The Jews of the Byzantine Empire did not normally write their spoken language, Greek. If they wrote at all, they tended to write in Hebrew, which is clearly the only language in which they were educated.1 What written Greek they have left us is generally couched in Hebrew characters, and is embedded within what can be considered as primarily Hebrew texts.2 The longest continuous piece of written Greek of this kind is the translation of the Book of Jonah as read during the afternoon service for the Day of Atonement; two different versions of this are preserved in prayer books for the festival.3 However, the language of this translation is highly artificial, as it follows its Hebrew original slavishly.4 A translation of the proclamation of the approaching New Moon, likewise preserved in prayer books, is also constrained by being a close translation of the Aramaic original.5 Against this background, particular importance attaches to some specimens [End Page 490] of Greek written in Hebrew characters preserved in the Cairo Genizah.6 While not very extensive (the longest continuous text in Greek has forty short lines), these fragments taken together add considerably to our stock of evidence about Jewish use of Greek in the Middle Ages, both in quantity and in range. Much of the new material comes under the heading of biblical translation and therefore presents the same difficulties as the Jonah translation mentioned above, but a few texts clearly reproduce the ordinary spoken Greek language of Byzantine Jews, and as such are deserving of close attention. It should be noted that we have relatively little in the way of written specimens of colloquial medieval Greek, and consequently these texts are important for historians of the Greek language as well as for those interested in Jewish linguistics.7 Leaving aside stray Greek words embodied in Hebrew letters, I shall focus in what follows on fragments of four Passover haggadoth. In three of them the instructions for conducting the seder are given in Greek; in the fourth the instructions are in Hebrew but there are some Greek words. Taken together they shed light not only on the Greek language spoken by the Jews who produced them, which is the main concern of this essay, but also on the ritual of the seder as practiced by these Jews, which has some distinctive features. The four fragments are significantly different from each other, in terms both of the contents and of the wording, from which we may deduce that we are dealing with witnesses to an actual practice that was not entirely standardized. The Greek rubrics generally appear to be free and spontaneous compositions, not fixed formulae, although there are some striking similarities in formulation. The date and provenance of these four fragments cannot be established with certainty. One (no. 5 in the corpus) is written on paper in an Oriental hand; the other three are written in Byzantine hands on parchment or a mixture of parchment and paper. Codicological criteria offer no reason to doubt that all belong chronologically in the "classical period" of the Genizah, defined by dated documents in the period from ca. 1000 to the early thirteenth century. On linguistic grounds they cannot be earlier than the tenth century (as indicated by the third-person plural verbal [End Page 491] ending –oun), and in one case (7) not earlier than the twelfth century (the earliest attested datable use of avgho for "egg"). They all clearly derive from a Greek-speaking milieu. Dialectal features are hard to pin down, but the use of the accusative for the indirect object (6,7) points to northern Greece, Constantinople, or Asia Minor. In what follows I will mention the elements of the seder in order, taking the four fragments (nos 5, 6, 7, and 8) together, and indicating at...