Pythagoras the Grammar Teacher and Didymon the Adulterer

a. SA Pag SEDLEY,D. Hyperboreus vol. 4 (1998) Fasc. 1 \a, a & Pythagoras the Grammar Teacher 123 Pora AX Pythagoras the Grammar Teacher and Didymon the Adulterer coating not only provided a suitable surface to write on. but could also be easily scrubbed off and reapplied’. The tablet has a small rounded protrusion at one end with a hole bored in it. Kenyon suggested that this was to allow it to be hung on a nail on the schoolroom wall. If so, the chances are that it was actually hung by a cord passed through the hole, since the hole itself seems too small to fit easily over a nail. [t is not entirely clear what the significance My text is a writing tablet (PBrLibr Add Ms 37516, 1), probably dating from the fourth century AD. It was acquired by the British Museum of this design might be, but there is at all events no necessity to infer early in the present century, its precise Egyptian provenance being unknown. Since its publication in 1909 by F. G. Kenyon!, it has apparently gone altogether unnoticed by historians of ancient beyond a routine piece of schoolwork. Most wooden tablets that have been found have two holes in the left-hand margin, enabling them to be ticd together in the form ofa codex, and this specimen is unusual in having a single hole, But the desirability of the hole, even in a freestanding tablet, is easy enough to explain: for example, without some means of suspending the tablet, it could not easily have been recoated simultaneously on both sides. The hole may have no more significance philosophy, despite the fact that it contains a story about Pythagoras?. It is almost perfectly preserved, and is in fact a rather beautiful object. [t is an approximately rectangular wooden school tablet, written on both sides. Both surfaces were painted with a white coating, and then both horizontal and vertical lines were scored with a pointed implement, freehand rather than ruled, dividing its entire surface into the exact number of boxes needed for the exercise that was to be attempted. Finally, the pupil has written in ink, mainly within the boxes, but occasionally crossing the horizontal lines separating them where they were not sufficiently straight. The tablet has at some point split horizontally and been carefully mended with dowels. It is certain that the tablet had already been repaired before the present text was written on it, both because the pupil’s writing diverges from the horizontal in order to avoid going over the join, and because the vertical line scratched to mark the right-hand margin is continuous across the two rejoined halves, even though they are not quite perfectly aligned. This is testimony to the value placed upon wooden tablets in the schoolroom. A significant number of the school texts found in Egypt are on wood, yet while papyrus was plentiful the country produced very little timber of its own. Hence, once a wooden tablet (known as a Aebkwpa or “whiteboard”) had been cut to shape and planed, it could expect to be recycled many times over. The white © F. G. Kenyon, “Two Greek school-tablets”, JAS 29 (1909) 29-40, pp. 29-30. Kenyon dates the hand to the third sentury AD, but a later dating has been urged on me by Guido Bastianini and Manfredo Manfredi. ? It will, however, be cited by D. H. Fowler in the second edition of his The Mathematics of Plato's Academy (ed. 1, Oxford 1987), to illustrate the tradition of ‘Pythagoras the philosopher’. that the tablet was intended to hold a display text‘, or indeed anything | than that. the of parts d conjugate has student the tablet On one side of the al grammatic a find we now, us concerns which side the On verb vırdtw. tranKenyon’s is Here s. Pythagora about story a on based exercise scription of the complete text’: I 2 6 MvBayópag prióoopos àroBàg Kai ypappata ddaoKwv svveßov- Aevev tolg Exvtod naßmroig Evarpóvwv Anexecdan od TvBayópov ptÀogópov anoßävres Kai ypóppara dtáokovro6 Aöyog Knonvnnovederar GvvPovAEbovtoG TOIG Eavtod pabytaic EVOILOVOV AMEXESBAL 3 4 5 6 | _rà MvOaydpa priooópw Anoßavıı Koi vpéppara fôgorovrt ÉbuËEv govvBovAedoar vois Eavtod pabytats Evayovov Anexesbon tov Modoyöpav pıRöcoyov dropóvra Kai ypäppara ödóokovrá pag svvBovAedcen toig Eavrod padytaic Evorpóvwv anexesdan © TvBáyope piddcoge àroBùg Kai ypopypata diSdcKov OÙ NOTE ovvefovaetoatov toig Eavtod pabrytaic Evayıövav ànéyeobar Kai duxd¢ >See R. Cribiore, Writing, Teachers and Students in Graeco-Roman Egypt (Atlanta 1996), on which I rely for much of my information about school practices and equipment. Our writing tablet is her no. 364, with plate LXI. 4 Cribiore, op. cit., classes the hand as a student’s, not a teacher’s, and in correspondence has reaffirmed and reinforced this judgement. 5 | have added accents, breathings and iota subscripts. In 3 Éboëev is added above the line.

es EN RAN PARTS NE AR RR”, We e a e n hho Bras y Pema nt ey pepe ui spot oH RE A AR E qe x Thin, SENSE TANT

Hyperboreus vol. 4 (1998) Fasc. 1 Adulterer dating ‘<a writing tablet (PBrLibr Add Ms 375 16, 1), probashbly Museum red by the Briti being ry AD. It was acquiEgypt Ron the fourth centu ian provenance nt century, its precise early in the prese by F. G. Kenyon , it has unknown. Since its publication in 1909 by historians of anctent apparently gone altogether unnoticed story Pythagoras | a philosophy, despite the fact that it containsfact a ratheabout object. beaut r in It is almost perfectly preserved, and 1s school tablet, writtiful on en en Itis an approximately rectangular wood white coating, and then both both sides. Both surfaces were painted with a with a pointed implement, d score horizontal and vertical lines were ce into the exact freehand rather than ruled, dividing its entire surfa to be attempted. was that ise exerc the number of boxes needed for the boxes, but n withi y mainl ink, in en Finally, the pupil has writt occasionally crossing the horizontal lines separating them where they were not sufficiently straight. carefully The tablet has at some point split horizontally anddybeen had alrea been repaired mended with dowels. It is certain that the tabletbecau se the pupil’s writing before the present text was written on it, both join, and diverges from the horizontal in order to avoid going over themargi n 1s the right-hand because the vertical line scratched to mark even though they are not continuous across the two rejoined halves, to the value placed upon quite perfectly aligned. This is testimony icant er of the school wooden tablets in the schoolroom. A signif papyrnumb plentiful the was us texts found in Egypt are on wood, yet while a wooden once , Hence country produced very little timber of its own. to shape cut been d”) had tablet (known as a KEVKWLLE OF “whiteboarmany white The over. times and planed, it could expect to be recycled pp. 29-30. ! FG. Kenyon, “Two Greek school-tablets”, JHS 29 (1909) 29-40, urged on been has dating later a but AD, Kenyon dates the hand to the third sentury edi. u Manfr edo me by Guido Bastianini and Manfr 2 It will, however, be cited by D. H. Fowler in the second edition of his The 123 coating not only provided a suitable surface to write on, but could also be easily scrubbed off and reapplied’. The tablet has a small rounded protrusion at one end with a hole bored in it. Kenyon suggested that this was to allow it to be hung ona nail on the schoolroom wall. If so, the chances are that it was actually hung by a cord passed through the hole, since the hole itself seems too small to fit easily over a nail. It is not entirely clear what the significance of this design might be, but there is at all events no necessity to infer that the tablet was intended to hold a display text*, or indeed anything beyond a routine piece of schoolwork. Most wooden tablets that have been found have two holes in the left-hand margin, enabling them to be tied together in the form of a codex, and this specimen is unusual in having a single hole. But the desirability of the hole, even in a freestanding tablet, is easy enough to explain: for example, without some means of suspending the tablet, it could not easily have been recoated simultaneously on both sides. The hole may have no more significance than that. | On one side of the tablet the student has conjugated parts of the verb vıxüo. On the side which concerns us now, we find a grammatical exercise based on a story about Pythagoras. Here is Kenyon’s transcription of the complete text’: 1 6 ITvboryöpas giAdcogas robos Kat ypaupata didcoKwv avveBob - 2 od Ilvdayöpov piaooopou Anoßävres Kai ypappata diddaKavto¢g Aevev toits EXVTOD nadrtoig Evaòvov GTÉYESB OL Aöyog Grouvnnovebetor GUVBPovAEDOVTOS toits éautod uabrtale EVALLOVOYV ATEXECdAL 3 | 1@ Iubayépe posó &roP&vtr Kat ypéuuota S8do0Kovt Ébobev ' cvvBovAedoaı roîç dato Halntats Evarıövov GTÉXET BA 4 tov IIv@aydpav puiócopov ànoBévra kai yodypata è dhokovid _paoiv gvvBovAedonn totg Exvt0d pabntats Évauuévov Amexeodon 5 à [MvOc&yope piócoqe Anoßäg Kai ypdppata Gôdokov ob note ouveBovledoatov tots Exvtod pabntats Évouévov d&méyecar 6 Kai dvuKds | | ’ See R. Cribiore, Writing, Teachers and Students in Graeco-Roman Egypt mr Mathematics of Plato's Academy (ed. |, Oxford 1987), to illustrate the tradition ‘Pythagoras the philosopher’. of (Atlanta 1996), on which I rely for much of my information about school practices and equipment. Our writing tablet is her no. 364, with plate LXI. * Cribiore, op. cit., classes the hand as a student's, not a teacher’s, and in correspondence has reaffirmed and reinforced this judgement. 5 I have added accents, breathings and iota subscripts. In 3 &Sotev is added above the line.

8 Pythagoras the Grammar Teacher tw IlvOayopa piiocopw Anoßävenv kai ypappata ÖLÖÜCKOVTE cuvBovievértnv TOTS éautod LAONTATS evaovav GRÉXESB totv ITvdaryöpoıv puaocépoiv AroPAVTOLW kai Ypeppata SdacKovtoiv MOYOS Anonvnnovederon GuVBOvAEvOVTOLY TOTS EALTOD pabntate EVALLOVaLV] AMEX ECHAL 9 10 11 cotv IIvOayéporv PLAocögYoıv ATOPÁVTOLV Kal YPÓJpOTo LOLI KÓVTOLV ¿Sotev ovvBovkedoat Tots Eavtod Labntats Evarıövalv] TELE Bon a ITvdayspa piiocépo Aroßävrnv kai YPÉLHLOTO SLÖÄTKOVTE PAOLV cvvBoviedoar Tots EXLTIOD HAANTAIG évaruóvov árexecbal à ITvdayöpa: puaooopo Anoßävenv Kal ypappata SaoKovte TOM note cuvBovievcártnv toig Eavtod Ladntats Evaovav anexyedbar | 12 «ai TANOVVTIKOS 13 ot Mobayöpaı piaéoopor Anoßävres Kai Ypappata ÊLÜGOKOVTES ouvepovievdoB8nv tots Éautod PAdnTATS évoiuóvov GTÉXECBOL tv IIvbayop@v pLıAocöpwv Anoßävrov Kat YPALLATA SLOAcKOVTMV Aöyog Anonvnuovederar ouvebovAevoävriov toi Eautod yoBntoie 14 EVOAYLOVOV ATEXETIAL 15 16 17 tots Ilu8ayépais piaocépois Anoßäcı Kai ypappata didacKkovtorg ¿dotev oovPovaAedoar tots Eautod pabntats Evaovav anexecda rodc [IvOaydpac puaocopovus &noPkvtas Kat ypépuata SLöückovräg pac svvpovAEdoat Tots ExvTOD pabNTats Evamıovav anexeodar & Ilvgaryöpaı piAdcogor Anoßävres kai Ypapporo SiddaocKovtes Dele note ovvBovieveniv toig Eavrod HaSntats Évoaiuôvov anéyeo[Oa] For now I shall translate fairly literally, leaving the actual meaning of the story to discuss later. First we get the base version, with a nominative subject: “The philosopher Pythagoras, when he had gone away and was teaching letters, used to advise his own pupils to abstain from blooded creatures”. The student then moves into the genitive: “The saying is remembered of the philosopher Pythagoras, when he had gone away [read äroBévroc] and was teaching letters, as advising his own pupils to abstain from blooded creatures”. There follows the dative, “It seemed good (£öo&ev) to the philosopher Pythagoras...”, the accusative, “They say that the philosopher Pythagoras...”, and even the vocative (5), “O philosopher Pythagoras, when you had gone away and were teaching letters, you once upon a time advised your own pupils to abstain from blooded creatures”. Then in line 6, with the heading kat dvtKdag (misspelt), the student launches into the dual. Thus in line 7 we find “Both the Pythagoras philosophers, when they had both gone away and were both teaching letters, advised his [sic] own pupils to abstain from blooded creatures”. And so on for the oblique cases of the dual, and finally, starting from the heading in line 12, for the entire plural. | ho In I shall come to the nature and purpose ofthe exercise in a minute. But first we might try to assign the student’s work a grade. It is hard to put it much above a bare pass. Although by the standards of comparable output from Egyptian schoolrooms it may seem high-quality work, by any absolute standard it is not at all distinguished. There are a number of grammatical mistakes, some more venial than others. One is a mere lapsus calami: in line 2, &noßa.vteg for «noßäü.vrog; other minor ones include missing augments, and the incorrect vocative of Pythagoras’ name in line 5. A bit worse is the termination of ovveßovAeüchnv in 13, a strange aberration from the correct cuveBovAevocy. In line 5, the student has also invented a vocative for the present participle, d1dacKov, which is not only a non-existent form to the best of my knowledge but also syntactically strained, since a nominative, not a vocative, would make better sense at this point. In the same line there is worse: svveßovAedoortov for what should be cvveBodievoac. Was the pupil so confused in grammatical matters as to think there must be a vocative form of the second person singular of a verb? This suspicion is confirmed in the last line, containing the vocative plural, where again a form seems to have been invented for the second person plural to accompany a vocative — although this time the precise reading is very hard to decipher (no doubt intentionally!). Almost equally significant is the retention of the singular éavtoÿ in the dual and plural versions, which suggests that the student was not fully in control of the passage’s meaning. Prima facie more shocking again is line 15, with its dative diéacKovtotg®, although in reality this may simply reflect the fact that in the spoken language by this date the participle had largely settled down to a single, uninflected form, -ovtac’, making the inflected dative -ovcı unfamiliar except from literary sources. In view of the range of errors, it may even be suspected that Greek was not the pupil’s first language®. This is always a serious possibility in the ethnically mixed population of Egypt. If nothing more exotic, the pupil may have been a Roman. The mediocre standard of his Greek will prove significant shortly, when we turn to another curious feature, namely the recurrently unorthodox order of the anecdote’s first three words. ° Kenyon notes that in the same line &noßäcı has been corrected from something longer, which he suggests may have been anofavtoig. 7 B. G. Mandilaras, The Verb in the Greek Non-literary Papyri (Athens 1973) * Cribiore (op. cit., 93 with n. 176) implicitly draws such an inference, on the basis of the student’s deviant forms for past tenses.

You might think that the pupil has, by way of compensation, at least shown some resourcefulness and linguistic skillin his devices for converting the core anecdote into the oblique cases. Alas, not so. The formulae which he uses for obtaining the oblique forms mechanically follow a rule book. In fact, Egyptian finds have given us at least two schoolbooks of similar date in which these very formulae are listed, as a guide to students carrying out the same exercise’. In both cases the only difference, enforced by the content of the story, is that for the verb of saying, einetv, our student has used the verb of advising, ovpBovNEVELV. But what is the purpose of the exercise? In particular, what merit is there in declining a personal proper name through the dual and plural, when this means inflicting on it forms of such exotic rarity that no pupil was ever likely to have recourse to them? Yet the survival in two schoolbooks of the same list of formulae, with the verb eizetv, shows — as we will shortly see confirmed — that this is an exercise which was standardly used on stories about what some named individual said. There is, moreover, one surviving grammatical exercise from Egypt where the names Priam and Hecuba are declined through singular, dual and plural!°. So the Pythagoras version cannot be dismissed as a case of a standard exercise comically misapplied to an inappropriate sample sentence. In fact, though, we can go much further than that. The exercise on the tablet is a well-documented one, which for at least five centuries was a Standard element in the progymnasmata, the training which preceded an education in rhetoric proper. The pupil who wrote on our tablet will have been in the second main stage of the educational curriculum. Probably a boy in his early teens, he will have been attending the classes of a Ypauatikôc or grammar teacher. Our earliest detailed source for this part of the curriculum is Aelius Theon of Alexandria, whose Progymnasmata is tentatively dated to thefirst century AD. The same exercise is briefly described by Quintilian!!, so we need have no II. 96-106 Spengel) explains in some detail how chreiai are used as an educational tool, in fact as an introductory exercise at the very start of the pre-rhetorical training. Pupils can be set a specific range of tasks all involvingthe manipulation of chreiai. They may be asked, for example, to recite a chreia, to expand a chreia into a full narrative, to reduce a full doubt that it was a standard fixture by the end of the first century AD. It still features in the Progymnasmata of Nicolaus of Myra as late as the and-so the action has achieved memory (10 mpay6év uvrung ÉTUXE)... fifth century AD. A chreia is a one-sentence anecdote about some specific individual achieved memory (tó ouupüv kvnung Eruxe)... recording a bon mot or memorable happening. Theon (Progymnasmata if it is an unprompted saying, you write “Of X the so-and-so the dictum has achieved memory (tó pnôëv uvnung Eruxe)...”; but if,in accordance with a further subdivision of the “saying” chreiai, it is a response to a question, to an enquiry or to some non-interrogative utterance, it is considered better style to write “Of X the so-and-so the saying is remembered (0 Aöyog árouvnuoveveral)”. To a very large extent, then, the chreiai seem to have been classified with the variants of the ? Kenyon, op. cit., 38; P. J. Parsons, “A school-book from the Sayce Collection”, ZPE 6 (1970) 133-49, at pp. 143-4. No. 372 Cribiore. "Inst. or. 1.9.5, 127 narrative to a chreia, to cap a chreia with an explanation, and even to criticise or refute a chreia. But especially prominent is the ritual called «Mo1c, in which a chreia is declined through all the cases and numbers, using a prescribed set of formulae which differ only marginally from the ones we have already met in the tablet. It is an exercise which Theon illustrates at some length, and without the least sign of embarrassment. Thus in Theon we get just the same kind of bizarre contortions as in the tablet. From the chreia “Isocrates the orator used to say that his gifted pupils were children of gods” we are presented with such bizarre derivatives as the dual “Both Isocrateses, that pair of orators, said that both their talented pupils were a pair of children of gods”. One sign of the special importance attached to the declension of the chreiais the following. Theon and other progymnasmata writers offer a classification of chreiai. The basic division is between the “saying” ones (Aoyırat), the “doing” ones (rpakrıkat), and the “mixed” ones which combine a saying and an action. The doing ones are further subdivided into “active” and “passive”. An example of a “saying” chreiais “Isocrates the orator used to say that his gifted pupils were children of gods”, and an active “doing” chreiais “Diogenes the Cynic philosopher, seeing a boy overeating, struck the boy’s paidagogos with his stick”, whereas a passive “doing” chreia ends up describing what befell the subject, not what he or she did (we will be meeting one of these later). Now these distinctions give every impression of having been dictated by the needs of the declension exercise, since the formulae listed have to be specific to the class of chreia selected. For example, the genitive version of a “doing” chreia, if it is active, is “Of X the sowhile ifit is passive you write “Of X the so-and-so what befell him has Ina “saying” chreia, on the other hand, the genitive formis attained by one of two formulae:

declension exercise in mind. This tends to confirm the central importance attached to the exercise. | | Looking back to our Egyptian teenager, we can now see grounds for lowering his grade a bit further. Since what Pythagoras said was not a response to something somebody else said, it would have been more correct in the genitive version to use tó pndev uvnung Eruxe than the phrase he actually opts for, (0) Aoyos árrouvnuovedezar!?. Also slightly worrying is whether this should count as a “saying” chreia at all. Its main verb ovvefobievev might rather be classed asa “doing” verb, so that it would be his deed (rpay0év), not his saying (A6yoc), that had been remembered. But, although the verb cuveBobAeve occurs in other chreiai too (e. g. DL I. 92), even Theon’s rather detailed rules offer no guidance on how to classify it; so no further marks should be deducted. What exactly is the point of this declension exercise? We are told that some considered it an ideal exercise for pupils making the transition from study of the poets to preparation for rhetoric!*. I am not at all sure why. Declining a chreia through the oblique cases in the singular does of course make sense, since it trains the speaker to introduce chreiai into a discourse with a satisfying range of syntactic variatio. And even though the pluralisation of personal proper names is a rarity, there might occasionally be chreiai with authentically plural subjects, whose “Isocrates the orator...”, “Plato the philosopher”, “Alexan der the king of the Macedonians”, and, in one case, “Pythagoras the philosop her...” So the opening of our chreia must be meant in the same way. The word order there is certainly abnormal — 6 Ilv@ayopag piAdécop oc, where [lv@ayopag 6 puaécopos would be the regular form. But in view of the standard format used in chreiai, this must be taken as meaning “Pythagoras the philosopher”, and we must resist any temptati on to manipulation could be facilitated by training in the declension ofa chreia through the plural. On the other hand, no one in the post-clasical era was likely ever to have to call on a dual, other than for purposes of recognition in the reading of classical texts. I find it hard not to get the feeling that, while the other exercises really were meant to help students master the chreia and make it do their bidding in the construction of.a rhetorical discourse, the declension exercise goes further than that. It is as if it were designed to show who is boss, not by merely mastering the chreia, but by humiliating it Into total submission. The popularity of this particular exercise will owe less to its usefulness than to the pride that could be taken in a display of sheer grammatical virtuosity". In the course of his explanation, Theon lists many examples of chreiai. Virtually all take the standard form ofa single-sentence complete anecdote, beginning “X the so-and-so...”, e. g. “Diogenes the Cynic...”, construe the sentence differently’. Here it is very importan t to notice that the student originally started by writing the name TTu@ayd only subsequently went back and added the article 6 in the margin, where it is in fact so far to the left that it has been '3 Nicolaus, Progymn. 18. 2 ff. Felten. | ‘* For the prestige attached to grammatical attainments, see Teresa Morgan, Literate Education in the Hellenistic and Roman Worlds (Cambridge 1998, forthcoming) chapter 5, “Grammar and the power of language”. pac, and left hand written on the sloping edge of the tablet, beyond the flat whitened area. This strengthens the likelihood that the chreia was originally assigned to him with the canonical form of opening, ITvdayöpag $ PLAOGogos... reordering, the simplest hypothesis for explaining it is that omitted the article by mere error, then added it in the left As for his he ori ginally margin, and thereafter for consistency retained in the oblique cases, the dual and the plural the word order that this had produced in his opening line. In view of what we have seen to be the mediocre standard ofhis Greek, he may have been totally unaware how unidiomatic the new word-or der wasté ___Nowat last we can turn to the chreia itself. Why this story? Where did it come from? Is it a genuine testimonium on Pythagoras? The story is no more likely to contain an actual historic al truth than most of the thousand or more other chreiai have come down to us or for that matter than the great majority of other Pythagoras testimon ia, most ofthem little more than manufactured folklore. But we ask whether, almost equally importantly, the anecdote is do need to evidence of the Pythagorean legend, as this developed among Pythago reans from the fourth century BC to the fourth century AD. As I remarke d earlier, ‘’ For example by taking piLócopoc Gros together and translating “Pythagoras, having turned out to be a philosopher...”, even though that would in itself be perfectly possible Greek. I owe to Manfredo Manfredi the observation that it could in principle be construed this way. The other possibility is that, whether or not so bidden by his teacher, after had started transcribing the first line he decided to move * For this same criticism, see already Parsons, op. cit., 144 n. 30. 129 he the definite article to the beginning of the sentence in order to mark out the sequence of cases and numbers down the left-hand side with complete clarity: the layout of the tablet itself tends to give the impression that, after he had finished writing the first line, he mapped out the remaining singular forms by adding the entire sequence Tod, tà, tov, © down the the left-hand side (three of these four are followed by an unexpected amount of space), before the remainder of each line was added in.

it does not seem to have been taken into consideration by any modern iscussion of the history of Pythagoreani | “seus fact, I have tracked down only one modem source Which even had disembarked and was teaching writings, used to counsel his students ical report about Pyth the legend. But the story of unparallelled. Nor for one expects ofthe Pyt lowly profession in the anci “blooded” for “ensouled [taly is the place with which he is most closely associated. nee Oy The standard expression now 7 hi ua imparting it. He began to teach only after his arrival in A ne other Den we to explain these oddi known be on shipboard. In a continuous narrative that would be fine. Sui chela! are not mere decontextualised quotations; they are, above alle se se contained. Nor ts it at all plausible that the reference to Yp&unoo i ws diti that Pythagoras wro ‘ote treatises;5 if e tradition there were such sucl ft vofetenee we would at least expect the definite article, and more probably the standard term cvyypauuata than the simple YPOppacta. (a rare usage, cf. SE M I. 48). Nor does either word, used in this sense, ma ea very plausible object for the verb d1daoxetv. | think we can do better than this. | BE First, it is hardly necessary to point out that PARENT DAO KEL y has an absolutely standard meaning, » {each Smart Où mare 1001-te ‘°. I find | it hard to to doubt do 1 that t | nerally, Ñ “to be a school-teacher”. srammar-teachet who set our student the chreia exercise wasIn ly aware ai own| pro profession. The opening part of the chreia, f this allusion to his part rela then means quite simply that Pythagoras the philosopher went off (from his philosophical school, one assumes) and became a schoolteacher, o more specifically a grammar-teacher. It then continues that, in this new life, he used to advise his pupils to practise vegetarianism. Y The Chreia in Ancient Rhetoric | (Atlanta 1986) 335-6. the variant “to chreia. How are | think, is that in évoruóvoy OnevyecOar oc) is a hapax legomenon , whose unique LX p. 194. 22 Littré). There nderstood only in | al aly in a context w here Pythagoras ts already tures” seems unique to this the adjective évaipov (-ov occurrence is in the This is not a bad try. Nor on the other hand does it Carry much conviction. If &noßüc meant “having “isemberked n cole Ni so is the use of an vegetarianism. in the Pythagoras biograph ies is EUWÜXDV ties? Diogenes Laertius 8. 9, where Diogenes makes a connection Pythagoras’ writings and his teaching. ent world. Another oddity a rather The crucial thing to notice, some traditions he Is supposed to have composed). See, “8 mp bet y as àrÉxE Bon, “to abstain fro m ensouled creatures”, and expression here which requires comment is YPérLotor ou ve ahh h may be that ypàppata here refers to Pythagoras’ own NES nd as reverential a stor schoolteaching was ” with regard to Pythagore abstain from blooded crea part of eaching is entirely that matter does it sou arobác is clearly intended to refer to Pythagoras’ arrival n pani dy and Crotona, it had been as one who was anne agoras, and a central him switching to grammar-t hagorean hagiography: although he had “disembarked” in many other places before rea " E | > 131 But where could this fit into the biographical trad ition? Advising his pupils — his philosoph y pupils, that is — to practise vegetarianism is a standard biograph | E. N N. discusses the meaning of the c hreia. R. F. Hock andI Ne | ing tr ion: “Pythagoras the philosopher, | ffer the following translation: 1a ne o to abstain from red meat.” They add the following comment: Teacher Hippocratic treatise On the nature of bon es (c. 19. however it refers, not to “bl ooded” creatures, but to “blood-bearing” vein s (Évatuovec sc. pAËPES). Everywhere else the term is not évatuov, thir d declension, but EVOLLOS, second declension. Now the medical lexicogr aphers, Starting with the great Alexandrian phys ician Herophilus in the pupils of the third cent ury BC, wrote numerous handbooks of Hippocratic Gre ek! Our main surviv ing specim en of the genre is the firs Erotian. Erotian can reel genre, both doctors and t-century AD Hippocratic off a long list of his own lexicon of predecessors in the grammarians. We lear n from him that, amo ng many other concerns, Hippocratic lexicogra phy was interested in collecting Hippocra tic hapax legomena!®, Mor bones is one treatise which eover, On the nature of the lexicographers definite ly covered, since Erotian quotes a citation of it from the first-genera tion Herophilean Bacchius?, Hence it is in principle likely that the Hipp ocra tic lexicographers noticed and commented on the VEC. hapax legomenon ÉVO ULO - This leads me to my sug geston about the origin of the Pythagoras chreia. The unique form Evaiwoves will have bee n identified by the lexicographers as an irre gular one, which should be avoided in favour of the standard Évauuor. If Pythagoras had been a grammar teacher, they may have quipped, he would have advised his pupils to “abstain Eee See H. von Staden, Herophilus (Cambridge 1989), part Il “The Herophileans”. ” Erotian, Gloss. p. 8. 13 Nach manson. * Bacchius fr. 38 von Staden (op. cit., 497) = Erotian e. 39, p. 38 Nachmanson.

from évatpovec”. This then became the anecdote on our writing tablet, which I would now translate as follows: “Pythagoras the philosopher went off and became a grammar teacher, and used to advise his pupils to abstain from évoiuovec” (1. e. from the word Evaioveg)?. That is my suggestion for the origin of the chreia. It is a grammarians’ in-house joke. But I can see at least four possible objections. (1) I have myself remarked that a chreia must be self-contained. Yet this one, thus interpreted, certainly needs some explaining. (2) If itis a joke from Hippocratic lexicography, could it still count as a chreia at all, as it has to do if it is to be chosen for the exercise of declining a chreia? (3) Did ancient grammarians make jokes, or even have a sense of humour? (As opposed to the countless jokes about them, that is.) (4) Can we allow the expression Évainôvov anéxec@at, “abstain from blooded creatures”, to mean “abstain from the word évotuovec”? This is a violation of the use-mention distinction — that is, in this case, the distinction between mentioning the word évaipovec, which is what the What does this mean? Theon, in describing the students’ exercise which consisted in criticising chreiai, lists this particular chreia as one Which deserves to be criticised, for its obscurity (104. 22-3). Nevertheless, he himself elsewhere uses the very same chreia to illustrate the declension exercise. That already deals with my first objection: a chreia which was acknowledged to be obscure could, on occasion, be used as material for the declension exercise. What does the obscure Didymon story actually mean? The explanation is that Sióvyuos can mean “testicles”. Being hung “by the name” means that Didymon was hung by the testicles, ék OLÓVLOV. There is one other version of this chreia extant, recorded by Diogenes Laertius (VI. 51) among the innumerable chreiai about Diogenes the Cynic: chreia means to do if I am right, and actually using the word, as it In fact does. The standard Greek formula for mentioning the word without using it would here be tod évatpovec onexeodon. That would have made the meaning clear, but also of course ruined the joke. The problem that faces us, however, is the converse one: could the simple &vaııövov anéxeo8or have been recognised by a Greek reader as referring to abstention from the word, not from the things"? | All four objections find rather satisfying answers in the text of Theon. Specifically, help comes from another chreia, which can be seen as the joker in Theon’s own pack (and the pun on joker IS intended). Theon several times quotes as an example the following chreia?: Artóvducov* è adANTAS AAOLS Ext poryeta EK TOD dvdpatog éxpeudo0n. Didymon the aulos-player, caught committing adultery, was hung by the name. * For a parallel joke about a celebrity who, forced to become a schoolteacher, carried on in his old habits, cf. Cic. Tusc. II. 27: Dionysius the tyrant, expelled from Syracuse, went to Corinth to become a schoolteacher, so addicted was he to wielding power. | u 133 Diogenes, hearing that Didymon the adulterer had been arrested, said “He deserves to be hung by the name”. This 1s actually a rather better version, since as far as I am aware being hung by the testicles was never an actual punishment for adultery anywhere in the ancient world. Hence itis more appropriate for someone in the chreia to comment that the adulterer deserves this punishment than for the chreia itself to record that he received it. This, then, may well be the original form of the chreia. Like all chreiai, it will have been designed for attachment to the name of some celebrity. The reduced version quoted by Theon, in which Didy mon himself replaces Diogenes as subject, seems to be accounted for by the need, for the sake of completeness, to include an example of a “doing” chreia which was passive in content, describing not what someone did but what happened to him. The Didymon story is, at any rate, Theon’s only proffered example of a passive chreia?. From its association with Diogenes we canno t infer anything about the original date of the chreia. Chreiai continued to be manufactured throughout the Hellenistic era, to satisfy an apparently insatiable market for anthologies of them. What is more, a chreia could readily travel from one historical figure to another, and Dio genes the Cynic in particular o * On the problem whether the neuter article may ever be legitimately omitted in such cases, cf. V. Bers, Speech in speech: studies in incorporated ‘oratio recta’ in so rare a name in antiquity (I have Attic drama and oratory (Lanham, MD 1997) 56. Rhodes) as to be almost certainly invented 2 E. g. Theon, Prog. 102. 34-5. 4 Aróduov is his name in DL VI. 51 and 68 (see below); most MSS of Theon have Aiôvuoc, but for the evidence supporting restoration of the obviously superior termination -wv see Hock and O’Neill, op. cit., 88 ap. crit. Didymon is, inciden tally, so far tracked down only two instanc es, both in °° Theon, Prog. 99. 2-4 Spengel. for the purposes of the chreia. “ Cf. J. F. Kindstrand, “Diogenes Laertius and 7 (1986) 217-43. the ‘Chreia’ tradition”, Elenchos

proved to be a magnet to whom stories of slightly scandalous witticisms would stick, much as in another age happened to Maurice Bowra. There is therefore not even any particular reason to assume that Diogenes was the original subject of this chreia. At all events, one reason for doubting a very early origin for this chreia, in either of its two forms, is that the use of 616 v.01 for “testicles” is post-classical, its earliest firmly datable literary occurrence being in a Philodemus epigram (AP. V 125. 6) from the first century BC. But an occurrence certainly no later than this, and probably two centuries older, is to be found in the Septuagint, translating an injunction of Moses at Deuteronomy 25. 11 — a fact which suggests that, at least in the (probably) third-century BC Alexandrian context in which Deuteronomy was translated, St9vuo1 was fashionable as a serious term for “testicles”, and not a scurrilous, colloquial or jocular usage. This evidence from the Septuagint fits rather neatly with a further ‘tem of information. Galen reports the same employment of didvpot, as an for what the mainstream medical tradition always called öpxeıg, an Staden), von 109 (fr. s Herophilu doctor the of usage tic idiosyncra enormously influential figure in third-century BC Alexandria”. Whether Herophilus’ introduction of $tòvpor to medical terminology merely reflected a local Alexandrian usage, or actually created that usage, must be a matter for speculation. But that he was making a serious contribution to medical jargon is confirmed by the fact that, even if the actual term did not catch on in later medical usage, it did leave its imprint by fathering the further term “epididymis” (used by Galen), for the adjunct of the testicle still today known by that name. There is, then, at least a fair chance that the original inspiration of this anecdote lies once again in medical lexicography. As I have said, many Herophileans wrote handbooks to Hippocratic terminology. What is more, in some cases these handbooks are reported as comparing Herophilus’ own usage with Hippocratic usage (fr. 267 a, cf. 268 von Staden). There is therefore good reason to guess that the Didymon chreia arose as another grammarian’s joke, springing from the tradition of medical lexicography. And if so we would have, in answer to my second difficulty, confirmation that the chreiai used in the declension exercise could indeed include anecdotes based on jokes about medical lexicography. The further fact that in the original version of the chreia Pythagoras the Grammar Teacher 135 Didymon seems to have been a doctor? may be a trace of this same origin: who could be a more appropriate butt for a witticism about medical terminology than a doctor? We do now at any rate have a probable answer to my third question. Yes: off duty, it seems that grammarians really did enjoy the occasional joke, even though they ruthlessly excluded any hint of humour from their formal grammatical treatises. And, returning to our school tablet, if an Egyptian grammar-teacher, in setting his pupils moral chreiai as the basis for the standard declension exercise, occasionally indulged himself by smuggling in one which was a vehicle for an in-house grammatical joke, rather than the usual bon mot, that need not altogether surprise us. What about my final listed problem, concerning violation of the use-mention distinction? Satisfyingly, the Didymon story contains a parallel violation. While the Pythagoras story seems to refer to the thing in place of the name (“abstain from blooded creatures” means “abstain from the word ‘évainovec”” — I. e. use in place of mention), similarly, though conversely, the Didymon story refers to the name in place of the thing: he “was hung by the name” (i. e. mention in place of use). This admittedly stretched usage (no pun intended) thus seems to be an established part of the same grammarians’ game. It appears, then, that some grammarians enjoyed consciously playing such use-mention games. | | Jokes ofthis kind are certainly well attested in the ancent collections of sophisms. For example (DL VII. 187), one standard sophism is designed to show that you have a big mouth: “If you utter something, it goes through your mouth; but you utter a waggon [Guagav dè Aadetc: note the case]; therefore a waggon [Gago nominative] goes through your mouth”. Clement (Strom. VIIL 9. 26. 5) knows a version where what goes through your mouth ts a house. Seneca (Ep. 48. 6) quotes with palpable contempt another sophism from the same family: “Mouse is a syllable. A mouse eats cheese. Therefore a syllable eats cheese”. These examples belong primarily to the discipline of dialectic rather than to that of grammar. But also among the many jokes about grammarians which have survived there is at least one which trades similarly on the use-mention distinction. An epigram of Palladas (fourth century AD) consoles a donkey who has just become the property of a 27 Here Herophilus is using the term for the female “testicles”, namely the ovaries. But there is no reason to doubt that he, like others, used a single term for both the ailment, and Diogenes said “Mind that in treating the maiden’s eye you don’t corrupt male and the female varieties. the xôon (‘pupil’, but also “girl”)”. # DL VI. 68: Didymon the adulterer was treating (iatpevovta) a girl’s eye

grammarian (Anth. Pal. XI. 383). Alluding to the grammarians’ regular citation of the Homeric «pt, used in place of epi01, “barley”, to exemplify the device of apokope, Palladas points out to the beast that grammarians offer “endless barley’”’. As well as being an epigrammatist, Palladas was an Alexandrian schoolteacher. Following in the same tradition as both the original author of the Pythagoras chreia and the grammar teacher who set it as a school exercise, he too proves to be a ypaupatırög prepared to indulgein linguistic jokes about the practices of his own profession. Philo’s usage elsewhere?!. Thus in Theon’s own immediate Alexandrian context, although nowhere else that | know of, the expression “abstain from blooded creatures” was immediately recognised as a description of vegetarianism. Without that recognition, the joke underlying the chreia would have had even less chance of being understood. In the apparent use of local parlance, we can perhaps begin to glimpse why this particular chreia about Pythagoras the grammar teacher has been lost to our mainstream literary traditions, and has survived uniquely through an exercise which, some three centuries later, was set as a routine class assignment by an Egyptian grammar-teacher’’. 136 I conclude, then, that the Pythagoras anecdote originated as a joke from medical lexicography: and, sadly, that it is not a genuine testimonium on Pythagoras, or even on the Pythagorean tradition. Scholars have, it turns out, been quite right to ignore this anecdote in constructing their histories of Pythagoreanism. But I doubt ifthey can be said to have ignored it for the right reasons, or even to have known that it was available for them to ignore. And now for a closing thought. So closely do the profiles of the Pythagoras chreia and the Didymon chreia match, that one might hazard the guess that in the pool of chreiai for pre-rhetorical exercises they were treated as a linked pair. Since the Didymon one, being already 137 David Sedley Christ's College, Cambridge B cTaTbe AETANbHO O6OCHOBBIBAETCH, MOYEMY mpouckonamaa 13 Erunta AepeBAHHan TaGIM4Ka CO LIKONBHLIMH YIPAPCHEHHAMH, B KOTOPOH YIIOMAHYT Iubarop n ero HaCTABNeHAA, HE MOXET paccMaTpuBaTbca HA Kak CBHACTENLfound in Theon of Alexandria, is probably not datable later than the CTBO 06 YUEHHH MOCNEAHETO, HH KAK CIEL PA3BHTHA JereHKapHoli TpanMunu first century AD, we might tentatively infer a similar terminus ante quem for the Pythagoras one, even though its actual attestation, in our writing tablet, belongs to the fourth century. Now a curious coincidence presents itself. The one available parallel for the expression “abstain from blooded creatures” ts found in the writings of Theon’s Alexandrian contemporary Philo. In De vita contemplativa (73%, Philo’s description of an ascetic diet includes: water rather than wine, and “a table pure of blooded creatures (tp&meCa Kabapa tHv éÉvatuwv), on which bread is the nourishment and salt the garnish”. Since bread is the only nourishment, it is clear that “pure of Evan” does not mean that only kosher meat is served, or for that matter that no red meat is served, but rather that no meat at all is on the table. This is in fact confirmed by o Hem. TekcT Ta6JIMUKU TOBOPHT O TOM, UTO, MEPECTAB 3AHUMATECA (DMSIOCOduet u OGpaTHBLUCE K IIPETIONABAHHIO CIIOBECHOCTH, IIud@arop yann Bo31ep- JKUBATECA OT “KPOBABOË” num (Évou1évov). Dpaza MOBTOPeHa yueHuKom, rpeuecknli A3bIK IIS KOTOPOTO efBa JIM GLi POAHBIM, 17 pa3 c HCTONL30Ba- *' Esp. De spec. leg. 1.255, à àrapyi TPÉTOVOO. tepedov GT OVÓEVOS TV EVALLOV, GAA ATO TOD Kado.peté.tofs |THC dvepantvnc zpopfig. As far as I have been able to establish from the TLG, in later antiquity the term is used this way only by writers directly echoing the passage fram Philo, Vit. cont.: Eusebius, Hist. eccl. II. 17, 22, and Sozomenus, Hist. eccl. 1. 12. 10, in the latter of which, uniquely, the actual expression évotuov anéyecbar occurs. 2 I am grateful to a number of people who commented on a version of this paper which was read to the Cambridge papyrological seminar “Reputations in tatters” # See Ineke Sluiter, Ancient Grammar in Context (Amsterdam 1990) 36, and, Easter Term 1998: to Colin Austin, David Blank, James Clackson, Sarah Clackson, for a full survey of ancient ridicule of grammarians, her article “Perversa subtilitas: Janet Fairweather, David Fowler, Eric Handley, Teresa Morgan and the editorial committee of the Corpus dei papirifilosofici greci e latini for helpful suggestions on individual points; and to Ineke Sluiter for very constructive comments and suggestions de kwade roep van de grammaticus”, Lampas 21 (1988) 41-65. Y T owe this reference to Alessandro Linguiti’s preparatory work for an edition of the Pythagoras testimonium (to appear in Corpus dei papiri filosofici greci e latini on a subsequent draft. A shorter version of this paper is appearing in Papirifilosofici: vol. [***), where he also makes the interesting suggestion that the term may reflect miscellanea di studi II (Accademia toscana di scienze e lettere “La Colombaria”, Philo’s Judaeo-Christian context. Florence 1998).

Huem umeHu [ludaropa BO BCEX BO3MOXHbIX UNCIIAX U Malexax H C COOTBETCTBYHOLUMMU H3MEHEHAAMH CHHTAKCHCA BCETO [IPEMIOKEHUI. OUEBAAHO, IIEpe. HAMH yMpaxHeHne, KOTOPOE PEKOMCHAOBAJIOCh ABTOPAMH PETOPAUECKAX YYEÖHHKOB INVIA MOAPOTOBHTENBHOFO Kypca, U B HEM, Kak YACTO OBIBAJIO, HCrIO/Ib30BaHa XPHA — KOPOTKHM AHEKAOT O 3HAMCHHTOM HEJIOBEKE. Hala Xpua COAEPXHMT urpy cnoB: IIndarop 3arıpeiuaet ye He YIIOTPEÖJIEHHE B TIHLLLY MAca, HO, kak 4 TIOAO6aET FPAMMATUKY, yroTpebieHne POPMBI Evaıımv BMECTO Gosee OÖbIUHOTO Évo.uos. PEAKOE COBO YKA3bIBACT, BO3MOXHO, HA CBA3b XPUH c anekcanapulickof nexcukorpapueli, 3a@ukcupoBaBlieñ BCTpeuamıeeca B "IT unnokpaToBoM Kkopnyce” évaipwv Kak OTCTYILIEHHE OT HOPMBI. Kak noKA3bIBACT AHEKAOT, MOCBALUEHHbIM HEKOEMY MIpenoböner Anınmy, rpamma- TUKU Mopol MO3BONANM cede LIYTKH Hal COÖCTBEHHOÄ Mpomeccueli, u M3 CO6paHua xpul Takue ACTOPAH TOTAHANH B LIKONBHYIO HIPAKTUKY. BepoaTHo, NOAOÖHLIÄ nyTb npozesana u xpua o IImbarope.