Is the 'default meaning' of opsis at DA 3.2 425b12-17 a capacity or faculty of sight, as Johansen claims, or can the term just as well mean an act of seeing, as Caston maintains? (Why is this important? Presumably: if Johansen is correct, then we should want to read the passage as talking about whether we need to postulate an additional faculty, to explain our ability to perceive that we see, and it seems that the passage is less plausibly taken to be concerned about the structure of consciousness, as Caston holds.)
It seems to me that Johansen is correct on this point:
1. Although Caston and Johansen both agree that opsis can by used by Aristotle to mean an 'act of seeing', they also both agree that, when used in contrast with horasis, the term means the faculty of sight. Yet it is clear that opsis is used by Aristotle in contrast with horasis throughout DA and especially in DA 3.2. This is shown not merely by the passages that Johansen cites, but also by a consideration of the usage of opsis in DA generally: a TLG search shows that of the approximate 30 occurrences outside the passage in dispute, the term is used consistently by Aristotle to mean the faculty of sight.
2. Even apart from the particular context of DA, it seems too strong to say, with Caston, that opsis for Aristotle "remains unmarked and can signify either a capacity or activity". It is unmarked syntactically, to be sure, but Aristotle typically uses the term for a capacity. Again, a TLG search reveals some 400 occurrences in the corpus. I won't be able to check Bonitz until later, but a correspondent wrote:
Bonitz lists for opsis as 'ipsa actio videndi' 369b9, 1118a3,16, 1167a4, 1230b26-1231a17, 503a34,b1, 872b13, 1011a28, 494b33, 680a3, 478a35.
It's not even clear that all of these are accurately classified, e.g. 1118a3 is the discussion in NE as to whether the virtue of moderation deals with pleasures derived from the sense of seeing (dia_ th~j o1yewj, as peri\ th_n a)koh&n at a7, means 'in the field of hearing', as Broadie and Rowe have it); likewise also the parallel from EE. 872b13 can be dismissed as from the Problemata. Etc.
A correspondent furthermore pointed out that opsis is used in contrast with horasis, as if a stock example of the distinction between an actuality and a capacity, to be taken for granted, in the following passages:
Met 1050a24-25. e0pei\ d' e0sti\ tw~n me\n e1sxaton h( xrh~sij (oi[on o1yewj h( o3rasij, kai\ ou)qe\n gi/gnetai para_ tau&thn e3teron a)po_ th~j o1yewj)...
EE 1219a17. tw~n me\n ga&r e0stine3tero&n ti to_ e1rgon para_ th_n xrh~sin, oi[on oi0kodomikh~j oi0ki/a a)ll' ou)k oi0kodo&mhsij kai\ i0atrikh~j u(gi/eia a)ll' ou)x u(gi/ansij ou)d' i0a&treusij, tw~n d' h( xrh~sij e1rgon, oi[on o1yewj o3rasij kai\ maqhmatikh~j e0pisth&mhj qewri/a.
3. Although Aristotle does say in DA that words for the senses and sensibles admit of being used either for the capacity or for the actualization, it seems incorrect to conclude from this, as Caston does, that DA 3.2 itself "contains a fundamental ambiguity in its terminology" and that therefore we may take a term like opsis as a capacity or actuality, as it suits us. The reason is precisely that Aristotle castigates his predecessors for using these terms without appropriate qualifications (ou) kalw~j e1legon, a)ll' e0kei=noi a(plw~j e1legon peri\ tw~n legome/nwn ou)x a(plw~j); and he draws attention to horasis as a word that exists precisely to avoid that sort of inaccurate and misleading ambiguity in the case of sight: o3rasij ga_r le/getai h( th~j o1yewj e0ne/rgeia. So Caston would have Aristotle not paying attention to any of this, precisely in the midst of Aristotle's warnings to pay careful attention. It doesn't follow from the fact that others use the terms without due care, that Aristotle does.
4. Johansen's Bayesian argument seems to me unanswerable. Given that Aristotle has been careful to draw a sharp distinction between the actuality and the capacity of vision; given that he has draw attention to a word (horasis) that exists precisely to distinguish the actuality; and given that he has made point (as we have seen) to reserve opsis for the capacity; then, if he had wanted to talk about the actuality, as Caston claims--in a context where he has been otherwise been investigating faculties of sensation, as Caston admits--he would have used the word for the actuality (horasis) or other phrases, such as verbs or infinitives, more clearly suggestive of the actuality. (On this last point compare NE 9.9.1170a33-b3, where Aristotle wishes to talk about activities, not faculties.)
5. And then there is an additional consideration, not raised in my earlier posts: Caston agrees that in 425b17-20 (the second half of the passage on 'perceiving that we perceive'), opsis should be understood as referring to the capacity or faculty. (That is why Caston has no distinctive translation for those lines; he is in basic agreement with Johansen there.) But the parallelism between the first and second half of the passage, and a back-reference, seem to require that both halves should be read in the same way.
Here is the argument, spelled out:
(i) At 425b18, 'to perceive by vision' (th~ o1yei ai0sqa&nesqai) means 'to perceive by the sense of sight.Of course, if we accept this resolution, then this is just to say that one would naturally take opsis in 3.2 to be about the faculty.
(ii) But 425b18 states an aporia meant to affect the conclusion just reached.
(iii) Therefore, 425b18 is meant to express the same thing as what was just considered (indeed, that is why the construction is exactly the same as at b13).
(iv) Thus at 425b13 too, 'to perceive by vision' (th~ o1yei ai0sqa&nesqai) means 'to perceive by the sense of sight'.
But Caston claims that the passage simply cannot be read in that way--that the regress argument and Aristotle's use of reflexives make no sense if the passage is taken to be about the faculty. And in that case maybe we have to say that Aristotle does, after all, fall prey to the same confusion of which he accuses his predecessors.
I'll have to turn to these arguments of Caston, and Johansen's rejoinders, next.