Alexander of Aphrodisias ’ Criticism of the Stoic Theory of Perception: typos and typ ō sis

: The Stoics identified the phantasia with the impression ( typos ) in the soul, or the impressing process ( typ ō sis ). Alexander of Aphrodisias engages directly with this account at De anima 68.10 – 21, and argues against the applicability of the impression in a theory of perception in Mantissa 10, especially 133.25 – 134.23. I analyse Alexander ’ s polemic account at De anima 68.10 – 21, I demonstrate that it differs from Chrysippus ’ criticism of Cleanthes (contrary to some commentators), and I show how it ﬁ ts in the context of his argument. From this analysis it will emerge how Alexander uses Stoic ideas to form his Aristotelian account. Then, I show that Alexander, by taking ‘ typos ’ metaphorically, not only prefers the term ‘ enkataleimma ’ over ‘ typos ’ in his theory of phantasia , but he keeps the ‘ typos ’ terminology only to remain faithful to Aristotle ’ s use (contrary to some commentators).


Introduction
Phantasia is a fundamental psychological concept in antiquity with a rich and complex history. 1 In Aristotle it is probably an activity of the perceptual soul, a kind of link between sense-perception and thinking, making a diverse set of activities possible beyond perceiving (including remembering, dreaming, imagination, moving by desire, and experience) for humans and for non-rational animals. As such, phantasia is distinguished from perception and thought emphatically in its official treatment in DA III 3, yet its nature and status is left unspecified. 2 What is clear, though, is that phantasia is related to phantasma, a sort of impression; phantasma is plausibly an internal bodily process in the blood that is required for representation as the cause of phantasia. 3 In the Hellenistic schools phantasia is a central epistemological concept, closely related to the criterion of truth, but also important in explaining action and aesthetic imagination. Thus, to cover a wide range of mental states (including sense-perception and thought for the Stoics) phantasia might be called appearance. Alexander of Aphrodisias, defending an Aristotelian theory of phantasia, supplements the account of Aristotle's De anima with tenets from works on memory and dreaming (Mem., Insomn.), and criticises Stoic views. In the paper I focus on one aspect of this polemics, relating to 'impressions', which remains on quite a physical level, and so does not directly touch on issues of epistemology or the purported rationality of phantasia. It will turn out that Alexander's contention about impressions is quite original, yet embedded within his conceptual framework; its analysis will illuminate Alexander's use of Stoic notions, and his way of forming his terminology.
The Stoics identified the phantasia with the impression (typos) in the soul, or with the impressing process (typōsis). Alexander directly engages with this account at De anima 68.10-21, and argues against the applicability of typos in a theory of perception in Mantissa 10, especially 133.25-134.23. Alexander's critique is not unique in the history of ideas. Close in time to him two sources are most important. Sextus Empiricus (M VII 372-387) identifies difficulties internal to the Stoic account of phantasia as typōsis and offers sceptical arguments against any view appealing to impressions. Plotinus, one generation after Alexander, argues in Enn. IV 6 that there is no explanatory role for impressions in an account of perceptual cognition or memory, so impressions can be discarded. 4 The Stoic account of perceptual phantasia was not a uniform theory within notably Watson (1982); Wedin (1988); Caston (1996); Osborne (2000). The main problem, apart from the negative treatment in DA III 3, for identifying the status of phantasia is that Aristotle apparently does not take it to have a specific object required for being a power, as he does not use the term phantaston, instead appeals to phantasmata, cf. Wedin (1988) 59-63. So, interpreters take phantasia quite differently, as: 1) a distinct power, e.g., Hamlyn (1968); Everson (1997) 157-65;Bloch (2007)  ; but also Alexander of Aphrodisias; 2) same as the perceptual power, but different in account/activity, e.g., Modrak (1986Modrak ( , 1987; Frede (1992); 3) a further activity of perception, e.g., Johansen (2012) 199-220;Turnbull (1994); 4) a sub-faculty for representation, e.g., Wedin (1988); Osborne (2000); 5) representing in virtue of causal powers, Caston (1996) and(1998). 3 See, e.g., Caston (1996Caston ( , 1998; Nussbaum (1978); and Aristotle's account of dream in Section 2.2. 4 Porphyry's subtitle already indicates the thesis: τὰς αἰσθήσεις οὐ τυπώσεις (Plot. 5, 50), see also Enn. IV 3.26.29-33; IV 4.22.28-34. At Enn. III 6.1. [1][2][3][4][8][9][10][11][12]IV 4.23) Plotinus is more permissive; he allows that perception involves impressions in the body, but identifies perception as such with the judging activity (krisis) of the soul. This view is arguably the history of the school either. As I revisit it in Section 2.1, already the first few deans disagreed in the correct interpretation: Cleanthes taking typos quite literally as physical impression, whereas Chrysippus moving to a more nuanced account taking phantasia as heteroiōsis, a kind of qualitative, yet physical, change. In Sections 2.2 and 2.3, I analyse Alexander's polemics at De anima 68.10-21, demonstrate that it differs from Chrysippus' criticism of Cleanthes (contrary to some commentators), and, in Section 2.4, I show how it fits in the context of Alexander's explication of phantasia. From this analysis it will emerge how Alexander uses Stoic ideas to form his Aristotelian account. In Section 3, I show that Alexander, by taking 'typos' metaphorically, not only prefers the term 'enkataleimma' over 'typos' in his theory of phantasia, but virtually drops the 'typos' terminology, using it only to remain faithful to Aristotle's use (contrary to some commentators). Going through Alexander's use of the 'typos' terminology, in Section 3.1, I show that he applies it mainly in two ways: in polemics against Stoics and citing Aristotle's account of memory. Even though in some passages Alexander seems to appeal to the terminology in his own voice, I show that we need not take this as an indication of admitting the 'typos' terminology in a strong sense.

Alexander against Stoic phantasia as Impression
Alexander argues on Aristotelian basis that phantasia should be a distinct power of the soul for judging (kritikē) 5 (De an. 66.9-68.4; cf. Aristot. DA III 3.427b6-428b9), differing, most importantly for us, from perception, aisthēsis (De an. 66.24-67.9). The power for phantasia needs to have its own activity, Alexander argues, as we shall see, against the Stoic view. This activity might be identified as a kind of judgement (krisis), just like the activity of perception. In the compatible with changes in the soul, though changes different from bodily changes, see Noble (2016). Plotinus seems to repeat Alexander's account, especially of simultaneous perception, on which see below. At Enn. IV 6.1.15-20 Plotinus goes further by appealing to the phenomenology of seeing, viz. "directing (prosballomen) our sight straight to where the object of sight is situated." This shows -Plotinus saysthat if there were incoming impressions, 'directing our sight' would be superfluous, so either should be dropped: and we know Plotinus' preference (cf. Enn. IV 6.1.32-40; for further arguments from phenomenology, see Enn. IV 6.1.20-29). On Plotinus on perception and impressions, see Emilsson (1988) 63-93; on memory and impression, see King (2010).
5 For an analysis of krisis and rendering it as judgement with propositional content, see Hangai (2020) esp. 112-5. Even though in the perceptual case 'discrimination' might be a proper rendering of krisis, it does not fit well with other cases like thinking, as Emilsson (1988) 121-5 points out.
judging activity a residueenkataleimma, a kind of impression generated in perceiving something, corresponding to Aristotle's phantasmais used; as the cause it triggers the phantasia-activity and determines the content for phantasia; though it is not an intentional object of phantasia. In some cases the residue is modified: internal mechanisms related to phantasia complete it when it is incomplete. 6 E.g., I saw my friend's green bicycle a few times; this left a residue in my central sense-organ; which is probably distorted, and some mechanism amended it; and so this modified residue triggers me to recall the bicycle (in a phantasia-related act) as red.
In turning to his positive account of what phantasia is, Alexander cites Aristotle's account of memory  where Aristotle compares the memory phantasma to a sort of impression (typos tis). 7 This leads to the aporia of 'presence in absence', viz. how it is possible to remember something absent (being past) by having something present (Mem. 1.450b11-18).
For clearly one must think about that which is so generated through perception in the soul, that is, in the part of the body which contains [the soul], as a sort of picture (hoion zōgraphēma ti), and the state of having this we call 'memory'; for the movement produced marks in 8 a sort of impression, as it were, (hoion typon tina) of the sense-impression (aisthēmatos), similar to what is done by people using their seals. 9 (Aristot. Mem. 1.450a27-32) Alexander writes: We must conceive [phantasia] as something becoming in us from the activities concerning perceptible objects as a sort of impression, as it were, (hoion typon tina) and a picture (anazōgraphēma), in the primary sense-organ […], being a sort of residue (enkataleimma ti) of the movement generated by the perceptible object, which remains and is preserved even when the perceptible object is no longer present, being like a sort of image of it (eikōn tis autou), which, by being preserved, is also the cause of memory in us. 10 (De an. 68.4-10) 6 Cf. De an. 69.5-70.22. On a more complete account of Alexander's view on phantasia, and especially these aspects, see Hangai (forthcoming). 7 The idea goes back at least to Plato's Theaetetusas well as the terminology, see Sorabji (1972) 5 n. 1;cf. Long (2002) 120-2; Ioppolo (1990) 438-40;and Togni (2013) for showing the dependence of the Stoic account on Platoand is used by Aristotle to describe sense perception too (DA II 12. 424a19-20). 8 Sorabji's (1972) rendering of ἐνσημαίνεται, instead of Bloch's (2007) 'stamped'. 9 Translation from Bloch (2007), adapted in terminology. δῆλον γὰρ ὅτι δεῖ νοῆσαι τοιοῦτον τὸ γιγνόμενον διὰ τῆς αἰσθήσεως ἐν τῇ ψυχῇ καὶ τῷ μορίῳ τοῦ σώματος τῷ ἔχοντι αὐτήν -οἷον ζωγράφημά τι [τὸ πάθος] οὗ φαμεν τὴν ἕξιν μνήμην εἶναι· ἡ γὰρ γιγνομένη κίνησις ἐνσημαίνεται οἷον τύπον τινὰ τοῦ αἰσθήματος, καθάπερ οἱ σφραγιζόμενοι τοῖς δακτυλίοις. 10 δεῖ νοεῖν γίνεσθαι ἐν ἡμῖν ἀπὸ τῶν ἐνεργειῶν τῶν περὶ τὰ αἰσθητὰ οἷον τύπον τινὰ καὶ ἀναζωγράφημα ἐν τῷ πρώτῳ αἰσθητηρίῳ […], ἐγκατάλειμμά τι ὂν τῆς ὑπὸ τοῦ αἰσθητοῦ γινομένης κινήσεως, ὃ καὶ μηκέτι τοῦ αἰσθητοῦ παρόντος ὑπομένει τε καὶ σώζεται, ὂν ὥσπερ εἰκών τις αὐτοῦ, As Alexander cites Aristotle almost verbatim, it is illuminating to point out the differences. Alexander employs Aristotle's similes with impression (typos) and picture (anazōgraphēma, 11 adding the prefix '-ana'), and connects memory to image (eikōn). 12 The divergence is the description of the impression as a "residue of the movement generated by the perceptible object", where Aristotle has "a sort of impression, as it were, of the sense-impression". That is, Alexander (i) replaces aisthēma with "movement generated by the perceptible object" (τῆς ὑπὸ τοῦ αἰσθητοῦ γινομένης κινήσεως); and (ii) clarifies 'impression' as 'residue' (enkataleimma) that captures the important features of remaining (hypomenei) and being preserved (sōzetai), which phantasia requires to represent external objects in their absence. This, however, does not mean that Alexander uses the terms 'typos' and 'enkataleimma' interchangeably, 'typos' emphasising the external cause, 'enkataleimma' the internal aspect. 13 Rather, as I argue, while probably 'typos' is a first approximation for the object of phantasia that captures the receptivity of this power, Alexander suggests already here that the proper term to capture the object is 'enkataleimma' 14 (while 'anazōgraphēma' refers to a specific type of object, e.g., of dreams). To motivate his preference for enkataleimma over typos (and from this passage it seems that Aristotle's phantasma is a non-starter, see Section 3.1), Alexander turns to arguing against accounts appealing to typos just after this passage: They identify phantasia with this sort of residue (enkataleimma) and, as it were, with this sort of impression (typon). This is the reason why they define phantasia as imprinting (typōsin) in the soul and imprinting (typōsin) in the ruling faculty (hēgemonikōi). (A1) However, phantasia might be not the impression (typos) itself, but rather the activity of the power of phantasia concerned with this impression (typon). For, if phantasia were the impression (typos) itself, we would be in [the state of] phantasia only having it without being active concerning it; and at the same time we would be in as many [states of] phantasiai as many things there were of which we have preserved an impression (typon). (A2) Again, they call phantasia either (i) the ongoing impressing (ginomenēn typōsin); 15 or (ii) the one that has already been completed (gegonuian) and exists. But if (i) the ongoing, they would identify phantasia in activity as perception, for perception is the coming to be of the impression (typou). But there are phantasiai also in separation from perceptual ὃ καὶ τῆς μνήμης ἡμῖν σωζόμενον αἴτιον γίνεται. All translations from Alexander's De anima are mine. I owe useful comments about the translations to István Bodnár and Péter Lautner. 11 On anazōgraphēma and anazōgraphēsis (picturing) in Alexander, see De an. 69.25 and 70.18,and Hangai (forthcoming). 12 See infra, n. 62. 13 As Modrak (1993) 185 interprets the passage. 14 See below, and Section 3. 15 Literally: coming to be, being generated. activities. And if (ii) the completed and preserved, they would identify phantasia as memory. 16   The main point for Alexander in the passage is to show that phantasia must have an activity specific to it. He argues that phantasia can be neither the impression itself (typos), nor the imprinting process (typōsis) in which the impression comes into being. If it were the impression (A1), then one would be in a state of phantasia even without being active concerning that impression, only having it (68.14-15). Moreover, one would be in as many states of phantasia at once as many impressions one stores (68.15-16). Alternatively, (A2), if phantasia was the impressing activity, it would be either (i) an ongoing activity (imperfect tense) or (ii) one that has been completed (perfect tense). Alexander admits neither, for the first defines perception, the second memory (68.16-21). I discuss the argument in two turns, the first part (A1) in Section 2.2, while the second part (A2) in Section 2.3; before that I revisit the target of the critique in Section 2.1.
Alexander's version of the Aristotelian view. Before analysing the argument, it is worth summarizing the Stoic views that are attacked.
According to Zeno, phantasia is typōsis in the soul (SE M VII 236 = SVF I 58; DL VII 45 = SVF II 53; DL VII 50 = LS 39A3; Cic. Luc. 77 = LS 40D3-4) in the literal sense of impressing physically, just like a signet ring impresses its shape into a piece of wax. 19 Presumably, the same idea is expressed by Cleanthes in claiming that phantasia is impression involving depth and protrusion, kata eisokhēn te kai exochēn (SE M VII 228, 372-373, VIII 400; PH II 70 = SVF I 484). Alexander too takes these terms as the basic meaning of impression . 20 This literalist theory is rejected by Chrysippus (SE M VII 229, 373 = part of SVF II 56) on at least two grounds. (1) If the hēgemonikon had phantasiai of two objects with contradictory attributes at the same time, the same body (the hēgemonikon 21 ) would have two contradictory attributes simultaneously (two different shapes, as the literal account of typos requires), which is impossible (SE M VII 229). Further, (2) there could be but one impression at a time (SE M VII 373; cf. DL VII 50 = LS 39A3). For the literal impression of one shape would obscure the literal impression of another, if they are impressed in the very same bodily part. But since impressions are taken to be into one body, the hēgemonikon, this consequence indeed follows (Argument from Obscuring of Affections). Moreover, as a corollary, accumulation of phantasiai would be impossible on Cleanthes's theory. And since memory, according to the Stoics, presupposes many similar phantasiai (SE M VII 373; cf. Aëtius IV 11, 2 = LS 39A2), it would be impossible too. Thus, instead of the literal account, Chrysippus proposes an alternative definition according to which phantasia is alteration (heteroiōsis: SE M VII 230-231, 372 = part of SVF II 56; SE M VII 376; alloiōsis: DL VII 50 = LS 39A3; SE M VII 400) of the pneuma constituting the hēgemonikon. This 19 As the signet ring analogy is exploited in other important waysindicating the richness of the content represented by the impression (see, e.g., SE M VII 250-251 = LS 40E6)perhaps Chrysippus too keeps it, though without implying that it is a model for the type of change involved, cf. DL VII 50 = LS 39A3. However, Løkke (2008) 41 argues that it is the 'noisy room' analogy where many speak simultaneously (SE M VII 230-231) that provides Chrysippus with a model for rich sensory content. 20 On Alexander's understanding of typos literally, and his critique of such an account for perception, see Section 3. 21 The hēgemonikon (the soul itself) for the Stoics is bodily, just like everything that exists (cf. LS chs. 44-45). It is pneuma in the heart in a certain state. For sources, see LS 45C, chs. 47 and 53, 65H; cf. 484,518;II 460,637,[714][715]774,787,[836][837][838][839] Soul-pneuma has a high level of complexity that can explain mental functions of animals, and even more complex pneuma explains rationality (logos) of humans. Cf. Annas (1992) 20-6, 37-70;Inwood (1985) 18-41; Long (1982); Todd (1976) 34-49;Hahm (1994).
is not a literal impressing, nevertheless it must be a physical affection, 22 for there is no room within Stoicism for non-physical change. Thus, Chrysippus' alteration is intended to be such as to allow for (1) simultaneous alterations by contradictory attributes and (2) consecutive alterations by multiple attributes so that the alterations are retained.
Notice that both of Chrysippus' arguments presuppose that there are not several physical parts of the hēgemonikon into which the (simultaneous) impressions of different or even contradictory qualities and shapes could be distributed to avoid the absurd consequences. That is, the Stoic theory apparently presupposes the Indivisibility of the Central Sense-Organ: impressions modify the central organ of perception as a whole rather than parts of it. 23 This is remarkable, as Alexander seriously considers in his investigation of simultaneous perception a problem analogous to (1) Quaest. III 9,). Moreover, in course of his discussion of simultaneous perception, Alexander himself poses the same problem for the Stoic account (In sens. 167.4-9; cf. Mantissa 4, 118.6-9). Accordingly, the hēgemonikon could not be in opposite states at the same time, only successively, so the Stoics cannot explain the possibility of simultaneous perception. 25 Thus, it seems that despite his objection to the literal account of impressing, Chrysippus' own solution is liable to the analogous Problem of Opposites (at least 22 This is the feature of the change that allows Sextus Empiricus (M VII 383-387) to attack the theory. Phantasia, being a physical change, is the effect of its object, so it must be different from the object, hence cannot have the similarity required for accurate representation. That is, according to Sextus' criticism, Chrysippus' theory introduces a veil of perception (see also Plotinus Enn. IV 6.1. 29-32). Alexander's argument at De an. 68.10-21 does not depend on the materiality of the change. Yet, Alexander too exploits this feature of impressions elsewhere considering the possibility of simultaneous perception of several perceptible objects, as indicated below. Plotinus reverses the argument at Enn. IV 7.6.44-49: if memory is possible, it cannot involve impressionas impression would be into body without allowing memoryso the soul cannot be corporeal. 23 Cf. SE PH III 188 = SVF II 96. 24 For a recent analysis of Alexander on simultaneous perception, see Hangai (2020). 25 Alexander's note may pick out the fact that the motion of the pneuma is tensional, that is supposed to be simultaneously inward and outward (cf. LS ch. 47). Alexander takes this to be successive phases, as Towey (2000) 187 n. 505 suggests. However, the argument may be construed without this reference. The crucial premise, then, is that it is impossible to do different things simultaneously with the same power (cf. Alex. Mantissa 4, 118.29-35).
according to Alexander). The problem for Chrysippus follows from two premises: (i) the Indivisibility of the Central Sense-Organ; and that (ii) the conception of alteration (the kind of modification Chrysippus proposes) does not allow the same thing to be modified in different ways simultaneously. Of these, Chrysippus clearly adopts (i). Perhaps his reason is to ensure the unity and simplicity of the soul and its ruling part. 26 Nevertheless, he wants to deny (ii), and allow that several of his alterations (even contradictory ones) may coexist in one and the same subject. Whether or not he succeeds does not concern us here. 27 Alexander probably believed that the Problem of Opposites is applicable to Chrysippus' doctrine, because the alterations are nevertheless physical changes. 28

Alexander's Argument (A1)
Turning to Alexander's argument at De anima 68.10-21, let us start with (A1) against identifying the phantasia with the impression. Alexander derives two unacceptable consequences: (a) we would be in a state of phantasia without being active, simply by having the impression; (b) moreover, we would be in as many states of phantasiawith various contentas many impressions we have. Let us see the two consequences in turn.
First, (a) is prima facie question begging, as it apparently supposes Alexander's own view that phantasia should be an activity of ours, different from perception. But, let us see if a better argument can be reconstructed. The consequence will follow if having an impression φ is sufficient for being in phantasia-state φ. Alexander does not specify what having an impression amounts to, but we can extract the sense from his initial account presented immediately before the argument (De an. 68.5-7, cited in Section 2). He claims that the impressions (or residues) are seated in the primary sense-organ, i.e. in the heart, as remnants of the perceptual motions that created them . This suggests that the residues remain in their place in the heart, without the possibility of moving elsewhere. So we might say: for s to have an impression φ is for impression φ to be present in the central organ of s. So, replacing 'having' with 'being present' in our 26 See, e.g., SE M VII 234-236. Annas (1992) 115-20 andInwood (1985) 33-41 connect the requirement of unity to the Stoic theory of action that should not allow conflict with the hēgemonikon; yet, they also emphasise that the powers of the hēgemonikon are distinct. 27 It is noteworthy that apparently Sextus also takes Chrysippus' second worry (Argument from Obscuring of Affections) to apply for Chrysippus' theory using alteration, cf. SE M VII 377. For an argument that Chrysippus can avoid the same objections against his own theory by appealing to the tensional motion of pneuma, see Ierodiakonou (2007)  initial formula we get: if an impression φ is present in the central organ of s, s is in phantasia-state φ (Sufficiency of Impression's Presence). 29 Thus we can reduce the vague 'having' of an impression to the quite straightforward 'presence' in a local sense.
Now, it seems that the Stoics accepted that impressions are present in the central organ. For they believed that impressions are in the soul, or in the hēgemonikon, or more properly impressions (i.e., phantasiai) are the hēgemonikon pōs echon, the ruling faculty in a certain state, which indeed applies to every mental state. 30 Now, since the hēgemonikon is the subject of the occurrences of phantasiai, phantasiai cannot be seated but at the very same place as the hēgemonikon. Since the hēgemonikon is the pneuma in the heart, phantasia-impressions are also in the heart. In this theory it makes no sense even to say that phantasia is elsewhere (either on the literal interpretation or Chrysippus' alteration). Again, if one identifies phantasia with impression (as the Stoics), then the presence of the impression will constitute the occurrence of phantasia with a particular content. Further explanation how a phantasia may occur is not needed. So, the Sufficiency of Impression's Presence is also accepted by the Stoics.
Alexander's problem with the Sufficiency of Impression's Presence is that it leaves the subject without an activity concerning the impression, it requires only the presence of a physical item. 31 That is, the Stoic account implies that for a psychological state (phantasia) to occur it is sufficient that a bodily item is located in a certain area in the body of the living being. For Alexander this does not constitute an explanation of the psychic phenomenon.
The other consequence (b) (we are in as many phantasia-states as many impressions we preserve) requires a further premise. This is because a preserved impression does not necessarily mean an impression that we have, viz. that is present in our central organ. We may preserve impressions somewhere elsee.g., in another organ, or in the vascular system as we shall see for Aristotle on dreamor being preserved in the central organ may constitute less than being present in it 29 The principle, in this context, should be understood generally to apply to Stoics, Aristotle and Alexander. Thus, 'impression' and 'central-organ' should be taken to cover any item which a theory posits to account for the relevant phenomena: for 'impression' besides the Stoic notion Alexander's residues or Aristotle's phantasmata; for 'central-organ' the Stoic hēgemonikon as well as Alexander's and Aristotle's primary sense-organ. 30 It is emphasised that the impression is in the ruling faculty (qua soul) and not elsewhere in the body, e.g., at SE M VII 232-236. Again, any power or occurrent mental state is the hēgemonikon pōs echon, 'other parts of the soul' are only instruments. Cf. LS 53L-M; SVF I 141; II 57, 806, 858; Inwood (1985) 36-7; Annas (1992) 71-102; Menn (1999); Brunschwig (2003). 31 Cf. SE M VII 237, 239, where Stoic phantasia is characterised as a passive reception rather than an activity.
(e.g., it may be present only potentially, to use Aristotelian terminology). To rule out such possibilities, it must be supposed that every impression preserved by s is always present in the central-organ of s (Constant Presence of Impressions).
As it is clear from the above description of the Stoic view, the Stoics adopted the Constant Presence of Impressions. For impressions may only be located in the hēgemonikon, so if an impression is preserved, it is in the hēgemonikon. Again, it is plausible to understand the Stoic view so as the impression is present in actuality (and not on a lower level like potentiality), since arguably phantasia (thus the impression) in itself involves awareness. 32 It is clear that this argument of Alexander's is not the same as that of Chrysippus' argument (1) in Section 2.1, against the literal interpretation of impressing, as it was identified as analogous to Alexander's Problem of Opposites concerning simultaneous perception.
Thus, Alexander's argument works against those who accept both the Sufficiency of Impression's Presence and the Constant Presence of Impressions, as the two premises together entail the unacceptable consequences. Alexander (plausibly) takes the Stoics, including Chrysippus, to accept both premises, 33 and continues with the assumption that the Sufficiency of Impression's Presence should be dropped and the Constant Presence of Impressions may be accepted. Indeed, as noted, Alexander has reasons to accept the Constant Presence of Impressions (cf. De an. 68.4-9, 97.11-14). 34 Before we move on to Alexander's considerations following the denial of the Sufficiency of Impression's Presence, it is instructive to see what it would look like to drop the Constant Presence of Impressions, so let us see Aristotle's account of dreaming.
First, Aristotle apparently denies the Constant Presence of Impressions, as his phantasmata (the items analogous to impressions) are not always in the heart or primary sense-organ (where they can appear), but they can be in potentiality somewhere in the vascular system or in the peripheral sense-organs (cf. Insomn. 32 Phantasia reveals itself and its cause, see, e.g., Aëtius IV 12, 1-5 = LS 39B. 33 Pace Lautner (1995) 35-6, who takes it to apply to Zeno and Cleanthes only. 34 Another reason could be forged based on the explanandum phantasia is introduced to explain: presence in absence. That is, many mental states occur in the absence of the intentional object, and so a representation is required to make this possible. In the generation of content for these mental states sometimes we play an active role, e.g., in imagining something, or thinking about something. In other cases, the content comes about without our efforts. In the former cases we need to have access to the representations that we 'add into' the content of our mental state (e.g., what we imagine). In the latter cases, the mental faculty responsible for generating the content needs to have access to the representations it uses as sources for the generation of content. Now, since representations are physical items in the central organ (for Alexander: residues; for the Stoics: impressions in pneuma; for Aristotle: phantasmata) that cause the relevant mental states, this access to the representations and contents is best explained by the physical (or local) presence of the items where the soul powers are located: in the central organ.
3.461b11-21), 35 and they are taken (down) into the heart by the movement of the blood in sleep (Insomn. 3.460b32-461a8). 36 Aristotle also denies the Sufficiency of Impression's Presence. Having arrived into the primary sense-organ, phantasmata do not automatically appear, but further physiological conditions are necessary to be met (ibid.): (A) absence of larger movements, (B) absence of disturbance. 37 That is, the presence of phantasmata is not a sufficient condition for appearing (for the dream-phantasia). Nonetheless, as Aristotle does not mention other factors, the physiological conditions taken together seem to constitute a sufficient condition. Hence, no further condition must be met for phantasmata to appear, in particular, no need for a specific activity of phantasia. 38 As we can see, Alexander requires precisely this: an activity of phantasia in the explanation why an impression appears in a mental state involving phantasia. In this regard, Alexander's explanation seems to be more psychological in nature than Aristotle's physiological account of dream appearance.
35 Even though Alexander also mentions residues in the peripheral organs, in sight, ἐν τῇ ὄψει (De an. 62.22-63.4), this might refer to after images rather than pick out the technical meaning of the term 'residue'. 36 Even though not explicit, the account is clearly about phantasmata. First, a dream is identified as phantasmata appearing in sleep (Insomn. 1.459a18-20; 3.462a29-31). Second, the passage is about the movements that come about from aisthēmata (Insomn. 3.460b28-29), which cannot be but phantasmata. Cf. Wedin (1988) 34-9. 37 (B) The absence of disturbance obviously refers to the absence of intensive motions caused by getting to sleep (Insomn. 3.461a5-6; cf. Somn. 3). If the heart was in such a disturbed state, nothing could appear what is in it, just as nothing is reflected in a rapidly rushing water-current; or what appears would be much distorted (Insomn. 3.461a8-25). One might argue, however, that (A) the absence of larger movements is needed for the phantasma to get to the heart in the first place. That is, the larger movements are impediment for smaller ones in arriving at the heart, for all movements compete with each other, and only the larger may win. This is certainly one plausible option. But, considering Aristotle's example and wording, it gains support that this condition applies also when the larger and smaller movements are in their proper place. Aristotle mentions smaller and larger fires next to each other, and also pleasures and pains (Insomn. 3.461a1-3). The important point is that the smaller fire may not be perceived even if it is present together with the larger (viz. they are next to each other). The reason is that the larger movement displaces (ekkryei) the smaller, so that the smaller is effaced (aphanizontai) (460b32-461a1). The same terminology is used by Aristotle in the Argument from Mixed Perceptibles (Sens. 7.447a14-b6), posing difficulties for the possibility of simultaneous perception of two proper objects in one sense-modality (e.g., two colours). The reason is also the same (as in the case of the dream), and it is presupposed that the two movements are co-present. 38 It might be objected that in memory, though, an activity seems to be involved on behalf of phantasia: taking something as an image (eikōn), Mem. 1.450b20-27; see King (2009) 78-80. Even if this is accepted, it is needed not merely for the phantasia to appear, but for it to constitute memory (see further Section 3.1). So far we have seen that Alexander's argument against the Stoic account of phantasia as impression differs from Chrysippus' polemic against Cleanthes' literal account. Indeed, the latter is analogous to Alexander's Problem of Opposites, presented in the context of simultaneous perception, which Alexander also takes to apply to any Stoic view. Again, the analysis of Alexander's argument shows that it aims at motivating to drop the Sufficiency of Impression's Presence while keeping the Constant Presence of Impressions; so it is not only against any Stoic account of impression, but implicitly at odds with Aristotle's account of dream appearance.

Alexander's Argument (A2)
Having denied the Sufficiency of Impression's Presence, Alexander goes on to see (in A2 of De an. 68.10-21) what other factors (activities of the soul) the Stoics could add as conditions for phantasia-states to occur. He considers two candidates for the phantasia-activity -(i) the ongoing impressing, (ii) the completed impressingand accepts neither. Let us see the two candidates in turn, and then what Alexander proposes instead.
If the activity is (i) ongoing impressing, the generation or creation of an impression, it will be identical with perception. For, according to Alexander, the impression is the residue that comes about as an effect of perception; i.e. the process in which the residue comes to be is perception itself. 39 Actually, this is not far from the Stoic theory, according to which perception (aisthēsis) is an experience by means of a phantasia; 40 even though they did not restrict phantasia to the perceptual case. 41 Alexander cannot accept this, since he has already 39 Although perception is not defined with reference to this effect. See also Section 3.1. 40 Aisthēsis for the Stoics can mean the perceptual apparatus (pneuma constituting the sense organs and routes from them to the ruling part), the capability of this apparatus to support perceptual experience (the reception of impressions, typōsis), the activity or exercise of it, as well as a reaction to phantasia as assent to aisthētikē phantasia (DL VII 52 = LS 40P-Q; Aëtius IV 8, 1 = SVF II 850; Cic. Acad. I 40 = LS 40B1; for assent see also SVF II 72-75). See Rubarth (2004) for these meanings and further references. Alexander's identification is closest to the "exercise of the capability for receiving phantasia" sense. Notwithstanding there are cases when one gives assent to a perceptual phantasia, i.e. one endorses the content presented to one by that phantasia. Indeed, we usually trust our senses in our ordinary actions. However, the Stoics are keen on distinguishing a stage when one is experiencing something without giving assent to it. Thus, perception may be called assent considering the ordinary cases, but we must be aware that strictly speaking perception is experiencing a phantasia. See e.g., Annas (1992) 75-8. 41 Apart from the perceptual ones, humans can have phantasiai about things that may be grasped only by reason (e.g., about incorporeals or universal concepts, DL VII 50 = LS 39A4). These distinguished phantasia from perception (De an. 66.24-67.9), and phantasia for him is not a process or activity that creates residues, 42 rather, one that uses them. Moreover, he wants to explain a wide range of mental phenomena by phantasia, which could not be done if phantasia were identical to perception, for perception is restricted to cases when its object is present.
Alternatively, if the activity was (ii) a completed activity (indicated by perfect tense: gegonuian), it would define memory. Again, memory is only one phenomenon that phantasia is to explain, and the remainder cannot be explained by memory, as it is restricted to the past, with experiences that have been perceived. Yet, even though we possess but some notes from Alexander's account of memory (for references, see Section 3.1), from those it is clear that a completed residue (or impression) is insufficient for memory. It is also required that the residue is an image (eikōn) of the perception of the past event. The present remark (De an. 68.20-21), however, seems to pick out the Stoic conception of memory, according to which it is the storing of phantasiai (SE VII 373 = SVF I 64; II 56). 43 The completed impressions constitute a set of impressions that remain still and supposedly available to the agent. So, if the remark is taken as specifically against the Stoic account, it suffices to be said that they themselves gave a wider role to phantasia than to memory, and so a definition of phantasia that picks out memory is inadequate.
Alexander concludes from the argument that phantasia must be a distinct activity; not (i) the creation of an impression (= perception), nor (ii) the completed impressing (= memory), yet necessarily related to the impression. He does not specify the relationship and the activity further, only implies its analogy to perception: the object of phantasia is a sort of perceptible (tina/hōsper aisthēta, cf. De an. 69.1, 10). It is not my aim here to analyse the activity of phantasia as understood by Alexander; 44 instead, let us see how Alexander proceeds from his polemic against the Stoic account, to better assess its import.

Alexander's Argument in Context
Alexander supposes the existence of an activity by invoking Aristotle's tripartite scheme of power/activity/object (De an. 68.21-30) which he uses as a framework phantasiai are not caused externally only, their formation depends on the creativity of the mind. The status of universals is complicated, for universals are supposed to be phantasmata, figments of the mind, without any real object corresponding to them (Stob. I 136, 21-137, 6 = LS 30A, DL VII 60 = LS 30C2). 42 Contrast Aristotle DA III 3.428a1-2: phantasia is that in virtue of which phantasmata arise for us. 43 On the Stoic account of memory, and further passages, see Ierodiakonou (2007). 44 For a summary, see Section 2.
for his psychological investigation (cf. De an. 32. 23-33.9, 40.15-19). 45 But once he has applied the scheme for phantasia, he needs to identify the object postulated for phantasia. He accomplishes this by saying it is an internal perceptible object (De an. 68.31-69.2). Alexander takes the object to be internal, as phantasia is supposed to occur when the corresponding perceptible object is not present, so that perception properly speaking is impossible. Since there must be an object to cause the state, the object has to be internal, present within the central organ. Thus, Alexander posits phantasta (objects of phantasia) to solve the problem of presence in absence or representation of absent objects. 46 Once the object of phantasia is identified and distinguished from the object of perception, it follows that the power of phantasia also differs from the power of perception, and so constitutes a distinct power of the soul. This move is noteworthy, as for Aristotle the relationship between phantasia and perception, and the status of phantasia is not that clearly specified. 47 Alexander concludes his argument at De an. 70.3-5 by opposing the Stoic view (τὸ λέγειν τύπωσιν ἐν ἡγεμονικῷ) that it fails to acknowledge the proper identity of phantasia as activity, for it would put phantasia in the residue (ἐν τῷ ἐγκαταλείμματι) instead. 48 Thus, it seems that Alexander first shows that (A1) there has to be an activity of phantasia related to an impression (A2) distinct from the generation of impression (perception) and the retaining of the impression to remember past events or perceptions. This he achieves through a quick polemic against a generic Stoic view, picking out only the essence of the theory; and so this passage (De an. 68.10-21) cannot be used as an independent source for Stoic theory (although part of it is printed as SVF II 59). The phantasia activity is, then, called for in cases when the external object about which the mental state has content is not present. In these cases, there must be an internal object (residue). Then, the power of phantasia is identified as the power of making such mental states as activities possible, so a power distinct from perception.

Residue and the Metaphorical Use of 'typos'
We have seen Alexander's problems with impression in phantasia, his reinterpretation of it as residue. To better understand his preference for residue, we  46 On Alexander's account of mental representation, see Hangai (forthcoming). 47 See supra, n. 2. 48 Alexander's final say against the Stoic view is provided in his account of truth and falsity of phantasia ) that can be taken as a criticism of katalēptikē phantasia. In this, arguably, Alexander reuses Carneadean arguments against the Stoics (cf. SE M VII 402-407), turning them to support his Aristotelian view, see Hangai (2017). should also consider his troubles with impression in perception, its inaccuracy to account for the wide range of perceptible objects, and its incompatibility with the physics of perception. Still, as it turns out from Section 3.1, Alexander is able to use the 'typos' terminology where Aristotle uses it, and where it only needs to indicate a physical change with an obvious cause. In his more robust account of representation, he uses 'residue' in a quite literal sense to emphasize remaining and preservation.
Alexander argues against a literal interpretation of impression in Mantissa 10, 133.25-134.27, in the context of his polemics with the Stoic account of sight 49 which implies that impression is sufficient for perception (cf. Mantissa 10,. 50 In the literal sense an impressionas illustrated with a signet ring pressed into melted wax that later is solidified (cf. De an. 72.7)is a persisting pattern in the surface of a quite solid receptor that actually has a shape corresponding in negative to the shape of the object producing the impression. Alexander picks up these features of impression in turn, and shows the inadequacy of the concept to explain sight as the Stoics did. First, the medium of impression is most apt if it is solid, in contrast to air (Mantissa 10, 133.25-28) which is fluid and can only receive confused impressions if any (134.9-10; cf. 133.31-38). 51 Again, impression intrudes into the receptor only superficially, even in apt materials, and by no means throughout the receptor (133.38-134.6). An impression is also something persistent even in the absence of its impressor, in contrast to perception that requires the presence of its object (134.6-7). Again, impression is a negative of the shape, a convex object creating a concave impression. And it is inadequate to claim that convexity is judged by concavity, for there are exceptions: some paintings are actually flat, though produce appearances of convexity . And most importantly, impression can represent only the shape of the object creating it .
49 Presumably Chrysippus', cf. DL VII 157 = LS 53N; Aëtius IV 15, 3 = SVF II 866; Sharples (2013) 99;Hahm (1978) 65-9;although Ierodiakonou (2019) questions that the opponents were early Stoics. On the Stoic theory of vision, see Løkke (2008) 37-9, andvon Staden (1978). 50 It is doubted that the author of the treatises of Mantissa is Alexander, though it is likely that it is someone from Alexander's circle. As the content of the treatises is for the most part consistent with what Alexander writes elsewhere, and often just the depth and the style of the discussions are reasons for the doubt, we can presume Alexander's authorship in this paper. Cf. Sharples (2013) 4-5. 51 SE M VII 374-375 cites the same worry for pneuma which is compared to other materials according to thickness and fluidity to show that it is unlikely that pneuma can receive or retain impressions. Also see SE PH III 188, Plutarch. Comm. not. 1084F-1085A = SVF II 847, cf. Ioppolo (1990) 434. Plotinus also uses this fact in arguing against the need for impressions, cf. Enn. IV 7.6. 37-44.
Thus, instead of taking it literally, Alexander claims that 'impression' can be used only metaphorically . 52 Since only a shape or figure can be impressed literally, in case of other perceptible features the residue (enkataleimma) may be called 'impression' only metaphorically. This suggests that Alexander, at the end, gives an explanatory role in his account of phantasia to the residue instead of the impression.

Occurrences of the 'typos' Terminology
This point is further justified by the fact that the terms of 'impression' rarely occur in Alexander outside the context of the arguments against Stoicism. Indeed, 'typōsis' never occurs in other contexts. 'Typos' is used elsewhere in its ordinary senses as 'mould' (e.g., In An. Pr. 6.16-18; In Metaph. 57.6 53 ), and mostly as 'telling something in outline' (e.g., Mantissa 25,186.11;De an. 60.3;In Metaph. 463.17;464.1;579.25,or several times in In Top.). In three passages, however, Alexander apparently replaces Aristotle's term 'phantasma' by 'typos', and uses the typos terminology in his own voice. These are discussed in the following paragraphs.
The first occurrence, at De an. 83.4, comes approximately 10 pages after Alexander's caution about the metaphorical meaning of the term, which could explain its use in itself, but more can be said. The context here is concept formation, which starts with perception, goes through memory (involving phantasia) and experience. 54 Alexander claims that in each case of perception an impression comes to be (τυπούμενος) which is preserved in memory (De an. 83.2-10). 55 Alexander puts forward the same account of perceptionas generation of impressionearlier in the treatise, and also elsewhere. (1) As we have seen in Section 2.3, in his polemic against the Stoic view he says that taking phantasia as the generation of impressions would actually define perception . Again, (2) in discussing phantasia, Alexander identified perception in activity as "possessing in itself this impression that [came to be] from perceptible objects that are external."   56 Here, perception is contrasted with phantasia insofar as the external object is present in perception, but absent in phantasia, which relies on internal objects that remain as residues (enkataleimma) of perception (69.2-4). The passage occurs in the context of Alexander's argument for showing that phantasia always has its origin in perceptions and hence can be called 'perception in activity' (68.31-70.5). Thus, he aims at keeping continuity between residues and perception through the impressions that are possessed in perception and kept in phantasia as residues. By identifying 'perception in activity' as 'possession of an impression', it is applicable to phantasia and perception alike; whereas 'generation of impression' would only be applicable to perception and not to phantasia (cf. 68.16-20). Finally, (3) in Quaestiones III 7his commentary 57 on Aristotle DA III 2.425b12-426a2 on self-awareness of perception -Alexander, in appealing to Aristotelian physics according to which the activity of an agent and that of the corresponding patient is one and is present in the patient, writes that "the being of perception in actuality consists in possessing the form of the thing perceived without its matter" 58 (Quaest. III 7,. Since the perceptible form is received in perception from the object through the object acting upon the perceiver and so assimilating the perceiver to itself (in the relevant way), 59 the object's acting upon the perceiver can be identified as the object producing an impression in the perceiver. Thus, impression is related to one aspect of perception: passive receptivity. Again, as the object assimilates the perceiver to itself so that the perceiver receives its form, the object produces an impression in the perceiver about itself and its form. The content of the impression is identical to the perceptible form received. So, since perception can be identified as the possession as well as the reception of form, it can also be described as the possession just as the generation of impression. Thus, referring to the generation of impression (τυπούμενος) in perception seems to be related to (a) the passive receptivity of perception, and (b) the continuity between perception and phantasia which is sustained by the remaining of the affection as a residue of perception. However, these aspects are better characterized by other terms: (a) receptivity as 'assimilation' or 'reception of form'; and (b) remaining as 'residue'. So, after all, even though Alexander uses the 'typos' terminology at De an. 83.4, on the one hand 56 τὸ κατ' ἐνέργειαν αἰσθέσθαι ἐστὶ τὸν τύπον τοῦτον σχεῖν ἐν αὑτῷ ἀπὸ τῶν αἰσθητῶν ἐκτὸς ὄντων. 57 The same doubts can be raised about the authorship of Quaestiones that about Mantissa in n. 50 supra, cf. Sharples (2014) 1-5. 58 Sharples' (2014) translation adapted in terminology. Κατ' ἐνέργειαν αἰσθήσει τὸ εἶναί ἐστιν ἐν τῷ ἔχειν τὸ εἶδος τοῦ αἰσθητοῦ χωρὶς τῆς ὕλης. Quaest. III 7, its import is very thin, and indeed better expressed by Alexander's preferred alternatives, and on the other hand the passage occurs 10 pages after the explication that 'typos' is to be used metaphorically.
The second problematic occurrence of 'typos', at In Metaph. 4.12 Golitsis (= 3.17 Hayduck), 60 is concerned with memory, occurring in the context of distinguishing different kinds of intelligence . The last of the senses identified is "the natural versatility in regard to the performance of actions that is found in animals capable of remembering" (In Metaph. 4.9-10 G = 3.13-15 H, translation by Dooley). This calls for commenting on memory, so Alexander cites Aristotle's definition: "memory is having a phantasma which is like an image of that about which the phantasia is" . 61 In Alexander's explanation "the impression according to the phantasia is not sufficient for memory, but the activity concerning the impression must also be concerned as with an image, 62 that is, it must be as from something else that has happened" (In Metaph. 4.11-13 G = 3.16-18 H). 63 What is relevant from this now is that Alexander uses 'impression' (typos) instead of 'phantasma'. 64 However, this should not be taken to imply that he adopted a 'typos' terminology in favour of the term 'enkataleimma'. Since Aristotle himself applies the term 'impression' in explaining memory (cited in Section 2) perhaps in the context of a commentary on Aristotle's account of memory it is appropriate to apply the same terms that one finds in Aristotle. Replacing 'phantasma' with 'typos', then, need not imply that 'typos' is the most appropriate term to use in this context, only that it is more appropriate than 'phantasma' due to the fact that Aristotle used it too, and presumably because at different times, as he has shown in On the Soul and On Memory and Sleep. 69  Here again, 'impression' is used for the effect of the activity of perception that comes about inside, which later can be used in different ways in different mental states. This effect is the residue of De anima, so why is it called 'impression' here? Since this passage is not connected to memory or to Aristotle's text, the previous explanation is not applicable here. But, in the context of truth and falsity in perception and in phantasia Alexander only wants to make a quick distinction between phantasia and perception, and perhaps this is appropriate without reference to residues (enkataleimmata). Moreover, he does refer to Aristotle's works (especially On Memory) in which the term 'typos' appears in relation to phantasia, which explains Alexander's use.
There is one particular compound with the term 'typos' that Alexander applies in his account at De an. 70.13 (occurring once in the whole Greek corpus): 70 prosanatypoun, 'impressing further'. The fact that the term is compounded from typoun renders it as not a simple application of the term 'impressing' (typoun). For this reason, I only mention that it is apparently used for a mechanism of phantasia to complete residues that are incompletely preserved to get an appropriate residue for phantasia to represent its content. This fits nicely with Alexander's account that the original generation of residues in perception can be called metaphorically as 'impression'. 71 Thus, Alexander, besides showing the inadequacy of the Stoic theory of phantasia appealing to impression insofar as they miss to posit a specific phantasia activity, replaces the terminology of 'impression' for 'residue', and uses 'impression' only metaphorically. Presumably he keeps the metaphor because Aristotle appealed to it too, but what remains from it is quite thin, amounting not much more than to being a physical change with an obvious cause.