Aristotle on the Transmission of Information: Receiving Form Without

15© The Editor(s) (if applicable) and The Author(s), under exclusive license to

Springer Nature Switzerland AG 2020

D. Bennett, J. Toivanen (eds.), Philosophical Problems in Sense Perception:

Testing the Limits of Aristotelianism, Studies in the History of Philosophy of

Mind 26, https://doi.org/10.1007/978-3-030-56946-4_2

Chapter 2

Aristotle ontheTransmission

ofInformation: Receiving Form Without

theMatter

VictorCaston

Abstract At the beginning of De anima 2.12, Aristotle compares perception to the

way sealing wax is imprinted by signet rings to explain how the senses receive form

“without the matter.” A close reading shows that the passage concerns the way

information is transmitted through causal interactions, specically through a kind of

transduction. Unlike other causal interactions where the form is embodied in what

is affected, replicating the agent’s active quality, the wax does not become a signet,

nor do the senses become colored, avored, and so on. Rather they receive the form

in question by embodying and replicating certain features that are essential to the

active quality, specically the ratios Aristotle thinks dene perceptible qualities,

and thereby receive information about the perceptible object. This distinctive natural

and material process explains the authority of perception: the senses receive the

identifying marks of objects and thus bear the “seal of reality.”

2.1 Introduction

Much of what has been put forward as evidence of Aristotle’s concept of intention-

ality is nothing of the sort. Brentano and others standardly rely on ve doctrines

from his theory of perception and understanding:

(i) Every psychological activity has an object (de An. 2.4).

(ii) In cognition, we become like the object (de An. 2.5; cf. 3.4).

The following paper was started in 2004 and read widely, but never published, apart from a brief

sketch in §§4.1 and 4.3 of Caston 2005. I have incorporated material from the latter, signicantly

revised and expanded here.

V. Caston (*)

University of Michigan, Ann Arbor, MI, USA

e-mail: [email protected]

(iii) The activity of the object is “one and the same” as the activity of cognition and

present in the subject (de An. 3.2).

(iv) Cognition is not an ordinary alteration (de An. 2.5).

(v) In cognition, the form is received without the matter (de An. 2.12; cf. 3.4).

Coming after Brentano, as we do, it is all too easy to hear familiar theses about

intentionality in these doctrines, especially the rst four. The rst sounds like the

thesis that every mental state is of or about something; the second, that something

in us is somehow similar, or even isomorphic, to the object of our mental state; the

third, that the object of cognition, as an object of cognition, is immanent in the

activity of cognition and is essential to its identity; and the fourth, that intentional

states cannot be accounted for naturalistically—that no material or physical event

exhibits intentionality.

If we view these doctrines in their original context, however, it is clear that this

impression is mistaken. Most of these doctrines do not even come close to having

the right extension for expressing a concept of intentionality.

The rst four doctrines are all instances of broader generalizations that are meant

to hold for nonintentional as well as intentional states. They are causal doctrines that

apply to agent-patient interactions quite widely, including noncognitive and

inanimate ones. In Aristotle’s view, (i) holds not only for cognition, but digestion.

The likeness at issue in (ii) is something he thinks is found in every agent-patient

interaction whatsoever. The patient undergoing change always comes to be like the

agent’s active quality, as result of its action. (iii) likewise applies universally to such

interactions. Aristotle believes the activity of every agent is “one and the same” as

the activity of the patient and present in it. (iv) extends beyond cognition as well. It

turns on the difference between two kinds of causal power, namely, passive powers

that are exhausted or used up when they are exercised—for Aristotle alterations in

an ordinary sense—and those where the power is preserved or even strengthened by

the activity, like building or digestion. These four doctrines cannot, then, serve as a

mark of the intentional. None of them individually constitutes a sufcient condition

for intentionality, or even all of them jointly, since they all apply to digestion.

Still less does any of them offer a necessary condition of intentionality, including

(v). This is due, once again, to their causal nature. Any intentional state to which

these doctrines apply would be about its immediate cause—it will be about what

brings it about. But for many intentional states, content and cause diverge, for the

simple reason that what these states are about has never existed or been the case, and

so cannot be a cause. The content of such states cannot be explained by any simple

causal account, whereas a mental state is about its immediate cause. Aristotle

recognizes this and because of it introduces a distinct new faculty he calls phantasia,

which plays a central role in his account of intentionality.

Perception and understanding, of course, are both intentional states. But they do

not serve as Aristotle’s model for intentional states generally, as Neoscholastics

sometimes appear to think. Perception and understanding are central for him, but

For discussion of these issues, see Caston 1996, 1998.

V. C a sto n

because of the role he takes them to play in our mental economy. Each involves a

key transition or “interface”: one between the world and our experience of it, the

other between that experience and rational understanding. What happens at each of

these junctures is obviously critical. Moreover, it is natural to think, as Aristotle

does, that the content of other intentional states derives in some way from the

content of perception and understanding. So for this reason alone, we need to have

some grasp of their intentionality too.

Neoscholastics like Brentano could reply with a scaled-back proposal. They

might concede that none of these doctrines provides a necessary condition of

intentionality, and also that the rst four do not provide a sufcient condition, even

jointly. But (v), they might insist, does constitute a sufcient condition, at least in

conjunction with the rest. Receiving form without the matter is something that

pertains only to intentional states, they might argue, even if it does not pertain to all

of them. It would thus mark a distinctive type of change—on their view, a “spiritual”

or “intentional” change, to use Aquinas’ terminology, which cannot be reduced to

underlying material changes. On a more recent version of this view, in fact, there is

no underlying material change: receiving form without the matter instead is a basic

and fundamental type of interaction with the world. On both versions, Aristotle still

draws attention to intentionality. But he does not offer an explanation of it. Instead

he takes it for granted: intentionality on this view is something sui generis, which

cannot be further analyzed or explicated. The phrase “receiving form without the

matter” is therefore construed in a purely negative way. In using it, Aristotle would

simply be claiming that in perception we do not receive form in the manner of

material changes; and this, it is assumed, must have something to do with

intentionality. On this view, though, little more can be said.

Part of this reply is no doubt right. But only part of it. Aristotle is trying to dis-

tinguish what occurs in certain intentional states from what happens elsewhere in

nature. The former essentially involves a transmission of information, whereas the

latter generally does not. But this does not commit him to a primitive conception of

intentionality, which is due to a fundamenal type of change, much less an immaterial

one. As Aristotle develops the notion, receiving form without the matter is not

purely negative. What he has in mind is a kind of transduction, where certain aspects

of an object’s form are transmitted, though not in exactly the same way. They are not

literally replicated, but transposed (as it were) to a different key: certain abstract but

essential features of the original form are literally preserved, but now realized in a

transduced form. This takes place, moreover, by means of ordinary material changes,

even though not reducible to them. It is implemented by such changes, just as the

activity of building is implemented by sawing and hammering, and by the material

changes underlying these (cf. de An. 2.5, 417b8–9). Thus, on the view I shall argue

for, Aristotle holds that intentionality is a natural phenomenon, which is fully

realized in material changes.

2 Aristotle ontheTransmission ofInformation: Receiving Form Without theMatter

2.2 Two Ways ofReceiving Forms

In De anima 2.12, Aristotle turns from his survey of the individual senses in the

previous ve chapters (2.7–11) to the power to perceive in general.

One of his

central aims in this chapter is to distinguish perception from other natural changes.

Plants can be warmed or cooled, but unlike us, they can’t feel or perceive it (424a32–

b1). What accounts for this difference? It is not the fact that we are acted on by a

perceptible quality, for plants are too: they are affected, he notes, “by the tangible

qualities themselves” (ὑπὸ τῶν ἁπτῶν αὐτῶν, a34), for example by heat or cold.

It’s also not the fact that we have a soul, since plants also have “an animate part”

(ἔχοντά τι μόριον ψυχικόν, a33), indeed a soul.

The example of plants being

warmed is thus very pertinent. It shows that for perception to occur, it is not suf-

cient that a perceptible quality affects something animate. There must be some fur-

ther difference that accounts for the fact that we perceive and plants do not.

The obvious, but unilluminating, answer is that animals are capable of percep-

tion whereas plants are not. But that only pushes the question back a step. For we

can still ask just what it is that enables animals to perceive, and how does it bring

this about? We feel that there ought to be something in an animal’s make-up, which

plants lack, that gives it this capacity; and that it makes a difference to the kind of

change that occurs, to the way in which they are affected by perceptibles, so that in

one case there is perception and in the other case there is not. Clearly animals have

the ability and plants do not. But the question is, what difference accounts for that

difference?

Of course, the fact that we can raise a question doesn’t guarantee that an answer

will be forthcoming. In the abstract, Aristotle could resist this demand for

explanation. He could stand his ground and repeat his earlier answer: animals have

the capacity to perceive and plants do not, end of story. It simply comes down to

The traditional chapter division between 2.11 and 2.12 slightly obscures this by splitting what was

originally a continuous sentence: “Now that we have spoken in outline regarding each sense indi-

vidually, | we need to understand about every sense in general that […]” (καθ’ ἑκάστην ὲν οὖν

τῶν αἰσθήσεων εἴρηται τύπῳ, | καθόλου δὲ περὶ πάσης αἰσθήσεως δεῖ λαβεῖν ὅτι […],

424a15–17). All translations are my own. Unless otherwise noted, Förster’s edition (1912) has

been used for the De anima.

De An. 1.5, 411b27–30; 2.2, 413b7–8; 2.3, 414a33–b1; cf. 1.5, 410a21–24; 2.3, 415a1–3.

It is not due to a difference in heat. In PA 2.2 (648b11–649b8), Aristotle discusses at some length

the different and sometimes conicting criteria we use when characterizing one substance as

warmer than another, and he expressly distinguishes feeling warmer—that is, being warmer to

touch—from heating more quickly or cooling more slowly. Boiling water, for example, is hot to

the touch, but cools more quickly than oil and would in fact be said to be cooler by nature. Other

criteria involve relative size, melting points, and combustibility. But Aristotle nowhere suggests

that these different effects are due to fundamentally different causal powers. It is rather a question

of how a single causal power, heat, affects the passive properties in different types of material. This

applies equally to de An. 2.12: he is contrasting the effect of a single quality like heat on the per-

ceptual organs of an animal with the effect it has on other bodies, including the body of a living

thing that lacks perception. For an excellent detailed discussion of the PA 2.2 passage itself, see

Lennox 2001: 191–95; for an attempt to enlist this passage in favor of a spiritualist reading, see

Johansen 1998: 277–80. On spiritualist interpretations generally, see reference in next note.

V. C a sto n

having the right sort of soul, one capable of perception, and there is no further

explanation that can be offered.

But this is not the end of the story for Aristotle. He is sensitive to the demand for

explanation and immediately moves to address it at 424b1–3. The reason (αἴτιον

γάρ) that plants do not perceive, he explains, is that they lack a balance or “mean”

(τὸ μὴ ἔχειν μεσότητα) and so do not possess the right basis (τοιαύτην ἀρχήν) for

receiving or being affected by perceptibles in the relevant way. Instead they are

“affected along with the matter” (πάσχειν μετὰ τῆς ὕλης). This explanation has two

parts. The rst part, about possessing a balance or mean, concerns what enables us

to perceive. The second, about how we are affected when we perceive, concerns the

activity of perception itself. In line with Aristotle’s general methodology (de An.

2.4, 415a16–22), we should look rst to the activity. Only afterwards can we return

to the capacity and what grounds it: the answer to the second question constrains the

answer to the rst.

What does it mean to say that plants are affected along with the matter? Aristotle

does not elaborate. But it is plainly meant to contrast with the way the senses are

affected. For he says that the senses receive the forms of perceptibles “without the

matter” (ἄνευ τῆς ὕλης, 424a18) just a little earlier. Here is how the chapter begins:

We have spoken in outline about the senses individually. But with regard to every sense in

general, we have to recognize that a sense is something that can receive perceptible forms

without the matter, just as wax receives the ring’s signet without the iron or gold: it takes on

the golden or brazen signet, but not in so far as it is a gold or bronze ring. The sense for each

[type of perceptible] is likewise affected in a similar way by what has color or avor or

sound, not in so far as it is said to be each of them, but in so far as it is of this sort, in virtue

of its logos.

Aristotle applies the notion of receiving form without the matter twice: rst to

the sealing wax, which is supposed to offer a clearer illustration of his meaning, and

then to perception itself, where he offers his most explicit formulation of the distinc-

tion. To receive a form without the matter, he claims, is not to be affected by an agent

(a) in so far as it is “said to be each of these things” (ᾗ ἕκαστον ἐκείνων

λέγεται, 424a23)

but rather

According to “spiritualist” interpretations, championed above all by M.F. Burnyeat, this in fact is

Aristotle’s response. For a detailed examination of spiritualist interpretations and the debate sur-

rounding them with full references, see Caston 2005, esp. §1.3 on the denition of spiritualism and

§2 for criticism.

De An. 2.11, 424a15–2.12, 424a24 (cf. 3.2, 425b23–4): καθ’ ἑκάστην μὲν οὖν τῶν αἰσθήσεων

εἴρηται τύπῳ. καθόλου δὲ περὶ πάσης αἰσθήσεως δεῖ λαβεῖν ὅτι ἡ μὲν αἴσθησίς ἐστι τὸ δεκτικὸν

τῶν αἰσθητῶν εἰδῶν ἄνευ τῆς ὕλης, οἷον ὁ κηρὸς τοῦ δακτυλίου ἄνευ τοῦ σιδήρου καὶ τοῦ

χρυσοῦ δέχεται τὸ σημεῖον, λαμβάνει δὲ τὸ χρυσοῦν ἢ τὸ χαλκοῦν σημεῖον, ἀλλ’ οὐχ ᾗ χρυσὸς

ἢ χαλκός· ὁμοίως δὲ καὶ ἡ αἴσθησις ἑκάστου ὑπὸ τοῦ ἔχοντος χρῶμα ἢ χυμὸν ἢ ψόφον πάσχει,

ἀλλ’ οὐχ ᾗ ἕκαστον ἐκείνων λέγεται, ἀλλ’ ᾗ τοιονδί καὶ κατὰ τὸν λόγον. Because the Greek

λόγος is a technical term for Aristotle and difcult to translate neutrally, I have left it transliterated

to avoid prejudicing the interpretation. I discuss its meaning below in §2.4.2.

2 Aristotle ontheTransmission ofInformation: Receiving Form Without theMatter

(b) in so far as the agent is “of this sort, in virtue of its logos” (ᾗ τοιονδί καὶ κατὰ

τὸν λόγον, a24).

Plants, presumably, are affected in the rst way (a) and possibly the second (b). The

senses, in contrast, are affected only in the second way (b). The difculty lies in

spelling out what each of these alternatives amounts to, and Aristotle’s use of pro-

nouns here is not very helpful.

On a fairly commonplace reading, Aristotle is drawing a contrast between the

underlying substance and its qualities.

Both, one might argue, are implicit in the

clause before, when we are told that the sense is affected “by what has color or a-

vor or sound” (ὑπὸ τοῦ ἔχοντος χρῶμα ἢ χυμὸν ἢ ψόφον, 424a22). The perceptible

object does not affect us in so far as it is said to be “each of these,” that is, in so far

as it is said to be some type of substance, but only in so far as it is “of this sort,”

namely, colored, avored and so on. To see or smell a gardenia is to be affected by

it, but not as a gardenia. It is to be affected by it only in so far as it is colored or

fragrant. The point of the contrast, on this reading, is that in perception we are

affected by objects only in a certain respect, namely, only in so far as they have

perceptible qualities. The assimilation that we undergo in perception is likewise

restricted. We become like the perceptible object only in so far as it has the

perceptible quality in question. To receive form without the matter, on this reading,

is just to be affected by objects in a more limited way.

This reading is especially natural for so-called literalist interpretations of

Aristotle.

According to such interpretations, our sense organs become “likened” to

their objects

in a very strong sense, by literally taking on the perceptible qualities

of their object. That is, every instance of the following schema will be true:

 Whenever I perceive a perceptible quality, F, my sense organ literally comes to be F—

the predicate ⌜F ⌝ applies to my sense organ in just the same sense that it applies to

the object.

Aristotle uses similar language when it comes to the understanding, drawing an explicit parallel

to perception (de An. 3.4, 429a15–18); see below, p.33.We will return to the interpretation of the

pronouns in these phrases (424a23–24) below, pp. 34–35. Lorenz (2007: 193 n. 29) reads ἢ at

424a24 with the later Ambros. H50 supp., rather than the ᾗ in the other MSS and accepted by all

the other editors (including Förster and Jannone, who often favors H50). But this reading does not

affect the difculty over how to construe the contrast, and indeed Lorenz in his translation still

understands this clause as governed implicitly by the preceding ᾗ in a23 (“except as being of this

or that quality,” emphasis mine).

W.D. Ross states this reading succinctly: “[…] so the sense for each kind of object (whether this

is coloured, avoured, or resonant) is affected by the object not in respect of the object’s being, for

instance, a chair or a table, but in respect of its being of a certain kind (i.e. coloured, avoured, or

resonant) […]” (Ross 1961: 265, emphasis mine). The reading is extremely widespread, but for

other clear statements see e.g. Hicks 1907: 416 ad loc.; Block 1960: 94; and possibly Shields

2016: 248.

Richard Sorabji has been a leading proponent of this view, though in fact he holds an idiosyncratic

version of it. The view more commonly associated with the label is in fact the one defended fully

and explicitly by Stephen Everson (1997). For discussion of how literalism is to be dened, see

Caston 2005, esp. §§1.1–2, and for criticism of literalist interpretations, §3.

De An. 2.5, 418a5–6 (cf. 417a20): πεπονθὸς δ’ ὡμοίωται καὶ ἔστιν οἷον ἐκεῖνο.

V. C a sto n

On this reading, when I look at a gardenia, part of my eye actually goes white; when

I smell it, part of my nose becomes scented with the same heavy fragrance. But no

part of me literally becomes a gardenia, in esh and blood. I only take on its

perceptible qualities or forms, without acquiring the gardenia’s underlying matter,

much less the substance. According to this interpretation, that is precisely the point

of the comparison with the signet ring. The wax doesn’t become metal, much less

another ring. It only takes on the ring’s surface contours—it comes to be shaped in

the same way.

Aristotle’s general principle that “man begets man” is thus upheld,

though only in a restricted form. In perception, we replicate the qualities of an

object, but not the object itself.

Part of this is undoubtedly correct. On Aristotle’s view, we are not affected by

perceptible objects in so far they are a certain kind of substance, but only in so far

as they have certain perceptible qualities—only in so far as they are white, or sweet,

or fragrant, and so on. This is not because we can’t perceive objects like tables,

chairs, or humans. We can, on Aristotle’s view. But they are not perceived intrinsi-

cally (καθ᾿ αὑτά). Rather, they are perceived extrinsically (κατὰ συμβεβηκός),

since being a substance of a certain sort is extrinsic to perceptibles like white.

fact, Aristotle says that the sense “is not affected at all by a perceptible in so far as

it is that sort of thing.”

We perceive the son of Cleon, he claims, “not because he is

the son of Cleon, but because he is light.”

What is principally perceptible (κυρίως

αἰσθητά) are qualities that are intrinsically perceptible (καθ᾿ αὑτά) to a single sense

exclusively (ἴδια): colors, avors, odors, tones, and tangible qualities like moisture

Here we have to ignore the fact that the depressions and projections will be reversed in the case

of the wax, as this is clearly not meant to be part of the tertium comparationis. All that is essential

on this reading is the absolute displacement from the original surface. Theophrastus, in contrast,

does press this point in criticizing Democritus, who also appeals to signet rings in his account of

vision (De sensibus 52, = Doxogr. Gr. 514.5–10). On Democritus’ theory, see below n. 69.

De An. 2.6, 418a20–23; 3.1, 425a26–27. This has been denied by Irving Block (1960: 95–97),

who argues that on the De anima account we cannot perceive physical objects either “directly” or

“indirectly.” But this ies in the face of Aristotle’s own words: he says that we perceive both

Diares’ son (κατὰ συμβεβηκὸς γὰρ τούτου αἰσθάνεται, 2.6, 418a21–22) and Cleon’s son (κατὰ

συμβεβηκὸς ᾐσθανόμεθα οἷον τὸν Κλέωνος υἱὸν, 3.1, 425a25–26), even if only extrinsically.

The English terms “intrinsic” and “extrinsic” capture the contrast between Aristotle’s technical

terminology here better than any of the traditional translations: “intrinsic” is a very close rendering

of καθ᾿ αὑτό (literally, “in virtue of itself”), while “extrinsic” permits contingency or arbitrariness,

without strictly implying it, just like Aristotle’s use of κατὰ συμβεβηκός, which notoriously

includes “necessary accidents.” For Aristotle, what is at issue is whether something holds in virtue

of a thing’s own nature as such, of some F qua F, or not.

De An. 2.6, 418a23–24: διὸ καὶ οὐδὲν πάσχει ᾗ τοιοῦτον ὑπὸ τοῦ αἰσθητοῦ.

De An. 3.1, 425a25–26: οὐχ ὅτι Κλέωνος υἱός, ἀλλ’ ὅτι λευκός. If we translate both occurrences

of ὅτι with “that,” instead of “because,” the sentence would read: “with regard to Cleon’s son, for

example, we do not perceive that he is Cleon’s son, but only that he is pale; and this happens to

belong to Cleon’s son.” But then Aristotle would be denying that we perceive extrinsic character-

istics, instead of offering an example of a characteristic that is so perceived, as he clearly is in

context (εἰ δὲ μή, οὐδαμῶς ἂν ἀλλ’ ἢ κατὰ συμβεβηκὸς ᾐσθανόμεθα οἷον […], 3.1, 425a24–25),

something he also afrms earlier (2.6, 418a21–22).

2 Aristotle ontheTransmission ofInformation: Receiving Form Without theMatter

and heat.

These qualities themselves act on the sense organs to produce perception,

and it is with respect to them that the essence of each sense is dened.

The percep-

tion of substances as such depends on this more basic form of perception.

But this cannot be Aristotle’s point in the passage above. To become like an object

only in a certain respect is not peculiar to perception. All nonsubstantial change is

like that. If I heat water on a stove, the water becomes like the stove only in so far as

both are hot. It does not become red, or hard, or metallic, much less a stove. Plants,

similarly, become like the sun only in so far as they become warm; they do not

become bright, ery globes. In both of these cases, the patient becomes a replica of

the agent, not in so far as the agent is a certain sort of substance, but only in so far as

it has a certain quality, namely, the active causal power that is presently affecting the

patient. Restricted likeness, then, is not something that distinguishes perception from

ordinary changes that plants can undergo. It is something they share in common. It

would dissolve the very contrast Aristotle is trying to draw in this chapter.

This becomes even clearer if we consider the comparison with signet rings.

Taking on an impression is a perfectly ordinary alteration for Aristotle. It is the kind

of change he can explain without difculty using his standard account of alteration.

If this was all he was after, he could have made his point as effectively with a stick

and clay. When the round end of a stick is pressed into wet clay, the clay takes on a

cylindrical shape—it is likened to the stick as result of the stick’s acting on it. But

the clay does not become wooden, much less a stick. Its likeness is thus restricted in

just the way required by the commonplace reading.

A moment’s reection should sufce to see that this comparison would not have

worked. Had De anima 2.12 opened with the example of a stick being pressed into

clay, it would have fallen completely at. It is not just that it would have been more

prosaic or less colorful. It is that something essential would have been lost.

2.3 The Wax andtheSeal

Sealing wax does not merely take on an impression of the ring’s surface. What it

receives or takes on, Aristotle says twice, is the seal or signet (τὸ σημεῖον, 424a20,

a21).

This goes beyond the sort of likening that occurs when a stick is pressed in

De An. 2.6, 418a24–25; 3.2, 426b8–12.

Sens. 6, 445b7–8. Cf. Sens. 1, 438b22–23; de An. 2.5, 417b19–21.

De An. 2.6, 418a25.

In the Categories, Aristotle characterizes alteration as a change in quality (14, 15b12), and

shapes as a kind of quality (8, 10a11–26). Sometimes, when something is completely transformed

in shape, it results in a new substance and so could not be considered an alteration in a substance

(Ph. 7.3, 245b9–246a9; cf. 1.7, 190b5–6). But this is a scruple about substantial changes and so

does not affect the issue here, which concerns nonsubstantial changes.

This applies to the notion of “liken-ness”—a likeness produced as a result of likening—found in

Charles’ account as well (2000: 114–16).

On this meaning of σημεῖον, see Spier 1990: 107; cf. Lacroix 1955: 92–93; Boardman 1970:

428–29; and on classical nger rings in particular, see Boardman 1970: 212–34. Plato also speaks

V. C a sto n

clay, and not just because signets are pictorial, as ancient Greek signets standardly

are. That sort of likeness is not central here.

A signet is a sign, as the Greek σημεῖον

makes clear. And a sign of a special sort. A signet produces a sealing, an impression

used to establish the identity of its owner. When placed on a document, especially

for legal or ofcial use, a sealing authorizes the claims, obligations, promises, or

orders made therein.

A sealing thus differs from most other impressions in that it

purports to originate from a particular signet and hence from a particular owner,

typically the author or source of the document. This is essential to the nature of seal-

ings. If sealings did not in general serve as tokens of the signet’s owner, they could

not vouch for the authenticity of a document in his absence.

The main point of comparison, then, is not the fact that the wax receives a set of

contours, but rather that it “takes on the golden or brazen signet” (λαμβάνει δὲ τὸ

χρυσοῦν ἢ τὸ χαλκοῦν σημεῖον, a20–21). It is true, of course, that the wax also

receives the surface contours of the signet. But the two changes differ in a crucial

respect. The contours of the ring are literally replicated in the wax. The signet is not.

The wax takes on the signet, but “not in so far as it is gold or bronze” (ἀλλ’ οὐχ ᾗ

χρυσὸς ἢ χαλκός, a21). The wax accordingly does not become another golden or

brazen signet, but a sealing. The signet and the sealing differ, most obviously, in

their material: one is metal, while the other is wax. But they also differ as signs. A

sealing must be present along with a document in order to authorize it. A signet, like

its owner, need not be. A sealing is a one-off effect. A signet, in contrast, can be used

repeatedly, to produce many sealings.

What these two signs, as signs, share in

common is their content. Both indicate the same thing, the owner of the signet, and

of signets being impressed at Theaetetus 191 (ὥσπερ δακτυλίων σημεῖα ἐνσημαινομένους, 8),

a passage frequently cited as a parallel for de An. 2.12. But Plato is describing memory there and

the impressions are taken from perceptions and thoughts (ταῖς αἰσθήσεσι καὶ ἐννοίαις, Tht.

1917), not perceptible objects, as in de An. 2.12. It thus has more in common with Aristotle’s use

of the signet ring comparison for memory at Mem. 1, 450a27–32. See the discussion below, §2.5.1.

The pictorial element does appear to be involved, on the other hand, when Aristotle uses the

comparison to signet rings to explain memory’s relation to perception in De memoria 1, 450a25–32

(for the text of the Parva naturalia, I use Siwek’s edition (1963)). He describes the resulting modi-

cation as “like a kind of picture” or representation (οἷον ζωγράφημά τι τὸ πάθος, a29). For dis-

cussion, see below, §2.5.1.

In addition to sealing private correspondence, legal documents, and ofcial public documents,

sealings were also used to indicate possession, a person’s identity in voting, and in religious sacri-

ce. For a good, brief summary of the uses of signets, with literary references, see Plantzos 1999:

18–22; also Richter 1968: 1–4 and esp. Boardman 1970: 13–14, 235–38, 428–30. Possession need

not be limited to letters or objects either. A fourth century BCE Athenian clay impression of a

signet, depicting a man and a woman embracing, has the legend: ΕΧΩ ΤΕ ΚΑΙ ΦΙΛΩ

ΑΡΙΣΤΟΤΕΛΗ[Ν] (“I possess and love Aristotle”). Christodoulopoulou-Proukake 1977 examines

the possibility that this love charm might have belonged to Aristotle’s mistress, Herpyllis, and

concludes that although there is no conclusive evidence in favor of such an identication, it cannot

be ruled out either. (I would like to thank Seth Schein for translating this article for me.) On the use

of magical love charms to maintain a lover’s affection (philia), as here, as opposed to acquiring a

new one by inciting sexual desire (erōs), see Faraone 1999: 27–30.

A point exploited later by Academics: see Cicero, Academica 2.86.

2 Aristotle ontheTransmission ofInformation: Receiving Form Without theMatter

thereby convey his authority. That is what the process of imprinting and the result-

ing contours are meant to secure: the wax takes on the owner’s insignia from the

signet and thereby signies him. The wax receives the signet “without the iron or

gold” (ἄνευ τοῦ σιδήρου καὶ τοῦ χρυσοῦ, 424a19–20), because it takes on the

owner’s insignia, and so comes to signify him, without becoming another signet.

If receiving form without the matter in general is to be understood on this model,

then it is different from other natural changes. For it essentially involves the

transmission of information. Ordinary natural changes do not. In ordinary natural

changes, an object is literally transformed, by taking on the agent’s form “with the

matter”: the patient embodies the form and so comes to be a replica of the agent, at

least with respect to the agent’s active quality. This is what happens when the stick

produces an impression in clay. The clay embodies the shape of the stick by

replicating its contours. In contrast, when something receives a form without the

matter, the object does not embody that form—whatever changes do take place, it

does not result in a esh and blood replica. The object is still informed by the agent,

if you like, as Augustine says in his De trinitate.

But what this amounts to in

Aristotle’s case is the object’s carrying information about the agent. The word

“information” is intended here in a very weak sense. Something can receive infor-

mation without having any cognition or indeed awareness of it at all. Aristotle’s

example makes this clear. Sealing wax is completely without cognition or

consciousness, much less intelligence.

What is essential is that the resulting state

is about the cause from which it originates. Intentionality in this minimal sense, of

carrying information, will extend to some inanimate things. But it does not occur in

the vast majority of natural changes.

Against Everson, who is contemptuous of representationalist interpretations of Aristotle’s

account of perception (1997, 98). In charging that such interpretations are anachronistic, he evi-

dently overlooks the pertinence of signets in the opening of de An. 2.12.

E.g. De trin. 11.2.3, 336.60–62 (ed.Mountain): “but that informing of the sense, which is called

sight, is imprinted solely by the body that is seen” (illa tamen informatio sensus quae uisio dicitur

a solo imprimatur corpore quod uidetur). For ‘sensus informatus’, see also 336.56–57 and earlier

11.2.2, 334.12–13.

I am thus taking the wax and signet ring to be a genuine example (οἷον, 424a19) of receiving

form without the matter, as it was for scholastic commentators such as Philoponus (In de An.

444.17–26, cf. 437.19–25) and even Thomas Aquinas, who describes it as a “tting example”

(conveniens exemplum, Sent. de An. 2.24, §554); more recently, see also Denyer 1991: 194. If it

were merely an analogy and not a genuine case, as some Neoscholastics claim (see n. 30 below),

then the parallelism Aristotle uses to develop his point here (οἷον, a19 […] ὁμοίως δὲ καὶ,

424a20–21) would not explain the nature of this sort of reception, but only gesture at it. Against

this reading, see n. 30 below.

Nothing could count as a sealing without social conventions, of course, but conventions are not

essential to the comparison, as I shall argue below (see §2.5.1). Greek and medieval commentators

commonly held that this kind of change occurred naturally on its own, apart from cognition or

awareness, in a medium like air or water, as well as in mirrors (Sorabji 1991). Aristotle himself

never expressly states that the medium receives the form without the matter. But he does very sug-

gestively compare the effect of a perceptible object on the medium to a signet ring on wax at de An.

V. C a sto n

The notion of information I am attributing to Aristotle is not, therefore, like mod-

ern information theories of content such as Fred Dretske’s, which hold that informa-

tion is transmitted in every nomic, or even every reliable, correlation; nor does

Aristotle regard information quantitatively, as the amount by which uncertainty is

reduced.

The notion of information in play here is a modest, pre-theoretical one,

of obtaining evidence about the features of things, one that will be claried through

further analysis as we go along. It is not meant to introduce independent leverage

over the texts, but arises instead directly from the comparison with sealing wax. A

sealing is a token, a sign that gives evidence that identies its owner and thus proof

of ownership, because of the way it is produced from the signet and thereby trans-

mits information about the owner’s relation to the object.

To claim, then, that in perception the forms of perceptible qualities are received

without the matter is to claim that all perception essentially involves the transmission

of information. This is not all there is to perception, obviously: in particular, it does

not say anything about consciousness or phenomenal awareness (cf. Ph. 7.2,

244b12–245a2).

But it makes intentionality, or at any rate the reception of

information, a necessary condition of perception. The opening of De anima 2.12

doesn’t make any stronger assertion than that. Above all, it does not claim that

receiving form without the matter is sufcient for perception or cognition, much less

a denition, as is sometimes claimed. For wax also receives the form without the

matter, but it does not perceive the signet ring.

3.12, 434b27–435a10, esp. 435a9–10, something Democritus had also done before him (ap.

Theophr. De sens. 51, 52=Doxogr. Gr. 513.28–514.1, 514.5–6=DK 68 A135).

I have in mind the classic account in Dretske 1981: see chs. 1 and 2in particular for the quantita-

tive notion of information. This is not to rule out that there might be other similarities between

Aristotle’s theory and Dretske’s (as Matt Evans has pointed out to me in conversation). On my

view, Aristotle does take information in perception to be transmitted through the ratios of certain

kinds (see below, §2.4); and both philosophers importantly also agree that more is required for

perception and intentionality in general than just the transmission of information (as each under-

stands it); see e.g. Dretske 1988: ch. 3.

For more on the question of consciousness in Aristotle, see Caston 2002.

Cf. Philoponus In de An. 444.17–20; Sorabji 1995: 218–19. Those who take it as a sufcient

condition or even a denition are forced to claim that the wax sealing comparison is nothing more

than an analogy, and a bad one at that, which “limps”: Owens 1981b: 77–78; Owens 1981a: 91; cf.

Brentano 1867: 81. But this does not follow, if we take the opening claim merely to express a

necessary condition of perception. This is, in fact, all the Greek states: it says that every sense is

receptive of the forms of perceptibles without the matter. It does not state or imply the converse,

that everything receptive of perceptibles’ forms without the matter can perceive.

Alan Code has suggested to me (in conversation) that Aristotle’s position might be even weaker

still. The rst sentence of De anima 2.12 speaks explicitly only about the senses and what they are

capable of, and not their activities; and from the fact that every sense is capable of receiving form

without the matter it does not follow that every act of perception involves the reception of form

with the matter. Second, the words “concerning every sense generally” (καθόλου δὲ περὶ πάσης

αἰσθήσεως) need not indicate a strictly universal generalization; the claim that follows need hold

true only “for the most part,” with the result that some senses might not be capable of receiving

form without the matter at all. Such a reading, though, while compatible with the letter of the text,

seems implausible in context. Aristotle does not at any point qualify the general character of his

2 Aristotle ontheTransmission ofInformation: Receiving Form Without theMatter

The reception of information is critical to the function of perception on Aristotle’s

view. An animal must have the two contact senses, touch and taste, he thinks, or it

could not survive: without them, it could not detect what should be eaten and what

should be avoided (de An. 3.12, 434b10–24; cf. 3.13, 435b16–25). It will be even

better off if it has one of the distance senses—sight, hearing, or smell—since then it

will also be able to detect such things from afar (ἀλλὰ καὶ ἄποθεν, b24–27). In

addition to survival value, perception also contributes to well-being, by laying the

foundation for practical and theoretical understanding. What makes perception

suitable for both of these ends is the reception of information, the fact that the

senses, to use Aristotle’s words, “report” (εἰσαγγέλουσι) the qualitative differences

(διαφοράς) between objects:

The senses [that perceive] through external [mediums]—smell, hearing, and vision—

belong to [animals] capable of locomotion. They are present (i) for the sake of survival in

all [animals] that possess them, so that by perceiving ahead they can pursue their food and

avoid what is bad and harmful; but in those [animals] that happen to possess comprehension

as well, they are (ii) for the sake of well-being. For they report many differences, from

which there arises comprehension of what can be understood and what can be done. Of

these, sight is intrinsically superior with regard to necessities, while hearing is extrinsically

superior with regard to understanding. For the power of sight reports many diverse

[differences], due to the fact that all bodies possess color, so that we also perceive common

[perceptibles] through this sense most of all—by ‘common’, I mean magnitude, shape,

change, rest, and number—whereas hearing only [reports] differences in sound and, in

certain [animals], voice as well. But extrinscially, hearing makes the greatest contribution

to intelligence, since language is responsible for learning due to its being audible, not

intrinsically, but extrinsically. For it is composed out of words and each word is a symbol.

comments or signal that he has reservations, as he often does (such as with the phrase “all or for

the most part”). On the contrary, the generalization which opens de An. 2.12 is very strong. It is

part of a μέν […] δέ construction linking it with the last sentence of the previous chapter (2.11,

424a15), in which his exhaustive survey of each sense individually (καθ᾿ ἑκάστην) in the preced-

ing chapters is contrasted with what “all” (πάσης) of them have in common “generally” or even

“universally” (καθόλου). His intended focus in the chapter, moreover, is not just the capacity to

perceive, but the type of activity that constitutes perception. This is clear when he develops the

comparison with the wax and the signet ring in the subsequent lines. He speaks generally, without

any qualication, of how “the sense for each [kind of perceptible] is also affected in a similar way”

(ὁμοίως δὲ καὶ ἡ αἴσθησις ἑκάστου […] πάσχει, 424a21–23).

Sens. 1, 436b18–437a15: αἱ δὲ διὰ τῶν ἔξωθεν αἰσθήσεις τοῖς πορευτικοῖς αὐτῶν, οἷον

ὄσφρησις καὶ ἀκοὴ καὶ ὄψις, πᾶσι μὲν τοῖς ἔχουσι σωτηρίας ἕνεκεν ὑπάρχουσιν, ὅπως διώκωσί

τε προαισθανόμενα τὴν τροφὴν καὶ τὰ φαῦλα καὶ τὰ φθαρτικὰ φεύγωσι, τοῖς δὲ καὶ φρονήσεως

τυγχάνουσι τοῦ εὖ ἕνεκα· πολλὰς γὰρ εἰσαγγέλλουσι διαφοράς, ἐξ ὧν ἥ τε τῶν νοητῶν

ἐγγίνεται φρόνησις καὶ ἡ τῶν πρακτῶν. αὐτῶν δὲ τούτων πρὸς μὲν τὰ ἀναγκαῖα κρείττων ἡ

ὄψις καθ’ αὑτήν, πρὸς δὲ νοῦν κατὰ συμβεβηκὸς ἡ ἀκοή. διαφορὰς μὲν γὰρ πολλὰς καὶ

παντοδαπὰς ἡ τῆς ὄψεως εἰσαγγέλλει δύναμις διὰ τὸ πάντα τὰ σώματα μετέχειν χρώματος,

ὥστε καὶ τὰ κοινὰ διὰ ταύτης αἰσθάνεσθαι μάλιστα (λέγω δὲ κοινὰ μέγεθος, σχῆμα, κίνησιν,

ἀριθμόν

), ἡ δ’ ἀκοὴ τὰς τοῦ ψόφου διαφορὰς μόνον, ὀλίγοις δὲ καὶ τὰς τῆς φωνῆς· κατὰ

συμβεβηκὸς δὲ πρὸς φρόνησιν ἡ ἀκοὴ πλεῖστον συμβάλλεται μέρος. ὁ γὰρ λόγος αἴτιός ἐστι τῆς

μαθήσεως ἀκουστὸς ὤν, οὐ καθ’ αὑτὸν ἀλλὰ κατὰ συμβεβηκός· ἐξ ὀνομάτων γὰρ σύγκειται,

τῶν δ’ ὀνομάτων ἕκαστον σύμβολόν ἐστιν.

V. C a sto n

At the beginning of the Metaphysics, Aristotle tells us that sight gives us knowledge

(ποιεῖ γνωρίζειν ἡμᾶς) more than any other sense because it “reveals many

differences” between objects (πολλὰς δηλοῖ διαφοράς, 1.1, 980a26–27).

But even

the sense of touch is said to “report” the stimuli it encounters (εἰσαγγέλλειν, Insomn.

3, 461b3), as its organ is capable of receiving all the tangible differences between

bodies.

In perception, then, we are not simply affected by qualitative differences.

The senses also tell us about them, because it is their nature to detect and register

such differences.

It is precisely this ability, which “discriminates the differences in the correspond-

ing perceptible” (κρίνει τὰς τοῦ ὑποκειμένου αἰσθητοῦ),

that makes perception

an essentially discriminative power (δύναμιν κριτικήν).

According to Aristotle, a

sense perceives “accurately” (ἀκριβῶς) if it can discriminate all the qualitative dif-

ferences along its range, to the greatest extent possible.

But it is also a question of

how nely it can discriminate within this range: the most accurate senses must be

sensitive, quite literally, to the most ne-grained differences—“the most minimal”

(τὰ ἐλάχιστα)—between them.

Such powers of discrimination have considerable

survival value, and animals with more rudimentary capacities will be at a disadvan-

tage in pursuing and avoiding the things they ought.

According to Aristotle, nature

For the Greek text of the Metaphysics, I use Ross 1924.

De An. 3.13, 435a21–24; cf. 2.11, 423b27–30.

The difference between (i) the affection produced in us (πάθος καὶ κίνησίς τις), which is differ-

ent and peculiar to each perceiver (τοῦ ἰδίου ἑτέρου ἀριθμῷ, εἴδει δὲ τοῦ αὐτοῦ), and (ii) the

single public object everyone perceives that triggers perception (τοῦ μὲν κινήσαντος πρώτου […]

τοῦ αὐτοῦ καὶ ἑνὸς ἀριθμῷ αἰσθάνονται πάντες), like a particular bell or incense or a re, is very

clearly distinguished at Sens. 6, 446b17–21, in response to a sceptical objection of Gorgias’ at On

Melissus, Xenophanes, and Gorgias, 980b9–14.

De An. 3.2, 426b10–12; also 2.6, 418a11–16; GA 5.1, 780b14–17, b29–31 (cf. b18–29); 5.2,

781a15–20. Cf. Sens. 4, 442b14–17; 7, 447b24–26.

APo. 2.19, 99b35; de An. 3.3, 428a3–5; 3.9, 432a16; MA 6, 700b19–21. Cf. Top. 2.4, 111a14–20;

de An. 3.12, 434b3–4; Metaph. 11.6, 1063a2–3: EN 10.4, 1174b34. On the rendering of κρίνειν as

“discriminate,” see Ebert’s seminal article (1983) also Nussbaum 1978: 334; and more recently de

Haas 2005; Corcilius 2014. For more detail on the conception of discrimination involved, see

Charles 2000: 112–13 n. 4, and Caston, “Aristotle on Perceptual Content.” (under review).

GA 5.2, 781a15–17: “To hear and smell accurately is, rst, to perceive the differences of the

underlying perceptibles as much as possible […]” (ἓν μὲν γάρ ἐστι τοῦ ἀκριβῶς ἀκούειν καὶ

ὀσφραίνεσθαι τὸ τὰς διαφορὰς τῶν ὑποκειμένων αἰσθητῶν ὅτι μάλιστα αἰσθάνεσθαι πάσας

[…]). GA 5.1, 780b15–17: “To see sharply is […] second, to discern the differences of the objects

seen as much as possible” (λέγεται γὰρ ὀξὺ ὁρᾶν […] ἓν δὲ τὸ τὰς διαφορὰς ὅτι μάλιστα

διαισθάνεσθαι τῶν ὁρωμένων). More generally, see GA 5.1, 780b14–33; 5.2, 781a14–20, b1–5;

cf. de An. 2.8, 420a9–11. A sense can also be said to perceive accurately or sharply, if it can per-

ceive objects at a great distance. But this depends more on the structures surrounding the sense

organ. For the Greek text of the GA, I use Drossaart Lulofs’ edition (1965).

Sens. 4, 442b14–15: “The most accurate sense is one that can discriminate the smallest [differ-

ences] in each kind” (τὰ γοῦν ἐλάχιστα τῆς ἀκριβεστάτης ἐστὶν αἰσθήσεως διακρίνειν περὶ

ἕκαστον γένος).

Aristotle’s notion is perhaps stronger than it need be, at least as regards survival, since the nest-

grained differences will not always matter, a point well-put at the beginning of Williamson 2013:

2 Aristotle ontheTransmission ofInformation: Receiving Form Without theMatter

compensates in such cases by wiring sensory powers more tightly to motivational

levers. Humans, for example, do not perceive odors without feeling a concomitant

pleasure or pain, and animals with hard eyes—certain insects and reptiles—do not

detect colors independent of fear or its absence. In both cases, Aristotle explains,

this is so precisely because their sense organs are not especially accurate (ὡς οὐκ

ὄντος ἀκριβοῦς τοῦ αἰσθητηρίου, de An. 2.9, 421a9–16). How well the senses

receive information about the world is thus crucial to the functional role of

perception.

2.4 How toReceive Form Without Matter

The doctrine of receiving form without the matter therefore is a locus of intentional-

ity, as Neoscholastic interpretations have claimed, at least in the sense that informa-

tion is received. But there is nothing magical about it. Receiving form without the

matter is not a basic type of change, which is primitive and unanalyzable, much less

an “immaterial” or “spiritual” change. This is clear from the example of the sealing

wax. It is a change that comes about entirely through ordinary material changes.

What distinguishes it is not materiality or immateriality, but whether it results in a

replica of the form received.

For Aristotle, an agent acts in so far as it has a given form F and this is the same

form that the patient receives. In ordinary changes, where the patient is “affected

along with the matter” (πάσχειν μετὰ τῆς ὕλης), the patient becomes F quite

literally: it comes to be F in just the same sense that the agent is F. This holds for

substantial and nonsubstantial changes alike, whether a man begets a man or the sun

warms a plant. The patient becomes a replica of the agent in the relevant respect: it

comes to be like the agent with respect to the agent’s form or active quality and so

is F in just the same sense that the agent is. Aristotle is thus committed to every

instance of the following schema:

If x receives F from y with the matter, x becomes a replica of y with respect to F—x comes

to be F in just the same sense that y is F.

By receiving a form in this way, an object embodies the form or active quality of the

agent in its own matter and so literally becomes the same kind of thing. The resulting

likeness may be partial in one way, in so far as it may be limited to just the active

quality or form. But it is total in another way, in so far as it embodies this form

literally. The water in the kettle literally becomes hot just like the burner and the

clay literally takes on the contours of the stick. Notice that even here the matter of

the agent and the patient may be different in kind, and not just different tokens of

the same type, especially with nonsubstantial change.

“Intelligent life requires the ability to discriminate, but not with unlimited precision.” I am grateful

to Matt Evans for pressing me on this point.

V. C a sto n

When a form F is received without the matter, on the other hand, the patient does

not literally become F—to change in this way is not to embody or replicate F. The

Neoscholastic interpretation emphasizes the negation. But we should give at least as

much attention to the qualication, which restricts the scope of negation. The claim

is not that there is not any replication. All that is denied is that the causal interaction

results in a replica of the active quality, F. It could therefore involve replicas of

other forms the agent possesses—G, H, and so on—forms that might be related to

F in relevant ways.

The case of sealing wax once again makes this clear. Even

though the signet is not replicated, its surface contours are; in fact, that is precisely

how the wax takes on the owner’s device or insignia. But taking on surface contours

is, as we have seen, a perfectly ordinary material change. So while the signet is

received without the matter, its surface contours are received with the matter.

Consequently, the sealing will be like the signet in various respects, respects that are

essential to the signet’s functioning as the particular signet that it is. But the wax

still does not replicate the signet: it becomes a sealing, not a signet. It thus receives

F and is also like F in relevant ways (namely, by sharing the same device), without

literally becoming F in the exact same way itself.

Receiving form without the matter, then, is not only compatible with other forms’

being received with the matter, it might even require this as an underlying change,

in just the same way that building a house requires activities like sawing and

hammering, and indeed the motions underlying them, such as gripping and swinging,

in turn. That is, it might well be the case that if a patient is to carry information

about F, without actually embodying F, it must nonetheless replicate the agent in

other, relevant respects—it must become G, H, and so on, in just the same sense that

the agent is. In fact, without any such correspondences, the content of the resulting

state would be a mystery: there would be nothing more that could be said to explain

why this particular state was about the agent or the agent’s being F; its aboutness

would be primitive and unanalyzable. But if there are underlying changes of an

ordinary sort, where other relevant forms are received with the matter, then there is

more that can be said. Let us examine more closely how such reception might work.

2.4.1 Transduction

When a form F is received without the matter, F is genuinely received and the

patient is genuinely likened to the agent, even though it does not become an exact

replica of F. Certain likenesses are preserved because F is received by receiving

Although it might seem supercially similar, Irwin’s distinction between a form of F that

“expresses F” and a form of F that “realizes F” (1988: §161, 308–9) is actually quite different.

First, he regards realization as a special case of expression and so a subset of it, whereas the two

kinds of receiving F I am distinguishing here are disjoint. Second, he appears to think of different

forms of F standing in these relations, whereas I have in mind a single form being received in dif-

ferent ways, which seems closer to the way Aristotle formulates the distinction.

2 Aristotle ontheTransmission ofInformation: Receiving Form Without theMatter

distinct, but relevantly related form(s). In this way, F is transmitted without itself

being literally replicated, because certain of its essential features are. Receiving

form without the matter is thus a type of transduction.

A familiar example may

help. Suppose I am listening to a vinyl record of Berg’s violin concerto.

What

makes this possible? A bow drawn across a violin string produces vibrations in the

string and then, as a result, in the air; these in turn are converted by a microphone

into changes in an electric current, which are then transposed into charges on

magnetic tape, and then later again into the grooves of a vinyl record; from there we

go back again, through the vibrations of a record stylus running in the record’s

grooves, to an electric current, which, after it passes through various circuits,

produces vibrations in a speaker cone and consequently in intervening air, until

ultimately these vibrations strike the listener’s ear. Only near the very beginning and

end of the process do we have sounds. But the latter is possible only because of what

intervenes. At each stage, a signal is transmitted, by causal means, through various

media: a certain abstract pattern of magnitudes and their variations are successively

embodied in different forms of energy. They share enough in common with the

original sound for this process to be regarded as a transmission of specic

information, which constitutes the signal. But this is achieved by the information

being embodied at each stage in certain material characteristics, often very different

from its original instantiation, that are peculiar to a given medium and the causal

processes that produce them.

Aristotle didn’t have a stereo system, of course. But he uses a similar example,

also involving sound reproduction, which is all the more pertinent because the inter-

vening “black box” in this case is lled not with electronics, but cognitive powers.

A student can repeat his teacher’s words: the sound he produces with his vocal cords

Transduction plays a central role in Pylyshyn 1984, who characterizes it as follows (151): “In its

most general sense, as used, for example, in electrical engineering, a transducer is a device that

receives patterns of energy and retransmits them, usually in some altered form. Thus a typical

transducer simply transforms or maps physical (spatiotemporal) events from one form to another

in some consistent way.” For more detailed discussion, see Pylyshyn 1984: ch. 6. His predominant

interest there is in constraining the notion of transduction in a way that will be adequate for his

computationalism. Accordingly, he is concerned with transduction as a bridge between physical

magnitudes in the world and their symbolic representation in the cognitive system, in a way that

excludes, for example, analog representation (see esp. 159 n. 2). Peter King (1994) regards this

narrower conception as posing a fatal problem for medieval Aristotelians, which is ultimately

responsible for the “collapse of their research program” and the “scientic paradigm” underlying

it, when they try to move from perception to understanding.

It is important to emphasize that transduction into symbolic representations is not what I have

in mind here. For although non-analog, symbolic representation might come into play with

Aristotle’s conception of the understanding of language (Int. 1, 16a3–16), it does not apply to

perception or to phantasia, which are not digital symbolic systems, but analog forms of representa-

tion (against, it seems, Silverman 1989: 279). Whether Aristotle’s theory of understanding found-

ers on the difculties King raises is something I hope to address elsewhere.

The example of an audio recording is famously exploited by Wittgenstein in Tractatus 4.014 (cf.

4.01). But I have more in mind than simply the sharing of “logical form.” What is crucial for trans-

duction is the way in which the transmission of form is effected, both in its particular material

instantiations and the causal processes that bring about the transitions from one to the other.

V. C a sto n

are of just the same sort as the sound he hears when his teacher speaks. It is, Aristotle

says, “as if they issued from one and the same stamp” (οἷον ἀπὸ χαρακτῆρος τοῦ

αὐτοῦ καὶ ἑνός).

In this case, the student’s utterances are actual replicas of the

teacher’s utterances, much as the sounds from my stereo speaker mimic the violin’s

sounds. But again they stand only at the beginning and end of the process. And what

interests us most is what happens in between: the intervening effect of the teacher’s

words on the student’s hearing and ultimately understanding. And hearing itself

cannot involve a replica, as Theophrastus explains: if hearing took place by means

of some further internal sound, we would face the same question all over again (τὸ

γὰρ αὐτὸ λείπεται ζητεῖν). Who would hear these internal sounds? And would they

be heard by means of still more interior sounds? To avoid a regress of homunculi,

hearing must receive sounds in a different way, without replication.

Aristotle

agrees. We do not hear sounds by making further sounds and then listening to them:

sounds inside the ears, Aristotle insists, would obstruct hearing (de An. 2.8,

420a7–11, a15–18). Instead, whenever we hear sounds, they are received without

the matter. They are received in a transduced form, where the sounds themselves of

the sounds are not replicated, just certain essential features of them.

The idea of transduction—that a pattern can be transmitted without the source

being fully replicated, by means of material changes that only share certain of its

essential features—is, I suggest, what the example of sealing wax is meant to

identify and what goes on in every case where form is received without the matter.

Aristotle, I claim, acccepts every instance of the following schema:

If x receives F from y without the matter, then for some relevantly related G

(i) x does not receive F from y with the matter, but receives it by receiving G

(ii) x receives G from y with the matter and so replicates G.

To say that F is received by receiving some “relevant” G is, of course, only to issue

a promissory note, which must be cashed out if the theory is to make any genuinely

substantive claims. But the relevant characteristic will be (i) something shared with

GA 5.2, 781a26–30: “For this reason learning spoken [words] comes about so that one might

repeat what was heard. For the change that enters through the sense organ is of the same kind again

as the change that issues from the voice, as though they had been produced from one and the same

stamp, so that one can say what one heard.” (διὸ καὶ ἡ μάθησις γίγνεται τῶν λεγομένων ὥστ’

ἀντιφθέγγεσθαι τὸ ἀκουσθέν· οἷα γὰρ ἡ κίνησις εἰσῆλθε διὰ τοῦ αἰσθητηρίου τοιαύτη πάλιν

οἷον ἀπὸ χαρακτῆρος τοῦ αὐτοῦ καὶ ἑνός διὰ τῆς φωνῆς γίγνεται ἡ κίνησις ὥσθ’ ὃ ἤκουσε τοῦτ’

εἰπεῖν.) I accept Barnes’ emendation of οἷα at a27, instead of the manuscripts’ οἷς.

De sens. 21, 505.12–15 Doxogr. Gr.: “But when [Empedocles] explains that hearing occurs by

means of internal sounds, making a sound inside, just like a bell’s, it is absurd to think that it is

clear how people hear. For we hear the external sounds because of it, but why [do we hear] when

it makes a sound? For the same thing still has to be found out.” (ἀλλὰ περὶ μὲν τὴν ἀκοὴν ὅταν

ἀποδῷ τοῖς ἔσωθεν γίνεσθαι ψόφοις, ἄτοπον τὸ οἴεσθαι δῆλον εἶναι πῶς ἀκούουσιν, ἔνδον

ποιήσαντα ψόφον ὥσπερ κώδωνος. τῶν μὲν γὰρ ἔξω δι’ ἐκεῖνον ἀκούομεν, ἐκείνου δὲ ψοφοῦντος

διὰ τί; τὸ γὰρ αὐτὸ λείπεται ζητεῖν, reading τὸ γὰρ αὐτὸ in l. 15 with the MSS, rather than

Wimmer’s correction, τοῦτο γὰρ αὐτὸ, which is printed in both Diels and Stratton.)

For convenience, I have used just a single variable here, G, but the quantier “some” is intended

to allow for the possibility that there may be more than one such form.

2 Aristotle ontheTransmission ofInformation: Receiving Form Without theMatter

the agent and (ii) essentially related to its being F, so that the patient will genuinely

and literally become like F in ways relevant to its being F. It will not be a replica of

F. But it will be a replica of G, and in a perfectly literal sense. It comes to be G in

just the same sense that the agent is G.

One might be tempted to think that, if this account is right, the reception of F is

“nothing over and above” the reception of G with the matter—that receiving F in

such cases might be reducible to receiving G, or even eliminable altogether—so that

talk of receiving F “without the matter” is little more than a façon de parler. But this

worry rests on an unfounded assumption: it presupposes that a form is genuinely

received only if it is replicated. And that is precisely what Aristotle is calling into

question in De anima 2.12, by distinguishing between two forms of reception, both

genuine. In receiving a form without the matter, the patient genuinely receives the

form and is genuinely likened and assimilated to the agent. It just doesn’t result in a

replica of that specic form. It is true that there is only a single token event involved

when receiving form in this way: receiving F and receiving G are, to use Aristotle’s

phrase, “one and the same in number.” But they are nevertheless two distinct types

of reception, since what it is to be each differs—in his terminology, they differ “in

being.” Receiving G is a perfectly ordinary reception with the matter, while F is

received without the matter; and the latter occurs through the former, by the relevant

Gs being received. Receiving F without the matter, then, will not be reducible to

receiving G, much less eliminable. Receiving G is simply how F is received; it is the

mechanism by which it is received. In short, receiving form without the matter is

implemented by receiving other forms with the matter. There can therefore be a

robust explanatory relation between the two.

This is clear in the case of the signet ring. There genuinely is such a thing as seal-

ing a document and it is not simply a matter of impressing shapes into wax, any

more than signing a document is simply making a scrawl. In neither case do we have

two separate token acts: a person does one in and by doing the other. In making the

relevant scrawl, one is signing a document, and in making the relevant impression,

one is sealing it. That is why one cannot sign a document without making a scrawl

or seal it without making an impression. But the two remain distinct types of act,

even if a single event instantiates both, since what it is to sign or seal a document is

different from what it is to make a scrawl or an impression. And in each case the

former is explained by the latter: just as an act is a signing because it is the making

of a certain kind of scrawl, so an act constitutes a sealing because it is the making of

a certain sort of impression.

Transduction occurs more widely than perception—as the case of sealing wax

makes clear, it only constitutes a necessary condition of perception, not a sufcient

one. But it may be especially well suited to cognitive processes. Aristotle himself

describes phantasmata as being like perceptual stimulations “without the matter”

(ὥσπερ αἰσθήματά ἐστι, πλὴν ἄνευ ὕλης, de An. 3.8, 432a9–10) from which they

On the social context that makes it possible for these material changes to be sealings and signa-

tures, see the beginning of §2.5.

V. C a sto n

are produced. In De memoria he says that the phantasmata retained in memory are

“imprinted like a kind impression of the perceptual stimulation, just like signatories

make with their rings.”

That might be hard to explain on some accounts: if the

perceptual stimulation was already the result of a form’s being received “immateri-

ally,” as Neoscholastic interpretations claim, it is not clear how the phantasma pro-

duced from it could get any more immaterial. In contrast, there is no problem with

iterations of transduction, at least in principle: successive transformations might

shave the signal down to ever more abstract patterns, while still preserving some

features from the original source. The notion of transduction might also help in

making sense of the analogy Aristotle draws between perception and understanding

at the beginning of De anima 3.4. He seems inclined to think of understanding, at

least in its most basic form, to be capable of receiving the form of its object (δεκτικὸν

τοῦ εἴδους, 429a15) as a result of being affected by its object or something like

being affected (ἢ πάσχειν τι ἂν εἴη ὑπὸ τοῦ νοητοῦ ἤ τι τοιοῦτον ἕτερον, a14–15).

What exactly is to be made of this depends upon how Aristotle conceives of the

object of understanding and the way it acts on our capacity to understand. But he

signals clearly enough that transduction is involved when in the next line he says

that the understanding itself is “potentially the sort of thing” its object is, though not

it (δυνάμει τοιοῦτον ἀλλὰ μὴ τοῦτο, a16). The understanding thus does not repli-

cate its object, much less embody it. But it is receptive of it, by becoming the sort of

thing the object is.

We might consider transduction to be an instance of an even more general idea

about how forms can be transmitted through certain changes without replication.

Aristotle’s account of sexual reproduction, for example, involves the transmission

of form in the opposite direction. Although the end result, the offspring, embodies

the form of the parent, the seed that produces the offspring does not, but only

transmits the form in virtue of certain “changes” (κινήσεις) it contains.

Cases of

production are similar in this respect too: the form in the mind of the doctor or the

builder leads to the instantiation of the form of health or the form of a house in

material bodies (Metaph. 7.7, 1032a32–b14; GA 2.1, 734b36–735a5). But the

Mem. 1, 450a30–32: ἐνσημαίνεται οἷον τύπον τινὰ τοῦ αἰσθήματος, καθάπερ οἱ σφραγιζόμενοι

τοῖς δακτυλίοις.

For more on both of these cases, see Caston 2005: 307–12.

While the seed is a part of a living human, Aristotle emphatically rejects the suggestion that the

seed itself is already a human being, a homunculus quite literally, as on preformationist accounts

(GA 2.1, 733b31–734b4). The seed is rather something that can produce a human being, in virtue

of the actual characteristics present in it, which are characterized as changes (734b4–735a4, esp.

734b8, 16–17, 22–23, 735a1–3). Aristotle thus thinks the seed possesses some relevant character-

istic G, distinct from the substantial form F it leads to, though essential to it, such that the matter

receives F by receiving G from the seed, analogous to transduction. But in this case, unlike percep-

tion, the matter becomes F in a literal sense, whereas the proximate agent, the seed, is not literally

F. In cases of receiving form without the matter, it is the reverse: the agent is literally F, whereas

the patient is not and will not as a result become F literally. In both processes, a form can be trans-

mitted in ways that do not involve replication. I would like to thank Alan Code for this suggestion

(in conversation).

2 Aristotle ontheTransmission ofInformation: Receiving Form Without theMatter

actions that transmit these forms do not themselves embody them. None of these

cases are transduction, as I have characterized it above, because of the direction of

causation. But they similarly involve the transmission of form without its exact

replica being embodied at every stage.

2.4.2 Ratios

We are now in a position to interpret the pronoun-ridden conclusion of the signet

ring passage at the beginning of De anima 2.12. To quote it again (424a21–24):

The sense for each [type of perceptible] is likewise affected in a similar way, by what has

color or avor or sound, not in so far as it is said to be each of them, but in so far as it is of

this sort, in virtue of its logos.

Matthen has recently criticized my use of the terminology of transduction (2019: 275–76) on the

ground that it undervalues later contributions to vision science as “merely a footnote to Aristotle”

(e.g. geometrical optics, the application of the theory of electromagnetism to the stimulation of the

retina, and the higher cognitive processing we now think happens subsequent to transduction in

vision). This is largely rhetorical, since I do not make any claims about Aristotle anticipating future

discoveries or about their relative importance. My point is merely that Aristotle’s explanation of a

second kind of reception in terms of sealing wax has similarities to our notion of transduction,

something Matthen does not contest and which is an integral part of our theories of perception. His

critique seems instead to rest on the assumption that I “equate visual sensing with transduction”

(276, my emphasis). But since I only claim that transduction is a necessary condition of percep-

tion, not a sufcient one, the conclusions he draws do not follow.

ὁμοίως δὲ καὶ ἡ αἴσθησις ἑκάστου ὑπὸ τοῦ ἔχοντος χρῶμα ἢ χυμὸν ἢ ψόφον πάσχει, ἀλλ’ οὐχ

ᾗ ἕκαστον ἐκείνων λέγεται, ἀλλ’ ᾗ τοιονδί καὶ κατὰ τὸν λόγον. It is useful to be fully explicit

about how each phrase is construed, since the literature generally has not been. I take the object

that has (τοῦ ἔχοντος) a given color, avor, or sound in 424a22 to be the implicit subject of the

verb “is said to be” (λέγεται) in a23 and “each of these” (ἕκαστον ἐκείνων) to be its predicate,

parallel to “is of this sort” (τοιονδί) in the next clause. The plural “these” (ἐκείνων) refers back to

the set of perceptible qualities just listed in a22: color, avor, sound, etc. If so, then the distinction

is between being affected by objects (a) in so far as they have a given perceptible quality and (b)

in so far as they are “of this sort” (τοιονδί), that is, in so far as they have a more general character-

istic that is not identical with the color, avor, etc., even though essential to it.

The alternatives are awkward. To read it as a contrast more favorable to the literalist, between

being affected by substances as such and being affected by them in so far as they are colored, a-

vored, and so forth, the plural pronoun “these” (ἐκείνων, 424a23) would have to refer back not to

color, avor, and sound, but rather to the singular expression “what has” (τοῦ ἔχοντος, a22) them,

and the phrase “each of them” (ἕκαστον) would also have to serve as the subject of the verb “is said

to be” (λέγεται, a23), as for example Hicks argues (1907: 416 adloc.). Hicks concedes that this

requires one to supply “each” implicitly a second time as the predicate, to get the proper antithesis

with “of this sort” (τοιονδί) in the next line (a24). One is not likely to nd such clumsiness prefer-

able unless one is already independently committed to literalism. Others, like Ward and Silverman,

read “these” (ἐκείνων) in the way I have suggested, as referring back to the perceptible qualities

just mentioned (Ward 1988: 220–21; Silverman 1989: 289 n. 9). We differ, though, in that they, like

Hicks, think that “each of these” (ἕκαστον ἐκείνων) is the subject of “is said to be” (λέγεται) and

take the subsequent contrast to be between being affected by the object in so far as it has a deter-

minable characteristic, like color, and in so far as it has a determinate characteristic, like crimson.

V. C a sto n

Though the sense is affected by “what has color, or avor, or sound,” it is not affected

in so far as the object is “said to be each of these” (οὐχ ᾗ ἕκαστον ἐκείνων λέγεται,

424a23), that is, not in so far as it is said to be of a certain “color or avor or sound”

(a22), for example, crimson, spicy, or shrill—in short, the determinate perceptible

qualities that we properly perceive by each sense. Rather, each is affected in so far

as the object is “of this sort, in virtue of its logos” (ἀλλ’ ᾗ τοιονδί καὶ κατὰ τὸν

λόγον, a24), so some kind to which these determinate qualities belong, having to do

with its logos, a term we need to clarify. Suppose I look at a red chili pepper. My eye

is affected by it and receives its visible quality, crimson. It does this, not by becom-

ing crimson itself, but by taking on a more general feature in virtue of which (κατά)

the object is crimson, namely, the proportion that on Aristotle’s view is an essential

feature of crimson and so belongs to its form and account. The phrase “in virtue of

its logos” in the passage above expresses all three of these senses—proportion,

form, and account—but especially the rst.

He is drawing a contrast between per-

ceptible qualities and the proportions or ratios that belong to their essence.

This might seem surprising. Proper perceptibles are qualitative for Aristotle and

he would certainly resist any attempt to reduce quality to quantity. He ridicules

theories that identify perceptible qualities like white, sweet, and hot with numbers

(Metaph. 14.5, 1092b15–16) and strongly rejects others, like Democritus’, which

reduce them to shapes (Sens. 4, 442b10–14). But he does not deny that qualities

exhibit quantitative features. On the contrary, he thinks the perceptible qualities

exclusive to each sense (ἴδια αἰσθητά) are dened by ratios of contrary qualities.

Perhaps this is most obvious in the case of tangible qualities, like hot and cold,

which differ in degree (μᾶλλον καὶ ἧττον): intermediate temperatures, in Aristotle’s

view, are to be explained straightforwardly in terms of proportions of the two

extremes (GC 2.7, 334b8–16, esp. b14–16). But there was also a widespread

tradition in Greece, which Aristotle knows and accepts, of treating the pitch of

sounds in terms of ratios.

Such an explanation might have been more easily

accepted in this case because it was possible to measure the quantities involved even

then.

But there were also theories of which Aristotle was aware, such as Archytas’

I do not see how this contrast would be relevant in context. I take Aristotle to have in mind deter-

minate perceptible qualities throughout (such as crimson, spicy and shrill), since they are what we

properly perceive by each sense, but he refers to them here simply by their genus.

This sense of logos is also crucial in the immediate sequel to our passage, where Aristotle char-

acterizes the sense itself as a kind of proportion (424a26–28) that, like the tuning of an instrument,

can be damaged by extreme changes that destroy the balance within the sense organ (a28–32; cf.

3.2, 426a27–b7).

For an excellent detailed discussion of Aristotle’s use of mathematics in his theory of perception,

see Sorabji 1972.

See e.g. APo. 2.2, 90a18–23; de An. 3.2, 426b3–7; GA 5.7, 786b25–787b20; Metaph. 1.2,

997b21; 10.1, 1053a15–17. For a clear and accessible introduction to the use of ratios in Greek

harmonics, see the introduction to Barker 1989, esp. 5–8.

The monochord, whose two sections were used to measure the ratios, appears to have been the

invention of a fth century BCE theorist, Simos (West 1992: 240–41). But ancient testimonia link

the discovery of ratios to other means of sound production as well: a smith’s differently weighted

2 Aristotle ontheTransmission ofInformation: Receiving Form Without theMatter

and Plato’s, that sought to explain pitch in terms of quantities that belong to the

sound itself, such as speed;

and Aristotle himself thinks that while different pitches

should not be identied with different speeds, differences in the former are to be

explained by differences in the latter (διὰ τὸ τάχος […] διὰ βραδυτῆτα, de An. 2.8,

420a26–b4). Aristotle is willing to extend quantitative analysis to the remaining

senses as well, explicitly relying on its success in music (Sens. 3, 439b30–440a6).

He discusses the ratios of contrary qualities in his account of both colors and avors

at some length.

And while he does not explicitly apply them to odors, he takes the

latter to be so closely correlated to avors that the specic differences within each

type are said to be ordered between their respective extremes in just the same way,

so that the quality spaces have analogous structures.

It is important to emphasize that these qualities are not dened solely by the ratio

or number on its own.

Qualities in different sense modalities might well share the

same abstract ratio. As with pitches, Aristotle recognizes only a nite number of

basic colors and avors, corresponding to certain whole number ratios,

and some

of these might well agree numerically. They will not be the same quality, though,

since proportions for Aristotle are always the proportions of something: the numbers

that stand in proportion to one another are the amounts of the items being compared,

such as constituents in a mixture.

Perceptible qualities will therefore be dened as

proportions of some specic pair of contrary qualities. So even if crimson and spicy

were to share the same numeric ratio, they would still be proportions of different

contraries: one is a proportion of white and black, the other a proportion of sweet

and bitter. The resulting qualities will therefore differ, even though they are

analogously the same (cf. κατ’ ἀναλογίαν, GC 2.6, 333a28–30), in virtue of the

abstract proportion they share.

To receive the form of crimson in the eye, then, does not entail that any part of

our eye literally becomes crimson, any more than that some part of us becomes a

stone when we perceive one (1.5, 410a8–13; 3.8, 432b20–432a1). But there must be

some relevantly related predicate that is true of both crimson and some part of our

eyes, and in just the same sense. What the passage at the beginning of this section

(424a22–24) claims is that the organ will embody the same logos or proportion of

contrary qualities: just as a crimson object has a certain proportion of light and dark,

so my eye will have the same proportion between two of its contrary

hammers, strings held taut by different weights, cymbals of different thicknesses, and vessels lled

with different amounts of liquid (for detailed critical discussion, see West 1992: 234).

Archytas 47 B 1 DK; Plato Timaeus 67–, 97–80.

Colors: Sens. 3, 440b14–26. Flavors: Sens. 4, 442a12–31; Metaph. 10.2, 1053b28–1054a13.

Odors’ relation to avors: Sens. 4, 440b28–30; 5, 443b3–20.

Against Silverman 1989, which emphasizes that what is received is only the abstract numerical

ratio (280, 289 n. 8, 290 n. 16); for criticism, see Everson 1997, 97, 99.

Whole number ratios of colors: Sens. 3, 439b30–440a2. Of avors: Sens. 4, 442a12–16.

See esp. Metaph. 10.2, 1053b28–1054a13. The same point underlies his critique of Pythagorean

and Platonic appeals to ratios: Metaph. 14.5, 1092b16–22; cf. 1.9, 991b13–20.

V. C a sto n

qualities—perhaps heat and cold, or viscosity and uidity, or some other pair.

One

could, if one likes, say that my eye is colored in a way (ὡς κεχρωμάτισται, de An.

3.2, 425b22–23), in so far as it embodies the dening proportion of a color. But it

will be the same only in an analogical sense (cf. Metaph. 5.9, 1018a13), since it will

not embody this proportion in the same contrary qualities and so will not become

crimson in the literal sense that the object is.

Aristotle is thus not drawing a distinction in the opening of De anima 2.12

between a substance and its qualities, as the traditional interpretation assumes. He

is drawing a distinction between two sorts of qualities: the perceptible quality itself

and some essential feature of that quality, such as its dening proportion or ratio.

This distinction enables him to rene the sense in which a colored object, for

example, affects us. It is not its color tout court, but rather an essential or dening

feature of the color, which makes it the very color that it is.

As a result, we receive

the perceptible quality in a transduced form, “without the matter,” by receiving its

dening proportion with the matter and embodying it in a different pair of contrar-

ies. We thus take on an actual feature of the perceptible object, its ratio, in virtue of

which it has that perceptible quality, and in so doing acquire information about the

character of things in our environment. We do not replicate the perceptible

quality itself.

Perceptual content arises, at least in part, because we are able to take on certain

abstract features essential to perceptible qualities and thus receive information

about them. The differentiating characteristics of perceptual states represent the

characteristics of perceptible objects because they stand in a strict analogical rela-

tion to them.

Against Everson 1997: 97, who assumes too hastily that the only available pair of contraries

would be the same pair underlying the external perceptible quality and so would entail a literal

replication of the quality. I should add that though Aristotle does not offer any details, I am assum-

ing that for each organ, there would be a specic pair of contraries in which the proportions of its

own perceptibles are embodied. There might be some story to tell as to why one pair is used rather

than another. But there also might not: it might simply be a brute fact that has to be accepted with

“natural piety.”

Lorenz (2007: 193 n. 29) dismisses my view in Caston 2005, on the grounds that it would com-

mit Aristotle to the sense’s not being affected by the perceptible as such, but only incidentally. This

would hold of Silverman 1989 (not cited by Lorenz), who states repeatedly that the ratio is an

incidental or extrinsic characteristic (271, 272–73, 280–81, 285). But it mischaracterizes my view,

which explicitly holds that the sense is affected by the perceptible in virtue of one of its essential

characteristics, namely, the dening ratio of the perceptible quality that makes it the specic qual-

ity that it is (Caston 2005: 314–15). Indeed, it is precisely because it is through an essential feature,

and not some extrinsic characteristic, that it counts as a reception of that very quality, even if it is

reception “without the matter.” It thus does not contradict Aristotle’s views about intrinsic percep-

tion, as Lorenz claims. For more on colors and other perceptible qualities, see Caston 2018).

2 Aristotle ontheTransmission ofInformation: Receiving Form Without theMatter

2.5 The Authority ofPerception

In De anima 2.12, perception is distinguished from other natural processes in so far

as the senses are affected in a very specic way, where perceptible forms are

received “without the matter”; and this, I have argued, is a kind of transduction

where only certain essential features of the forms are preserved, transmitting

information about the characteristics of objects in our environment. But it is worth

reecting further on Aristotle’s choice of signet rings and sealing wax as a model,

and the implications this has for intentionality.

2.5.1 The Seal ofReality

The example of sealing wax is meant to be a familiar one his readers would have

been acquainted with. The strategy is similar to the one Aristotle employs in Physics

2in developing his own distinctive conception of teleology. We will more readily

understand nal causes, he thinks, if we start by considering artifacts, since they are

better known to us (ἐκ τῶν γνωριμωτέρων ἡμῖν), even though in his own view the

best examples of teleology are not artifacts at all, but occur in nature without design

or deliberation—things like spider webs, swallows’ nests, and the leaves of plants

(2.8, 199a20–32). In such cases, we see nal causation in its purest form, independent

of the aims of a conscious being, and thereby arrive at what he believes is “naturally

clearer and better known” (ἐπὶ τὰ σαφέστερα τῇ φύσει καὶ γνωριμώτερα). So too

with a familiar artifact like sealing wax, where we can more easily see how informa-

tion is transmitted by ordinary material changes. This allows us to understand how

the same thing can occur in a case like perception, where information is transmitted

naturally without social conventions or practices. It is plausible to think that here as

well Aristotle regards the natural case as the more fundamental one, which in itself

can be better known, even though it is not so initially to us.

Comparing mental states to impressions in wax is not new with Aristotle, of

course. There are many references earlier in Greek literature, for example, to

inscribing words on “the tablets of the heart.”

The implication here is also that

these mental states come to possess content, by using an example familiar from

everyday life in antiquity, of inscribing words on wax tablets. But this comparison

also differs signicantly from Aristotle’s.

In the rst place, it typically concerns memory, not perception, with storing a

record of events that can be retained over a long period and later accessed when

E.g. Aeschylus, Supplices 179, Prometheus Vinctus. 789, Agamemnon 80, Choephori 450,

Eumenides 275; for close discussion of the use of the metaphor in these passages, see Sansone

1975: ch. 4; see now Agócs 2019, esp. Section 2. Cf. also Pindar, Odes 10.2–3; Sophocles,

Philoctetes 1325, Trachiniae 683. Plato’s imagery in the Philebus and Theaetetus is related to

these (see below). On its use in the later Platonic tradition, see Sheppard 2017.

V. C a sto n

needed. But second, the wax tablet image here is essentially linguistic. The content

is expressed in words and so naturally suggests representation in a “language of

thought.” No such thing is required by Aristotle’s account. A sealing will signify the

owner of the ring, by means of the owner’s emblem or insignia. But this is typically

pictorial, not linguistic. For Aristotle the senses are not inscribed,

as they are in

Plato’s playful comparison of the soul to a book, written by an inner scribe (Philebus

38–39).

Rather they are impressed by the perceptible qualities themselves and

thereby come to possess the insignia, as it were, of the perceptible object, the

outward signs by which the underlying substance makes itself known.

Plato does

famously speak of signet rings and wax in the Theaetetus (191–196), which is

often cited in this context as a parallel. But what is not usually appreciated is that

Plato’s comparison there is not to perception, but to memory, and accordingly for

Plato the sealing is not produced by perceptible objects, but by perceptions and

thoughts (ταῖς αἰσθήσεσι καὶ ἐννοίαις, 1917, 1944, cf. 1956). So even when

Plato uses sealing wax as a model for mental states, it is more closely aligned to the

literary gure of the tablets of the heart, in so far as both concern memory. Aristotle

is alive to this difference and in fact helps himself to a number of the details in his

own account of memory, especially regarding the relevance of the underlying

material conditions (Mem. 1, 450a27–b11).

In comparing sealing wax to perception and not memory in De anima 2.12,

Aristotle is departing from this tradition. In this respect, he is much closer to other

predecessors. Gorgias, for example, says in the Encomium of Helen that “the soul is

impressed by the objects we see through sight” and the “images of objects seen are

etched into the mind.”

His contemporary, Democritus, is likewise reported to have

compared the effect of a perceptible object to “the sort of imprinting you might

make in wax,” although in his case the effect is not the one made directly on the

sense organ, but the prior impression made on the intervening medium, air, which in

turn transmits its effect to the organ.

Against Kalderon 2015: 172.

Plato also mentions an inner painter (ζωγραφόν) who provides illustrations, but this is subse-

quent to the scribe’s writing (μετ

ὰ

τὸν γραμματιστὴν τῶν λεγομένων, 39b6). If Sedley 2004:

137–38 is right, the wax block model in the Theaetetus should also be understood in line with this

passage, and so as “discursive” and involving “internal verbalization.”

To use Matthen’s phrase, what they provide perceptions with is precisely a “place-of-origin

stamp” (2019: 281). But such a mark does not have to be recognized by the subject in order for the

perception to be about the object (pace Matthen). On this account, the fact that it originates in this

way is sufcient: it is a natural sign. There is no homunculus. (Similar remarks apply to Matthen’s

criticisms at 283.)

Gorgias, Encomium of Helen, 15 (ed. Donadi): ἃ γὰρ ὁρῶμεν […] διὰ δὲ τῆς ὄψεως ἡ ψυχὴ κἀν

τοῖς τρόποις τυποῦται. 17: οὕτως εἰκόνας τῶν ὁρωμένων πραγμάτων ἡ ὄψις ἐνέγραψεν ἐν τῷ

φρονήματι.

Theophrastus, De sens. 51 (= Doxogr. Gr. 514.1–2=DK 68 A135): παραβάλλων τοιαύτην εἶναι

τὴν ἐντύπωσιν οἷον εἰ ἐκμάξειας εἰς κηρόν. Cf. also De sens. 52 (= Doxogr. Gr. 514.5–6=DK 68

A135): ὁ ἀὴρ ἀπομάττεται καθάπερ κηρός. Aristotle may have in mind Democritus’ views about

the effect on the medium and subsequent effect on the sense organ, when he discusses impressions

being made in wax right through to the other side at de An. 3.12, 435a1–10, esp. a10: ὥσπερ ἂν εἰ

τὸ ἐν τῷ κηρῷ σημεῖον διεδίδοτο μέχρι τοῦ πέρατος.

2 Aristotle ontheTransmission ofInformation: Receiving Form Without theMatter

The third and by far the most important difference, though, is the epistemologi-

cal dimension of Aristotle’s model, which derives from the legitimating function of

seals. Sealings are used to verify that a document stems from a person with the

relevant authority, who sanctions its contents. In the case of perception, this amounts

to the suggestion that objects in the world give their seal and thereby their authority

to the messages the senses report: our perceptions have the backing of the world and

so provide us with a warranted basis from which to form beliefs and act. In this

respect too, Aristotle’s use of sealing wax as a model differs critically from Plato’s.

If the memories Plato is concerned with in the Theaetetus carry any authority, it is

only because the perceptions that stamped them already possessed some authority.

Aristotle, in contrast, aims to explain this prior fact by applying the signet ring

model to perception itself. Perceptions carry authority because the world legiti-

mates them: they are not mere appearances, because of the way they come about.

They are the true coin issued by the world.

Of course, it is one thing to claim this and another to have established it. As

Athenian law well recognized, there are fraudulent sealings, even if it is very

difcult to counterfeit the signet itself:

a given sealing might not originate from the

person it purports to have originated from, even if it comes from the right signet,

since someone else might have stolen it or used it. In much the same way, one might

reasonably worry that some of our perceptual experiences do not come from the

objects they seem to and thus misrepresent their source. So even if there is an

authoritative way of receiving information, like sealings, it cannot completely rule

out the possibility of fakes. It cannot provide a panacea for all of our epistemic ills.

Aristotle would fully acknowledge the justice in this complaint. He does not

wish to deny the existence of error or its salience. Just the opposite. Error, to use his

words, is “endemic” (οἰκειότερον) to animals, who spend “a great deal of time” in

this state (πλείω χρόνον). In fact, he regards this as a datum that any adequate

psychology must account for, and he ridicules his predecessors precisely because he

thinks they cannot account for it, while he believes that his own theory can.

But he

explains it by appealing to a distinct new ability he calls phantasia.

So it does not

belong to perception, at least not in its most basic form. To his mind, then, the prob-

lems raised by error do not undermine the ultimate authority of perception itself.

In this respect, Aristotle’s comparison has much in common with the Stoics’. See Caston, The

Stoics on Content and Mental Representation (in progress).

This law, like many others, is ascribed to Solon: Diogenes Laertius 1.57; Diodorus Siculus 1.78.

De An. 3.3, 427a29–b6: “Yet they should have also said something at the same time about mak-

ing errors, for this is endemic to animals and the soul spends a great deal of time in this state. For

it is necessary [on their theory] that either (i) all appearances are true (as some have said); or (ii)

contact with what is unlike is error (since that is the opposite of recognizing like by like). But both

error and knowledge of contraries seem to be the same.” (καίτοι ἔδει ἅμα καὶ περὶ τοῦ ἠπατῆσθαι

αὐτοὺς λέγειν, οἰκειότερον γὰρ τοῖς ζῷοις, καὶ πλείω χρόνον ἐν τούτῳ διατελεῖ ἡ ψυχή· διὸ

ἀνάγκη ἤτοι, ὥσπερ ἔνιοι λέγουσι, πάντα τὰ φαινόμενα εἶναι ἀληθῆ, ἢ τὴν τοῦ ἀνομοίου θίξιν

ἀπάτην εἶναι, τοῦτο γὰρ ἐναντίον τῷ τὸ ὅμοιον τῷ ὁμοίῳ γνωρίζειν.)

For a close examination of these arguments in context, see Caston 1996.

V. C a sto n

It is easy to understand the motivations behind such a move. Counterfeits are

effective only if sealings are for the most part genuine. The success of counterfeits

depends on the presumption that the tokens in question are genuine, and this

presumption cannot survive for long if there are too many counterfeits.

Fraud is

viable only if there is widespread reliablity, and part of what ensures this reliability

in the case of seals is the extraordinary precision and accuracy of their details.

the case of perception, this must be true as well. Perception could not perform its

proper function if it did not succeed “for the most part” in informing us about quali-

tative differences between objects—any process that didn’t simply wouldn’t count

as perception. Just as wax sealings must be hard to counterfeit if they are to authen-

ticate documents reliably and serve as sealings, so too the stimulation of our percep-

tual organs cannot be informative in the way its function requires unless it is difcult

to simulate in ordinary circumstances. It is not that it cannot be mimicked, or that

we cannot be taken in. Obviously, it can and we often are. But this is only because

perceptions in general are reliable and so standardly reveal genuine features of the

surrounding world to us.

Aristotle insists repeatedly that perception has this sort of authority. In

Nicomachean Ethics 6.2, he identies perception as one of three faculties in the soul

that have authority over action and truth (τρία δή ἐστιν ἐν τῇ ψυχῇ τὰ κύρια

πράξεως καὶ ἀληθείας). But unlike the other two (understanding and desire), per-

ception is not “the source of any action” (οὐδεμιᾶς ἀρχὴ πράξεως, 1139a17–19). It

is only in charge of truth—its business, if you will, is the acquisition of truth.

Archaeologists have discovered glass seals that all share the same insignia, sometimes referred

to as “look-alikes.” But something analogous holds in their case too: they function as seals only if

they represent a single ofce or institution, where more than one person is authorized to act on its

behalf (Younger 1999). The sealings thus still purport to come from a single source, even if not

from a single person.

See Sines 1992, who provides the data to show that the level of precision was typically “greater

than what could be attained by normal vision unaided by a lens” (53); see esp. 67–68.

This general line of argument will be familiar to many readers from Donald Davidson’s argument

against radical scepticism based on the principle of charity (2001). But the underlying intuition is

not original with Davidson. Aristotle makes a similar assumption about people’s beliefs, the

“appearances” or phainomena from which he thinks all investigation must start. He rules out the

possibility that anyone’s beliefs could be wholly wrong: in order for a proposition to be found cred-

ible by a believer, there must be something to recommend it to a being naturally suited to learning,

even if what is believed is wrong in other ways, perhaps signicantly so (Rh. 1.1, 1355a15–18;

Metaph. 2.1, 993a30–b11). But there are also important differences from Davidson about the

underlying intuition, which arguably favor Aristotle. When it comes to beliefs Aristotle does not

hold anything as strong as Quine and Davidson’s principle of charity, since for Aristotle it is not

the case that most of a person’s beliefs, simply taken as such, will be true—most beliefs will be

true only in part and only when construed in a certain way (beyond just differences in reference).

At the same time, Aristotle holds something stronger when it comes to perception, at least with

regard to some. As I will presently show, he holds that all of a person’s most basic perceptions must

be true without exception, though this is not essential to the present point, which is that the com-

parison with sealings suggests their general reliability.

It is clear that perceptual beliefs, on the other hand, at least when they function as part of a practi-

cal syllogism (EN 7.3, 1147a25–26), can exercise authority over our actions (ἡ τελευταία

πρότασις δόξα τε αἰσθητοῦ καὶ κυρία τῶν πράξεων, b9–10).

2 Aristotle ontheTransmission ofInformation: Receiving Form Without theMatter

Elsewhere in the Ethics Aristotle is even more specic. Perception, he says, has

authority over particulars (περὶ τῶν καθ’ ἕκαστά ἐστιν, ὧν αἴσθησις ἤδη κυρία,

7.3, 1147a26): when it comes to apprehending how things stand in our environment,

it is perception that provides the decisive information. In the opening of the

Metaphysics, he goes even further and declares that our perceptions constitute the

most authoritative knowledge of such things (καίτοι κυριώταταί γ’ εἰσὶν αὗται τῶν

καθ’ ἕκαστα γνώσεις, 1.1, 981b11), even though they do not tell us why anything is

the way it is (οὐ λέγουσι τὸ διὰ τί περὶ οὐδενός), such as why re is hot (b12). But

they do tell us that (ὅτι) it is hot and more generally how things are (b13). The task

of natural science, he says in the De caelo, is to give an account of “what consis-

tently appears authoritatively to perception” (τὸ φαινόμενον ἀεὶ κυρίως κατὰ τὴν

αἴσθησιν, 3.7, 306a16–17).

The senses do not all possess the same authority in all cases, though. When they

disagree, some are more authoritative than others (κυριωτέρα, Insomn. 3, 461b4–5).

Sight, for example, has more authority than touch as regards the number of objects

perceived (2, 460b20–23). But within its own domain, the authority of each sense is

supreme: when I spy something sweet on the counter, it is still taste, and not sight,

that is the nal arbiter of its avor (Metaph. 4.5, 1010b14–17).

Aristotle actually holds something even stronger. The senses have authority, in

part, because each sense is infallible about certain objects, namely, the qualities that

are intrinsically perceptible to it exclusively (ἴδια αἰσθητά). It is “not possible,” he

says, for a sense “to be mistaken about” such things (περὶ ὃ μὴ ἐνδέχεται

ἀπατηθῆναι, de An. 2.6, 418a12); consequently, the perception of them is “true” or

“always true.”

He is not saying merely that we do not make any general mistake

The Greek text is taken from D.J. Allan’s OCT edition (1965).

Aristotle thinks we have a general predisposition to believe what is presented through the senses

if there is no report from another sense that conicts with it; and even when there is this sort of

conict, we may accept the deliverance of a given sense anyway, if our judgement (τὸ ἐπικρῖνον)

is impaired (Insomn. 3, 461a25–462a8, esp. 461b4–7).

Kenny (1967: 192–93) rightly notes that for Aristotle sight can be mistaken about the color of an

object, but nonetheless it can only be corrected by itself on another occasion, and not by another

sense. But while Aristotle maintains that we can be wrong about the color of a given object, we are

not mistaken about the color itself: see below, pp.43–44.

De An. 3.3, 427b11, 428b11, b18; 3.6, 430b29; Sens. 4, 442b8–10; and also Metaph. 4.5,

1010b1–3, if we insert <μὴ> before ψευδής at b2, with both Ross and Jaeger. On one occasion,

Aristotle qualies this universal generalization (“or has the least falsehood”, de An. 3.3, 428b19);

see below, pp.49–50.

These strong formulations rule out two attempts to construe Aristotle’s view more weakly, as

holding either (a) that the senses have a kind of incorrigibility about their own objects or (b) that

perceptions of these objects are not truth-apt or evaluable. For (a), see Kenny 1967: 193 and the

comments in n. 80 above. For (b), see Kalderon 2015, who argues that there is no possibility of

error in such cases “because the sensing of a primary object fails to be evaluable as correct or

incorrect” (67), although he acknowledges the “potential embarrassment of explaining away”

Aristotle’s repeated use of “true” in these contexts “as merely loose talk” (68). On his view,

Aristotle is a naïve realist, who thinks that we “simply confront” the perceptible quality; one can-

not confront something correctly or incorrectly (68). Embarrassment isn’t the issue, though, but

V. C a sto n

regarding the types of objects each sense perceives—for example, that what we see

are colors and what we taste are avors.

Rather he is claiming that we do not make

mistakes about which specic quality we are perceiving—his go-to examples are

white and sweet—even though he thinks we do make mistakes about which object

possesses it or about its location (418a15–16; Metaph. 4.5, 1010b19–21).

In fact,

he is even more precise. He allows that (i) perception is not always true about which

sort of object that has a color, for example, about whether what is white is a human

(de An. 3.6, 430b29–30). But (ii) it can also be mistaken as to which particular

object in our enviroment is white, whether “this, or something else, is white” (εἰ δὲ

τοῦτο λευκὸν ἢ ἄλλο, 3.3, 428b21–22).

It follows from this, importantly, that we

can be mistaken about the color of an object: when the distant mountains look pur-

ple, we are correct in seeing purple, but mistaken that the mountains are purple;

their “proper” color (τὸ οἰκεῖον χρῶμα) is different, perhaps a mottled brown and

green.

In neither case, however, do we make a mistake about the colors them-

selves: the perception that it is white is not mistaken (ὅτι μὲν γὰρ λευκόν, οὐ

ψεύδεται, 428b21); and seeing something exclusively perceived by sight, like

inconsistency. It is not just that the passages cited above all expressly mention truth, but that one

of them further offers an extended discussion which correlates the truth and falsehood of percep-

tions of the three types of perceptibles with the truth and falsehood of the representations generated

from them (de An. 3.3, 428b18–429a2) and thus concerns truth evaluability throughout.

As Hamlyn argues on conceptual grounds (1959: 12; Hamlyn 1993: 106 ad 418a11), with Kirwan

(1993: 111) concurring; possibly also Ross (1961: 238), Osborne (1983: 406) and Vasiliou (1996:

123). Hamlyn himself acknowledges that itis difcult to know what to make of the view or its

import; though for a more positive attempt, see Vasiliou 1996. Against this reading, see Kenny

1967: 191; Denyer 1991: 186–87; Johnstone 2015: 314; and Koons 2019, who treats the issue at

very great length.

In addition to the textual evidence above, I would add two objections. First, Hamlyn cannot

interpret the other passages we are considering here in an analogous way and so is forced to say

that De anima 2.6 is making a distinct and independent point (1959: 12), which is highly implau-

sible in light of the close parallels. But secondly, we cannot make sense of what intrinsic percep-

tion is supposed to get unerringly right, on Hamlyn’s interpretation. For Aristotle does not think

that a single sense can intrinsically perceive the genus of its own exclusive perceptibles: he says in

Metaphysics 13.10 that although sight does see the universal, color, it does so only extrinsically,

because the particular color it sees is in fact a color (ἀλλὰ κατὰ συμβεβηκὸς ἡ ὄψις τὸ καθόλου

χρῶμα ὁρᾷ ὅτι τόδε τὸ χρῶμα ὃ ὁρᾷ χρῶμά ἐστιν, 1087a19–20). (This poses a problem for Koons

2019 too, given his incautious formulation of “specic infallibilism” at 419.) If sight is infallibly

right about anything, it is about which particular colors it happens to be seeing.

Hamlyn in fact acknowledges (Hamlyn 1993: 106, 134) that Aristotle mentions a specic color,

white, in passages outside of De anima 2.6, but does not revise his interpretation in light of this,

instead holding (134–35) that De anima 2.6 makes a different point than 3.3, 428b18. See also

Hamlyn 1959: 12, 15.

A point rightly noted by Krips 1980: 83.

Aristotle speaks of the proper color of an object at De anima 2.7, 419a2, 6 (cf. HA 5.19, 551b9);

he also denies that transparent materials, like water, have their own distinctive (ἴδιον) color; rather

they exhibit different colors at different distances, which they possess only extrinsically (κατὰ

συμβεβηκός). He contrasts it with solid bodies with determinate boundaries, where “the appear-

ance of their color” (ἡ φαντασία τῆς χρόας) is similarly deteminate (Sens. 3, 439a18–19, b1–6,

b12–14).

2 Aristotle ontheTransmission ofInformation: Receiving Form Without theMatter

something white, is always true (τὸ ὁρᾶν τοῦ ἰδίου ἀληθές), though it is not always

true with regard to whether it is human (οὐκ ἀληθὲς ἀεί, 3.6, 430b29–30).

The

sense of taste likewise always reports truly about what is sweet (ἀεὶ ἀληθεύει περὶ

αὐτοῦ [sc. τὸ γλυκύ], Metaph. 4.5, 1010b23–26): when something tastes differ-

ently at different times, “there is actually no dispute about the characteristic, but just

about that to which the characteristic belongs” (περί γε τὸ πάθος ἠμφισβήσεν, ἀλλὰ

περὶ τὸ ᾧ συμβέβηκε τὸ πάθος, 1010b19–21).

Though one can be deceived about

common perceptibles, “regarding those exclusive to a sense one is not deceived, for

example sight about a color or hearing about sounds” (περὶ δὲ τῶν ἰδίων οὐκ

ἀπατῶνται, οἷον ἡ ὄψις περὶ χρώματος καὶ ἀκοὴ περὶ τῶν ψόφων, Sens. 4,

442b8–10; cf. de An. 3.3, 428b23–25). For convenience, we can call this most basic

form of perception, of perceptibles intrinsic to a single sense exclusively, “sensa-

tion,” to distinguish it from other types of perception.

Ross marks the passage as corrupt in both of his editions, but the MSS are in unanimous agree-

ment. He does not explain his reservation, moreover; he might be concerned about the verb ὁρᾶν

taking an objective genitive (τοῦ ἰδίου), following Beare (1906: 90 n. 2) and Hicks’ (1907: 524

adloc.). But a genitive object, while rarer, is not without parallel: cf. LSJ 9, ὁράω, II.1.e.

Against Hamlyn, when he claims that sight can be in error over which color is being seen

(1993: 106).

It bears emphasis that this is solely for convenience, since while Aristotle draws terminological

distinctions between these different kinds of perception, he uses a single verb, αἰσθάνεσθαι, for

them all. Moreover, I strenuously disagree with commentators, going back to Alexander of

Aphrodisias (De anima 41.9), who think that for Aristotle the most basic form is the only form of

perception, strictly speaking. (For discussion of Alexander, see Caston 2012: 15–16 and 148–49 n.

366.) When Aristotle distinguishes three types of perception in De anima 2.6, he regards them all

as genuine forms of perception. What I have called “sensation” is simply the most central, funda-

mental kind, which is exactly what Aristotle means when he says that in their case, perception is

said κυρίως, not “strictly”; and the essence of each sense dened in relation to it (de An. 2.6,

418a24–25). I defend this inclusive approach at greater length in Caston, “Aristotle on Perceptual

Content” (under review).

I also disagree with Matthen 2019, who thinks that Aristotle’s view should be framed in terms

of sensation for substantive reasons. He takes Thomas Reid’s distinction between sensation and

perception as critical to our own understanding of these words, where sensation “lacks any essen-

tial intrinsic signicance beyond itself,” while perception “essentially and by its intrinsic nature

intimates the presence of something outside itself” (271, original emphasis). Even though he

acknowledges that Aristotle was unlikely to have made this distinction, he nonetheless believes

that Aristotle’s notion of aisthēsis is captured more accurately by “sensation” in Reid’s sense, and

that it is a mistake to characterize it as perception, so understood. This couldn’t be further from the

truth: Aristotle standardly characterizes the activity of our senses as being of qualities of objects in

the external world, which they independently possess (see above, p.43) and a number of times

even says that the senses report on the qualities of external objects. Matthen, it should be noted, is

consciously and deliberately being uncharitable: his central thesis is that Aristotle’s theory is hope-

lessly awed in a way that was only corrected much later by geometrical optics and modern vision

science. (He does not mention Aristotle’s remarks about geometrical optics or his own views in

Meteorology 3). In the abstract, of course, an uncharitable reading might be correct. But surely it

has to answer to the full range of texts. Otherwise it is just cherry-picking and prejudicial.

V. C a sto n

2.5.2 The Basis ofInfallibility

Given the subsequent tradition, it might seem tempting to take Aristotle’s claims as

based on a contrast between the immediate contents of consciousness and our

knowledge of the external world: even if we cannot know the causes of our

experience, we can be certain about how things appear to us within experience, a

contrast we nd in the Cyrenaics and Pyrrhonists on down.

But this cannot be

what Aristotle has in mind. For if we were infallible in that way, it would hold for

perceptual contents quite broadly: I could not be mistaken that there appears to be a

man in the distance or that he appears to be 100 yards away any more than that there

appears to be something white.

Aristotle, however, is drawing a distinction between

these kinds of appearances; and, as he vigorously maintains against Protagoras, not

all appearances are true (οὐ πᾶν τὸ φαινόμενον ἀληθές, Metaph. 4.5, 1010b1–2).

Aristotle is therefore not concerned with just the face value content of our experience,

about what we seem to be experiencing, but with whether that content accurately

matches up against the world. It is for just this reason, in Aristotle’s view, that man

is not “the measure of what is and what is not”—the world, rather, is the measure of

man. Knowledge and perception can only be called “measures,” he believes, because

they do measure up the way things are. But most appearances do not (Metaph. 10.1,

1053a31–b3).

Cyrenaics: Plutarch, Adversus Colotem 24, 1120–; Sextus Empiricus, Adversus Mathematicos

7.191–94. Pyrrhonists: Diogenes Laertius 9.104–6; Sextus Empiricus, Outlines of Pyrrhonism 1.4.

Graeser 1978 may be a fellow traveller, when he says he is tempted to treat sensibilia as sense data

because it would help explain why they are “practically liable to error,” though he adds that

Aristotle’s view “as it stands does not really qualify as a sense-data theory” for unstated reasons

(91). Sorabji 2004: 82 and Schoeld 1978: 119 (cf. 122) note that what Aristotle calls aisthēmata

are a near equivalent to sense-data.

As Block 1961: 3; Krips 1980: 83; and Ben-Zeev 1984: 120 correctly note. Kenny 1967 makes

this point against what he calls the “Rylean” interpretation, though not, curiously, against the

“Berkeleian” one (191–92).

In this passage, Aristotle mischievously claims that what Protagoras meant in putting forward his

measure doctrine was only that knowledge and perception are measures, the latter being something

Aristotle takes to be obviously true and nothing to fuss about. But Book Kappa of the Metaphysics

recognizes that the measure doctrine makes a more radical claim, namely, that whatever seems or

appears to be the case is true (Metaph. 11.6, 1062b12–15). And while on that view such states

would also qualify as knowledge and veridical perception, it would not be in contrast to other

appearances’ being false (as it would for Aristotle), since for Protagoras there are no false appear-

ances. Therefore the most that Aristotle can claim in good conscience would be that his interpreta-

tion is the only charitable interpretation of the measure doctrine, because it captures the only part

of Protagoras’ claim that has a hope of being true.

Turnbull argues that Aristotle is nonetheless committed to a moderate form of Protagoreanism,

at least as regards colors and other qualities that are exclusive to one sense, holding that they exist

as such only in so far as they are being perceived, much like the “Secret Doctrine” in Plato’s

Theaetetus (155–157); and that for this very reason our perception of them is “logically” infal-

lible, unlike our perception of spatial and temporal features (1978: 4–5). I argue against Protagorean

interpretations at length in Caston 2018.

2 Aristotle ontheTransmission ofInformation: Receiving Form Without theMatter

When Aristotle says in Metaphysics 4.5 that we never make a mistake about the

quality itself (περί γε τὸ πάθος), but only about which thing it belongs to (περὶ τὸ

ᾧ συμβέβηκε), he is making a claim about the quality present in the world

(1010b19–21), not the quality or affection of the perceiver (τοῦ αἰσθανομένου

πάθος), which he identies as the perceptual stimulation (τὸ αἴσθημα, b33–34).

The perceptible quality is something that is “different from and in addition to the

perception” (ἔστι τι καὶ ἕτερον παρὰ τὴν αἴσθησιν, b36) and produces it (ἃ ποιεῖ

τὴν αἴσθησιν, b34). In Aristotle’s view, then, the most basic form of perception,

sensation, gives us unerring information not about our own state of mind, but about

some feature of the world around us. Its report is true of some part of our environ-

ment, to which the quality sensed genuinely belongs.

This claim would be extremely implausible if it concerned the distal objects of

perception. As Aristotle recognizes, objects often appear to have different colors at

different distances no less than different sizes (Metaph. 4.5, 1010b5–6). The same

wine can likewise taste different to a single perceiver at different times, depending

on one’s condition (1010b21–23), since the same objects do not seem sweet to both

a healthy person and a sick one, or hot to both someone who is debilitated and

someone who is in good shape (EN 10.5, 1176a13–16). In all of these cases, the

immediate conditions around the sense organ heighten or mask our perception of

certain qualities. The bitter uid that coats a sick person’s tongue makes everything

taste bitter, in the same way that strongly avored food temporarily makes it difcult

to discern other tastes (de An. 2.10, 422b7–10). If this is how the conditions of the

organ affect its sensitivity in general, then the errors involved will be just like what

happens with rose-tinted glasses. Something similar can obviously be said about

how the quality of the medium affects the mistakes we make about the colors of

distant objects, such as the sun appearing crimson through mist or smoke;

or how

differences in illumination lead weavers to choose the wrongly colored threads by

lamplight, or how an object’s being placed next to different colors (even in perfectly

normal conditions) can lead one to mistake the true color of the dyed threads.

The

presence of perceptible qualities close to the organ or in the surroundings alters or

masks the effect of qualities further away and prevents us from perceiving an

object’s proper qualities. But our sense will still be accurately picking up the

perceptible qualities of something external and will correctly instantiate its

The stimulation (αἴσθημα) is the modication of the perceiver (τοῦ αἰσθανομένου πάθος,

Metaph. 4.5, 1010b33) or, as Book Kappa says, what is produced (ποιοῦντα) by perceptibles in the

perceiver (11.6, 1063b4).

Sens. 3, 440a10–12, cf. 439b5–6; Mete. 3.4, 370b10–11.

Mete. 3.4, 375a22–28.

V. C a sto n

proportions.

The mistakes perception is liable to make only concern whether a

The only possible exception is from Book Kappa of the Metaphysics, which describes cases

where the quality an object appears to have is due solely to the condition of the organ. But the

authenticity of Kappa is a matter of some debate. It undoubtedly stems from Aristotle’s school, if

not Aristotle’s own hand. But for that very reason, it may be no more of a secure guide to Aristotle’s

views on precise details than Theophrastus is, who agrees with Aristotle extensively, but also

develops and inects his views in distinctive ways. So we cannot put great weight on evidence

from Kappa without conrmation from other parts of the Aristotelian corpus, and it should not be

used as the controlling element in an interpretation. For the debate over authenticity, see the valu-

able summary, with references, in Nielsen 2017: 304 n. 3. For a discussion of Theophrastus’

Aristotelianism, see Caston 2019.

Chapter 6 of Kappa maintains that the same object does not appear (φαίνεται) sweet to some

people and the opposite to others, unless someone’s organ and “criterion” (κριτήριον) for avors

is “decrepit or damaged” (διεφθαρμένων καὶ λελωβημέωνων, 1062b36–1063a3); it later suggests

that the very same object, without undergoing any change, can appear dissimilarly if the perceiv-

er’s condition is not similar to a healthy person’s (τὸ μὴ ὁμοίως διακεῖσθαι τὴν ἕξιν καὶ ὅθ’

ὑγίαινον) and so will produce different stimulations in those who are ill (1063a35–b6). This would

t with an interpretation that many have found tempting, starting with Alexander of Aphrodisias

(De anima 41.13–42.3), that Aristotle thinks that sensation is infallible only in normal conditions,

construed quite broadly: where one’s organs are functioning as they ought, given their nature, in

the external conditions for which they are naturally suited (situated at the right distance, without

obstructions, in appropriate illumination, and so forth). For contemporary interpretations, see

Block 1961, perhaps the earliest defense of a “normality theory of perception” on teleological

grounds (5–9); also Gaukroger 1978: 106 (cf. 91–92); Gaukroger 1981; Ben-Zeev 1984; Charles

2000: 122–24; Johnstone 2015; Koons 2019; for some criticism, see Krips 1980. For discussion of

Alexander’s view with full references, see Caston 2012: 149–51; see also Johnstone 2015, who

argues that Aristotle’s position is “essentially the same” as Alexander’s.

But this is largely wishful thinking. Even though one might have expected Aristotle to embrace

a normality theory, given his emphasis on teleology, there is not a single passage in the rest of the

corpus that explicitly says that sensation is true only in normal conditions, despite his having the

language to do so, or even a passage that entails it. Aristotle, for example, never says that any

perception is true “always or for the most part” (ἀεὶ ἢ ὡς ἐπὶ τὸ πολύ), nor does he explain the

reliability of the senses by reference to the end or function of perception; nor is there any reason to

think that a teleological justication is entailed by the infallibility of sensation about which specic

perceptible is being perceived (as Koons 2019 claims without argument at 436, 437), even if the

converse were true. The best that can be done is to argue, as Barnes 1987 does (56–64), that a

“schematic argument” can be reconstructed from Aristotle’s teleological commitments. But it is a

long way from saying that the senses are necessary for an animal’s survival (e.g. de An. 3.12,

434a30–b2; Sens. 1, 436b18–a3)—or even the conjecture that they must be generally reliable for

this purpose—to saying that they are infallible with respect to their sensations (as Barnes recog-

nizes (1987: 66–68, 70–74; cf. 62–63), but Koons does not (2019: 431–34)), at least if “normal

conditions” are specied in a substantive and non-circular way.

Kappa’s claims about abnormal internal conditions, then, are something of an outlier. It is out

of step in other respects too. (1) In claiming that the same object never tastes differently to different

perceivers, except when an organ is decrepit or damaged (1062b36–1063a3), the rst passage

directly conicts with Metaphysics 4.5, which acknowledges that there are disagreements about

which avor belongs to a given object, even for a single perceiver over time (1010b19–26); it only

denies that one is mistaken about which avor we are tasting. De anima 2.10, moreover, explains

the mistakes made when ill as due to an external condition, namely, the bitter uid suffusing the

tongue (2.10, 422b8–10); and the second passage from Kappa (1063a35–b6) is actually compati-

ble with this explanation. (2) The word κριτήριον, moreover, though commonplace in Hellenistic

epistemology, is a hapax legomenon in Aristotle. Its use here is likely inuenced by one of its two

instances in Plato, at Theaetetus 178b6, whose context likewise concerns Protagoras’ relativism.

But one would still expect a term Aristotle uses more commonly like κρῖνον. It would not be sur-

prising for a member of his school.

2 Aristotle ontheTransmission ofInformation: Receiving Form Without theMatter

given quality belongs to this or that distal object, the same mistake Aristotle ags

later in Metaphysics 4.5 (1010b20–21) and in De anima 3.3 (428b21–22).

The most that Aristotle can count on, then, is the character of the proximal stimu-

lus, the condition of the medium where it is in contact with the organ and directly

affects it. And it is easy to see why he might think that perception is infallible here.

When a sense is affected intrinsically by perceptible qualities exclusive to that sense

(de An. 2.6, 418a24–25), the form it receives in that causal interaction is the very

same active quality in question. There is no room for deviation in genuine sensa-

tions: the reception of a given form, F, in sensation could not come about from any

other perceptible quality—a sensation of F could only come about by the action of

F on the organ.

What justies Aristotle’s commitment to the infallibility of sensa-

tion is not formal or teleological aspects of perception, but efcient causal ones: it

does not derive from doctrines about the essence of the individual senses or their

Johnstone (2015: 317) confuses these two questions, when he infers from (1) the claim at

1010b3ff. that we can make mistakes about “the true size or color of a thing” or “the true avor of

a thing” (emphases mine) that (2) we can misperceive perceptibles like white or bitter. Aristotle

clearly afrms that we can make a mistake about which properties belong to particular objects, but

he denies that we are mistaken about the quality itself (Metaph. 4.5, 1010b19–21). A similar analy-

sis applies to Mete. 3.4, 374b14–15 (cf. a18–19) and Sens. 3, 439b3–5, where Aristotle is again

speaking of which colors in fact belong to certain distant objects, and other passages where a

specic object appears to have a color other than its actual one due to surrounding colors or the

condition of the medium (374b10–11, 375a22–28; Sens. 3, 439b5–10), or the sound of a given

object appears differently due to disturbances in the medium (6, 446b6–9).

Johnstone dismisses my alternative as a stretch and as less charitable than his own (318 n. 18).

But in fact mine closely mirrors the contrast in the Metaphysics 4.5 passage just cited (περὶ τὸ

πάθος vs. περὶ τὸ ᾧ συμβέβηκε, 1010b19–21), which his discussion overlooks; and it preserves

the consistency of the Metaphysics and De anima passages without having to supply tacit quali-

cations (such as “for the most part”), something his interpretation does not (319).

Kenny (1967) comes close to this when he says “there is no room for a mistake between the

detecting of the quality of sweetness in a thing, when it is there, and the identication of this qual-

ity as sweetness” (195), except for the phrase “in a thing.” As we saw above, Aristotle thinks that

judgements that assign a quality to an object or a location can be mistaken. But with genuine sensa-

tions we are never mistaken that there is such a quality in the world.

This may not be at odds with the thesis of “Fungibility” Matthen attributes to Aristotle, which

holds that a sense could be in the exact same state as a result of a different object acting on it in

sensation (2019: 277), since Matthen seems to be primarily concerned with different tokens of the

same quality, whereas I am concerned with different types. But he may intend a broader version of

Fungibility that would conict: at 278, he asserts that it is possible for a green object to produce

the effect a blue one would, due to the interference of certain unspecied conditions; but he fails

to take into account the sorts of conditions Aristotle considers, discussed above. Matthen’s obser-

vation that we can be mistaken about which object a sensation comes from, on the other hand, is

something Aristotle himself emphasizes (see p.43 above) and is not in question here. It is also not

evidence that Aristotle is only concerned with sensations in Matthen’s sense. Aristotle emphasizes

that what we perceive are public, external objects at Sens. 6, 446b22–23 (see n. 34 above). The

most Matthen can claim is that Aristotle is not entitled to think this, not that he doesn’t think it. On

whether the same qualititative state can be produced outside of sensation by other means, see n.

101 and p.50 below.

V. C a sto n

normal functioning,

but rather his view that sensation is just an instance of agent-

patient interactions more generally, where the exact same quality active in the agent,

F, is received by the patient as a result of the interaction.

100

This is no less true where

F is received “without the matter”: even though we do not ourselves embody F in

such changes, but only other forms essentially related to it, in Aristotle’s view we

nonetheless receive F. Sensation is always about what brings it about, namely, the

intrinsic and exclusive perceptible quality that is present and directly acting on the

sense organ; it thus covaries strictly with its proximate cause.

101

There is no ambigu-

ity in the signal in sensation: it yields unequivocal information about some part of

the surrounding world. A sensation will therefore always be true of some proximal

stimulus; and if conditions are right, it might tell us something accurate about the

distal object as well. Aristotle seems to be fairly optimistic that this is what gener-

ally happens: the way we are affected by the medium by and large gives us evidence

that identies the character of distal objects. But as we have seen, he also acknowl-

edges that it is not universally the case. We do make mistakes about which proper-

ties belong to a given distal object, because we sometimes take it to have

characteristics due in part to how we are affected by intervening conditions.

The only phenomenon Aristotle discusses that might cause hesitation here would

be afterimages, where after prolonged exposure to something bright we seem to see

changing patches of color that are not due to any external stimulus currently acting

on our eyes. Such a case might lead someone to question the claim that sensations

are always true, since it can occur in “normal” conditions, where the perceiver’s

sensory apparatus is in working order, the medium well-illuminated and

unobstructed, and surrounding objects at a good distance to be observed. It is to

address this sort of case, I suggest, that Aristotle adds an eirenic qualication

towards the end of De anima 3.3 when discussing the truth of sensations, so as not

to have to go into the issue more deeply there: perception of exclusive perceptibles,

he says, is true “or possesses the least possible amount of falsehood” (ἢ ὅτι

ὀλίγιστον ἔχουσα τὸ ψεῦδος, de An. 3.3, 428b18–19). For he continues as though

no qualication had been made at all: he says of the next case, perceiving an object

to have an extrinsic characteristic, that “it is at just this point that it is possible to be

mistaken” (καὶ ἐνταῦθα ἤδη ἐνδέχεται διαψεύδεσθαι, 428b20). As Jonathan

For an appeal to formal causes, see Marmodoro 2014: 85–86; for an appeal to nal causes, see

the list of commentators who interpret Aristotle’s view in terms of normal conditions in n. 95

above; and for discussion pitting the two against each other, see Koons 2019, esp. 430–40. As

Koons points out, appeals to the essence of the senses cannot justify more than infallibility about

the genus of a sense’s perceptibles (on which, see n. 82 above). But teleological justications far

overshoot their mark: while teleological considerations might plausibly require a certain reliability

for survival, they in no way require infallibility (against Koons 2019: 431, 432, 437). See n.

95 above.

100

For a detailed examination of Aristotle’s views about the efcient causal role of perceptible

qualities, see Caston 2018.

101

Because there may be similar states, produced by other causes, that are not sensations (see

below, p.50), the relation between cause and effect will be many–one; hence the covariation will

be in one direction only.

2 Aristotle ontheTransmission ofInformation: Receiving Form Without theMatter

Barnes points out, Aristotle’s choice of words “strongly suggests that falsity is

impossible in the case of proper objects.”

102

If so, it is a modal claim much like the

one in De anima 2.6 (418a12–13). So there is no departure from his view elsewhere.

That Aristotle had phenomena like afterimages in mind seems to be conrmed by

the sort of case he describes a few lines later: a representation generated from a

sensation will be true while the sensation is present (παρούσης τῆς αἰσθήσεως);

falsehood is possible only if it persists after the current sensation ceases (de An. 3.3,

428b27–28; cf. 3.2, 425b24–25). So it is no surprise that when he discusses

afterimages explicitly in De insomniis (2, 459b7–13), he does not treat them as

sensations at all, but he takes them to be evidence of phantasia: they are a quasi-

perceptual experience, due to the original stimulation persisting in our organs after

the external stimulus has ceased and possibly undergoing further alterations.

103

Afterimages do not, therefore, constitute a genuine exception to the unqualied

claim he makes on every other occasion, that sensations are always true, even if they

might seem so at rst glance. When we are actually perceiving, and not merely

having some quasi-perceptual experience, the only mistakes we make about

perceptibles exclusive to a single sense concern where they are located in our

environment and whether they in fact belong to some given object.

This is not to say that the same pattern of stimulation of the sense organ could not

be produced by other means than the direct action of perceptible qualities in the

environment on our senses. Aristotle plainly thinks it is possible to bring this type

of stimulation about in other ways. If that happens, it will result in an experience

like sensation, even though nothing in our vicinity has the perceptible quality in

question. Such states would not be genuine sensations, though, but merely like

them, in that they would be qualitatively similar to sensations and purport to be from

external objects; and for just this reason we can be misled by them (Insomn. 3,

461a25–b2).

104

They are like a fraudulent sealing: although it purports to be from

the proper authority, its origins are in fact deviant and its authority void, even if

someone accepts it as genuine. Aristotle not only acknowledges this possibility, he

exploits it in order to explain error (Insomn. 2, 460b22–25). For him the crucial

point, though, is that this sort of experience is not a genuine sensation, but involves

phantasia, and so it is not a counterexample to his claims about infallibility. This

will raise sceptical worries of the sort that would later come to be called the “problem

of the criterion,” of how we can tell which of our experiences are the veridical ones.

But for Aristotle the importance of infallibility is not anti-sceptical. He never

attempts to build a certain and indubitable foundation out of the infallible reports of

sensation. Its importance lies rather in sensation’s role as a transducer. For however

102

Barnes 1987: 55 n. 14. Shields (2016: 291) likewise recognizes that this claim “implicitly

rescinds the qualication” at 428b19, which he had earlier regarded as “a signicant divergence”

(290). Hicks seems cognizant of the difculty when he says that the second case is the point where

“serious error becomes possible” (1907: 471 adloc., my emphasis). But the Greek does not hedge

in the way Hicks needs it to.

103

See Caston 1998: 272 n. 56.

104

Against Matthen 2019: 279.

V. C a sto n

limited reports of proximal stimulations are, they are still a source of genuine and

unequivocal information about our immediate environment. And this is necessary if

his account of intentionality is to get off the ground.

105

2.6 Conclusion

When Aristotle characterizes each sense as a capacity for “receiving form without

the matter,” he is concerned with intentionality after all, as the Neoscholastics had

alleged. The senses are capable of receiving information about the world and this, in

his view, sets their activity apart from most other natural processes. But there is

nothing ghostly or immaterial about it: perception is something fully realized by

material events in an animal’s organs. It is a form of transduction, where the

perceptible form of an object is received not by embodying and replicating it, but by

preserving certain essential features of it, exemplied in different material qualities.

Such transduction carries authority, like the sealing impressed in wax, because of

the way in which it is produced. In a genuine sensation, a particular effect on the

organ can be brought about only by the perceptible quality itself; therefore, if we are

having a sensation, we cannot be misinformed about the presence of the quality in

the environment. This, Aristotle believes, provides a sufcient basis for perception’s

more general authority about particulars, even though we can and sometimes do

attribute such qualities to the wrong things. For when we attend carefully to the

deliverances of sense more broadly, they still provide a reliable guide for both

survival and knowledge. Such epistemological optimism is not his primary concern,

though, but intentionality. The infallibility of sensation anchors the content of our

intentional states. For however widely these may range, they are tethered to actual

features of the world.

106

Bibliography

Primary Sources

Aristotle. (1907). De anima (R.D. Hicks, Ed.). Cambridge: Cambridge University Press.

105

Johnstone 2015 says he sees “no principled reason why” Aristotle could not allow error in such

cases. But presumably this counts as a principled reason.

106

I am grateful to many people for the comments and criticisms on earlier drafts, especially the

detailed remarks of Rachel Barney, Matt Evans, Sean Kelsey, Baron Reed, and Michael Wedin, as

well as to audiences at Santa Clara University, Humboldt-Universität zu Berlin, University of

Victoria, University of Michigan, Yale University, University of Arizona at Tucson, University of

Colorado at Boulder, Oakland University, Northern Illinois University, University of Pittsburgh,

and Renmin University (Beijing).

2 Aristotle ontheTransmission ofInformation: Receiving Form Without theMatter

Aristotle. (1912). De anima libri III (A.Förster, Ed.). Budapest: Academia Litterarum Hungarica.

Aristotle. (1924). Metaphysics. A revised text with introduction and commentary (W.D. Ross, Ed.,

2 Vols). Oxford: Clarendon Press.

Aristotle. (1961). De anima (Edited, with Introduction and Commentary, by W.D. Ross). Oxford:

Clarendon Press.

Aristotle. (1963). Parva naturalia, graece et latine (P.Siwek. Desclée & C.i Editori Pontici, Eds.,

Vol. 5). Rome. (= Collectio Philosophica Lateranensis)

Aristotle. (1965a). De generatione animalium (H.J. Drossaart Lulofs, Ed.). Oxford: Clarendon Press.

Aristotle. (1965b). De caelo, libri quattuor (D.J. Allan, Ed.). Oxford: Clarendon Press.

Augustine. (1968). De Trinitate libri quindecim (W.J. Mountain, Ed., CCSL 50/50A). Turnhout:

Brepols.

Diels, H., & Kranz, W. (Eds.). (1952) Fragmente der Vorsokratiker (6th ed.). Berlin: Weidmannsche

Buchhandlung. [= DK]

Gorgias. (2016). Helenae encomium (F. Donadi, Ed., Bibliotheca scriptorum Graecorum et

Romanorum Teubneriana, BT 2019). Berlin: De Gruyter.

Plato. (1900–1907). Platonis opera (J.Burnet, Ed., 5 Vols). Oxford: Clarendon Press.

Theophrastus. (1879). De sensibus. In: Doxographi Graeci (H.Diels, Ed.). Berlin. [= De sens.]

Thomas Aquinas. (1984). Sentencia libri De anima (Opera omnia iussu Leonis XIII P.M. edita

45.1). Rome/Paris: Commissio Leonina/J.Vrin. [= Sent. de An.]

Secondary Literature

Agócs, P. (2019). Speaking in the wax tablets of memory. In L.Castagnoli & P.Ceccarelli (Eds.),

Greek Memories: Theories and Practices (pp.68–90). Cambridge: Cambridge University Press.

Barker, A. (Ed.). (1989). Greek musical writings (2 Vols. Vol. 1: Harmonic and acoustic theory).

Cambridge: Cambridge University Press.

Barnes, J. (1987). An Aristotelian way with scepticism. In M. Matthen (Ed.), Aristotle today:

Essays on Aristotle’s ideal of science (pp.51–76). Edmonton: Academic Printing & Publishing.

Beare, J.I. (1906). Greek theories of elementary cognition from Alcmaeon to Aristotle. Oxford:

Clarendon Press.

Ben-Zeev, A. (1984). Aristotle on perceptual truth and falsity. Apeiron, 18, 118–125.

Block, I. (1960/1961). Aristotle and the physical object. Philosophy and Phenomenological

Research, 21, 93–101.

Block, I. (1961). Truth and error in Aristotle’s theory of sense perception. The Philosophical

Quarterly, 11, 1–9.

Boardman, J. (1970). Greek gems and nger rings: Early bronze age to late classical. London:

Thames and Hudson.

Brentano, F. (1867). Die Psychologie des Aristoteles, insbesondere seine Lehre vom NOUS

POIHTIKOS. Mainz: Verlag von Franz Kirchheim.

Caston, V. (1996). Why Aristotle needs imagination. Phronesis, 41, 20–55.

Caston, V. (1998). Aristotle and the problem of intentionality. Philosophy and Phenomenological

Research, 58, 249–298.

Caston, V. (2002). Aristotle on consciousness. Mind, 111, 751–815.

Caston, V. (2005). The Spirit and the letter: Aristotle on perception. In R.Salles (Ed.), Metaphysics,

soul, and ethics: Themes from the work of Richard Sorabji (pp.245–320). Oxford: Oxford

University Press.

Caston, V. (2012). Alexander of Aphrodisias, on the soul, Part I: Soul as form of the body, parts

of the soul, nourishment, and perception (Translated, with an introduction and commentary.

Ancient commentators on Aristotle series). London: Bristol Classical Press.

V. C a sto n

Caston, V. (2018). Aristotle on the reality of colors and other perceptibles. Res Philosophica,

95, 35–68.

Caston, V. (2019). Theophrastus on perceiving. Rhizomata, 7, 188–225.

Caston, V. [in progress] The Stoics on content and mental representation.

Caston, V. [under review]. Aristotle on perceptual content.

Charles, D. (2000). Aristotle on meaning and essence. Oxford: Clarendon Press.

Christodoulopoulou-Proukake, A. (1977). Πήλινο σφράγισμα δακτυλιολίθου ἀπὸ τὴν Ἀθήνα.

Ἀρχαιολογική Ἐφημερίς, 116, 164–170.

Corcilius, K. (2014). Activity, passivity, and perceptual discrimination in Aristotle. In J.F. Silva

& M.Yrjönsuuri (Eds.), Active perception in the history of philosophy: From Plato to modern

philosophy (Studies in the history of philosophy of mind, vol. 14) (pp.31–53). Cham: Springer.

Davidson, D. (2001). A coherence theory of truth and knowledge. In: D.Davidson (Ed.), Subjective,

intersubjective, objective (pp.137–157). Oxford: Clarendon Press; reprint of D.Henrich (Ed.)

(1983). Kant oder Hegel? (pp.428–438). Stuttgart: Klett-Cotta.

de Haas, F.A. J. (2005). The discriminating capacity of the soul in Aristotle’s theory of learning.

In R.Salles (Ed.), Metaphysics, soul, and ethics in ancient thought (pp.321–344). Oxford:

Clarendon Press.

Denyer, N. (1991). Language, thought and falsehood in ancient Greek philosophy. London:

Routledge.

Dretske, F. (1981). Knowledge & the ow of information. Cambridge, MA: The MIT Press.

Dretske, F. (1988). Explaining behavior: Reasons in a world of causes. Cambridge, MA: The

MIT Press.

Ebert, T. (1983). Aristotle on what is done in perceiving. Zeitschrift für philosophische Forschung,

37, 181–198.

Everson, S. (1997). Aristotle on perception. Oxford: Clarendon Press.

Faraone, C.A. (1999). Ancient Greek love magic. Cambridge, MA: Harvard University Press.

Gaukroger, S. (1978). Explanatory structures: A study of concepts of explanation in early physics

and philosophy. Highland Park: Humanities Press.

Gaukroger, S. (1981). Aristotle on the function of sense perception. Studies in History and

Philosophy of Science, Part A, 12, 75–89.

Graeser, A. (1978). On Aristotle’s framework of sensibilia. In Aristotle on mind and the senses.

Proceedings of the seventh symposium Aristotelicum (pp. 69–97). Cambridge: Cambridge

University Press.

Hamlyn, D.W. (1959). Aristotle’s account of aesthesis in the De anima. Classical Quarterly, n.s.,

9, 6–16.

Hamlyn, D.W. (1993). Aristotle De anima, Books II and III (with passages from Book I) (Translated

with an introduction and notes by D.W. Hamlyn, with a report on recent work and a revised

bibliography by Christopher Shields). Oxford: Clarendon Press.

Hicks, R.D. (1907). Aristotle, De anima. With translation, introduction, and notes. Cambridge:

Cambridge University Press.

Irwin, T. (1988). Aristotle’s rst principles. 1988. Clarendon press, Oxford

Johansen, T.K. (1998). Aristotle on the sense-organs. Cambridge: Cambridge University Press.

Johnstone, M.A. (2015). Aristotle and Alexander on perceptual error. Phronesis, 60, 310–338.

Kalderon, M. E. (2015). Form without matter: Empedocles and Aristotle on color perception.

Oxford: Oxford University Press.

Kenny, A. (1967). The argument from illusion in Aristotle’s Metaphysics (Γ, 1009–10). Mind, 76,

184–197.

King, P. (1994). Scholasticism and the philosophy of mind: The failure of Aristotelian philosophy.

In T. Horowitz & A.I. Janis (Eds.), Scientic Failure (pp.109–138). Lanham: Rowman &

Littleeld Publishers, Inc.

Kirwan, C. (1993). Aristotle Metaphysics, Books Γ, Δ, and Ε (Translated with notes, 2nd ed.).

Oxford: Clarendon Press.

Koons, B. (2019). Aristotle’s infallible perception. Apeiron, 52, 415–443.

2 Aristotle ontheTransmission ofInformation: Receiving Form Without theMatter

Krips, H. (1980). Aristotle on the infallibility of normal observation. Studies in History and

Philosophy of Science, Part A, 11, 79–86.

Lacroix, L. (1955–1956). Les ‘blasons’ des ville grecques. Études d’archéologie classique

1:91–115.

Lennox, J. (2001). Aristotle, On the parts of animals. Translated with a Commentary. Oxford:

Clarendon Press.

Lorenz, H. (2007). The assimilation of sense to sense-object in Aristotle. Oxford Studies in Ancient

Philosophy, 33, 179–220.

Marmodoro, A. (2014). Aristotle on perceiving objects. Oxford: Oxford University Press.

Matthen, M. (2019). Is the eye like what it sees? A critique of Aristotle on sensing by assimilation.

Vivarium, 57, 268–292.

Nielsen, K.M. (2017). Spicy food as cause of death: Coincidence and necessity in Metaphysics Ε

2–3. Oxford Studies in Ancient Philosophy, 52, 303–342.

Nussbaum, M. (1978). The role of phantasia in Aristotle’s explanation of action. In Aristotle’s De

motu animalium. Text with translation, commentary, and interpretive essays (pp. 221–269).

Princeton: Princeton University Press.

Osborne, C. (1983). Aristotle, De anima 3.2: How do we perceive that we see and hear? The

Classical Quarterly, 33, 401–11.

Owens, J. (1981a). Aristotelian soul as cognitive of sensibles, intelligibles, and self. In J.R. Catan

(Ed.), Aristotle: The collected papers of Joseph Owens (pp.81–98). Albany: State University

of NewYork Press.

Owens, J. (1981b) Aristotle: Cognition a way of being. In: J.R. Catan (Ed.) Aristotle: The col-

lected papers of Joseph Owens (pp. 74–80). Albany: State University of New York Press;

reprint of Canadian Journal of Philosophy 6 (1976), 1–11.

Plantzos, D. (1999). Hellenistic engraved gems. Oxford: Clarendon Press.

Pylyshyn, Z.W. (1984). Computation and cognition: Toward a foundation for cognitive science.

Cambridge, MA: The MIT Press.

Richter, G.M. A. (1968). The engraved gems of the Greeks, Etruscans, and Romans. Vol. 1: A his-

tory of Greek art in miniature. London: Phaidon.

Sansone, D. (1975). Aeschylean Metaphors for Intellectual Activity. Wiesbaden: Franz Steiner

Verlag. (= Hermes Einzelschriften, Heft 35.)

Schoeld, M. (1978). Aristotle on the Imagination. In G.E. R.Lloyd and G.E. L.Owen (Eds.),

Aristotle on mind and the senses (pp. 99–141). Proceedings of the Seventh Symposiuim

Aristotelicum, 27 August–5 September 1975, Cambridge. Cambridge: Cambridge

University Press.

Sedley, D.N. (2004). The midwife of Platonism: Text and subtext in Plato’s Theaetetus. Oxford:

Oxford University Press.

Sheppard, A. (2017). The soul as a writing tablet, from Plato to Proclus. In J. Finamore and

S. Klitenic Wear (Eds.), Dening Platonism: Essays in honor of the 75th birthday of John

M.Dillon (pp. 218–27). Steubenville, Ohio: Franciscan University Press.

Shields, C. (2016). Aristotle De Anima. Translated with an introduction and commentary. Oxford:

Clarendon Press.

Silverman, A. (1989). Color and color-perception in Aristotle’s De anima. Ancient Philosophy, 9,

271–292.

Sines, G. (1992). Precision in engraving of Etruscan and archaic Greek gems. Archeomaterials,

6, 53–68.

Sorabji, R. (1972). Aristotle, mathematics, and colour. The Classical Quarterly, 22, 293–308.

Sorabji, R. (1991). From Aristotle to Brentano: The development of the concept of intentional-

ity. In H.Blumenthal, &H.Robinson (Eds.), Aristotle and the later tradition (pp.227–259).

Oxford: Clarendon Press. (= Oxford Studies in Ancient Philosophy, suppl. vol. 1991.)

Sorabji, R. (1995). Intentionality and physiological processes: Aristotle’s theory of sense-

perception. In M.C. Nussbaum & A.Oksenberg Rorty (Eds.), Essays on Aristotle’s De anima

(pp.195–225). Oxford: Clarendon Press (First paperback edition, including an additional essay

by M.F. Burnyeat; originally published 1992).

V. C a sto n

Sorabji, R. (2004). Aristotle on memory. Second edition. London: Bristol Classical Press. (First

edition, 1972.)

Spier, J. (1990). Emblems in Archaic Greece. Bulletin of the Institute of Classical Studies, 37,

107–129.

Turnbull, R.G. (1978). The role of the ‘special sensibles’ in the perception theories of Plato and

Aristotle. In P.K. Machamer & R.G. Turnbull (Eds.), Studies in perception: Interrelations in

the history of philosophy and science (pp.3–26). Columbus: Ohio State University Press.

Vasiliou, I. (1996). Perception, knowledge, and the sceptic in Aristotle. Oxford Studies in Ancient

Philosophy, 14, 83–131.

Ward, J.K. (1988). Perception and λόγος in De anima ii 12. Ancient Philosophy, 8, 217–233.

West, M.L. (1992). Ancient Greek music. Oxford: Clarendon Press.

Williamson, T. (2013). Identity and discrimination (Reissued and Updated Edition). Oxford:

Wiley-Blackwell.

Younger, J.G. (1999). Glass seals and ‘Look-alike’ seals. In P.P. Betancourt, V.Karageorghis,

R.Lafneur, & W.-D.Niemeier (Eds.), Meletemata: Studies in Aegean archaeology presented

to Malcolm H.Wiener as he enters his 65th year (pp.953–955). Liège: Université de Liège. (=

Aegaeum, vol. 20.)

2 Aristotle ontheTransmission ofInformation: Receiving Form Without theMatter

Studies in the History of Philosophy of Mind

Volume 26

Series Editors

Prof. HenrikLagerlund,Department of Philosophy, Stockholm University,

Stockholm,Sweden

Prof. MikkoYrjönsuuri,University of Jyväskylä,Academy of Finland

and University of Jyväskylä,Jyväskylä,Finland

Editorial Board Members

Emeritus Prof. LilliAlanen,Uppsala University,Uppsala,Sweden

JoëlBiard,Professor of Philosophy, University of Tours,Paris,France

MichaelDellaRocca,Departement of Philosophy,Yale University,

New Haven,CT,USA

EyjolfurEmilsson,Professor of Philosophy,University of Oslo,Oslo,Norway

PatriciaKitcher,Professor of Philosophy, Columbia University,New York,USA

SimoKnuuttila,Emeritus Professor,Helsinki University,Helsinki,Finland

BéatriceM.Longuenesse,Emerita Professor,New York University,

New York,USA

CalvinNormore,Professor of Philosophy, University of California,

Los Angeles,CA,USA

David Bennett • Juhana Toivanen

Editors

Philosophical Problems in

Sense Perception: Testing the

Limits of Aristotelianism