1  

Luminous Mind: Self-Luminosity vs. Other-Luminosity in Indian Philosophy of

Mind

Matthew MacKenzie

Colorado State University

DRAFT 2014

Luminosity

The metaphor of consciousness as light (prakāśa) or luminosity (prakāśatā) is at

the heart of Indian thinking about the nature of the mind going back at least as far as the

early prose Upaniṣads. In a well known dialogue in the Bṛhadāraṇyaka Upaniṣad

between King Janaka and the sage Yājñavalkya, the King asks, "Yājñavalkya, what is the

source of light for a person here?" In response, Yājñavalkya mentions the external

sources of illumination, such as the sun, moon, and fire, as well as the illumination

provided by a voice in the darkness. The King then asks,

“But when both the sun and the moon have set, the fire has died out, and

the voice is stilled, Yājñavalkya, what then is the source of light for a

person here?"

"The self (ātman) is then his source of light [svayaṃjyoti]. It is by

the light of the self that a person sits down, goes about, does his work, and

returns."

"Which self is that?"

  2  

"It is this person—the one that consists of perception among the

vital functions (prāṇa), the one that is the inner light within the heart”

(BAU 4.3.1-8; Olivelle 1998, p. 111).

Like a light, consciousness has (or is) the capacity to shine forth (prakāśate) and

illuminate (prakāśyati) its object. Indeed, just as, without illumination, no objects could

be visible, without the light of consciousness, no object could be experienced. Thus

luminosity comes to denote the capacity to disclose, present, or make manifest. Physical

light, of course, can reveal or make visible objects, but, as later Indian philosophers

pointed out, it can only do so to a perceiver. The luminosity of consciousness, on the

other hand, is that original capacity to make experientially present some object to some

subject. Yet this inner light that makes possible all experience and knowledge is itself

quite elusive. Earlier in the text, Yājñavalkya remarks,

You can't see the seer who does the seeing; you can't hear the hearer who

does the hearing; you can't think of the thinker who does the thinking; and

you can't perceive the perceiver who does the perceiving. The self within

all is this self of yours. All else besides this is grief! (BAU 3.5.2; Olivelle

1998, p. 83).

On this view—widely, but by no means universally held in Indian thought—the

conscious subject (or consciousness itself) is not knowable in the same way as its objects.

Despite its association with the elusiveness of the subject, however, luminosity also

  3  

comes to be associated with the distinctive flavors (rasa) or qualitative features of

conscious experience—that is, with the phenomenality consciousness (Ram-Prasad 2007,

p. 54).

By the classical period, luminosity comes to denote the distinctive mark or feature

of consciousness as that which reveals or discloses (to a subject), particularly in the

context of distinct episodes of conscious cognition. As Chakravarthi Ram-Prasad

characterizes it:

Luminosity is the rendering of an event as subjective. It is that by which

there is an occurrence, which it is like something to undergo. The

subjective is the having of the experience (anubhava). Luminosity is the

Indian metaphor for phenomenality, the undergoing by the subject of

something else (it object). The philosophers are agreed on all sides that

consciousness is phenomenological; it is luminous. The debate is over the

constitution of the phenomenality of consciousness. The debate is about

what it is for there to be subjectivity (Ram-Prasad 2007, p. 54).

While we should not simply identify phenomenality and subjectivity, it is important to

note here that the question of luminosity constitutively involves questions of

intentionality, phenomenality, and subjectivity, as well as their interrelations.

The basic divide in Indian accounts of the luminosity of consciousness is between

other-illumination (paraprakāśa) and self-illumination (svaprakāśa) theories. For

advocates of other-illumination, the luminosity of consciousness consists in its capacity

  4  

to present a distinct object. Thus, transitive, object-directed intentionality is the mark of

consciousness. Conscious states, in order to be states the subject is conscious of, must be

presented by a distinct, higher-order cognition. Hence, consciousness illuminates that

which is other than itself, and conscious states themselves are apperceived by another

state. In contrast, for advocates of self-illumination theories, the luminosity of

consciousness consists in its being reflexive or self-presenting. Consciousness presents

itself in the process of presenting its object. Moreover, just as light does not need a

second light in order to be revealed, so consciousness does not need a distinct state to

present itself—it is self-intimating.

In what follows I will discuss some of these different accounts of luminosity

developed and defended in the Indian context, as well as some relevant connections to

contemporary issues in the philosophy of consciousness.

Other-Luminosity

As mentioned above, the basic divide with which we are concerned here is

between other-luminosity (paraprakāśa) and self-luminosity (svaprakāśa) accounts of

consciousness. However, these are broad classifications that cover a variety of distinct

and often incompatible views, each of which is developed within the context of a

particular, systematic philosophical viewpoint (darśana). So, for instance, while the

Nyāya, Bhaṭṭa Mīmāṃsā, and Madhyamaka schools have defended other-luminosity, the

specific views defended are often quite different and form only a part of dramatically

different integrated philosophies of consciousness. For my purposes in this chapter, I will

  5  

concentrate on the Nyāya tradition’s account of other-luminosity as a paradigm case,

though I will also discuss some relevant features of the Bhaṭṭa Mīmāṃsā view.

The Nyāya school defends a resolutely realist, intentionalist, and first-order view

of consciousness. It is realist in three relevant respects. First, Nyāya is fully ontologically

realist about consciousness. Consciousness is a real, unique, and irreducible capacity of

the self.1 Second, Nyāya is epistemologically realist in that veridical episodes of

conscious cognition disclose real, mind-independent objects and properties in the word.

Third, the school defends a direct realist account of perception in that veridical episodes

of perception reveal mind-independent physical objects, without the intermediary of

conscious mental images or representations (ākāra). Nyāya realism is, in turn, tightly

bound up with their intentionalist analysis of consciousness. It is the very nature of

consciousness to be of or about an object (svābhāvika viṣayapravaṇatvam). Episodes of

conscious cognition are individuated by their object, and the intentional object of

cognition is (in the standard case) an object in the world, not any kind of mental

intermediary. Thus, intentionality is the mark of the mental and the intrinsic intentionality

of consciousness is understood in externalist terms. Moreover, since consciousness is

transparent (nirākāra, ‘without image/aspect’), the phenomenal features of a conscious

state are grounded in (and perhaps reducible to) the way the object is presented through

the state. Finally, Nyāya defended a first-order view of consciousness. The luminosity of

a cognition consists in its revealing its object (arthaprakāśobuddhi), not itself. That is, it

illuminates that which is other than itself (its object) and requires a distinct, second order

cognition for a first-order cognition to be revealed.

  6  

On the Nyāya view, then, when one has a perception of a tree, one is in a

conscious state with the content ‘there is a tree’. The first-order cognition derives its

content from the worldly object it discloses and it makes no reference to either the

cognition itself or its subject. The term for a first-order cognition in this context is

vyavasāya. It presents the tree in a particular way (in terms of color, shape, and so on),

but this is ultimately a function of the properties of the tree. But in what sense is this

cognition conscious? It is conscious in that it makes its subject aware of the object. When

one has a perception of a tree, that perception makes one conscious of the tree. On this

view, one need not be aware of the cognition in any way in order for it to make one

conscious of its object. More formally, we can call this the independence condition: the

presence of a conscious mental state does not depend on the subject’s being aware of that

state (or aware of being in it). As Fred Dretske remarks, “There are, to be sure, states in

(or of) us without which we would not be conscious of trees and pianos. We call these

states experiences. Since these experiences make us conscious of things . . . the states

themselves can be described as conscious. But we must be careful not to conclude from

this that because the states are conscious, we must, perforce, be conscious of them”

(Dretske 1995, p. 100). The Nyāya school is in complete agreement with this point.

How, then, do we come to be aware of our conscious states, when we are?

According to Nyāya, our first-order cognitions (vyavasāya) are cognized by a subsequent,

higher order cognition (anuvyavasāya) or apperception. In the typical case of perception,

a first-order cognition triggers a second-order cognition that takes the first-order

cognition as its intentional object. There are several features of this view worth noting.

First, this is thought to be an automatic causal process. It is causal in that, in minds like

  7  

ours, the occurrence of a first-order conscious perception triggers (causes) an

apperceptive meta-cognition. This implies, on the Nyāya view of causality, that first- and

second-order cognitions (qua vehicles of content) are distinct, because causation occurs

between distinct existences. Second, anuvyavasāya as automatic meta-cognition should

be distinguished from voluntary reflective introspection. While we do typically

apperceive our basic perceptions, we do not usually go around introspecting our

cognitions. Yet, in so far as the relation between vyavasāya and anuvyavasāya is causal,

the Nyāya school holds that this process can be blocked or simply fail to occur. For

instance, when one is fully absorbed in an object, one might have no meta-cognition of

the perceptual cognition, or when one hears a loud noise, the sudden shift in attention

might disrupt the formation of a meta-cognition directed at a prior visual perception.2

Third, while the vyavasāya and the anuvyavasāya are distinct cognitions, their content is

intimately related. When a perception with the content ‘there is a tree’ is apperceived, the

content of the apperception will be ‘I perceive a tree’. The direct object of the meta-

cognition is the first-order cognition, while the indirect objects of the meta-cognition are

the first-order cognition’s object (the tree) and the self. Hence, while the meta-cognition

is about another cognition, it retains reference to the first-order cognition’s object of

perception. In addition, the meta-cognition involves self-reference in that it indirectly

refers to the self as the subject of the state. It does not, however, refer to itself, as all

cognitions reveal only that which is other than the cognition itself.

As I pointed out above, Nyāya is centrally concerned with articulating and

defending a robustly realist account of consciousness and cognition. With this in mind,

we can see why the various elements of their other-luminosity theory are tightly

  8  

connected. As intrinsically intentional, consciousness is essentially world-directed.

Indeed, on their view, a conscious cognition is, in a sense, exhausted by its revealing its

object. It exists only as this directedness at the world. On the other hand, it is also no less

than the unique capacity to make the self aware of the world. So while Nyāya (like most

Indian schools) gives a causal account of episodes of cognition, it is an account of the

causes and conditions of the occurrence of an intrinsically intentional (luminous) state of

the self, irreducible to any physical or otherwise non-intentional quality. Further, if the

nature of intentionality is to reveal transcendent objects—that is, objects that are

independent of the cognitive state itself—then it would seem to follow that consciousness

is transparent or formless. Anything revealed by a conscious state, on this account of

intentionality, is ipso facto other than that state. As simply the illuminating of an object,

consciousness itself has no form of its own. For Nyāya, then, because consciousness is

essentially saviṣayatā (‘with-an-object’, i.e. intentional), it must also be nirākāra

(‘formless’ or ‘without aspect’).

This line of thinking is strikingly similar to one developed by Jean-Paul Sartre. In

the introduction to Being and Nothingness, he writes, “All consciousness, Husserl has

shown, is consciousness of something. This means that there is no consciousness that is

not a positing of a transcendent object, or if you prefer, that consciousness has no

‘content’” (Sartre 1956, p. 11). For Sartre here the paradigm case of ‘content’ would be

something like an immanent mental image that serves as an intermediary between mind

and world, such as the ‘ideas’ of the British empiricists. For both Nyāya and Sartre, the

theory of intentionality as pure world-directedness leads to a rejection of these

intermediaries (the ākāra in the Indian context) and therefore to the view that

  9  

consciousness is radically transparent or formless. That a cognition is formless in this

sense does not preclude its having a contentful intentional structure (viṣayatā). Rather,

the point is that this form of intentionality is direct and presentational as opposed to the

indirect and re-presentational account given by proponents of sākāravada (‘with-aspect-

ism’). For Nyāya, then, a proper understanding of the luminosity of consciousness entails

the rejection of its supposed self-luminosity, because, if a cognition only reveals a

transcendent object, it cannot reveal itself and it must be revealed by another cognition.3

We may call this the intentionality argument against self-luminosity.

In summary, we can discern three interwoven strands in the Nyāya case for other-

luminosity.4 There are causal considerations in support of other-luminosity. The

Naiyayikas are asatkāryavādins concerning causality, meaning that they take cause and

effect to be ontologically distinct. A sensory-perceptual cognition arises from an

appropriate causal link between an object in the cognizer’s vicinity, the sense organs, and

the self. A cognition, being caused by its object, must be distinct from its object. In the

case of apperception, a meta-cognition arises when there is an appropriate causal link

between the prior cognition, the inner sense (manas), and the self. So again, there must be

two distinct cognitions. Call this the causal argument for other-luminosity. There are also

broadly phenomenological considerations in support of the view. That is, Nyāya points

out ways in which perception and apperception can diverge in our own experience, such

as sudden shifts of attention. Call this the phenomenological argument. Finally, as we

have seen in the intentionality argument, the Nyāya analysis of the nature of cognition

takes it to be pure directedness at a distinct object. These three strands make for a

powerful and integrated approach to the nature of consciousness.

  10  

The Bhaṭṭa Mīmāṃsā school also defended an other-luminosity account of

consciousness, but one that was in some ways quite different from Nyāya. Bhaṭṭa

Mīmāṃsā, like Nyāya, holds that cognitions are other-luminous in that their function is to

reveal a distinct object. However, Nyāya holds a first-order view of consciousness in that

an object-directed first-order cognition is sufficient for the subject to be in a fully

conscious cognitive state. That is, a first-order cognition yields a personal-level conscious

state5, such as seeing a tree. In contrast, for Bhaṭṭa Mīmāṃsā there is a two-stage process

leading to full-blown person-level conscious states. In the first stage, one has a perceptual

state that presents an independent object. In the second stage, one has a meta-cognition

that makes available on the person-level the object of the first cognition. Hence, we might

think about this meta-cognition as yielding a form of access-consciousness of the

contents of the first-order state. In this way the Bhaṭṭa Mīmāṃsā view is similar to

higher-order representation (HOR) views of consciousness. However, it is important to

note that, unlike standard HOR views, Bhaṭṭa Mīmāṃsā holds that the first-order

cognition is luminous and therefore both intentional and phenomenal. Hence, the

luminosity of the first-order state is not explained by the occurrence of the higher-order

state. Rather, the higher-order state explains the subject’s conscious access to the contents

of the first-order state—the incorporation of that content into the person-level stream of

consciousness. Moreover, as distinct states, one could have a first-order perception of a

tree in the absence of a higher-order cognition, yielding a case of phenomenal

consciousness without access-consciousness. Here again we see a difference between

Nyāya and Bhaṭṭa Mīmāṃsā views of other-luminosity. For Nyāya, a first-order

  11  

perceptual state is a full-blown conscious state in the subject’s mental life and is available

for belief-formation, attention, and action-guidance. For Bhaṭṭa Mīmāṃsā, however, the

first-order state in the absence of meta-cognition will not be fully integrated with other

functions and would therefore be more like a subliminal or very marginal perception.

This puts their view somewhere between pure first-order and higher-order views of

consciousness.

Perhaps the most unique feature of the Bhaṭṭa Mīmāṃsā view is its account of the

nature of the meta-cognition. On their account, the meta-cognition does not take the first-

order cognition as its object. In fact, first-order cognitions never become direct objects of

cognition in the causal-perceptual sense. Rather, the intentional object of the meta-

cognition is the object of the first-order cognition. When one has a first-order perceptual

cognition of a tree, the meta-cognition’s object is also the tree, not the perceptual

cognition. But if the meta-cognition also takes the tree as its object, what is its function

over and above mere perception of the tree? For the Bhaṭṭas, the meta-cognition serves to

ascribe the awareness of the perceptual object to the subject, thereby incorporating it fully

into the subject’s explicit mental life. Note, then, that we have here a distinction between

the conscious intentionality (luminosity) of a state and its subjectivity, its for-me-ness. It

seems that the role of the meta-cognition here is to bring the intentional state into the

domain of full-blown subjective experience. Again, this is similar to higher-order views

of consciousness. But standard higher-order views still take the meta-cognition to be

about another cognition, whereas the Bhaṭṭas hold that it too is about a worldly object.

The basic idea here is that the meta-cognition grasps its object as cognized

(jñātatā) and having the quality of being cognized presupposes a cognition of the object.

  12  

Thus the subject becomes aware of the object directly (non-inferentially) and the first-

order cognition indirectly by way of analytic presupposition (arthāpatti) or by inference

(anumiti) (Ram-Prasad 2007, pp. 62-63). Thus Bhaṭṭa rejects the inner perception model

of self-awareness in favor of an indirect model. The meta-cognition reveals that one has

cognized the object, but is not directly of the first-order cognition. Now, in the standard

model of inference in Indian philosophy, inferential warrant is based on universal

concomitance (vyāpti), for example, ‘where there is smoke, there is fire’. In this case,

however, a problem arises because the cognition known on the basis of the cognizedness

of the object is never itself perceived. What then could establish the existence of the

cognition? According to Kumārila, the move from the cognized object to the cognition is

based on analytic presupposition. The property of ‘cognizedness’ is definitionally or

conceptually related to the occurrence of cognition. Thus, the cognizedness of the object

already presupposes the cognition. Other Bhaṭṭa thinkers (and their critics) take the move

to be a form of abductive inference, wherein one infers from the observable to the

unobservable. In either case, the upshot is that, when one sees a tree, on does not see

one’s seeing, but one comes to know that she is seeing by way of the ‘seenness’ of the

tree itself. Therefore, self-knowledge of this kind is parasitic on knowledge of the world.

Yet what is this property of ‘cognizedness’? It has been compared to the

‘cookedness’ of rice, though I am not sure how helpful is that analogy. More importantly,

it is explained in terms of the object’s availability for speech and action. When I have a

perceptual cognition of a pot, the pot is experientially available for reference (‘there’s a

pot’) and actions such as grabbing the pot. Recently, in his contemporary appropriation of

the Bhaṭṭa model, Mark Siderits (2011) has suggested that cognizedness be understood in

  13  

terms of global availability. That is, when an object is available for sensation, speech,

action, memory, belief-formation, etc., it has the higher-order property of cognizedness.

On this view, a first-order cognition has both phenomenal character and intentionality,

but it is not yet fully subjective or fully conscious. It does, in some minimal sense, make

its subject aware of its object. It also triggers a meta-cognition that directly takes the

object as qualified by cognizedness (global availability). This meta-cognition also

indirectly takes it that the subject is cognizing the object. The result is a cognition of the

form ‘I am cognizing the object’. In contemporary terms Bhaṭṭa Mīmāṃsā view appears

to be a unique form of a non-reductive higher-order thought account of consciousness.

Self-Luminosity

In the Pramāṇavārttika, the Buddhist philosopher Dharmakīrti remarks that, “the

mind is by nature luminous cognizance (prabhāsvara)” (PV 2.210cd-211ab; Dunne 2004,

p. 372). Just as light reveals those objects upon which it falls, consciousness has the

unique capacity to make experientially present those objects to which it is directed. It is

that by which there can be any phenomenal appearance. As the capacity for experiential

manifestation, luminosity, as we have seen, is linked to the phenomenality of

consciousness. Consciousness has no color, shape, size, weight, and so on, yet it is the

condition of the possibility of experiencing these qualities. Indeed, when one searches for

consciousness itself as an object or content, one is unable to find it.6 The cognizance of

consciousness is its capacity to grasp or apprehend its object. This aspect, then, is linked

to the intentionality of consciousness. Conscious states not only present phenomena, but

  14  

also are intentionally directed to objects (viṣaya). Further, consciousness involves not just

the capacity to reveal or phenomenally manifest, but also to identify, re-identify, and

understand its objects.

According to the Buddhist proponents of self-luminosity, the luminosity of

consciousness consists in its being reflexive or self-presenting. Consciousness presents

itself in the process of presenting its object. In some Buddhist schools, the term

‘svasaṃvedana’ (self-awareness) denotes this self-luminosity or pre-reflective self-

awareness that is an invariant aspect of conscious experience. On this view, individual

conscious states simultaneously disclose both the object of consciousness and (aspects of)

the conscious state itself. Thus, when a subject is aware of an object, she is also (pre-

reflectively) aware of her own experiencing. Buddhist philosophers such as Dignāga,

Dharmakīrti, and Śāntarakṣita defended the idea that consciousness is reflexive or self-

presenting in this way.

According to Dignāga, “Every cognition is produced with a twofold appearance,

namely that of itself (svābhāsa) and that of the object (viṣayābhāsa)” (PS(V) 1.9a;

Hattori 1968, p. 28). The object-appearance or object-aspect is the presentation of the

intentional object in cognition. It is what the experience is as of. Whatever the further

status of the intentional object, in so far as it is given in experience, there is an object-

appearance. There are then further questions such as the relation between the object-

aspect and the object that is the cause of the cognition. Dignāga argues for a causal theory

of experience whereby the intentional object of an experience (it’s object-aspect) is

logically independent of its cause and wherein the experience can be as of an object that

does not in fact exist. Further, the identity of the cognition is partly constituted by its

  15  

object-aspect and so it can be individuated in terms of its intrinsic intentionality. That is,

a cognitive episode’s being as of a tree is intrinsic to it, whether or not there is a

cognition-independent tree. Yet a cognition is not exhausted by its presentation of an

intentional object. It also presents a subject-aspect (svābhāsa), which for Dignāga means

the way the cognition presents itself. When I have an experience as of a tree, on this view,

the experience presents both the tree (the object-aspect) and the experiencing of the tree

(the subject-aspect). And since I grasp both the object-appearance and the self-

appearance of the cognition in which the object is presented, the dual-aspect structure of

cognition implies pre-reflective self-awareness (svasaṃvedana). Importantly, the

viṣayābhāsa, svābhāsa, and svasaṃvedana are features of a single episode.7 Hence, the

pre-reflective self-awareness here is not a distinct higher-order cognition, but rather an

intrinsic feature of the first-order cognition itself. This means that Dignāga’s view can be

classified as a same-order theory of self-awareness.

Dignāga’s dual-aspect view of cognition, then, is similar to that of Colin McGinn.

On McGinn’s view, experiences are ‘Janus-faced’. He writes:

[S]ubjective aspects of experience involve reference to the subject

undergoing the experience—this is what their subjectivity consists in. But

we can also say that . . . experiences have a world-directed aspect: they

present the world in a certain way, say as containing a scarlet sphere

against a blue background. This is their representational content, what

states of affairs they are as of. Thus . . . experiences are Janus-faced: they

point outward to the external world but they also present a subjective face

  16  

to their subject; they are of something other than the subject and they are

like something for the subject (McGinn 1991, p. 29).

The world-directed aspect or objective face here corresponds closely to Dignāga’s

viṣayābhāsa, while the subjective face corresponds to the svābhāsa. Note further that this

involves a rejection of the transparency of experience. For Dignāga, as for McGinn, there

is more to experience than how the object is presented. The phenomenal character of

experience also involves how the experience itself is presented. Thus we should make a

phenomenological distinction between what the object is like and what it is like to

cognize the object. For instance, what it is like to see a bright yellow lemon may be

different from what it is like to imagine the lemon or to remember it, even if the lemon

itself is presented in the same way (as bright yellow, etc.) in each cognition. Jonardon

Ganeri characterizes the subjective-aspect simply as, “whatever it is in virtue of which

attending to one’s experience does not collapse into attending to the world as presented in

experience” (Ganeri 2012, p. 170).

Indeed, Dignāga’s main argument for the dual-aspect view (dvyābhāsatā) is that if

cognitions are transparent, there would be no distinction between a cognition and the

cognition of that cognition (PS 1.11; Kellner 2010, p. 210). That is, if a cognition of a

tree (C1) has no form other than its object and a meta-cognition of C1 (C2) is also

transparent, then, while C2 has as its object C1, C1 has no form other than the form of the

tree. Therefore, C2 will collapse into just another cognition of the tree. On the other hand,

if C1 has two faces, then we can make sense of what C2 grasps when it cognizes C1,

  17  

namely the subject-aspect of C1. C2 will also grasp C1’s object in grasping C1, but it will

not collapse into it. Call this the meta-cognition argument.

Dignāga does not say very much about the nature of the subject-aspect, but he

does say that we are aware of the various ‘mental factors’ (caitta) that are built into our

cognitive episodes (PS(V) 1.6ab; Kellner 2010, p. 207). In much Buddhist philosophy of

mind, cognitive episodes are analyzed according to a citta-caitta model. On this model,

the primary cognition is the citta (consciousness or conscious event) and this basic event

can be characterized in terms of a variety of mental factors (the caitta). One common

taxonomy lists 51 distinct mental factors, divided into six groups. Within this taxonomy,

however, there are five omnipresent factors: feeling (vedanā), recognition (saṃjñā),

intention (cetanā), attention (manasikāra), and contact (sparśa). Each episode of

conscious cognition involves some permutation of these five factors. In addition to being

luminous and cognizant, an episode of cognition may be pleasant or unpleasant, focused

or unfocused, calm or agitated, etc. (Dreyfus 2011, p. 119). These mental factors

contribute to the global phenomenal character of the episode and are available for

reflection by the subject—that is, they are pre-reflective aspects of the experience. These

features of the experiencing, rather than the object experienced, may be part of what

Dignāga terms the svābhāsa, the ‘self-appearance’ or subject-aspect of the cognition.

On this interpretation of Dignāga’s reflexivist theory of consciousness, the

object- and subject-aspects of experiences are phenomenal modes of presentation. The

object-aspect presents the object (e.g., as being red and spherical), while the subjective-

aspect presents the experience itself (e.g., as being a pleasant, focused, visual experience,

as well as being a cognition of a red sphere). They are phenomenal modes of presentation

  18  

in that the object and the experience are presented qualitatively. In modern parlance, we

can say that there is something it is like to be aware of a red sphere and there is also

something it is like to live through an involuntary, attentive, pleasant, visual experience

of a red sphere.8

Note that, in the above McGinn passage he remarks that the subjective face is

presented to the subject. Likewise, in Dignāga, the dual aspects of experience are

appearances (ābhāsa) and, arguably, it is a feature of the conceptual grammar of

‘appearance’ that it is always ‘appearance-to’. Indeed, in so far as our discussion of

luminosity has taken it to be a form of phenomenal presentation, the very idea of

consciousness as presentation seems to presuppose presentation-to. This fits well with

views that accept the self (ātman) as the enduring subject of experience, but seems to

present a problem for Buddhist philosophers for whom anātmavāda (the doctrine of no-

self) is foundational. In short, if both the objective and subjective aspects are

presentations, to whom or to what are they presenting?

For Buddhist reflexivists like Dignāga, the answer is svasaṃvedana (‘self-

awareness’, svasaṃvitti). Here ‘self-awareness’ or ‘reflexive awareness’ denotes the

primitive, direct acquaintance one has with one’s own experience. To be presented an

object in experience is to be aware of the object as it is given in and through that

experience, whether or not one thematizes the object as experienced. To experience an

object is also to live through the experiencing directly. In both cases, we may say, there is

something it is like to have the experience and that is a function of how the experience

presents the object and how it presents itself. Reflexive awareness, then, is the direct

awareness one has of that which is presented (objectively or subjectively) in experience.

  19  

It is, in other words, the basic awareness of the objective and subjective faces of each

cognitive episode.

The Buddhist reflexivists, then, are committed to something like what David

Rosenthal (1997) has called the Transitivity Principle, according to which a subject is in

a conscious state M only if the subject is, in some suitable way, aware of M (or of being

in M). The subject need not be reflectively or attentively aware of M, but must only be

aware of it ‘in some suitable way’. The Transitivity Principle does not entail reflexivity,

because one might, like Rosenthal, hold the higher-order representationalist (HOR) view

that M is represented by a distinct, second-order state. It does, however, entail a rejection

of the independence condition maintained by Nyāya. What is distinctive about the

reflexivist view is that self-awareness is not a distinct higher-order state, but an intrinsic

feature of consciousness. It is a same-order view of pre-reflective self-awareness.

Dignāga’s central argument for self-luminosity has come to be known as the

memory argument. Dignāga states that one cannot remember what one has not

experienced before. But we can remember our own previous experiences—that is, we can

remember not just the object of a previous experience, but also the previous experience

itself. Thus the experience must itself have been experienced. If the prior cognition is

cognized by a distinct cognition, as in the higher-order view, then there would occur an

infinite regress. This is because, according to Dignāga, the higher-order experience too

can be remembered and so must itself have been experienced, and so on. To avoid the

regress, he argues, we must hold that experiences are reflexive—that is, that the

awareness of the experience is not separate from the experience itself.

  20  

For example, in order for me to genuinely remember a particular sunset, I must

have actually seen it—‘memory’ here is a success term. On Dignāga’s view, I can also

remember my experiencing the sunset. I can remember the vividness of the perceiving,

the way my attention was grabbed by the explosion of color and shape, the mix of joy and

melancholy that accompanied seeing the sunset, and so on.9 But if I can only remember

that of which I have previously been aware, then I must have been (however pre-

reflectively or inattentively) aware of my perceiving the sunset. I have access through

memory to both the object-appearance (the beautiful red sunset) and the self-appearance

(what it was like for me to be perceiving the sunset on that occasion). That access

presupposes awareness of both aspects of the experience. Note, then, that Dignāga is

arguing for something similar to the transitivity principle of consciousness, but with

regard to memory. A subject can only remember a state M if the subject was in some

suitable way aware of M. But if my awareness of seeing the sunset is a separate state M*

and I can also remember my awareness of my seeing of the sunset then M* would need a

further state M**, and so on. Dignāga concludes that reflexivism wins by eliminating the

alternative. However, it is open to the anti-reflexivist to respond that the regress is

blocked by making M* unavailable to memory. Perhaps, following HOR theories of

consciousness, M* is an unconscious cognition of M. Dignāga does not address this

response.

The upshot of the meta-cognition argument and the memory argument, for

Dignāga, is that experiential episodes are complex events10 involving an object-

appearance, a subject-appearance, and the basic awareness of these appearances, where

these features are inseparable aspects of experience. Now, in one respect, the role of

  21  

svasaṃvedana is epistemic. Indeed it comes to be treated by Buddhist epistemologists as

the most basic and secure means of knowledge (pramāṇa). The idea here is that we have

a direct (i.e., immediate, non-inferential) and perhaps even infallible acquaintance with

the phenomenal contents of our experience. To have a conscious pain is to be aware of

the qualitative pain directly, just by having it. Moreover, even if there is no scarlet sphere

in my immediate environment, I am still directly aware that I am having an experience as

of a scarlet sphere. On this view, then, there is no phenomenal presentation—no

presentation of either the subjective or objective face of an experience—without the basic

awareness of those faces. Thus Dharmakīrti argues, “The seeing of objects is not

established for one whose apprehension thereof is itself imperceptible” (PV 1.54). That

is, if one is not at all aware of the experience in and through which the object is

presented, then the object is not phenomenally present at all. Furthermore, note that the

svābhāsa and viṣayābhāsa are given to or given within a conscious, first-person point of

view. For the Buddhist reflexivists, svasaṃvedana constitutes the conscious point of view

within which the two faces of cognition are given. Reflexivity, therefore, constitutes a

minimal form of subjectivity in the phenomenological sense of a ‘dative of

manifestation’, that to which the phenomenally present is presented. Yet, crucially, this

point of view is not a distinct or enduring subject of experience existing over and above

the interconnected episodes of experience constituting individual streams of

consciousness (cittasantāna). Rather, as we have seen, reflexive awareness is a basic

feature of each individual experiential episode. At bottom, each episode of experience is

its own subject.

  22  

In addition to the epistemic role of svasaṃvedana, we also find it playing a

transcendental role—that is, self-luminosity comes to be seen as the distinguishing mark

(svalakṣana) or very nature of consciousness. In Madhyamaka lam ka ra 16, S antaraks ita

(2005, p. 53) argues:

Consciousness rises as the contrary

Of matter, gross, inanimate.���

By nature, mind is immaterial���

And it is self-aware.

On this view, matter is inherently inanimate and insentient (jad a), while consciousness is

inherently luminous and cognizant – that is, reflexive and intentional. There is nothing it

is like to be a stone, and it has no states that are intentionally directed toward an object. In

contrast, dynamic sentience is the very mode of being of consciousness, and for

S antaraks ita, the sentience or phenomenality of consciousness is understood in terms of

its reflexivity. As his Tibetan commentator, Jamgon Mipham, remarks in this context:

Objects like pots, being material, are devoid of clarity [luminosity] and

awareness [cognizance]. For them to be cognized, it is necessary to rely on

something that is quite different from them, namely, the luminous and

knowing mind. The nature of consciousness, on the other hand, is unlike

matter. For it to be known, it depends on no condition other than itself. ...

In the very instant that consciousness arises, the factors of clarity and

knowing are present to it. Although other things are known by it, it is not

  23  

itself known by something else and is never without self-awareness (it is

never ‘self-unaware’) (S antaraks ita 2005, p. 202).

The distinction drawn here is similar to Searle’s distinction between those things with a

first-person ontology and those with a third-person ontology. On Searle’s view,

“consciousness has a first-person ontology; that is, it only exists as experienced by some

human or animal, and therefore, it cannot be reduced to something that has a third-person

ontology, something that exists independently of experiences” (Searle 2002, p. 60).

Objects like pots do not have experiences and apparently exist independently of their

being experienced. Conscious states, on the other hand, do not exist independently of

being experienced—their very mode of being is to be experienced

Mipham goes on to argue, following the Indian Buddhist reflexivists, that:

It is thanks to reflexive awareness that, conventionally, phenomenal

appearances are established as the mind, and the mind [i.e. a cognitive

episode] is in turn undeniably established as the object-experiencer. If

reflexive awareness is not accepted, the mind would be disconnected from

its own experience of phenomena and the experience of ‘outer objects’

would be impossible (S antaraks ita 2005, p. 123).

That is, for these thinkers, the experiential object is recognized to be a phenomenal

appearance (ābhāsa) or representation (ākāra) that is not distinct from the cognition

within which it is presented. In other words, one sees that the supposed external object is

  24  

in fact merely the objective-face of an experience and thus an aspect of the experience

itself. Further, on this reflexivist view, absence of pre-reflective self-awareness would

yield a kind of mind-blindness (‘the mind disconnected from its own experiences’)

wherein at any given time one might be having any number of phenomenal experiences

without any awareness that one was having them, in absence of which their intentional

objects would not be phenomenally present. In such cases, one’s cognition would be

more like blindsight than phenomenal consciousness. That is, if conscious states have a

first-person ontology—they only exist in so far as they are experienced or undergone—

then they presuppose a subjective or first-person point of view within which they appear.

This basic conscious point of view is reflexive awareness, in the absence of which one

would have no access to one’s own states and their contents. Hence, on this view,

reflexivity or luminosity constitutes the necessary condition of any phenomenal

appearance, subjective or objective. In this sense, reflexivity or self-luminosity comes to

be seen as transcendental.

Whereas the Buddhist epistemologists recognise both reflexivity and

intentionality as fundamental aspects of consciousness, the Advaita Vedānta school takes

only reflexivity to be the essence of consciousness.11 Consciousness (cit) in its

fundamental nature is pure reflexive subjectivity. What we normally think of as the

intentionality of consciousness itself actually arises from the association of pure non-

intentional consciousness with certain non-cognitive mental states (vṛtti). As Ram-Prasad

characterises it:

  25  

This Advaitic conception of consciousness as essentially reflexive in fact

is tantamount to saying that it is purely reflexive. Indeed, this is the idea

behind the conception of consciousness as ‘witness’ (sākṣin …). Just as

onlookers do not engage in the events they are witnessing, so witnessing-

consciousness does not engage with objects. It is present, but it is

transparent to content, not itself intentionally directed towards (i.e.

‘engaged with’) objects (Ram-Prasad 2007, p. 80).

Moreover, it is important to note that, for the Advaitin, the ātman is not a substantial self

such as was defended by the Naiyāyikas, but rather, the self is pure reflexive

transcendental subjectivity. Thus, in contrast to the Buddhist theory of reflexive

awareness, for the Advaitin it makes no sense to say that individual mental states or

events are self-luminous. Conscious mental states are immediately present, not needing

an additional second-order mental state to reveal them. They are, therefore, self-

presenting in a derivative sense in that they need no further cognitions to reveal them.

They are not, however, present to themselves, as in the Buddhist view. Rather, conscious

states are immediately present to the self as pure witnessing subjectivity. The ātman, as

pure consciousness, is the self-luminous source of illumination for any phenomenon

whatsoever, ‘internal’ or ‘external’, and cannot itself become an object of cognition. As

the condition of the possibility of any presentation of an object, consciousness is not one

object among others, yet it is indubitably present. So, while the Buddhist view of

luminosity focuses on the internal structure of individual, empirical cognitive events, the

  26  

Advaita account of luminosity focuses on a strongly transcendental notion of

subjectivity.12

The Advaita philosopher Citsukha, for instance, carefully defines self-luminous as

that which is immediately evident (aparokṣa), but not an object of knowledge (Ram-

Prasad 2007, p. 75). It is immediately evident in the sense that we cannot be mistaken

about whether we are conscious. The earlier Advaitin, Śrī Harṣa, unpacks this familiar

Cartesian point in terms of three absences (Ram-Prasad 2007, p. 75). First, there is the

absence of doubt over whether one is having an occurrent conscious cognition. Second is

the absence of a meta-cognition that one has no first-order cognition, when one does have

a first-order cognition. Third, is the absence of a meta-cognition that one does not have a

first-order cognition when one has no first-order cognition. Whatever we make of Śrī

Harṣa’s view, the basic point is straightforward: we have an immediate and indubitable

acquaintance with at least the existence of our occurrent consciousness. For the Advaitin,

we can be mistaken about the contents of consciousness in a variety of ways, but the fact

that we cannot be mistaken about whether we are conscious at all, just supports the

distinction between consciousness itself (cit) and its various contents (viṣaya) and

modifications (vṛtti). Citsukha is keen to maintain that consciousness is never its own

object, because it is never an object at all. On this view, an object is that which is

revealed by consciousness, and that which is revealed distinct from that which reveals.

Indeed, on this account, what it is to be an object is to be presented as distinct from the

experience of it. In this way it is similar to the phenomenological notion of an object as a

Gegenstand—that which stands ‘over-against’ one’s awareness of it. If consciousness

were to know itself as an object, it would thereby falsify itself by occluding the revealing

  27  

of that object in the knowing itself. Hence, cit is self-conscious without positing itself as

an object. As the source of all revealing, then, consciousness is pure unobjectifiable

subjectivity.

On the one hand, consciousness cannot be known as an object, while, on the other

hand, its occurrence cannot coherently be doubted. If consciousness is known as an

object, they argue, there would be a infinite regress, owing to the distinction between

awareness and objects of awareness. If it is denied that consciousness is self-evident, then

one would, it seems, be forced into the performative self-contradiction of denying that

one has experience. In fact, Śaṅkara argues that because consciousness is the condition of

experiencing any object of knowledge, its existence cannot be overturned—that is, it

cannot be denied on the basis of any future experience, since any future experience will

itself presuppose consciousness. Any object of consciousness can be doubted, but

consciousness itself cannot.

So, whereas the Yogācārin view of self-luminosity plays a central role in an

account of experience that unifies reflexive awareness and the subjective and objective

aspects of experience, the Advaitins sharply distinguish the self as pure reflexive

consciousness from all objects and modifications of experience. Moreover, while the

Buddhists hold that episodes of experience are impermanent occurrences within an ever-

changing stream of consciousness (cittasantāna), the Advaitins maintain that self-

luminous consciousness is changeless and permanent. As the changeless background of

all changing experience, this witness-conscious is taken to account for the diachronic

unity of experience. According to Advaita, the Buddhist view of experience as made up

of causally connected moments of experience—self-luminous or not—cannot account for

  28  

either the diachronic cognition of objects or of the stream of consciousness itself.13

Śaṅkara argues:

The perception of similarity takes the form of ‘This is like that’. ‘That’

refers to the remembrance of something seen: ‘this’ to the perception of

something present. If after remembering the past experience denoted by

‘that’, consciousness should linger till the present moment referred to by

‘this’, then the doctrine of momentariness would be gone. If, however, the

remembrance terminates with the notion of ‘that’, and a different

perception relating to the present (arises and) dies with the notion of ‘this’,

then no perception expressed by, ‘This is like that’, will result, as there

will be no single consciousness perceiving more than one thing. Moreover,

it will be impossible to describe our experiences. Since consciousness

ceases to be just after seeing what was to be seen, we cannot use such

expressions as, ‘I see this’, or ‘I saw that’, for the person who has seen

them will not exist till the moment of making these utterances (BAU-B

4.3.7; Deutsch & Dalvi 2004, p. 138).

The Buddhist is faced with a dilemma. Because perception of similarity (or identity)

requires comparison between an earlier and a later perception, either there is a single

enduring consciousness that has both perceptions and the doctrine of momentariness

(kṣaṇikavāda) is false, or there is no enduring consciousness and each perceptual event is

locked in the solipsism of the present moment. In the latter case, no perception of

  29  

similarity is possible. Third, Śaṅkara argues that if consciousness is momentary, then

there can be no diachronic continuity of the first-person perspective. Moreover, note that

Śaṅkara sees very clearly the deep phenomenological connection between the experience

of persisting objects and the experience of oneself as a persisting subject. Thus, on

Śaṅkara’s view, one must either accept an enduring self or consciousness, or be faced

with an experientially disconnected series of mental events.14

The standard Buddhist response to this objection is to appeal to the causal and

functional connections between mental events within a single stream of consciousness.

Of course, the mere fact that one event causes another (even when those events are

mental) does not entail that the two events are experientially continuous. In order for the

Buddhist response to be plausible, the causal-functional connections must ground

phenomenal continuity. So phenomenally continuous mental events must not only be

causally connected, they must both be part of the same phenomenal point of view. In the

Advaita account this point of view is provided by witness-consciousness as an enduring

subject of experience over and above the various mental events (vṛttis) that constitute the

stream of experience. On the other hand, Buddhist philosophers will appeal to a

continuous, but not enduringly self-identical, point of view as a feature of the connection

mental events. For Buddhist reflexivists, svasaṃvedana can be pressed into service here.

First, reflexive awareness constitutes the synchronic phenomenal point of view in that it

is that feature of awareness to which the subject-aspect and object-aspect are present.

Second, as Dignāga argues, reflexive awareness plays a central role in the diachronic

relations of access-consciousness in memory. That is, svasaṃvedana allows for the

apprehension of both aspects of experience by a later experience and this later

  30  

apprehension is from the inside (Ganeri 2012). Hence, on this view, it is the self-

luminosity of consciousness that constitutes the synchronic and diachronic phenomenal

point of view, not an enduring subject.15 The Advaitin’s mistake, it might be said, is to

slide from a continuous point of view within a stream to an enduring subject above or

behind the stream (MacKenzie 2012).

Conclusion

By way of conclusion, it will be useful to draw some further connections between

the Indian debates around the nature of luminosity and some contemporary issues in

philosophy of mind. The first and most obvious connection involves the long-running

Western debates over the connection between consciousness and self-consciousness. This

debate goes at least as far back as Aristotle and continues up through the present (Caston

2002). In the classical phenomenological tradition (including Brentano (1995)) it is the

orthodox view that consciousness constitutively involves pre-reflective self-awareness

(Zahavi 2006). Further, the connections between phenomenological and Indian accounts

are the subject of a thriving current literature (MacKenzie 2007, Thompson, et al 2011).

In particular, the connection between reflexive awareness, subjectivity, and the self are

ripe for cross-traditional exploration. Similarly, in analytic philosophy debates between

first-order representationalist, higher-order, and self-representationalist views of

consciousness present a fruitful area of connection (Kriegel and Williford 2006). Indeed,

the self-representationalism of Kriegel (2009) bares remarkable resemblance to that of

Dignāga, while the first-order representationalism of Dretske is in some important

respects similar to that of Nyāya. Also worth exploration would be the connections

  31  

between Bhaṭṭa views and the dispositional higher-order views of Carruthers (2000) and

others. Further areas of connection include debates over the nature of phenomenal

consciousness, the transparency of experience, the relation between self-awareness and

intentionality, the temporality of experience, the senses of ownership and agency,

personal identity, externalism and self-knowledge, and bodily self-awareness.

Despite the antiquity of some of the Indian views and thinkers discussed above, I

believe they offer important insights and resources for contemporary thinking. Not only

do we find interesting—and in some cases, remarkably contemporary—views and

arguments among classical Indian accounts of the luminosity of consciousness, we find

well-integrated models, covering a range of important aspects of the mind and its

operations. These models are worthy of considerations as part of the global history of

philosophy, but also, in some cases, as live options for contemporary philosophy of mind.

Moreover, cross-traditional philosophy of mind, it seems to me, is especially valuable

insofar as it allows us to engage with models and traditions that may have quite different

background assumptions and commitments, and deals with a different or perhaps wider

range of experiences, than our current discourse.

  32  

                                                                                                               1 This capacity is in the Nyāya ontology technically a quality (guṇa), as distinct from a

substance or action.

2 It is important to note that while anuvyavasāya is a function of the inner sense (manas),

which is also responsible for voluntary selective attention, apperception here is not a case

of voluntary attention.

3 Sartre himself does not draw this conclusion and endorses the idea that consciousness is

self-luminous.

4 There are, of course, other strands in the Nyāya view, particularly having to do with

epistemology.

5 By ‘person-level’ here I mean, roughly, a fully, attentively conscious mental state, as

opposed to either a subliminal or merely background state, such as inattentively hearing

the hum of a refrigerator.

6 This is one of the phenomenological uses of śūnyatā.

7 Note that the relation between svābhāsa and svasaṃvedana is controversial. On some

views, self-awareness just is the subject-aspect’s apprehension of the object-aspect. On

my view, self-awareness is the apprehension of both faces. Yet, insofar as the subject-

aspect is the presentation of the cognition as that very cognition itself, there will be an

intimate connection between the svābhāsa and svasaṃvedana.

8 For an interpretation of the subject-aspect that does not appeal to phenomenal character

see Ganeri 2012. Though, I must admit that, on my notion of phenomenal character,

anything that presents a ‘subjective face’ to a first-person point of view is, by definition,

an aspect of phenomenal character.

9 These are, of course, my examples, not Dignāga’s.

  33  

                                                                                                                                                                                                                                                                                                                                         10 The event are not mereologically complex, but rather aspectually complex.

11 Actually, things are a bit more complicated here. Because the Yogācārins hold that,

ultimately, there are no objects apart from consciousness, it can be said that reflexivity is

more fundamental than intentionality. The point here, though, is that for the Yogācārin,

both grasper and grasped are part of consciousness, whereas for the Advaitin,

consciousness only appears to be intentional.

12 The ātman or witness here is the dative, the ‘for whom’, of any phenomenal

manifestation.

13 The following draws from MacKenzie 2012.

14 As Śaṅkara argues later in the same passage, mere causal connection between mental

events is not sufficient to give experiential continuity.

15 Note here that for the Buddhist reflexivist, the phenomenal point of view is parasitic on

the moments of consciousness that form the causal continuum, rather than belonging to

an enduring subject—like a series of beads arranged so that the hole in each bead is

aligned with others to form an opening through the whole series.