I want to call attention to Johansen’s understanding of one line, which I put in bold:
‘Since we perceive that we see and hear, it is necessary [b12] that one perceives that one sees either by sight (opsis) or by some other [sense]. But the same [sense] will be [b13] of sight and the underlying colour, so that either there will be two [senses] [b14] of the same thing or it [the sense] will be of itself….
As we saw, Johansen (in an “interpretative note”) remarked that the highlighted line is meant to apply to both the option that we perceive that we see by sight, and the option that we perceive that we see by some other sense. This seems reasonable. Since seeing consists of a sense faculty receiving a perceptible form, then a perceiving of this (however accomplished) will somehow have to take both of these under its scope.
In contrast, if we take the highlighted line to apply to only one option, say, the second, then various difficulties arise. There’s a lesson in this, as I shall explain.
Recall that I was puzzled by how, if capacities and activities are correlative, a ‘capacity reading’ could give a different result from an ‘activity reading’.—Presumably they cannot imply different results, but a philosophical view will be more naturally suggested by the one than by the other.
Nonetheless, that reminder, about how these views must in some sense be equivalent, serves as a useful check. It should lead us to examine, in the case of any objection raised against the one interpretation, whether it doesn’t also apply to the other interpretation; and, if there is a device that the one interpretation can usefully exploit, whether it can’t also be exploited by the other—because we would expect a fundamental parity between the two.
So, for instance, we saw in an earlier post that Caston wanted to maintain that a capacity reading of DA 3.2 was prima facie inconsistent with De Somno 2, and he presented this as if it were a difficulty peculiar to it. --But then later we saw that Caston effectively conceded that this was just as much a difficulty for an activity reading; and we saw that the same sort of harmonization that Caston suggested for the activity reading, could be used for a capacity reading as well. So on that point there is actually parity between the two views.
Yet the same thing happens here, in connection with the line I’ve highlighted above. When Caston gives an exposition of the capacity reading (on p. 765) he takes the highlighted clause to apply only to the option that it is a different sense from sight which perceives that we see. This leads to a variety of difficulties, which Caston draws out and appears to treat as if they applied only to the capacity reading. Yet later (on p. 771), when giving an exposition of the activity reading, Caston construes the line as applying to both options: “perceiving that we see is a perception both of our seeing and of the object seen”, he says. Yet that way of dealing with the line was also available to the capacity reading (as Johansen shows).
Presumably Caston took the capacity reading in that way because many of its defenders have done so. As he notes in footnote 33 on p. 765: “The restriction to cases where there is a distinct sense is not actually in Aristotle’s text, but is supplied by many translations and commentaries…The translations of Hicks and Smith are more faithful to the text; they represent the inference as following instead from the initial premise—namely, that we perceive that we see—which is neutral between the options Aristotle is considering”, and then he cites his own page 771 (and Osborne) as taking that better approach.
That’s all fine: but then one needs to add that the various objections Caston raises on p. 765 are an artifact of a faulty way of developing the capacity reading. In this respect, there’s parity between the interpretations.