| |

Received: 28 March 2018    Revised: 6 July 2018    Accepted: 24 April 2018

DOI: 10.1111/rati.12202

SPECIAL ISSUE ARTICLE

Attention and self in Buddhist philosophy of mind

Jonardon Ganeri

College of Arts and Science, New York

University and Philosophy Programme, New
York University Abu Dhabi, New York, NY
Correspondence
Jonardon Ganeri, College of Arts and
Science, New York University and Philosophy
Programme, New York University Abu Dhabi,
19 Washington Square North, New York, NY
10011, USA.
Email: jonardon.ganeri@nyu.edu
Abstract
Buddhist philosophy of mind is fascinating because it de‐
nies that there is a self in one of the ways that has tradi‐
tionally seemed best able to make sense of that idea: the
idea that the self is the agent of actions including the
thinking of thoughts. In the Buddhist philosophy of mind
of the fifth century thinker Buddhaghosa what does the

Funding information
Research for this article was partially funded
by the University of Cambridge New Directions
in the Study of Mind Project and by the New
York University Virtues of Attention Project.
explanatory work is instead attention. Attention replaces
self in the explanation of cognition’s grounding in percep‐
tion and action; it does this because it performs two func‐
tions at once, a function of placing and a function of
focussing. A comparison between the thought of
Buddhaghosa and Brian O’Shaughnessy helps to clarify
what is at stake in the residual employment of the concept
self as a derived rather than a primitive notion.

KEYWORDS
attention, Buddhaghosa, Buddhist philosophy, O’Shaughnessy,
self

1 |   B U D D H I S T S AG A I N S T S E LF

Buddhist philosophy of mind is fascinating because it denies that there is a self in one of the ways that has tradition‐
ally seemed best able to make sense of that idea: the idea that the self is the agent of actions including the think‐
ing of thoughts. It is in this sense that experience might be said to have a subject. In Buddhist philosophy of mind
what does the explanatory work is instead attention. Buddhaghosa, for instance, is emphatic that there is no self as
normally conceived: ‘The self of the sectarians does not intrinsically (sabhāvato) exist’ (Dispeller 77). A fifth century
philosopher who worked at the university of Anurādhapura in Sri Lanka, Buddhaghosa has been described as pos‐
sessing ‘one of the greatest minds in the history of Buddhism’. A cosmopolitan intellectual, he moved between India
and Sri Lanka, between Sanskrit and Pāli, between Hinduism and Buddhism, in search of a fundamental theory of

Ratio. 2018;1–9. wileyonlinelibrary.com/journal/rati © 2018 John Wiley & Sons Ltd  |  1
 |
2       GANERI

mind. Buddhaghosa’s ideas would, indeed, influence conceptions of the human throughout South and South‐East
Asia for a millennium and a half, and they continue to do so today. One may say that his writings are the distillation
of a thousand years of observation and reflection in the context of a research programme initiated by Śākyamuni,
the Buddha himself. They are more than that, though: they are also testimony to a true innovator, a pioneer and a
creative thinker. Having reviewed rival hypotheses of three of his predecessors he wonders who is right and an‐
swers, pointedly, ‘Nobody: we should accept what is right in the claims of each’ (Fount 287). He was an innovator, and
self‐consciously so, sometimes openly declaring that he was going beyond anything that can be found in the older
commentaries, even acknowledging that his new thoughts had not yet gone far enough. Many of his original ideas
are in commentaries on the canonical Abhidhamma‐piṭaka, especially his Fount of Meaning (Atthasālinī, his commen‐
tary on the Dhammasaṅgani) and Dispeller of Delusion (Sammoha‐vinodanī, his commentary on the Vibhaṅga)—and
perhaps he permitted himself a little greater philosophical licence there than in his manual, the Path of Purification
(Visuddha‐magga), a synthetic and comprehensive description of the Buddhist path, or in ‘his’ synoptic commentaries
on the Sutta‐piṭaka (the Majjhima‐, Saṃyutta‐, Aṅguttara‐ and Dīgha‐Nikāya aṭṭhakathās).
As I have said, Buddhaghosa denies that there is any such thing as a self: ‘[the Wheel of Existence] is devoid of any
self as an experiencer of pleasure and pain conceived as “this self or mind which speaks and feels” (M.i.8). This is how
it should be understood to be without any maker or experiencer’ (Dispeller 190); and ‘For this is said with reference
to such feeling as is accompanied by clear comprehension of [the question]: “Who feels? Whose feeling is it? For what
reason do these feelings come to be?” Herein, who feels? No being or person feels. Whose feeling is it? Not the feeling
of any being or person.’ (Dispeller 263). Instead of self there is only the ‘minded body’ (nāma‐rūpa):

In many hundred suttas it is only minded body that is illustrated, not a being (satta), not a person
(puggala). Therefore, just as when the component parts such as axles, wheels, frame poles, etc., are
arranged in a certain way, there comes to be the mere term of common usage ‘chariot’, yet in the
ultimate sense when each part is examined there is no chariot,—and just as when the component
parts of a house such as wattles, etc., are placed so that they enclose a space in a certain way,
there comes to be the mere term of common usage ‘house’, yet in the ultimate sense there is no
house,–[similarly for ‘fist’, ‘city’, ‘tree’ etc],–so too, when there are the five aggregates [as objects] of
clinging, there comes to be the mere term of common usage ‘a being’, ‘a person’, yet in the ultimate
sense, when each component is examined, there is no being as a basis for the assumption ‘I am’ or
‘I’; in the ultimate sense there is only minded body (Path 593–4 [xviii.28]).

A standard argument from grammar is swiftly refuted:

It was also asked: ‘Since there is no experiencer of it, whose is that fruit?’ Herein: ‘For mere arising
of the fruit/The common term “experiencer” is used,/Just as one says ‘It fruits’/when a fruit arises
on a tree.’ For just as it is simply owing to the arising of tree fruits which are one part of the states
called a tree, that it is said that ‘the tree fruits’ or ‘has fruited’, so it is simply owing to the arising of
the fruit consisting of the pleasure and pain called experience, which is one part of the aggregates
called ‘deities’ and ‘humans’, that it is said that ‘a deity or a human being experiences or feels plea‐
sure or pain.’ There is therefore no need at all for another [i.e. separate] experiencer (Dispeller 164;
Path 555 [xvii.171–2]).

The basic idea here is that a sentence containing a non‐agentive active verb can replace a sentence with an agent.
Surface grammatical form is no guide in answering deep metaphysical questions.
GANERI |
      3

2 |  TH E CO N CO M ITA N C E M O D E L O F M I N D

Buddhaghosa’s Buddhist philosophy of mind contains two different descriptions of a mental life without selves. Let
me say more about these two accounts, in each of which notions of attention play a central role. The first of the two
accounts is that there are certain ‘concomitants’ (cetasika) always accompanying every moment of worldly experi‐
ence (citta). Consciousness is never barely consciousness‐of; it is always consciousness‐with: ‘Just as when it is said
“The King has come” what it means is not that the King has come alone, leaving his courtiers behind, but that he has
come together with his courtiers’ (Fount 67). Consciousness, as we might put it, has commitments. The kind of as‐
sociation in question is specifically mental, and does not obtain between the mental and the physical or between the
physical and the physical (Fount 70). Rather, the idea is that concomitants are co‐emergent (sahajāta), co‐depend‐
ent (sahagata) and conjoined (saṃsaṭṭha) with experience (Kathāvatthu 337; Path 589 [xviii.8]), so that they ‘arise to‐
gether, run together, cease to exist together and thus exhibit a harmonious unity’ (Karunadasa, 2010, p. 72). A single
moment of consciousness—a citta‐cetasika complex—is a unified system. The components of the system that is a
single moment of experience are (Path 589 [xviii.8]):

• aboutness or directedness (citta, having an object/goal/opportunity, ārammaṇa),

• presence (phassa, ‘touch/contact’),
• felt evaluation (vedanā, ‘feeling’),
• identificatory type (saññā, ‘label’),
• effortful control or intending (cetanā),
• sustaining of cognitive boundaries (arūpa‐jīvindriya),
• attentional placing (ekaggatā; samādhi; appanā), and
• attentional focusing (manasikāra).

One might picture the account by thinking of each concomitant as a finger, and intentional aboutness as the palm,
their combined action being firmly to grasp something. A conscious awareness is an interdependent complex, the
ingredients of which cannot exist independently. Karunadasa (2010, p) rightly observes that,

both citta and cetasikas show how a multiplicity of mental states combine to produce a single unit
of cognition. What we call an instance of cognition is neither a single isolated phenomenon nor
a substantial unity. Rather, it is a complex of multiple mental states each representing a separate
function and all combining towards the cognition of the object. Their internal combination is not
based on the substance‐quality distinction: citta is not some kind of mental substance in which the
cetasikas inhere as its qualities. As mental dhammas or basic factors of psychological experience
they are co‐ordinate.

I have elsewhere argued that the dhamma‐ontology of early Buddhism is a trope‐theoretic ontology (for details,
see Ganeri, 2001, pp. 99–111; Ganeri, 2012, pp. 127–129). If that is correct then what we should say is that the con‐
comitants are mental tropes which are bound together by a similarity relation in a unit of conscious experience. Each
one is a mental qualitative particular, a unique mental aspect, and they combine trope‐theoretically into items of richly
articulate intentional experience.
 |
4       GANERI

3 |  TH E C AU SA L CO N D ITI O N S O F CO N S C I O U S N E S S

The term citta in this context indicates one aspect abstracted from the complex that is a complete moment of con‐
scious awareness, the aspect carrying intentional content, and it signifies something like ‘cast of mind’ or ‘mental
stance’ or ‘mind‐set’ (Harvey, 1995, pp. 115), or the ‘cognizance’ component of a complete moment of conscious
awareness. Buddhaghosa’s second account of consciousness describes not the commitments of consciousness
but the causal conditions under which consciousness arises. This second account is by and large an innovation
over earlier Buddhist sources. Motivating the second account is the idea that consciousness is what happens when
‘mind’ (mano) takes up a cognitive task. It does this by being interrupted from its default status. Unless called to act
by a world, mind in its default status ‘goes on occurring endlessly, in periods of dreamless sleep, etc., like the cur‐
rent of a river’ (Path 458 [xiv.114]). The interruption drags mind to consciously attend, ‘like some of the water that
flows a little after a boat going upstream’ (Path 460 [xiv.122]). There is a succession of tasked activities, separated
by revisits to the rest state (Gethin, 1994; pp. 14, 15; Karunadasa, 2010, pp. 139). Buddhaghosa’s second account
of consciousness is therefore meant to describe what it is for something in the world so to impress us that we
become aware of it; the mind is tasked to take up a view on the world.
What the second account states, more specifically, is that consciousness arises at the end of a series of cog‐
nitive activities (Dispeller 356; Path 458–60 [xiv.115–124], 21 [i.57]), each member a condition for the next (Path
488 [xv.36], 617–8 [xx.44]). Each activity is indexed to one of four discrete cognitive modular elements (dhātu),
so‐called because they assort (dhīyate) and because they carry (dhārenti) their own nature (sabhāva) (Dispeller 77),
this definition being what implies modularity. Roughly, an element of ‘mind’ (mano‐dhātu), in its two varieties, is
responsible for low‐level sensory processing, while the element of ‘mind‐cognition’ (mano‐viññāṇa‐dhātu), also in
two varieties, is charged with higher levels of perceptual processing and with the ‘running’ (javana) of what we
would now call working memory. The series is described first for the case of vision, and then progressively ex‐
tended to the other senses and on to other modalities of thought. In the case of vision, it begins when a stimulus
impinges the eye. If the stimulus is sufficiently strong to capture attention and so interrupt the default state, an
instruction (āvajjana) to orient a sense modality produces a first visual acknowledgement (cakkhu‐viññāṇa), a ‘mere
seeing’ (dassana‐matta). Having a first visual acknowledgement of the presence of an object, is not yet, however,
being conscious of it. Two further stages are required. There is, first, a complex cognitive activity involving the
‘receiving’ (sampaṭicchana), ‘investigating’ (santīraṇa) and ‘determining’ (votthapana) of what is seen. New with
Buddhaghosa, the import of these terms have to be deduced from the similes he gives to illustrate the entire pro‐
cess. These three cognitive activities are the take‐up by progressively more advanced aspects of perceptual pro‐
cessing: there is a selective ‘reception’ of stimuli into early vision and on into higher‐order visual processing, where
they are first ‘investigated’ to identify spatial properties and boundaries, and then categorically ‘determined’ as
an object of a certain sort. These processes stand in an approximate correspondence with phases of early, middle
and late vision identified in modern empirical work in the psychology of perception.
In a normal case of vision another cognitive activity will now take place: javana, ‘running’ (commonly translated
as ‘impulsion’). Taking as long as the whole of what has come before, the process is one of ‘running across’ the ob‐
ject time and again; it a sort of combing over and ‘ordering’ (anuloma) of the object file that has come from percep‐
tual processing. The final stage in this combing and ordering is said to involve taking ‘as an object that [which has
been so clarified]’ (tadārammaṇa). Gethin describes this as a residual moment of hanging on, in which ‘the mind,
no longer in its active mode, nevertheless momentarily holds on to the object it has just savoured’ (1994, p. 18).
So the stream of consciousness is wave‐like, each wave rising up to take an attended view on the world and
then sinking to a default status before rising again in an oscillation between untasked state and tasked activity
(see Thompson, , pp. 29–34, 40–45 for empirical studies). The second account describes the wave‐like rise and
fall of awareness, the way it is drawn into a moment of conscious attention which then fades out. Between these
moments there is a reversion to the default state (bhavaṅga), a discontinuous mode of consciousness that is out‐
of‐sync with the stream of conscious attentiveness.
GANERI |
      5

Buddhaghosa’s philosophy of mind thus contains two different descriptions of a mental life. According to the
first, an ever‐shifting directedness is associated with concomitants, some of which are necessary for conscious
experience, others individuate it by attitude and type. Conscious states are those which exhibit the following
feature: one is conscious when, in the context of an ongoing mental life, one is open to the presence of an object
in one’s surroundings, an object one selectively attends to and identifies as falling under a stored categorical type.
This description is of the manner by which conscious states ‘bend’ onto world, and it belongs within a philosophy
of thought, in which attention is described in terms of coordinate conscious activities of placing and focusing. The
second description belongs rather to the philosophy of cognition. What it states is that each of our perceptual
encounters with the world sets in motion a series of cognitive activities. Subliminal orienting, the activation of a
first cognitive module to a sensory stimulus gives rise to a primary visual acknowledgement. A second cognitive
module selectively receives the content of primary sensory acknowledgement into early perception (seeing an
orange array, say), a third works to investigate what has been received, and a fourth further to determine the
output of the investigating module. Investigation locates spatial boundaries (seeing something ovoid), while deter‐
mination categorizes an object (seeing as a mango). The perceptually processed content is then subject to further
cognitive processing, the ‘running’ of working memory, resulting finally in conscious experience. The content of
that experience can then itself be taken up by various consumer systems, such as deliberation (vicāra) and verbal
report (vohāra). This is a cognitive scientific description of how mind is ‘tasked’, and it consists in a description
of the operation of a variety of discrete cognitive modules. The cognitive role of attention is to orient the sense
modalities and to gate‐keep the flow of information.

4 |  CO R E S E LV E S ?

I said in Section 1 that typical understandings of the concept self are rejected. We should understand this to be a
rejection of any concept of personal or psychological identity based on agency or ownership: there is no ‘doer’, as an
agent of speech; and there is no ‘owner’, a possessor of feelings and experiences. The ‘self of the sectarians’ is a self
that owns experience and that performs actions including mental actions.
Rune Johansson, nevertheless, has proposed that we can see a role for a sort of ‘core self’. He says that.

citta, generally translated ‘mind’ [is] the core of personality, the centre of purposiveness, activity,
continuity and emotionality. It is not a ‘soul’ (atta), but it is the empirical, functional self. It is mainly
conscious but not restricted to the momentary conscious contents and processes. On the contrary,
it includes all the layers of consciousness, even the unconscious: by it continuity and identity are
safeguarded. It has a distinctly individual form. (1969, p. 30)

It ‘is by nature a centre of emotions, desires and moral defilements. It is partly conscious, partly unconscious. It has
also intellectual capacities and is capable of being transformed’ (ibid., p. 107). It is ‘not simply the mind and also not
simply personality but something of both: the organizing centre, the conscious core of personality, often described as
an empirical and functional self (but not atta), perhaps ultimately analysable into processes’ (ibid., p. 131). In his earlier
article, he refers to ‘a centre within personality, a conscious centre for activity, purposiveness, continuity and emotion‐
ality’ (1965, p. 179); insofar as it is not ‘an inner core … very much similar to all individuals’ but rather ‘an individually
formed centre’ (ibid., p. 174), it ‘comes very close to the psychological concept of personality’ (ibid., p. 178). Johansson
observes that ‘only once is it explicitly denied that citta is the self (S.ii.94), while it is very often denied that viññāṇa and
the other khandhā are the self’ (ibid., p. 168).
Is this redescription of citta as ‘core self’ consistent with the evident Buddhist denial of self? A necessary
requirement is that citta is neither the owner of feelings and experiences nor the agent of acts of speech, deed
or thought, and it does seem right that citta does not own the cetasikas: they are concomitants, not attributes. It
 |
6       GANERI

seems right too that citta is not an agent cause, for although it is a cause, it is itself causally conditioned, so it is not
the ‘first cause’ that the idea of agent causation implies. Peter Harvey agrees with much of Johansson’s descrip‐
tion, but corrects him for neglecting to emphasise that citta is ‘not only active but also acted upon’ (1995, pp. 114,
115). Yet Johansson does say that citta is capable of being transformed, and although it may not be the ‘basis’ of
the concomitants, it is nevertheless not itself a concomitant but a part of the citta‐cetasika system.
What function or aspect of citta might one point to as source of entitlement to describe it as self? Certainly
aware of the chariot metaphor and of general Buddhist rhetoric against self, Johansson must have had something
in mind in what he surely intended as a provocative claim. It is not Dan Zahavi’s ‘minimal self’, because the minimal
self experiences and feels, indeed experiences and feels with a sense of ‘mineness’ (Zahavi, 2005). Yet is anything
even more minimal than Zahavi’s minimal self rightly described as a self? I think that the most promising way
to understand his idea is by appeal to the idea that attention consists in the systematicity or structuring of the
stream of consciousness. Sebastian Watzl says that ‘consciously attending to something consists in the conscious
mental processes of structuring one’s stream of consciousness so that some parts of it are more central than oth‐
ers’ (2011, p. 158). The claim that citta is a core self might now be rephrased as being that citta consists of those
parts of one’s stream of consciousness (santāna) that have been made more central in the course of consciously
attending. So then citta is not a mere collection, nor is it the mere totality, but rather it is, at any given moment,
those specific elements which attention centralizes. Insofar as what is distinctive of attentional structuring is that
it enables a centre/periphery distinction to be drawn in relation to the stream of consciousness, we can use that
distinction to explicate the idea that citta is a core self. In this case, what is meant by ‘self’ is ‘those aspects of
the structure of the stream of consciousness which dynamically take centre stage when there is attention’. The
core self is the way attention shows up in the organization of the stream of consciousness. This makes sense of
the important idea in early Buddhism that citta can be modified and transformed in the training and cultivation
of skills of attention, a claim developed in detail in Buddhaghosa’s Path of Purification. The reason it is possible to
modify citta by training one’s attention is simply that citta consists in the way attention structures the stream of
consciousness. The ‘core self’, citta, is a surrogate self, something that performs many of the principal cognitive
tasks of the ‘self of the sectarians’ but has none of the metaphysical baggage, and cannot properly be called ‘self’
if it is fundamental to the functional role of that concept that it provides experience with agency. Insofar as there
is a role for the concept ‘self’ it is as a derived, not a fundamental notion.

5 |  T WO RO LE S FO R AT TE NTI O N

Attention is in fact given two distinct functions in this account. The Buddhist claim is that two functions of atten‐
tion are jointly necessary for conscious experience: a producing in mind (manasikāra), which is a kind of focusing;
and a singleness of mind (ekaggatā), a state of concentration (samādhi) that either approaches or consists in ab‐
sorption (appanā). As its very name implies, attention in the form of ‘bringing‐to‐mind’, is what is responsible for
experience having an intentional content:

‘Bringing‐to‐mind’ is a mode of work, working in the mind. It makes mind, so to speak, different
from the previous mind… It has the characteristic of driving associated states towards an object,
the function of joining associated states to an object, the manifestation of facing an object (āram‐
maṇa‐abhimukkha). It is included in the aggregate of constructing activities, and should be regarded
as the charioteer of associated states because it regulates an object (Path 466 [xiv.152]; Fount 133).

This is the focal aspect of attention which has a controlling role in directing or driving at an object. Concentrated,
absorbed, singleness of mind, on the other hand, is defined as follows:
GANERI |
      7

It is the centring (ādhāna) of consciousness and concomitants equally (sama) and correctly (sammā)
on a single object; placing (ṭhapananti) is what is meant. So it is the state in virtue of which conscious‐
ness and its concomitants remain equally and correctly on a single intentional object (ekārammaṇe),
undistracted and unscattered—that should be understood as concentrating. It has non‐distraction
as its characteristic. Its function is to eliminate distraction. It is manifested as non‐wavering. (Path
84–5 [iii.3, 4])

In this capacity, attention removes unclarity but does not itself present the object, just as the wind may dispel the
clouds but does not itself see the moon (Fount 281; Path 673–4 [xxii.7–10]). That is to say, it does not itself produce for
thought an intentional object but instead has an ancillary constitutive role. It is the placing of attention on a part of the
phenomenal array, and fixing that region in phenomenal space by not wavering between attributes. Its functional role
is that of stabilising by eliminating distractors; one might think of the manner in which a thing may be held in place by
a series of springs attached in different directions, each spring slack unless there is an attempt to move in the opposite
direction, when it taughtens and pulls the object back.
There is here a fundamental distinction between two roles for attention: attention as driving mind onto an
object and attention as stabilizing an object in mind. Govinda astutely ventures that singleness of mind is a sort of
limiting principle and bringing‐to‐mind a directing principle. He says that singleness of mind is what ‘distinguishes
one object from another and prevents its dissolution and merging into other objects’. Suggesting that such atten‐
tion is ‘spontaneous’ because it ‘is not forced by will but rather aroused by the immanent qualities of the object
itself which ‘attracted’ the attention’, he goes on to argue that ‘singleness of mind and bringing‐to‐mind can be
defined as the negative and the positive side of the same function: the former one eliminates (or turns away from)
everything that does not belong to the object, the latter directs itself towards the object which has been isolated
in this way’ (Govinda, 1961, p. 116, my italics).
The philosophical significance of these claims of Buddhaghosa is best brought out through a comparison
with Brian O’Shaughnessy’s very important discussion in Consciousness and the World (O’Shaughnessy, 2002).
O’Shaughnessy begins by arguing that attention is necessary for consciousness, or, more precisely, that con‐
sciousness ‘necessitates the accessibility of the perceptual attention’ (2002, p. 10). Attention is necessary for
consciousness because consciousness has a necessary truth‐orientation; it is a ‘reality‐detector’, and indeed
that is what it means to say that consciousness puts us in contact with reality, that ‘whereas dreams merely
putatively are of Reality, consciousness is ‘in touch with’ Reality’ (ibid., p. 12). There are already echoes here
of two claims Buddhaghosa has made, that presence (phassa) and attention (manasikāra) are concomitants
(O’Shaughnessy speaks rather of commitments) of consciousness. O’Shaughnessy likewise identifies dual
functions for attention, ‘that of providing psychic space for certain select phenomena, and of bringing certain
phenomenal existents to consciousness’ (2002, p. 293; my italics). He adds that these are two functions of a
unitary phenomenon, for ‘perception is an event in which the attention acquires an object and is at the same
time an event that calls upon or uses attention’ (ibid.). As for the nature of attention, O’Shaughnessy says that
a natural and appropriate imagery can easily turn into a myth. The natural imagery is of attention as mental
‘life‐blood’, as ‘a sort of mental “space” of awareness present in the mind, which is occupied exclusively by the
experiences it enables to exist’ (ibid., p. 285), for ‘if (say) emotion or thought or perception are to so much as
exist, attention needs to be available’ (ibid., p. 277). The point of the imagery is that it captures that sense in
which attention is limited (‘occupied’), a sense that O’Shaughnessy illustrates with the example of driving a
car through a narrow pass, one’s attention to the driving precluding one from attending to a difficult conversation
at the same time. O’Shaughnessy cautions that this imagery can easily be misunderstood, for it may lead to the
impression that the experiences which attention enables to exist are one thing and the ‘mental space’ of awareness
is something else:
 |
8       GANERI

The myth in question takes the following form. It is of a mental existent (which I shall call S), a par‐
ticular mental ‘space’ that is of type awareness (in some sense), which coexists with and is distinct
from contemporaneous experiences. Those experiences relate to that awareness-space, not as its
objects, but as its occupants, and that property enables them to exist. (ibid., p. 285)

And again,.

That to which these various expressions refer [‘the attention’, ‘awareness’] is something that is
closely akin to a psychic space. And yet as we have just seen in the recent discussion of the myth‐
ical S, it cannot be something that, like the space of a canvas or stage, precedes and outlives its
occupants. (ibid., p. 288)

Instead of falling into the myth, O’Shaughnessy says that we should realise that ‘what we have in mind in speaking
of “The Attention” … is nothing less than Experiential Consciousness itself…To repeat, it is what we frequently refer
to as ‘the stream of consciousness’ (of literary fame)’ (ibid., p. 288). ‘Experiential Consciousness’ is O’Shaughnessy’s
term for the stream of consciousness (ibid., p. 15). His idea is that to find our attention occupied by a certain given
experience is for the experience to partly constitute the attention, much as a single piece does a jigsaw puzzle; what
is occupied by experiences is, as he puts it, a system of those experiences, the system being the network of interrela‐
tionships which experiences need in order to exist (ibid., p. 288). He concludes:

Denuded of the above array of [systemic] properties, they [sc. experiences] would be like so many
psychological atoms wandering in a void. Endowed with them, they constitute a continuous ongo‐
ing phenomenon which is a sort of circle or centre of awareness. This awareness is the Attention.
(ibid., p. 289, 290)

I need hardly add that ‘the mythical S’—this unspeakable nothing whose possible existence is acknowledged not
even in the index of Consciousness and the World—is the self. For O’Shaughnessy, as for Buddhaghosa, attention re‐
places self in the explanation of perception, thought, and emotion. What O’Shaughnessy does so brilliantly is to
demonstrate how the natural imagery of attention is what itself gives rise to the Myth of Self as Detached from
Experience; the ‘self of the sectarians’ is a bad attempt to formulate a good insight about attention. Of course nothing
can prevent us, should we so wish, from stipulatively defining the word ‘self’ to refer to the effects of attention, and
this I think is just what Johansson’s claim that citta is ‘core self’ ultimately comes to.
So, to conclude, I have noted how Buddhists like Buddhaghosa distinguish between a role for attention at a
functional level of cognitive processing and a different role at the psychological level of awareness. At this second
level, a further distinction is made between the attentional roles of focussing and placing, and this means that
attention does double duty in the dynamic centring of the stream of conscious psychological events. We might
call the transient and effervescent centre a ‘self’ if we want to, but that does not imply any departure from the fun‐
damental Buddhist refusal to commit to self as the agent of thinking and acting. This recovered ‘self’ is dependent
on the structure of attention, not the detached cause of that structure.

AC K N OW L E D G E M E N T S

Many thanks to Shalini Sinha for her wonderful organisation of the Ratio Annual conference in Reading, 2016
and for editing this special issue, to David Oderberg, the editor of Ratio, and to the participants and audience of
the conference for their many very helpful insights. Thanks also to Oxford University Press for permission to use
material from Attention, Not Self (Oxford University Press, 2018).
GANERI |
      9

ORCID

Jonardon Ganeri  http://orcid.org/0000-0001-7778-2344

REFERENCES

Buddhaghosa. (1991). Path (Visuddhimagga): Ñāṇamoli, Bhikkhu (trans.). The path of purification: Visuddhimagga by
Bhadantācariya Buddhaghosa (5th ed.). Kandy: Buddhist Publication Society.
Buddhaghosa. (1996). Dispeller (Sammoha‐vinodanī): Ñāṇamoli, Bhikkhu (trans.). The dispeller of delusion (Sammohavinodanī),
Revised for publication by L. S. Cousins, Nyanaponika Mahāthera, and C. M. M. Shaw (2 vols). Oxford: Pāli Text
Society.
Buddhaghosa. Fount (Attha‐sālinī): (1) Ñāṇamoli, Bhikkhu (ms.), transl. The Fount of Meaning (Atthasālinī) by Buddhaghosa
Ācariya (Island Hermitage Library, Dodandūwa). Partially typed, partially handwritten, ms. on Asl. 36–114. (2)
Nyānaponika, Bhikkhu (2005 [1942]), transl. Darlegung der Bedeutung (Atthasālinī), Edited for publication by Sven
Bretfeld and Rainer Knopf (Oxford: Pāli Text Society). (3) Tin, Pe Maung (1920), transl. The Expositor (Atthasālinī):
Buddhaghosa’s Commentary on the Dhammasangaṇī, the First Book of the Abhidhamma Piṭaka. Oxford: Pāli Text
Society, 2 vols.
Ganeri, J. (2001). Philosophy in classical India: The proper work of reason. London: Routledge.
Ganeri, J. (2012). The self: Naturalism, consciousness, and the first‐person stance. Oxford: Oxford University Press.
Gethin, R. (1994). Bhavaṅga and Rebirth according to the Abhidhamma. In V. Pagel & T. Skorupski (eds.), The Buddhist
forum III: 11–35. Reprinted in Buddhism: Critical Concepts in Religious Studies. Edited by P. Williams. Volume IV:
Abhidharma and Madhyamaka (2005, pp. 159–181). London: Routledge.
Govinda, L. A. (1961). The psychological attitude of early Buddhist philosophy. London: Rider & Co.
Harvey, P. (1995). The selfless mind: Personality, consciousness and Nirvāṇa in early Buddhism. London: RoutledgeCurzon.
Johansson, R. E. A. (1965). Citta, mano, viññāṇa—A psychosemantic investigation. University of Ceylon Review, 23 (1&2),
165–215.
Johansson, R. E. A. (1969). The psychology of Nirvāṇa. London: George, Allen and Unwin.
Karunadasa, Y. (2010). The Theravāda Abhidhamma: Its inquiry into the nature of conditioned reality. Hong Kong: Centre of
Buddhist Studies, The University of Hong Kong.
Ñāṇamoli, B. , trans. (1996). The dispeller of delusion (Sammohavinodanī) (2 vols). Revised for Publication by L. S. Cousins,
Nyanaponika Mahāthera and C. M. M. Shaw. Oxford: Pali Text Society [Vibh‐a].
O’Shaughnessy, B. (2002). Consciousness and the World. Oxford: Oxford University Press.
Thompson, E. (2015). Waking, dreaming, being: New light on the self and consciousness from neuroscience, meditation and
philosophy. New York, NY: Columbia University Press.
Watzl, S. (2011). Attention as structuring of the stream of consciousness. In C. Mole, D. Smithies, & W. Wu (eds.).
Attention: Philosophical and psychological essays (pp. 145–173). New York, NY: Oxford University Press.
Zahavi, D. (2005). Subjectivity and selfhood: Investigating the first‐person perspective. Cambridge, Massachusetts: MIT
Press.

How to cite this article: Ganeri J. Attention and self in Buddhist philosophy of mind. . 2018;00:1–9.
https://doi.org/10.1111/rati.12202