devices, called painting"silent poetry,"equating the visual aspect of the two arts that Horace summarizes as ut pictura poesis. For the writers of the three Latin memory texts, memory is a signet ring leaving its impression on wax. Aristotle, in his De Memoria et reminiscentia* , puts down the old waters of Mnemosyne using virtually the same image: "Some men in the presence of considerable stimulus cannot remember owing to disease or age, just as if a stylus or a seal were impressed on flowing water." Martianus Capella, a contemporary of Augustine, goes even further. It is Capella who once and for all replaces the cut stone of a sealing ring with the stylus, the image impressed on the wax of memory by letters traced on an invisible tablet. The three-dimensional pictogram of Classical memory thus appears as the arrangement of logograms on the slate of the mind. Capella's Marriage of Philology to Mercurywas read in the Middle Ages; the monastic curriculum built around the seven liberal arts has been shaped in part by Capella's fanciful summary of antique learning. He served as one of the bridges between Cicero and Alcuin, to Aquinas, over which the conception of memory as a store has reached us.
And while in antiquity this image of memory as an archive referred primarily to a device used by the rhetor, scholasticism made of memory a faculty of every soul, like will and intelligence. Thus, each soul was also burdened with a conscience—a record of its own doings that could be read and examined by clergy and laity, literate and illiterate alike. The rhetorical device provided the foundation for a new activity, confession, the verbal manifestation of a secret kept in one's own heart. And not only deeds left traces that could be admitted; past words and even past thoughts that inspired the deeds could soon be read in an examination of conscience.