One often hears extravagant claims made for the Aristotelian doctrine that "what understands and what is understood are the same" (De anima iii.4; 430a4). This identity between knower and what is known, or between percipient and what is perceived, is often said to offer a way out of the familiar skeptical arguments against the possibility of our having knowledge of the external world. Typically such claims are made by students of Thomas Aquinas, who in this way seek to render Aquinas's theory of knowledge immune from the skeptical and idealist controversies of modern philosophy. In this paper I argue that Aquinas put this doctrine of identity to no such use, that in fact he explicitly rejects any such use of the doctrine, and that furthermore no plausible reading of the doctrine could conceivably produce such dramatic results.
Saint Josephs University
Robert Pasnau: The Identity of Knower and Known
One often hears extravagant claims made for the Aristotelian doctrine that "what understands and what is understood are the same"1 or, elsewhere in the De anima, that "knowledge is in some way the knowable things, while sense is those that are sensible."2 This identity between knower and what is known, or between percipient and what is perceived, is often said to offer a way out of the familiar skeptical arguments against the possibility of our having knowledge of the external world. This doctrine, we are sometimes told, allows us to bridge the problematic gap between our ideas or impressions and the outside world. By providing for the identity of our thoughts and their objects it guarantees immediate access to reality.
Typically such claims are made by students of Thomas Aquinas, who in this way seek to render Aquinas's theory of knowledge immune from the skeptical and idealist controversies of modern philosophy. Etienne Gilson is typical of those who make this sort of move:
Gilson invokes the identity of knower and known as the only
condition under which we can hold that the external object itself is "present in thought." On this identity depends "the whole objectivity of human knowledge." Now precisely what Gilson means by such claims is not evident. But before evaluating the importance of this cognitive identity, as claimed by Gilson and others, we should consider what, for Aquinas, this identity really is.
Bernard Lonergan has noticed that there is a certain tension between the identity doctrine as it appears in Aristotle and in Aquinas. Aristotle, at least sometimes, speaks of an identity between the activity of cognizer and cognized - for instance at De anima iii.2 (425b26-28) where he says that "the activity of the sensible and the sense is one and the same, although their existence is not the same."4 Perhaps he means this to be an instance of a general point he makes elsewhere: that acting and being affected (e.g., moving and being moved) are identical activities or actualizations under different descriptions.5 However Aristotle is to be understood, it is clear that (at least sometimes) he takes the identity of knower and known to be an identity of two activities. In other words, knowing is the same as being known.
Aquinas, interestingly enough, follows this way of speaking without remark when he comments on the De anima text. His gloss on the above passage from De anima iii.2 closely follows the text, holding that "the activity of any sense is one and the same in subject with the activity of the sensible, but they are not one in definition."6 Nowhere in this part of his commentary does he give any hint of being uncomfortable with that reading of the identity doctrine. Nevertheless when not constrained by the letter of Aristotle's text Aquinas consistently reads the identity of knower and known in a quite different manner. Instead of equating the act of cognizing with the act of being cognized, Aquinas speaks of a formal identity between the cognizer and the object cognized. Even in the De anima commentary itself Aquinas prefers to speak of the identity in this way.
The identity at issue, in other words, is taken to be a matter of the identity of species. Here the term 'species' is meant in the cognitive sense: i.e., it refers not to logical or biological kind, but rather to sensible and intelligible species. The species informing intellect is the same as the object's species. As he says elsewhere in the De anima commentary, "the species of the thing actually cognized by intellect is the species of intellect itself."8 Outside of the De anima commentary Aquinas invariably treats the identity in question as formal identity:
the very form or species of the external object is received in sense or intellect. He writes that a mental representation "must be of the same species [as the external object], or rather, must be its species."9 This does not mean that the senses or intellect actually take on the characteristics of the object being
cognized. For the sensible or intelligible species exists in a different way in our cognitive faculties - it exists there intentionally, not naturally. Nevertheless, despite this difference in its manner of existence, the species or form is identical internally and externally.
If the identity of knower and known is to do any epistemological work for us it will be when it is understood as formal identity. This is how Gilson takes the doctrine - as meaning that the species "is the very object" (1). On this identity depends "the whole objectivity of human knowledge." The idea must be something like this: We are able to have knowledge of the external world through our interior impressions and ideas because the two are somehow identical. There is no gap between appearances and reality, and no veil of species or ideas, because the species are - at least in a way - identical with external objects. Now, to be sure, Aquinas never uses the doctrine of formal identity for this end. As is often noted, Aquinas never confronts the sort of skeptical problems that are characteristic of modern philosophy. But, all the same, readers of Aquinas often believe that this doctrine can be used for just that purpose. Indeed it might be suggested that one reason Aquinas was never troubled by the question of our knowledge of the external world is that he had in place this doctrine of formal identity.
It's not hard to see, however, that the notion of formal identity does not give us a general reply to the main forms of skeptical argument. How, for instance, could the identity doctrine answer the dreaming doubt? If we are dreaming, then plainly our ideas are not formally identical with external objects (unless by chance it happens that things are the way we dream them to be). But how do we know, at any given moment, that we're not dreaming? Likewise, if we are being deceived by an evil demon, then our ideas are not formally identical with external objects. But how do we know we are not being so deceived? Of course if one is allowed to presuppose the truth of the identity doctrine then these possibilities vanish. But the skeptic isn't going to tolerate our supposing from the outset that reality corresponds to our ideas.
I don't think that Gilson and others who rely on formal identity intend the doctrine to solve these sorts of skeptical worries. Indeed Gilson devotes an entire book to the Thomistic treatment of skepticism without even discussing the formal identity doctrine.10 What Gilson and others rather seem to have in mind is to use formal identity to rebut a kind of idealism, according to which human knowledge is not of the external world but of our inner ideas or impressions. This fits Gilson's claim that "the whole objectivity of human knowledge" depends on this formal identity. Evidently what he wants to avoid is a kind of subjectivism according to which our knowledge never reaches outside ourselves.
Care has to be taken to distinguish this claim for the identity doctrine from a weaker claim, that formal identity is responsible for giving our mental states their intentional
content. This latter claim seems to be what Peter Geach has in mind when he explains the formal identity doctrine and says that "it is thus that our mind 'reaches right up to the reality.'"11 This is perfectly true in the sense that Aquinas explains mental representation and intentionality in terms of formal identity. My thoughts about x are about x insofar as they are a likeness of X, and Aquinas explains likeness in terms of formal identity. In this rather modest sense the identity doctrine does explain how our thoughts reach right up to reality.
Gilson and others want to make a stronger claim for the identity doctrine. They hold that our thoughts, in virtue of this formal identity, reach right up to reality in a special way. It's not just like the way my words 'the president' reach right up to the president, or even like the way a picture of the president reaches right up to the president. Moreover, our thoughts would not reach reality in this special way if they were mere likenesses. The point Gilson and others want to make is that Aquinas's theory of knowledge is superior to an account on which ideas and impressions are no more than pictures (of some kind) of the external world. Our ideas are, somehow, the objects themselves, and so they are no mere representations of reality.
One way to put this stronger claim made for the identity doctrine is that it gives Aquinas a direct realist account of perception and knowledge in general. Norman Kretzmann takes this position, writing that our access to the external world "is utterly direct, to the point of formal identity between the
extra-mental object and the actually cognizing faculty."12 This matches Gilson, who we earlier saw claim that "it is not the species of the object that is present in thought but the object through its species." What is being rejected by these writers is representationalism, the view that the immediate objects of perception and knowledge are our inner ideas and impressions, and that the external world is known indirectly (if at all).
How does the identity doctrine secure direct realism? The answer isn't entirely clear. The most obvious motive for emphasizing the identity between species and object would seem to be that this allows one to admit that the species is itself apprehended, but nevertheless deny that this entails representationalism, since the species is identical to the object. schematically the reasoning would run as follows: even though X is apprehended in virtue of Y's being apprehended, one can still maintain that X is apprehended immediately because (a) X=Y, and (b) if X=Y then apprehending Y just is apprehending X. This seems the most natural way to understand the move being made by Kretzmann and Gilson above.
This is a surprising conclusion to reach, however, because most commentators would deny that Aquinas thinks species are in any sense the objects of cognition. Kretzmann, for instance, claims on the very next page that the relationship between species and object is causal, not representational - i.e., we shouldn't think of the species as literally representing the external world to some inner eye. But if species are merely
causal intermediaries between our cognitive faculties and external objects then it is hard to see why it should matter that they be identical with those external objects. Surely any modern direct realist theory of perception will allow causal intermediaries between object and percipient: no one would dream of denying the title of 'direct realism' to a theory of perception merely because it tolerates causal intermediaries. And it is not clear what more is to be gained by holding that those causal intermediaries are in some sense identical with the external object. The debate between direct realists and representationalists isn't about the causal connection between percipient and object, and n if species are mere causal intermediaries then it is hard to see how the doctrine of formal identity contributes to the argument for direct realism. To put this same point another way, Kretzmann, by taking species as merely causal intermediaries, is already ascribing direct realism to Aquinas. It's not clear what more he gains by formal identity.
So formal identity seems relevant to direct realism only if species are somehow themselves apprehended. Now I don't think that saying this is enough to end discussion of the identity doctrine, since I argue elsewhere that there is a sense in which Aquinas does treat species as the objects of cognition. But we should notice that even if the identity doctrine gives us a way to reject representationalism this still isn't going to help us refute the associated skeptical difficulties. The skeptic makes the following line of argument against the representationalist:
If all you see directly are your inner ideas and impressions, then how can you have any knowledge of the external world? How can you know that things really are as they seem to be? From the outset the identity doctrine can do no more than beg these questions by presupposing that things are as they seem to be. The skeptic might concede for the sake of argument that if our perceptions are veridical, then our ideas are formally identical with external objects. But the skeptic will want to know how we are entitled to assume that this formal identity holds. The doctrine of formal identity itself offers no answer to this question. The skeptic has to be answered in some other way.
Again we see that formal identity provides no help against the skeptic. But the doctrine may still serve a useful purpose if it can show us how to be direct realists. On my analysis formal identity appears to give us a way of having our cake and eating it too. We can treat species as the internal objects of cognition, and still reject representationalism. Because the species is identical with the object, apprehending the species is apprehending the object. This line of argument, however, rests on an invalid move. The argument assumes that we can substitute identical objects into claims about perceiving and apprehending while preserving the truth of those claims. The schema was that if X=Y then apprehending Y just is apprehending X. It's not clear that this schema holds even when limited to objects that are numerically identical. Phrases such as 'S apprehends Y' or 'S sees Y' might, at least arguably, be instances of an opaque
conclusion. The objection supposes that, however we take the claim that intellect is the thing cognized, that identity will entail that the thing cognized is in intellect. But the only candidate for being in intellect is the intelligible species. Therefore the intelligible species is itself cognized.
One way Aquinas could deal with this objection is to accept the conclusion, but remind the reader of the identity doctrine:
that the species in intellect is the species of the external object (2). Hence cognizing the species is the same as cognizing the external object. (This is what I referred to above as the strategy of having your cake and eating it too.) Aquinas doesn't answer the objection in that way, which isn't surprising since -as we've just seen - the move is invalid when the identity in question is mere formal identity. Rather, Aquinas answers the objection (3) by clarifying his interpretation of the identity doctrine:
Here Aquinas isn't giving up the doctrine of formal identity. On his view one thing can be a likeness of another only if there is a formal identity between them: in his words, "a likeness between two things occurs to the extent that there is an agreement in form."16 The point of his speaking of likenesses here is to emphasize that the identity in question is mere formal identity;
it's not as if the very object being apprehended is inside intellect. Hence the identity doctrine doesn't entail that the species is the object of cognition. Elsewhere he makes this point even more plainly, saying that "the application of what is cognized to the one cognizing... shouldn't be understood as a kind of identity, but as some kind of representation."17 Again the point is not that no kind of identity is involved, but that it is merely formal identity, not actual numerical identity. Hence (contrary to 3) external objects could be cognized directly without species being cognized at all. And, vice versa, species might be cognized directly without external objects being cognized at all. Aquinas of course holds that the former possibility in fact obtains and that the latter does not (except in introspection), but we've seen that the doctrine of formal identity cannot help him reach this conclusion.
1. De anima iii.4 (430a4). All quotations from the De anima are based on the medieval Latin translation by William of Moerbeke.
2. De anima iii.4 (430a4).
3. The Christian Philosophy of St. Thomas L.K. Shook (tr.), New York: Random House, 1956, p.227.
4. "Sensibilis autem actus et sensus idem est, et unus; esse autem ipsorum non idem."
6. "Sed etiam actus cujuslibet sensus est unus et idem subjecto cum actu sensibilis, sed ratione non est unus" (Sentencia libri De anima [=InDA] II.26 [III.2.590]).
7. "Et per hunc modum dicitur intellectus in actu esse ipsum intellectum in actu, inquantum species intellecti est species intellectus in actu" (InDA 111.7 [111.13.789]).
8. "Species igitur rei intellectae in actu est species ipsius intellectus" (InDA 111.3 [III.9.724]).
9. Summa contra qentiles III].49.2266.
10. Thomist Realism and the Critique of Knoweledqe Mark A. Wauck (tr.), San Francisco: Ignatius, 1986.
11. Anscombe and Geach, Three Philosophers, Ithaca, NY: Cornell University Press, 1961, p.95.
12. "Aquinas's Philosophy of Mind" in Kretzmann and Stump, eds., The Cambridge Companion to Aquinas, Cambridge: Cambridge
University Press, 1993, p.138.
13. See, e.g., Quaestiones de veritate [=QDV] 2.5 ad 5.
14. "Intellectum enim in actu est in intelligente: quia intellectum in actu est ipse intellectus in actu. Sed nihil de re intellecta est in intellectu actu intelligente, nisi species intelligibilis abstracta. Ergo huiusmodi species est ipsum intellectum in actu" (Summa theoloaiae [=ST] la 85.2 obj.l).
15. "Intellectum est in intelligente per suam similitudinem. Et per hunc modum dicitur quod intellectum in actu est intellectus in actu, inquantum similitudo rei intellectae est forma intellectus; sicut similitudo rei sensibilis est forma sensus in actu" (ST la 85.2 ad 1). Cf. ST la 87.1 ad 3.
16. QDV 8.8c.
17. "Applicatio cogniti ad cognoscentem, quae cognitionem facit, non est intelligenda per modum identitatis, sed per modum cujusdam repraesentationis..." (QDV 2.5 ad 7).