Contrasting Nyaya-Vaisheshika and Buddhist Explanations of Attention
by Alex Watson
In contemporary Cognitive Science and Philosophy of Mind, “attention” is a burgeoning field, with ever-increasing amounts of empirical research and philosophical analysis being directed toward it.1 In this essay I make a first attempt to contrast how Nyaya-Vaisesikas2 and Buddhists would address some aspects of attention that are discussed in that literature. The sources of what I attribute to “Nyaya-Vaisesikas” are the sections dealing with the manas (“internal organ”, “organ of attention”) in the Nyayabhasya, Nyayamanjarf and Prasastapadabhasya. The words “Buddhist” and “Buddhism” in this essay refer
specifically to the Sautrantika Buddhism of Vasubandhu's Abhidharmako- sabhasya (350-430 CE).3 A comparison involving a later phase of Nyaya- Vaisesika thinking on this matter,4 a later phase of Buddhism, or a different branch of Buddhism may well yield different results.
Section I lays out the ontological postulates that Nyaya-Vaisesikas and Buddhists deemed necessary for the explanation of attention. Section II looks at three arguments that the Nyaya-Vaisesikas gave for their principal postulate, the manas, and three corresponding Buddhist responses to these arguments.
Sections III and IV look at contrasting Nyaya-Vaisesika and Buddhist explanations of, respectively, “shifts of attention” and “competition for attention.” Sections V and VI consider whether the Buddhist model can adequately account for voluntary or endogenous attention, and whether the Nyaya-Vaisesika model can adequately account for involuntary or exogen¬ous attention. In the closing section I identify three things that are commonly attributed to attention and that may seem impossible in both the Nyaya- Vaisesika and the Buddhist models; I show how the two Indian models can account for them.
I. The Ontological Postulates of Attention
In the Buddhist view, we need postulate no new faculty to explain attention, at least nothing that is separate from awareness, certainly no substance. The Nyaya-Vaisesikas, by contrast, to explain attention postulate a substance, an organ responsible solely for attention: the manas. Before the self can become
Philosophy East & West Volume 68, Number 4 October 2018 1292-1313 © 2018 by University of Hawai ‘ i Press aware of a sense-object, it must first direct the manas to the location of the sense-faculty that is receiving the object's data: to the eyes for visual data, to the nose for smells, to the mouth for tastes, et cetera. The metaphors used are that of a boy (= self) sitting in the corner of a room throwing a rubber ball (= attention/manas) to different
locations of the room (= different parts of the body), or a master directing his servant to certain tasks. The manas—and the power of attention that it carries with it—is an unconscious5 instrument that is under the control of the self and, in accordance with the latter's executive commands, can move from sense¬faculty to sense-faculty in order to enable the self to focus on and thus receive information from
different modalities. Functioning as an intermediary bottleneck between the self and the data received by sense-faculties, its limited capacity explains why the self is aware of only one thing at once, and is not flooded by data from all of the sense-faculties simultaneously.6 We have so far been speaking of the manas as that which enables a connection between the self and the sense-faculty whose object needs to be attended to,
thereby explaining the self's limited capacity. But its domain is not restricted to data coming in through the five external sense-modalities; it extends to internal sense data: ideas, pleasures, pains, desires, et cetera. It is the instrument that enables these, too, to be attended to. We have a contrast between a relatively inflationary Nyaya-Vaisesika account (according to which the explanation of attention requires the postulation of
an attention-organ—which is a substance—and an attention¬process, the movement of the manas) and a relatively deflationary Buddhist account (according to which it does not). For Buddhism it is simply the case that every cognition, that is, every act of awareness, pays attention to its object.7 This was elaborated in terms of awareness, by its nature, “turning toward” (abhogah.) its object. For Buddhism we need to accept the existence only of “awareness-
events” or cognitions (which necessarily involve atten¬tion); to postulate a manas separate from these is to postulate a ghost in a machine (or a mantra in some medicine; see below) that works quite effectively without it.
We find this kind of reasoning in the Abhidharmakosabhasya in a passage where Vasubandhu argues that what the Nyaya-Vaisesikas seek to explain by appeal to four things—self, manas, cognitions, and latent impressions—can easily be explained by appeal only to the last two of these four; self and manas play no useful explanatory role, and there is thus no reason to believe in their existence.
The question being addressed there is why consciousness changes, why it does not always occur in the same form.8 An obvious answer that changes in consciousness can be brought about by changes in sense-objects is not given; indeed the passage gives the impression that what is being spoken of are times when one's eyes are shut (e.g., meditation, dreaming), or when there are no changes in the objects within the range of the senses, or when such changes do
not affect one's awareness (e.g., intense thinking or day¬dreaming). The Vaisesika maintains that consciousness changes because it is caused by specific conjunctions (samyogavisesa) of the self and manas9 In other words consciousness will not always stay the same because the present conjunction of self and manas will give way to another, different one.
But this reply does not satisfy Vasubandhu, because both the self and manas are unchanging according to the Vaisesika, so how could there be qualitatively different conjunctions between them? The Vaisesika replies that different conjunctions of self and manas are associated with different cognitions (buddhi).
The idea here is presumably that in one moment the manas may be conjoined with the self while the latter has a particular cognition inhering in it; at other moments the manas will be conjoined with the self while the latter has different cognitions inhering in it. This answer predictably prompts the Buddhist to ask: but what causes the cognitions to differ from each other? Here the Vaisesika appeals to the ripening of different latent impressions
(samskara). A conjunction of self and manas that is caused by a particular latent impression will be associated with a particular cognition; a conjunction of self and manas that is caused by a different latent impression will be associated with a different cognition.
But, argues Vasubandhu, this means that self and manas are not contributing at all to the explanation of changes of consciousness. It is the latent impressions that are carrying all of the weight. Why not, therefore, jettison self and manas from one's ontology? The Vaisesikas are deriving changes from
self-manas conjunctions depending on latent impressions, but they could derive them from the stream of consciousness depending on latent impressions, thus ridding themselves of two unperceived entities whose contribution to change, given that they are both unchanging, is mysterious. If they do that, they arrive at the Buddhist model.10
It is in this context that Vasubandhu gives the example of a mantra with which a fake doctor enchants his herbal medicines, claiming that they only work because of the mantra (fearing that otherwise his patients will obtain the herbs themselves and dispense with him).11 Just as the curative power of the herbs can be explained entirely by their own potency, without any appeal to that of mantras, so changes in consciousness can be explained entirely by changes in its latent impressions, without any appeal to self and manas.12
Alongside this contrast between a heavy and light ontology surrounding attention, the following contrast can be made regarding the relationship between attention and awareness: For the Nyaya-Vaisesikas the organ of attention, the manas, is an instrument necessary for the rise of awareness. The operation of attention is in that sense temporally prior to, and ontologically distinct from, awareness. For Buddhism attention cannot be separated from awareness; it is one aspect of it. For the Nyaya-Vaisesikas the process of attention is a cause of awareness. For Buddhism attention is simply a feature of awareness, the fact that awareness is turned toward its object.13
II. Arguments for the Manas
We will now look at some of the arguments for the existence of a manas and see how Buddhism can respond.
The Argument from Absent-mindedness
Frequently, when a person with functioning sense-faculties is within the range of a sense-object, awareness of the sense-object will arise. A non-deaf person in the presence of a sound will frequently be aware of the sound. But this does not always happen. A person may be entranced by a picture in an art gallery and be completely unaware of the sounds around them. This leads us to infer that some factor is present in the first kind of case but absent in the
second. This factor, for the Nyaya-Vaisesikas, is the presence of the manas at the location of the ear. The “absent-mindedness” of the second person is to be interpreted as “absence of the manas”.
Since the Buddhists do not postulate an organ of attention, how can they account for the difference between these two cases? They can assert that in the second case the person's attention (manaskara), th at is, their conscious¬ness' turning toward an object, is pointed elsewhere—at the picture and not the sound. Consciousness can only point in one direction at once. If it is pointed toward the picture, then it will be fully occupied with the picture; it cannot simultaneously point toward, that is, be receptive to, the sound.
The Argument from Capacity Limitation
As alluded to above, one of the main arguments for the existence of a manas is that, without one, the self would be flooded simultaneously by data from all five senses. But it is not flooded in that way. In any one moment we are aware only of a minute selection from all of the sounds, tastes, smells, and colorful forms that are stimulating our sense-faculties, and all the sensations on the inside and outside of our bodies. The manas provides an explanation of this fact: allowing the self to connect with only one piece of information at once, it acts as a bottleneck.15
How can Buddhists explain this fact, given that they do not posit a manas as a bottleneck? A partial explanation has already been given above: attention, for Buddhism, is consciousness' turning toward or pointing to an object, and a pointer cannot point in more than one direction at once. Thus each individual consciousness/cognition (lasting for just a moment) has only one object. If one object is grabbing consciousness' attention, that will block the way for other potential objects. Buddhism appeals to this idea rather than to the existence of an unperceived organ of attention.
I said that this is only a partial explanation. Let us distinguish between the existence of a bottleneck and the determination of what precisely will pass through the bottleneck. The manas explains not only the former but also the latter: the reason why data from one particular sense-faculty rather than another makes it through to consciousness is that the manas is in connection with that faculty. The Buddhist idea that consciousness can only point in one direction at once accounts only for the existence of the bottleneck. Section 4 below (Competition for Attention) will look in detail at how Buddhism can account for what passes through the bottleneck.
TheArgumentfortheNecessityofanInstrumentofInnerAwareness As we saw above, the manas is required to bring about awareness not just of sense-content16 coming from outside the person, but also of thoughts and feelings and other “inner sense-content.” The main argument found in Nyaya and Vaisesika texts for the necessity of such an instrument of inner awareness is as follows. All perception, such as perception of external objects, requires an instrument.
Perception of pleasure and pain et cetera is a kind of perception, so it requires an instrument. The instrument in this case cannot be any of the five external faculties, since they are restricted to their own particular kinds of object. So we arrive at the manas as the instrument of this kind of perception.17
One could challenge that all perception requires an instrument, or that awareness of pleasure and pain et cetera is a kind of perception. But if it is granted that external sense-perception requires an organ of attention, it would be very strange to maintain that inner awareness does not. If awareness of
sounds and smells requires paying attention to them, so, too, surely, does awareness of inner “objects”. And it is a fact of experience that internal sense-content and external sense-content can compete with each other for one's attention, with one losing out to the other, so we need the same organ of attention that is responsible for admitting data from only one of the five external senses to be also the determiner of whether internal sense-content will
make it through the bottleneck. We need one entity to be potentially in contact with both internal and external data and to choose between them (or to carry out the subject's choice), that is, to connect with one of these two and to thereby exclude connection with the other. If we were able to become aware of internal sense-content without this being enabled by the manas, then attending to data from an external sense-faculty (as a result of the manas'
presence there) would not preclude simultaneous awareness of a huge—unlimited—amount of internal data. This, being contrary to experience, is an undesirable consequence. It seems, then, that the self will need the manas in order to be able to pay attention to, and become aware of, cognitions, pleasures, pains, and the like.
But from the Buddhist point of view the picture has become clunky, suffering from a multiplication of entities. For internal sense-content occurs in the self: pleasures, pains, desires, thoughts, et cetera are qualities that reside in the self for Nyaya. Why, then, should the self require an instrument —unconscious and external to the self—to become aware of pleasures et cetera that belong to it, having it as their locus, substrate, and subject? Why should it have to reach outside itself in order to see what is within it?
III. Shifts of Attention
Let us now look at some aspects of mental life where one might think that there is more involved in attention than simply cognition's turning toward its object—where attention seems not to be reducible to a mere feature of awareness—and consider how the Buddhist view can account for them without the concept of a manas o r any separate faculty of attention.
(1) There is the deliberate moving of attention away from a present to a new object. Surely attention here opposes itself to the current act of awareness, pulls away from it, and is thus something over and above awareness. The difference between the Buddhist and Brahminical explanation of this is as follows. For the Brahminical thinkers, there is one thing called attention
that exists throughout the process; in the first moment it points in one direction, and then in order to bring about a shift to a new object, it points in another (see figure 1). For Buddhism what we have is first one awareness with its particular object, and then the rise of a new awareness with a new
object. Valuing ontological parsimony, the Buddhist sees no warrant for postulating the thing on top in the diagram (See figure 2). We experience just a sequence of perceptions, cognitions, et cetera. Since we apprehend no second layer, no unitary attention over and above this sequence, why grant reality to it? (We are dealing with an ontology that aims to stick very closely to experience. We find this same reluctance to grant reality to anything other than what can feature as an object of concept-free perception in Buddhism's denial of a self and denial of universals.18).
(2) If attention were inseparable from awareness, then there would be no instance of attention preceding awareness. How, then, can Buddhism explain those cases that we might term anticipatory attention, that is, the
placing of one's attention somewhere in anticipation of an object's arrival in that place? Here, too, Buddhism sees no need to accept the existence of some manas-like entity that travels to a certain location, let's say the eye, waits there, and
then encounters an object that appears in the line of vision of the eye. Rather what we have are just two separate, but temporally contiguous, awarenesses. In the first one, one sees whatever falls in the location to which the eye is pointed, possibly just empty space (or in certain psychology experiments concerned with attention, a blank screen), and in the second, one sees whatever new object then appears in that location. So again, instead of a two-tier
model, with both a sequence of awarenesses and on top of them a separate (and a unitary) thing called attention, we have just a one-tier model. But how does Buddhism capture the sense of something here waiting and then receiving? It claims just that the first moment is characterized by alertness and expectation, and in the second moment an object is perceived at the expected location. (
3) But there still seem to be aspects of these two examples that the Buddhist model has not explained. In the first example there is a sense of a decision being made to withdraw from the present sense-object and shift to a new one. Similarly, in the second example surely something directs the
attention to the expected location. There is a sense of one's awareness or one's attention being placed there. The Buddhist model seems quite able to explain a sequence of awarenesses characterized by passivity, where a sequence of objects impinges on them, but not so suited to explain those times when
the subject consciously directs their awareness in certain directions. We will see below how Buddhism explains these cases, but let us first ask whether this sense of consciousness or the subject of consciousness directing
attention/awareness is perhaps an illusion. Let us here introduce a third Indian tradition, that of Sarikhya, which answers this question in the affirmative. For the Sarikhyas, though we (i.e., the part of us that is conscious) may think we are controlling the direction of our attention, it is in
fact determined wholly by unconscious habits. The self, which is characterized by consciousness, looks on, but contains no agency whatso¬ever: impulses leading to physical or mental action, including the directing of attention, do not issue from it. But the self, though in reality capable only of purely passive observation, superimposes onto itself, falsely arrogates to itself, responsibility for these actions, such as the directing of attention.
Unconscious habit directs the attention in a certain way, we become aware of our attention having shifted, and we say to ourself “Ididthat”19 (see figure . This Sarikhya view reminds one of Benjamin Libet's experiments in Berkeley. Subjects were asked to perform simple gestures while their neuronal activity was measured. What he found was that a split second before a subject makes the decision to perform the gesture, neurons already transmit the order to the
hand to do it. This has been taken to suggest that when our consciousness thinks that it is freely and autonomously performing a task, rather what is happening is that on becoming aware of an impulse that started at the neuro-biological level, quite independently of its will, consciousness gives itself
credit for initiating that impulse. In this view then, as in Sarikhya, consciousness retrospectively explains an impulse as having issued from itself, despite that impulse having a quite different origin.
How does the Buddhist model differ from this Sarikhya model? A comparison of figures 3 and 4 will highlight the two main differences. (1) In Buddhism there is no self. (2) Consciousness in Buddhism is not passive and impotent, leaving all shifts of attention to be brought about by habit. Intentions (“I” in figure 4) within the stream of consciousness can and do bring about shifts of attention. That provides the answer to the question we
posed earlier of how Buddhism can explain times when the subject consciously directs their awareness in certain directions. How does the Buddhist model differ from the Nyaya-Vaisesika model, in particular with regard to the amount of conscious control over the directing of attention? In the Nyaya-Vaisesika model the self's control over the direction of attention is total (see figure 5). In the Buddhist model, by contrast,
conscious¬ness' control over the direction of attention is far from total, since conscious¬ness' intentions are merely momentary phenomena, which, though they may determine what is attended to in the very next moment, cannot guarantee that attention will stay focused on the object in question after that. Thus, although both the Nyaya-Vaisesika and the Sarikhya views, unlike the Buddhist, assert the existence of a self, there is a sense in which they occupy the two extremes, with Buddhism falling in the middle.
Setting aside the Sarikhya view, let us further develop the contrast between the Buddhist and the Nyaya-Vaisesika views. In the latter case we have one thing, the self, which is responsible for every shift of attention, as well as every instance of attention staying in the same place. We have a top-down
model in which the self's attentional commands are carried out by the manas, an instrument that automatically obeys these commands. The Buddhist model, by contrast, is characterized not by vertical hierarchy, but by horizontality: shifts in attention are not brought about by commands from above, but by the
previous moment in the stream of consciousness (or by the forceful arrival of a new sense-object within the range of the individual). One could add a further contrast between a “personal” and an “impersonal” model. Shifts of attention are explained by the Nyaya- Vaisesikas on the analogy
of a master ordering a servant; attentional causation issues from a person or a person-like self. Buddhism, by contrast, constitutively denies the existence of any self or person (atman, pudgala, purusa) either in the mind-stream or as the substrate of the mind-stream. Shifts of attention within this
mind-stream result from an impersonal causation that operates between mental events. These mental events may include conscious intentions, but these are not the conscious intentions of a self or person; they are just events within a mind-stream.
IV. Competition for Attention
We have been looking at the issue of “shifts of attention,” that is, at what precisely happens when one's attention moves away from one object to a different object. Now I would like to “shift attention” to a related but slightly different issue, that of “competition for attention.” When any number of potential things are available to be attended to, what is it that determines which one “wins,” that is, succeeds in being attended to?
The Nyaya-Vaisesika answer is a relatively straightforward one: it is the self that arbitrates, and comes down on the side of one among the various candidates. It then puts its decision into action by dispatching the manas to the location of the sense-faculty whose object it has chosen to attend to. For Buddhism these competitions for attention can be decided without any top-down arbitration. Let us separate out two different kinds of case.
(A) Perception. There are all sorts of things in one's visual field at any one moment; there are also potential objects of other senses—noises, smells, et cetera. What is it that determines what will be attended to? Here it is quite easy to come up with some principles that obviate the need to appeal to top-down arbitration. A sense-object that poses a threat to the individual will grab its attention more than a simultaneous one that does not. An object of desire will attract attention to itself more than one for which the subject has no strong feelings.
(B) Cognitions other than perception. In the Buddhist model, the stream of consciousness unfolds moment by moment. Whatever arises—a memory, a daydream, a thought, et cetera—will be what is attended to. Thus, the question of what determines what the attention will shift to is equivalent to the question of what determines what kind of moment of consciousness will arise, that is, what the content of the next moment of consciousness will be
To give an indication of Buddhist thinking here, we can observe a passage from the Abhidharmakosabhasya.20 Vasubandhu asks us to imagine that the idea of a particular woman arises in someone's mind-stream; what will be attended to next? It will depend on what latent impressions are contained in that mind-stream, as these will determine what the primary associations are to that woman. If we are talking of a monk, perhaps abhorrence of her body will arise. Or if we are talking of a layperson who is a friend of the woman's husband or son, perhaps an image of the husband or son will emerge.
Vasubandhu indicates some principles as to which out of a number of different associations are likely to come to the surface. Out of the many associations to the particular woman, those which are common, intense, or recent are most likely to be activated, since those latent impressions will be more powerful
than the latent impressions of less common, less intense, or less recent associations. But even if there is a common, intense, or recent latent impression available and on the point of reaching fruition, it will be interrupted by the advent of a sufficiently powerful bodily sensation or external sense-object; in that case the emergence of an association to that woman will be inhibited.21
So when there is a competition between various potentially available objects of attention, Buddhism feels no need to appeal to any top-down arbitration. There is neither a manas in contact with the chosen object of attention nor a self making a choice about where to send the manas. There is simply a competition in which the most forceful candidate emerges as the winner.
In order to make this absence of top-down control plausible, I give here two analogies. In a running-race the winner is simply the one who crosses the finish line first; it is not that competitors have to stop before the finish line and be ushered across at the discretion of the referee. Similarly, it is just built into our psychic system that whichever latent impression advances toward the level of consciousness most forcefully will make it through and block the way for others. It is unnecessary to postulate some agent at the threshold—whether the person, or some homunculus-like entity at a sub¬personal level—that is needed to give its approval before entry to conscious¬ness can be achieved.
Or one could compare the process to a conversation, where each of the participants in the conversation stands for a latent impression seeking actualization (the actualization of a latent impression will result in a particular memory, emotion, judgment, inference, dream-image, etc.). Talking stands for the
actualization of a latent impression. There is usually only one person talking at once. And what they say will be most compelling if it relates in some way to what has just been said. (I.e., latent impressions whose content closely relates to the content of the present moment of consciousness will be imbued with more power to emerge than those whose content does not.) Usually there will be more than one person wanting to talk at once. But a person will not say
what they want to say at exactly the same time that someone else is talking, for then they will not be heard. The stronger-willed people will persist at trying to jump in until they are heard. The weaker-willed ones will give up if too much noise and energy is coming from other people. The point is that there is no arbiter standing over all these people and saying, “Okay, now you,” “Okay stop, now time for someone else.” The agency comes
from the participants alone. What unfolds is simply the result of the interplay between these participants and their energies; it is not planned by an overseeing authority, or monitored by a superintendent. Similarly, a plurality of potential contents of consciousness can jostle with each other and yield a winner without a self playing the role of arbitrator.
V. An Objection to the Horizontal Buddhist Model
There is a consensus among experimental psychologists working on attention that it can be both involuntary (i.e., unintended) and voluntary (consciously intended), both exogenous (stimulus-driven) and endogenous (internally controlled).22 The Buddhist model seems well placed to explain the involuntary and
the exogenous, but how about the voluntary and endogenous? Surely there are occasions when the subject of consciousness does get involved, considers different options, comes to a decision about what to attend to, and then attends to it. Does not the absence of top-down control in Buddhism prevent it from explaining such cases?
Buddhism can explain such cases. What happens there is that the content of one, or more, moment(s) of consciousness is the weighing of different options, the content of the next is an intention to attend to one out of the options, and the content of the next is a focus on the chosen option. But this is not top-down control; it is a horizontal process of unfolding.
In Buddhism's horizontal functionalism there is nothing “above” the stream of consciousness, such as a self, that could exert “downward” control. But individual moments of consciousness can and do exert control over what is attended to in the next moment. The phenomenology associated with times of conscious control of attention is surely compatible with both of these metaphysical pictures and so cannot decide decisively in favor of one over the other. But it is arguably to the advantage of Buddhism that it explains voluntary and endogenous attention without appeal to any homunculus-like entity.
VI. Objections to the Top-Down Nyaya Mode
Here are three objections that might be leveled against the top-down Nyaya- Vaisesika model. All three are related to the fact that while this model has no problem explaining voluntary or endogenous movements of attention, it seems not so well suited to explaining the involuntary or exogenous.
(1) Both phenomenological investigation and empirical research reveal that attention is not always, in fact not usually, directed by conscious decisions on the part of the subject. When a sense-object arrives with force within the range of the sense-faculties, attention seems to operate like a reflex, without the intervention of the knowing subject. And even when the mind is occupied with thoughts, ideas, daydreams, worries, memories, surely it is a small minority of shifts of attention that are preceded, and seemingly brought about, by a conscious decision and command on the part of the subject of consciousness.
(2) If every shift of attention, as well as every instance of attention staying in the same place, were the result of the will and command of the self, surely there would never be any discrepancy between what is being concentrated on and what the subject wants to be concentrating on. If attention were under the sole control of the self, it would never move away from an object until ordered to do so by the self. How, then, can we account for, for example, the disappointment felt by a subject at not having been able to keep their attention on something, for example while meditating, writing a paper, learning Sanskrit paradigms. The explanation of this is difficult if there is a self keeping the manas i n the same place until it is satisfied.
(3) There is a problematic duality in the concept of the manas, as both that which responds to the self's commands and that which enables the self to have the information on the basis of which to issue its commands. The self needs it to become aware of the objects of the senses. Yet surely the self already needs to have that awareness in order to make informed decisions about where to dispatch it.
When an object erupts into the range of the subject with force—a loud noise for example—surely we do not want to say that before the self becomes aware of the noise it first has to direct the manas to the ear. How would it know to direct the manasthere unless it first knew that there was a noise to be listened to? But if it does somehow detect the noise before dispatching the manasthere, why dispatch the manasthere at all?
Nyaya does have an answer to these three objections. Vatsyayana and Uddyotakara mention the following three pertinent examples. One may be wakened from sleep by a loud noise or by being shaken. While awake, one's attention may be dragged, against the will of the self, which is enjoying a certain object, to a new object that has arrived with force. A person feels a sudden pain in the foot caused by inadvertently stepping on a thorn or a pebble.23 How can Nyaya explain these cases? Surely it cannot say that the self directs the manas to the relevant sense-faculty. For none of these examples involves the
premeditated allocation of attention to an object that the self would like to focus on. But if the manas does not travel to the relevant sense-faculty, then one of the conditions held by Nyaya-Vaisesikas to be required for awareness—contact between the manas and the sense- faculty—would be lacking; so the person would not feel the shaking, hear the noise, or experience the pain in the foot. Both commentators assert that in these three cases the manas moves to the sense-faculty as a result of adrsta, the “unseen” force that is equated with a self's dharma and adharma, that is, its stock of good and bad karma.24
So here we have an answer not only to the first two objections about nonconsciously controlled shifts of attention, but also to the issue in the third of how the self can have any awareness at all of loud noises, say, before it makes the decision to send the manas there. The answer is: the manas can be dragged there, independently of any decision on the part of the self, by adrsta.
Those who are not attracted by Nyaya-Vaisesika's wider ontological commitments may find this an unilluminating explanation of the workings of attention. But setting that aside, there is a difficulty with it even from the point of view of Nyaya ontology. A common Nyaya-Vaisesika argument for the self is that the manas, being unconscious, requires something conscious to direct it. To attribute shifts of the manas to adrsta, that is, karma, seems to contradict this principle that the manas requires a conscious director; for karma is unconscious. Indeed, its unconscious nature is stressed by Naiyayikas when they argue for the existence of God on the grounds that karma, being unconscious, requires a conscious being to bring about its implementation.
I see two ways in which Nyaya could neutralize this objection. (1) They could maintain that the adrsta moving the manas i s governed by God, for that would be consistent with the Naiyayika and Vaisesika argument that the movement of atoms at the beginning of creation, even if it has adrsta as its immediate cause, must be governed by something conscious, namely God. (2) They could point out that adrsta is a quality of the self, so that a movement of the manasbrought about by adrstaisdirected by the conscious self at least in thesensethatitisdirectedbyaqualityoftheconsciousself.
The dilemma that faces Nyaya here is that, on the one hand, it needs exogenous movements of the manas to be directed by something conscious because of its general principle that movement of anything unconscious must be brought about by something conscious. On the other hand, what is required here is an explanation of exogenous shifts of attention; and if the explanation is that such shifts are brought about by the conscious self, they no longer seem to be exogenous.
VII. Final Remarks
Before finishing, I point to three things, sometimes asserted of attention, which may at first sight appear to be contradicted by both the Naiyayika and Buddhist models.
1. We can pay attention to more than one thing at once.
2. We can be simultaneously attentively aware of one thing and non- attentively aware of other things (e.g., data being received by other sense faculties).25
3. Attention admits of degrees; we can pay a lot of attention or a little bit of attention.
It might look as though both the Naiyayikas and the Buddhists will have to deny all of these three. For they both assume models in which awareness has only one object. Since the manas is the size of a mere atom, it cannot be in contact with more than one sense-faculty at any one moment. It is precisely this
fact that is emphasized by Naiyayikas when they put forward the manas as a bottleneck that explains why we are not simultaneously flooded with data from more than one faculty. If the manas is only ever in contact with one faculty at a time, and there is no awareness without the involvement of the manas,
then surely 1 and 2 are impossible; in any one moment full attention is paid to one object, and there is no awareness whatsoever of any other. Neither can attention admit of degrees, because the manas i s either in contact with a sense-faculty or not; it cannot be partly in contact with it, strongly in contact with it, or weakly in contact with it.
In the Buddhist model, too, each moment of consciousness has only one object; it certainly cannot be simultaneously directed at data from two different senses (1). If awareness of an object has arisen, cognition must be paying attention to it, that is, turned toward it; there can be no non-attentive awareness of an object (2). And attention, being a dharma, is either completely present or completely absent; it cannot come in different strengths (3).
We need to distinguish between what, phenomenologically, seems to be the case and what is actually the case. Then we can see that denials of 1, 2, and 3 can take two different forms. A denial of 1 targeted at the phenomenological level will claim that if we focus closely on our experience, we will find that we only ever seem to be attending to one thing at once. A denial of 1, targeted at the level of reality, will claim that though we may appear to ourselves as paying attention to more than one thing simultaneously, this is not what is actually happening
. Both Naiyayikas and Buddhists deny 1 on the level of reality, but do not deny it on the level of phenomenology. How, then, do they explain that the situation is phenomenologically presented to us in a way that misrepresents reality? The Naiyayikas appeal to the speed of the manas.26 Though it may
appear to me that I am simultaneously delighting in the taste of food and the sound of music, the manas is in fact rapidly traveling back and forth between my mouth and my ears. It travels so fast that the appearance of simultaneity results. The Buddhists appeal to the brevity of each moment of consciousness. Every second is composed of hundreds of discrete aware¬nesses, some having as their object the food and some the music. Let us say that thirty awarenesses
in a row are focused on the food, followed by thirty on the music, followed by thirty on the food. The period in the middle during which the food is not being focused on amounts to such a short period of time that it appears to us as though our awareness never left the food. Thus, the appearance of simultaneity results.
So much for the Naiyayika and Buddhist attitude to 1. I have not encountered discussions of 2 or 3 in primary sources. But their responses could easily resemble their responses to 1. They could affirm 2 and 3 phenomenologically, but deny them on the level of reality. What would they say is really going on
when we seem to be simultaneously attentively aware of one thing and non-attentively aware of other things? The Naiyayikas could assert that when, for example, we are paying attention to a painting, but dimly (i.e., non-attentively) aware of the background noise in the art gallery, the manas is spending the bulk of the time at the eye and a small proportion of the time at the ear. The Buddhists could assert that the vast majority of moments of
consciousness are aware of the painting, but a small proportion are aware of the noise. How would the two traditions explain seeming variation in the intensity of attention? A period of extremely intense attention could be explained by the Naiyayikas as resulting from the uninterrupted presence of the manas at the location of one sense-faculty, and less intense attention as resulting from
presence interrupted by absences. Similarly, intense attention for the Buddhists would result from a long sequence of unidirectional awarenesses, and less intense attention from interruptions in such a sequence by awarenesses focused on data from different senses. Thus, to argue against the Naiyayika and the Buddhist views on the grounds that they implausibly deny 1, 2, and 3 neglects that they are quite capable of
explaining the phenomenology that leads us to assert 1, 2, and 3.27
I am grateful to the following people who took the time to read drafts of this essay and from whose comments I benefited enormously: Amit Chaturvedi, Christian Coseru, Matthew Dasti, Jonardon Ganeri, Martin Lin, Philipp Maas, Roy Perrett, Kranti Saran, Mark Siderits, Sebastian Watzl, and the two anonymous reviewers for PhilosophyEastandWest.
1 - For overviews of the field, see Mole 2013, Watzl 2011a, and Watzl
2 - I am using this hyphenated form “Nyaya-Vaisesikas” simply as a short way of referring to both Naiyayikas and Vaisesikas. I do not mean to imply that there was a single composite school comprising both. But on the issues discussed in this essay, I do not find any significant divergences between the two.
3 - For this date see Vol. 2 of Deleanu (2006: 186-194).
4 - Garigesa and Raghunatha, e.g., introduce innovations in their treatment of the manas.
5 - See NyayakandalT, p. 175, 6: manas cetanam na bhavati, karanatvad ghatadivad iti, “The manas is not conscious, because it is an instrument, like a pot or such like.”
6 - Cf. Broadbent 1958, Posner and Boies 1971, and Posner and Warren 1972.
7 - Attention (manaskara) is named (e.g., at Abhidharmakosabhasya 2.24) as one of ten mental factors that accompany every cognition (citta).
8 - Abhidharmakosabhasya, p. 1222, 9 ff.: idam sphutam codyam apa-dyate kasman na nityam tadrsam evotpadyate. . . , “Clearly the following question arises here: why does [[[consciousness]]] not arise always in the same form?” The passage has been translated by Sanderson (1995, pp. 47 ff.), Duerlinger (2003, pp. 102 ff.), and Kapstein (2001, pp. 370 ff.).
9 - Abhidharmakosabhasya, p. 1222, 10.
10 - Abhidharmakosabhasya, p. 1223, 4-1224, 2: nityam avisiste manasi
katham samyogavisesah.? buddhivisesapeksa iti cet, sa eva paricodyate: katham buddhivisesa iti? samskaravisesapeksad atmamanah.samyogad iti cet, cittadevastu samskaravisesapeksat (Lee edition, Dwarika Das SastrT edition, supported by Yasomitra's commentary; -apeksatvat Pradhan edition), “Since the
manas is permanently unchanging [like the self], how can there be different conjunctions [between the two of them]? If [the Vaisesika] says, ‘[the conjunctions] are associated with different cognitions,' the same question [with which this whole passage started] should be put: how can the cognitions
differ? If [the Vaisesika] answers that specific cognitions are derived from connections between the self and the internal organ that depend on specific latent impressions, then why should he not derive them from consciousness itself depending on specific latent impressions?”
11 - Abhidharmakosabhasya, p. 1224, 2-3, and Yasomitra's Sphutartha, ad loc. The Bhasya reads: na hi kimcidatmana upalabhyate samarthyam ausadhakaryasiddhav iva kuhakavaidyaphuh.svahanam (Pradhan edi¬tion; -phuhsvahanam Dwarika
Das SastrT edition), “For we do not see that the self would have any power [to influence the production of consciousness], just as [there is no evidence that] the phuh.svahas uttered by quack doctors [contribute] to the efficacy of the herbs [they prescribe].” I follow Sanderson 1995 for both the choice of readings and the translation.
12 - The term manas is used by Vasubandhu, and in Abhidharma more generally, to elucidate the Buddhists' own Philosophy of Mind, but it never refers there to a substantial entity like the Nyaya-Vaisesika organ of
attention. It is used as a synonym of citta and vijnana, that is, to refer to cognition or consciousness. See Abhidharmakosabhasya 2.34ab (cittam mano 'tha vijnanam ekartham); Dhammajoti 2007, pp. 101 -102, citing A bhidh arm a va tara; a n d Coseru 2012, p. 69 n. 46). Cf. Galloway 1978.
13 - Manaskara (standardly translated as “attention”) is, for Buddhist Abhidharma, as stated above, that feature of a cognition that turns or orients it toward an object (cetasaabhogah.,Abhidharmakosabhasya,ad 2.24). See Dreyfus
2011 for an elaboration of the way that mindfulness (smrti) and concentration/absorption (samadhi), which both contribute to the continued attentiveness of
a mental state toward its object, are also part of what we mean in English by “attention.” All three of these, according to Abhidharmakosabhasya 2.24, necessarily accompany every mental state. Awareness, for Buddhism, consists of on the one hand the awareness itself (citta), and on the other the various features that necessarily accompany it and qualify it, e.g., its hedonic tone (vedana), th e fact that it is turned toward its object (manaskara).
14 - See Prasastapadabhasya, p. 17, 1-2; K. K. Chakrabarti 1999, p. 104;and A. Chakrabarti 2005, pp. 126-127. The latter (2005, p. 126) sees the first instance of this argument as Brhadaran.yakopanisad1.5.3.
15 - Nyayasutra 1.1.16, Nyayabhasya and -varttika ad loc., and K. K. Chakrabarti 1999, pp. 103-104.
16 - Since the connotations that the term “sense-data” has acquired in contemporary Philosophy of Perception mean that readers familiar with that literature will likely take it to refer to data that is by definition mind-dependent, I am avoiding that term.
17 - Nyayabhasya, ad 3.1.16; K. K. Chakrabarti 1999, pp. 104-105.
18 - There were a few things, though, that Vasubandhu and other Buddhist ontologists granted reality to, despite being imperceptible, e.g., the sense-faculties.
19 - These comments about Sarikhya's metaphysics have to be reconciled with the fact that Sarikhya-Yoga teaches a spiritual path that involves the gradual training of attention. A hierarchy of levels of development of cognition (cittabhumi) is given (ksipta, mud.ha, viksipta, ekagra, niruddha) at the very outset of the Patanjalayoga- sastra (1.1). Maas (2009) has
demonstrated that the rationale behind this categorization is the ability of the mind to deliberately choose the object of its attention: this ability does not exist in the two first- mentioned stages; it appears to some degree in the third, viksipta; and it is fully developed in the ekagra and niruddha
stages. I do not know how Sarikhya-Yoga would reconcile its teaching of ever greater control over attention with its metaphysical commitment that the self, lacking any agency whatsoever, plays no role in directing attention. Perhaps the training of attention takes place not at the leveloftheconsciousself(purusa), but at that of unconscious cognition (citta).
20 - Abhidharmakosabhasya, pp. 1220-1222. The passage has been (differently) translated by Sanderson (1995, pp. 46-47), Duerlinger (2003, pp. 101-102), and Kapstein (2001, pp. 369-370), and summarized by Gold (2011, pp. 13-
21 - After having given this indication of some principles that govern what contents of consciousness arise, that is, what is attended to, Vasubandhu cites a verse to the effect that this is a mere indication of the workings of the mind. A complete knowledge of the causes that determine the unfolding of mind-streams is the domain only of omniscient ones. Without omniscience we cannot
know even all of the causes of a single eye in a peacock's tail, let alone the much more subtle matter of all of the causes ofthe particular unfolding of a stream of consciousness.
22 - Watzl 2011b, p. 846: the two distinctions (voluntary-involuntary and endogenous-exogenous) are not equivalent; daydreaming is arguably endogenous but involuntary.
23 - The first two examples are given in the commentaries to Nyayasutra
2.1.27; the third in the commentaries to Nyayasutra 3.2.32. I am very grateful to Matthew Dasti and Amit Chaturvedi for directing me to these passages.
24 - Nyayabhasya (p. 72, lines 8-13) and Nyayavarttika (p. 199, line 18, to p. 200, line 3) argue that if adrsta were not able at the beginning of creation to move both the atoms and the many manas, then bodies, sense¬faculties, and
objects could not be formed. The implication is that since adrsta can move the various manas at the beginning of creation, then it can also be the cause of those shifts of the manas during an individual's life that take place independently of the self's desire and effort.
25 - On the question of whether we can be aware of something that we are not paying attention to, see, e.g., Mack and Roch 1998, Simons and Chabris 1999, Braun 2009, and the many references given by Watzl (2011a, pp. 726-727).
26 - Nyayabhasya ad 3.2.57-8 and NyayamanjarT, Vol. 2, pp. 409-410.
27 - If Naiyayikas and Buddhists accept 1, 2, and 3 at the phenomenological level—if they do not deny that multifocal attention, inattentional awareness, and degrees of attention are all phenomenologically pre¬sented to us—does the burden of proof not lie with them? They may have an account of what is happening at the level of reality that, despite contrasting with the phenomenology,
is compatible with it; but given a choice between two versions of what is going on at the level of reality, why not opt for the one that honors what is suggested by the phenomenology? Here it is relevant to point to recent experimental work in psychology that calls into doubt our pre-theoretical confidence
in the reliability of the phenomenology of attention as a guide to its real nature. See, e.g., Blackmore 2004, pp. 51-77; Lavie 2007; Noe 2007; and Schwitzgebel 2008. This research does not, of course, provide support for the Nyaya or Buddhist views, but it does imply that phenomenological acceptance of 1, 2, and 3 itself provides little or no evidence for 1, 2, and 3 atthelevel of reality. (I owe this footnote to Roy Perrett, whom I thank.)
Abhidharmakosabhasya. Page numbers refer to the Dwarika Das SastrT edition; readings of the Lee edition and Pradhan edition are reported when they differ: Dwarika Das SastrT edition: Abhidharmakosa and Bhasya of Acarya Vasubandhu with Sphutartha Commentary of Acarya Yasomitra. 1981. Edited by Swami Dwarika Das SastrT. Bauddha BharatT Series 5, 6, 7, 9. Varan. asT. Lee edition: Abhidharmakosabhasya of Vasubandhu:ChapterIX:Atma- vadapratisedha. 2005. Edited by Jong Cheol Lee. Bibliotheca Indologica et Buddhologica 11. Tokyo: Sankibo Press.
Pradhan edition: Abhidharmakosabhasya of Vasubandhu. (1967) 1975. Edited by P. Pradhan, Tibetan Sanskrit Works Series 8. Jayaswal Research Institute, Patna. Second, revised edition by Aruna Haldar, without improvement of the text. Nyayasutra. In Nyayabhasya.
Nyayabhasya. 1997. GautamTyanyayadarsana with Bhasya of Vatsyayana. Edited by Anantalal Thakur. Nyayacaturgranthika 1. New Delhi.
NyayakandalT. (1895) 1984. In The Padarthadharmasahgraha of Prasastapada with the NyayakandalT of SrTdhara, edited by V. P. Dvivedin. Viziana- garam Sanskrit Series 6, 1895.
NyayamanjarT. 1969 and 1983. NyayamanjarT ofJayantabhatta, with T.ippan.i —Nyayasaurabha by the Editor. Edited by K. S. Varadacharya. 2 volumes. Oriental Research Institute Series 116, 139. Mysore.
Nyayavarttika. 1997. Nyayabhasyavarttika of Bharadvaja Ud¬dyota¬kara. Edited by Anantalal Thakur. Nyayacaturgranthika 2. New Delhi. Prasastapadabhasya. 1994. Word Index to the Prasastapadabhasya: A Complete WordIndex to the Printed Editions ofthe Prasastapadabhasya. Edited by Johannes Bronkhorst and Yves Ramseier. Delhi.
Sphutartha. (1932-1936) (1971) 1989. Sphutartha Abhidharmakosavyakhya: The WorkofYasomitra. Tokyo: Sankibo Press. Secondary Sources
Blackmore, Susan. 2004. Consciousness: An Introduction. New York: Oxford University Press. Braun, Jochen. 2009. “Attention and Awareness.” In The Oxford Companion to Consciousness, edited by Tim Bayne, Axel Cleeremans, and Patrick Wilken, pp. 68-72. Oxford: Oxford University Press.
Broadbent, D. E. 1958. Perception and Communication. New York: Pergamon Press. Chakrabarti, Arindam. 2005. “Manas: In Defence of the Inner Self.” In Towards a Spiritual Psychology: Essays in Indian Psychology, edited by K. Ramakrishna Rao and Sonali Bhatt Marwaha, pp. 123-137. Delhi: Samvad India Foundation.
Chakrabarti, Kisor Kumar. 1999. Classical Indian Philosophy of Mind: The Nyaya Dualist Tradition. Albany: State University of New York Press. Indian edition, Delhi: Motilal Banarsidass, 2001.
Coseru, Christian. 2012. PerceivingReality:Consciousness, Intentionality, and CognitioninBuddhistPhilosophy. New York: Oxford University Press. Dasti, Matthew. 2013. “Nyaya's Self as Agent and Knower.” In Free Will, Agency, and Selfhood in Indian Philosophy, edited by Matthew R. Dasti and Edwin F. Bryant, pp. 112-136. Oxford: Oxford University Press. DOI: 10.1093/acprof:oso/9780199922734.003.0006
Deleanu, F. 2006. The Chapter on the Mundane Path (Laukikamarga) in the Sravakabhumi. A Trilingual Edition (Sanskrit, Tibetan, Chinese) Anno¬tated Translation and Introductory Study. 2 Vols. Tokyo: The Interna¬tional Institute for Buddhist Studies.
Dhammajoti, Bhikku K. L. 2007. Abhidharma Doctrines and Controversies on Perception. Third revised edition. Hong Kong: Centre of Buddhist Studies, University of Hong Kong.
Dreyfus, G. 2011. “Is Mindfulness Present-Centered and Nonjudgmental? A Discussion of the Cognitive Dimensions of Mindfulnesss.” Contemporary Buddhism 12, no. 1:41-54. DOI:10.1080/14639947.2011.564815
Duerlinger, James. 2003. Indian Buddhist Theories ofPersons:Vasubandhu's “Refutation ofthe Theoryofa Self.” London: RoutledgeCurzon. Galloway, Brian. 1978. “Vijnana, Samjna and Manas.” The Middle Way 53, no. 2:72-75.
Gold, Jonathan C. 2011. “Vasubandhu.” In Stanford Encyclopedia ofPhiloso- phy. http://plato.stanford.edu/archives/sum2011/entries/vasubandhu/. Kapstein, Matthew T. 2001. Reason's Traces: Identity and Interpretation in Indian and Tibetan Buddhist Thought. Studies in Indian and Tibetan Buddhism. Boston: Wisdom.
Lavie, Nilli. 2007. “Attention and Consciousness.” In The Blackwell Companion to Consciousness, edited by Max Velmans and Susan Schneider, pp. 489-503. Oxford: Blackwell Publishing.
Maas, Philipp André. 2009. “The So-called Yoga of Suppression in the Pätanjalayogasästra.” In Yogic Perception, Meditation and Altered States of Consciousness, edited by Eli Franco, in collaboration with Dagmar Eigner, pp. 263-282. Vienna: Verlag der Österreichischen Akademie der Wissenschaften, 2009.
Mack, Arien, and Irvin Rock. 1998. InattentionalBlindness. Cambridge, MA: MIT Press. Mole, Christopher. 2013. “Attention.” In Stanford Encyclopedia of Philoso¬phy. http://plato.stanford.edu/entries/attention/. Originally written in 2009, substantially revised in 2013.
Noë, Alva. 2007. “Inattentional Blindness, Change Blindness, and Conscious¬ness.” In The Blackwell Companion to Consciousness, edited by Max Velmans and Susan Schneider, pp. 504-511. Oxford: Blackwell Publish¬ing. Posner, Michael I., and Stephen J. Boies. 1971. “Components of Attention.” PsychologicalReview78, no. 5:391-408. Posner, Michael I., and R. E. Warren. 1972. “Traces, Concepts, and Conscious Constructions.” In Coding Processes in Human Memory, edited by Arthur W. Melton and Edwin Martin, Chichester: Winston and Wiley.
Sanderson, Alexis, ed., trans., and annot. 1995. “Passages in the 5th and 9th Chapters of Vasubandhu's Abhidharmakosabhäsya.” Oxford: Unpub¬lished Teaching Material for M.Phil in Classical Indian Religion.
Schwitzgebel, Eric. 2008. “The Unreliability of Naive Introspection.” PhilosophicalReview117:245-273. Simons, Daniel J., and Christopher F. Chabris. 1999. “Gorillas in Our Midst: Sustained Inattentional Blindness for Dynamic Events.” Perception 28:1059-1074.
Watzl, Sebastian. 2011a. “The Philosophical Significance of Attention.” Philosophy Compass 6, no. 10:722-733. ———. 2011b. “The Nature of Attention.” Philosophy Compass 6, no. 11:842-853.