This book will be of interest to advanced students in philosophy, historians of logic and medieval philosophy, as well as logicians who wish to understand the significant contribution that Robert Kilwardby's thirteenth-century commentary on Aristotle's Prior Analytics made to those fields. Paul Thom's analysis treats Kilwardby's views on propositions, syllogisms, reduction, necessity, and contingency with depth and finesse, revealing what he takes to be the Aristotelian ontology underlying them. Helpful summaries remind the reader of the important formal and ontological results of each chapter. The book is a genuine logical treat, despite a misleading typo of '(4)' for '(3)' at page 108, l.26, and the fact that most readers will need a good, strong magnifying glass for some of Brill's printed type and superscripts. I found only a couple of drawbacks.
First, Thom's discussion of Kilwardby's distinction between restricted and unrestricted assertoric propositions, upon which the notions of perseity and per se-necessity depend, does not clearly bring out the ontology of time that underlies it. Restricted assertoric propositions are supposed to capture the presentist view of time, and unrestricted assertoric propositions, the eternalist view—the difference between what is "real" existing in the present and existing trans-temporally or eternally, respectively (19). Thom, however, only utilizes 'is', as opposed to 'is', to mark the difference between unrestricted, eternal, and "unending necessity" and the present tense (26–27, 37–41). Both he and Kilwardby seem to think that only an eternalist view of time, combined with eternal Aristotelian "general essences," is compatible with either logical or nomological necessity. But some fourteenth-century thinkers (e.g., Albert of Saxony) disagreed.
Second, Thom's treatment of Kilwardby's syllogistic theory saddles Kilwardby with a very strong form of de re modal actualism, the view that objects in the actual world have not only real characteristics (e.g., "being a horse," "being red") but also modal characteristics (e.g., "possible redness," "necessary horse-kind"). This much is clear from Thom's discussion of the following propositions: [End Page 482]
i. Everyx : Bx is Ax
ii. Everyx : Bx is necessarily Ax
iii. Everyx : Bx is per se Ax
iv. □ Everyx : Bx is □ Ax
v. □ Everyx : □ Bx is □ Ax
According to Thom, Kilwardby distinguishes two sorts of necessity-statements: per se-necessary and per accidens-necessary (19–20). If (i) is "Every man is an animal," it is true and per se-necessary. So is its modal counterpart, (ii), "Every man is necessarily an animal." Because perseity is involved with (i), "Every man is an animal" (which, given Thom's discussion, really seems to amount to perseity in Aristotle's third way, as opposed to the first or second), (i) becomes semantically equivalent to (ii) and (iv) (21), thereby making (i), (ii), (iii), and (iv) all semantically equivalent (56–59, 157). On the other hand, if (i) is "Every white thing is colored," it is true and per accidens-necessary. So is its modal counterpart, "Every white thing is necessarily colored." Given Thom's analysis, Kilwardby's ontology turns out to involve much more than merely superimposing the ontology of Aristotelian category-theory or the four causes onto his syllogistic theory. Thom claims (i), (ii), (iii), and (iv), actually become semantically equivalent to (v), (20–22, 158, and 218). This strong de re modal actualist result seems to surprise him, as he previously appears to have thought Kilwardby was committed to only the weaker modal position that (i), (ii), and (iii) are semantically equivalent to (iv), though not (v), (22 n. 44). What is philosophically troublesome here is Thom's perhaps all too-eager transformation of the nomological (or perhaps epistemological) necessity of "perseity" into logical necessity, in order to make (iii) and (ii) semantically equivalent (18–22, 38, 148). Thom's own commitment to de re modal actualism emerges throughout his treatment of Kilwardby's per se...