The Shibboleth and translation technique

A dialectal difference in Hebrew between the Ephraimites and the other tribes gave rise to the English term 'a shibboleth' as a means of distinguishing an out-group from the in-group, whether by a mispronunciation specifically - as in the original instance - or, more generally, by a custom or fashion, by a catchphrase or password, or by a characteristic manner of speaking (and sometimes by a moral standard).

Until I can read Hebrew, here is the English Standard Version's rendering of the incident in Judges 12: 5-6:

"And the Gileadites captured the fords of the Jordan against the Ephraimites. And when any of the fugitives of Ephraim said, “Let me go over,” the men of Gilead said to him, “Are you an Ephraimite?” When he said, “No,” (6) they said to him, “Then say Shibboleth,” and he said, “Sibboleth,” for he could not pronounce it right. Then they seized him and slaughtered him at the fords of the Jordan. At that time 42,000 of the Ephraimites fell."

The Old Greek versions differ in their treatment of 'shibboleth', which apparently means "a flowing stream" or "an ear of corn". The codex Vaticanus (showing so-called καί γε Hebraising features) takes a verbum e verbo approach and employs στάχυς, "an ear of corn" (note the first aorist forms of εἰπεῖν).

καὶ προκατελάβετο Γαλααδ τὰς διαβάσεις τοῦ Ιορδάνου τοῦ Εφραιμ, καὶ εἶπαν αὐτοῖς οἱ διασῳζόμενοι Εφραιμ Διαβῶμεν, καὶ εἶπαν αὐτοῖς οἱ ἄνδρες Γαλααδ Μὴ Εφραθίτης εἶ; καὶ εἶπεν Οὔ. (6) καὶ εἶπαν αὐτῷ Εἰπὸν δὴ Στάχυς· καὶ οὐ κατεύθυνεν τοῦ λαλῆσαι οὕτως. καὶ ἐπελάβοντο αὐτοῦ καὶ ἔθυσαν αὐτὸν πρὸς τὰς διαβάσεις τοῦ Ιορδάνου, καὶ ἔπεσαν ἐν τῷ καιρῷ ἐκείνῳ ἀπὸ Εφραιμ τεσσαράκοντα δύο χιλιάδες.

However, the codex Alexandrinus (with Antiochene or Lucianic features) takes a sensus de sensu approach (note the pluralisation too):

καὶ προκατελάβοντο ἄνδρες Γαλααδ τὰς διαβάσεις τοῦ Ιορδάνου τοῦ Εφραιμ, καὶ ἐγενήθη ὅτι εἶπαν οἱ διασεσῳσμένοι τοῦ Εφραιμ Διαβῶμεν, καὶ εἶπαν αὐτοῖς οἱ ἄνδρες Γαλααδ Μὴ ὑμεῖς ἐκ τοῦ Εφραιμ; καὶ εἶπαν Οὔκ ἐσμεν. (6.) καὶ εἶπαν αὐτοῖς Εἴπατε δὴ Σύνθημα· καὶ οὐ κατηύθυναν τοῦ λαλῆσαι οὕτως. καὶ ἐπελάβοντο αὐτῶν καὶ ἔσφαξαν αὐτοὺς ἐπὶ τὰς διαβάσεις τοῦ Ιορδάνου, καὶ ἔπεσαν ἐξ Εφραιμ ἐν τῷ καιρῷ ἐκείνῳ δύο τεσσαράκοντα χιλιάδες.

That is, 'and they said to them, "Then say 'password'!"' (cf. 2 Maccabees 8.23, 13.15).

There was no attempt in either version to reflect the phonetics involved, altough ἄσταχυς existed (and was known in Hellenistic Jewish circles - at least, those of Philo and Josephus). Something along the lines of the <σι-> for <θε-> found in Laconian might have been possible.

The Vulgate gives a version with an attempt to reflect the word requested and the word spoken in answer.

occupaveruntque Galaaditae vada Iordanis per quae Ephraim reversurus erat cumque venisset ad ea de Ephraim numero fugiens atque dixisset obsecro ut me transire permittas dicebant ei Galaaditae numquid Ephrateus es quo dicente non sum. (6. ) interrogabant eum dic ergo sebboleth quod interpretatur spica qui respondebat tebboleth eadem littera spicam exprimere non valens statimque adprehensum iugulabant in ipso Iordanis transitu et ceciderunt in illo tempore de Ephraim quadraginta duo milia.

Note also the explanation of sebboleth, quod interpretatur spica, "an ear of corn". All this seems closer to Vaticanus. Jerome may have drawn on Symmachus' translation, which combined fidelity to the Hebrew with natural Greek idiom and constructions.

There is one further detail of interest. Field's Hexapla has a note on στάχυς: σεβελω * και ειπεν ταχυς z. The * is a hexaplaric asterisk (an 'x' with a dot in each wedge) indicating an addition to the LXX text from another version. The Larger Cambridge Septuagint notes from the margin of cod. 85 : καὶ εἶπεν τάχυς. Perhaps, someone somewhere wanted to reflect the difference of pronunciation with a minimal pair... but note that the accent on "swift" would be ταχύς.