On the Alleged Three Mediaeval Metagrammatisms Young, Douglas Greek, Roman and Byzantine Studies; Spring 1966; 7, 1; ProQuest pg. 5

Notes on the Text of Pindar 

Douglas Young 

I 

On the Alleged Three Mediaeval Metagrammatisms 

I N HIS Histoire du texte de Pindare (Paris 1952), J. Irigoin consecrates 
a chapter (XI, pp. 123-134) to his theory that the text of the odes 
underwent in the Middle Ages three successive transliterations 

or metagrammatisms from uncials to minuscules. If true, this 
would be most unusual for a pagan author, for a pagan text normally 
was transmitted through a single metagrammatism. Irigoin writes 
(p. 125): "Les fautes graphiques causees par la confusion des lettres 
onciales ne sont jamais nombreuses. Pour la recension ambrosienne, 
dans Ie texte des Olympiens, une faute est certaine: 0 10.51 EACAC: 
EACAC: une autre faute se retrouve dans quelques manuscrits de 
la recension vaticane: 0 6.180 AE'Z': l::J.E'Z': c' est probablement, dans 
ces manuscrits, une correction independante de la recension ambro­
sienne; la priere de demande (Oloot) se termine par l'ofIrande de l'ode 
(€j1lvv 8' vJLvwv 8Eg' EVTEP7TES av8os)." (Note that Irigoin cites by cola: 
the Bowra line references are 0 10.43 and 06.105). 

Now the corruption of aEg' to 8'g' at 0 6.105 occurs not only in A, the 
leading manuscript of the so-called "Ambrosian" recension, but in 
L M Nand 0 post corr. Certainly the confusion of lambda and delta is a 
common uncial error; but it can also occur in a purely minuscule am­
bience, e.g. at Theognis 847, where N writes 8ag for the word Aag in its 
minuscule antigraph, D. In that place there may be some unconscious 
association of thought, namely rhyming, an occasional source of error 
in copying. Irigoin himself notes (p. 125) that the scribes' sense of the 
context may have influenced the corruption of aEf to 8'f at 0 6.105. 
The other corruption on which he founds, that of El\uas to eauas at 
o 10.43, need not have arisen through visual confusion of uncial letters : 

5 



6 NOTES ON THE TEXT OF PINDAR 

it could spring from mere trivialisation, the substitution of a com­
moner word of similar appearance. Collation of the Pindar text in A 
shows that it is frequently a careless apograph: e.g. at 0 4.8 A writes 
the trivial i7T7TOV for 11Tov; at 06.91 ac/>06yywv for ayac/>O€YK'TWV; at 0 7.59 

X>..wpav for xwpa~. At 0 10.98 A offers ava1T>"aO'O'Et for ava1TaO'O'Et, and 
such an error would more readily arise in uncials; but even so it would 
not help towards proving a separate transliteration for A, because five 
other manuscripts offer the same error. 

It seems, then, that there is no adequate basis for Irigoin's theory 
that A is the product of a separate and belated transliteration from 
uncials to minuscules in the later thirteenth century (p. 246). Irigoin 
writes of "L'abondance, dans Ie manuscrit A, des fautes dues a la con­
fusion des lettres onciales ... ", though he had previously written 
(p. 125) that "Les fautes graphiques causees par la confusion des lettres 
onciales ne sont jamais nombreuses." It is a pity that Irigoin did not 
list the allegedly abundant uncial errors he found in A, which I myself 
cannot find in the text. Irigoin does not even list many errors from 
the scholia of A, and he himself admits (p. 125): "mais on sait que, 
jusqu'a la fin du Xe siecle, les scholies ont ete volontiers ecrites en 
petite onciale." Thus uncial errors in scholia, in A or other Mss, can 
derive from eleventh or twelfth century scribal work on Mss all de­
rived from a single transliteration of the text and scholia in the later 
stages of the Photian renaissance, say around A.D. 1000. 

It is of dubious help to Irigoin's thesis to remark (pp. 246f) that the 
Theocritus text K, bound with A of Pindar, derives from a separate 
transliteration: " ... Ie meme Ambrosianus C 222 inf., avec Ie sigle 
K, tient une place analogue dans l'histoire du texte de Theocrite; il 
remonte a une translitteration distincte ... " lrigoin refers to C. 
Gallavotti, Theocritus, pp. 243-245. Gallavotti remarks that the hand of 
Theocritus K in the volume is that of the accompanying Aristophanes 
text, but different from that of Pindar A. Gallavotti gives only three 
readings suggestive of uncial errors in K. One is Theocr. 15.68: 8!-'wa 
for a!-'wv, which could originate from uncial AMWN being misread as 
AMWA, by a double error. But one should note that the Theocritus 
Mss G and P have the reading D!-'wI,s; and both this and K's Dp.wa could 
have arisen from a gloss (meaning 'servant-girl') on the preceding 
word EVv6a, which gloss had supplanted &'JL(;JV in later copying. In 
Theocritus epigram 11.4 K's error &:>..tp.wvw~ for the correct Datp.ovtws is 
shared with the Perusine Mss and thus cannot prove a separate meta-



DOUGLAS YOUNG 7 

grammatism of K. Thirdly, Gallavotti cites from Simias Securis 6 

K's variant OVa'YJAE~~ for ovaKA~~. This could arise from an uncial con­
fusion of eta and kappa, but confusions occur also of minuscule forms 
of these letters. In this connection one may note that at Pindar 0 6.98, 

for firr}paTOL~ we find fVKpaTOL~ in A, and in Nand 0 by their first 
hands. In A it is a yp( a~ET(XL) variant. One must reckon with the possi­
bility that the transliterated archetype, or master-copy, of all our ex­
tant Pindar Mss had some variants in its text, apart fom those in its 
scholia. Even were the thesis valid that K of Theocritus derived from 
a separate metagrammatism, substantially more evidence would still 
be needed to make attractive Irigoin's theory that A of Pindar 
stemmed from a late thirteenth century metagrammatism. 

Irigoin's contention receives no support from the Aeschylus text (A) 
that accompanies K ofTheocritus and A of Pin dar. Study of the beauti­
ful collation by Dr Roger D. Dawe shows no sign of separate meta­
grammatism of Aeschylus's A, but reveals some indications of its 
antigraph's having been minuscule: e.g. at Septem 759, for 7TlTVOV 
A offers 7TlTVOV, and at Persae 922 for eyyalav one finds ayyalav in A 
and three other Mss. 

Irigoin's case for there having been a third mediaeval transliteration 
of Pindar, that of his "recension vaticane raccourcie," (p. 127), seems 
also to be inadequate. Apart from two examples in the scholia, which, 
as said above, are not helpful because scholia continued to be written 
in uncials after the text had been transliterated into minuscules, 
Irigoin cites only one text variant to prove his thesis: P 3.27 P:'lAOOOKCP] 
f1-710000KCP' This is in C ante corr. E G V2. It might be an old uncial 
error that had been corrected in the rest of the family ofMss descended 
from the single metacharacterized copy of the Photian renaissance; 
but it might also exemplify a common type of error in minuscule 
copying, whereby a scribe anticipates a syllable, here writing in second 
place the syllable 00 due to come in third place. Alternatively, the 
repetition of the vowel omicron brought with it a repetition of the 
second consonant accompanying it: schematically C1 V1C2 Vl> 
C2V1C2Vl . Or again, to pursue Irigoin's own argument elsewhere 
(p. 125) about the influence of context on the scribe's mind, note that 
in P 3.27 the next word is IIv(}wvt, so that a scribe might well have 
written f1-710006KCP through some latent notion of the Medes at Delphi, 
as described by Herodotus. At any rate, the miswriting of f1-71A006K<p as 
f1-710006K<p will not go far to establish Irigoin's thesis of a third mediaeval 



8 NOTES ON THE TEXT OF PINDAR 

metagrammatism of Pindar. In general, to demonstrate more than a 
single transliteration from uncials to minuscules of any classical text 
one would require to see a sizeable constellation of indubitably uncial 
errors in the text; and for Pindar none such is visible. 

Irigoin further theorizes that there was an ancient metagrammatism 
of Pindar into the Ionic alphabet (pp. 22-25). On this G. P. Goold 
writes (TAPA 91 [1960] 284): "The evidence adduced is pitifully 
trivial." 

THE UNIVERSITY OF ST ANDREWS 

February. 1966 



Word-division at Verse-end in Pindar Young, Douglas Greek, Roman and Byzantine Studies; Spring 1966; 7, 1; ProQuest pg. 9

DOUGLAS YOUNG 9 

II 

Word-division at Verse-end in Pindar 

M ODERN EDITORS follow August Bockh in removing, by what 
they consider emendation, the examples of word-division at 
verse-end offered by the Pindar paradosis. Doubts about the 

justification for their proceedings arose strongly in my mind when 
considering a particular instance, at N 1OAlf, where Bockh's Emenda­
tionsversuch involved the assumption of a complex series of transposi­
tions, which moved Hermann to protest. The paradosis there runs: 

" A.."~' t ',/..." 'n' 0fkl-I,(XTWV. V£Ka,!"opL(xt~ yap oaaL~ L7T'7T'OTP0'!"OV aaTV TO pOt-

e '\ TT' e " ~ TOLO aI\TjaEV.H.OpW OV T EV fkVXOt~, ••• 

41 

42 

The transmitted text makes good sense and scans correctly, but the 
word-division llpo{-ITOLO at the verse-end broke the rule of Hephais­
tion, 7T'av fk€TPOV Et<; T€A€{av 7T'€paTOVTaL Mgw, "Every measure (of 
verse) ends in a complete word or phrase." Bockh therefore elabo­
rately transposed the words, and altered one, to run: 

" ,/,."" II' ,~, r , ,/,. op,p,aTWV. VLKa'f'opLaL<; yap oacu<; POtTOLO TOU L1T1TOTpO'f'OV 

" f} '\ 17' f} " ~ aaTV al\T}a€v H.0PLV ov T €V p,VXOL<; • ••• 

41 

42 

Bockh here assumed that his order ABC D had been corrupted in the 
paradosis into C DBA, and that the word TOO' had been altered to TO. 
He claimed (1,329) that he could show many examples in Pin dar of 
disturbed order of words, but in fact he cites none so complicated as 
he assumes here. 

Hermann himself permitted Pindar to divide a word at the end of a 
verse, but at I 8A2f he removed an example by re-writing a passage, 
his motive being to secure exact responsion with other strophes­
what Wilamowitz (Pindaros 9) was later to term the "petitio principii 
strengster metrischer Gleichmassigkeit." At I 8A2f the paradosis 
runs: 

\ \" , .... f}' 
TO fk€V €fkOV 7T"Y}1\€L €a-

" , fkOLpOV 07T'aaaL yafkOV 
, I~ , 

aLaKLUC(- yEpa<; • •.• 



10 NOTES ON THE TEXT OF PINDAR 

Hermann altered that to read: 

Hermann here assumes, in the last six words, that his order ABCDEF 
had been transposed in the paradosis to the order ACDEFB, with the 
second word, yEpa!> , transposed to sixth place. Something similar 
occurred at 0 10.73, where N transposes the word p.Eyav to follow 
€vw7Tl8o!> five words later. But Hermann's change of order also in­
volves two further changes, fh,ap.otpov to O€6p.opov and 07TaUat to 
07TaUUat. These are small changes, indeed, but more than is needed. 
D's text can stand, with the single change of O€ap.oLpov to O€ap.opov, 

in this form: 

That scans as 4th paeon+ 1st paeon+ resolved diiamb, followed by 
glyconic, which responds satisfactorily to the corresponding verses 
in other strophes. 

Before considering other places where the Pindar paradosis exhibits 
word-division at verse-end, it may be noted that Bockh allows Pindar 
to elide at verse-end and to divide phrases, for instance by separating 
a preposition at verse-end from the noun it governs in the next verse, 
or by putting at the start of a verse an enclitic word leaning back on a 
word at the end of the preceding verse. 

Bockh rightly stresses Pindar's taste for what we call enjambement, 
used for emphasis, e.g. at 02.92-95, with its culminating phrase: 

" ... -\\ " "J..l1" , .. allopa p.aNl.ov EVEpyeTav 7Tpa7TtUtV a'fJuovEuTEPOV TE XEpa 

@~pwvo!>. 

Here the name of Theron is first word of the epode, and Pindar 
stops abruptly after it, to give emphasis. Likewise, at P 2.72fhe sets an 
emphatic word at the start of a new triad, thus: 

\1 '0 \ '" KaI\O!> TOt 7Tt WV 7Tapa 7Tatuw, atE' 
Ka,\6!>. 

Sir Maurice Bowra remarks (Pindar [1964] 319) that "for the greater 
part of his career Pindar made his syntactical units run counter to his 
metrical, presumably because he saw the poem as a whole and did 



DOUGLAS YOUNG 11 

not wish it to fall into separate and easily distinguishable sections." 
It is perfectly in keeping, therefore, with that vision of the ode as a 
whole to reinforce enjambement with elision, as occurs notably at 
03.25f: 

C'I '" ~ J () I tJ , 07] 'TO'T E~ ')Iatav TTOPEVEV VfJ-0~ wpfJ-aw 
'T , 
.J.U'Tptav vtv. 

"Straightway his spirit was eager to bring him to the Danubian land." 
On this place Bockh remarks (1,318): "Versum exire posse in vocaL 
bulum apostropho mutilatum certum est ab aliis poetis, estque unum 
in Pindaricis exemplum certissimum Ol.II1,26, ubi vox wpfLaLv' in fine 
antistrophae posita est." Modern editors seem to prefer A's com­
moner word wpfLa, which has the disadvantage of requiring1TopEvEv to 
be taken in an unparalleled intransitive sense. 

At N 8.37ff the paradosis gives us 

,,, "" """ xpvuov EVX0JI'TaL, 1TEOLOV 0 erEpOL 

, , , , "'" ~ < '" , 'e' ~ ,\ '.1. ' a1TEpaJI'TOV, EyW 0 aUTOLS aowv KaL X OVL YVLa Ka V'f'aLfL ' 
" , " ..l.' <:'" , ',\ A aLVEWV aLVT}Ta, fLofJ-'f'av 0 E1TLU1TELPWV a LTpOLS. 

Bockh, and later Bergk, printed the elided Ka,\vifJaLfL', and those with 
whom the Muses are not angry can see how much more emphatic is 
Pindar's first person optative than the "immendation" by \\Takefield, 
Ka'\vifJaL, which depends on a supplied EVXOfLaL. 

Other examples of elision at verse-end affect particles, as at P4.179f: 

, <;" 
'TaXEES U 

afLcP~ n ayyatov OEfLEfJ,\OL'> VaLE'TaOV'TE'> E{3av. 

Bockh keeps the D', which is needed to avoid an unsuitable asyndeton. 
At I 6.31-32: 

'..l. "'" , M' , TTE'f'VEV OE uvv KELV4;J EP01TWV 'T 

l8vea Ka~ T6v {3ov{3oTav . . . 

Bockh keeps the T', but prints it at the start of a new verse, thus: 
T' ;()vea. It means <both', and is desirable, though not necessary. At 
I 8.19f the witness is D, which offers: 

1TaTpos OVVEKa otovfLaL YEYOJI'TO ()vya'rEpES, 'AaW1TtOWY e' 
o1T'\oTaTa, Z7]vt ()' aooy {3auLM"i. [Read TE aooy, E. Schmidt.] 



12 NOTES ON THE TEXT OF PINDAR 

One may keep D's dual, and its ()' as an example of appositional TE 

(cf Dennislon, Greek Particles 502,e): cr ••• because they were born 
twin daughters of their father, and the youngest pair of Asopides, and 
they pleased King Zeus." At I 8.34f the manuscripts have Ef7TE 8' I 
Ev{3ovAos €V p.EaOtat fJEP.tS, ... The connective is needed. Bockh com­
ments (1,318): "insititiae possunt particulae videri; nolui tamen in re 
ambigua mutare quidquam." 

With the conservative caution of the august Bockh here one may 
contrast the radical dogmatism of the venerable Paul Maas, in Greek 
Metre (tr. Lloyd-Jones [Oxford 1962] §139, pp. 87f): "Elision at the end 
of the line is avoided ... there is none in Pindar (on the interpolation 
of these particles see o. Schroeder, Pindar, 1900, 9 ... )." Maas says 
nothing about the two clear cases of elision of non-particles, 0 3.25 
WPWXLV' and N 8.38 KCtAvifJaLP.'. Yet he writes deprecatingly (p. 92) about 
"a metric which, in order to impose its laws, has to change the style 
for the worse or do violence to the transmitted text," and again "the 
metre of a poem is nothing apart from the poetry it is there to serve." 
On poetic grounds the wpp.Ctw' and KCtAvI/JCttp.' offered by the paradosis 
are greatly preferable. 

Bockh defends Pindar's practice of allowing an enclitic at the start 
of a new verse, thus dividing the phrase to which it belongs, e.g. 
N 4.63f: 

and 18.1Of: 

" "c I ) I ••• ovvxCts 0sVTCtTOVS CtKP.CtV 
~ , , ,~ , 

TE oEWOTCtTWV O'XCtO'Ctts OOOVTWV •.. 

• • . TOV 1J7T€P KEcpCtAas 
yE TCtvTlxAov A.{()ov ... 

Bockh also condones Pin dar' s setting a preposition at a verse-end and 
the noun it governs at the start of the next verse, as at 0 13. 112f, where 
Bockh accents 

,- , . . • KCtt 7TCtO'CtV KCtTCt 
tE'\ \ ,~, t , I\I\Cto EVpTJO'EtS... 

Pindar is also willing to leave other prepositive words at verse-end, 
like ws, 7} and KCtl. For examples: 

o 10. 18: "IAf!- cpEPETW X&.ptV 

tAY7JO'l8Ctftos, WS 
, AXtAEt n&'TpOKAos. 



DOUGLAS YOUNG 

P 9.99: 7TapOEVLKat 7TOOW 1} 
( , )f , 

VLOV EVXOVT , 

O 14 5 \ ..... " \ " \ \" , \ \ . : KI\VT , E7TH EvxofLat. GUV yap vfLJuV Ta TEp7Tva Kat 
, \ "" , fJ ~ Ta YI\VKE aVETa, 7TaVTa POTO'S, 

a 9.65: t.I7TEpcpaTov uv8pa fLoPCPo/- TE Kat 

EpyoLa£. 

Fr. 36.16 Bowra: (9paLK{av yatav afL7TEAOWaaV TE Kat 

" EVKap7Tov· 

13 

In view of such phenomena, and of Bockh's conservative respect for 
the textual facts regarding elision at verse-end and the separation of 
enclitics and prepositions, it is curious that he allowed himself to 
override the textual data in the matter of word-division at verse-end, 
in deference to Hephaistion's dogma that every metron ends in a 
complete word. That was in 1811. Contrast the words written in 1962 
by the Regius Professor of Greek at Oxford, Hugh Lloyd-Jones, in his 
preface to his translation of Maas's Greek Metre: "Ancient theories 
about Greek metre are of little or no value; and modern theories are 
valuable only in so far as they are grounded upon the evidence of the 
texts." 

What, then, is the evidence of the Pindar manuscripts in the matter 
of word-division at verse-end? I am here understanding verse-end as 
that established by the occurrence of hiatus or syZZaba brevis in Zonga in 
some stanza of the ode. The case of N 10.41, IIpo{-ITOLO, has already 
been discussed, and the case of I 8.42 D7Ta-laaL, where both have been 
removed in modern texts by the combined assumption of trans­
position of words and corruption. There is another example at P 4.211f 
where the best manuscript, B, offers the perfectly appropriate read­
ing: 

€s cJ>aaLJI S' E7TEtT' av-

~AVOOV, . . . [E~AvOov G2 C V, EV I ijAvOov EJ 

E7TEtTEV ijAvOov of Gl and the later Mss Hand cp is preferred by Bockh 
and modern editors. At 0 6.53 the majority of the veteres offer a"-A' 
Ey-IKEKpv7TTo yap axo{vCfJ ... , which Bockh changes to aM' EV I 
KEKPV7TTO to avoid the word-division at verse-end. At N 8.40f the para-



14 NOTES ON THE TEXT OF PINDAR 

dosis offers the division of a non-compound word at verse-end, thus: 
"1:. <:'" 1 \ ~, I aV<;£Tca 0 ap£Ta, Xl\wpaL~ ££puaL~ 

• "<:'1<:' , ... 
w~ OT£ O£VOPEOV at,u-

UEL, uocpo~~ dVDpWV d£p8£~u' iv DLKaLoL~ TE 7T'po~ vypov 
'8 I aL Epa. 

dtu-IUEL scans right, with Homeric initial long alpha, and makes good 
sense; but Bockh, to conform to Hephaistion's dogma, alters to 

• " <:' I <:' " 
w~ OT£ O£Vop£OV ctUU£L, 

, .J.. ~ '<:' ~ < £v > UO'f'O L~ avopwv . . . 

In view of the lability of EV, this is not a difficult change, were any 
needed. 

At I 6.7fI we find: 

<C And may it be ours, preparing a third bowl for the Olympian 
Saviour, to honour Aigina with libations of honey-voiced songs." 
KaTaU7T'lvDw occurs in an appropriate sense at Eur. Or. 1239, DaKpvoL~ 
KaTau1TlvDw U£. Bockh's K&Ta I u7T'lvDEW is termed by Fennell a <C metrical 
tmesis." The sense would be <C ••• to pour libations (absolutely) over 
Aigina." 

At I 3.18, if one takes I 3 as having the same colometry as 14, the 
view adopted by Turyn and Snell, then we find a compound word 
divided at verse-end, thus: 

, \ <:' \ \ <:' I • I "\ \' "\ \ "1:. aLWV OE KVI\WOO/LEvaL~ a/LEpaL~ allll allllOT £~-

&AAa~£v. CI.TPWTOL yE /LaV 7T'a~o£~ 8£wv. 

Turyn avoids word-division by running the two verses into one. 
Snell prints ig as a separate word at the end of the verse. Bowra writes 
the compound as a single word, and makes one long verse in 13, while 
dividing the end of the epode into two verses in I 4. 

That makes, then, a total of seven places where manuscripts offer 
apparent word-division at verse-end by Pindar. Three of them are of 
uncompounded words, llpot-ITOLO, o7T'&-luaL, dtu-luEL and four of 
compounds, av-HAv8ov, iy-IKlKpV1TTO, KaTa-lu7T'lvOEW, ig-I&AAagEV. 
Only one involves a proper name. To remove the divided simple 



DOUGLAS YOUNG 15 

words from the verse-end involves somewhat more surgery than is 
needed for the compound forms. 

Now Bockh had to defend himself against an accusation by the re­
viewer G. F. Grotefend, that he had been guilty of petitio principii in 
emending away examples of word-division at verse-end. See B6ckh's 
excellent polemic, Narratur historia litis de vocabulis inter duos versus 
non dividendis motae nuperrime (in Pindari opera I [1811] 324ft). Hermann 
too objected in particular to the violent transposition assumed in 
Bockh's emending away of the divided JIpot-1 TOtO (apud Bockh, op.cit. 
1,329). Indeed, Bockh himself had earlier been disposed to allow word­
division for proper names and compounds. But later he wrote (1,85): 

"nunc et perfectiorem video esse Pindarum, quam qui hoc potuerit 
sustinere et pauca exempla, ubi divisum poterat vocabulum videri, 
certa sublata sunt emendatione." He claimed that only four cases 
needed emendation, and he asks (1,313): "Quis vero ob quaternos 
locos in tam corrupto scriptore hoc praeceptum damnaverit, nisi rei 
criticae imperitus, quum praesertim coniecturae, quibus in iis usi 
sumus, non admodum sint audaces?" 

Now there are in fact seven cases requiring emendation, not four; 
and we have only four of the seventeen books of Pindar known to 
antiquity. If we had all seventeen, we might expect to find, not the 
seven cases of our four books, but perhaps twenty-eight cases of 
word-division at verse-end. It may be thought unscientific to emend 
away these seven cases, even if most of the emendations are slight. 
Moreover, we must remember Bockh's defence of cases of elision at 
verse-end, and of separation of enclitics and prepositions from the 
rest of the phrase in which they form part. In general, it may be con­
sidered that Pindar's usages at verse-end in these regards fit in per­
fectly with his zeal for enjambement. 

Moreover, though in 1811 Bockh could excusably write of Pin dar as 
"tam corrupto scriptore," today, thanks to the labours of Mommsen, 
Schroeder, Turyn, Snell and others, we can see that the paradosis is 
substantially very sound. 

THE UNIVERSITY OF ST ANDREWS 

February, 1966 



Emendations and Defences of Readings in Pindar Young, Douglas Greek, Roman and Byzantine Studies; Spring 1966; 7, 1; ProQuest pg. 16

16 NOTES ON THE TEXT OF PINDAR 

III 

Emendations and Defences of Readings in Pindar 

FROM ANALYSIS of errors in sixteen manuscripts it appears that a 
comparatively small proportion involve more than one syllable 
and that they are classifiable into types that recur frequently. 

The following attempts to emend apparently corrupt places proceed, 
in principle, on the basis of making the minimum change from the 
paradosis. 

Non-elision is an occasional source of corruption, and may have 
caused the trouble at I 8.44, where the paradosis is: 

" , , .Q ' '..I. 'J. \ ~ , ..I. '" , OV T EVUE/"EUTaTOV 'fJautv aOI\KOV Tpa'fJEtV 7TEO£OV. 

The metre is Aeolic, and the final verse of the strophe elsewhere 
consists of iambic metron+ 2nd paeon (= resolved bacchius)+ poly­
schematist choriambic dimeter. A satisfactory responsion (with 
bacchius for resolved bacchius) results from eliding cpauLv to make the 
verse 

" , , Q ' '..I.' 'J. \ ,.. , ..I. '" , OV T EVUEfJEUTaTOV 'fJau aOI\KOV Tpa'fJEtV 7TEOLOV. 

The textual problem at I 4.50 perhaps derived from an error in in­
flection, the change of {jTJPCf. to the {jTJpfiv of the paradosis. The gentlest 
medicine seems to be to read: 

1\ '" TOl\fLC! yap HKW!; 

OVfLOV €P,{JpEfLETfiv OYJpif. AE6VTWV 50 
" ,.. "'" \, l:. EV 7TOVCfJ, fLYJTtV 0 al\w7TYJs, • . • 

"For like in spirit to the boldness of roaring lions he goes hunting in 
the athletic contest, but in cunning a vixen ... " 

At N 4.16 one might re-interpret VfLvOV of the paradosis as VfLVWV and 
make the passage run: 

, "',,, r ,..,.", • \ , 
H 0 En t:,afLEVEL .1 LfLOKPLTO!; al\LCfJ 

\ \ '{j '\ '\ {j 'r 
UO~ 7TaTY]p E al\7TETO, 7TOLKLI\OV KL apLt:,WV 

{jafL&' KE T00E fL'AEL KAL{jEIS 

VfLVWV KEA&'OYJUE KaAA{VLKOV ••• 16 



.. 

DOUGLAS YOUNG 17 

«If your father Timokritos were still being warmed by the powerful 
sun, with variously wrought accompaniment on the lyre he would 
often have reclined and celebrated with this strain of hymns the 
glorious victor ... JJ 

At P 6.14 TV7TTOP,€VO' of the paradosis could be re-interpreted as 
'1'V7T'1'OIL€-IICP, used in the Ab Vrbe Condita construction, giving this 
sentence: 

'" ,,, f3 ' "'(} , TOV OU'1'€ X€LIJ.€PL05; o~ P05;, €7T'CXKT05; €I\ wv 10 

€pt{3pOf.LOU V€cPEAas 
, '" " ., " , , aTpaT05; af.L€LI\L Xo5;, OUT aV€f.L05; €5; f.LUXOUS 

<" "e. ,/..' '<:;' al\os asOtaL TTaf.L't'0PCfJ X€paO€L 
, 

TUTTTOf.L€VCfJ • 

HIt (Pindar's treasury of songs) neither a wintry downpour, arriving as 
the inbrought (= mercenary, hostile) brutal army of a thundering 
cloud, nor a wind shall drive into the recesses of the brine with the 
smiting of the all-carrying scree." Pindar, familiar with Delphi, had 
the image of a scree-slope being set in motion by heavy rain or 
squalls. 

At 0 1.87 manuscripts give us, and editors accept, 
Of<:;, '" , ,/.. , ~, ., , ., 
€OWK€V OL't'POV '1'€ xpua€OV TT'1'€pOLaLV '1' aKaf-taV'1'as LTTTTOU5;. 

I have always been sceptical about the notion that Poseidon gave 
Pelops winged horses to compete with Oinomaos. It would look so 
visibly unfair. Also, the wings would not help unless the horses 
soared into the air, in which event the car would capsize. I suggest that 
a nu has fallen out and we should read 

"<:;' '" ',/.. , ", , , ., 
€UWK€V OL't'POV T€ xpua€OV TTT€pVOtaLV '1' aKaf-taV'1'as LTTTTOUS. 

This would make a licit form of glyconic, followed by a polysche­
matist choriambic dimeter catalectic. At Theognis 551 the epithet of 
horses, TaxuTT'1'EpVOWt, 'swift-heeled: is trivialised in some deteriores to 

, 
'1'aXUTTT€pOLaL. 

A syllable has dropped out at N 4.62 in the word offered by the 
paradosis as (}paauf.Laxav. I would amend to (}paavw:xxavos, referring to 
Peleus, as at 0 6.67 the same epithet applies to Herakles. The passage 
would then go: 

~ "" '() " " TTUP O€ TTaYKpaT€5; paauf.Laxavos T€ I\€OV'1'WV 
II 't I ., I 
ovuxas o!, UTa'1'OUS aKf.Lav 

<:;, , , '<:;' , 
T€ U€LVOTaTWV axaaats OOOVTWV ••• 

2-G.R.B.S. 



18 NOTES ON THE TEXT OF PINDAR 

"Boldly resourceful, having subdued all-mastering fire and the ex­
ceedingly sharp claws of lions and the strength of their dreadful 

h " teet ... 
The problem of gloss substitution is raised by 0 1. 63f, where the 

veteres offer, unmetrically, oraw acpfh'Tov (Uuav athov. Much favour 
has been accorded to Mommsen's proposal to read 0 rULV acpOt'Tov (Uv VLV. 

Mommsen's assumption would be that his Olv VtV was glossed by 
Oluuv aV'Tav. It might, however, be a gentler medicine to write ors rv 
acpO£'TOv (Nuuav and assume that OIIIN, written continuously, was 
taken as OrULV and aVTav added for an object, with singling of the 
sigmas of O/.uuav. 

There is a related problem at N 4.68, where the paradosis offers, 
unmetrically, 8wpa Kat Kp&'TOS Eg/.cpavav ES Y€V€cts aV'Tij>. For most 
of the past four centuries the learned seem to have been content with 
Fulvio Orsini's ES y/.vos aV'Tij>; but corruption of ES ylvos to ES Y€V€cts is 
not very likely. I suggest that aVTij> is a gloss, and that the true reading 
might be ES Y€V€&s ot or tv. iv (tv) is not necessarily reflexive, cf 
C. D. Buck, The Greek Dialects (Chicago 1955) §1l8.4; §121. Maas re­
stores it at N 7.98 in a similar context where divine favour is sought for 
a man and the generations of his posterity. 

A crux at N 6.43 may derive from confusion caused by a gloss. The 
paradosis offers, unmetrically in 43: 

Q' " fl'" , ,...OTava T€ VLV TTOU a I\€OVTOS 

tVLK&uaVT' EP€tP€t 8auKLoLS 43 
m' - .,' , ., 'VI\€£OVVTOS V7T WyV'}'LOLS 0P€ULV. 

Hermann produced acceptable metre by writing 

- ,., .J.. c;, , VtKWVT 7JP€'f'€ oaUKLOLS •.. 

The assumption that VLKWVT' ifp€cp€ was corrupted to VLK&UUVT' EP€tPE 
involves a double change of tense. Change of tense is indeed a common 
enough phenomenon, but two adjacent changes would be un­
paralleled in Pindar. It occurred to me that Pindar might have written 
vtKas aVUT€cp€ (= aV/.UT€cP€) , "wreathed for a victory," and that some­
body glossed aVUT€cP€ by EP€tPE, so that in uncials some copy appeared 

EPE't'E 
to a later copyist to present NIKA~NITE<1>E. Taking EP€tPE as a 
correction for what he read as U'T/.cPE, he then made NIKALAN into the 
participle VLK&uaVT'. Misdivision would thus be an additional factor. 



DOUGLAS YOUNG 19 

Misdivision may have led to the crux at N 4.90. 

\ E' A.. ' '8 " , , TOV v'f'aVTJ~ E EI\WV yEpaLO~ TTpOTTaTWp 

t 0 ao~ aELaETaL, TTat t 90 

Pin dar is excusing himself for not writing a special ode for the de­
ceased Kallikles (80), maternal uncle of Tim as arch os. At 90 he perhaps 
wrote <> aws aELa' ETL, TTat. "Old grandfather Euphanes, who is still 
alive, celebrated him (Kallikles) with a will in song." 

Misdivision in uncials could have led to the crux at I 6.46, in the 
prayer of Herakles for the baby Ajax: 

, ,,, A..'" '''' "" t' t ,~ TOV /LEV appTJKTOV 'f'vav, waTTEp TOOE OEp/La /LL/LVOL 7TEpL7Tl\aVaTaL 

8 ' '" , '8' TJpos, ov TTa/LTTpWTOV aE I\WV 
...., "7\T' 

KTELva TTOT EV HE/LECf. 

Stephanus proposed /LE vvv, but its corruption to /LL/LVOL would be hard 
to explain. Perhaps the paradosis arose from misdivision of the uncials 
representing o EP/L' aIL' E/LoL LlEPMAMEMOI could have been divided as 
LlEPMA MEMO\. Then, having regard to the frequent resemblance of 
the narrow uncial epsilon to an iota, a scribe might go from what 
looked like MI MOl to MI MNO\. In any event, this corruption seems to 
involve two stages, as a few do. 

Visual confusion and misdivision in uncials may be invoked also in 
solution of the difficulties at 0 13.114, presented by the words of the 
paradosis aAAa KOVcP0UILV EKVEvaaL TTOULV. Assuming that an uncial 
mu has been misread as double lambda, and that two words have been 
read as one, I would print the passage thus: 

~ ',.I.. ~,-- , 
a/La. KOV'f'0L~ LV EKVEvaaL TToaw. 114 
ZEV TEAEL', alow SLOOL 

Kat TvXav TEp7TVWV yAvKEtav. 

Pindar has uttered a wishful allusion, at 105ff, to Xenophon's hopes 
for future Olympic victories and then gone on to catalogue miscel­
laneous victories of the Oligaithidai. Summing them up, he reverts to 
Xenophon's hopes, thus: "And all over Hellas, if you seek, you will 
find (their victories) more numerous than can be taken in together at 
a glance. May he (Xenophon) corne swimmingly out with his nimble 



20 NOTES ON THE TEXT OF PINDAR 

feet .... " In an ode for a stadiodromos, a reference to nimble feet is 
more probably to those of the victor than to those of the poet, the 
assumption in Maas's conjectures, &y€ KOVCPOLULV EKV€VUW 7TOUtV. The 
accusative and infinitive of prayer at 114 seems to suit the context. 

Dropping of a letter in uncials may explain the crux at 0 1.104, 
where the paradosis runs: 

7TE7TOLea OE ~€1I01l 
P/r] 'TLII' &P46T€pa KaAwII T€ 'tOpLV t afLa t Kat ovvafLLV KVpU.VT€POV 

TWV y€ vvv KAvTai:uL OaLoaAwulfL€V VfLVWV 7TTVXai:S. 105 

The metre needed is: trochaic dimeter+ polyschematist choriambic 
dimeter+ anaclastic dochmius. For metre a gentle medicine is 
Wilamowitz's afLij., but in sense it is superfluous with &fLcp6T€pa. 

Logically, it would imply that Pindar could find somebody either 
(1) ignorant of KaA& and stronger than Hieron, or (2) expert in KaA& and 
equally powerful with Hieron. Pindar is likely to have conveyed his 
compliment on the lines of that paid to Theron in 0 2.92ff, where we 
find two comparatives co-ordinated: 

,~ I , I \ I '\ e A I •.• avoauofLaL €VOPKWV I\oyov al\a €L VOCfJ, 
A I ,t I " I\.J.. '\ ,,'" A \ \ T€K€LV fLTJ TLV €KaTOV y€ €T€WV 7TOI\LV 't'LI\OLS allopa fLal\l\ov 

" " .J..e I I I €V€py€Tall 7Tpa7TLuLv a't' OIl€UT€POII T€ x€pa 

@-r7Pwvos. 

In 0 1.104, to balance the comparative KVpU.fJT€POV there is needed a 
comparative in the corrupt place, after KaAWV T€ 'tOPLV, where the 
older mediaeval Mss offer afLa Kat and later Mss had &.\.\ov Kat 

(Triclinian) or &.\.\OV ~ (Moschopoulean). Metre and sense are well 
satisfied if one writes 104 thus: 

"I am sure that I shall not adorn with noble folds of hymns any other 
friend, among men now living, who is, in both respects, more expert 
in fine things and in power more authoritative." 

The first half of a polyschematist choriambic dimeter can have four 
long syllables. For five long syllables successively in a verse of Pindar 
cf. 09 epode 5 (Snell); P 5 strophe 7 (Snell); and for six longs P 8 epode 
6. Cf. Corinna 1 Page (= Poetae Melia Graea [Oxford 1962] 654) col. 
iii 18, 23, 28, 32; also Sophocles E1. 121, 122; Phil. 204. 



DOUGLAS YOUNG 21 

The Triclinian and Moschopoulean a,uov may descend from original 
f.La,uov, which at the uncial stage had lost its initial mu in the colloca­
tion I~PINMAAAON. AAAON, misread as AMON, might have given 
rise to afLa as an attempt to emend. Possibly the variants a,uov and 
a/LC1. stood in the first minuscule copy of the Photian renaissance, 
and thence travelled to separate branches of the family. 

From attempts to mend faults in the paradosis I proceed to attempts 
to defend some places. 

At P 2.11 editors print 8' apfLaTa, the reading of the manuscript E 
and the Roman edition. But one may consider the variant T' apf.LaTa, 

offered by C prim. D G and V. There is a word apfLa (B) in LSJ, ap­
parently derived from apap{aKw and meaning 'union, love: It is a 
Delphic word, and thus not unsuited to Pin dar, who frequented 
Delphi, for use in a Pythian ode. The sentence would run thus: 

. . . g€aTOV OTav U¢>pov 
" ." 1\ r 1 €V T apf.LaTa 7T€LaLxaI\LVa KaTa",€V'Yvvn 

aOlvos i7T7TLOV, •.. 

"whenever he joins together a polished chariot and the strength of 
horses into unions that obey the reins" (= chariot-teams). At N 7.83 
editors print Hermann's text: 

f3aaLMja O€ 8diJv 7TP€7T€L 
~ 1 ~ .\,~ 1 <, 
Ua7T€OOV av TOO€ yapv€f.L€V afL€pc!-
, , 

07TL. 

But D offers 8€f.L€pa, B 8afL€pa, and the true reading may be 8€f.L€p~ J.rrt. 
Hesychios glosses 8€fL€p6s with f31f3aws, a€fLv6s, €vaTa8-rys. One may 
compare Aesch. PV 134TeXv 8€f.L€pwmv alow. But the metrical problem 
is not easy. Snell thinks the line (epode 4) is analysable as glyconic+ 
cretic, the glyconic having a tribrach ending. But, comparing the 
iambic metra at the ends of lines 1, 2 and 7 of the strophe, it seems as 
if epode 4, and epode 3, can be analysed as glyconic (with final anceps) 
+ diiamb. If so, D's reading might stand as an example of choriamb 
equivalent to diiamb. 

In N 6 elisions at 13b and 50b have troubled the learned with 
metrical perplexities. Snell describes the metre as: "aeolica, dimetra, 
choriambica ad dactylos vergentia." Perhaps we should run together 
what in modern editions are the last two lines of the strophes and 
antistrophes, and regard them as constituting polyschematist iambo-



22 NOTES ON THE TEXT OF PINDAR 

choriambic tetrameters, based on a rhythm equivalent to the chori­
amb, thus: 

~ '.J."" '.J.'" , VVV 7'E 7TE",aVT OVIC a(-lp,opos afJ-",t 7Tal\C!- ICvvayE7'as. 

(choriamb+ ionicus a maiore+ choriamb+ diiamb) 
~ " "A' \ , '·RR ' , .J.' C I VEtlCOS EfJ-7TECT XLI\I\EVS XafJ-aL lCa,..,..as a", apfJ-a7'Wv. 

(ditrochee+ antispast+ molossus+ diiamb) 
At N 6.35 a boxer is described in the paradosis of the veteres as XELpaS 

'(-I-aVTw(JEIS, which Triclinius changed metri gratia to XELpaS 'WXVTL 

O€OE~S. The assumed corruption does not seem plausible. What is the 
metre? Turyn noted: "Discriptio metrica non paucis locis dubia est." 
Snell's description is: "aeolica, dimetra, choriambica ad dactylos 
vergentia." In other strophes verse 6 runs: 

- v v-I v v - -11- v v-I v v u --
We seem to be confronted with a pair of polyschematist choriambic 
dimeters. If so, then ifJ-aVTw(J€IS would scan, on the assumption that a 
molossus - - - can respond to an Ionicus a minore v v - -. Hesychius 
has the verb i(-l-aVTOW, <furnish with straps: the sense required. 

At 0 10.46, where Heracles is laying out the Olympian precinct, the 
veteres offer: 

, <:" I {; "A' , .. " (J ~ • • 7TEpt OE 7Tas aLS I\7'W fJ-EV oy EV lCa apc.p 

OtEICPWE, 7'0 OE ICVICAc.p Oa7T€OOV 46 
(JijICE OOp7TOV AVCTW, • • . 

Some Byzantine of the Palaeologan age altered 46f to 
..- , , "-' I, '''-OLEICPLVE, 7'0 OE ICVICI\c.p 7T€OOV 

;O'Y]IC€ OOP7TOV AVCTLV, • . • 

But how likely is it that 7Tloov would have been corrupted to Oa7TEOov? 

Now it may be remembered that, in Aesch. PV 829, the paradosis 
offers ... 7TPOS MOAOCTCT<X oa7TEoa, which Paley retained, allowing the 
initial alpha in Oa7T€Oa to be scanned long. If the same prosody ob­
tained here, line 46 would scan as 3rd paeon+ 4th paeon+ cretic, 
"v-v I "vv- 1 -"-, which would be an acceptable responsion to other 
strophes. 

Once again the basic soundness of the Pin dar paradosis in the 
veteres appears from close examination. 

THE UNIVERSITY OF ST ANDREWS 

February, 1966