1. Johansen begins by conceding that opsis can be used by Aristotle to mean the act of seeing, citing Bonitz' Index Aristotelicum, which gives passages in Historia Animalium in which the term evidently has this meaning. But what is at issue, he says, is whether it can naturally mean this in the specific context of DA.
2. He then points out that, in DA, Aristotle explicitly uses the term to indicate the faculty of sight and contrasts this with horasis, a term which Aristotle reserves for the act of seeing. Johansen maintains that (outside the passage that is in dispute) opsis is never used in DA to mean an act of seeing. Johansen cites 2.1.412b27-413a2 and 3.3 428a6-7. (He might also have cited--which, strangely he doesn't-- 3.2,426a13-14, which occurs in the very same chapter!) Here are all three pasages:
DA 2.1.412b27-413a2. As, then, the cutting of the axe or the seeing of the eye (horasis) is full actuality, so, too, is the waking state; while the soul is actuality in the same sense as eyesight (opsis) and the capacity of the instrument. The body, on the other hand, is simply that which is potentially existent. But, just as in the one case the eye means the pupil in conjunction with the eyesight (opsis), so in the other soul and body together constitute the animal. (Hicks)
w(j me\n ou}n h( tmh~sij kai\ h( o3rasij, ou3tw kai\ h( e0grh&gorsij e0ntele/xeia, w(j d' h( o1yij kai\ h( du&namij tou~ o)rga&nou, h( yuxh&: to_ de\ sw~ma to_ duna&mei o1n: a)ll' w3sper o)fqalmo_j h( ko&rh kai\ h( o1yij, ka)kei= h( yuxh_ kai\ to_ sw~ma zw|~on.
DA 3.3.428a6-7. Sensation (aisthesis) is either a faculty like sight (opsis) or an activity like seeing (horasis).
ai1sqhsij me\n ga_r h1toi du&namij h2 e0ne/rgeia, oi[on o1yij kai\ o3rasij
DA 3.2.426a13-14. The actualization of sight(opsis) is called seeing (horasis).
o3rasij ga_r le/getai h( th~j o1yewj e0ne/rgeia
3. Johansen then additionally gives a Bayesian argument to the effect that, given what Aristotle says in DA, the term opsis is not what one would expect, if Aristotle had wanted to indicate an act of seeing: "Given this precedence, if Aristotle meant to refer to the actuality of sight at DA III.2 425b13, it is quite unlikely that he would use the word 'opsis'."