² See Sáenz-Badillos, A. and Targarona, J., Gramáticos hebreos de Al-Andalus (siglos X–XII). Filología y Biblia, (Córdoba, 1988)Google Scholar and Tene, D., ‘Ibn Janah, Jonah’, Encyclopaedia Judaica 8, 1971, 1181–1186Google Scholar.

³ For instance, in his edition of the Kitāb al-'Uṣūl, A. Neubauer states that ‘These three MSS. have more or less vulgar forms, especially the plural în for ûn and the conditional form for the indicative. As I felt persuaded that Abu ՚I-Walîd did not employ vulgar forms, I tried to restore the classical form wherever I could’. (Neubauer, A., The book of the Hebrew roots by Abu ՚l Walid Marwan Ibn G'anah otherwise called Rabbi Jonah (Oxford, 1873–1875), vm). This last comment would be inconceivable in a modern editionGoogle Scholar.

⁴ J., and Derenbourg, H., Opuscules et traitès d'Abou ՚l-Walid Merwān Ibn Djanāh (Rabbi Jōnāh) de Cordove, Paris, 1990Google Scholar.

⁵ The Oxford manuscript (Poc. 134, Uri 458) is located in the Bodleian Library and is dated to 1316.1 was able to consult the Firkowitch manuscripts in the microfilmed copies of the Hebrew Institute of Hebrew Microfilmed Manuscripts. These manuscripts bear the numbers 4587 (14 ff), 481 (2ff) and 1014 (1 f). The first is a complete copy of the original and dated to 1126 C.E. I thank Dr Abraham David for his assistance with my research on the manuscripts of the Firkowitch collection.

⁶ J. and H. Derenbourg, Opuscules, cxm: ‘Nous disposions d'abord d'une copie de ce manuscrit que M. Neubauer s'était fait pour son usage et qu'il nous a gracieusement abandonnée. Plus tard, pendant le cours de l'impression, les curateurs de la Bibliothèque nous ont confié pendant un certain temps, le manuscrit lui-meme.’

⁷ ibid., cxrx: ‘Le manuscrit, qui est fort bien conservé, a cependant souffert aux derniers feuillets, et certaines parties étaient devenues tout à fait illisibles. Nous avons pu heureusement les rétablir d'après un manuscrit du Kitâb at-taswiya qui s'est trouvé récemment dans la collection Firkowitch […].’

⁸ The only reference I have found regarding this inaccurate rendition is D. Tene's brief comment on the fact that Sefer ha-hašwa'ah was an incorrect translation of Kitāb al-taswi'a, yet he offers no further comment about this: ‘At Beziers in the mid-13th century, Salomon b. Joseph b. Job translated Ibn Janāh's Kitāb al-taswi'a (‘The Book of Rebuke’) under the incorrect title Sefer ha-Hashva'ah.’ Tene, D., ‘Hebrew linguistic literature’, Encyclopaedia Judaica 16, 1359Google Scholar.

⁹ Concrete phrases from the KT are referred to by folio number and lines of the Oxford manuscript.

¹⁰ I follow the Petersbourg variant here (ןמ). The Oxford manuscript, which the Derenbourgs followed, has אמ instead.

¹¹ MS. 5460 of the Biblioteca Nacional, Madrid, contains the second half of this translation. It was translated into Spanish by J. M. Camacho, ‘Rabi Yonâ ben Gannach. La segunda mitad del “SEFER HAHAXUA”, versión hebraica de su “ KITAB AT-TASUIYA ”, por Salomón bar Yosef ben Ayyub’, Boletin de la Real Academia de Ciencias, Bellas Letras y Nobles Artes de Córdoba, 1929.

¹² ‘ Er nannte die Schrift Taswîja, “ Vergeltung ” weil unter den Gründen zu ihrer Herausgabe auch der war, seinen Kritikern “ ihr hässliches Thun zu vergelten, indem sie mit einer Disciplin sich befassten, die sie nicht verstūnden.” ’ Bacher, W., Leben und Werke des Abulwalîd Merwân Ibn Ġanâh (R. Jona) und die Quellen seiner Schrifterklärung (Leipzig, 1885), 18Google Scholar.

¹³ In ‘ Notice sur Abu ᾿l Walid Merwān ibn Djanah… suivi de l'introduction du Kitab al-Luma’ d'lbn Djanaḥ, en arabe, avec une traduction française et des notes', Journal Asiatique, serie IV, 16, 1850, 5–50, 47Google Scholar.

¹⁴ The linguistic features mentioned here are exclusively those not reflected in the Derenbourgs' edition. It should be noted that these features represent only a small proportion of the significant number of Judaeo-Arabic characteristics in the text. There are also some interesting linguistic differences between the Oxford and the Petersbourg manuscripts which are not commented upon here.

¹⁵ ‘ It seems possible that this orthographical usage came into being by analogy: ᾿alif mamdûda occurs relatively often after i (…) and was accordingly, owing to ᾿imâla, spelled with yâ. Then, by orthographical analogy, yâ was used to denote ᾿alif mamdûda in general, even when not pronounced with ՚imâla՚, Blau, J., The emergence and linguistic background of Judaeo-Arabic (Oxford, 1965), 74Google Scholar.

¹⁶ Cf. Blau, J., A grammar of Mediaeval Judaeo-Arabic (Jerusalem, 1981), 24–25Google Scholar.

¹⁷ ibid., 63–4.

¹⁸ ibid., 19.

¹⁹ ibid., 150 2.

²⁰ ibid., 235 7.

²¹ ibid., J. and H. Derenbourg, Opuscules, 353.

²² ibid., 254.

²³ ibid., 373.