Al-Sijistani represents the Neoplatonist school of the Ismai-liyya. This development was initiated by al-Sijistani's predecessor, Muhammad al-Nasafi, but, according to Walker, it is al-Sijistani who expresses the high water-mark of the Neo-platonist approach. Walker endeavors to demonstrate that the importance of al-Sijistani, "perhaps the foremost Muslim Neo-platonist theologian and philosopher of his era," extends well beyond the sectarian and doctrinal issues involved (p. 17). As an example of the impact that al-Sijistani's thought may have had on the development of medieval Islamic philosophy, Walker dwells, in particular, on its possible influence on Avicenna. As another, not less important example, one may mention the Longer Version of the so-called Theology of Aristotle. Walker alludes to the striking similarities between al-Sijistani's emanational theory and that of the Longer Version, and in particular to the concepts of the divine will or the Logos which both compositions have in common, and which are not to be found in the Enneads or in the Shorter Version of the Theology. The question of the origin of the Longer Version is still open to debate: did it originate in Ismaili circles, as suggested by Shlomo Pines ("La longue recension de la theologie d'Aristote dans ses rapports avec la doctrine ismaelienne," REI 22 [1954]: 7-20), or was it only adopted by the Ismailis, as suggested by Fritz Zimmerman ("The Origins of the Theology of Aristotle," in Pseudo-Aristotle in the Middle Ages: The Theology and Other Texts, ed. Jill Kraye et al. [London: The Warburg Institute, 1986], 109-295), or perhaps it is the product of Jewish Neoplatonists who were influenced by the Ismailis, as suggested by Paul Fenton ("The Arabic and Hebrew Versions of the Theology of Aristotle," in Psuedo-Aristotle in the Middle Ages, 241- 64). The study in depth of Ismaili texts, as the work reviewed here, is an essential step on the way to a sound reexamination of these theories.

The Wellsprings is a philosophical book, intended by its author for "someone long practiced in the premises of demonstrative reasoning" (p. 42). Although it is not a systematic work, its forty discourses, called "wellsprings" (Yanabi, sing. yanbu) cover the ground of the Neoplatonist emanational theory, as well as that of Ismaili theology. In his introduction, Walker discusses the philosophical sources of the work (e.g., The Theology of Aristotle and the Pseudo-Ammonius), and points to its background in a syncretistic religious milieu - a milieu reflected in its use of the Christian Gospels. Walker dates the composition of the work to around 350/961, i.e., under the reign of the Fatimid Caliph al-Muizz, and discusses its authenticity, the general outline of its doctrine, and its relation to al-Sijistani' s other works.

Walker intends this book as "a study of an individual work of Ismaili philosophy, of this sect's theology and doctrine, and of a form of literature that its author chose as his preferred intellectual vehicle" (pp. x-xi). The analytical part, incorporated in the introduction and the commentary, complements and deepens in an exemplary way the synthesis provided already in Early Philosophical Shiism.

The bulk of the study revolves around the translation of Kitab al- Yanabi, and it is on this translation that I would like to offer a few comments. Walker rightly introduces al-Sijistani's work as belonging to "an obscure, quasi-philosophical tradition from a minority, sectarian position, where the meaning and context of words and phrases are not necessarily those of well-known authors" (p. x). The English translation provided by Walker, however, makes al-Sijistani much less obscure. Compared to Henry Corbin's French translation of parts of the Kitab al-Yanabi (Le Livre des sources, in Trilogie ismaelienne [Paris, 1961], 3-126), Walker's translation is refreshingly straightforward. (A typical example is the word muqaddas, rendered by Corbin as "sacrosaintete, " and which Walker translates as "hallowed" - a word as hieratic but less ornate, not to say less pompous.) Walker aims at a translation that will offer the reader a coherent, intelligible English text. It should be said from the outset that he succeeds to an impressive extent in achieving this difficult goal.

Nevertheless, it seems to me that the desire to give precedence to fluent English, and to limit the reader's need to refer to footnotes, produces at times infelicitous translations. In al-Sijistani's Neoplatonic epistemology, the sources of knowledge are four: intellect, soul, speaking-prophet and founder. The role of the first source, the intellect, is tayid, which, as Walker notes correctly, stresses "concepts of support, confirmation, collaboration, strengthening" (p. 7). Yet, instead of choosing one of these words, Walker translates tayid throughout as "inspiration." It is indeed true that, for al-Sijistani as well as for other Neoplatonizing Ismailis, "the benefit of intellect is inspiration, a kind of infusion of knowledge" (p. 7). Corbin, who usually translates tayid as "confortation spirituelle," also suggests that sometimes "inspiration" is preferable.

In the translations of medieval philosophical texts, however, "inspiration" usually renders another Arabic term. According to the freethinkers of medieval Islam (e.g., Ibn al-Rawandi and Abu Bakr al-Razi), human beings, like ducks and geese, are endowed from birth with all the knowledge essential for their salvation. This innate, intuitive knowledge, called ilham, is the freethinkers' answer to the authoritarian epistemology of prophetic religions, and it has become common practice to translate this term as "inspiration." It cannot be argued that the freethinkers' concept of ilham is irrelevant to Ismaili doctrine, or that tayid and ilham belong to two different circles, whose thought may never meet. The Ismailis perceived the freethinkers' concept of ilham as a direct threat to their own hierarchical soteriology. It is no mere chance that the few extant fragments of the freethinkers' heretical works are preserved by Ismaili authors, such as al-Muayyad (!) fi' l-Din and al-Sijistani's predecessor, Abu Hatim al-Razi. In the texts of these authors, tayid and ilham appear as contradictory concepts. In this context, it would seem preferable to translate tayid in a manner that would stay closer to its literal meaning.

Another problematic translation relates to the role of the speaking- prophet (natiq), who "functions first and foremost by verbal compilation, the codification of law, the scripting of truth into words" (p. 8). His specific role is thus that of talif, which means literally "composition." Accordingly, Walker translates it sometimes as "scriptural compilation" (e.g., on p. 4). True, this literal (and accurate) translation does not wholly convey the profound meaning of the prophet's role, whose scripture "is a linguistic incarnation of intellect (aql mujassam)." Walker's decision, however, to translate it in most cases as "incarnation" gives precedence to the refinements over the essentials. It seems evident to me that, for any reader of the Arabic text, including educated Ismailis, talif continued to indicate first and foremost the Prophet' s connection to the composed text of the Scriptures.

The translator of Arabic medieval philosophical texts (as, indeed, the translator of any text) has to make a difficult choice between an accurate, literal translation, which may sound wooden, and a freer, more elegant rendering, which may lose much of the flavor of the original. In general, it seems that Walker has opted in favor of fluent English, a choice which I find wholly justifiable. In the case of key terms, however, the original meaning becomes of prior importance. Terms like "inspiration" and "incarnation" may seem familiar to the English-speaking, mostly Christian reader, and thus make the text more accessible, but they miss too much of their original semantic value, and are therefore misleading.

Besides the translation and analysis, the Arabic text itself claims our attention. Walker rightly acknowledges Corbin's edition of the Kitab al-Yanabi as "a usable, intelligently edited version." Walker' s translation, however, is based on a re-examination of the manuscripts and clearly reflects what amounts, in fact, to the preparation of a new edition of the text. In a review of Walker's Early Philosophical Shiism (JAOS 115.3), I expressed the hope that Walker's book would open the way for scholarly editions and translations of al-Sijistani' s works. The publication of the translation of The Wellsprings fulfills part of this wish, and we are grateful for it; one does not usually see one's hopes realized so quickly. This publication, however, whets the appetite for more. Al-Sijistani's Kitab al-Maqalid, for example, is yet unedited, but also a new edition of the text of Kitab al-Yanabi would not be superfluous.

SARAH STROUMSA THE HEBREW UNIVERSITY