This is a collection of complex and intriguing essays on Aristotle and Plato. The range of subjects is broad, and includes a number of pieces on Aristotle’s treatment of knowledge, some essays on nous, perception and consciousness, works on love, beauty and justice in Plato, and others. Kosman’s approach is precise and also expansive. He has a nice way of using small interesting puzzles about Aristotle’s thought as hidden passageways into exploration of the deepest Aristotelian concepts. He is very careful in considering language, context of obscure statements and all scholarly concerns. But he also allows himself to roam around the periphery of his subjects, almost always bringing us back to the center at the end. This occasionally leads to a feeling that one is tumbling from one concern to another without being able to see clearly where the essay is heading, and at least once or twice Kosman concludes an essay with a paragraph I wish he had put on the front end. But these are barely complaints; the richness of each essay is tied to the latitude he gives himself, and I found every essay elucidating, enriching and provocative.
Though the collection is organized chronologically, with no attempt to group the essays topically, such an organization might have been productive. I read through the essays twice, first in the order presented, and then grouping them topically, and found much instructional complementarity in the latter.
I will follow one such thread in this review, to give the reader a sense of Kosman’s approach. A number of essays address questions in Aristotle concerning the nature of epistêmê, and the related concepts, nous, apodeixis and epagôgê. Kosman begins this exploration in the opening essay of the book, “Understanding, Explanation, and Insight in Aristotle’s Posterior Analytics”, with the unusual translation of the word apodeixis as ‘explanation.’ The importance of this is that even Aristotle’s works on logic should not be regarded simply as laying out rules for proper inference, but Aristotle here, as elsewhere, always has ti esti? questions in sight. Kosman claims that both Analytics are governed by a single interest: the nature of explanatory science (86). We are by nature drawn not just to learn how things happen, but what they are. Kosman quotes Aristotle: “what something is and why it is are the same” (Post. An. 90a15). But Kosman understands this in an interesting way: cause, he says must not be understood as something separate from the entity, but the search for a cause is the search for an understanding of the entity “under that description that reveals some of its kath’ auto predicates.” To explain something, then, is by nature to understand what it is, and to know what something is we must know why it must be what it is; that is, we must know its cause.
To explain means to find a “description K of some entity that is L, which description reveals the cause of that entity being L” (9). Thus, a syllogism can be seen not simply as a proof that something is the case, but as an exploration of why it is the case. For instance (my example): Jack is a bachelor. Jack is also Caucasian, 5' 9" and a philatelist. No syllogism can be drawn from these descriptions. But if we notice that Jack is a Catholic priest, then we might have a syllogism, and we have one precisely because we have discovered the aspect of who Jack is that, viewed as a genus, accounts for his bachelorhood. As Kosman says in a later essay: “the first essential task of scientific explanation will be the noetic task of correctly describing particulars” (90).
As cause and essence turn out to be mutually entailing, like concave and convex, so does the connection between epagôgê and apodeixis. These are usually thought of as separate, complementary processes. Epagôgê is seen as the function of nous, or the process by which we discover the universal major premise of a syllogism, and apodeixis is thought of as a technical procedure by which we draw inferences about particulars from the universals that epagôgê has discovered. Kosman argues first that nous is not a mysterious act of ‘seeing’ principles (19) which is a distinct way of coming to know, complementary to the dependent apodeixis. In fact, “the way in which we come to . . . recognize principles as principles, just is the act of explaining.” Nous is to epagôgê as epistêmê is to apodeixis. But further, epagôgê is not a radically different process from apodeixis, but rather (if I understand Kosman correctly) the ‘concave’ perspective of the process of explanation that reveals principles, in their employment, as principles. The ‘convex’ perspective is apodeixis, which draws from principles to complete an explanation, showing a necessary connection between a cause (the explanandum viewed through the specific characteristic that accounts for this aspect of its being) and the phenomenon under investigation.
This investigation of the nature of explanation continues in another essay, “Necessity and Explanation in Aristotle’s Analytics.” Here the animating puzzle is that Aristotle seems to claim wrongly that mixed syllogisms with a necessary major premise and an assertoric minor premise produce necessary conclusions. This leads Kosman to consider more deeply the role that necessity plays in explanation. The problem begins to dissolve when we shift from a more post- Cartesian understanding that necessity is primarily of value in giving certainty to our conclusions, to an Aristotelian quest for explanation: if the cause does not necessitate the consequence, then it does not explain the consequence. The Aristotelian quest is not for certainty, but rather for understanding. This is a very interesting and useful line of inquiry, deftly explored. Kosman applies it to Aristotle’s mixed modal syllogism problem by claiming that syllogisms of mixed modality represent the first step in the process of explanation, the step of revealing the universal “by making clear the particular.” If I discover a piece of blue plastic with straight edges, measure its interior angles to discover that they are equal to two right angles, and then ask why that is so, my answer will come with the recognition that the shape is a triangle, and it is necessarily true of the triangle that its interior angles sum to two right angles. I have ‘made clear the particular.’ But blue plastic does not necessarily come in the shape of a triangle, so I have also created a mixed modal syllogism: (a) All the interior angles of all triangles necessarily sum to two right angles; (b) this happens to be a triangle; (c) therefore the interior angles of this necessarily sum to two right angles. Subsequent logicians (starting with Theophrastus) would deny that the necessity in the major premise alone can convey necessity to the conclusion. If this just happens to be a triangle, then it just happens to have interior angles that sum to two right angles. Kosman argues, I think, that this may be true of necessity as a guarantor of certainty, but if we are more interested in necessity as the sine qua non of explanation, then the syllogism helps us to identify the universal that should be read through this particular if we are to understand its possession of this characteristic: “having two right angles belongs necessarily to this figure qua triangle” (90). Again, it is through processes such as these that we perform the ‘noetic task’ of discovering the appropriate universal in the particular, and through this kind of apprehension of particulars that we begin to understand the universal. Of course, truly to explain something, Kosman argues, both premises must be necessary. Drawing on my previous example, if Jack just happens to be a priest, then we haven’t fully explained why he is a bachelor. If, however, it was somehow necessary for Jack to be a priest—if for instance Jack needed to be a priest to be fully Jack—then we have understood the cause in a way that reveals the ti esti? of the phenomenon.
But Kosman has a larger point in this. Aristotle resists the obvious pull to surrender to contingency in the mixed modal syllogism precisely because he wants to “deny that the world of the necessary is anything other than the world of the contingent understood properly . . . [H]e wants, in other words, to resist a Platonist separation of contingency from necessity” (91). It is of course not necessary that this piece of blue plastic be shaped as a triangle, nor that we focus on that aspect of it. But for Aristotle, the goal of explanation is to connect the proper necessity to the relevant contingent fact about a particular—e.g. it is not by virtue of its blueness that its angles sum to two right angles, but by virtue of its triangularity. This is what Kosman calls the ‘noetic task’ that is performed in the mixed modal syllogism with a necessary major and contingent minor premise.
In “Saving the Phenomena” Kosman uses the question of whether Aristotle is a scientific realist or instrumentalist to further investigate the nature of explanation. Kosman attacks the question by investigating what Aristotle means by prior and better known têi phusei. He finds both the ‘logical’ and the ‘ontological’ interpretations lacking in their capacity to ground Aristotle’s conception of adequate explanation. His critique of the ontological interpretation, as far as I could understand it, seems weak. But finding both ways of thinking about principles inadequate, Kosman concludes that Aristotle is both a realist and an instrumentalist. The principles he seeks are neither simply ontologically prior nor mere creatures of logic. This adds one more piece to Kosman’s understanding of Aristotelian explanation: we have only explained something if we have fit the syllogism “into the entire explanatory body of a science.” We cannot understand logical entailment “in terms of an individual, isolated piece of explanatory reasoning” (148). This seems to lean towards the instrumentalist view, resembling modern assertions that all explanation is at best the creation of a cognate model of natural phenomena. But the deeper point for Kosman, I think, is that Aristotelian explanation is not a bottom-up architectonic structure. Just as nous is not a separate faculty from epistêmê, and epagôgê is not a process entirely separate from apodeixis, in the Aristotelian conception we are always going back towards universal principles and forward towards inferential application to particulars at the same time. It is in that spirit that the idea of an explanatory nexus is perhaps a more appropriate way of thinking of Aristotle’s analytics.
Kosman’s overturning of the usual architectonic understanding of Aristotle’s search for principles has even stronger explanatory power in considering the interactive role of character and deliberation in the Nicomachean Ethics (last essay: “Aristotle on the Virtues of Thought”). He applies to ethical thinking the claim that “understanding principles comes not from our thinking directly about them,” but through trying to employ them in the process of explanation. By doing so he derives a nifty resolution of a seeming contradiction in Aristotle’s thought, that Aristotle both claims that reason shapes desire, and denies that we can deliberate about ends. In Kosman’s analogue to the understanding of principles in Posterior Analytics, reason shapes desire in ethical action not because we think through proper ends and then desire dutifully falls in line, but rather because, in deliberating on what to do in this concrete instance, we clarify the principles that underlie our desires, and desire can refine and reshape itself in light of that clarification.
I have not done justice to the richness and interconnectedness of these investigations into explanation, and I admit that Kosman’s complex and sometimes elliptical style leaves me uncertain that I have fully understood him. But contemplating his arguments was extraordinarily productive for me. The other essays display the same level of intricacy and insight. This collection not only demands reading by serious students of Plato and Aristotle, but re-reading interspersed with contemplation.