Martha C. Nussbaum and Amélie Oksenberg Rorty (edd.), Essays on Aristotle's De Anima. Oxford: Clarendon Press, 1992. Pp. viii + 439. ISBN 0-19-824461-4 (hb); 0-19-823600-X (pb).
Reviewed by Cass Weller, University of Washington.
This volume is without doubt a valuable collection of fine essays well representative of the range of topics in Aristotle's De Anima. Many of the essays, though, will be of far greater value and interest to those already initiated in the mysteries of a certain scholarly controversy: Did Aristotle subscribe to a theory of mind which broadly anticipates contemporary functionalist materialism? There is a perennial temptation to rediscover, or shall we say refresh one's interest in, figures in the history of philosophy by illuminating their dark texts with the brightest burning lights of the day. (I wonder if any functionalist interpreter of Aristotle thinks functionalism is false.) It is the contention of the flagship article of this collection that many have succumbed to this temptation regarding Aristotle's philosophy of mind by seeking, misguidedly, to father on Aristotle a contemporary solution to Descartes' mind-body problem. Those scholars who find their pursuit of Aristotle vindicated by the eminently reasonable and philosophically respectable things he turns out to say are always disturbed to hear one of their own denounce his views as literally incredible. This is what M.F. Burnyeat in the role of provocateur has done.
Some might think that my referring to his article, "Is Aristotle's Philosophy of Mind Still Credible: A Draft," as the flagship article of this collection gives it undue prominence. However, of the eighteen articles written after 1978 (there are twenty papers in the collection) I count at most only six untouched by the concerns of Burnyeat's influential paper which has circulated in typescript since 1983 (it is cited in pre-publication form by five authors in this collection) and which he publishes here with great reluctance and some annoyance. In large part this volume represents the efforts of the editors, Martha C. Nussbaum and Amélie Oksenberg Rorty, to provide a permanent public record of a debate which until now those without could only judge to be of moment from the noise and clamor within. Consequently, the ensuing discussion will be rather narrowly focused on this debate and make little if any mention of the many valuable essays in this collection with other concerns.
Part of the difficulty for a reader is just sorting out the players and their evolving positions on this particular dominating field of play; and I promise no charts. To get some idea of what the reader must confront consider the following. Burnyeat attacks the materialist functionalist interpretation under the description 'The Putnam-Nussbaum thesis'. In response Putnam and Nussbaum complain, in a philosophically bold and sweeping yet rather rhetorically bizarre essay, that Burnyeat is hostage to bad metaphysics and misrepresents their view or at least pays no mind to its more evolved version. On their view Aristotle succeeds in sustaining both the explanatory priority of the mental/intentional and its natural and organic unity with its constituent matter. Negatively put, he (Aristotle) preempts what they as professing fin de siècle Aristotelians eschew as the reductionist and bad metaphysical tendencies of functionalism. He thus, properly understood in their terms, fulfills Wittgenstein's desire to have a natural history of man (p.56). This makes for an odd reversal. Burnyeat denounces Aristotle's psychology as unworthy of credence precisely because it fails to be a version of contemporary functionalism or at least because of the way it fails. Putnam and Nussbaum defend Aristotle's philosophical psychology as the humanist way of the future precisely because of the way it is not a version of fast fading contemporary functionalism. Sic transit. From another quarter Cohen attacks Burnyeat's rival anti-materialist interpretation and, in the name of defending a materialist functionalist interpretation, argues for Sorabji's claim of 1974 that receiving sensible form without matter is to be understood as literal qualitative assimilation. It is ironic to note that Sorabji, however, now, even as he redoubles his efforts on behalf of literal qualitative assimilation, openly regrets (p.208, 210n) any inadvertent encouragement his paper of 1974 may have given to functionalist interpreters. He names Cohen as one of the misled.
A large part of the difficulty of getting a handle on the debate, aside from its fluid character and claims and counterclaims of misrepresentation, is the extreme character of Burnyeat's view which goes beyond what is needed to keep Aristotle out of the functionalist camp. The central point he makes is that to be a materialist functionalist regarding the mind one must subscribe to a minimally Cartesian conception of matter whose nature does not include thought. And Aristotle doesn't because he holds on the contrary that it is of the nature of some material bodies, i.e., animal bodies, that they perceive, desire, and think. A dog's body or its matter is essentially alive and percipient. Thus Burnyeat's Aristotle is a vitalist, to use the older terminology in which he has been contrasted with mechanist philosophers. Well and good. But Burnyeat also holds that according to Aristotle an act of perception may occur in an animal without any collateral physiological change in the percipient body. It is this latter claim that has drawn the most fire from Burnyeat's critics. It is largely the focus of the attack by Putnam and Nussbaum as well as Cohen.
The paper by Moravcsik and Code, "Explaining Various Forms of Living" in this volume, however, avoids this particular distraction by concentrating on Aristotle's lack of a conception of matter suitable for functionalism. Therefore, before taking up Burnyeat's case against functionalist interpretations it might be helpful to look at the relatively unencumbered negative case they make and then the alternative view of Jennifer Whiting. But first I want to set out in broad terms the view Aristotle would have to subscribe to in order to be a functionalist materialist.
There are creatures who think, perceive, and desire and they are composed of material and only material bodies. There is no immaterial mind-substance inside them that ultimately does their thinking, perceiving, and desiring. No state, process, act, or event involving only material bodies or objects composed only of material bodies is such that a token of that type, i.e., an occurrence of it, even though a case of thinking, perceiving, or desiring, is necessarily a case of thinking, perceiving, or desiring. Thinking, perceiving, and desiring really occur but are not definable in a purely physicalistic or non-intentionalist behaviorist vocabulary. The capacity for thought, perception, and desire only contingently characterize the physical systems that exhibit them. Thinking, perceiving, and desiring are functional states of the creatures they occur in on analogy with artifacts being functional objects. Just as door knobs qua material compositions are not necessarily door knobs and can be made of different materials (they are compositionally plastic), so too episodes of thinking, perceiving, and desiring qua constituting material events are not necessarily episodes of thinking, perceiving, and desiring.
Code and Moravcsik make the central point that materialist functionalism offered as a solution to the mind-body problem presupposes a Cartesian notion of body in the broad sense of extended substance none of whose states is essentially mental. The idea of functionalism is to eliminate mind as a substance-kind and accommodate mental phenomena as functional states to the states and processes of only those entities acknowledged by Cartesian or post-Cartesian physics. They argue, rightly in my opinion, that Aristotle can't be prefiguring such a solution to the mind body problem because he doesn't have the appropriate Galilean/Cartesian concept of body. He has no univocal notion of ubiquitous matter whose nature is everywhere the same and includes neither being alive nor being percipient. In contrast to such a conception just consider Aristotle's that includes on the one hand the sublunary material elements ingredient in corruptible organic animal bodies and, on the other, the special stuff composing incorruptible immortal heavenly bodies.
The only dissenting quibble I would register is that their historical explanation of why Aristotle's psychology is incompatible with materialist functionalism tends to obscure the logical incompatibility of the two positions. Baldly put, Aristotelian subjects of cognition intrinsically possess the power of cognition; functionalist subjects of cognition do not. Further, the occurrence of an Aristotelian mental act is such that it is essentially mental while the occurrence of a functionalist mental act is not. We don't need to mediate the inference from functionalism to its incompatibility with Aristotelian psychology through the latter's incompatibility with the Cartesian conception of matter.
Here is a slightly different way of putting the point that brings into relief the two related but distinct incompatibilities. Functionalism with respect to the mental is compatible with any ontology whatsoever (even an ontology of immaterial souls) just so long as the power of cognition belongs essentially to no object of that ontology, and no occurrence of an event (even an event in the soul) is such that it is essentially a mental event (an event with mental aboutness, for example). On Aristotle's theory, however, sentient bodies are essentially sentient and the particular occurrence of an act of perception is such that it is both essentially mental, pace functionalists, and essentially material, pace Descartes, albeit in a suitably Aristotelian sense. That is, the body of a sentient organism is, on Aristotle's view, essentially percipient and its acts of perception are essentially mental as well as essentially material. Therefore, Aristotle's theory entails i) that animal bodies are not Cartesian material substances and ii) that functionalism is false.
Thus a central problem confronting functionalist interpreters, as we have seen, is one of establishing the contingency of the vitality and mentality of an animal's constituent matter. Once it is granted that the organic body of an animal is essentially ensouled, i.e., alive and percipient, another material body must be found which is not essentially alive and percipient upon which the organic one is founded, something continuous with the living body and its corpse; otherwise Aristotle's view won't be compatible with materialist functionalism. An earlier version of this problem that predates the current functionalist controversy took the form of a complaint issued by Ackrill.1 Led by Aristotle's use of artifacts, e.g., statues and spheres made of bronze, as illustrations of the relation of form and matter Ackrill supposed that the matter of any object of a sort was only contingently informed by the form corresponding to that sort. Thus, since the body of an animal of some kind is essentially informed by a soul/form of that kind, Aristotle, Ackrill claims, inconsistently applies his form-matter analysis to animals, and to living things in general. More recently Bernard Williams,2 endorsing Ackrill's criticism, has called this the body-BODY problem of hylomorphism. It is what Jennifer Whiting ("Living Bodies," chapter 5) sets out to solve on behalf of functionalist interpreters of Aristotle.3 Her view, as I understand it, commits her to defending, within Aristotle's system of concepts, the intelligibility and contingency of such statements as 'This eighty five pounds of earth, air, fire, and water is a German Shepherd' and 'The earth in the tissues of this German Shepherd is alive'. She grants that the homoiomerous tissues ingredient in the working organs and other anhomoiomerous parts of a living animal are essentially ensouled but argues that the term 'flesh', for example, is ambiguous. Sometimes it designates one of the essentially ensouled functional matters of the organic body and sometimes one of the contingently ensouled compositional matters of the organic body, where a compositional matter is just so much earth, air, fire, and water. Here is what I hope is a fair representation of her argument (pp.76-85).1) The elementary natural bodies of earth, air, fire, and water retain their original character even as they constitute the homoiomerous tissues of the living body, i.e., the earth, air, fire, and water in the tissues of a living animal, however mixed and combined, are portions of the same stuffs other portions of which can be found in unmixed lifeless heaps.One might question various aspects of this argument, for example, the acceptability of (2) and antecedently whether it means that every bit of earth in the tissues of the German Shepherd is alive or only that this eighty five pound mixture of earth, air, fire, and water is alive. One might also wonder how, on this interpretation, one avoids turning Aristotelian substances into accidental unities. That is, how does one avoid turning living organisms into functionally defined entities that supervene on ontologically prior material kinds (especially since this is exactly what functionalists want)? On the view under consideration it looks like being a dog is an accidental property of the package of earth, air, fire, and water which is the dog's compositional matter.
2) The portions of these four elements found in a living animal are characterized contingently by the properties which necessarily belong to the organic body of the living animal.
3) The portions of the four elements found in a living animal make up the inorganic compositional matter as opposed to the functionally defined matter of an organic body.
4) Therefore, every living animal has both an organic body which is essentially ensouled and an inorganic one which is not and can thus serve as the body whose career is continuous with the life of the animal and with its lifeless corpse.
Whiting addresses this problem with ingenuity, though without success, in my opinion. Her way of preserving the integrity and intrinsic unity of an animal so that its form is not an accidental property of persistent inorganic matter is twofold. [What follows is more my interpretative reconstruction rather than a rehearsal of her explicit reasoning.] First she turns the tables on matter and observes, in good Aristotelian fashion, that matter is to play the role of predicate not subject. Thus, Rin-tin-tin is earth-en inter alia -ens. The other move is legislative and is presupposed by the first move. We must have a distinction between accidental changes in underlying matter and substantial generation de novo of individual members of biological kinds. Therefore, we are to treat "a form's coming to be embodied in some matter (inorganic matter I presume) as yielding a product which is not a property of what persists and which is itself an intrinsic unity -- that is, a unity neither component of which is separable from the other in a way such that it could serve as subject in some other unity" (p.86). But this product is a functionally defined compound of an organic body and its soul, "the set of capacities in virtue of which the organic body is capable of performing its functions" (p.87). Thus, qua functionally defined individual a living organism can be viewed as standing as subject to its inorganic matter: This dog is earthen (because its flesh is earthen). But now the game is up. She has avoided turning the emergence of individual dogs into nothing more than accidental changes in persistent matter by stipulation. An Aristotelian individual substance is a functional entity whose functional properties belong to it and to its organic body essentially. Such an individual cannot be construed as a property contingent or otherwise of persistent inorganic matter. In addition to what she calls a thick compound, "a compound of form and the portion of compositional (inorganic) matter constituting it at a given time" (p.87), which is surely a case of form being accidentally predicated of inorganic matter,4 we also have what she calls a thin compound, the functionally defined intrinsic unity of soul and organic body. But now I have several related questions. Why isn't a thin compound some kind of logical construction upon thick compounds? Why aren't thick compounds ontologically more basic that thin ones and why isn't the inorganic component of a thick compound ontologically independent of and therefore ontologically more basic than the compound of which it is a component? This is to repeat my earlier question: How if Aristotle's view is consistent with materialist functionalism can he maintain the ontological priority of his favored individual substances and their forms over the inorganic material elements in the universe?
Let me conclude this section by making the same point in abstraction from the details of her particular account. I propose to let the case for functionalism stand or fall on the question of whether it is a contingent fact that portions of earth, air, fire, and water regularly find themselves in the essentially ensouled tissues of living organisms.
Functionalist interpreters must accept, on Aristotle's behalf, the contingency of this fact insofar as functionalism is a form of materialism. But this requires countenancing a world in which earth, air, fire, and water exist in their elemental characters but are nowhere constituents of living organisms. I am not directly concerned with the difficulties that the necessary existence of the species poses for functionalist interpreters but with the idea of earth, air, fire, and water having the kind of ontological independence typically ascribed to Democritean atoms and essentially ascribed to Cartesian matter. Can Aristotle really deny that it is of the nature of the four elements that portions of them are the recyclable ultimate constituents of living organisms? The difference I am assuming between Aristotelian matter and Democritean matter is that Aristotelian matter exists for the sake of form and is to be thought of as material subject to the designs of craftsmanship. Thus a world of unwrought heaps of earth, air, fire, and water with no further prospects would be radically incomplete and unintelligible.
The fact that each of the elementary bodies also admits of a form-matter distinction poses another, perhaps the ultimate, difficulty for materialist functionalist interpretations of Aristotle. Consider the following inconsistent set of propositions.(1) Every compound individual of the form F is F [e.g., Every dog is a dog.]Aristotle is clearly committed to (1), (3), and (5). Whiting's Aristotle also subscribes to (2) and (4).
(2) Every compound individual of form F is F only if its compositional matter is contingently F and essentially something else.
(3) Every determinate individual member of the kind F is a compound of the form F and matter.
(4) Portions of earth are determinate individuals of the kind earth.
(5) The matter of a portion of earth, while contingently a portion of earth, isn't anything essentially.
I now turn to Burnyeat's case against functionalist interpretations. His general claim is the one we have been considering, that Aristotle's conception of the physical, because it allows that some bodies qua body are essentially ensouled and possessed of psychological attributes, is incompatible with the one required by functionalism. He focuses on Aristotle's account of sense perception to show that animal bodies are essentially percipient and thus not physical bodies of the sort required by materialist functionalism. He argues that for Aristotle taking on the sensible form F-ness without matter (De Anima, II.12) and the sense organ's becoming like what is perceptibly F (De Anima, II.5) is nothing more nor less than a sensory awareness of F-ness. And this, he claims, has implications which are incompatible with functionalism.
He is particularly eager to argue against Sorabji's influential interpretation of these doctrines. Sorabji thinks that they concern only the physical side of sense perception. When a normally sighted animal looks at something red in normal conditions, the jelly in its eyes turns red. Thus Sorabji takes these doctrines as quite literally describing the qualitative assimilation of perceiver to object perceived. Now Burnyeat contends that these doctrines, whatever their content, describe the most basic level of interaction between a perceiver and its object. He, therefore, concludes that if, as he hopes to show, the doctrines in question are to be interpreted along intentionalist lines and thus Sorabji is wrong in interpreting them in terms of literal as similation, then Aristotle leaves no room for the physiology of perception. In the case of perception, sensory awarenesss would be the formal side but there would be no material side parallel to the physiological processes pertaining to the material side of emotional states such as anger. If Sorabji is wrong, perception isn't even contingently physical (though it requires a sentient body); and if not even contingently physical then not amenable to a functionalist interpretation. It is from this standpoint that Burnyeat attacks Sorabji's interpretation as essential to functionalist interpretations.
Since I don't share Burnyeat's all-or-nothing assumption about the prospects of a material account of perception in Aristotle, I don't regard Sorabji's view as crucial for functionalist interpretations, though I agree that it is quite mistaken. In particular, one could interpret the doctrines in question generally along the intentionalist lines favored by Burnyeat and argue that the intentional character, the mental aboutness, of perception is formal rather than material and conclude that since the formal is the functional and the functional is, well, functional, the intentionality of an act of perception only contingently characterizes the material state that happens to realize it. The fact that Aristotle rules out the Platonic/Cartesian conception of an immaterial soul some of whose states are intrinsically intentional helps to makes this line of thinking, to my mind, the most attractive version of the functionalist interpretation. Since functionalist interpreters are already under obligation to establish in general that the materials in which psychological attributes are realized are only contingently so characterized, they lose nothing and only gain if the disputed doctrines fall on the formal/intentional side and not on the material/physiological side. The less Aristotle has to say about the physiology of perception the fewer the apologies that have to be made for him. The more he says to register his recognition of the intentionality of cognitive states the more he looks like someone with a functionalist's concerns.
I don't endorse this view. But it seems to me that functionalist interpreters needlessly weaken their position by giving ground here and supposing that an intentionalist reading of the disputed doctrines is incompatible with their view. Perhaps, though, they sense that if receiving the sensible form, F-ness, of an object is understood as simply perceiving it as F their case is lost for the following reason. Once Aristotle calls attention to the fact that perception involves some version of classificatory awareness it's highly unlikely that he would be willing to regard such awareness as anything but essential to each occurrence of perception as the very thing that it is.
The best way to begin to assess Burnyeat's view and a specimen response to it is to bear in mind the following:A theory of mind is a functionalist theory of mind only if it entails that no token of the type, i.e., no occurrence of, thinking, perceiving, or desiring is necessarily an occurrence of thinking, perceiving, or desiring.
According to Burnyeat's rival interpretation of various Aristotelian doctrines pertaining to sense-perception, the occurrence of an act of sense-perception is nothing more nor less than an act of sensory awareness, is identical to the occurrence of no physical process or event, and does not require any physiological change in the sense organ in which it occurs.
Cohen helpfully focuses on the logical relations between the following claims of Burnyeat's rival interpretation of Aristotle.5i) A sense-organ's taking on a sensible form is an act of awareness rather than a physiological change.(I shall continue to assume throughout that what goes for a sense-organ's taking on a sensible form goes for a sense-organ's becoming like its object.) What Cohen says about the relations between (i) and (ii), and Burnyeat's understanding of them is illuminating.
ii) It is possible for perception to occur without any associated physiological change.Burnyeat uses (i) as the leading premise in his argument against the functionalist interpretation. It has solid (albeit disputed) textual credentials. (ii)'s credentials, however, are less clear, as is the relation Burnyeat supposes it bears to (i). He nowhere argues that (ii) follows from (i). His arguments are devoted to proving (i); then (ii) puts in a sudden appearance. This suggests that Burnyeat may have the following sort of argument in mind: perception is nothing more nor less than a sense-organ's reception of sensible form and the reception of form is not a physiological process. So since there's nothing more to perception than reception of form, it is possible for perception to occur without any corresponding physiological change. (p.63)
The strategy Cohen imputes to Burnyeat is one of first establishing (i), that receiving a sensible form is an act of sensory awareness and not a physiological change, then deriving (ii), that perception can occur without any associated physiological change, which is inconsistent with functionalism. On Cohen's assessment he fails to establish (i) and neglects to derive (ii) from it, which is just as well, since (i) doesn't entail (ii) anyway. So, for all Burnyeat has shown and in the absence of further argument, functionalist interpreters of Aristotle are free to accept or reject (i).
It would be fairer to represent Burnyeat as inferring (ii) from (i). He says as much, assuming, as everyone does, the equivalence of taking on sensible form and becoming like the sensible object (p.19). Now when Cohen claims that Burnyeat fails to derive (ii) from (i) and that (ii) is not a consequence of (i), he suggests, as we have seen, an argument Burnyeat may have had in mind. Thus(0) Perception is nothing more nor less than reception of sensible form.Therefore,
(i) Reception of sensible form is an act of awareness and not a physiological process.
[Therefore, perception is [sensory awareness] and not a physiological process.](ii) It is possible for perception to occur without any corresponding physiological change.Now in arguing that (0) and (i) do not license the inference to (ii) Cohen simply observes that while vision, to take a specific sense, may be identical to no physiological process, it does not follow that there is no physiological (physical) process at all essential to the exercise of vision. After all, each act of vision might require some physical realization or other even if there is no single kind of physical realization required for vision.
To begin, all parties can at least agree that the identity of perception and reception in (0) which establishes that the latter is the form or essence of perception can be reformulated as(0.1) Necessarily, x is an act of perception if an d only if it is an act of receiving sensible form without matter.Everything depends on how (i) is to be read. But it seems to me that there are readings of it other than the one, due to Cohen, we've been relying on, from here on (i.1). Here are three readings.(i.1) Taking on a sensible form without matter as a repeatable type is an act of sensory awareness which constitutes the formal/mental aspect of an act of sense-perception and is identical to no type of physiological process. The fact that Jones is now visually aware of red is not the same fact as the fact that his nervous system is now in state R. The former constitutes the formal side of Jones' seeing red, the latter the material side of his seeing red. On this view Jones' datable seeing of red admits of two identifying descriptions so that we might say that Jones' visual awareness of red is identical to his nervous system's being in state R in the troublesomely analogous way that this instance of sphericity is identical to this piece of bronze.(i.1) is, in effect, a parochial claim about which of Aristotle's doctrines are supposed to capture the mental/functional part of sense-perception and which the merely physical part. It leaves open the question of metaphysical dependence, that is, whether the token-physiological event is only contingently a mental act, or whether it is essentially an act of awareness. Put in other terms, (i.1) says only that the type Taking-On-Sensible-Form-Without-Matter is a mental type and not a physiological type. It is thus consistent with token-physicalistic functionalism, and does not imply (ii).
(i.2) Taking on a sensible form without matter is as a repeatable type an act of sensory awareness which constitutes the form or essence of perception and is identical to no type of physiological process. Moreover, sensory awareness is such that any act characterized by it is necessarily characterized by it. That is no act is only contingently an act of sensory awareness. And thus, while every act of perception requires a sentient body -- in fact, by definition, only occurs in a sentient body -- it is such that it is essentially an act of sensory awareness.
(i.3) Taking on a sensible form without matter is an act of awareness which is the form or essence of perception and is identical in no sense, either at the level of types or tokens, to any physiological change, process, or state of a material system, although, by definition, it occurs only in a sentient body.
(i.2) is a broad denial of physicalistic functionalism. It claims that token-acts of perception are essentially mental events. Physicalistic or materialistic functionalism, however, is committed to the thesis that tokens of mental event types are not essentially mental events; rather they are essentially the physical events with which they are identical. [Functionalism doesn't deny de dicto 'Nec (a mental event) is a mental event'; it denies de re that any mental event (token) is such that it is necessarily a mental event.] However, since (i.2) simply doesn't specify the relation of token-acts of sensory awareness to tokens of physiological types, it doesn't imply (ii). Moreover, insofar as physiological processes are essentially psychological (in Aristotle's sense), (i.2) is compatible with token-acts of sensory awareness being identical to occurrences of physiological events.
(i.3) is a form of immaterialism with respect to acts of sense and is clearly inconsistent with functionalism. Now if such immaterial acts were acts of an immaterial substance whose existence depends on no body, then presumably such acts of sensing could occur without any corresponding physiological change in the body of the sensing subject, though as a matter of contingent fact they don't. However, Aristotle's sensing subjects are material substances; so we'll need another argument, if we're to get from (i.3) to (ii). One might, nevertheless, be inclined to accept (ii) on the strength of (i.3) simply in the absence of a principle that metaphysically ties the occurrence of determinate acts of sensory awareness to the occurrence of determinate material processes in the sense organ. This would be a kind of burden of proof defense. Since each act of perception is in its very nature essentially mental and no physiological process is essentially mental, you have to tell me why acts of perception require, causally or metaphysically, corresponding physiological changes in the sense organs. One might, in casting about for an explicit deductive argument, help oneself to the background assumption that if token-events of kind M are identical to no token-events of kind P then an event of kind K could occur without a corresponding occurrence of an event of kind P. But this is questionable. A 4-6-3 double play in the fifth inning of a particular game last Tuesday, the particular occurrence of an event of kind D, may be identical to no physical event, to no occurrence of an event of kind P. -- Is every movement of the batter and the fielders included in the physical event or only the movement of the baseball from the time it leaves the pitcher's fingertips until it lodges in the first baseman's mitt? If the spin of the ball had been slightly different it would have been a different physical event; would it have been a different double play? -- And yet we believe this particular double play couldn't have occurred without the occurrence of some physical events or other, because we believe that at some level or in some sense physical events are constituents of the double play in virtue of the fact that baseballs and baseball players are physical objects.
What's a bit strange about trying to argue from a token-act of perception being identical to no physiological occurrence to the possibility of its occurring without any physiological change is that it reverses the order of a familiar line of argument. One is accustomed to hearing arguments for the possibility of a existing without b in order to establish that a is not identical to b, rather than arguments from non-identity to the possibility of a existing without b.
Moreover, in the case at hand, it seems clear that when Burnyeat claims that for Aristotle perception does not require the occurrence of a physiological change in the sense-organ he means something stronger than the mere logical possibility of perception occurring without a physiological change in the sense-organ; namely, that in fact perception, as Aristotle sees it, regularly occurs without physiological changes occurring in the sense-organ at all. Let's call this the strengthened version of (ii). What this suggests is an argument that draws on resources hearkening back to a claim Burnyeat makes in the opening paragraphs of his paper."...whatever the meaning of the phrase 'taking on form without matter', it picks out the most basic level of interaction between a perceiver and the object perceived." (p.16)
Here's an argument for the strengthened version of (ii), with an interim argument for (i.3), depending on the claim just quoted. I think Burnyeat would accept it.(1) Taking on sensible form without matter is the most basic form of interaction between perceiver and perceptible object. [first assumption of basic interaction]
(2) If taking on sensible form without matter is the most basic form of interaction between perceiver and perceptible object, then in perception if anything occurs in the sense-organ it is an act of taking on sensible form without matter. [second assumption of basic interaction]
(3) Each occurrence of taking on a sensible form F without matter is such that it is essentially an act of sensory awareness of F.
(4) No occurrence of a physiological process is such that it is essentially an act of awareness.
(i.3) Therefore, no occurrence of taking on sensible form without matter is a physiological process. [3,4]
Therefore, in perception what occurs in the sense-organ is not a physiological process, i.e., perception occurs without physiological changes in the sense-organs. [1,2,i.3]
With these options on the table let's return to Cohen's version of Burnyeat's implicit argument with (i.1), in particular, in mind. Grant that perception as a type of mental act is identical to reception of sensible form. It doesn't follow that it's possible for acts of perception to occur (in animals) without any corresponding physiological change. It doesn't follow because the identity of perception and reception, neither of which is identical to a physical type, is perfectly consistent with particularized cases of animal perception being token-identical with physiological processes. (Compare: Pain as a type of mental state is nothing more than a state of feeling that causes various behaviors. But it doesn't follow that particular cases of pain aren't token-identical to brain states.) If this is Cohen's argument, then he understands (i) as (i.1), the simple denial of type-type identity.
One can't help but suppose that in ascribing (i) to Burnyeat Cohen has (i.1) in mind, because it does what he says it does and doesn't do what he says it doesn't. It does yield consistency with functionalism and it doesn't entail (ii). However, what it also doesn't do, I'm afraid, is get Burnyeat right. Burnyeat surely doesn't take his rival interpretation of receiving sensible form without matter to be the simple denial of type-type identity, if for no other reason than that it so obviously doesn't entail (ii) let alone the strengthened version of it. And while it may be that he is entitled to no more than (i.1), that would have to be settled independently. My best guess, as I've indicated, is that he holds (i.3) which is inconsistent with functionalism and which along with the two questionable assumptions of basic interaction imply both versions of (ii).
I don't think he's entitled to (i.3), because I don't think he's entitled to assume on Aristotle's behalf that physiological processes are not essentially psychological. I suspect Burnyeat assumes that granting any dependence of sensory acts on associated physiological changes in the sense-organs raises the specter of materialism and gives functionalist interpretations, as he understands them, a toe hold, as if a specification of physiological processes in the eye would provide sufficient conditions for acts of visual awareness. However, that fear is easily allayed if we don't give the term 'physiological' a post-Cartesian physicalistic sense. Instead, we can just extend the peculiar dependence of matter on form in the case of ensouled substances to their psychological states and acts. Thus, as the matter of a living organism is its essentially ensouled body so too the matter of a sensory act is a physiological process in the appropriate sense-organ, ultimately the heart, whose nature is similarly tied to the form of the item of which it is the matter or at least tied to the form/soul of the organism in which it occurs. This effectively blocks any lingering temptation to ascribe to Aristotle reductive sounding accounts of psychological phenomena of the sort, 'Anger, in humans, although it, of course, involves desire for retaliation, is really just blood boiling up around the heart', as though blood boiling up around the heart were merely a Cartesianly physical affair. Thus, I am recommending (i.2) as Aristotle's understanding of taking on sensible form without matter and becoming like the object perceived.
 J. Ackrill, "Aristotle's Definitions of psuche", Proceedings of the Aristotelian Society, 73 (1972-73), pp.119-33.  Bernard Williams, "Hylomorphism" in J. Annas (ed.), Oxford Studies in Ancient Philosophy, 4 (1986).  See G.E.R. Lloyd, "Aspects of the Relationship between Aristotle's Psychology and his Zoology" (chapter 9 of the volume), for a refreshingly non-partisan sifting of the evidence in the biological works. He finds Aristotle drawn in both directions, toward and away from vitalism.  Since her account depends crucially on how form is predicated of matter it is unfortunate that her discussion does not take account of Driscoll, "EIDH in Aristotle's Earlier and Later Theories of Substance" in Studies in Aristotle, ed. O'Meara, Catholic University Press, 1981, pp.129-159, and Loux, "OUSIA: A Prolegomenon to Metaphysics Z and H", History of Philosophy Quarterly 1 (1984), pp.241-266.  I follow Cohen in focusing exclusively on token-identity versions of functionalism. For weaker versions of functionalism, see Richard Boyd, "Materialism Without Reductionism: What Physicalism Does Not Entail" in Readings in the Philosophy of Psychology Vol. I, ed., Ned Block, Harvard 1980, pp.99-103; and John Haugeland, "Weak Supervenience", American Philosophical Quarterly Vol 19, No. 1 (1982), pp.93-102.