↵ select ↓ ↑ navigate esc close

Paper Feed: Attention

Paper Feed ·

Listen: https://serve.buildtall.systems/b11e72b8d78db40261490a36381b4372273a80483f9264c1cc5c9d0d0db24d6a.mp3

Source: https://serve.buildtall.systems/5caffa61e41e1294fa4bb82e1458ac59276579a9f50fcfc80b1b1fba25b1eb58.pdf


This episode covers Attention, an entry from the Stanford Encyclopedia of Philosophy that surveys how philosophers and psychologists have tried to explain attention. It traces the subject from early modern philosophy through twentieth-century psychology to current debates in consciousness, reference, knowledge, value and social theory.

A note on the record first: OpenAlex, the index of scholarly works, has no record of this document, so this episode makes no claim about how often it has been cited. The document is a survey rather than a study. It reports no new experiment and no dataset of its own, and its method is a historical and critical review of the literature. The passage headings show it moving from Descartes, Berkeley and Locke through the bottleneck theories that followed Donald Broadbent, then to feature integration, coherence, precision and competition theories, spotlight and motor theories, and finally to attention's roles in consciousness, demonstrative reference, other minds, knowledge, voluntary action, aesthetics, ethics and the attention economy. Its strength is breadth and a fair hearing for rival positions, supported by quotations from primary sources. Its limits are ones it states itself: it describes many debates as unresolved, and it says that experiments such as inattentional blindness and change blindness cannot settle the questions they were used to address on their own. Neither the front matter nor the passages state any funding or data disclosures, so none can be reported here. The copy carries a print header dated 6/19/23, and it presents itself as an entry from the Stanford Encyclopedia of Philosophy.

In the early modern period, a number of writers gave attention a range of explanatory jobs to do. The entry's first example is Descartes' Meditations. Descartes faced what the entry calls an "apparent tension": his First Meditation concludes that everything can be doubted, yet his Third Meditation claims that clear and distinct ideas are beyond doubt. Clear and distinct ideas were Descartes' name for the thoughts that, when grasped properly, seem to leave no room for error. To resolve the tension, he introduced a claim about attention. In his reply to the seventh set of objections, he says that clear and distinct ideas provide a place where doubt does not take hold only when we are paying attention to them. In his words:

So long as we attend to a truth which we perceive very clearly, we cannot doubt it. But when, as often happens, we are not attending to any truth in this way, then even though we remember that we have previously perceived many things clearly, nevertheless there will be nothing which we may not justly doubt so long as we do not know that whatever we clearly perceive is true.

The entry notes that this passage is usually cited for what it says about memory. Descartes is saying that remembering past clear perceptions is not enough to keep doubt at bay. Without a general rule that whatever is clearly perceived is true, a thinker who is not currently attending to a truth can still be justly doubtful about it. The entry argues that the same passage also gives attention an epistemic role. Clarity and distinctness only realize their potential for knowledge when attention is directed at the ideas that have them.

The entry draws out the consequence. Ideas that are clear and distinct can still be doubted, as the policy of the First Meditation requires, but a thinker who is properly attending to them cannot sustain that doubt. On this reading, the crucial first move in Descartes' epistemology, the move from radical doubt, which is the method of questioning everything to find what cannot be questioned, to certainty about particular clear and distinct ideas, is a transition mediated by attention.

Berkeley assigns attention a quite different job. Like Locke before him, he was trying to solve an epistemological problem, and attention enters the picture as the tool for solving it. In the Introduction to his Principles of Human Knowledge, Berkeley rejects Locke's claim that there exist such things as abstract ideas. But he keeps Locke's core empiricist commitment, that thinking is always a matter of handling ideas received from experience. Taken together, those two commitments seem to lead to an unacceptable conclusion: that it is impossible to think about abstracta, the general features that many particular things share. Berkeley does not accept that conclusion. He holds that it is perfectly possible to think about the properties of triangles in general.

The second edition of the Principles adds sentences to the Introduction that make the role of attention explicit. According to them, it is attention, and in particular the withholding of attention, that explains how we can think about abstracta without positing abstract ideas. The added text reads:

[It] must be acknowledged that a man may consider a figure merely as triangular, without attending to the particular qualities of the angles or relations of the sides. So far he may abstract, but this will never prove that he can frame an abstract general, inconsistent idea of a triangle.

Berkeley does not develop this into a full theory of attention. He says nothing more about how attending to some features and withholding attention from others might make abstract thought possible. Even so, the added passage makes clear that he expects attention to play an important part in his picture of the mind.

The link between attention and abstraction was later taken up in the second half of the nineteenth century by William Hamilton. Hamilton did not treat the connection as a starting point for explaining either attention or abstraction. In his view the two are too intimately related to explain one another. In a passage from 1876 he writes:

Attention and Abstraction are only the same process viewed in different relations. They are, as it were, the positive and negative poles of the same act.

Locke gives attention a positive account, which Descartes and Berkeley did not. The entry places it in a catalogue of what Locke calls modes of thinking, set out near the start of chapter nineteen of book two of his Essay Concerning Human Understanding. Locke moves quickly through a series of mental states. The first is a state in which ideas "float in our mind without any reflection or regard of the understanding", which the French call reverie. Next comes attention. Here the ideas that present themselves are "taken notice of, and, as it were, registered in the memory", and that registration, Locke says, "it is attention". The last state is the one where the mind "with great earnestness, and of choice, fixes its view on any idea", considers it "on all sides", and will not be pulled away by other ideas. Locke calls that intention or study.

The same sentence also covers remembrance, recollection, contemplation, sleep, dreaming and ecstasy. The entry points out how little room Locke gives attention, and that it sits among these other states, because Locke is not improvising a hasty theory. He is saying these phenomena need no new substantive theory. In his view reverie, study and the rest are simply "modes of thinking", the names thinking takes when it happens in different ways, not independent phenomena in their own right.

The entry illustrates this with the classic example of modes: once you have a theory of walking, you need no separate theory to explain limping, pacing or ambling. It draws out two consequences of Locke's view. Once a theory of thinking is in place, no very substantive theory of attention is needed, and no substances or processes specific to attention have to be postulated. And attention cannot figure in explaining how thinking is possible at all, since that would reverse the order of explanation, like explaining walking by assuming that strolling is already possible.

Locke saw attention as explanatorily slight, but over the eighteenth century that view gave way. Attention came to be treated as a phenomenon with explanatory work to do, and so one that needed its own independent theory. The entry dates the start of that effort to 1738, when Christian Wolff's psychology textbook was the first to devote a whole chapter to attention. The explanatory scope then widened in two directions.

First, attention moved from acting on ideas already received to being involved in their initial reception. By 1769 Henry Home Kames, in an appendix to his Elements of Criticism, treated attention's role in regulating what comes into the mind as defining it:

Attention is that state of mind which prepares one to receive impressions.

Kames added that attention is required even for the simple act of seeing.

Second, theories began giving attention a role in producing behaviour. Dugald Stewart's 1792 Elements of the Philosophy of the Human Mind keeps Locke's view that attention decides what is stored in memory, and adds that it also decides which memories get recalled, writing that

Some attention [is] necessary for any act of memory whatever.

Stewart also gave attention a part in skilled action, pointing to "the dexterity of jugglers", which he said

merits a greater degree of attention from philosophers than it has yet attracted.

By the end of the eighteenth century, then, attention was seen as needing its own explanation, with a role in perception, in skilled action and in memory. The entry says that the range of phenomena attention was expected to explain kept growing through the next century, from perception to thought to action. By the end of the nineteenth century, as scientific psychology was taking shape, three camps had emerged. E.B. Titchener took attention's role in perception and in "sensory clearness" to be its essential feature. Alexander Bain located its essence in action. G.F. Stout argued that the main job of a theory of attention was to explain attention's role in reflective thought.

Because these psychologists disagreed about what attention was supposed to explain, and because empirical psychology had no settled methodology, the debates between them fell into what the entry describes as a state acknowledged to be "chaotic", a phrase it credits to Pillsbury in 1906.

The end of the nineteenth century inherited a mess. The range of jobs assigned to attention in the eighteenth and nineteenth centuries had left theorists with no agreed picture of what attention is. Writers in the late 1800s and early 1900s set out to bring order to this, and their typical move was to show that attention is less mysterious and less complex than earlier accounts had made it. The clearest example is William James's The Principles of Psychology, published in 1890, which tries to reduce attention to something simpler.

One part of James's strategy is to downplay attention's role in complex perception. His chapter on attention discusses experiments on what is now called subitizing, the ability to tell how many items are in a small set without counting them. James dismisses these experiments, writing that they "is obvious that such observations decide nothing at all about our attention, properly so called" (407).

James does give attention an important job in producing behaviour. He writes that "Volition is nothing but attention" (424). But the text makes clear that this claim serves a deflationary project. James links attention to volition as a way of showing how volition could be treated in simple terms, not as a way of enlarging attention's explanatory reach.

This approach shapes James's account of what he calls the "intimate nature of the attention process". He names two fairly simple processes that, he says, "probably coexist in all our concrete attentive acts", and that "possibly form in combination a complete reply" to the question (411). The first is the accommodation or adjustment of the sense organs, such as turning your ears toward a sound, bringing your eyes into focus, or taking a sniff. The second is the anticipatory preparation from within of the ideational centres concerned with the object attended to. In plain terms, this second process is imagination. When attention does not involve moving the sense organs, James holds, it consists in imagining the thing or action one is attending to, or looking for.

James supports this with an example from Hermann von Helmholtz, which the text presents as important for showing features of attention that later theorists tended to neglect. Helmholtz asks the reader to sit at a piano, play a G, and then, imagining the sound just heard, play a low C. Doing this, he claims, lets one hear that the G is audibly present as the third overtone within the sound of the C. The attention needed to hear that overtone, on Helmholtz's account, consists in imagining what the overtone would sound like. James extends the point, arguing that many cases of paying attention to what one is doing consist in this same kind of preparatory imaginative engagement.

The text notes that this is typical of James's method. In his treatment of emotion as well as attention, he tries to explain a large personal-level psychological phenomenon in a realist but somewhat revisionary way, using relatively simple and unmysterious resources. He contrasts his approach with a different deflationary one, that of F.H. Bradley in 1886.

Bradley held that attention is not something for which an independent and substantive theory can or needs to be given. The text says Bradley did not develop this in much detail, and that he would later change his mind. In his 1886 article

Is there a special activity of attention?

, he argued that James's project of picking out particular processes as the ones that constitute attention was misguided. His reasoning was that no particular attention-processes can be identified:

Any function whatever of the body or the mind will be active attention if it is prompted by an interest and brings about the result of our engrossment with its product. There is no primary act of attention, there is no specific act of attention, there is no one kind of act of attention at all.

(316)

The text observes that Bradley does not use Locke's vocabulary, and that James himself does not seem to have read Bradley this way. Even so, Bradley's position has much in common with Locke's claim that attention is a mode. For both, what makes something an instance of attention is not which processes are taking place, but the facts about how those things happen. So listing processes is, for Bradley, the wrong form for a theory of attention to take.

Other writers of the same period took still different routes to a deflationary account. In his 1888 book La Psychologie De L'Attention, Théodule Ribot explained attention in what would now be called a behaviourist way. Behaviourism was not yet an established general approach in 1888, but Ribot's view that attention's behavioural manifestations are essential to it reads as an early and strong statement of behaviourism. He asks whether the movements of the face, body and limbs, and the changes in breathing that accompany attention, are simply effects and outward marks, as is usually assumed, or whether they are instead the necessary conditions, constituent elements and indispensable factors of attention. His answer:

Without hesitation we accept the second thesis.

(19)

Alexander Bain took a more moderate version of this behaviour-centred view. He did not identify attention with the behaviour itself, but with truncated versions of the motor-control processes that normally produce it, processes "stopping short of the actual movement performed by the organ" (Bain, 1888, 371). If Ribot's view is an early form of behaviourism, Bain's is an early form of the motor-based theories of attention that the entry takes up in Section 2.8.

The twentieth century opened with a period in which one deflationary tactic dominated the study of the mind: behaviourism. Behaviourists tended to neglect attention, but they did not ignore it entirely. John Dashiell's 1928 Fundamentals of Objective Psychology, for example, is a behaviourist work that attempts to account for attention

as a form of posturing.

The project of identifying a behaviour that could explain attention was, however, an unpopular one, and the entry explains why. Gilbert Ryle observed that attention, and what he calls heed concepts more generally, resist simple behaviourist analysis. When a man is described as driving carefully or eating absentmindedly, Ryle writes, "the special character of his activity seem to elude the observer, the camera and the Dictaphone." Knitted brows, taciturnity and a fixed gaze may be signs of intentness, but, in Ryle's words, "these can be simulated, or they can be purely habitual."

In the middle of the twentieth century behaviourism's dominance waned, cognitive psychology established itself, and a new theoretical approach to explaining attention took shape. The three developments were closely linked, and the entry treats the publication in 1958 of Donald Broadbent's Perception and Communication as central to all three. The year before, B.F. Skinner had tried, in Verbal Behaviour, to apply a behaviourist approach to distinctively human aspects of cognition. That project failed, and Noam Chomsky's review of the book in 1959 made the failure plain. Chomsky's own Syntactic Structures, published in 1957, helped establish the new paradigm by showing how internal processing could be theorized by describing transformations on representations, without asking how those representations were physically realized.

Broadbent's distinctive contribution was to show how the move from behavioural data to the postulation of a particular cognitive architecture could be disciplined by borrowing the intellectual resources used in thinking about information technology. The entry notes that 1958 was an important year for those technologies, since the integrated circuit was invented that year, and Subscriber Trunk Dialling was introduced to UK telephone exchanges. The telephone exchange was the image that most naturally suggested itself as a metaphor for attention at the time.

Near the end of Perception and Communication, Broadbent sets out the claim that the theory of information transmission through telephone exchanges can provide the basis for an alternative to behaviourism, and he attacks the positivist methodology that had motivated many behaviourists. The entry argues, though, that the central lesson for the theory of attention does not depend on that attack. Early in the book Broadbent writes that

perhaps the point of permanent value which will remain in psychology if the fashion for communication theory wanes, will be the emphasis on problems of capacity.

He adds that "the fact that any given channel has a limit is a matter of central importance to communication engineers, and it is correspondingly forced on the attention of psychologists who use their terms." The introduction of capacity limitations into discussions of perception and attention was, as he predicted, hugely and permanently influential.

Broadbent claimed more than that the human brain has capacity limits of the kind communications engineers had learnt to theorize. He also claimed that these limits are clustered, so that a single bottleneck in capacity is especially critical to the brain's handling of perceptual information. The bottleneck sits at the junction of two systems operating in series. The first has a large capacity and operates automatically on all the stimuli the perceiver is presented with. The second has a much smaller capacity and so has to be deployed selectively. Those who followed Broadbent took the bottleneck created by joining these two systems to correspond to attention: when a representation of a stimulus passes through the bottleneck, the stimulus counts, just by that fact, as one to which attention has been paid.

Broadbent himself was cautious about presenting his bottleneck claims as a theory of attention. The word 'attention' occurs rarely in Perception and Communication. His later book, Decision and Stress, published in 1971, describes his earlier experiments as studies of attention, but there too he prefers to speak of selective perception or vigilance. In a 1982 article, 'Task Combination and Selective Intake of Information', he concedes that

the topic of this paper is one that is often termed 'attention', and it may seem unduly artificial to have given it a more cumbrous title.

He goes on to restate his reservations. 'Attention' is a word of ordinary language, he says, that can reasonably label a area of experiments, but it has also been used as "a mysterious asset or energy which is sometimes attached to human functions and sometimes not." Avoiding the word, he concludes, is a step towards clarity.

When Broadbent does use the word, it is mostly in discussions of attention shifting. The entry reads his early view as holding that wherever there is a bottleneck in information-processing capacity, additional mechanisms are needed to control how the limited resources are deployed. Broadbent seems to have thought of these bottleneck-control mechanisms as the attention mechanisms. He never claimed to have given a theory of them, only a theory of where they would be needed. Nor were these mechanisms the subject of the debates Broadbent set off. Those debates concerned the nature and location of the bottleneck itself, not the factors that decide what, on a given occasion, gets through it.

In the decades that followed, many psychologists set out to locate the attentional bottleneck he had proposed. Nearly all of them were guided to some degree by his picture of two serial systems joined by a bottleneck. On that picture, whether a task demands attention depends on whether performing it requires the small-capacity system that lies past the bottleneck. Research into the attention demands of particular tasks therefore became another route to finding where the bottleneck lies. The two-systems-and-a-bottleneck model was often questioned, but for most attention research in the second half of the twentieth century it was very much the orthodox view. The entry observes that this has since changed, and that in psychology's current paradigm only a few aspects of Broadbent's picture remain orthodox.

Through the nineteen sixties and seventies, psychologists building a theory of attention were strongly shaped by Donald Broadbent’s picture of attention as a bottleneck. In that picture, attention corresponds to a bottleneck in information-processing capacity, created because two separate perceptual processing systems are connected. The first task for these researchers was to find where the bottleneck sat: which kinds of processing happen in the large-capacity system before it, and which happen in the small-capacity system after it. The disputes over that question produced several theories, each defined by a claim about where the bottleneck is located.

The first is early selection theory, which is Broadbent’s own account. He held that only very simple properties are picked up by the large-capacity system. Anything semantic, or anything about the particular identity of a stimulus, is detected only after the stimulus has passed through the bottleneck. The personal-level consequence is that we can recognize what things are and what they mean only when we attend to them, while we can detect their simple physical properties even when we are not attending. The entry describes this as a communications-theoretic version of two intuitive points that Broadbent’s research had put on an empirically sound footing. The first is that "one has no immediate control over one’s awareness of simple features of one’s environment, such as the fact that there are people talking in the next room." The second is that the details of things, such as the content of that chatter, can be detected only for the one or two things we attend to.

The entry notes a more problematic consequence. If early selection is right, the semantic properties of an unattended item must remain unrepresented in the nervous system, so they could have no psychological effects. On this view, the meaning of an unattended item could not explain why that item attracts attention to itself on the occasions when it does. Opponents of the theory objected most often to exactly this point.

The chief rivals of early selection were the late selectionists. They claimed that all, or almost all, perceivable properties are detected automatically by a large-capacity system that operates on every stimulus a person is presented with. On this late selection view, passing through the bottleneck does only two things: the subject becomes conscious of contents the large system has already encoded, and those contents are stored in working memory. The entry says this theory should not be dismissed as obsolete. Jesse Prinz shares the late selectionists’ view that attention’s primary role is not managing limited perceptual resources but projecting already-processed representations into working memory. Drawing on the work of Stanislas Dehaene, Prinz holds that this projection is what makes represented content come to conscious awareness. The entry adds that critics have asked whether this projection can account for everything attention does, and that Prinz accepts that other researchers may define attention differently. In his words, "‘other researchers may choose to define attention differently’."

The central claim of the original late selection theory was that withdrawing attention from a stimulus makes it processed without the subject’s awareness, rather than not processed at all. The entry says this has partly been vindicated. It is now uncontroversial that unattended stimuli undergo some processing the subject lacks awareness of, including processing of their semantic properties. Negative priming shows this: the meanings of unattended items produce effects even though the experimental participants are typically unaware of what those meanings are.

The entry argues that the traditional late selection theory went too far in a further respect. It held that attention affects only what gets remembered and experienced, and that the direction of attention has no effect on initial perceptual processing. The entry says this is false. A study by O’Connor and colleagues used functional brain imaging to compare neural activity while participants saw high and low contrast checkerboard patterns in one or the other half of the visual field. In some conditions the patterns mattered to the task, in others they were irrelevant, and in a third no pattern appeared while participants watched the screen in anticipation. The results showed differences in baseline neural activity, and in the response to stimuli, even in the lateral geniculate nuclei, which lie early on the route from the retina toward the visual cortex. Attention’s effects, then, are not confined to cortical stages that come after an initial encoding. The entry concludes that this refutes any version of late selection on which selectivity is entirely a late process.

Nilli Lavie’s load theory keeps some late selection ideas while accepting the bottleneck. According to this theory, attention does correspond to a capacity bottleneck, but the demands of the current task determine where that bottleneck acts. When perceptual demands are high, resources are soon used up, the bottleneck operates early, and unattended stimuli are processed very little. When perceptual load is lower, the bottleneck operates later, and unattended stimuli are processed more. The theory predicts that peripheral stimuli will produce less neural activation and be less distracting during perceptually demanding tasks, and the entry reports that behavioural and neurological observations have borne these predictions out. Many psychologists remain wary, however, of Lavie’s return to elements of the Broadbentian framework.

Other capacity-limitation theories keep Broadbent’s idea that selectivity results from limited capacity but drop the single bottleneck. Some propose multiple bottlenecks in processing capacity. Others see capacity limits operating throughout the whole processing stream, not clustered into bottlenecks at all.

By the beginning of the nineteen nineties, the entry explains, the debate between early and late selection had come to look fruitless, which suggested it rested on some mistaken assumption. Several diagnoses have been offered, and

little consensus has been reached as to which of these diagnoses is correct.

One diagnosis is that the words ‘early’ and ‘late’ are themselves problematic. In a parallel architecture with no prevailing direction of information flow, there is no sense in calling one stage earlier than another.

The most influential version of this diagnosis is Alan Allport’s, which locates the problem in an assumption about linearity. Allport characterizes the problematic assumption as the claim that

[T]he processing of nonsemantic attributes (i.e. the processing of attributes other than symbolic or categorical identity) occurs earlier in a logical/causal sequence of operations than does any semantic or categorical processing.

The entry argues that this assumption does not look problematic on its face. When we read a written word, our sensory system responds first to simple properties of the lines on the page, and that information passes on to stages where more complex, meaningful properties are processed. Neuroscientists still routinely call occipital processing early and frontal processing late. Sabine Kastner and Leslie Ungerleider describe "an increase in the complexity of processing as activity proceeds anteriorly through the ventral stream into the temporal lobe."

Hierarchical Bayesian theories of perception break from the older picture of one-way flow, placing weight on top-down predictions that are tested against incoming error signals. But they still keep earlier and later stages, understood as more or less abstract layers of inference. So the entry concludes that we cannot coherently blame this degree of linearity for the collapse of the early and late debate.

The entry suggests the real problem lay in an additional assumption. A communications engineer, concerned with limited channels, would naturally suppose that once the physical properties of a thing have been represented, they need not be represented again for the person to become aware of its meaning. The brain, however, represents physical properties in multiple parallel and overlapping systems, and only some of these put the subject in a position to think about the properties they represent. Access to a property may require that it be represented all over again in a parallel system. The entry calls this strangely profligate from Broadbent’s point of view, and it says that the problems with the debate come from an assumption about the linearity of the processes supporting personal-level thinking.

Even so, the entry holds that the collapse of the debate cannot be credited entirely to this assumption. The falsity of the assumption undermines no more than the methodology of the debate. One cannot infer from a lack of personal-level awareness of some content that no representation of that content exists, and Broadbent was wrong to suppose that the psychological effects of unattended stimuli depend only on properties the inattentive perceiver can be aware of. Yet these are methodological problems, and they do not by themselves show that the debate was misconceived.

The entry sums up the situation as unhappy. Everybody agrees that something was misconceived about the debate that dominated the decades after Broadbent reintroduced attention to the psychological agenda. Broadbent’s influence on current thinking has often been noted, and sometimes lamented. As the entry puts it, "nobody is clear about whether the elements of those theories that are retained in current theorizing are problematic ones."

In the early nineteen eighties, Anne Treisman and her collaborators identified what became known as the binding problem, and described a process that could solve it. Treisman proposed that attention be identified with that process. This proposal is known as the Feature Integration Theory of attention. The entry notes that the theory has been hugely influential, both as a theory of attention and as the framework that introduced and organized research into the binding problem itself.

To show what the problem is, the entry starts with Treisman's own description.

Sensory information arrives in parallel as a variety of heterogeneous hints, (shapes, colors, motions, smells and sounds) encoded in partly modular systems.

Modular here means that different kinds of information are handled by separate, specialized parts of the system. Because typically many objects are present at once, the system must "collect the hints, bind them into the right spatial and temporal bundles, and then interpret them to specify their real world origins."

The entry makes this concrete with a red tomato and a green apple seen together. The visual input is broken up for specialized processing. One set of brain areas detects shapes and another detects colours. The shape areas register that a tomato-shaped thing and an apple-shaped thing are present. The colour areas register that a red thing and a green thing are present. A typical perceiver knows more than those four facts, though. He knows which red thing is the tomato and which green thing is the apple. Binding is the problem of linking properties detected in separate centres to the right objects.

Treisman's proposed solution rests on two points. The first is that spatial representations are ubiquitous. The second is that there is only one thing to be seen in any one place at any one time. Given those, if the colour centre detects red at place one and green at place two, and the shape centre detects a tomato shape at place one and an apple shape at place two, the binding problem can be solved. Treisman proposes that the system reaches a correctly bound representation by moving a variously sized "window" from one location in the scene to another. The window blocks out information from all but one location, and the features found there can then be bound as belonging to the same object. The area inside the window is taken to correspond to the region to which attention is being paid.

The entry points out that there is more than one way to read the explanatory claim here. In Treisman's early work, the window of attention is presented as identifying the processes of attention itself, so that attention is feature integration. In her later work, the explanation runs differently. Solving the binding problem is one role for the kind of selectivity that attention makes possible.

Some later work has pressed the theory into more ambitious service. The entry notes that John Campbell, in his 2002 book Reference and Consciousness, relies on it in an attempt to explain how the reference of demonstrative expressions gets fixed by their producers and understood by their consumers. Treisman herself suggests, tentatively, that descendants of the theory may provide part of the explanation for

the bound, unitary, interpreted, personal view of the world of subjective experience.

She goes on to suggest, again tentatively, that such an explanation

should give us all the information there is about the conditions that create consciousness.

The entry then turns to the opposing view. A number of philosophers, and a smaller number of psychologists, deny that feature binding creates a problem needing any serious cognitive apparatus to solve. They make this case for a variety of reasons, and generally as part of large-scale proposals to revise the conceptual framework of cognitive neuroscience. Kevin O'Regan and Alva No\u00eb reject the binding problem as "a pseudo-problem", as part of a broader attack on the idea that perception requires representing the thing perceived. Vincent Di Lollo argues that the assumptions about neural coding that generate the binding problem have been superseded, so that the problem is ill-defined. M.R. Bennett and P.M.S. Hacker, in a book-length critique of the philosophical foundations of neuroscience, argue that the notions of representation and information in use in neuroscience are subject to confusions. In their account these confusions make it "widespread" for neuroscientists to make

confused statements of the so-called binding problem.

The entry observes that these debates now turn on foundational issues for the cognitive sciences as a whole, and have ceased to be specifically concerned with explaining attention.

Even if these critics are right that a perceiver can typically experience several properties as belonging to a single object without any special binding process, attention may still determine how different streams of sensory information come to be combined in some particular cases. Even there, the entry notes, characterizing that binding-like role raises its own difficulties. Psychologists studying the integration of information across sensory modalities have described the role of attention as "curious" and as "a paradox", especially where speech perception is involved. Philosophers who have questioned the general binding problem have not yet extended that critique to these narrower cases. A preliminary philosophical examination by Mole suggests that the appearance of paradox may come from questionable assumptions about the predominantly bottom-up nature of perceptual processing, or from metaphysical assumptions about the localizable way in which attention's causal influence is exerted.

The section turns to what is called the coherence view, which pushes back on the idea that attention exists mainly to manage a limited capacity for processing information. That capacity-limitation view has been the standard account among psychologists since Donald Broadbent proposed it in the 1950s. The paper notes that philosophers have tended not to endorse any straightforward version of it, and that there are empirical reasons to think the simplest versions are mistaken.

The first piece of evidence comes from the 1970s, when Ulric Neisser and his collaborators ran a series of experiments. They showed that, once trained, a subject can carry out two attention-involving tasks at the same time without much interference between them. Neisser read this as meaning that where attention is needed to manage a bottleneck, the bottleneck is often one of behavioural coordination rather than information-processing capacity. The paper's examples of such bodily limits are simple ones: we cannot look in two directions at once, and we cannot throw a ball and write with the same hand. On Neisser's view, cognitive processing becomes selective mainly because the body requires it to be.

In 1987, Odmar Neumann and Alan Allport published two papers in the same volume that developed a similar idea. Where Neisser stressed what a single body imposes on cognition, Neumann and Allport stressed what is imposed by the need to keep action coherent and directed at a goal. They called their position a "selection-for-action" theory.

If action is taken to mean only bodily behaviour, this view is close to Neisser's. The paper argues that the point keeps its force when action is extended to deliberate mental activity, such as solving a puzzle, where no bodily limit is involved. It cites G. F. Stout, who drew the same parallel in 1891. Stout wrote that

Thinking is action directed towards intellectual ends. Intellectual ends are attained by an appropriate combination of movements of attention just as practical ends are attained by an appropriate combination of movements of the body.

He concluded that explaining thinking requires a clear account of active attention.

The paper's broader point is that in deliberate thought and action, we almost always have to choose among several possible sources of information, and then among several things to do with what we have chosen. Wayne Wu has returned to this need in a series of works, calling it the "many many problem." In Wu's hands the idea becomes part of a larger effort to understand the connections between attention and agency. For Wu, attention's role in action is more basic than its role in perception. Even the involuntary capture of attention by a striking stimulus, on his account, involves a readiness to act on whatever we involuntarily attend to. Wu's aim is to avoid drawing an artificial line between perceptual attention and the selectivity that agency demands, and to move toward a single account of ourselves as both agents and perceivers.

On these selection-for-action views, the function of attention is not to manage a shortage of capacity. It is to manage an excess of it. We can be distracted by many stimuli because we are able to process many of them, and the mechanisms of attention provide the selectivity and focus that this excess makes necessary. Attention, on this account, gives coherence to the activity of several components of a system that can handle far more input than the current task needs, and keeps those components from interfering with one another. The paper presents this as a sharp contrast with Broadbent's picture of attention as the management of a capacity bottleneck.

The tradition perpetuated by Wu understands the selectivity of attention as providing a solution to one particular problem. Several courses of action may be open at any time, and taking any one of them requires the others to be suppressed. The section points out that solving this problem is not the only cognitive need that selectivity could serve, even in a mind with no bottleneck of processing capacity, meaning no limit on how much information can be handled at once. So the selection-for-action account is not the only alternative to Broadbent's bottleneck theory. Other functions for attentional selectivity might also be proposed.

The section then offers a different option. Rather than choosing one function as the one attention essentially serves, we might characterize attention as "a quite general process of selectivity." On this view, attention could be called on whenever one thing needs to be prioritized over another. That might be because a processing bottleneck demands it, because a coherent action plan has to be carried out, or because of some quite different reason. The work of Sebastian Watzl develops the idea that attention might be such a non-specific process of prioritization.

In Watzl's treatment, the idea rests on an analogy. Just as a binary "is greater than" relation can order a set of items, a number of these pairwise prioritizations can be considered together. Taken in sum, they give the several items that make up one's mental life a prioritization structure. For Watzl, attention is the process by which one ’s life comes to have such a structure.

Watzl also argues that a mind lacking such a structure would lack a particular perspective on the world, and so would not have the subjective point of view that characterizes conscious experience. From this he derives the claim that every conscious creature is attentive. The section notes that Carolyn Jennings has suggested that certain kinds of highly engaged expert performance may be a counterexample to this claim.

A different line of argument holds that there can be reasons for attentional selection that have nothing to do with processing bottlenecks. Psychologists who hope to understand perception as Bayesian inference have pursued this idea, as have philosophers influenced by them.

The most philosophically developed version of this Bayesian approach treats cognition as a whole as Bayesian updating. A hierarchy of hypotheses is constantly being tested, and each hypothesis is updated in light of evidence from the level below it. Its advocates try to fit perception, thought and action into this one framework. A central claim is that the information passed up the hierarchy is encoded as signals that represent the errors in the predictions made by the hypotheses one level higher. On this view, the content of experience at any moment is given by whichever hypothesis makes the least erroneous predictions. Different versions of the theory use different techniques to calculate the relative size of these prediction errors.

Because the theory codes information as prediction error, it gives the job of the senses a different description. The entry puts the point this way:

instead of providing us with the information that it has started to rain, our senses provide only the information that it is more rainy than we had expected.

The advocates of this theory claim that it

allows us to see attention in a new light and to provide alternative conceptualizations of its functional role in our overall mental economy.

On these alternative conceptions, attention adjusts the weighting of incoming prediction-error signals according to their expected precision. The entry takes care to separate precision from accuracy. Accuracy measures the difference between the value a signal indicates and the actual value. Precision measures the random fluctuation in a signal even when the actual input is held constant. A miscalibrated instrument might therefore be highly precise without being especially accurate.

The entry says that several empirical results make it plausible that the brain takes account of expected precisions when it processes perceptual signals. In the ventriloquist illusion, for example, sounds are heard to come from the place where their source is apparently seen, because vision is expected to give a more precise indication of location than hearing. Another case concerns changing context. If vision is expected to be noisy because a thick fog has come down, we may place more weight than usual on the signal from hearing. The entry suggests that attention plausibly plays a role in bringing about this change in weighting.

The stronger claim is more controversial. The advocates of this theory say that

attention is nothing but precision optimization in hierarchical inference.

The entry reports that Ransom and colleagues have argued that, if attention is always taken to be precision optimization, the theory struggles to account for certain forms of voluntary attention. Clark has suggested that this challenge can be met if the sources of voluntary attention are identified with beliefs rather than desires.

Earlier, the episode touched on the view that attention exists to manage limits on how much the mind can process. The authors show that other accounts have been offered. Hohwy and Clark, working within prediction-error coding, and Neisser, Neumann, Allport, and Wu, in the coherence theories, have each suggested functions for attention other than managing limited processing capacity. Even so, the article reports that the older idea remains widely held. Many psychologists still treat the claim that attention's selectivity exists because the visual system cannot process everything at once as beyond dispute. The article quotes a 2000 paper by Vecera as an example:

Because the visual system does not have the capacity to process all inputs simultaneously, there must exist attentional processes that help the visual system select some inputs.

The article says sentences like this are treated as a platitude, a safe opening line for a research paper, and that many more examples can easily be found. For theorists who reject that platitude and think attention serves some other function, the natural next question is whether the mechanisms of selection might also be something other than capacity bottlenecks. The article's answer is that the clearest such mechanisms are competitions. Its reasoning is that “well-organized competitions can always select one winner, however good and however numerous the competitors”, so a competitive mechanism's selectivity need not depend on any bottleneck or other limit on processing capacity. The article describes two kinds of competition. In a simple race, each competitor completes a process that can be compared with the others along some dimension, and the one with the highest value wins. In a struggle, competitors actively suppress one another as part of their own processing. Simple race models of attention have been proposed by Shibuya and Bundesen, and by Bundesen alone. The article says the best current theories add some elements of mutual struggle, or top-down control, to the simple race. One such supplemented model is the biased-competition model, developed in several works by Desimone, Duncan, and Reynolds. On this model, many attention effects, both at the level of the whole person and at the level of underlying neural processing, arise from numerous struggles among the stimuli that fall within the receptive fields of neurons throughout the perceptual hierarchy. Each struggle is biased, though not settled outright, by a top-down signal specific to attention. The article notes that competition views of mechanism do not force a non-Broadbentian view of function. Reynolds and Desimone present their theory in recognizably Broadbentian terms:

The visual system is limited in its capacity to process information. However, it is equipped to overcome this constraint because it can direct this limited capacity channel to locations or objects of interest.

Scholars then disagree about where attention lies within the model. Some stress the top-down biasing signals and take their source to be the real locus of selection. Others, following Duncan's original presentation of the theory, read it as placing selectivity in the whole integrated competition, in which any part can bias any other. On that second reading, the article says, the theory fits a proposal that attention has a metaphysical status like unison in an orchestra. No single process is the attention process. Attention consists in the unified activity of whichever processes are relevant to a person's current task. Those processes differ from task to task, so attention does not require any particular processes to occur. Christopher Mole develops this view and calls it the “Cognitive Unison Theory.” The article notes that it revives the old view of Locke and F.H. Bradley that attention is a mode of thinking, has a precedent in Ribot's idea of attention as monoideism, and draws on Alan White's 1964 monograph on the many forms of attention, based on the study of ‘heed’ concepts in natural language. The theory has a costly consequence. If operating in unison requires that no task-irrelevant processing takes place, full attention is rare, and most attentiveness is partial or divided. Mole accepts this. Yair Levy argues that it leaves no adequate account of listening or observing, which always involve some attention yet can be done more or less attentively. Philipp Koralus proposes a better treatment of divided attention, in which attentive pursuit of a task shares formal properties with answering a question. Naomi Eilan argues on phenomenological and epistemological grounds that attention is “the means by which we answer questions about the environment”, a thought the article credits to Rowland Stout. Koralus builds a formal model of this questioning on Hamblin's 1958 semantics of questions. The article then raises questions that do not depend on the unison theory. One is how partial or divided attention relates to full attention. Another is whether attention that merely distracts us should be distinguished from attention we keep concertedly focused. Aaron Henry uses constitutive norms to argue that distraction always involves some shortcoming while still counting as genuine attention and not merely a defective version of it. Zachary Irving argues that a wandering mind poses its own puzzles, and offers an empirically informed account on which mind-wandering is “attentive but unguided”. The article observes that a theory like the Cognitive Unison Theory, in which attention depends on a task guided by understanding, will need some other account of the phenomena Irving identifies.

The spotlight metaphor does a different job from the bottleneck metaphor. Bottleneck images have guided theories that look for the cognitive resources that operate only on attended stimuli. Spotlight images have guided theories that ask which features of a stimulus decide whether it is being attended at a given moment. The entry puts it this way: "it has been spotlight metaphors that have guided the theories that attempt to say which features of a stimulus determine whether attention is being paid to that stimulus at any given moment." The core idea is that a stimulus's location ultimately decides whether it receives attention. The entry is careful to say this does not deny that we can attend to something because it is brightly coloured or interesting. Its point is that "one pays attention to brightly coloured or interesting things only by directing one's attention to the location of those things."

The entry explains why the metaphor is so appealing. Imagine an array of differently coloured shapes appearing and disappearing in various places. We might attend to whatever is at the top of the screen, to all the red shapes, or to all the triangles. Some of these ways of attending look more basic than others. Attending to the top part of the screen looks like a primitive task. Attending to the triangles cannot be primitive in the same way:

we cannot simply attend to the triangles.

To attend to the triangles, we first have to work out where they are. To attend to the top of the screen, we do not need to check whether the things there are triangles. So attending by location seems more basic than attending by shape.

This leads to a tempting thought: that attending by location is absolutely basic, so that attention is always allocated by location. If that were right, a theory of the spatial allocation of attention, a theory of how the attentional spotlight moves and focuses, would be a large and central part of any account of attention. The entry notes that this would be good news for anyone seeking a single unified theory, and for the scientific project of explaining attention, since the spotlight-like behaviours of attention are among its best understood aspects.

The entry grants that location does seem to play a special role. But it argues that the role is probably not as straightforward as the most parsimonious spotlight theories would need. Two points are made. First, location is not the only property with such a role. Second, the role of location is more complicated than a simple spotlight metaphor suggests. Empirical debate has focused mostly on the second point.

One source of evidence is the study of patients with unilateral neglect, who show different patterns of attention failure. These studies suggest that several frames of reference are involved in allocating attention spatially. Sometimes attention follows location in egocentric space, meaning space defined relative to the body, and sometimes it follows location in more complex frames. The entry notes that Allport made this point in his critique of the Broadbentian debates. There are also well-known experimental effects. John Duncan showed that attention shifts more readily between two locations that fall within a single object than between equally distant locations separated by an object boundary. Experiments with three-dimensional stimuli add another wrinkle. Participants cannot differentially attend to depths in empty space. They can shift attention forward and backward, relative to the point their eyes are fixating, only when the relevant locations contain objects.

The entry draws these sources together: "the spotlight of attention is not allocated simply on the basis of coordinates in any single spatial frame of reference." This matters for anyone trying to locate attention's mechanisms in the brain, because different brain areas map space differently and represent spatial information about objects in different ways. It has led psychologists to move away from simple spotlight metaphors. Still, the entry argues these findings do not undermine the core of the spotlight idea. What a person is attending to might still depend only on where that person is attending, even if many factors, some of them complex, determine which location that is.

Other findings are more damaging to a pure spotlight view, though they do not refute it. The entry points to experiments by Dwight Kravitz and Marlene Behrmann. In these studies, a brief cue appears at one end of a shape on a computer screen. The shape might be a plain rectangle, a rectangle with bulbous ends, or an H or h shape. It appears beside another shape that sometimes matches it in both shape and colour and sometimes does not. Attention is then measured by reaction times to stimuli presented on and near these shapes. Kravitz and Behrmann found that all of these properties interact to shape how attention is allocated across the screen. Most strikingly, attention differed between trials showing an upper and a lower case H and trials showing an upper case H and the number 4. The 4 is exactly the shape of a lower case h turned upside down. The entry concludes that attention, even when triggered by a simple spatial cue, "is not allocated merely on the basis of location, but also on the basis of shape, colour, and arbitrary learned meaning."

The entry notes that this does not rule out a special role for location. Several lines of research have sought a single representation in the brain that determines how much attention each item receives. The oldest of these, starting with work by Koch and Ullman, treats that representation as a map in a quite literal sense. Every location in a region is represented, in a format that directly encodes the spatial relations between locations. Each object is represented as being at some location, and the properties of those objects determine how likely attention is to land there. On this view, which the entry calls the theory of salience maps, location has a unique role even though other properties also contribute.

More recent versions of this idea, such as the work of Pessoa, add subjective properties to these maps, including emotional value and current motivational relevance, alongside objective properties that make an item stand out. The more such non-spatial information is added, the less spotlight-like the allocation of attention will seem. But as long as the representation is still map-like, location can keep an important role, perhaps even a uniquely important one.

The entry also points out that these debates may matter beyond the spotlight metaphor. Aaron Henry has argued that the viability of some recent philosophical theories depends on how they turn out. Those theories tend to emphasize the role of agency in attention, which leaves them open to counterexamples from involuntary attention. Henry suggests that the threat of such counterexamples can be defused if involuntary capture is understood to involve attention being allocated on a map in which the agent's own priorities for action are represented.

Finally, the entry offers an argument against pure spotlight views based on experience rather than laboratory data. A pure spotlight theory, on which attention always follows location, cannot account for the case that Hermann von Helmholtz and William James discussed. A listener can attend to the pitch of a note or to its timbre and overtones, and this difference does not correspond to any difference in the location attended. A purely location-based theory also struggles with tasks that depend on time, such as finding out which of a group of shapes appears on screen for the longest. The entry notes that temporal aspects of attention were a late addition to psychological research. They show a diverse range of effects with diverse neural underpinnings. Philosophical work on this topic has mostly addressed our understanding of the experience and perception of time, rather than attention itself. The entry refers readers to a separate entry on that subject.

The section closes with a qualified conclusion. Because some shifts of attention do not change the location attended, a theory of which locations a subject is attending to cannot tell the whole story about attention. It may still be a central component of some of attention's forms.

Motor theories take a more physical route into the same territory. Their starting point is neurophysiology in macaque monkeys. Researchers used electrodes, placed in living animals, to stimulate brain areas known to coordinate eye movements. The result, the paper's sources report, is strong evidence of an anatomical overlap between the regions that control eye movement and the regions that decide which stimuli produce the strongest responses in visual cortex. The work also points to a functional link. Moore and Armstrong found that the connections from frontal eye-movement areas to attention-like effects in occipital visual areas are spatially specific. A stimulation that elicits an eye movement toward one part of the visual field, when applied at a higher level, produces at a lower level changes in neural responsiveness that are exactly similar to the changes seen when attention is shifted to that same location.

These findings are the basis for what is called a motor theory of attention. On this view, the processes underpinning attention are, in at least some cases, truncated versions of the processes that coordinate the movements of the sense organs. Rizzolatti and colleagues state the central idea directly:

the program for orienting attention either overtly or covertly is the same, but in the latter case the eyes are blocked at a certain peripheral stage.

Overt attention involves looking, while covert attention shifts without the eyes moving, so on this account both draw on one program, and covert attention simply has the eyes held back.

Advocates present this as a revival of an older picture. Alexander Bain, at the end of the nineteenth century, identified attention with truncated versions of the motor processes that bring about its behavioural manifestations, processes

stopping short of the actual movement performed by the organ.

The chapter separates these strong versions from a weaker family of views. In the weaker views, attention and motor control are closely linked, but the links are not specifically to the movement of sensory organs. Heiner Deubel and his collaborators provide one line of evidence. They show that changes in how an object figures in action shift attention in a way that affects the perceptual discrimination of stimuli on or near that object. In one example, Deubel and Schneider found that attention spread across the space around a wooden X shape about 6.5 centimetres across depended on the hand used to reach for it. A subject reaching with the left hand to grasp the top-left and bottom-right arms attended differently from one reaching with the right hand to grasp the top-right and bottom-left arms. In each case attention went not only to the shape but to the parts that were about to be gripped.

The authors are clear that action-based theories have limits. Eye and limb movements are always movements to a location, so motor theories have difficulty explaining shifts of attention that are not tied to a difference in location. They face the same problem that pure spotlight theories face. The paper also notes that it is unclear how motor theories apply to sensory modalities where the orientation of the sense organs is less straightforward than in vision. Even so, the motor theory is presented as a plausible and well-supported account of the location-based visual attention found in typical laboratory experiments.

The success of motor theories with those laboratory experiments can be read in two ways. The optimistic reading takes the theory to explain the processing behind some of attention's simpler and more central forms. The pessimistic reading treats the theory as exposing a problem with the experimental paradigms themselves. Participants in these tasks direct attention to different parts of the visual field while keeping their eyes fixed on a central spot. The behavioural signs of attention are usually faster reaction times to attended stimuli, and their neural correlates are biases in which stimuli most strongly drive the circuitry that processes them. On the pessimistic reading, these signs are consequences of the truncated eye movements the fixation requirement induces, so the phenomenon studied in attention labs may not be the psychologically important one it has been taken to be.

Motor theorists are not alone in risking an overly general theory built from a narrow set of phenomena. Henry Taylor argues that while philosophers writing recently on attention often aim at a general theory, that ambition is rare in the psychological literature. There, researchers have followed Michael Posner by treating attention's roles in vigilance, perceptual orientation, and the focused deployment of processing resources somewhat separately. Taylor suggests that philosophical theories should reflect that tolerance for pluralism.

The first question this part takes up is whether attention is necessary for consciousness at all. The entry describes a spectrum of answers. One group holds that many unattended items appear in consciousness. A second holds that only some do, and a third holds that unattended items figure in consciousness only in special circumstances. The entry notes that people differ, independently of any philosophical commitments, in which of these views they find plausible, citing a 2007 study by Schwitzgebel. Some theories go further. Watzl's theory, for instance, makes attention necessary for "creature consciousness", meaning that every conscious creature must be an attentive one. Such views, the entry points out, need not entail that attention to an object is necessary for perceiving that object.

At one extreme of this spectrum is the view that attention to a thing is strictly necessary for consciousness of it, "so that the things to which we are not paying attention do not figure in our consciousness at all." The entry says that a number of psychologists and some philosophers endorse this view, and that it has historical precedent. William James began his chapter on attention with the remark "my experience is what I agree to attend to", and Kames wrote that

Attention is requisite even to the simple act of seeing.

The evidence offered for the strong view comes from experiments showing how little people notice of what they have not attended to. In Arien Mack and Irvin Rock's inattentional blindness paradigm, participants given an attention-demanding task about one part of their visual field often fail to notice shapes or words flashed elsewhere. In Rensink and colleagues' change blindness paradigm, participants viewing an alternating pair of pictures may need ten to twenty repetitions before they spot a large but narratively unimportant difference between them. The most memorable case involves a large number of participants who fail to notice a man in a gorilla suit appearing in the middle of a scene, because they are busy counting passes in a basketball-style game.

The entry argues that no single experiment can settle the matter. The claim that attention is always necessary for consciousness is a universal one, and no one instance establishes it. It is also unclear whether these experiments show unattended objects dropping out of consciousness altogether. The participants' ignorance of the gorilla is compatible with the gorilla making some phenomenal difference to their experience, which they simply cannot use when answering the experimenter's question. That difference might be forgotten at once, or it might be too unstructured to be mobilized. In the second case the effect would be better called inattentional agnosia than inattentional blindness. Absolute amnesia or thoroughgoing agnosia for unattended items could look the same as inattentional blindness, both in behaviour and in retrospective introspection. The entry says that neuroimaging may therefore be among the few sources of evidence available, and it cites a study that used fMRI to argue against the amnesia reading for unattended written words.

Change blindness raises a parallel problem. Only narratively insignificant differences go unnoticed, so these effects might show not that there is no consciousness without attention, but that there is consciousness without attention whose contents are limited to the scene's overall gist. Defenders of the strong view can reply that attention was paid to the gist features after all, either in a dispersed, non-focal way or because those features briefly and involuntarily caught attention. Without an experimental way to achieve complete inattention, the two interpretations cannot be separated. The result, the entry says, is a kind of deadlock. Experimenters can claim only to have shown perception in the "near absence" of attention, which gives inconclusive support to the hypothesis that attention and consciousness can come apart.

This methodological puzzle also casts doubt on our everyday beliefs about unattended items. Those beliefs vary from person to person and shift with features of the context in which they are probed. To the extent that we find it natural to suppose unattended items do figure in consciousness, the entry suggests this may come from what it calls the "refrigerator-light illusion". The idea is that the light seems always on because we only ever look and find it on. If our belief that we are conscious of unattended things rests on finding those things whenever we check, the belief is on shaky ground, since checking itself shifts attention to the things checked. A checking procedure, in the entry's words, "cannot provide us with good evidence about the status of unattended items".

Since a direct check is ruled out, the entry says, the dispute must be assessed through theory. The two sides hold that inattention either removes an item from experience or leaves the subject in a poor epistemic position regarding it. The entry judges it premature to settle this by inference to the best explanation, because neither side has a theory of the relation between attention and consciousness developed enough to compare explanations. It does identify the considerations each side might invoke. Those who think attention is necessary for consciousness can point to theories that give attention a leading role in explaining consciousness, and to evolutionary arguments about how the two came to be linked. Those who favour the epistemic-deficit reading of inattentional blindness can point out that it fits a view, defended by Campbell, on which attending to an item gives the acquaintance needed to use a demonstrative concept of it. On that view, a person who did not attend to the gorilla could still have the gorilla in consciousness without being able to form a thought like "That's a gorilla", which would explain why inattentive people cannot answer whether there was a gorilla.

The entry notes that the gorilla case is hard to read because most people who do not attend to the gorilla could attend to it, and would have if something had prompted them. Ned Block has proposed a case where the failure to attend has a more systematic explanation: identity crowding. A peripheral letter such as a 'T' or an 'F' can be seen reasonably clearly by itself, but when other characters are placed alongside it, their properties become confused with its own. Attention seems to bind together properties of all the characters in the vicinity, so attention cannot be aimed at a region that fits the single character tightly. Block claims that even when crowded characters share all their features with the characters crowding them, they are consciously discriminated from their surroundings and figure as objects in consciousness. Since attention cannot be allocated more finely than the crowd, Block takes identity crowding to be a counterexample to the claim that we are conscious of objects only when we attend to them. Taylor and Richards have contested the claim that the characters figure as individuated objects rather than as parts of a feature-rich texture. Block has replied that the best explanation of several related phenomena favours his reading.

The entry concludes that inference to the best explanation is always hazardous, especially near an explanatory gap, and that the unresolved question may have to wait for a better understanding of how attention relates to demonstrative reference, and perhaps of attention's role in epistemology more broadly.

The second question is the reverse. Is attention sufficient for consciousness, so that whatever we attend to is consciously attended to? Declan Smithies has argued that every instance of attention is consciousness-involving. His reasons are that every instance of attention must connect to its subject's rational access to the contents of experience, and that wholly unconscious processing has no connection to rationality. On this view attention is essential to the rational access that conscious experience provides. The entry notes that this connection is itself contested. Johannes Roessler argues that John McDowell's account of how perceptual experience gives reasons cannot account for attention's role. Susanna Siegel and Nicholas Silins claim that an experience can influence belief in a way that counts as rational access even when the experience receives no attention. Terry Horgan and Matjaž Potrč argue that the justification of belief by experience goes beyond both the contents of attention and the contents of consciousness. The entry suggests that a theory of attention and a theory of epistemology may constrain each other, and so offer a route to settling whether unconscious attentiveness is possible.

The claim that all attention is conscious attention sits uneasily with some empirical results. In one experiment by Yi Jiang and colleagues, participants saw stimuli through a set of mirrors arranged so that each stimulus reached only one eye. A more vivid image shown to the other eye won the competition for consciousness in this binocular rivalry, so participants did not consciously experience the weaker stimulus. The unseen stimuli included erotic photographs, and these seemed to attract attention regardless. Consciously seen stimuli placed where the unseen photographs appeared were responded to more accurately in a task that required detecting the orientation of Gabor patches. A striking finding was that the way the unseen photographs attracted and repelled attention depended on the sexual orientation of the participants. The entry says that the natural reading is that the pictures captured attention without being experienced. This suggests, though it does not conclusively show, that a person can attend to a thing without being conscious of it. A related case involves a blindsight patient, who has no experience of cues in the blind area of his visual field, yet those cues produce the speeding of processing and reduction in reaction times that usually mark attention.

One complaint against both lines of evidence is that they do not distinguish attention to a thing from attention directed at a region of space in which the thing happens to lie. They therefore cannot show attention to a thing without consciousness of that thing. Liam Norman, Charles Heywood and Robert Kentridge address this complaint in part. Their normal participants viewed an array of small, rapidly flickering Gabor patches, in which objects were defined only by changes in patch orientation. Participants saw only a flickering screen, not the objects. Cues still drew attention to those unseen objects, as shown by differences in response times to items presented inside and outside their boundaries. The entry says such results show an influence running between unconscious processing and the direction of attention, and perhaps show that attention can be paid to objects that never appear in conscious experience. These results, it says, place a limit on how close a relation between attention and consciousness can be claimed.

The entry connects these disputes to questions about attention's metaphysics. The claim that attention is not sufficient for consciousness usually belongs to a view on which attention and consciousness rest on two distinct brain processes that can occur independently. The claim that attended things are a proper subset of the things present in consciousness is what those who follow Locke require, since for them conscious thinking constitutes attention when it occurs in a certain way. The claim that attention is both necessary and sufficient for consciousness is what those who identify the process by which things reach consciousness with the process by which attention is allocated must hold, as Prinz does. Some scientific theories also carry strong commitments here. Taylor Webb and Michael Graziano use their Attention-Schema theory to argue that consciousness emerges where control of attention requires a schematic representation of how attention is being allocated.

The third question is how attention changes what experience is like. Whatever attention does to the presence of things in consciousness, the entry says, there is good reason to think that shifts of attention make some difference to the character of conscious experience. Phenomenologists and Gestalt psychologists describe this difference in various ways. Maurice Merleau-Ponty emphasized how attention can shape the figure and ground structure of perceived space. Aron Gurwitsch developed a theory in which the structure given by attention has three parts. In the second half of his 2017 book, Sebastian Watzl proposes that attention imposes a priority structure on our mental lives, and he develops the phenomenological side of that proposal.

Whether the differences made by attention can be captured as differences in the content of the attended experience is currently disputed. If Watzl is right that they cannot, the entry says, the phenomena of attention present a serious difficulty for any fully general attempt to explain the character of experience through the contents it represents. Such explanations might still work within narrower domains.

The entry also reports a series of studies by Marisa Carrasco and her collaborators. These found that the phenomenal differences produced by shifts of attention include differences in perceived brightness, contrast and colour saturation, but not differences in hue. Ned Block argues that these differences should not be understood as differences in content, because treating them that way would imply that attending to an item creates an illusion about its brightness and contrast, which he finds implausible. Block takes Carrasco's findings to show an aspect of experience that a content-based theory of phenomenal character cannot account for.

Similar disputes concern how attention affects the perception of ambiguous figures such as the Necker cube, based on introspective considerations first raised by Fiona Macpherson. Bence Nanay proposes that these effects can be explained by those who hold that every difference in consciousness is a difference in content. On his account, attention modulates the specificity of content, making representations of determinable properties more or less determinate. The entry notes that this resembles debates over how blurriness enters visual experience.

Finally, Mattia Riccardi gives attention a different phenomenological role. Opposing a proposal by Matthen, he suggests that attention explains the presentness that distinguishes perception of real objects in one's immediate surroundings from experience of objects in photographs or objects pictured by the mind's eye during vivid imagination.

Consciousness is only one of the puzzling mental phenomena that the entry says may be explanatorily tied to attention. Demonstrative reference is another. That is the way we pick out particular things with words like "that" or "this". One advocate of linking attention to demonstrative reference is John Campbell, who writes that "the notion of conscious attention to an object has an explanatory role to play: it has to explain how it is that we have knowledge of the reference of a demonstrative."

The idea has an older form. Bertrand Russell explored it in a manuscript on the theory of knowledge that he abandoned in the summer of 1913, under the influence of Wittgenstein. Russell wrote that

At any moment of my conscious life, there is one object … to which I am attending. All knowledge of particulars radiates out from this object.

Campbell develops two lines of thought in support of this. The first starts from conversations in which one speaker says 'that woman' while several women are present. To know which woman is meant, the listener must attend to her and know that the speaker is attending to her. Campbell uses this to argue for more than a causal link. He says attention and reference stand in a particularly intimate relationship, one he sums up as "knowledge of the reference of a demonstrative is provided by conscious attention to the object".

The second line concerns deductive arguments whose premises use demonstratives, such as: that is F; that is G; therefore, that is F and G. Such an argument is valid only if the word 'that' keeps one referent across both premises. Campbell holds that this single fixing of the referent happens if and only if attention is not redirected between the premises. On this view attention plays a role in fixing a demonstrative's reference that is comparable to the role a Fregean sense plays in fixing the reference of a proper name.

The explanatory approach has long faced a charge of circularity, raised by Peter Geach in his 1957 book Mental Acts. Geach argued that it is quite useless to say the relevant sense-perceptions must be attended to. Either that does not give a sufficient condition, or the phrase 'attended to' simply names the very relation between judgment and sense perception that needs to be explained.

Imogen Dickie has tried to avoid this charge. In a 2011 paper she draws on cognitive psychology, which offers a theory of how attention tracks objects over time. She suggests that this theory can explain how attending to an object removes any knowledge-defeating element of luck from inferences about it, which makes the object a possible topic of demonstrative thought. Her wider view is that "a thought comes to be about a thing when the justificatory practices in which that thought figures are anchored to the object in question."

The entry leaves open whether a theory of attention can be turned into a non-circular account of demonstrative reference. Even so, the link between attention and reference seems to clarify what happens when we understand referring expressions. Campbell's examples suggest that what we attend to and what we refer to are often the same. Developmental psychology adds that attention-related abilities help infants come to understand a caregiver's demonstrative use, a point the entry returns to when it discusses other minds. Some of the implications reach beyond language, to how thoughts come to have particular objects as their contents.

Finally, Michael Barkasi has suggested that the frontal eye fields, through their pattern of neural projections, could show how attention contributes to fixing the reference of thoughts. Where Campbell describes attention in epistemic terms, Barkasi describes it as sub-doxastic, meaning it operates below the level of belief. On his account attention 'sets targets' and so regulates the flow of visual information.

The entry also turns to how people come to understand one another. Here it draws on developmental psychology, which it says offers results suggesting an intimate link between the growth of attention-related abilities and the growth of the capacities needed to understand the mental states expressed when an infant and its caregiver interact. An ability to appreciate what others are attending to looks like a crucial stage in a normal infant's development toward understanding that a caregiver's utterances carry referential intentions. More generally, attentional capacities develop at milestones on the route to some distinctively human cognitive abilities.

The entry describes a specific developmental pathway. Normal human infants first develop an ability and willingness to attend to their mother. Next they develop an ability and willingness to attend to whatever she is attending to. Most importantly, they come to take part in episodes in which a third object is attended to jointly by mother and child, "with mutual understanding of the fact that their attention is shared." In these episodes, infants and caregivers rely on complex cues, including speech, gaze-following and pointing.

The entry says there is good evidence that a child's progress along this pathway is tied to two things: its ability to respond appropriately to the mental states of others, and its ability to learn new vocabulary by grasping what its mother's words refer to. It draws on a longitudinal study which suggests that earlier stages on the pathway predict the age at which infants reach later ones. The most striking finding concerns pointing. Mastering pointing as a way to direct another's attention at twelve months predicts understanding of false beliefs at fifty months, a grasp that someone can hold a belief that is untrue. This holds even when temperament and language are factored out. The entry says this gives "some evidence for the hypothesis that this sequence is a causal one."

A second line of evidence comes from autism. The arrested development of these attention-involving abilities is, in the entry's words, "a revealing marker of the impairments suffered by autistic children." A third, more contested, line concerns non-human primates. A lack of such abilities is related to limits in the mental-state attribution of these animals, though the entry calls this last claim "more controversial."

The entry then considers what philosophers can learn from this empirical work. Some progress has been made. Joint attention appears to help explain distinctively human aspects of cognitive and social development. Once that development is complete, it also appears to help explain how some elementary social phenomena are carried out. For adult cases of shared intention, Fiebich and Gallagher are cited as arguing that jointly attending has a basic, elementary status, as "the most minimal instance of a joint action."

A central philosophical difficulty is characterizing the sharedness of the attention involved in this developmental trajectory. Christopher Peacocke argues that joint attention involves a reciprocal recognition: each person recognizes the other's recognition of their own attention. This creates a mutual openness that the entry compares to the openness of "common knowledge", the notion that figures in philosophical accounts of communication and convention. Young children achieve joint attention, so its achievement cannot plausibly make sophisticated intellectual demands. For that reason Peacocke argues that awareness of reciprocated openness must be more fundamental than common knowledge. He suggests it is "a variety of perceptual awareness that makes such common knowledge possible."

John Campbell shares Peacocke's view that joint attention is a perceptual phenomenon, rather than one reached by reflecting on one's perception in a certain way. Campbell differs, however, in taking the person with whom attention is shared to be a fellow subject of the joint experience. On his view that person is not an object figuring in the content of the experience each joint attender has. Lucas Battich and Bart Geurts reject Campbell's attempt to move the explanation of sharedness from the cognitive to the perceptual domain. They argue that his view suffers from problems rooted in his relational conception of perception. They claim these issues shed light on the long-running debate between "representation-based and relation-based theories of perception."

The entry argues that the links it has just described, between attention and demonstrative reference and between attention and knowledge of other minds, might be special cases of a more general connection between attention and the making of epistemic moves, the steps by which a thinker comes to know or justifiably believe something. Descartes may have had such a connection in mind when he suggested that attention to clear and distinct ideas is a necessary condition for those ideas to realise their special epistemic potential. Stewart may have meant something similar when he suggested that

Some attention [is] necessary for any act of memory whatever.

Kant's anthropological writings also give attention a crucial epistemic role, explaining the control a person can exert over representations in thought and in imagination, including the kind of imagination Kant takes to be involved in turning experiences into experiences of an objective spatiotemporal world.

The entry notes that each of these philosophers frames the idea in a way that reflects their other commitments. Still, it says, some version of the idea can be made plausible before any particular theory is adopted. The inattentional blindness experiments are its example. Participants are shown large changes while they attend to something else, and inattentive people miss things that attentive people would find obvious. The entry then argues that the same holds for a priori reasoning. No item is so large and central that a sighted, conscious observer could miss it, and no step in reasoning is so simple and compelling that every thinker, attentive or not, can be expected to recognise it. If the blindness effects show that perception carries attentional demands a thinker must meet before perceiving things can give knowledge of them, then similar demands may have to be met before grasping a thought gives a justified belief in that thought's consequences. The entry adds that this reading of the blindness effects is controversial.

Nicholas Silins and Susanna Siegel suggest that even where inattention explains why a perceiver fails to form a belief about what is happening in their visual field, unattended aspects of experience can still provide justification for such beliefs. Émile Thalbard argues to the contrary.

In both the a priori and the perceptual case, the entry says, acquiring knowledge seems to require attention on at least some occasions, and that requirement seems to be a practical one. It may be that knowledge of some propositions is needed for one's experience of a gorilla to give knowledge of the gorilla's presence, where those propositions could serve as a major premise in an inference by which the experience was interpreted. But the requirement cannot be that one must be bearing such propositions in mind. The same holds in the a priori domain: one cannot be required to entertain some prior knowledge of a proposition from which an entitlement could be deduced. Requiring occurrent propositional knowledge for these moves would, the entry says, embark on the regress set out in Lewis Carroll's dialogue of the tortoise and Achilles.

What epistemic move-making requires is not occurrent knowledge of a proposition. The thing that it requires is the right sort of active attention.

These attentional requirements, the entry stresses, are not requirements merely for alertness, and they are not captured by saying that the thinker has to pay some attention to the relevant ideas. A thinker may be attending to a syllogism and still fail to see that the conclusion follows, if the attention is directed to the syllogism's rhythm. The entry calls a non-question-begging characterisation of the attentional requirements of knowledge acquisition in general an important contribution to epistemology. Current work on attention, focused as it is on perception, may supply only part of the general theory that is needed. Work on purely intellectual forms of attention remains scarce.

Mark Fortney is cited as emphasising this shortcoming. He argues that neglect of intellectual attention has led philosophers to misinterpret some claims made in the empirical literature on attention and consciousness. Fortney also argues that considerations about intellectual attention should play a central role in debates about the transparency of phenomenology, and about what that transparency means for the epistemology of self-attributed thoughts. Those debates are illustrated by Byrne's work. Fortney suggests that once attention is given this role, certain results from cognitive psychology shed light on the debates, and that some claims about the transparency of occurrent thought start to seem dubious.

The entry turns to attention's role in voluntary action, and it begins by noting that this is much harder to study in a controlled experiment than attention's role in perception. As a result, the theories that emerged from experimental psychology have focused almost exclusively on attention's perceptual instances. The entry contrasts them with theories developed before psychology and philosophy split apart, when the action-involving side of attention was far more prominent.

William James is the key example. The entry notes that in his Principles of Psychology he suggests that

volition is nothing but attention.

At one point James went further. In an 1880 statement, he proposed that

Attention, belief, affirmation, and motor volition, are ... four names for an identical process, incidental to the conflict of ideas alone, the survival of one in spite of the opposition of the others.

The entry observes that this reads as an anticipation of the competition-based theories it discusses earlier.

The entry argues that the question of attention and voluntariness runs parallel to the question of attention and conscious perception. Attention seems necessary for the voluntary performance of finely tuned behaviour, just as it seems necessary for consciously perceiving fine detail. But it does not seem natural to think we must attend to the execution of every voluntary act, any more than we must attend to something in order to perceive it at all. On commonsense grounds, then, one could claim that attention figures in some voluntary behaviour but not all of it.

A rival theorist could push back, holding that attention is necessary for any action to count as voluntary. On this view, an illusion, analogous to the refrigerator light illusion, makes inattentive actions seem voluntary when they are not. The entry describes the same kind of methodological puzzle that surrounds consciousness here, one with deep epistemological roots: it is unclear what sort of evidence could settle the matter. The entry suggests this may be a topic for future research.

The entry then reports positions that resist both extremes. Denis Buehler, it says, rejects the idea that attention is either necessary or sufficient for an act to be voluntary, yet argues that attention has a central role in explaining flexible, goal-directed behaviour. Where that behaviour involves a skill, Alex Dayer and Carolyn Jennings argue that no simple account of attention's role can be given. They claim instead that a theory of attention in skilled performance must be pluralist.

Finally, the entry argues that separating attention's role in action from its role in perception is somewhat artificial, especially when the perception concerned is of one's own moving body. Research into that bodily form of perception has largely developed apart from research on attention in vision and hearing. The entry points to a review by Gregor Hochstetter, which covers this bodily research and, following Marcel Kinsbourne, argues that an internal representation of a "body image" must play some role in guiding attention during ordinary bodily awareness.

Attention’s contribution to our understanding of values has been relatively underexplored in the recent literature, the entry notes, but the range of value-theoretic contexts in which attention makes an explanatorily significant appearance is nonetheless broad. The first of these is aesthetics.

Aesthetics is often described as the branch of philosophy concerned with one particular sort of experience. The entry points out that this description only becomes informative once that sort of experience is specified, and that the specification cannot simply be a list of the objects experienced. The same wine may be tasted by a connoisseur and by a thirsty man, with only the connoisseur’s experience counting as aesthetic. The same painting may be seen aesthetically by an art lover and from a purely financial point of view by its dealer. One response is to individuate experiences more finely than their objects, by identifying the properties that are experienced as belonging to those objects in the aesthetic case. An alternative is to consider how the subject experiences those contents, by identifying an attitude that characterises the subject in the aesthetic case, or a mode of attention that the subject instantiates.

Attempts to identify such an aesthetic mode of attention are well represented in the philosophical tradition. Kant’s claims that judgements of beauty involve a first moment of disinterested pleasure are usually read as an example of the attitude-identifying approach. Jessica Williams has argued, however, that Kant must also be working with a notion of aesthetic attention, understood as a mode of attention involving a distinctive combination of imagination and understanding.

A similar notion plays a central role in R.G. Collingwood’s account of how art, and literature in particular, contributes positively to a society’s culture. Collingwood describes attention as

the assertion of ourselves as the owners of our feelings.

He takes this self-assertion to be crucial in enabling us to exercise volition over those feelings, so that

Their brute power over us is thus replaced by our power over them.

On this account, a society whose artistic culture has diminished becomes vulnerable to emotive influences. Collingwood himself suggests that this might best be read not as a broad philosophical claim, but as a point about Europe in the nineteen thirties.

Attempts to demarcate the aesthetic through an aesthetic attitude or an aesthetic mode of attention have been attacked, both for their details and for their more general orientation. George Dickie’s 1964 paper is the most famous instance of such an attack. As a result, many philosophers now widely accept that it is wrongheaded to specify necessary and sufficient conditions for something to count as aesthetic by identifying a particular aesthetic attitude. Many still hold that informative things can be said, for less strictly defined theoretical ends, about the attitudes often found in aesthetic cases. In this spirit, some ideas about distinctive forms of aesthetic attention have recently been revived.

Servaas van der Berg proposes that a distinctive mode of attention can be seen in cases of aesthetic appreciation. He claims that this mode is distinguished by a motivational structure similar to that of game playing, as theorised by Bernard Suits and Thi Nguyen. That structure inverts the normal order of priorities between means and ends. In ordinary action, one takes means in order to bring about ends. The game-playing golfer, for example, adopts the end of getting his ball into a hole so that he can take the means to reaching that end. Van der Berg argues that a similar inversion occurs in aesthetic appreciation, where the cognitive means required to reach understanding supply the reasons for adopting the end of understanding the item being appreciated.

Bence Nanay sets out a position with similar aims, though without aspiring to give necessary or sufficient conditions for the aesthetic. He aims instead to cast light on how certain central cases of aesthetic experience operate, through a theory of the distinctive way attention is engaged during them. He distinguishes focused attention from distributed attention, and uses examples from a range of arts to suggest that aesthetic experiences often require attention of both sorts. He also proposes that this contributes to the value we place on such experiences.

Simone Weil gives attention a central place in ethics, and in value theory more broadly. She writes that

The authentic and pure values – truth, beauty and goodness – in the activity of a human being are the result of one and the same act, a certain application of the full attention to the object.

On this view, attention is not a side ingredient of moral life. It is the act from which truth, beauty and goodness come. The entry notes that this raising of attention to a near-supreme status is the basis for Weil's claims about priorities in education.

The entry says that a number of other thinkers working where philosophy meets literature share her position. Christopher Thomas, for example, suggests that the role of attention in Weil's "ethics of affliction" can be read as an effort to apply central ideas from Kant's aesthetic theory to ethical questions.

The influence of Kant, and the interplay between ethics and aesthetics, moved to the foreground in the work of Iris Murdoch, who developed Weil's claims further. Murdoch holds that certain forms of attention play an essential role in exercising the virtues, and that the capacity for these forms can be cultivated through experiences of beauty. Dorothea Debus takes up these themes and puts particular weight on Weil's claim that it is full attention to which cardinal value, meaning the highest moral worth, is attributed.

Antony Fredriksson and Silvia Panizza argue that the aspects of attention Murdoch considers important become clearer when seen through Maurice Merleau-Ponty's account of attention, which stresses the openness of the attentive mind. Nicolas Bommarito has built an account of modesty along broadly Murdochian lines, describing modesty as a "virtue of attention". The entry reports his argument that this approach avoids epistemic difficulties facing other accounts, which define modesty as requiring false beliefs about the extent of one's own strengths.

Peter Goldie and Michael Brady approach related questions from another direction. Both relate attention and virtue to the epistemic significance of a person's emotional life. Both hold that an emotion's way of modulating attention is essential to it, and that this modulation plays an ineliminable role, meaning one that cannot be removed, in our full appreciation of the evaluative properties we meet in the world. They disagree about how this epistemic role should be understood. Goldie takes the experience of one's own emotion to be reason-giving, with an epistemic authority like that of one's experiences of things in the visual field. Brady holds that the emotional subject's experience of items in the world has an epistemic authority that their experience of their own emotional state does not share. On his account, the epistemic significance of an emotion must be exhausted by its influence on outwardly directed attention.

Many theorists have noticed that some of the largest businesses, Google and Facebook among them, receive very little money from the end-users of their best-known products, yet receive an enormous amount of attention from those same users. The account says that writers who have noted this often treat it as a sign of a condition that helps shape the current zeitgeist, the prevailing spirit of the age, and that this contribution is typically judged to be harmful.

Michael Goldhaber seems to have been among the first to describe this as a shift from an "information economy" to an "attention economy". The account then builds on Herbert Simon's suggestion that attention becomes scarce when information is abundant, and notes that one can give this an economic reading.

The idea that the present century has seen a distinctive commodification of attention has become widespread, and several social and political philosophies have been used to theorize it. Jonathan Beller takes a broadly Marxist approach, which reads value through the lens of labour and class, and offers what he calls an "attention theory of value". Clinton Castro and Adam Pham draw on Debra Satz's notion of "noxious markets", markets whose goods are harmful to those who take part in them, to argue that technologies designed to maximize attentive engagement should be regulated in ways analogous to the regulation of the tobacco industry. Georg Franck draws on the work of Levinas and Heidegger to explain how the attention economy mediates our transactions of mutual recognition and respect.

The technologies behind this social change do have effects on attention. Empirical studies that measure these effects, the account explains, have tended to frame them by their downstream social consequences rather than by their immediate cognitive basis. Rosen, Carrier and Cheever, for example, observe how social media use affects the study habits of school and university students, relate this to educational performance, and suggest pedagogical strategies that might reduce the harm.

The account observes that such theorists have made relatively little theoretical contact with the accounts of attention offered by philosophers and cognitive psychologists. Some, such as Atchley and Lane, do refer to a bottleneck theory of attention, the idea discussed earlier in the entry, to explain how attention-demanding technologies reduce the attention available for other purposes. Still, the account states, "claims about the psychological effects of the attention economy have often been framed with reference to theories of the psychological basis of addiction", as in the work of Bharvaga and Velsquez.

The entry's main takeaway is that attention has been asked to do many explanatory jobs, from making perception and memory possible to fixing reference and shaping value, and that no single account has settled which of those jobs it is for. The limits and open questions it states include whether attention is necessary or sufficient for consciousness, whether unattended items can figure in experience at all, how spatial and object-based selection relate to one another, whether voluntary action always involves attention, what intellectual attention consists in, and whether a theory of attention can be turned into a non-circular account of demonstrative reference. Its final section notes that the attention economy has been discussed with little contact with psychological theory, so connecting those fields is also left open.

  1. Attention