Crivelli’s monograph is devoted to Plato’s account in the Sophist of how false yet meaningful sentences can be possible. The discussion takes the form of a detailed philosophical commentary on the middle part of the dialogue (236D-264B). Crivelli quotes, translates, and discusses the text almost in its entirety, including issues of text and translation. We get a thorough and lucid reconstruction of Plato’s argument. At contested points, Crivelli offers a classification of the available interpretations, meticulously ponders the arguments for and against each class, and makes a choice between them. This procedure makes it difficult to do justice to the subtleties of individual contributions, but this price must be paid if a complete review of interpretative possibilities is being aimed at. Crivelli’s book is bound to become a standard reference in Plato scholarship because of its comprehensiveness, philological rigor, and philosophical clarity.
Chapter 1 (The Sophist Defined) briefly discusses the frame of the dialogue which portrays the Eleatic Visitor’s and Theaetetus’ search for the definition of the art of the sophist. Crivelli, adopting the most straightforward reading, takes it that the first six attempts to define the sophist fail, while the final definition is meant to be accepted. In the discussion leading up to the seventh definition, Plato puts sophistry and imitative arts into the same class. According to Crivelli, the image-making art of the sophist consists in producing false sentences that are images of true sentences (p. 23 f.). In my opinion, both Plato’s formulation1 and the analogy of painting and sculpture suggest that the original-image relationship obtains between objects of the real world and objects of the world of appearance created by the sophist’s discourse rather than between linguistic items, true sentences and deceptive false sentences.
Chapter 2 (Puzzles About Not-Being) addresses the problems concerning not-being, images, and falsity – concepts involved in the final description of sophistry. In Crivelli’s view, the fundamental problem about not-being is that what is not is unsayable. He interprets ‘not-being’ in terms of non-existence in the initial puzzles (237B7-239A12), and the first two puzzles can easily be understood along these lines: (1) whoever tries to say what is not, does not accomplish an act of saying, because there is no object to which the act of saying could be directed; (2) no expression can be applied to what is not, because ‘not-being’ or ‘not-beings’ imply numerical attributes which in turn imply existence. However, I have some reservations concerning this interpretation: in the last puzzle (3), the incomplete use of the verb ‘to be’ clearly surfaces. For it turns out, says the Visitor, that argument (2) is incoherent because its conclusion (namely, ‘what is not is inexpressible’), simply in virtue of its grammatical form, attributes being and unity to what is not. The modern notion of existence was defined in contradistinction to the sense of the copulative use of the verb ‘to be’. In fact, on Crivelli’s own interpretation, Plato never distinguishes between the existential and the copulative verb ‘to be’ in the Sophist; moreover, as he acknowledges, the complete and incomplete uses are semantically close in Plato’s usage. Finally, ‘non-existence’ and ‘not being anything at all’ seem to be inter-entailing. In view of these considerations, there seems to me little point in insisting on a purely existential reading of the puzzles.
Crivelli argues that the puzzles about falsity are also based on the impossibility of saying what does not exist. In the description of falsity in terms of ‘saying things that are not’ the latter phrase either directly means nonexistent items, or the argument fallaciously substitutes negations of existence for some other use of the negated verb ‘to be’. According to Crivelli’s interpretation, the latter strategy is adopted in the puzzle about images.
Chapter 3 (Puzzles About Being) treats the stretch of argument about previous ontological theories more briefly. I missed here an explanation from Crivelli of why it is necessary to raise the question about being along with the problems about not-being.
Chapter 4 (The Communion of Kinds) discusses the first part of the constructive metaphysical argument. In the proof of the distinctness of the five ‘great kinds’, being, change, stability, identity, and difference, the sentence ‘change is stable’ is treated as false.2 This is problematic, because change as a kind must be stable, and the statement rejected should be accepted as true. Crivelli’s solution is that the sentence must be interpreted in terms of its definitional reading (‘change is in virtue of its own nature stable’ – an obviously false claim) rather than in terms of ordinary predication (‘change instantiates stability’).3 He suggests, following an interpretative tradition started by Michael Frede,4 that the ontological distinction between ‘things that are said to be on their own’ and ‘things that are said to be relatively to something different’ (255C14-15) is also based on the distinction between ordinary predicative sentences and definitional sentences. In Crivelli’s version, in affirmative definitional sentences the subject and predicate term signify the same kind, and the predicate offers a complete description of the nature of the kind signified by the subject term. The second condition sets apart definitional sentences from identity statements.5 Another consequence is that predications of the genus, like ‘angling is an acquisitive art’, do not qualify as definitional sentences (127, n. 80). Crivelli fails to explain what kind of sentences, then, predications of the genus are and how they relate to definitional sentences.
255E-257A contains an argument leading to the conclusion that ‘not-being’ is necessarily applied to every kind, including being, because each of them is different from every other kind. The Visitor examines pairs of affirmative and negative sentences with the same subject term (‘change’) and predicate terms. He argues, for instance, that ‘change is identical’ and ‘change is not identical’ are both true, since the two sentences are to be understood differently. The affirmative members of the pairs are doubtless normal predications expressing that the subject participates in the kind signified by the predicate. As for the negative members, Crivelli classifies them as negative definitional sentences. He considers the interpretation in terms of negative identity statements a viable (although less economical) option (p. 166). Change is not-identical ‘because of its communion with the different, whereby, being separated from the identical, it becomes not that but different’ (256B2-4). If this remark is not meant merely as an argument for the truth of the sentence ‘change is not-identical’ but as a statement of its meaning that traces back negation to participation in difference, then the interpretation in terms of negative identity might be closer to Plato’s intentions than Crivelli’s definitional reading.
Chapter 5 (Negation and Not-Being) scrutinizes 257B1-258C6, where the Visitor turns, according to Crivelli, to negative predication. Crivelli argues for an extensional interpretation of negative predication at 257B1-C4 (‘x is not F’ = ‘x is different from everything that is F / an F’). In 257C5-D13, ‘the parts of the different’ (for example, ‘not- beautiful’, the part of the different set against beauty) qualify, in Crivelli’s view, as kinds in the full sense. He argues that ‘not-being’ defined at 257D14-258C6 is a collection of indefinitely many kinds like ‘not-being-beautiful’ and that it is explained in terms of difference from everything that falls under ‘being (an) F’, which amounts to an analysis of incomplete uses of the negated verb ‘to be’. The interpretation developed in this chapter is coherent and well- argued, but it cannot be regarded as cogent in every respect, because the text is underdetermined to such an extent that even the widely accepted assumption that Plato’s aim here is to give an account of negative predication can be reasonably called into question.6
Chapter 6 (Sentences, False Sentences, and False Belief) identifies two moves in the solution to the problem of falsity. (1) Crivelli emphasizes that, in Plato’s account, a sentence is not directed to a single object (like propositions in modern theories). Instead, it is directed to two things, an object referred to by the name contained in the sentence and an action (a kind) signified by its verb. The object and the action are put together in a single act of saying like the violin and the bow, but they do not form a new composite entity. (2) A false sentence says what is not about an object to be about that object, but this does not mean that the action in question does not exist. Take the false sentence ‘Theaetetus is flying’. ‘Flying’ is not about Theaetetus, but this means only that flying is different from every kind that is about Theaetetus. The upshot is that false sentences need not be directed to non-existent objects. Crivelli also examines whether and how this analysis can be extended to negative false sentences.
Let me finally make some more general comments. Crivelli refrains from going into broader questions like the historical context of the problems raised in the Sophist, or its relationship to Plato’s earlier ontology or to the treatment of falsity in the Theaetetus. That seems to me a reasonable decision, given the advanced state of the scholarly debate about the Sophist’s argument itself. Less fortunate is that he does not offer a general philosophical discussion of the results of his exegesis, remaining as he does within the confines of a commentary. 7
If we read Crivelli’s monograph as an interpretation of the inner core of the Sophist, we might complain about a certain imbalance: the discussion focuses on the problem of falsity, while the ontological theory advanced to make a solution possible remains in the background. Of course, it is perfectly legitimate to focus on the problem of falsity and not-being. But even so, we might wish to see what the connection is between this complex of problems and the examination of being. An alternative interpretation of the initial puzzles about not-being might be helpful here. In my view, the fundamental assumption governing the argument is that being and not-being cannot be co- instantiated. The puzzles are designed to show that this assumption is problematic and must be qualified. A joint examination of being and not-being is needed because the problematic premiss concerns both not-being and being.
My overview obviously cannot do justice to the subtlety and richness of Crivelli’s analysis; moreover, it frequently highlights points where I see room for questions. Crivelli’s monograph synthesizes the solid results emerging from the vast research literature produced in the last five decades on the Sophist. He soberly chooses among the interpretative possibilities, never yielding to speculative interpretations. The virtues of his account are comprehensiveness, detailed and clear presentation, and the philosophical coherence of the interpretation embraced. The book is excellently produced, with an index of names and passages cited. I recommend the book to philosophers, classicists, and graduate students – to anyone who is seriously interested in this masterpiece of the late Plato.
1. I have in mind the occurrence of the word onta, ‘things that are’, at 234D6 and 235A2.
2. 255A4-B7. The same problem arises in other passages as well.
3. The generalizing reading of the sentence in question (‘everything that instantiates change also instantiates stability’) seems to work as well. Crivelli rejects it because in the proofs at 255D-E normal predications also play a role (p. 122). This is not fatal, provided that the alternation of the sentences of different types is systematic.
4. Prädikation und Existenzaussage (Göttingen, 1967).
5. So ‘the good is the most highly praised kind in the Republic’ expresses a true identity statement, but it does not have a definitional reading. A further difference is that the subject of identity statements can be kinds and particulars alike, whereas definitional sentences relate exclusively to kinds.
6. Job van Eck, ‘Falsity without Negative Predication’, Phronesis 40 (1995), 20-47.
7. The Introduction merely summarizes the exegesis, and the Appendix, a formal presentation of the theory of false and true sentences attributed to Plato, makes for difficult reading for non-logicians.