Dr O’Neil * Analysis of the Pella Curse Tablet (Katadesmos).

I translate the inscription as follows:

I forbid by writing the ceremony and the marriage of Dionysophon and Thetima, and of all other women, and widows and virgins, but especially Thetima, and I assign them to Makron and the daimones. And whenever I shall unroll and read this again, after digging it up, then Dionysophon may marry, but not before. May he not take any wife but me, and may I and no other woman grow old with Dionysophon. I am your suppliant; pity me. dear daimones, for I am weak and bereft of all friends. But protect me so this does not happen and evil Thetima will perish evilly, [undecipherable] mine, but may I be fortunate and blessed, [undecipherable].

I will comment on points which throw light on the dialect of the curse tablet as well as on those needed to understand its meaning.

line 1: Θετί]μας loss of the omicron from θεο- is common in Doric, especially in Megarian. Final long alpha instead of eta shows this is not Attic-Ionic.”

καταγράφω is a Dorian form. Attic uses other verbs for this purpose.

τάν άλλάν πασαν. -άν is the regular first declension plural form in Doric and Aeolic, which contract long alpha and omega differently from Attic-Ionic.

line 2: παρκαττίθεμαι (= Attic παρακατατίθεμαι) apocope of prepositions is common in Doric, and even more so in Aeolic dialects.

Makron is the deceased, with whom this defixio was buried.

line 3: δαίμοσι is the standard Greek form. The neighbouring Greek dialects. Thessalian and Northwest Greek would have δαιμόνεσσι and δαιμόνοις respectively. όπόκα is Doric, corresponding to Attic όποτάν. Doric regularly has κά in place of Attic άν.

Optatives in place of subjunctives are also found in Doric. διελέξαιμι corresponds to Attic διελίξαιμι “unroll”. It may be an error, or the substitution of epsilon for iota may show a more open pronunciation of the vowel. This is found twice in SEG 38 (1988) 649.‘ a late fourth century B.C. epitaph from Pella, which reads εσστέ for εστί. and Ίφεκράτης for Ίφικράτης. πάλειν for Attic πάλιν. Confusion of long iota and the diphthong -ci is also found as early as the fourth century in Attic. Voutiras reads πάλ{L}ιν, which he explains as the writer starting to write an E. recognising it as an error, but failing to erase it.

άνορόξασα corresponds similarly to άνορΰξασα, just as in the case of διελέξαιμι above. There are some parallels in the Mace­donian glosses for the substitution of omicron for upsilon. Confusions of upsilon and omicron are found extremely rarely in Attic inscriptions. Without more examples we cannot be sure these cases are typical of Macedonian, or rare errors like the Attic examples. The writer of the curse is not weakening it by this exception, but confirming it will not be reversed. Similar examples of reinforcing a curse by an impossible alternative can be found, but there is no precise parallel for using the impossi­bility of the curser changing their mind.

line 4: γάμαι aorist infinitve with -ms- to -m-. Aeolic would have -mm-, so this cannot be Thessalian.

line 6: ...]αν feminine name of three letters plus -an is probably not an Attic form. The Macedonian name Phila, read by Voutiras, would fit. but so would other short names. If the conjecture of Phila is correct, it is unfortunate that we do not have the name in full, so we could see whether the aspirate was written as voiced.

line 7: έμΐν is a West Greek form of the pronoun. ΔΑΓΙΝΑΓΑΡΙΜΕ: Dubois suggests pi should be read for the first gamma, and that the initial delta is a voiced tau. Aspirates are more commonly voiced in Macedonian than unaspirated unvoiced plosives but the latter are occasionally found voiced. The two iotas may represent the dipthongs -ei, and the epsilon the diphthong -ai, the reverse of the case of πάλειν above. The passage would then read ταπεινή γαρ εΐμαι on Dubois’ inter­pretation ταπεινή “weak” makes good sense in the context,

line 8: γενέσται is a Northwest Greek form corresponding to Attic γενέσθαι .

ΨΜΝ is not a possible combination of consonants in Greek. Voutiras reads YMN. but this still cannot be read to give meaningfiil Greek in his context.

line 9: τψ is also not a possible combination of consonants in Greek. It may be that the curse tablet had words which could not be read as Greek as some kind of magic. Unfonunately these occur where the tablet is damaged, and there are no sign of such “magic words” where wc can read it clearly.

The curse tablet can definitely be seen to be in some form of West Greek, with forms corresponding to Doric and Aeolic dialects, but quite clear differing from Attic-Ionic. It does seem in its use of the form Δαιμοσι to be distinct from either of the known West Greek dialects spoken in the areas to the south of Macedon., Thessalian and Northwest Greek. So it seems unlikely that its use at Pella is due to the migration of an individual from the areas immediately In the south of Macedon or to the adoption of one of the dialects spoken in those areas by a Macedonian native. On the other hand, the form άνορόξασα, with its omicron in place of the regular upsilon. does have parallels with attested Macedonian glosses and there is reason to think Macedonian may have had more open pronunciation of short upsilon and iota. The simplest explanation for the dialect forms in the curse tablet is thait it has been written in the original Macedonian dialect, and that that dialect is a West Greek one, related to, but distinct from its more southern neighbours.

“Doric Forms in Macedonian Inscriptions” James L. O’ Neil (University of Sydney)