Augustine on Active Perception, Awareness, and Representation


It is widely thought that Augustine thinks perception is, in some distinctive sense, an active process and that he takes conscious awareness to be constitutive of perception. I argue that conscious awareness is not straightforwardly constitutive of perception and that Augustine is best understood as an indirect realist. I then clarify Augustine’s views concerning the nature and role of diachronically unified conscious awareness and mental representation in perception, the nature of the soul’s intentio, and the precise sense(s) in which perception is an active process.


Introduction
Augustine is often thought to hold an important place in the history of philosophical thought about perception. Thus, for instance, several medieval proponents of 'active' theories of perception took Augustine to offer a valuable corrective to what they deemed to be the excessive passivity of Aristotelian accounts,1 and it is nowadays still often thought that Augustine took perception to be an active process in some substantive and historically distinctive sense.2 What the active nature of Augustine's account of perception amounts to is somewhat less clear, but it is often thought that Augustine is distinctive-and historically significant-in thinking that conscious awareness is constitutive of perception,3 and that bodies do not act upon souls in perception (instead souls act upon themselves).4 In this paper, I clarify several central features of Augustine's account of perception. I first (Section 2) examine how Augustine defines perception in some earlier works and outline two possible worries his definition faces concerning the objects of perception and the nature of awareness in perception. I then (Section 3) suggest that Augustine's mature account of perception-most notably in De Trinitate-offers an appropriate response to the two worries identified and clarify the nature and role of conscious awareness in perception. I argue that-contrary to what is sometimes thought-substantive conscious awareness is not constitutive of perception and that Augustine is best understood not as a direct realist (Matthews 2002;Silva 2014) or an anti-realist (O'Daly 1987;1999), but as a representationalist or indirect realist who takes subjects to perceive the external world by means of a series of inner representations which act as a window onto the outer world. Finally (Section 4), I clarify the nature of the various processes involved in perceptual experience and clear up some misunderstandings about the nature of the soul's intentio (a term which I leave untranslated for now), thereby offering a more precise account of the sense(s) in which Augustine's account of perception should be regarded as active.

Explicating Perception
In his early De Quantitate Animae and De Musica, Augustine characterises perception as a process wherein something is 'not hidden' (non latere) from the (Quaestiones in Secundum Librum Sententiarum Q58,Q72,Q73,; Robert Kilwardby (97)(98)(99)(100)(101)(102)(103)123 Lewry). For discussion of these philosophers, and others who were inspired by Augustine's active account of perception (such as William of Auvergne, John Peckham, Matthew of Aquasparta, and others), see Pasnau (1997, 125 ff.); Silva and Toivanen 2010;Toivanen 2013, 141-61;Adriaenssen 2017, 53-6. 2 E.g. Gannon 1956Miles 1979, 22 ff.;Caston 2001, 38 ff.;Silva and Toivanen 2010;Silva 2014;Kalderon 2017, 32 ff. 3 E.g. O'Daly 1987Caston 2001, 39;Silva 2014, 86-7. 4 E.g. Caston 2001Silva 2014, 79-80, 87. 6 That is to say, 'quod patitur corpus' could have the same meaning as: (i) 'what the body is affected by' (which, ordinarily, would suggest that the thing in question is an external object of perception); or (ii) 'what the body undergoes' (which would suggest that the relevant thing is an affection of the body). Several readers prefer (i) (for discussion, cf. Brittain 2002, 275-6). However, (ii) seems to be a better rendering for at least two main reasons. First, Augustine offers a protracted discussion of whether growth is a counterexample to the definition ) and this is difficult to make sense of on the (i) reading. Secondly, the revised definition of sense-perception offered at Quant. An. 25.48 (see below) is an improvement on the first and-like the (ii) reading-defines sense-perception as a passio corporis (which is consistent with (ii) but not with (i)). Given that no reasons have been offered for changing one's mind over what is not hidden from the soul, and that the dialectic requires that the definition be improved by revisions, it seems that the second definition makes explicit or unambiguous (rather than radically revises without comment) what is assumed in the original definition. 7 For instance: 'sensation is "a bodily experience of which the soul is not unaware" ' (McMahon 1947, 104); perception is 'the soul's awareness of that which the body experiences' (O'Daly 1987, 86); 'an experience of the body noticed by the soul' (Rist 1994, 93); 'perception is the soul being aware of what affects the body' (Brittain 2002, 275). However, one should not simply assume that Augustine takes perception to be necessarily conscious (as some explicitly do, e.g. Brower-Toland 2020, 42-3, 65, 70). Moreover, as has been noted by other scholars (e.g. Emilsson 1988, 8), it is worth keeping in mind that in discussing αἴσθησις or sensus, the ancients do not distinguish between sensation and perception in the way moderns typically do.
In what follows, Augustine promises to refute this definition. After some preliminary questions, Augustine turns to examine the kind of cognising (cognoscere) attained through the senses (Quant. An. 24.45). It is established that, in seeing some smoke, one may come to cognise that there is a fire by means of inference. The fire is thus 'not hidden' from the soul but the fire is not perceived (sentire). Augustine then goes on to further show how (SENSE-PERCEPTION) is inadequate by considering counter-examples such as growth and aging. These are processes undergone (pati) by the body and-since we are capable of cognising their existence through inference (coniectare, colligere, cognoscere coniectatione)-they are not hidden from the soul. However, contrary to what (SENSE-PERCEPTION) implies, such cognising is not sense-perception. We do not typically perceive or sense growth itself (24.46, 48), but instead merely the effects of growth (25.48;cf. Mus. 6.5.15). Thus, while we infer the existence of growth we do not perceive it and (SENSE-PERCEPTION) is an inadequate definition.
To improve the definition, Augustine suggests that, even if not every affection of the body not hidden from the soul is an instance of sense-perception, it is the case that every instance of sense-perception is an affection of the body not hidden from the soul (sensus est certe omnis passio corporis non latens animam, Quant. An. 25.48). Thus, a 'per se' specification must be added. This yields the following improved definition: (SENSE-PERCEPTION*) sense-perception (sensus) = def an affection of the body which is not in itself hidden from the soul (passio corporis per seipsam non latens animam, 25.48; cf. 30.58-9).
(SENSE-PERCEPTION*) aims to rule out counterexamples of the kind previously adduced in the discussion of growth and aging by insisting that, in perception, we are able to non-inferentially cognise something undergone by the body. The new definition is tested (25.49) and, after some apparent difficulties are resolved,8 Augustine and his interlocutor are ultimately happy to accept (SENSE-PERCEPTION*) as a satisfactory definition (30.58-9). In his De Musica (begun c. 387, but emended c. 408),9 Augustine assumes something very much like this account of sense-perception (although it is perhaps the 8 Augustine raises a difficulty by suggesting that when something is not hidden, it is known (cum autem non latet aliquid, utique scitur, Quant. An. 26.49). However, the main difficulties are resolved when Augustine points out that one should not think that everything which is not hidden is known (Quant. An. 30.58). 9 Cf. Epistulae 101.3-4; Retractationes 1.5.6; Jacobsson (2002, pp. x-xxviii).

Nawar
Phronesis 66 (2021) 84-110 soul's actions which are 'not hidden' from the soul, Mus. 6.5.9-11; cf. 6.5.15). However, whether it is an affection of the body or the soul's activity upon itself that is not hidden from the soul, (SENSE-PERCEPTION*)-and accounts like it-face at least two potential worries.
First, there is a potential worry about the relata of the relation invoked in (SENSE-PERCEPTION*). More concretely, if (SENSE-PERCEPTION*) simply invokes a dyadic relation between an affection of the body (passio corporis) and the soul, then one might worry that Augustine makes no explicit mention of real relations to any external items that exist independently of the perceiver (e.g. cats, hats, mats) and which are typically taken to be the objects of perception. Accordingly, one might worry-as certain medieval readers of Augustine did (cf. Olivi, Quaestiones in Secundum Librum Sententiarum Q58, p. 484)that Augustine's account of perception exhibits a certain kind of idealism or antirealism wherein perception may occur without any external object(s) and perception does not put us in touch with the world but merely with certain features or elements of ourselves.
Secondly, one might worry about the relation invoked in (SENSE-PERCEPTION*) and the fact that it is not clear why Augustine should characterise sense-perception in terms of x not being hidden from y. There are two sides to this worry. On the one hand, if 'not hidden' in (SENSE-PERCEPTION*) denotes or entails a relation of actual awareness (such that y is not hidden from x iff x is actually aware of y), then it is puzzling why Augustine should choose to articulate his account in the seemingly roundabout way he does. That is to say, if Augustine meant actual awareness, then why not simply say so?10 On the other hand, if 'not hidden' (non latere) denotes or entails a relation of merely potential awareness (such that y is not hidden from x iff x is potentially aware of y), then one might worry that sense-perception (sensus) turns out to be something more like (e.g.) sunburn instead of the complex psychological experience we ordinarily take to accompany and be caused by or partly constituted by such physiological processes. 10 Brittain 2002 suggests that Augustine's 'rather contorted formulation' is due to a desire to causally isolate the soul from the body (277-8). My own reading does not turn upon causal isolation but instead upon the thought that x not being hidden from y does not entail that y is consciously aware of x to any significant degree.

Representation and Conscious Awareness
I think that, properly understood, Augustine's account of perception-at least in his mature De Trinitate (the relevant books of which were seemingly written between 414 and 418)11-offers an appropriate response to the worries noted above and that appreciating how it does so yields a better understanding of some central but frequently misunderstood elements of Augustine's views concerning perception. An appropriate response to the first worry-about the relata of the relation invoked in (SENSE-PERCEPTION*)-requires appreciating the nature of Augustine's representationalism and why Augustine thinks perception requires a real relation to mind-independent objects even if such objects are not explicitly mentioned in (SENSE-PERCEPTION*). I will discuss this issue in more detail momentarily, but for now it suffices to note that Augustine's account of perception is best understood as a form of indirect realism according to which we perceive mind-independent external objects by standing in a certain relation to inner representations of such objects. An appropriate response to the second worry lies, I think, in appreciating that 'not hidden' denotes or entails a relation of merely potential substantive awareness and that, for Augustine, perception does not in itself require conscious awareness (or at least not substantive conscious awareness).12 This might seem counterintuitive and seemingly goes against influential readings of Augustine which take him to be distinctive in making conscious awarenessand not the mere informing of one's senses-constitutive of perception (e.g. Caston 2001, 39). However, while it is somewhat difficult to specify what conscious awareness amounts to in any precise way (presumably it has certain phenomenological characteristics, but beyond that much is unclear),13 it is 11 For the dating of the various books of De Trinitate, see Hombert 2000;Kany 2007;Ayres 2010, 118-120. For discussion of other works, such as De Genesi ad Litteram, see below. 12 Brittain 2002, 278 and, following him, Silva 2014, 93-6 briefly suggest that there might be degrees of awareness in Augustine and that perception merely requires low-level awareness. This suggestion finds precedent in Louis de la Forge (e.g. 'il paroit que St. Augustin a reconnu pour le moins deux degrez dans la sensation' , Traitté de l'Esprit de l'Homme pref.) and I take my account to be one possible way of developing this suggestion. 13 It is difficult to give a clear account of consciousness or awareness (some philosophical accounts equate the two and take the relevant terms to be interchangeable while others do not). In discussing Augustine's views, Brower-Toland suggests that 'to say that a given psychological act or state of awareness is conscious, then, is just to say that it is experienced by its subject; there is, as the familiar refrain goes, "something it is like" for the subject to have or be in that state ' (2020, 44), but I myself struggle to see how such remarks are informative. Alternatively, one might suppose that x's being suitably sensitive to the Nawar Phronesis 66 (2021) 84-110 important to appreciate that Augustine seems to think that a certain kind of conscious awareness-which we shall see may be glossed as diachronically unified conscious awareness-is not required for perception. There are at least two grounds for thinking that Augustine does not take conscious awareness to be constitutive of perception. First, in De Trinitate, Augustine claims (11.8.15): The will turns away memory from the sense when it is intent on something else and thereby does not allow present things to sink in (cum in aliud intenta non ei sinit inhaerere praesentia). This can be easily grasped because it often happens that someone is speaking to us in person but, due to thinking about something else, we seem not to have heard. However, that is false. We did hear, but we did not remember because the speaker's utterances slipped immediately through the sense of our ears, due to the alienation of the will's command, by which they [the speaker's utterances] are usually affixed to memory (audivimus enim, sed non meminimus, subinde per aurium sensum labentibus vocibus alienato nutu voluntatis, per quem solent infigi memoriae).14 As Augustine describes things here (and seemingly elsewhere, e.g. Mus. 6.8.21; De Genesi ad Litteram 7.20.26),15 there are instances of perception which do not manifest conscious awareness or which manifest only a low degree of conscious awareness. Thus, a person whose focus or will is directed elsewhere presence of α and acting accordingly suffices for x's being aware of α. However, this kind of 'awareness' seemingly occurs in so-called 'blindsight' (wherein agents are sensitive to the presence of α in action but sincerely do not report being aware of α; cf. Milner and Goodale 2006, 67-86) and falls short of so-called conscious awareness. For discussion of the difficulties in defining conscious awareness in this sort of context, see Dretske 2006, 153-9. 14 As noted by MacDonald 2012, 241 n. 6, 'remember' need not denote the calling to mind of something in the past; instead, here it seemingly denotes the calling to mind of something in the present.

15
'In hearing even the shortest syllable, we can say that we heard nothing unless memory helps us so that at that moment of time in which it is not the beginning but the end of a syllable that is sounding there remains in the soul that motion which was produced when the beginning of the syllable was sounding. That is why, generally, when we are occupied by another conscious thought we seem not to have heard those who are speaking in person' (in audienda itaque vel brevissima syllaba, nisi memoria nos adiuvet-ut eo momento temporis quo iam non initium, sed finis syllabae sonat, maneat ille motus in animo, qui factus est cum initium ipsum sonuit-nihil nos audisse possumus dicere. Hinc est illud quod plerumque alia cogitatione occupati, coram loquentes non nobis videmur audisse, Mus. 6.8.21). may nonetheless be said to hear and thus to perceive a sound even though the sound did not get an adequate foothold in their conscious awareness (cf. Plotinus, Enneads 4.4.8.1 ff.). In such instances, one has been alienated from one's senses (cf. Gn. ) and one cannot call to mind or remember what one has perceived. Nonetheless, one has received information from the relevant object through one's senses and one has perceived even if it does not seem to be the case to oneself that one has perceived. In such cases, one can seemingly act as if one were listening or manifest other kinds of sensitivity to the relevant stimuli but the perceiver's awareness of the relevant objects is significantly impoverished. It is not entirely clear to what degree the information received by the senses makes its way to the soul, but it is clear that, on Augustine's view, in such cases memoria does not fulfil its usual function (cf. Plotinus 4.4.8.1 ff.). More concretely, Augustine elsewhere emphasises that memoria is always employed in hearing a sequence of syllables as a unified word, seeing something in motion as something moving, or even simply seeing an object as a single object.16 What seems to occur in the instances of perception Augustine describes (e.g. in De Trinitate 11.8.15) is that memoria does not unite the relevant experiences into a diachronically unified experience and, as a result, the subject's experience of the object does not manifest substantive or diachronically unified conscious awareness.17 Insofar as Augustine had similar views about the role of memoria in his earlier works (e.g. Mus. 6.8.21), it seems plausible that Augustine articulates (SENSE-PERCEPTION*) as he does because he has such cases-wherein the object is not hidden from us but we do not enjoy substantive, conscious awareness of the object and are unable to call it to mind-in view and thinks they should nonetheless count as instances of perception.
There is also a second reason for thinking that Augustine does not take conscious awareness to be constitutive of perception. In discussing perceptual experience in De Trinitate 11, Augustine discusses two salient processes: a lower-level process (which Augustine calls 'perceiving' or its subspecies, e.g. 'seeing') and a higher-level process which Augustine calls 'cogitation' (cogitatio). However, conscious awareness seems to occur primarily at the level of cogitatio. Thus, Augustine characterises seeing (videre)-an activity he attributes to the outer man (Trin. 11.1.1) and an instance of perception-as being constituted by a polyadic relation involving three elements: (α) the thing which we see (res quam videmus, corpus quod videtur, 11.2.2); (β) vision (visio); and (γ) what holds one's eyesight on the thing being seen as long as it is being seen, i.e. the intentio of the soul (quod in ea re quae videtur, quamdiu videtur sensum detinet oculorum, id est animi intentio, 11.2.2). Augustine identifies (α) as the corporeal object perceived (e.g. 11.2.2) or the form ( forma, species) of the body which is seen (species corporis quae videtur, 11.2.5). Augustine identifies (β) as the form in the sense (eamdem formam esse visionem, 11.2.3; cf. 11.2.2-3, 5), or the informing of the sense (informatio sensus, 11.2.3, 5), i.e. a process which occurs in the sense ( fit in sensu, 11.2.2).18 This item serves as an intermediary between the corporeal and the incorporeal realms (11.2.3, 5, 5.9; cf. Plotinus 4.4.23.15-31) and is numerically distinct from (α), the form of the external object (11.2.3).19 It is a representation (imago) and likeness (similitudo) of the external mind-independent object (11.2.3-5) and it is impressed 'only from the body which is seen' (a solo imprimatur corpore quod videtur, 11.2.3, but cf. 11.5.9). Finally, Augustine identifies (γ) as the desire to see (videndi appetitus) (11.2.2) or 'the will of the soul' (voluntas animi, 11.2.5) which 'moves the sense to the sensible thing' (rei sensibili sensum admovet, 11.2.5).20 This is a purely psychological item (solius animae est, 11.2.5).
In addition to the seeing which takes place in the outer man, Augustine describes a further, higher-level process. Cogitation (cogitatio), i.e. conscious thought, takes place in the inner man, i.e. the rational soul (anima rationalis, cf. Trin. 11.3.6).21 This activity, which does require conscious awareness (e.g. De

18
The fact that Augustine calls (β) visio might incline one to think that (β), on its own, constitutes seeing or perception (Kalderon 2017, 23-4 seems to take this route). However, this would be a mistake. Seeing is not constituted of (β) alone, but of (α)-(γ) jointly (and visio is ambiguous between (β) and the process composed of (α)-(γ), i.e. seeing). 19 Thus, for instance, he says 'the form made from a ring is to be distinguished from that form which is in the ring' ( formam factam ex anulo quae distinguenda est ab ea forma quae in anulo est, Trin. 11.2.3). 20 Kalderon takes (γ) to be 'the attention of the mind' (2017, 23-4), i.e. 'conscious attention' (2017,24) and MacDonald 2012 takes (γ) to be a seemingly conscious 'act of will' (240). As a result of my account of the soul's intentio, I disagree with these readings. See Section 4 below. 21 In addition to the process (i.e. seeing) composed of (α)-(γ) and the process (i.e. cogitation) composed of (δ)-(ζ), Augustine claims there is a further process by which the form in the sense, (β), comes to be in memoria and (δ) comes about (Trin. 11.9.16; cf. Ep. 162.3-6). However, Augustine says almost nothing about it.
Immortalitate Animae 4.6; Trin. 11.3.6, 8.13; 14.6.8 15.9.16), is also constituted by a polyadic relation involving three elements: (δ) the form ( forma, species) in memoria; (ε) internal vision (interna visio), i.e. the form in the inner gaze of the soul (acies animi, cf. acies cogitantis, Trin. 11.7.11; contuitus cogitantis, 11.9.16); and (ζ) what binds together the form in memoria and the inner gaze, i.e. the will (voluntas) (11.3.6) or the intentio of the will (intentio voluntatis, 11.4.7). As Augustine describes things, when the soul's inner gaze-i.e. something like an inner theatre by which we consciously imagine or contemplate absent corporeal things as well as consciously see present corporeal things (11.3.6)settles upon (δ), i.e. a form in memoria, it is informed by it. Element (ε) is thereby produced and cogitation (cogitatio) takes place. The form impressed upon inner vision, i.e. (ε), lasts only so long as the inner conscious thought does and the moment one consciously thinks of something else, one's inner gaze is in turn impressed by another form. The form in memoria, i.e. (δ), and the form in the inner gaze (ε) are extremely similar (simillimae, 11.3.6; cf. Civ. Dei 11.26). However, they are numerically distinct and (ε) lasts only so long as the relevant conscious thought or experience (cogitatio) is taking place.22 In contrast, (δ) continues to exist in memoria afterwards (Trin. 11.7.11).
It is important to appreciate that this higher-level process, i.e. cogitation, is not something which occurs merely in imagining things while closing one's eyes or in remembering events of days gone by. Instead, it seemingly occurs whenever we have fully conscious or diachronically unified conscious experience of things. In such instances, one collects or joins together (cf. colligere) several different forms or representations into one diachronically unified experience in which we (e.g.) see objects as unified objects, hear speech as words, and so on. Given that much of our ordinary perceptual experience seems to be of this (diachronically unified) nature, it seems that cogitation occurs extremely frequently in our ordinary perceptual experience even if this kind of awareness is distinct from perception as Augustine characterises it.23 22 Since (δ)-(ζ) are all purely psychological and incorporeal items, Augustine regards them as not differing in their fundamental nature and thereby thinks (δ)-(ζ) form a more cohesive unity than do elements (α)-(γ) (Trin. 11.3.6-4.7, 7.12). 23 Silva 2014, 88-90 advances a dual-process reading of corporeal and spiritual vision in Gn. Litt. according to which they are simultaneous. While I agree that ordinary perceptual experience is composed of at least two processes, my own account of these processeswhich focuses on De Trinitate-is somewhat different from Silva's (e.g. he does not discuss cogitatio or the role of memoria and construes the nature and role of conscious awareness rather differently). In a recent paper (which was not available at the time this paper was Nawar Phronesis 66 (2021) 84-110 Thus, in De Trinitate, Augustine sees much ordinary perceptual experience as being made up of at least two processes. On the one hand, there is the seeing (videre) or perceiving Augustine attributes to the outer man, i.e. the process composed of elements (α)-(γ). This seems to be a lower-level process which does not require diachronically unified conscious awareness and yet it suffices for perception to occur.24 On the other hand, there is the cogitatio Augustine attributes to the inner man, i.e. the process composed of elements (δ)-(ζ). This is where diachronically unified conscious awareness is primarily 'located' . Insofar as ordinary perceptual experience typically manifests diachronic unity and conscious awareness, process (δ)-(ζ) typically occurs at the same time as (lower-level) perception. (Augustine's other works provide similar but not always clearly identical accounts.)25 submitted), Brower-Toland 2020 offers a similar dual-process view of perception in De Trintiate. Brower-Toland and I are in agreement on numerous points, most notably in taking conscious awareness to be primarily located in cogitatio. However, we also differ on several issues. Thus, for instance, Brower-Toland assumes that perception is necessarily conscious (e.g. 2020, 42-3, 65, 70, 73) whereas I deny this claim. Thus, although her use of the term 'paradigmatically' complicates matters, she seemingly denies that seeing, i.e. the process composed of (α)-(γ), is perception and instead thinks that the process composed of (α)-(γ) and cogitatio, i.e. the process composed of (δ)-(ζ), jointly constitute perception. In contrast, I think that (α)-(γ) on its own constitutes perception-while nonetheless agreeing that ordinary perceptual experience is constituted by these two processes together-and that this is important for understanding Augustine's definition of perception in other works (something which Brower-Toland does not discuss). For further differences, see below. 24 Silva says that 'the physiological process … whereby the organ receives the form of the sensible thing … is not seeing ' (2017, 47). Equally, Brower-Toland thinks that (α)-(γ), without conscious awareness, is not an instance of perception. However, Augustine clearly identifies the process of (α)-(γ) as seeing (videre, Trin. 11.1.1) and I take it to be an instance of perception. Moreover, note that Brower-Toland takes (α)-(γ) to be non-conscious (e.g. 2020, 67-8). Even though I do not hold that perception is necessarily conscious to any significant degree, I (more weakly) claim that process (α)-(γ) does not, on its own, manifest diachronically unified consciousness, and whereas Brower-Toland takes memoria to be responsible for all conscious awareness, I (more weakly) claim that it is responsible for diachronically unified conscious awareness. 25 In De Libero Arbitrio, Augustine draws a distinction between the senses perceiving their proper objects (e.g. sight's sensing of colours, etc.), and the actions of an 'inner sense' (sensus interior Although drawing distinctions between different 'levels' or 'grades' of sensation or perception has notable precedents, such as Plotinus, and consequents, such as Descartes,26 Augustine's account of ordinary perceptual experience might seem unusual to some modern readers. However, there are, I think, at least three important things to appreciate about Augustine's account. First, contrary to what is sometimes thought (e.g. Kalderon 2017, 23-4), conscious awareness-or at the very least substantive conscious awareness-does not seem to be a constituent of perception and someone whose attention has wandered may nonetheless be hearing (Trin. 11.8.15). In such cases, the activity Augustine attributes to the outer man is taking place. Yet, although the subject hears the sounds produced by a speaker, the subject is not consciously aware (at least to any significant degree) of what the speaker is saying. The relevant forms have made their way from the external world into the perceiver's sense(s) but-due to the agent's will or focus being turned elsewhere-have (The third kind, intellectual vision, is directed at intelligible items and is not relevant). Corporeal vision takes place through the eyes (12.6.15) and shares significant similarities with the process composed of elements (α)-(γ) which Augustine attributes to the outer man in De Trinitate. Spiritual vision is a process which occurs through the soul (or the imaginative part of the soul) during mental imaging of corporeal things. It deals with imagines and similitudines and is associated with conscious thought (cogitatio, e.g. Gn. Litt. 12.11.23,12.25). It may occur independently (as when one closes one's eyes and uses one's imagination) (12.24.51) but also-we may suppose, in agreement with Silva 2014occurs simultaneously with corporeal vision. Precisely how it should be characterised, especially relative to process (δ)-(ζ), merits independent discussion. 26 Most saliently, Plotinus seems to distinguish between the body's 'awareness' or reaction to perceptual stimuli and the soul's awareness of the relevant stimuli while seemingly wanting his account of perception or sensation to encompass both processes. not adequately made their way through memoria so as to impinge upon the inner gaze of the soul as unified, conscious experiences (11.9.16). Secondly, Augustine's account of perception involves various forms ( formae, species). In ordinary conscious perceptual experience (which involves seeing an object as unified) at least the following forms or sets of forms are involved: (α) the form(s) of the external object; (β) the form(s) impressed upon the sense; (δ) the form(s) in memoria; and (ε) the form(s) impressed upon the inner gaze of the soul. These four forms (quattuor species) are numerically distinct (I emphasise the point because direct realist readings seem to be committed to its denial). With the exception of (α), each form in some way causally depends upon the prior (e.g. Trin. 11.9.16;cf. 11.2.3,. Moreover, while (α) and (δ) are fairly stable items, (β) and (ε) are transient and last only so long as the sense and the inner gaze are fixed upon the relevant objects (11.7.11). Although Augustine does not explicitly discuss why so many forms are needed, he might be assuming that each salient 'organ' or 'faculty' (in some loose sense of these terms) involved in perceptual experience (and neighbouring phenomena) must have a copy of the relevant form produced in it.27 Thirdly and finally, neither those who see Augustine as an idealist or antirealist who holds that the objects of perception are mind-dependent items (e.g. O'Daly 1987, 95-6, 106;1999, 411),28 nor those who see Augustine as a direct realist (e.g. Matthews 2002, p. xxi, 64;Silva 2014, 81, 91, 96)29 seem to get Augustine right. Idealist or antirealist readings (such as O'Daly's seems to be) think that the objects of perception are mind-dependent items such as mental images or inner forms. However, such readings are mistaken. The inner forms are not the objects of perception but are instead the means by which we perceive external items. The objects of perception are external, mind-independent things which are common objects (i.e. not private items, resembling sense-data) and which exist prior to and independently of our perceiving them (Trin. 11.2.2 ff.). They impress forms in our sense organs (et ex uno visibili multorum cernentium 27 William Ockham, who criticised earlier theories for extravagantly multiplying species and formae beyond necessity, plausibly suggests they might be motivated by the desire to avoid action at a distance (Quaestiones in Libros Sententiarum Q13; cf. Pasnau 1997, 161 ff.).

28
'Augustine repeatedly stresses that sense-perception is perception of images of bodies (imagines corporum), and not of the bodies themselves ' (1987, 95); 'when we perceive, we perceive the likenesses of the objects perceived' (96); 'sense-perception is perception of images of objects, not of the objects themselves ' (1999, 411). 29 E.g. 'this is Augustine's statement of direct realism, it is to the object that the intentio is turned in the act of perception; we only perceive the external thing' (Silva 2014, 81).
formatur aspectus, 11.11.18) and are items about which we can fruitfully intersubjectively agree or disagree (e.g. Gn. Litt. 12.24.51;Trin. 12.12.17).30 Contra direct realist readings, although Augustine describes the external object perceived as a constituent of seeing (i.e. the process attributed to the outer man), Augustine holds that in ordinary perceptual experience we perceive mind-independent, external items in virtue of something in the perceiver standing in a suitable relation to certain intermediary forms and, ultimately, to the mind-independent object in the external world.31 The relevant intermediary forms in the perceiver are numerically distinct from those of the extramental object perceived (e.g. Trin. 9.11.16; 10.5.7; cf. De Civitate Dei 11.26).32 That is to say, Augustine thinks that a perceiver, S, perceives (α) (some mindindependent external object or its features) by means of (β) and has conscious and diachronically unified perceptual experience of (α) by means of (β), (δ), and (ε) (and perhaps also other items). The form in the sense, (β), is a kind of likeness (similitudo) or representation (imago) of (α), i.e. the form of the external object perceived (Trin. 9.11.16; 10.5.7; 11.2.3). So too (δ), the form in memoria, is also a likeness of (α) (and of (β)) (e.g. detracta specie corporis quae corporaliter sentiebatur, remanet in memoria similitudo eius, 11.3.6; cf. 11.9.16). The form in memoria, (δ), stands in some appropriate relation to S so that it impresses 30 E.g. 'although one sense is mine and the other is yours, it can happen that what we see is not one thing as mine and another as yours, but instead a single thing in front of each of us, seen simultaneously by each of us' (Lib. In support of attributing direct realism to Augustine, Silva 2014, 90-1 adduces Gn. Litt. 12.11.22. There, Augustine says: 'when something is discerned by the eyes, its representation immediately comes about in the spirit; however, its coming about is not recognised unless the eyes are removed from that which we were seeing through the eyes, so that we may then see that thing's representation in the soul' (nam cum aliquid oculis cernitur, continuo fit imago eius in spiritu; sed non dignoscitur facta, nisi cum ablatis oculis ab eo quod per oculos videbamus, imaginem eius in animo invenerimus). While I agree with Silva that external, mind-independent items are the objects of perception, Augustine is not claiming that we perceive such external objects without any inner representation(s). Instead, as Augustine goes on to repeat in De Trinitate (11.2.3 ff.), he is merely claiming that we only become aware of the coming about or the existence of such imagines by reflection. As far as I am aware, there is no evidence in Augustine for the kind of direct realism advocated by some later medieval philosophers (cf. Pasnau 1997, 168 ff.). 32 'We perceive colours, for example, by seeing, sounds by hearing, smells by smelling, tastes by tasting, hard and soft things by touching. In all these cases it is representations extremely similar (imagines simillimas) to the sensible objects, but not the corporeal objects themselves, which we turn over in conscious thought (cogitatione versamus) and retain in memory and, by means of these, desires for such things are stirred' (Civ. Dei 11.26).

Nawar
Phronesis 66 (2021) 84-110 S's inner gaze. It thus results in (ε), which is numerically distinct from (δ) but extremely similar (simillima, 11.3.6) to it, and (ε), which determines the object of the perceiver's conscious awareness, somehow reveals (α).33 Insofar as (ε) and (δ) and (β) represent or reveal (α), and (ε) and (δ) and (β) have the relevant roles in perception described above, and (ε) and (δ) and (β) have basic semantic properties (e.g. they signify, represent, etc.; cf. Gn. Litt. 12.11.22 ff.) which play a crucial role in perception, then Augustine seems to be best understood as an indirect realist or representationalist.34 While it is not entirely clear whether Augustine thinks representation can be reductively explained, he does makes clear that he thinks that representation is at least partially susceptible to philosophical analysis (De Genesi ad Litteram imperfectus liber 16.57): Every representation is similar to that thing of which it is a representation, and yet not everything that is similar to a thing is its representation (omnis imago similis est ei cuius imago est nec tamen omne quod simile est alicui etiam imago est eius). For instance, in a mirror and a picture, because there are representations, they must be similar to that of which they are representations. Equally, even if two men are similar to each other, if one was not born from the other neither of them can be said to be a representation of the other. A representation exists, then, when something is printed out from something (imago enim tunc est cum de aliquo exprimitur).
Augustine thus thinks that if x is a representation (imago) of y, then x is similar to y (in a certain way) and that x must be adequately caused by y. In this respect Augustine is more informative than some of his medieval successors whose accounts of representation have received more attention (cf. King 2005;Brower and Brower-Toland 2008), and it seems that Augustine does not think that the representational properties of the relevant items are determined 33 On the basis of Plotinus' remarks about συμπάθεια, action at a distance, and εἴδωλα (cf. Enn. 4.5.2.1-9; 6.3.15.24 ff.), Emilsson 1988  purely by their intrinsic properties. However, there are at least two important things to appreciate about representation as Augustine understands it. On the one hand, similarity or isomorphism does not suffice for representation (and, by implication, neither would other symmetric relations) and there is little reason to think that representations, especially mental representations, represent through resemblance. As Augustine points out, while twins or eggs may look alike, they do not thereby represent one another (cf. De Diversis quaestionibus 74). Moreover, while some kind of similarity between a representation and what it represents might be required , a high degree of similarity is not necessary and the similarity may be fairly 'loose' or abstract.35 Representations often lack many of the properties inhering in the thing from which they are printed out (and which they represent) (multa desunt imagini quae insunt illi rei de qua expressa est, Div. Qu. 74) and incorporeal representations in the soul will clearly lack many features of the corporeal items they represent (cf. Gn. Litt. 7.21.29;12.4.10 ff.).
On the other hand, an appropriate causal connection is seemingly required between a representation and what it represents (but is also clearly not sufficient). The representation must in some sense be 'printed out ' (exprimitur,Gn. Litt. imp. 16.57;Div. Qu. 74),36 or come about (gignere, Gn. Litt. imp. 16.58) from the thing it ultimately represents.37 This causal connection between the agent's representations and the mind-independent objects in the world explains why Augustine's account of perception and perceptual experience does in fact put us in touch with the world and provides, I think, an adequate response to the 35 The degree to which various kinds of mental representations are caused by and resemble their objects had been discussed by the Stoics (SE, M. 7.228-32;PH 2.70-6;DL 7.49-51;Nawar 2014;2020, 152-3)

Nawar
Phronesis 66 (2021) 84-110 worry raised above concerning Augustine's earlier definition of perception and the connection between subject and world (see Section 2 above). The form impressed onto the soul's inner gaze is thus: a representation; what the soul is immediately aware of; the means by which it perceives mind-independent things in the external world; and a kind of window onto the external object (Trin. 11.9.16) in something like the manner that a mirror (or perhaps a television showing a live programme) is a 'window' onto some remote object.38 Insofar as conscious thought (cogitatio) results from the suitable combination of elements (δ)-(ζ) and the object of a person's conscious awareness is determined by what their inner gaze is informed by, it seems that Augustine holds some kind of 'Cartesian theatre' view (of the kind much maligned by Rylean disciples).39 Diachronically unified conscious perception and other perceptual-like processes share the fact that what the inner gaze is immediately fixed upon is (δ), the form(s) in memoria (cf. Trin. 11.7.12).
However, although (δ) is what we are immediately aware of, it is typically that through which we perceive (in something like the way in which one's spectacles are items through which we see) rather than, properly speaking, the object of our awareness or of perception.40 In diachronically unified conscious perception, the forms in memoria (impressed onto the inner gaze) stand in a suitable relation to some external object (α) and are suitably sensitive to changes in (α) so that changes in (α) are suitably reflected by changes in (δ).41 38 Olivi-seemingly with Augustine on his mind (Quaestiones in Secundum Librum Sententiarum Q59, pp. 536-7)-objects to this idea (e.g. at 536: et quando aspicit speciem, non est ibi dare alium aspectum quo a specie transeat ultra ad rem aspiciendam … quia species non est aliquid quod posit realiter pertransiri, nec ipse aspectus est aliquid quod posit in proposito tales transitus habere). For recent discussion, see Adriaenssen 2017, 41-9. 39 However, Augustine is not claiming that we see (e.g.) a black cat by having the form of a cat within memory which is itself literally black impressed upon our inner gaze and which would in turn require some further process in order to perform its function. Accordingly, it is not clear to what extent Augustine's view is defectively homuncular or leads to objectionable regresses. 40 As Augustine describes things, the will (voluntas) can come to be so focused upon (δ) that the inner gaze (acies) is wholly fixed upon the inner appearance (ad interiorem phantasiam tota confluxerit, Trin. 11.4.7), takes (δ) to be the object of perception, and fails to see 'beyond it' (to (α)) (cf. Gn. Litt. 12.25.52). Cf. Olivi,Quaestiones in Secundum Librum Sententiarum Q74, In contrast, in cases of illusion or other processes which fall short of perception, we may suppose that changes in (α) are not suitably reflected by changes in (δ). Thus, for instance, in hallucinations or sleep, wherein there is no external object making an impression upon one's outer senses, there is no suitable connection between some external object and the form impressed on the inner gaze and thus such episodes are not instances of perception. In perception one perceives external, mind-independent objects by means of the relevant imagines. In hallucination (where the object experienced is what is sometimes called a In sum, Augustine thinks that ordinary perceptual experience is constituted by at least two salient processes: a lower-level process, perception, attributed to the outer man; and a higher-level process, cogitation, attributed to the inner man. The lower-level process attributed to the outer man is itself constituted by a polyadic relation between: (α) the external mind-independent object perceived (or its relevant form); (β) the form impressed in the perceiver's sense; and (γ) the intentio of the perceiver's soul. This lower-level process, on its own, suffices for perception but does not suffice for substantive awareness, i.e. for conscious and diachronically unified perceptual experience.
The higher-level process attributed to the outer man is constituted by a polyadic relation between: (δ) the relevant form in memoria; (ε) the form impressed in the perceiver's inner gaze; and (ζ) the will (voluntas) or the will's intentio. This higher-level process results in diachronically unified conscious awareness. These two processes jointly involve at least four numerically distinct forms, and (β), (δ), and (ε) are representations. When the lower-level process and the higher-level process are suitably connected-such that (β) is caused by, sensitive to, and represents (α) and (δ) and (ε) suitably track (β) and thereby (α)-then conscious diachronically unified perceptual experience occurs.

Active Perception and Intentio
As noted above, it is often said that Augustine's account of perception should be regarded as distinctively active. However, despite verbal agreement on this issue, insofar as readers have attempted to render precise what is meant by 'active' (beyond saying, e.g., that perception is something we do), they have variously maintained that Augustine's account of perception is distinctively active in virtue of the fact that: (A) Augustine thinks that conscious awareness is constitutive of senseperception (e.g. O'Daly 1987, 85 ff.; Caston 2001, 39;cf. Silva 2014, 79-80, 87); or (B) Augustine holds an extramissive account of vision, according to which rays of light are projected from the eyes (e.g. Miles 1979, 16;Toivanen 2013, 135-7); or 'non-existent object') one is having an experience of some mind-dependent imago (cf. Gn. Litt. 12.2.3 ff., 18.39 ff.).

Nawar
Phronesis 66 (2021) 84-110 (C) Augustine thinks that in perception the soul is not acted upon or causally influenced by bodily things but instead acts upon itself (e.g. Caston 2001, 40;Silva 2014;2017, 44). We are now in a better position to appreciate that none of these views is entirely accurate. As regards (A), we have seen that Augustine does not take conscious awareness to be constitutive of perception in any straightforward sense. As regards (B), while it is true that Augustine seems sympathetic to an extramissive theory of vision, and may have thought that something very roughly analogous holds in the case of some of the other senses (cf. Mus. 6.5.10, 12; Gn. Litt. 3.4.6 ff.;4.34.54;), this does not offer much insight into what occurs within the perceiver's soul and is not an area where Augustine's views are distinctive. In fact, Augustine here readily defers to what he takes to be the relevant authorities without much in the way of creative or critical input.42 As regards (C), this is partially but not entirely correct. Thus, it is true that, on Augustine's view, it is the soul that often initiates the process of perception by directing the senses towards their objects. It is also true that the soul does indeed act upon itself, most saliently in the process Augustine attributes to the inner man, i.e. cogitatio, which is constituted of elements (δ)-(ζ) (cf. Mus. 6.5.9 ff.; Gn. Litt. 12.16.33). However, contrary to what is often claimed,43 Augustine does not think that the soul is causally isolated from the corporeal things it perceives or that bodily things cannot act upon the soul.44 Given 42 Augustine often signals deference to authorities when he remarks that there is an emission (effusio, Mus. 6.8.21; emissio, Gn. Litt. 1.16.31;cf. emicare,Gn. Litt. 12.16.32) of corporeal rays from the eyes which, like a radar system, interact with perceivable objects before returning to the eyes and bringing about a change in them (cf. Gn. Litt. 7.13.20, 19.25;12.16.32;Mus. 6.8.21;Trin. 9.3.3;11.2.4;O'Daly 1987, 80-4). Such extramissive theories found favour in Empedocles (Theophrastus, De Sensibus 7-11), Plato (cf. Meno 76c-d; Timaeus 45b-d), the Stoics (DL 7.157), Galen (De Placitis Hippocratis et Platonis 7.5.1.1 ff.), according to whom πνεῦμα is emitted from the eyes (cf. Ierodiakonou 2014), and others. 43 E.g. 'Augustine firmly denies that any body can act on the soul and so that the visible object can affect the soul in any way' (Caston 2001, 40). 'It is impossible for something material to act upon something immaterial, or in other words for something lower on the scale of being to act upon something higher on that same scale' (Silva 2014, 85); 'perceptual experience is not caused by the sensory stimuli' (Silva 2014, 87). Cf. Markus 1967, 376;Miles 1979, 16. O'Daly, who thinks there is 'reciprocal influence of body and soul ' (1987, 45, cf. 84, 122-3), is-I think-correct (see below), but is an outlier. 44 A detailed account of Augustine's views of causation and the causal relations between body and soul remains to be written (cf. Stead 1986). Here it suffices to note that in spite of the occasional stark remarks (e.g. Mus. 6.5.9-10) and the fact that Augustine claims that bodies do not bring about anything ( facere aliquid) in souls (Mus. 6.5.9; Gn. Litt. 7.19.25;12.20 that Augustine is a realist who thinks that we perceive mind-independent objects (see above), that the relevant inner forms involved in perception are representations (imagines), and that representations are caused by what they represent Div. Qu. 74), there is good reason to think that-even though the soul might act upon itself in perception-corporeal, mind-independent objects nonetheless figure among the causes of the relevant inner forms and representations.
In what sense(s), then, should Augustine's account of perception be regarded as active? At the most general level, it is true that for Augustine conscious perceptual experience is something which manifests the soul's agency and something the soul does (e.g. Gn. Litt. 10.12.21). Moreover, it is true that-as I have emphasised above-memoria (a part of the soul) is seemingly always actively involved in diachronically unified perceptual experience. However, it is possible to offer a somewhat more precise account of the active nature of perception and there are, I think, five principal aspects which deserve attention.
First, the soul is continuously acting upon the body-including its perceptual organs-in order to sustain them, maintain their unity, and grant them their functions (e.g. De Quantitate Animae 33.70).45 This includes rendering the perceptual organs susceptible of taking on the forms of perceivable objects Gn. Litt. 3.4.6;7.15.21,18.24). The soul performs these activities by means of a vital intentio (Mus. 6.5.9;Gn. Litt. 7.19.25) and in these contexts, it seems that 'intentio' (a term which is difficult to translate and seems highly polysemous)46 does not denote something akin to conscious attention or not thereby maintain that bodies cannot act upon or causally influence souls simpliciter. Instead, he is merely claiming that they cannot act upon or causally influence the soul in certain ways (e.g. as efficient causes, cf. Civ. Dei 5.9.4; Imm. An. 8.14; i.e. as active, sustaining, and productive items whose continued causal influence is necessary to sustain their effects in existence, cf. Cicero,18.41 ff.;Seneca,Ep. 65.4;SE,. This leaves open that bodies may act upon souls as antecedent, proximate, cooperative, and other kinds of causes (e.g. Civ. 15;cf. Cicero,De Fato 18.42;SE,Clement Stromata 8.9.25.1 ff.;. Thus, Augustine speaks of souls being affected by the things they perceive (e.g. omnis enim anima viva … visis movetur, Gn. Litt. 9.14.25;cf. Trin. 10.2.2,7.10;11.5.9,8.14; and bodies acting on souls (e.g., Civ. Dei 14.5,15.2;cf. Trin. 3.3.8,8.15). 45 Cf. Imm. An. 8.14,De Natura Boni 5,15,18;Div. Qu. 54;Mus. 6.4.7,5.11,14.44;Gn. Litt. 3.16.25;7.18.24;Civ. Dei 21.3;Ep. 166.2.4. 46 Discussions of 'intentio' in Augustine are offered by O'Daly 1987, 29, 43-5, 84-7, 108-11, Di Martino 2000, Caston 2001, 38-45, and Alici 2008. However, I differ from these treatments on several points. For instance, the aforementioned are often inclined to render intentio as 'intentionality' or 'attention' (e.g. Silva 2014, 82 ff.;Kalderon 2017). I take this to be infelicitous. As regards the former (i.e. 'intentionality'), it is not clear that we find the term being used in the way that Brentano found inspiring but which seems mysterious to some Nawar Phronesis 66 (2021) 84-110 awareness (a point which may mislead some readers). Instead, it is comparable, perhaps, to digestion and is: a means by which the soul sustains the body's functions; something of which we are rarely, if ever, aware; and something over which we exercise little or no direct control. More concretely, here intentio seems to denote something like the tension of the finer elements which the soul uses to animate and control the body (Gn. Litt. 7.19.25;cf. Ep. 166.2.4) and the activity by which the incorporeal soul directs the body (Gn. Litt. 8.21.42). Secondly, by means of its intentio, the soul is responsible for priming the senses to be formed (et sensum formandum admoveat ei rei quae cernitur, Trin. 11.2.5). It often directs the senses towards their objects (e.g. rei sensibili sensum admovet, 11.2.5) (and also away from them, 11.8.15) and also sustains the senses upon their objects.47 The relevant priming goes beyond rendering the senses susceptible to being formed (see above); instead, it renders them inclined to being formed. This seems to manifest itself more broadly as something like an openness to the world (cf. McDowell 1994, 29 ff.) or some kind of striving-comparable, perhaps, to Spinozan conatus-towards perception which is at least partially satisfied when perception occurs (Trin. 11.5.9-6.10; cf. Mus. 6.13.42).48 The directedness and sustaining features of intentio most obviously manifest themselves in control of the senses (e.g. turning one's of us (as when Aquinas says the perceiver's eye receives the intentio coloris of the colour it sees but does not thereby become coloured, e.g. Aquinas, Summa Theologiae 1a2ae.22.2 ad 3). As regards the latter, as we shall see below, there is often intentio without awareness, attention, or voluntary control. The polysemy of intentio was long ago noted by Scotus (Reportatio Parisiensis 2, d. 13, q.1) but seems to not be sufficiently appreciated by many modern readers. By my count, uses of intentio in the relevant prior Latin literature may denote: (i) stretching out (as per the Stoic theory of vision, Aulus Gellius Noctes Atticae 5.16.2); (ii) the 'tuning' or harmonious arrangement of a thing (e.g. Cicero, Tusculan Disputations 1.10.19 (used to describe the soul as per Aristoxenus' theory)); (iii) tension, tensing or concentration (e.g. Cicero, TD 2.23.54, 27.65; Seneca, Quaestiones Naturales 2.8); (iv) the attention of the soul or mind more generally (e.g. Cicero, TD 4.2.3; Quintilian 1.1. 34,2.11;Celsus,De Medicina 2.10.4;3.18.11); (v) raising or tensing one's voice (e.g. Quintilian 1.10.25); (vi) intensity (e.g. of fever, Celsus, Med. 3.18.7); (vii) intention or intent (Quintilian 3.6.14); and (viii) going forth or attacking (e.g. Quintilian 3.6.21, 9.1). Cf. ἐπίτασις ('stretching' , 'tensing') (Gellius, NA 6.7). 47 The directing, sustaining, and joining are noted by Caston 2001, 38-9 andKalderon 2017, 29, but are understood somewhat differently and taken to be more straightforwardly consciously volitional. 48 A further parallel, brought to my attention by an anonymous reader, may be drawn with the account of cognition sketched in the Timaeus (wherein the active soul seems to always be reaching out towards its potential objects; for recent discussion, see Corcilius 2018). eyes, etc.),49 and the senses' sensitivity to the forms of external things (cf. Gn. Litt. 3.4.6,7.15.21,18.24).50 In these contexts, intentio seems to denote something like the soul's will (voluntas animi, Trin. 11.2.5; cf. intentio animi, 11.4.7, 6.9-10) or its desire to see (videndi appetitus) (11.2.2). It is something-comparable, perhaps, to breathing, love, or lust-which the soul may come to be aware of and control, but often operates subconsciously (cf. Ennaratio in Psalmarum 54.1).51 Thirdly, the ensouled and functioning sense organ takes on the perceivable form ( forma, species) or likeness (similitudo) of the object perceived. In the case of vision there is extramission (rays are projected from the eyes), but in all cases the sense organ's taking on of form is the passio corporis identified in (SENSE-PERCEPTION*) (Quant. An. 25.48; 30.59) and the process described as visio in De Trinitate 11 (with something analogous holding of the other senses). As Augustine describes things, the external object is a 'quasi parent' (quasi parens) of this process which occurs in the outer man and it seems that this process is the result of the perceivable object acting upon the perceiver.52 Even though the object's continued presence is typically required for the sense to be impressed with the relevant form, the cooperation of the perceiver's soul and its intentio are required, and there thus seems to be a significant difference here between Augustine and some later Aristotelian accounts, such as that of Aquinas.53 The perceived object does not, Augustine emphasises, by itself bring about the impressing of the relevant form in the sense (Trin. 9.12.18;11.2.5,3.6,49 This was also noted by some later Aristotelian accounts, notably that of Aquinas (e.g. Summa Theologiae 1a2ae.9.1c.), which emphasised the role of the will in directing the sense organs. 50 Olivi, inspired by Augustine, says that the aspectus suae intentionis needs to be directed towards the object (Quaestiones in Secundum Librum Sententiarum Q72, p. 9) but how such remarks should be understood is controversial (cf. Pasnau 1997, 132 ff.;Toivanen 2013, 153 ff.). 51 MacDonald takes Augustine's account of cognition to be voluntarist in that an agent's 'act of will is essential to, partly constitutive of, any act of sense perception ' (2012, 245) and worries that that is highly implausible and cannot apply to non-human animals, who lack the faculty of will (2012, 244-5). As one possible response, he suggests that perhaps 'sense perception sometimes occurs automatically, apart form will's activity ' (2012, 246). My own account supports and develops this suggestion (but, in contrast to MacDonald, I am not even initially inclined to regard the relevant voluntas as requiring awareness and less inclined to posit a faculty-like psychology). 52 'And the form itself which is impressed by it, which is called "vision" ' (et ipsa forma quae ab eodem imprimitur sensui, quae visio vocatur. Trin. 11.2.2); 'the informing of the sense, which is called "vision" ' (sensus informatus sit, quae visio vocatur, 11.2.3). 53 Aquinas-who himself explicitly follows Augustine in emphasising that the soul's intentio is required for the act of any cognitive power (e.g. Quaestiones Disputatae de Veritate 13.3c; Summa Contra Gentiles 1.55.4)-nonetheless sees perception as a process involving an agent (the perceived object) which in itself is sufficient for inducing its form in what it acts Nawar Phronesis 66 (2021) 84-110 4.7, 5.9),54 but it does nonetheless play a robust causal role in the perceiver's sense taking on the relevant form. Fourthly, the soul-or, more precisely, the will (voluntas)-is responsible for 'joining' (coniungere) the senses to their objects (Trin. 11.2.2,3.5,4.7,5.9,cf. Mus. 6.13.42). In the lower-level process attributed to the outer man, it is not entirely clear what the relevant joining amounts to. However, in the higher-level process attributed to the inner man, the soul brings about a spiritual form within memoria which matches that impressed upon the perceptual organ (e.g. Gn. Litt. 12.11.22,16.33,33.49;Trin. 11.9.16) and combines imagines within memoria (e.g. Trin. 11.10.17; cf. 10.5.7) to provide diachronically unified experience or the raw material from which such experience is fashioned. The spiritual form or imago is brought about by the soul and the relevant imagines are processed by the soul into a unified representational state and the soul thus seems to be acting upon itself (cf. Mus. 6.5. 13.39;Gn. Litt. 12.16.33,18.40;Trin. 10.5.7).55 Fifthly, in the conscious awareness which typically accompanies perception (i.e. cogitatio, the higher-level process attributed to the inner man) and is a component of much ordinary perceptual experience, the gaze of the soul or mind (contuitus animi, acies animi, conspectus animi; cf. acies cogitantis) fastens upon some form(s) or imagines in memoria. In doing so, the soul's gaze is informed by the relevant form in memoria and a further form is thereby brought about, this time in the soul's gaze (Trin. 11.3.6,7.11,9.16). Once again, some 'joining' takes place and in this case it seems to involve making one conscious and diachronically unified appearance or an appearance as of one thing (utrumque in cogitando ita copulat ut tamquam unum singulariter appareat, 11.7.12; cf. 11.8.13). For Augustine, then, imagination and memoria are always involved in conscious perceptual experience and here too the soul acts upon itself. The process by which the soul's inner gaze turns to some particular form or another is described as being directed by the will (voluntas, 11.3.6), and it is only when the soul's gaze is informed in the relevant way that the relevant upon and a patient (the perceiver) which in no way cooperates with the agent (patiens quod in nullo cooperatur agenti, Quodlibet 8.2.1c). 54 Augustine regards (α) as being some kind of cause from which the impressing of the form upon the sense, i.e. (β), occurs (e.g. ex corpore quod videtur gignitur visio, Trin. 11.2.3; cf. speciem corporis ex qua sensus formatur, 11.4.7). However, while (β) comes about from the visible thing, it does not come about from it alone (gignitur ergo ex re visibili visio, sed non ex sola, 11.2.3). 55 However, note that insofar as the relevant inner forms are representations of external items, the mind-independent perceivable objects still count as some kind of cause in this process.
form is consciously considered (cogitare, 11.7.11, 8.15) and conscious thought (cogitatio) takes place. In sum, as Augustine describes things, the inner gaze of the soul (acies animi)-which determines what the soul is consciously aware of-seems a bit like a set of roving spotlights controlled by the will (voluntas) or the soul's intentio (Trin. 11.8.12;cf. Gn. Litt. 12.13.27). As the above account should make clear, the soul is active in various ways in the relevant perceptual processes and Augustine's use of the term intentio is complex. In the relevant contexts, intentio denotes: (i) a vital 'tension'-of which we are not aware-by which the incorporeal soul acts as an efficient cause to the body and sustains its life, unity, and functions (e.g. Gn. Litt. 7.19.25;8.12.42; Epistulae 166.2.4); (ii) a directedness, of which we may be aware, closely associated with the will (voluntas), which directs the senses to their objects and primes them to receive forms; and (iii) something closely associated with the soul's conscious attention which may extend beyond the perceiver's body (e.g. Trin. 11.8.15;Gn. Litt. 12.13.25,27,23.49; cf. De Immortalitate Animae 10.17).56 There is a limit to what the gaze of the soul can take in at once and so it must be selective (Trin. 11.8.12). If all the 'spotlights' are focused exclusively within one particular aspect of oneself, then one may come to be alienated from one's bodily senses and fail to be aware of what is occurring externally (Trin. 11.4.7,8.15;Gn. Litt. 12.12.25 ff.;cf. Plotinus 4.4.8.8 ff.). In such cases, a lower-level process may occur, i.e. the process, described as visio and attributed to the outer man, in which the relevant form is impressed in the senses. As a result of this impression in the sense, the relevant form (and, through it, the relevant mindindependent external object) is not hidden from the soul and perception occurs (cf. Trin. 14.6.8; 15.9.16; Ep. 147.1.6 ff.), even if the soul is not substantively aware 56 The soul's intentio is closely associated with the soul's attention, the object of its attention, what focuses its attention (Gn. Litt. 12.20.42;Trin. 11.8.15;cf. Civ. Dei 19.19 where σκοπός is rendered as intentio), or the tensing or concentrating of its conscious thought (e.g. cogitationis intentio, Gn. Litt. 7.20.26;12.12.25;Trin. 11.8.15). However, even here it seems that one should be cautious about construing intentio as 'attention' . I leave it to readers to decide the precise extent to which intentio is polysemous (Augustine would happily accept the result as he thinks all words are equivocal and exhibit a significant degree of polysemy, cf. De Dialectia 9-10; Nawar forthcoming), but insofar as a single translation of intentio may be offered it seems that what is at issue is more a tensing but even so this is often metaphorical. There are possible parallels with the Stoic account of perception, according to which πνεῦμα may reach out from the perceiver to the object of perception (DL 7.52;Gn. Litt. 8.21.42;12.23.49;cf. Caston 2001, 43-5;Verbeke 1945;O'Daly 1987, 48;Solère 2008).

Nawar
Phronesis 66 (2021) 84-110 of the relevant form (as the soul's inner gaze is not suitably turned to 'face' it) and no diachronically unified conscious thought about the relevant object is taking place (although, of course, some conscious thought about something else might be taking place).

Conclusion
In this paper, I have sought to clarify some central features of Augustine's account of perception. In De Quantitate Animae, Augustine defines perception as 'an affection of the body which is not in itself hidden from the soul' (passio corporis per seipsam non latens animam, 25.48; 30.59). I argued that this should be taken at face value while noting that this definition faces at least two potential sets of worries, one set concerns antirealism or idealism (i.e. that it is not clear that perception puts us in touch with the world) and the other concerns the nature of the 'not hidden' relation (e.g. that it is not clear why Augustine does not speak more straightforwardly of awareness). I argued that Augustine characterises perception as he does because he does not take substantive awareness, i.e. diachronically unified and conscious awareness, to be constitutive of perception; instead, he takes such awareness to be constitutive of cogitation, a process which ordinarily accompanies the lower-level process he describes as 'perception' but is not required for it. I further argued that Augustine is best understood as an indirect realist who thinks we perceive mind-independent external objects by standing in a certain relation to inner representations of such objects which require an appropriate causal relation with the objects they represent. Finally, I clarified the manner in which Augustine's account of perception and perceptual experience should be regarded as active and offered a more nuanced account of precisely what the soul does in the various relevant processes that make up ordinary perceptual experience, while also clarifying the nature of the soul's intentio in each case.57