P.Mich. inv. 346, dated to the fourth century CE, was republished as SB 14 12173 following Youtie’s editio princeps.

Youtie (1978).

It is a private letter, and on the back, we read the address “house of Apa Kyri in the amphodon of Hierakeion”

See TM Geo 257.

in the capital of the Arsinoite nome. The letter’s writing location cannot be determined on internal grounds.

The APIS record for this papyrus is available at https://quod.lib.umich.edu/a/apis/x-1980.

Fig. 1: SB 14 12173. Image courtesy of the University of Michigan Library Digital Collections.

After the standard epistolary formulae at the beginning of the letter, the text in ll. 13–17 runs as follows: πᾶν ποίησον, Ἀφρείνγ[ιε], | σπαθείν (l. σπαθίον) μοι γάρου ὁ̣[μο]ῦ κ̣αὶ ἀνήσ[ου] | πέμψε (l. πέμψαι) μοι, καὶ ἔπιτα̣ (l. ἔπειτα) [α]ὐτοῦ ἀνάγ|κῃ οἴεις (l. οἴ<σ>εις) τὰ θεραπ̣[ ̣ ̣ ̣] Ἡρᾷ τῶι | πεδίον (l. παιδίῳ) μου. πέπεμφές (l. πέπομφάς) μοι φάσιν | ὅτι ἔρχεται Ζηνοβία [κ]αὶ ὁ ἀδελ|φὸς αὐτῆς· μεμενήκαμεν καρ|τεροῦντες ἐ[π]ά̣νο (l. ἐπάνω) το[ῦ] ποταμοῦ, which was translated by Youtie, “make every effort, Aphreingios, to send me a spathion of fish-sauce as well as anise, and then you will bring here without fail the medicaments for my child Heras. You have sent me word that Zenobia and her brother are coming; we have patiently waited for them on the river.” Youtie contemplated whether “the writer intended ὀμοῦ = ὠμοῦ” (as a result of which one might consider reading ὠ[μο]ῦ) and, on θεραπ̣[...], that “the lacuna is too short to accommodate the neuters θεραπευτικά and θεραπεύματα. Very tempting is θεραπίδια, but this diminutive is extremely rare (see LSJ, s.v.), and I have not felt justified in recovering it from a lacuna.” T.E. Grafton outlined the difficulties with Youtie’s readings as follows: “similarly, in SB 14.12173 (fourth c. AD), the writer of the letter expresses religious devotion, stating at the start of the letter that she makes daily obeisance for her mother. There is also a request to bring the medication for the writer’s child Hēras. There is a hole in the document that obstructs half the word for medication, but the start of it is τα θεραπ-. Whatever the rest of the word is, it is fairly clear from context that the request is for some kind of medicine or related material that the child needs for an ongoing illness of some sort. The exact nature of the medicine or illness is never stated outright. While it is serious enough to require some medication, it does not require a special trip just for the medicine.”

Grafton (2017): 77.

Reconsideration of the text allows for καὶ ἐπίγε[ι (l. ἐπείγει) α]ὐτοῦ in l. 15, instead of καὶ ἔπιτα̣ (l. ἔπειτα) [α]ὐτοῦ. For the form of the letters γει cf. the same letters in l. 28. The phrase ἀνάγκη (without article) ἐπείγει is found in Homer (Il. 6.85 and Od. 19.73 ἀναγκαίη γὰρ ἐπείγει) and in prose texts, such as in Joseph., Vit. 162 κἂν μεγάλη τις ἐπείγειν ἀνάγκη δοκῇ, Artem. Onir. 2 ὅπου γε μὴ μεγάλη ἀνάγκη ἐπείγοι, Men.Rh. 432 (Spengel) ἀνάγκη γὰρ ἐπείγει. We find a similar expression using the compound verb κατεπείγει in PSI 7 804.2 (337 CE) [ἐπεὶ ἀνάγ]κ̣η κατεπίγει (l. κατεπείγει) μ̣ε γενέσθαι (although ἀνάγκη is restored). This document dates to approximately the same period as the Michigan papyrus, and both private letters convey a sense of urgency. The word ἀνάγκη can be construed with the simple genitive (here the pronoun αὐτοῦ), as an equivalent of χρεία, e.g., in P.Mich. 3 213.7–10 (III CE) ἐὰν δέ \σοι/ ἀνάγκη | γένηται σείτου (l. σίτου), δώσει σοι ὁ παραδούς σοι ταῦτά μου τὰ γράμμα|τα, “if you need wheat, the man who delivers my letter to you will give it to you,” or in P.Mert. 1 38.22–23 (mid-IV CE) ἀνάγη (l. ἀνάγκη) μοι <ἐ>καίνε|τω (l. ἐγένε|το) τῶν ἀδελφῶν μου, “there arose a need for my brothers;” cf. also the phrase τῆς δὲ χρείας ἐπειγού|σης τάχιστα τὴν κατακομιδὴν τοῦ σείτου (l. σίτου) γενέσθαι in P.Oxy. 18 2128.32–33 (165 CE).

The pronoun αὐτοῦ in l. 15 refers to something singular, presumably a spathion of fish sauce and aniseed. However, these are two different products and should be kept and sent in different containers. Therefore, αὐτοῦ refers either to the dispatched products as a single thing, or to one of these products. In addition, at the end of l. 14 we may restore ἄνησ[ον] and not the genitive, since we cannot say with certainty whether aniseed was referenced as a product without an indicated quantity, or as a substance stored and transported in a spathion or another type of container.

Then, although one may compare δώει for δώ<σ>ει in P.Oxy. 3 522.26, I think that understanding οἴ<σ>εις for οἴεις is difficult.

Gignac (Gram. I, 130) refers only to omissions of medial σ before a stop, especially a dental or a nasal.

A better reading is ὁς (l. ὡς) εἰς, with the upper part of sigma obliterated (cf. the sigma of της in l. 5, less its upper part). The incorrect spelling of ω as ο can be also be seen in ll. 12 ( Ἀπολλόνιον) and 20 (ἐ[π]ά̣νο). At the end of l. 16 we should read τον and not τωι; cf. the shape of omicron in τον (l. 12), where omicron is left open at its top. Omega, by contrast, is less likely, based on the shape of the letter in ὑμῶν (l. 9) and Σεραπίωναν (l. 11).

If the reading ὁς (l. ὡς) εἰς τά is the correct understanding of the text, ὡς functions as an adverb to indicate real intention before a preposition (see LSJ s.v. C IIa): we may compare, e.g., P.Oxy. 7 1063.5–6 (II–III CE) ὀνόματα τρία ὡς εἰς ἀρχεφοδεί|αν (l. ἀρχεφοδί|αν); P.Flor. 1 9.8–9 (255 CE) ὡς εἰς τὴν προ|κειμένην κώμην Θεοξ̣ε̣νίδα.

As for θεραπ̣[±2] ̣ηρα, the letter after the lacuna ends with a horizontal stroke, which could be the end of an α, ε, σ, or τ. The forms of these letters are not consistent in this papyrus and their finishing strokes reach the following letter at different heights in the line. The prepositional phrase εἰς τά requires a neuter noun in the accusative plural, therefore, we could read either a word θεραπ̣[±2]α̣, as Youtie assumed and attempted to restore, or another word which included (and ended with) ηρα, that is, θεραπ̣[±2] ̣ηρα. However, we do not know any noun beginning with θεραπ- and ending in -ηρόν (in the nominative singular). We could propose εἰς τὰ θεραπ̣[ευ]τ̣ήρ<ι>α τȏν πεδίον (l. τῶν παιδίων) μου, which requires a spelling error, that is, the omission of iota; cf. similar errors in Gignac, Gram. I 304, who also noted κεντηνάρα (for κεντηνάρια) in P.Cair.Isid. 56.3, 5 (315 CE).

Cf. also the omission of ι in the feminine noun βακτηρία in BGU 13 2361 Fr B Col. 1.7 (IV CE) βακτηρας (= βακτηρίας; l. βακτηρίαι).

When Youtie published this papyrus in 1978, the word θεραπευτήρια had not yet appeared in the papyri, and we do not know what he would advise based on his rule “iuxta lacunam ne mutaveris.” The noun θεραπευτήρια was attested for the first time in P.Oxy.Hels. 50.16–17 (III CE; found in Oxyrhynchos; place of writing is unknown), which was published in 1979 by P. Hohti, and then in two invitations to a festival therapeuteria for girls (P.Oxy. 66 4542 and P.Oxy. 66 4543), published by D. Montserrat in 1999.

The purpose of this festival is not clear, and Montserrat (in the introduction of the edition, and separately in an article

See Montserrat (1991): 43–49.

) proposed that they may have related to the puberty of the girls, and probably it was a practice associated with a temple ritual which somehow involved the girls. Later, S. Huebner

See Huebner (2009): 149–171.

suggested that therapeuteria could be a female recovery post-circumcision.

So also Malouta (2012): 300; el-Mofatch (2016): esp. 2001.

On the other hand, A. Nifosi suggested that therapeuteria “is probably a temporary service of unmarried girls for Isis, which included a sort of vow to guarantee their chastity. We cannot rule out completely the interpretation of the θεραπευτήρια as the ‘recovering’ of the girl from circumcision, as Huebner claims, but there are many reasons to question this theory,”

Nifosi (2019): 24.

and “there are no attestations of coming-of-age rituals in the Hellenistic Period, nor is there any information about coming of age rituals outside Oxyrhynchos. Therefore, at the moment we can only safely say that the θεραπευτήρια was a ritual performed by Roman citizens living in the area of Oxyrhynchos in the 3rd century AD.”

Nifosi (2019): 25.

Finally, Alba de Frutos García

de Frutos García (2022): 333.

stated that “direct sources do not tell us very much, so both explanations [that is, by Monteserrat and Huebner] seem plausible. Whatever the ritual, the θεραπευτήρια could have celebrated a newly acquired girl in marriageable and fertile condition.” Ιf my suggestion of the reading εἰς τὰ θεραπ̣[ευ]τ̣ήρ<ι>α τȏν πεδίον (l. τῶν παιδίων) μου is correct and is accepted, the evidence of the therapeuteria reaches into the fourth century CE (at least mid-fourth century CE).

H.C. Youtie dated the papyrus to the fourth century CE, and W. Clarysse to the mid-fourth until mid-fifth century CE, because of the name Apa Kyri, which is hardly found before the fifth century (information retrieved from the date entered in TM 32954).

The proposed attestation of the noun in a Christian context in our papyrus may serve to invalidate the connection to Isis. Oxyrhynchos could be only one of the many possibilities as the place of writing of the letter. The writer of the letter does not write his/her name in the opening formula or elsewhere in the letter, but only refers to his/her παιδία,

Based on the new reading the person Ἡρᾶς with TM Per 373993 should be removed as evidence.

and (s)he is the person who seems to be responsible for their therapeuteria. The plural παιδία is interesting. Are these παιδία his/her children or his/her slaves?

Certainly, the word included female slaves, as P.Ammon 1 3 Col. 4.10–11 (348 CE), a private letter, γ γ̣ὰρ τοῦ Παχὼν ἀ̣π̣ῆ̣[λθε]ν κ̣α̣τ̣έλει|[π]εν δέ μοι ἐντα̣[ῦθα] τ̣ὰ̣ παιδία αὐτοῦ ἐπεὶ θ[η]λυκά ἐστιν, and other documents show.

Are these παιδία male or female? It is obvious that these questions cannot be answered with any certainty, therefore, the problem of the nature and purpose of the therapeuteria cannot be addressed on the basis of this papyrus. As we know from P.Oxy.Hels. 50.15–17, the feast was postponed until wine was supplied. Here, it is not certain if both fish-sauce and aniseed are needed (see above discussion concerning the pronoun αὐτοῦ in l. 15). Like wine, fish-sauce is more appropriate to the context of a feast than to medical treatment, but aniseed can be used both in cooking and medicine.

See Manniche (1989): 135.

Therefore, I would prefer to read the text as follows: πᾶν ποίησον, Ἀφρείνγ[ιε], | σπαθείν (l. σπαθίον) μοι γάρου ὁ̣[μο]ῦ (or, l. ὠ̣[μο]ῦ?) κ̣αὶ ἀνήσ[ου] (or, ἄνησ[ον]?) | πέμψε (l. πέμψαι) μοι, καὶ ἐπίγε[ι (l. ἐπείγει) α]ὐτοῦ ἀνάγ|κη ὁς (l. ὡς) εἰς τὰ θεραπ̣[ευ]τ̣ήρ<ι>α τȏν | πεδίον (l. τῶν παιδίων) μου “make every effort, Aphreingios, to send me a spathion of fish-sauce as well as aniseed, and a necessity for it is pressing for the therapeuteria of my children.”