DEGREES OF ATTENTION IN EXPERIENCING ART

This paper examines gradients of attention in relation to aesthetic appreciation. My main claim is that we should leave open the possibility that aesthetic response might be triggered by stimulations taking place far from the centre of one’s focused attention. In support of this claim I first discuss the notion of ‘periphery of attention’ and the challenges that it poses to contemporary psychological theories of aesthetics. I provide four criteria for differentiating between several types of attentional processes and then proceed to single out the characteristics of non-focal types of attention(-related) processes with varying intensity such as pre-attentive processing, the mere exposure effect and psychic overtones. Finally, I reassess the periphery of attention in the light of its relation to aesthetic appreciation. I hold that given certain constraints such as repeated exposure, perceptual learning, encoding in long-term memory, and possibility of retrieval, subdued, inconspicuous forms of stimulation can elicit aesthetic responses.


I. INTRODUCTION
Attention and aesthetic appreciation of the arts share a common fate deeply rooted in the tradition of aesthetics and the philosophy of art. Aesthetic appreciation was often described as engaging a particular type of attention; for example, Wolff, Lessing, and Beardsley 1 famously held that works of art command our attention and provide privileged access to significant aspects of the world, by the careful selection and detachment of a set of stimuli from their environment. Clearly, by 'attention' they mean focused attention, understood as a form of perceptual attention. But the ways of addressing works of art are not limited to this exclusive state of deep concentration pursued for its own sake. My aim here is to question the notion of focal attention understood in the context of aesthetics and to give an account of episodes of aesthetic response constituted independently of focal attention. I am going to rely mostly on a paper by Thomas Munro, presented at the Third International Congress of Aesthetics in 1957, entitled 'Types of Form as Related to the Psychology of Attention' , 2 where he gives valuable insight into various conditions under which we perceive and attend to works of art. Munro's ideas fell unjustly into temporary obscurity and this particular text, highly significant but barely accessible, deserves closer analysis.
A motivation for salvaging this text is to acknowledge that there is an array of possibilities for the study of the role that attentional processes and marginal perception play in aesthetic experience. Considering that contemporary psychological aesthetics 3 takes for granted the idea that focal attention is a necessary condition for aesthetic appreciation without even considering alternative views, raising the question of degrees of attention in experiencing art brings to light psychological aspects of this type of experience which have been so far neglected. The temporal structure of aesthetic response is one such aspect. Endorsing the view that attention engaged in appreciative practices can be deployed in varying intensity rather than be reduced to an all-or-nothing affair enables us to consider instances where aesthetic response comes about gradually. It also begs the questions of how aesthetic response unfolds over time, how it gets initiated and comes to completion.
Here is how I propose to address these questions. In the first part of the paper I discuss Munro's views with respect to the perceptual conditions 4 under which we experience art and more specifically, the phenomenon of the 'periphery of Despite their attractiveness, none of these theories takes seriously the hypothesis that aesthetic response may occur or be initiated at the periphery of attention. 4 I thank Tereza Hadravová and Jakub Stejskal for suggesting that I keep separate the experiential problem of experiencing artworks under certain atypical perceptual conditions (that is, peripheral perception, non-focal attention, and so forth) and the ontological problem of whether some objects are better suited for such a perception than others. attention' (TF,p. 306) and the challenges that it poses to contemporary psychological theories of aesthetics. In the next section I briefly present a few criteria for differentiating between several types of attentional processesnamely, selectivity, duration, intensity, and agency; I will further use these criteria in a focused manner in order to single out the characteristics of non-focal types of attention (related) processes such as pre-attentive processing, the mere exposure effect and psychic overtones, which are central to my purpose of challenging the thesis that only focal attention is relevant to discussions about aesthetic experience. In the following section I further elaborate on the question of gradients of attention by presenting in greater detail these non-focal  (TF,p. 303). In this paper Munro goes on to reinforce the standard treatment of art appreciation as involving selective and sustained attention. He lists a number of psychological principles that are responsible for holding attention and can be put at the service of art. Here he mostly recapitulates the Gestalt laws of perceptual organization: thus, selectivity of a limited set of stimuli, their detachment from a context that would induce a 'different attitude' (that is, an aesthetic attitude) towards the proper object of contemplation, emphasis and intensification of the selected properties, variety and contrast, and, finally, internal organization or unity of form (TF, are in turn considered to play an important role in motivating sustained attention. In the light of these principles, works of art appear as high saliency stimuli. An important discussion is presented, however, in the second part of his paper, where Munro challenges this saliency condition and casts doubt on the universality of focal attention as a desired effect of all forms of art: 'Some art is adapted to attract and hold the attention of the observer; to be conspicuous and highly interesting; while another kind of art does not. The latter is content to play a more subdued, inconspicuous role' (TF,p. 306). He further rightly observes: Many kinds of art are made to be perceived marginally, not with focused attention: to recede somewhat into the background or periphery of attention, while the observer carries on other activities. This function of art has been ignored by aestheticians. Consequently, they ignore the ways in which many kinds of art are adapted to appeal through inattentive, marginal perception. (TF,p. 306) Art is understood here in a broad, loose sense, as the product of organizing various sensuous materials into a relatively coherent whole. Examples of such forms of art which recede into the background are Tafelmusik, decorative art, a cathedral façade commonly perceived on many occasions (TF, pp. 306-7), and any other form of aesthetic stimulation not interfering with current goal-oriented actions. 5 The quality of being inconspicuous or unobtrusive is considered both at the level of the work of art as a whole (Tafelmusik) and at the level of individual parts of complex works (film music playing an incidental [TF, p. 308] role in some experiences of watching a movie).
A limit to this view is that it is far from clear what exactly counts as the 'background or periphery of attention' and 'marginal perception' but Munro certainly puts his finger on a matter that hasn't yet been studied thoroughly enough by aestheticians -namely, that art (or any sensory configuration of potential aesthetic interest) can be experienced with varying degrees of attention, which leaves open the possibility that an aesthetic response could be evoked without focused attention. Here's another quotation of Munro's that supports this view: It must not be assumed, however, that the aesthetic power of art is weakened in proportion as forms become simple and unobtrusive. […] Subdued, inconspicuous art,  Oxford: Phaidon, 1979). Significantly, a potential title for this book was The Unregarded Art, that is, art experienced without focused scrutiny: 'The wallpaper, the pattern on the curtains, the scrolls on the picture frame […] rarely invite conscious scrutiny. True, we also may fail to notice a picture in a room or see it merely as a patch on the wall, but we know all the time that it is meant to be focused and contemplated. Painting, like speaking, implicitly demands attention whether or not it receives it. Decoration cannot make this demand. It normally depends for its effects on the fluctuating attention we can spare while we scan our surroundings ' (ibid., 116). Other examples, provided by Anthony Savile, include manuscript illumination or Pompeian frescoes; according to Savile, however, these are mere additives or even nuisances to more absorbing activities. Anthony Savile, Aesthetic Reconstructions: The Seminal Writings of Lessing, Kant and Schiller (Oxford: Blackwell, 1987), 67. like a gently insistent person, may have a deeper effect in the long run because one can enjoy its continuous or repeated presence. One's conscious attention is usually elsewhere, but one is vaguely aware of a shifting sensory field or background of sights and sounds, tactile sensations, occasional tastes and odors.
[…] A work of art which is seen or heard marginally on many occasions, such as a church or garden which one passes daily, may have deeper aesthetic effect in the long run than another which is seen once only with undivided attention, such as a motion picture film. (TF, What, I would argue, is important in this discussion of unobtrusive forms is the distinction between foreground and background in processing sensory information and the affective valence connected to it, the assumption being that an aesthetic effect can be initiated without focused, undivided attention. Equally important here is the regularity with which we are exposed to such unobtrusive forms (to their 'continuous or repeated presence'). My contention is that the regularity condition would enable such dim stimulations to be constructed in time into complete percepts, through successive memory encodings, thus securing them a persistent role; unlike mere background noise that is ignored and immediately forgotten, unobtrusive stimuli are less likely to be suppressed by higher order mechanisms since they do not impede the progress of the current activities in which we may be engaged. 6 I will come back to this question in the following sections with more substantial evidence from cognitive psychology. Now, what does it mean when we say an item recedes into the background of attention? One difficulty is to establish whether the polar opposite of focused attention is couched in terms of attention, perception, or awareness, all of them thorny questions in contemporary philosophy of mind. Moreover, the property that qualifies whatever this process might be is not easy to understand either.
Is this type of attention/perception/awareness marginal, peripheral, faint, or something else? One way of reading the just-quoted passage by Munro is to say that these inconspicuous forms of art appear somehow to touch us without getting our focused attention. Here, Munro does seem indeed to merge vague awareness with marginal perception, leaving out at least one kind of conscious attention. But I wouldn't go so far as to assume marginal perception to be inattentive. As I shall argue, it is questionable whether we can do without some 'Putatively, when unattended information is strong, attentional mechanisms detect and successfully suppress this distraction; but when it is weak, suppression is not elicited and learning occurs. ' the more so because it is precisely the question of degrees of attention in relation to art appreciation that is central to Munro's paper, as he states here: How much and what kind of attention is desired by artists for their products? How much and what kind of attention do works of art attempt to attract and hold? How much and what kind is it desirable to give, from the standpoint of maximum enjoyment? There can be no single, specific answer to these questions, for many different kinds and degrees of attention are sought and given. (TF,p. 306) It remains to be seen whether the indeterminacy of the relationship between attention and aesthetic appreciation rests on the multifold reality of artistic productions or whether the question of the varying strength or degrees of attention in experiencing art lacks substance altogether.
Before coming up with a preliminary hypothesis as to what capacities are concerned when experiencing a background of sensations, it could be helpful to look at available answers to this problem, which are to be found in the work of two art historians interested in the psychological implications of the perception of the arts -namely, Ernst Gombrich and Michael Baxandall. Gombrich and Baxandall gave several analyses of different modes of seeing and attending, which may help us capture the meaning of the 'periphery' or 'background' understood in the context of art perception. Gombrich, in particular, in The Sense of Order refers to the decorative motifs that we 'take in as background' 7 though they lie outside the focus of our attention. According to Gombrich, we 'see' these motifs without 'looking at' them and, at best, spare them 'fluctuating attention' . 8 When regularities in an environment are taken for granted, such as the decorative motifs of the interior of the Alhambra, they are scanned through undifferentiated, peripheral vision: Our confidence in the orderly interior allows us to make use of peripheral and provisional inspection where there is no contrary evidence in sight. And perceiving the further orders of decorative enrichment we can continue in the same way, narrowing the visual span while still relying on the continuities vaguely perceived on its periphery. 9 Importantly, this type of perception, which Gombrich calls 'global perception' , is not to be regarded as 'careless' since it can capture much better the general appearance of certain artistic configurations than deep scrutiny. Gombrich thus presents a dynamic approach of the peripheral perception of order and the aesthetic effect that it might create. In other words, the elements sensed at the periphery of attention, far from being discarded, can fuse to create the impression of a coherent whole, such as the architectural interior of the Alhambra palace and fortress in Granada. As integral parts of a complex piece of work, these subdued elements will ultimately be integrated in a global percept. As we have seen, Munro also considered the inconspicuousness of individual parts of complex works but what is more intriguing in his theory about the periphery of attention is that it allows the integration of various elements into a percept to be deferred; the integration and the aesthetic effect that it might trigger don't necessarily occur in a singular experience but unfold over a much longer time.
For instance, to paraphrase Munro, the decorative motifs on the façade of a cathedral vaguely perceived on many occasions on one's way to work could have a 'deeper aesthetic effect in the long run' than the singular experience of the architecture of the Alhambra palace. Munro doesn't provide any reason why this hypothesis should hold true but there is promise that his intuition will now be tested by considering evidence from cognitive psychology and, more specifically, memory studies.
Baxandall also stressed the importance of elements sensed marginally in art perception. Unlike Gombrich, however, he is interested in the periphery of a more restricted field of vision, comprising singular pictorial compositions, a Cubist painting, for instance, and the way in which we allocate our attention to different parts of such compositions. In many cases, artists choose to give a pictorial representation of different modes of vision, for instance, peripheral and focal vision; they provide, so to speak, a sense of what different perceptual experiences might look like: take, for example, Rembrandt's chiaroscuro technique or, more recently, Pepperell's blurry images aiming at depicting the peripheral visual space. 10 It is not surprising then that writers on art might feel the need to elaborate a psychology of art by drawing on specific case studies. For instance, in a study of Braque's painting Violin and Pitcher, Baxandall asks whether the psychology of attention can help us to understand the functioning of an intriguing pictorial element -namely, the nail depicted at the top middle of the canvas. 11 Well-informed by vision science, he makes a fine distinction between different gradients of perceptual processing, such as central foveal, parafoveal, and outer peripheral registrations. 12 What is particularly interesting is the way he reveals the virtues of coarse information sensed at the periphery and its importance to our experience of painting. Again, as with Gombrich, peripheral attention to individual elements is integrated in the global perception 13 of the composition as a whole, occurring in a singular experience.
Paradoxically, in this particular Cubist painting, it is only in a 'coarse, peripheral register' that the violin and pitcher 'survive' as full-blown, identifiable objects: 'If one fixates the nail, and makes oneself attend to the pitcher parafoveally nearby it becomes a regularized and trompe l'oeil pitcher, shimmering with vivid light effects. ' 14 Note that here what he actually describes is covert attention, which is by definition dissociated from the direction of the gaze and can be as vivid and intense as any eye fixation. 15 In other words, what is literally at the periphery of perception may happen to be in the foreground of attention; focal attention need not be perceptual.
With these notions of periphery in mind, we can venture to say that the periphery of attention, understood as a form of perceptual attention, generally refers to the region of the visual field where the power of resolution of the eye is low; the boundaries of this region are broadly defined as near periphery, immediately surrounding the centre of gaze or far periphery, reaching further away. Thus, coming back to Munro, a work of art or any type of artistic form 'receding into the background or periphery of attention' could be interpreted in terms of being perceived in the extreme periphery of the visual field.
The periphery of attention then is where the apprehension of sensory impressions is coarse-grained; it is an elusive zone, which we can hardly describe without appealing to elusive terms.
In what follows, I will try to substantiate the claims concerning the relation between the periphery of attention and art appreciation by drawing on studies in cognitive psychology and I will argue that the range of attention phenomena is much wider than an all-or-nothing response. Attention comes into play in varying degrees: we can pay more or less attention to various stimuli, sustain it for as long as necessary, or shift it transiently from one location to another. But before delving into the complications related to this grading process, I present four criteria that help to differentiate between several types of attention. More specifically, these criteria concern some geographical questions related to the reach of attention, its temporal span, qualitative feel, and the presence of subjective control.

13
I am indebted to Jakub Stejskal for this observation.

III. CRITERIA FOR DIFFERENTIATING ATTENTIONAL PROCESSES
The first and less contested characteristic of attention is no doubt its selectivity.
Attention is known to have a limited capacity to process information; not all incoming data can reach its focus; as for the exact length of this focus there is no common agreement: the focus of attention has been compared either to a spotlight, 16 a lantern, 17 or a landscape, 18 whose size could be adjusted at will, 19 thus presenting a truly irregular geographical span. When selectivity is poor, incoming information is coarse grained and the attended region is larger. This criterion helps, for instance, in distinguishing between distributed and focused (that is, selective) attention, two processes that differ with respect to the attended content, that is, to how many objects we are attending to, 20 and with respect to the resolution or granularity of the content attended to, that is, to what extent we can discriminate various entities composing this content. As we saw in the introduction, fine art was generally thought to elicit focused attention, selecting either the fine-grained properties of artistic compositions or art products as coherent wholes. 21 Secondly, duration or temporal span is a criterion for distinguishing sustained from transient forms of attention. Thus, endogenous attention, controlled by top-down processes, is employed at a late level of processing and can be sustained 'for as long as is needed to perform a task' . 22 Endogenous, sustained attention may be at work in aesthetic experiences defined in terms of a contemplative mode (broadly, the museum experience). On the other hand, exogenous attention, controlled by external stimuli, rises and decays quickly, peaking at early levels of processing and occurring 'even when the cues are known to be uninformative and irrelevant and when they impair performance' . 23 The pendulum-beat of a clock, for instance, may briefly arrest attention every now and then. Likewise, objects or entities that strike us immediately as beautiful Another characteristic of attention, which was widely considered to be its essential feature, is vividness; according to Titchener, 25 for instance, the primary role of attention was, allegedly, to enhance the clearness of sensory contents.
This feature was brought into question by subsequent studies in cognitive psychology: intensity, rather than vividness (which refers only to high intensity), would be a more appropriate feature, since more inclusive. It is established that attention comes in varying degrees, expressed at the subjective level by intensity variations. 26 We thus abandon the idea that vividness is an unquestioned attribute of attention. In this regard, attentional engagement with stimuli is presumably more or less active rather than an all-or-nothing response; one can pay more or less attention to an object or to a location and this variation of intensity is liable to make a phenomenological difference, which means that the activity of attending will modify our overall experience by reason of its force. It is on this particular point that I will dwell in my overall argument.
Finally, agency is the last property of attention that I would like to mention here, where agency refers to the quality of any activity initiated by an agent's mental states. Attention can be under the voluntary control of a mental agent when a stimulus becomes interesting in association with some goal, or, on the contrary, rise involuntarily, as a consequence of the stimulation itself, without relation to anything else. The concept of 'disinterested attention' 27 was particularly significant in the aesthetic debates of analytic philosophers of art, where disinterestedness referred to an attitude of total engagement with an object or configuration of forms, pursued for its own sake. 28 The typical characteristics of attention roughly sketched in this sectionnamely, selectivity, duration, intensity, and agency -will prove useful in establishing a provisional typology of attention phenomena, to which I shall now turn.

IV. NON-FOCAL TYPES OF ATTENTION AND OTHER RELATED PROCESSES
This section provides a typology of attentional phenomena and other related processes, assuming that they admit of some kind of ordering. With this, I explore the capacity of attention to take degrees and to be employed in varying intensity.
I will consider only attentional (related) processes that comprise low degrees of intensity, in the hope that they might reveal important aspects of the information processed at the periphery of attention and its potential aesthetic impact.

IV.1. PRE-ATTENTIVE PROCESSING
The first phenomenon that I consider is pre-attentive processing. This type of information processing is automatic, global, and holistic, producing the units to which attention may subsequently be directed in a more focused way; 29 it helps the perceiver to crudely structure his or her perceptual environment while going about his or her business. Pre-attentive processing has also been reframed in terms of anticipatory schemata 30 embedded in our cognitive systems (therefore beyond the reach of voluntary control), which allow taking on information relevant to our immediate perceptual actions while ignoring the rest.
Furthermore, pre-attentive processing is a preliminary stage to further processing but it is not yet a full-blown attentive process; we can be sensitive to information outside the current original focus of attention but, even though some features or global properties of the environment are detected in this preliminary stage, they have to be passed on to subsequent stages of processing in order to be identified as parts of 'fleshed out' perceptual objects; the selectivity and the clearness of sensory contents are thus kept to a minimum. For instance, an absentminded person can walk along a path without noticing the details and still not bump into obstacles that he or she may come across. Another characteristic is that it affects only the immediate present; it has a limited time span, and could hardly give rise to perceptual learning (that is, the capacity to distinguish progressively more fine-grained aspects of the perceptual environment). Finally, it appears that pre-attentive processing doesn't provide emotional content either 31 and this particular property will be significant when engaging in aesthetic questions.
Given these limited functions (such as crudely structuring one's environment, possibly guiding shifts of attention, short-term storage of information, lack of emotional content), it would be highly unlikely for an aesthetic response to be initiated at this stage without further information processing.

IV.2. THE MERE EXPOSURE EFFECT AND PERCEPTUAL FLUENCY
Nevertheless, information analysed without focal attention and not always accompanied by full awareness is not necessarily lost. Consider next the mere exposure effect, which refers to those situations in which repeated exposure to indeterminate stimuli (with low selectivity) generates enhanced affect ratings.
Repeated exposure to such stimuli helps to preserve them in long-term, implicit memory, thus preventing them from disappearing unnoticed for good. Like preattentive processing, the mere exposure effect is a passive automatic process (that is, requiring no subjective control), 'fleetingly conscious' (that is, with low intensity), and it can give rise to perceptual fluency or ease of processing. 32 This means that the perceptual encoding processes that involve previously encountered stimuli will be facilitated due to the effect of habituation or familiarity that is created through repeated exposure. Ease of perception, coupled with a certain complexity of the observed stimuli, which would prevent boredom, is thought to enhance preference for them but it is still a matter of debate whether the valence of the experience is positive of negative. 33 Unlike pre-attentive processing, the effect has been related to perceptual implicit learning, a process by which we acquire the ability to discriminate between different stimuli without being aware of doing so: for instance, the ability to detect pitch relations and regularities can be acquired through mere exposure to the musical system of a culture. 34 It is well known that when an object (such as a property, scene, or location) is integrated in the usual routine of performing certain acts, we barely pay attention to it, if at all. A good illustration of the effect of habit upon attention is given by James, who, talking about himself in the third person, remembers how, on revisiting Paris after ten years' absence, and, finding himself in the street in which for one winter he had attended school, he lost himself in a brown study, from which he was awakened by finding himself upon the stairs which led to the apartment in a house many streets away in which he had lived during that earlier time, and to which his steps from the school had then habitually led. 35 As this passage suggests, objects or scenes lurking in the background and falling below the threshold of our full awareness may subsequently reach the centre of focused attention either when they appear unexpectedly in a given context or when they altogether cease to be available, that is, when they cease to be part and parcel of our perceptual habits. 36 In other words, the transition from an automatic record of a set of stimuli to full awareness happens either when their unexpected presence disrupts our habitual, present actions, or when they are summoned up through some introspective, reflexive act (such as remembrance); James's story beautifully illustrates both ways to recover dimly attended information, although he says nothing about the affective valence that might accompany this act. Now, how is James's story connected to the mere exposure phenomenon? His example is relevant if we consider that when psychologists test the role of mere exposure in determining aesthetic response they in fact allow subjects to access dimly attended stimuli by presenting them out of their original context of appearance, more specifically in an experimental setting that disrupts the subject's perceptual habits. In other words, the investigation of the mere exposure effect is not disconnected from an actual perceptual awareness of the previously encountered stimuli, even if the subjects are blind to having been systematically exposed to such stimuli. Despite the possible limitations of such measurements, what interests me here is that the results revealed by the studies of the mere exposure effect provide empirical evidence for Munro's intuition that inconspicuous forms experienced regularly could have an aesthetic impact in the long run. As we have seen, mere exposure to stimuli may in some cases entail developing preferences for them. The effect has been observed both for mere exposure to representations of artworks belonging to the canon of art 37 and for stimuli we encounter in everyday life. If this effect is sufficiently robust, then even objects or scenes to which we only dimly pay attention could end up progressively building up aesthetic experiences (or at least aesthetic preferences).

Degrees of Attention in Experiencing Art
The problem with such a view is that, in everyday situations, we are exposed to a profusion of sensory impressions (the noises of the humming fridge, the traffic lights, and so forth); how would we know which ones are liable to give rise subsequently to aesthetic experiences? Unless there is a hidden import or value 38 that is susceptible, hypothetically at least, of becoming salient later, the vast majority of the impressions that the environment offers will be discarded. In studies of the mere exposure phenomena, psychologists have laid heavy stress on the idea that unattended information can leave its marks on consciousness 39 in general and influence affective judgements in particular. But then again, it must be borne in mind that in most of the cases examined it is not exposure as such that is investigated but a perceptual judgement (preference rates and so on) of a number of stimuli that a subject happens to have been systematically exposed to. In other words, what lies outside the focus of attention is not the perceptual stimulus as such, since rating a stimulus entails being aware of it, but the source of the stimulation (its repeatability, constant presence in an environment, and so forth). A possible consequence of all this is that the only way of gaining some insight into the aesthetic effect triggered at the periphery of attention is ultimately to bring into focus the original toned-down stimulations, thus altering the very nature of the phenomenon that we set out to investigate in the first place. The transition from the mere record of sensory stimulations to a minimal subjective awareness of the import of such stimulations might be a necessary step for determining whether an aesthetic effect obtains 'in the long run' .

IV.3. PSYCHIC OVERTONES
With the notions of fringe of consciousness and psychic overtone we move towards conscious phenomena that have a qualitative feel, where the subjective mental states are no longer missing from the picture. Unlike pre-attentive processes, fringe experience does contain an affective component. In this particular case, psychic overtone consists of a dim awareness (low intensity) of relations and objects, which we gain by mere acquaintance (that is, bare impression) and which comes with, to quote William James, a 'feeling of harmony or discord' , which accompanies our thoughts. 40 James describes overtones in music as follows: 'different instruments give the same note, but each in a different voice, because each gives more than that note, namely, various upper harmonics of it which differ from one instrument to another. They are not separately heard by the ear; they blend with the fundamental note, and suffuse it [low selectivity], and alter it.' 41 Using this musical comparison of the auditory perception of harmonics, which is always contextual, he goes on to call 'psychic overtone' , 'suffusion' , 'halo' , or 'fringe' 'the influence of a faint brain process upon our thought, as it makes it aware of relations and objects but dimly perceived ' . 42 He illustrates fringe experiences with an example from word comprehension in the process of uttering a phrase: No word in an understood sentence comes to consciousness as a mere noise. We feel its meaning as it passes; and although our object differs from one moment to another as to its verbal kernel or nucleus, yet it is similar throughout the entire segment of the stream. The same object is known everywhere, now from the point of view […] of this word, now from the point of view of that. 43 The same process holds true for a sequence of fugitive visual impressions: 'Illuminate a drawing by electric sparks separated by considerable intervals, and after the first, and often after the second and third spark, hardly anything will be recognized. But the confused image is held fast in memory; each successive illumination completes it; and so at last we attain to a clearer perception. ' 44 What James seems to describe here is the temporal dynamics of making a sensation or idea distinct (a phrase to be understood, a tone to be heard, an image to be recognized, and so forth) through a series of inward and outward activities: recollection and perceptual expectations as well as immediate perceptual experiences concur in giving shape to fringe experiences.
The reason for discussing these three mechanisms -namely, pre-attentive processing, mere exposure effect, and psychic overtones -is to give a sense of some dimensions of intensity along which attention phenomena might differ, with a particular focus on low or feeble degrees of intensity, in the hope that they might eventually prove useful in understanding the workings of the 'periphery of attention' and its relation to aesthetic appreciation. Of course, the proposed typology also presents difficulties to the extent that the phenomena discussed are not on the same level of description; for instance, it might be questionable to compare a functional account of attention (Neisser's account of pre-attentive processing) with a rather phenomenological one (James's psychic overtones). Despite its drawbacks, this provisional typology has at least the merit of pointing out that attentional processing takes place at 42 Ibid., 259. 43 Ibid., 281-82. 44 Ibid., 440-41. The explanation continues as follows: 'We hear a sound in which, from certain associations, we suspect a certain overtone; the next thing is to recall the overtone in memory; and finally we catch it in the sound we hear; […] the impression awakens the memory-image, which again more or less completely melts with the impression itself. In this way every idea takes a certain time to penetrate to the focus of consciousness. And during this time we always find in ourselves the peculiar feeling of attention.
[…] The phenomena show that an adaptation of attention to the impression takes place. ' various stages, that it is not an all-or-nothing affair, and, arguably, that this temporal dynamics might be relevant to revealing important aspects of appreciative response.

V. AESTHETIC APPRECIATION AND THE PERIPHERY OF ATTENTION
Building on the theoretical framework developed in previous sections, I now suggest that the periphery of attention be treated in line with the Jamesian approach to fringe experiences, as a phenomenon concerning the qualitative difference in the apprehension of certain stimuli (that is, dim or faint awareness), irrespective of their specific content. This is where the characteristics of attention sketched above -selectivity, duration, intensity, and agency -will come in handy.
To what extent do these characteristics manifest themselves at the periphery of attention? Bringing together the analyses of art theorists (Munro), art historians (Gombrich, Baxandall), and modern and contemporary psychologists (James, Reber, and so forth), which, I trust, complement each other, we could reasonably describe the periphery of attention as involving coarse-grained, vaguely sensed information; it is the exact opposite of the state of deep absorption and concentration usually associated with focused attention. Moreover, the periphery of attention is to be distinguished from multifocal, or distributed, attention, 45 which is mainly concerned with quantitative rather than qualitative questionsnamely, the amount of sensorial content (number of objects, properties, and so on) which can enter the focus of attention. For instance, when attending an opera, spectators may have to distribute their attention amongst a large number of properties and widen their field of interest in order to fully grasp the performance; they may of course ignore the heavy attentional demands (or the performance altogether), but the two phenomena of distributed attention and peripheral attention are different. Furthermore, whatever may happen at the periphery of attention is to be distinguished from covert attention in that it is not under the control of the observer nor does it have the same resolution or intensity: for instance, following Baxandall, we can deeply, covertly, attend to Braque's pitcher in the corner of our eyes without looking to obtain that trompe l'oeil effect.
Whereas the type of processing with which we are concerned here is momentary and does not require cognitive effort or subjective control in order to be maintained, it is more likely that it engages the early stages of perception during which only minimal operations, such as figure-ground segregation or detection of basic features (colour, orientation, and so forth), are taking place. This early phase, dealing with input from the periphery, has sometimes been called 'ambient processing' , 46 as opposed to a focal phase dealing with fine-grained information. I wish to advance the hypothesis that subdued forms sensed at the periphery of attention, coupled with the phenomena of repeated, diffuse exposure, are liable to provide the basis for generating an aesthetic response.
When a stimulus bearing a potentially hedonic value is perceived marginally and through repeated exposure, allowing information to be stored in long-term memory, it can give rise to aesthetic effects that we may eventually become aware of. In making these considerations, which, as they stand, are still highly speculative, I am mostly relying on intuitive examples of aesthetic episodes such as those singled out in Munro's conference paper. Again, the example of toneddown music comes to mind here, for instance, some main musical theme of a film, which gently accompanies the moving pictures and towards which we are pulled back only after the movie ends. I provide below some empirical grounding based on memory studies that could support these intuitive claims.
What I am suggesting is that in this type of encounter with non-intrusive stimulations, the aesthetic effect acquires its force only gradually, through repeated exposure to items held constantly in the background of attention; it is not a matter of a once-experienced event. The development of a definite aesthetic experience is thus deferred and aesthetic experience cannot be explained in terms of a momentary stimulation experienced through direct perception. This raises an interesting question about the temporal structure of aesthetic response: just how long does it take for an aesthetic response to be initiated and then to fade away? The temporal complications in this dynamic process are virtually absent from the conceptual apparatus developed within the field of aesthetic inquiry.
A tentative account, which is closest to the existing theories that I briefly discussed in the previous sections, would be to say that the aesthetic effect of stimuli experienced at the periphery of attention roughly resembles the perception of overtones described by James: in other words, we first experience some fugitive sensorial impressions that we cannot hear alone and then we recall them in memory, thus allowing them to reach the focus of this type of marginal stimulation. Thus, aesthetic response starts as a vague dispositional state and then becomes actualized through a subsequent act of remembering; it is not immediately felt, but is constructed in time by a series of endogenous acts that operate on the impressions we received at particular moments in time. A question that one may well ask is why would we need to appeal to memory retrieval in the first place when speaking of the aesthetic effect triggered at the periphery of attention? Why not stop at the moment of the original, crude stimulation? A reasonable reply to this is that just as we do not have immediate access to fringe experiences without altering their very nature, 48 so too the aesthetic disposition originating from sensory stimulations peripherally attended to would not be manifest to us, much less transparent. It is then difficult to know whether aesthetic experience can be merely dispositional, implying no subjective access whatsoever. What exactly is the process that takes place when evaluating inconspicuous stimulations? Are we merely reporting an aesthetic feeling felt in the past, experienced on the fringe of consciousness so to speak, of which we only become aware later through actual reminiscence or is the aesthetic response, rather than being elicited in the immediacy of every individual weak sensory stimulation, formed in the very late act of recall, thus appearing only in the memory mode?
The temporal complications described here may account for a particular kind of experience likely to become aesthetic, such as nostalgia, 49 which does not rely primarily on intensely perceived sensorial contents. In nostalgia, which can be triggered both by an external cue (like a postcard or a tune conjuring up past events) and by a voluntary act of remembrance, these sensorial contents reach the focus of attention only after the event remembered has taken place, they are not fully apprehended immediately. If this is a legitimate example, aesthetic experience could then be considered to extend beyond the primary sensory encounter; its content would outstrip the original perceptual experience; it would rely on memory traces of objects or scenes improving and becoming more precise over time rather than on deep immediate apprehension. Bence Nanay, who speaks of the lingering effect of some aesthetic experiences, also expresses this idea, although in a very different context: he mentions the continued presence of certain aesthetic experiences in our memory immediately after the object of contemplation ceases to be available, for instance, after leaving a concert hall or the cinema; 50 in the examples I have considered so far the time span of the lingering effect would be much wider and would not be caused by a powerful aesthetic encounter that would subsequently alter our perception of the world.
Another interesting example that illustrates the force of the lingering effect is provided by Svetlana Alpers in a powerful description of remembering the 9/11 attacks on the Twin Towers, which she saw from her loft windows in New York: Truth to tell, by the brilliant morning light, that majestic and silent disintegration was a beautiful sight to see. […] But there was horror mixed with it.
[…] the photographs I made on the roof that day don't match the image in the mind. The photographs show two tiny buildings and some smoke and then an empty place.
[…] From the visual point of view, there is little sense of loss. What I have is an image in the mind and now this account. 51 Alpers's narrative of this episode more than ten years after its unfolding points to important aspects of the active and constructive nature of memory as well as of the changing affective valence connected with particular sensory impressions that undergo transformation with distant recollection. As she rightly observes, photographs may fail to capture the sense of the past and most of the time the 'mental images' continuously revised in the act of remembering are more in accordance with the feel of a past event (for example, the 'feeling of loss'). However, in the cases I have in mind, the experience acquires a rich content only in the memory, not in the initial perceptual act.
Subdued forms of aesthetic stimulation are not supposed to create a powerful impression from the outset.
The trouble with this explanation is that in order to make it plausible and convincing, these inconspicuous stimulations have to be in our long-term memory store in the first place; they have to become available for further analysis.
We know that powerful perceptual events are very likely to be encoded in memory, but could there be memory storage of dimly attended perceptual information? From the perspective of the psychology of perception, 52 there is some evidence that support the idea of implicit learning with respect to poorly attended information; in the realm of aesthetics, however, there are few memory assessment studies that test the success of encoding and retrieving non-salient aesthetic stimuli. Repeated exposure should in principle facilitate implicit learning and one can understand why Munro mentioned it as an essential ingredient of aesthetic response elicited by marginally perceived items. Of course, Munro couldn't have referred to the mere exposure effect because studies on this topic were published later but his intuitions were nevertheless very insightful. Another hypothesis put forward by psychologists is that perceptual learning may be greater when poorly attended information is unobtrusive, because it is less liable to be suppressed by higher order mechanisms; 53 thus, rather than being parasitic upon more important activities, subdued, inconspicuous aesthetic stimuli may after all have a positive effect in the long run and enrich our perception of the world. An argument in favour of this idea comes from studies on liking and memory as a function of exposure. For instance, Szpunar et al. found that for incidental listening to musical stimuli ecologically valid memory ratings as well as liking ratings increased linearly with increased exposure. 54 On the other hand, the effect of exposure on liking was not produced for focused listening; in this case, the stimuli were recognized with increased accuracy over repeated exposures, eventually leading to satiation. The quality of the attending experience -incidental or focused -does seem therefore to have an effect on preference ratings and it is not always focused attention that leads to increases in affective response.

VI. CONCLUSION
In this paper I have addressed the question of aesthetic appreciation that does not rely primarily on focal attention and direct sense perception. Although I reject the idea that we can have an aesthetic experience with no awareness whatsoever, occurring at a pre-attentional level, I would keep open the hypothesis that in some circumstances the roots of aesthetic response might be found in stimulations taking place far from the centre of one's focused attention. I have argued that given certain constraints, such as repeated exposure, perceptual learning, encoding in long-term memory, and the possibility of retrieval, subdued, inconspicuous forms of stimulation can elicit aesthetic responses. Whether marginal perception and other phenomena of marginal sensory stimulation qualify as attentional processes is still controversial and further conceptual analysis should be made in this regard in order to refine these questions.