Vijnaptimatrata and the Abhidharma context of early Yogacara
by Richard King
Vol. 8 No. 1 Mar.1998
Contemporary accounts of early Mahayana Buddhist schools like the Madhyamaka and the Yogacara tend to portray them as generally antithetical to the Abhidharma of non-Mahayana schools such as the Theravada and the Sarvastivada. This paper attempts to locate early Yogacara philosophical speculation firmly within the broader context of Abhidharma debates. Certain key Yogacara concepts such as alayavijnana, vijnapti-matrata and citta-matra are discussed insofar as they relate to pre-existing concepts and issues found in the Vaibhasika and Sautrantika schools, with specific reference to the Abhidharmakosa and the corresponding bhasya of Vasubandhu. Finally, some remarks are made about the benefits of approaching the history of religious ideas without the benefits and distortions of hindsight, particularly as this relates to the attribution of an idealistic position to the early Yogacara literature.
Contemporary accounts of Mahayana Buddhist schools like the Madhyamaka and the Yogacara tend to portray them as generally antithetical to the Abhidharma of non-Mahayana schools such as the Theravada and the Sarvastivada. There are a number of reasons for this. Firstly, it often reflects the tendency to conceive of internal Buddhist doctrinal controversies as the source of schisms in a manner akin to the disputations of Christian history. As Heinz Bechert has pointed out, the principle of schism (sanghabheda) in Indian Buddhism was based upon disputes over monastic code (vinaya) and not differences in doctrinal position . Secondly, increasing examination of Tibetan commentarial materials has allowed Buddhist scholars to provide a much fuller account of the history of Mahayana thought. However, one consequence of this has been the tendency to accept interpretations of Buddhist scholastic thought which often derive from the work of authors with allegiances to other perspectives. In the case of the Yogacara school, for instance, scholars have generally tended to accept the accounts provided in Sankara's Brahmasutrabhasya or in the commentarial works of Tibetan Madyamaka schools like the dGe lugs pa without questioning the status of these accounts.
Working with the benefits of hindsight, it is sometimes difficult to see the wood for the trees and it often proves fruitful to return to the early stages of a tradition and its literature in order to gain a fresh perspective upon the material. Indeed it would seem that the search for new conceptualisations and understandings of original teachings is precisely what the history of religious doctrine has been about. In the case of the Yogacara school it is easy to follow established commentarial traditions in interpreting this school's distinctive philosophy. In most cases this would seem to be both the most pertinent, fruitful and humble course of action to take. However, the danger of hindsight and allegiance to specific hermeneutic traditions is that they quickly tend to become institutionalized. As I have argued elsewhere, we should avoid projecting later debates and controversies between Buddhist schools of thought (for example between the Madyamaka and Yogacara trends of Mahayana Buddhism) into the early literature of such movements .
One way to redress the balance in the study of Buddhist thought in India is to make a simple point that is so often overlooked. Buddhist philosophical debate in India took place within an Abhidharmic context. After all, in what other theoretical and literary context could such debate have occurred? Once one acknowledges that the differentiation between Buddhist schools of thought was not a crucial factor in the schismatic development of sanghas (sanghabbeda) one realises the sense in which this must indeed be the case. When doctrinal disputations bring forth issues relating to Buddhist monastic ethics and practice, schism, of course' may very well result. However, in the history of Buddhist thought it seems that in most cases philosophical disputations between rival schools left fundamental issues of religious practice largely intact. For this reason, in India at least, Mahayana and non-Mahayana adherents could remain members of the same sangha without any fundamental conflict of interest or danger of schism.
It is clear that the portrayal of Mahayana Buddhists as generally antithetical to the Abhidharma of non-Mahayana schools such as the Theravada and the Sarvastivada remains something of an overstatement in an Indian context. Abhidharma, broadly speaking, is Buddhist philosophy. Madhyamika and Yogacara thinkers both established and contested their theories from within a theoretical framework which was unquestionably Abhidharmic in style, content and presentation . This should not be a surprise, particularly once one realises the sense in which the group-identity of a given sangha is established by the Pratimoksa and the rules of the Vinaya rather than any particular philosophical doctrine or position.
Simplistically speaking, the history of religious ideas can be approached from two basic directions: backwards and forwards. The former approach involves interpreting a particular movement or school of thought utilising subsequent elaborations and commentarial expositions of it. The latter approach involves beginning before the movement has even arisen and examining the antecedent conditions, movements and potential influences upon that movement as a way of contextualising its earliest stages. This paper is a brief attempt to approach early Yogacara philosophy from the latter perspective. In taking this approach to the material I wish to place much greater emphasis upon the Abhidharma context of Mahayana philosophy by suggesting a number of ways in which the Sautrantika analysis of perception may have functioned as an important precursor of the Yogacara's own position. I will attempt to demonstrate this by discussing certain Abhidharma antecedents to the Yogacara notions of alayavijnana and vijnaptimatrata.
The Abhidharmakosabhasya and Early Yogacara
The Sautrantika and Yogacara epistemologies are similar despite the through-going realism of the former and the apparent idealism of the latter. The Sautrantika accept that it is only the form (akara) or representation (vijnapti) of an object which is perceived. Where the schools differ is in the Yogacara refusal to accept the validity of discussing external objects as causes (nimitta) given that an external object is never (directly) perceived.
The thesis of the similarity between the two schools can be illustrated if we examine the views of the two Vasubandhus, the author of the Abhidharmakosa and its commentary (Abhidharmakosa-bhasya), and the author of various early Yogacara works such as the Twenty Verses (Vimsatika) and the Thirty Verses (Trimsika). These two thinkers may or may not be one and the same person as tradition suggests , but given the similarity between the Sautrantika and Yogacara epistemologies, movement from the former position to the latter is not totally incomprehensible. In discussing the doctrinal position of the author of the Abhidharmakosa-bhasya, Thomas Dowling suggests that
Vasubandhu occupies a unique doctrinal position vis a vis the question of what constitutes a dharma. On the one hand, he is unwilling to accept the over inflated dharma list of the Vaibhasikas, and on the other, he is not committed to the radical perspective of the ultimate voidness of all dharmas (in the Kosa, at any rate.) His lack of recourse to such terms as paramartha satya/samvrti satya in their specifically Mahayana sense is evidence that the Kosa is a straightforward text grounded in the tacit assumption that in some sense there 'are' dharmas .
In the Abhidharmakosa-bhasya Vasubandhu criticises the Sarvastivada notion of `possession' (prapti) arguing that the difference between possession and non-possession (aprapti) is merely the state of having destroyed or not destroyed the defilements. This is a classic example of the Sautrantika critique of the Vaibhasika position, curtailing the Vaibhasika tendency to postulate theoretical and unexperienced entities (this is seen for example in the Sautrantika denial of the category of citta-rupa-viprayukta-samskaras, that is `formations neither associated with consciousness nor form').
The Sautrantikas criticised the postulation of such entities as prapti and avijnaptirupa, replacing this scheme with a causal model based upon the notions of karmic seeds (bija) and the transformation of particular streams of consciousness (samrana-parinamavisesa) . This suggests another strand of continuity between Vasubandhu qua Sautrantika and Vasubandhu qua Yogacarin in the emphasis both place upon the transformation of consciousness as a means of explaining the fluctuating nature of samsara. Again, such notions as shape (samsthana), taken to be substantial and real (dravya-sat) by the Vaibhasikas, are purely mental conceptions (parikalpam kurvanti) arising from visual perception according to Vasubandhu in the Abbidharmakosa-bhasya . This account may prove to be a precursor of the attack upon the notion of a six-sided atom in Vasubandhu's Twenty Verses (Vimsatika), verses 12-14.
In the critique of 'possession' (prapti) in the Abhidharmakosa-bhasya Vasubandhu qua Sautrantika seems to utilise a notion which becomes of crucial importance in the subsequent Yogacara elaboration of the path to liberation, viz. asraya-paravrtti, the conversion of the basis. He states that
Verily, the physical basis of the Noble One has undergone transformation by virtue of the path of vision and the path of cultivation such that those defilements that are allayed no longer have the ability to shoot forth. As rice seeds that are in a non-germinal (or impotent) state, just so one is called a 'destroyer of the defilements' with reference to the defilements of the physical basis (bhutasaraydh). (my italics) 
Thus, through the 'conversion of the basis' one may be called a `destroyer of defilements' (prahinaklesa). Vasubandhu also goes on to argue that upon realisation of the supreme goal of yogic attainment (nirodha-samapatti), an untainted stream of consciousness (nirmala santati) is produced:
For the one who has returned from the Path of Vision (darsana-marga), as a result of destroying all of the defilements that can be destroyed by Vision, without remainder, there occurs a fresh stream [of consciousness] that is without blemish and characterized by revulsion of the physical basis .
The idea of a `conversion of the basis' asraya-paravrtti), that is a purification of consciousness through the eradication of all defilement (klesa), becomes an important theme in the subsequent development of the Yogacara school. The classical formulations of Asanga and Vasubandhu tend to portray this conversion as a destruction of an essentially phenomenal store-consciousness (alayavijnana), the repository of karmic seeds (bija). However, later interpretations within the Yogacara (for example the work of Paramartha) envisaged this transformation as an eradication of defilements which leaves behind an essentially undefiled consciousness (amala-vijnana). This pure consciousness was seen as the foundation or support (asraya) which originally formed the basis for the activities of the now defunct defilements. On this view, the conversion of the basis no longer means the cessation of the store-consciousness, but rather its transformation and re-turn (paravrtti) to its former pristine condition. This 'pure mind' tradition within Yogacara Buddhism has clear antecedents in early Buddhism  and is perhaps best represented in the early Yogacara literature by such texts as the Mahayanasutralamkara. As such it reflects not only the open-endedness (ambivalence?) of many Yogacara terms, but also the assimilation of ideas usually associated with the Mahayana notion of tathagatagarbha . The Abbidharmakosa-bhasya thus provides interesting source-material for all of these subsequent Yogacara developments.
Taking the argument one step further one might wish to argue that the notion of the alayavijnana, as utilised by Vasubandhu the Yogacarin, is little more than an elaboration of concepts already expounded in the Abhidharmakosa-bhasya, which of course expounds the Dharma from a Sautrantika perspective. This view, in fact, is propounded by Asanga in Mahayanasamgraha I. 11 where he argues that the notion of the alayavijnana far from being a Mahayana innovation 'is mentioned in the sravakayana by means of various synonyms (paryaya)'. It is tempting to point to the Theravada notion of bhavanga-citta as influential in this regard, though Lambert Schmithausen argues that this is unlikely precisely because it is not mentioned by Asanga at this point . Nevertheless, Rupert Gethin is surely right to suggest that 'these two concepts are to be understood as having a certain affinity and that they belong to the same complex of ideas within the history of Buddhist thought' .
In the Abbidharmakosabhasya Vasubandhu asks,
ko 'yam bijabhavo nama atmabhavasya klesaja klesotpadanasaktih
And what is called the seed state? It is the power to originate defilement, produced by the defilement of one who has attained existence .
This characterisation conforms to the early Yogacara conception of the alayavijnana as sarva-bijaka-vijnana, the store of (defiled) karmic seeds. Schmithausen offers a list of twenty uses which the concept of alayavijnana provided (14 'philosophical' and 6 exegetical) for the early Yogacarins . Most of these cluster around the explanation of personal continuity given the absence of an abiding-self, and providing a link between karmic action and subsequent fruition. The Sautrantika metaphor of the seed (bija) became central in the case of the latter issue once the Vaibhasika conception of the existence of dharmas in past, present and future (the sarvastivada position) was rejected. However, as Schmithausen points out, although the Sautrantika postulated the notion of a karmic seed to establish causal continuity over time, the Yogacara seems to have felt that this required the further postulation of a store (alaya) consciousness as the repository of these seeds. Nevertheless, it is important to note at this point that the store-consciousness is by no means considered to be an ultimate reality in the works of either Vasubandhu the Yogacarin or Asanga, as has sometimes been suggested. In the Viniscayasamgrahani section of the Yogacarabhumi, Asanga describes the alayavijnana as the root of defilements (samklesamula) which ceases through the cultivation of wholesome dharmas . Equally, in Trimsika v.18 Vasubandhu suggests that the alayavijnana is nothing more than a collective term for the seeds themselves. For the Sautrantika author of the Abhidharmakosa-bhasya the notion of the seed (bija) is equally nothing more than a power (sakti) within the five aggregates--specifically it is a designation (prajnapti) for an uncognised process (asamjnayamanah) within the skandhas . This also appears to be the view put forward in Vasubandhu the Yogacarin's other major works . For the Sautrantika author of the Abhidharmakosa-bhasya, this means that the adoption of conceptual categories other than the analysis of the pudgala in terms of the five skandhas is both scripturally and logically questionable. This is pure Sautrantika (the name 'Sautrantika' itself designating those who believe that the Sutra-pitaka alone contains the actual words of the Buddha). Thus, upon closer examination we find that many of the most important 'new' Yogacara concepts (such as vijnana-parinama and alaya-vijnana) as utilised in the various Mahayana sastras attributed to Vasubandhu, seem to be philosophical elaborations or extensions of concepts and themes already found in the Abbidharmakosa-bhasya.
Abhidharma Antecedents of Vijnaptimatrata
Let us consider, for instance, another notion which is central to the Yogacara school--that of Vijnaptimatrata or `Cognitive-Representation Only'. It is no surprise, given our previous discussion, to note that the term `vijnapti' also has a historical background in Abhidharma scholasticism. Vijnapti is a technical term of Abhidharma Buddhist philosophy, often translated as `intimation' or `information'. `Avijnapti', as its negation, would therefore be rendered as `non-intimation' or `non-indication'. The Vaibhasika Abhidharma distinguishes between two forms of 'intimation' (vijnapti) .
(1) vagvijnapti (or vacika-vijnapti)--verbal intimation or expression, and
(2) kayavijnapti (or kayikavijnapti)--corporeal intimation or gesticulation.
In early Mahayana literature `vijnapti' came to designate any cognitive act which is carried out by the mind (manas). In the Yogacara school all experience is claimed to be fundamentally 'mental' in nature. More specifically, an individual's experience is constituted by a series of projections externally intimated from the store-consciousness (alayavijnana) of karmic impressions (vasana). There appears to be no explicit distinction here between types of vinapti as we find in the Vaibhasika texts.
In direct contrast to the notion of vijnapti is the Vaibhasika notion of avijnapti-rupa . avijnapti-rupa, or 'matter which is not-manifested in consciousness', is matter (rupa) of the highest subtlety . It offers no resistance to the sense organs and does not even allow itself to be `touched' (sparsariyate) by consciousness (vijnana-dharma). Thus the postulation of its existence remains purely inferential. It is the subtle residue left over by the physical vijnaptis. avijnapti-rupa retains the moral quality of the vijnaptis from which it originates in terms of the passively accumulated seeds (bijas) or perfumes (vasana) which later reach fruition in the form of future karmic retribution. This residue is described as material since it is compounded by the four basic elements (mahabhuta) . Takakusu thus describes the avijnapti-rupa in the following terms:
Of the eleven [[[rupa]] dharmas], the first five are sense-organs and the next five are sense-objects. The four gross elements--Earth, Water, Fire, Air--are represented by the sense-objects. In addition to these, there is a peculiar one. That is the 'form-element not manifested' outwardly (avijnapti-rupa). When we will to act the mental function itself is called will (cetana). Inis called will-action. This is usually expressed in words or in body, and is called word-action or body-action respectively. These two actions manifested outwardly, whether they are good or bad, present a corresponding and similar action in mind, and form an abiding impression or image. They are then called unmanifested action (avijnapti-karma). These actions being taken as form-elements are considered to be sense-objects though not manifested (avijnapti-rupa) .
`Avijnapti' as a category therefore represents the result of (karmic) actions which has yet to manifest. As such it is the notion of a latent impression caused by karmic activity. For the Vaibhasikas this 'entity' was postulated to account for the temporal discrepancy between an action and its subsequent (karmic) fruition. Here again we see that vijnapti denotes the manifested result of karman. One hardly need point out that this is also a fundamental aspect of the Yogacara usage of the term and may be said to illustrate further the continuity of thought between the Vaihasika/Sautrantika complex of Abhidharma notions under discussion in the Abhidharmakosa-bhasya of Vasubandhu and the Yogacara philosophy expounded by the author of the same name in various Mahayana sastras.
Vijnapti in the Yogacara context is the manifested fruition (vipaka) of traces of past karmic activity (vasana) in the constructed form (parikalpita) of an apparently new experience. Thus our entire network of perceptions is perpetually conditioned by our own past choices and actions. The responses that we make to these 'new' experiences themselves condition the nature of future experiences through the establishment of karmic traits (vasana) deposited in the store-consciousness (alayavijnana). Vijnapti for the Yogacara then is not simply the `cognitive-representation of sense-objects' (vijnaptir visayasya, Trimsika v.2), as is usually understood by the term, but is more fundamentally a representation of the agent's own subliminal karmic predispositions (anusaya). In other words, in the Yogacara system vijnapti comes to be seen as a direct reflection of one's own state of mind .
This point can be further illustrated by an examination of the cognitive process as outlined in Asanga's Bodhisttva-bhumi. Asanga attempts to explain in this text the manner in which our cognition of samsara is perpetuated. Accordingly, Asanga argues that the 'pure given-ness' (vastu-matra) of perception is conceptualised (vikaltyate) in sensory apprehension resulting in the construction of an objective-support (alambana) for consciousness. Attachment to these objective-supports (as independently existing entities) perpetuates our experience of samsara through the appropriation of karman in the form of habitual, subconscious forces (samskaras). This position is not as unique doctrinally as some have suggested. In a Buddhist context the doctrine of karman necessitates an acceptance of the view that experience is conditioned in some fundamental sense by the seeds of past actions. What is distinctive about the Yogacara account of the dynamics of karman is the refusal to extend the discussion beyond a purely phenomenological account of karmic appropriation.
Whether Asanga's explanation of the processes of cognition leads to a form of idealism is certainly an interesting question. However, for the Buddhist, as for the believer in karman in general, the important point is the realisation that the objective world which confronts us is, or at least was at some stage in its manifestation, a product of our own intentional actions (karman). This is a central Buddhist idea, expounded for instance in Abhidharmakosa-bhasya IV.1:
It is said that the world in its variety arises from action (karma). It is because of the latent dispositions (anusaya) that actions accumulate (upacita), but without the latent dispositions [they] are not capable of giving rise to a new existence. Thus, the latent dispositions should be known as the root of existence (mulam bhava) .
The Yogacara notion of vijnaptimatrata is an attempt to reformulate this basic Buddhist insight through a comprehensive phenomenological analysis of the activities of the mind. Vijnapti therefore refers, not just to the immediate cognitive-representation of an object in one's mind, but also, more fundamentally, to the representation of those subliminal forces (samskara) which have caused such an object to be presented to consciousness in the first place. In taking this stance the Yogacarin is not thereby committed to a form of subjective idealism since it is abundantly clear that we do not simply imagine the world of external objects that confronts us here and now. Nevertheless, the Yogacarin argues, our experiences are still causally dependent upon the traces (vasana) of past karmic activity.
One should note then that the early Yogacara position at least seems to involve an acceptance that our experience has a certain pre-determined aspect to it. The world that we perceive is not simply `imagined' since we can neither wish it away arbitrarily nor transform it into something else at the slightest whim. Nevertheless, for the Yogacarin the world that we experience is fundamentally the product of our past karmic actions and our reactions to it will determine whether future circumstances will continue to confront us in a similar manner. The world then is not `real' in the sense of containing objective and independent entities (parikalpita), but is real in the limited sense of being a really existing (though causally inter-dependent) flow of perceptions (paratantrastita). This realm is `dependent' insofar as our experiences are dependent upon past karman.
So, the Yogacarin might wish to say something like the following. The world of our experience is actually `there,' if not in the form of a subject-object dichotomy which exists independently of our experience of it. We experience a certain pre-given reality (vastu-matra) in sensory experiences but these are in actual fact a stream of interdependent sensory impressions dependent for their appearance upon the consciousness which is perceiving them. The manifestation of these sensory impressions is ultimately dependent upon the defiled nature of the mind (klista manas). As a result of this we tend to attribute independent existence (parikalpita-svabhava) to what we believe to be the subjective and objective correlates of our experience. The stream of perceptions that we actually perceive, however, is `just mind' (citta-matra), being the conceptualisation (vikalpa/vijnapti) of the pre-given (vastu) into an objective support of consciousness (alambana). Despite this, it must be stressed that for the Yogacarin there is `something' there (viz. the paratantric flow of dharmas) which constitutes the `raw material' of our experience, although in the final analysis this is merely a fruition of seeds caused by past conscious activity (karman).
Thus, it is possible to interpret the Yogacara discussion of the perpetuation of samsara through the appropriation of karman not so much as a denial of the external world but rather as a restriction upon the parameters of legitimate discourse to a phenomenological context (that is a context which does not attempt to postulate entities beyond the pure givenness (vastu-matra) of experience).
It is important to bear in mind that the Yogacara conception of citta/vijnana denotes a whole complex of events and processes which cannot be adequately rendered by English terms such as `consciousness' or `mind'. The `citta' of cittamatra includes within it the conscious apprehension of sensory objects (six in all including the mano-vijnana). This is a crucial point to acknowledge since, for the Yogacara school, the sensory apprehension of objects cannot be divorced from one's consciousness of it (though it is possible to make a purely abstract and theoretical distinction between vedana on the one hand and vijnana, samjna and samskara on the other when discussing the skandhas). In a sense the Yogacara position offers the flipside to the standard Abhidharma position that citta is intentional, that is, that to be conscious is to be conscious of an object. For the Yogacara, to postulate an object requires that it is first apprehended by a citta. The emphasis here is no longer on the suggestion that citta is intentional but rather on the fact that objects of consciousness are just that. Thus, the thesis of the intentionality of citta becomes displaced in the emerging Yogicara philosophy by an emphasis upon the `phenomenalistic' nature of objects. Objects are really dharma-constructs and representations (vijnapti), dependent upon the complex processes of citta for their appearance. Thus, one can talk of apprehending a sensory object only after one has become conscious of it. Sensory apprehension is thereby subsumed by the Yogacara analysis under the broader domain of `citta,' which, now more clearly than ever, remains too rich and all-embracing a term to be rendered by `mind' or `consciousness'. As well as an awareness of sensory objects, citta also denotes the organising faculty of the manas, the affective distortion of that process by the defiled mind (klista manas) as well as the subliminal karmic seeds (samskaras) and latent dispositions (anusaya) that are collectively known as the alayavijnana. The complexity of terms like citta, therefore, when combined with the Yogacara endorsement of the category of rupa-dharma and the acknowledgment that vijnana remains only one of five skandhas suggests that it is problematic to interpret the early Yogacara literature as propounding a form of idealism at least in the sense in which this has commonly been understood in the West.
For the Yogacara school any discussion which transgresses the experiential boundaries of citta leads to the utilisation of conceptual distinctions (vikalpa), idle speculation and conceptual-proliferation (prapanca). It should be noted, however, that this view of the limitations of language and appropriate dialogue is again not without its precedent. The Sautrantika analysis of perception denied that external objects were given in perception; only the images (akara) of the external world are actually perceived by consciousness. The Sautrantika of course did not take this to mean that one could not thereby make veridical statements about an external world. Indeed, by stating that external objects (nimitta) can be inferred from the experience of mental images (akara), the Sautrantika were explicitly accepting that such discussion was appropriate. Although clearly in conflict with the Yogacara position one can see that Sautrantika epistemology is only one step away (albeit a significant one!) from the more radical `phenomenalism' of the Yogacara school. As we have seen, for the early Yogacara of Asanga and Vasubandhu discourse about the nature of an external world is inappropriate precisely insofar as it goes beyond the realm of that which is empirically (in this context experientially) given.
At the risk of labouring the point, we should note that for the Yogacarin it is not the case that we simply `imagine' our experiences. They are `real' to the extent that they are `given' without our conscious intervention. The pure given-ness (vastu-matra) of our experiences is thus beyond our conscious control. The question of `externality,' however, is prevented from entering the Yogacara account since it is a quality which cannot be a veridical aspect of our experience (since if x is really external to our consciousness then it cannot be within its perceptive range). There may or may not be an external world beyond our perception, but this will have nothing to do with our actual experience which can only be `internal' and subjective. Such, according to the Yogacarin, is the nature of conscious experience. Attachment to the objects of experience (alambana) as if they were independent and external to the subject is the primary cause of the perpetuation of one's cognition of samsara. Ignorance and attachment (based upon past karman) thereby cause the bifurcation of consciousness into subjective and internal and objective and external. This is the `myth of the transcendent object'--that is the fallacious belief that one is having a veridical experience of an external world; the myth (maya) under which all unenlightened beings are labouring.
In the final analysis one can say that the main philosophical difference between Vasubandhu qua Sautrantika (as exemplified in the Abhidharma-kosa-bhasya) and Vasubandhu qua Yogacarin (as exemplified in many Mahayana works) is that in the former case the inference from experience to external cause (nimitta) is accepted, whilst in the latter this is seen to be logically unestablished (asiddha), a source of suffering (dukha) and delusion, and philosophically superfluous.
The Yogacara Path and the Abhidharma of Non-Mahayana Schools
As stated at the outset of this paper, the context of Mahayana philosophy in India was provided by the Abhidharma conceptual framework and modus operandi. This is not to say that the Mahayana had no critical responses to make to established Abhidharma theories, only that the context remained unquestionably Abhidharmic in both form and orientation .
Mahayana attitudes towards the Abhidharma of the Hmayana schools can be determined if we consider Mahayana conceptions of the Buddhist path (marga). The cultivation of the practice of yoga in the emerging Yogacara school leads at its highest levels to the cessation of the notions of `subject' and `object' (i.e. prapanca). The meditative path is hierarchical and progressive and each new stage involves the renunciation or cessation (nirodha) of what came before. From a Mahayana perspective this eventually came to mean that the Abhidharma analysis of the Hmayana schools remains appropriate for as long as one is proceeding along the mundane path (laukika) but that it should be relinquished once one enters the supramundane (lokottara) stages of the path. To rid oneself of the obscurations of knowledge (jneyavarana) as well as the obscurations of afflictions (klesavarana) that are overcome on the mundane (i.e. non-Mahayana) path, some of the techniques, doctrines and concepts of the Hinayana Abhidharma must eventually be relinquished. Equally, progress in yogic practice and the Mahayana path (marga) eventually leads to a cessation of notions of an `objective' dimension to experience since this type of realism is deemed inappropriate as soon as one realises `object-less concentrations' (niralambana samadhi). Progression along the path leads to an increasingly analytical scrutiny (prajna) of the images or representations (vijnapu) which occur in samadhi. This is the realisation of vijnaptimatrata in the meditative sphere--namely that that which is manifested (vijnapti) in perception is merely an image (akara) and not an independently-existing object. Finally, even these images (akara) must be relinquished as one realises the full import of sunyata and dharma-nairatyma. This final step amounts to the realisation that all dharmas are the same (sama), quiescent (santa) and indistinguishable from one another.
The notion of a progressively deconstructive path would seem to be the import of such verses as Trisvabhavanirdesa 36:
Through the perception that there is mind-only (citta-matra), there arises the non-perception of knowable things, through the non-perception of knowable things, there arise the non-perception of mind also .
This may prove to be a brief allusion to the progressive nature of the path and the gradual mastery and cessation of different levels of yogic attainment (and their concomitant conceptual frameworks). Note, for instance, that in the same way as the Madhyamaka interpretation of dependent-co-origination subverts the notion of a substantially originated entity, the insight into the fact that all perceptions are representations (vijnaptimatra/cittamatra) eventually subverts the notion of `representation' itself, since one realises that no truly external object can be presented to consciousness. Thus, Madhayanta-vibhaga I.6 states that,
Depending upon perception, there arises non-perception, and depending upon non-perception, there arises non-perception .
To which Vasubandhu explains,
Depending upon the apprehension that there are only cognitive-representations (vijnaptimatra), there arises the non-apprehension of things. Depending upon the non-apprehension of things, there arises the non-apprehension of cognitive-representation-only as well.
So even the notion of vijnaptimatrata is to be relinquished at the highest levels of attainment. This statement alone suggests that it would be extremely misleading to take either `vijnapti' or `vijnana' as designations of an ultimate reality in the early Yogacara literature (as enshrined in the doctrinal epithets `Vijnaptimatrata' and `Vijnanavada' which are often used by scholars as alternative nomenclatures for the Yogacara school) since both notions are relinquished in nirvikalpa-jnana. One cannot even rely upon the notion of alayavijnana for solace in this regard. In the Yogacarabhumisastra Asanga discusses the cessation of the alayavijnana . Again in Mahayanasamgraha I.61.3, Asanga declares that the alayavijnana is `like maya, like a mirage, like a dream and an optical illusion'; as such, it is `the seed for the imagination of the non-existent (abhutaparikalpa)' . As we have already noted Vasubandhu (the Yogacarin) also seems to have accepted that the alayavijnana is little more than a metaphorical concept (upacara) . This in itself follows on from the Sautrantika designation of the `seed' (bija) metaphor as a nominal existent (prajnapti-sat) . The Madhayanta-vibhaga also seems to argue that consciousness is something which is eventually relinquished. Thus, I.3 states that,
Consciousness arises with the appearance of objects, sentient beings, self and Cognitive-Representations. Nothing exists as its object, therefore that [[[object]]] being absent that [[[consciousness]]] too is non-existent .
To which Vasubandhu adds in his commentary,
As material form etc. [[[consciousness]]] appears as objects, and as the five sense-organs, it appears as sentient beings. These five senses refer to one's own as well as other consciousness-streams. The appearance as a self is the defiled mind (klistam manah), since it is associated with self-delusion. Cognitive-representations (vijnapti) appear as the sixfold consciousness. 'Nothing exists as its object' because the appearances of objects and sentient beings are without a fixed image; and because the appearances of self and cognitive-representations are false appearances. Thus, 'because of that [[[object]]] being absent, that [[[consciousness]]] also is non-existent'. That is the four kinds of graspables--form, etc., the five sense-organs, mind, and the sixfold consciousness, are absent. On account of the graspables being absent, the grasping consciousness also is non-existent .
Thus, upon closer analysis, there is much evidence which questions some of the central assumptions of a straightforwardly idealistic interpretation of Asanga and Vasubandhu. Certainly one cannot hope to cling onto the concept of consciousness-only (citta-matra) or representation-only (vijnapti-matra) as evidence of subjective idealism in the early Yogacara. In the highest states of attainment both the mind and cognitive-representations (vijnapti) are to be relinquished, 'external objects' being relinquished at a much earlier stage.
Perhaps by drawing attention to the continuities of thought between pre-Mahayana and early Mahayana Abhidharma thought in India one can gain a much greater appreciation of the philosophical context of early Yogacara thought. Clearly, interpreting early Yogacara through the eyes of later commentators can greatly enhance the comprehension of Indian Mahayana thought. At times, however, ignoring later controversies and focusing upon the philosophical and conceptual continuities to be found in the incipient stages of a particular school of thought can bring to light new insights in our understanding of that school. If we examine the early literature of the Yogacara the possibility emerges that long established interpretations of the school by Buddhist and non-Buddhist commentators and even subsequent developments within the school itself do not exhaust the hermeneutical options, ambiguities and alternate avenues that could have been followed as the school evolved. The early Yogacara literature provides many interesting avenues for future exploration and development and these have by no means been fully explored or exhausted by contemporary western scholarship on the subject. It may yet prove to be the case that many of the long cherished and well established interpretations of the Yogacara school (for instance that it expounds an uncompromisingly idealistic position) are founded upon more ambiguous philosophical beginnings than is generally acknowledged. By approaching the history of ideas from the rear, as it were, a different picture emerges of the early Yogacara position. It is a picture which suggests a great deal more in the way of philosophical continuity between Yogacara and mainstream Abhidharma thought than is often suggested and points to a doctrinal situation which is decidedly more complex than the stereotypical representation of citta-matra in the works of later Yogacarins and their opponents might at first suggest.
Richard King, Department of Religious Studies, University of Stirling, Stirling, FK9 4LA, Scotland, UK; e-mail
 BECHERT, H. (1982) `The importance of Asokas so-called schism edict in Indological and Buddhist Studies, Volume in honour of Professor J. W. de Jong, Canberra, pp. 61-68.
 KING, R. (1994) Early Yogacara and its relationship with the Madhyamaka school, Philosophy, East and West, 44(4), pp. 659-686.
 A recent example of the contextualisation of the Yogacara concept of alayavijnana in terms of its prevailing Abhidharma background is WALDRON, W. (1994, 1995) How Innovative is the Alayavijnana? The alayavijnana in the context of canonical and Abhidharma vijnana theory, journal of Indian Philosophy, 22, pp. 199-258 and 23, pp. 9-51.
 FRAUWALLNER, E. (1951) On the Date of the Buddhist Master of the Law Vasubandhu (eerie Orientale Roma m). See also FRAUWALLNER, E. (1961) Landmarks in the history of Indian logic in Wiener Zeitschrift fur die Kunde Sud-Ostasiens, Vol. v, p. 131.
 DOWLING, T. (1976) Vasubandhu on the Avijnapti-Rupa: a study in fifth-century Abhidharma Buddhism, unpublished Ph.D thesis, Columbia University, p. 52.
 Ibid., p. 73.
 Ibid., p. 85.
 asrayo hi sa aryanam darsanabhavanamargasamarthyat tatha paravrtto bhavati yatha na punas tatpraheyanam klesanam prarohasamartho bhavati ato `gnidagdhavrivad abijibhuta asrayah klesanam prahinklesa ityucyate. Ibid., p. 59, and esp. p. 94.
 darsana marga vyutthitasyavisesa darsana prahata vyaprahanat pratyagrasraya parivrtti nirmala santati vartate. Translation by Dowling, ibid., p. 106.
 HARVEY, PETER (1995) The Selfless Mind: personality, consciousness and Nirvana in early Buddhism (London, Curzon Press) pp. 155-179, 217-226.
 KEENAN, JOHN, P. (1982) Original purity and the focus of early Yogacara, journal of the International Association of Buddhist Studies, 5, pp. 7-18.
 SCHMITHAUSEN, L. (1987) Alayavijnana: on the origin and the early development of a central concept of Yogacara philosophy (Tokyo, International Institute of Buddhist Studies) pp. 7-8.
 GETHIN, RUPERT (1994), Bhavanga and rebirth according to the Abhidhamma, in: The Buddhist Forum, Vol. III (London, School of Oriental and African Studies) p. 35.
 DOWLING, T. op. cit., note 4, p. 64.
 SCHMITHAUSEN, op. cit., note 11, pp. 4-7.
 For details and full references see WALDRON, W. (1995), How Innovative is the Alayavijnana ? The alayavijnana in the context of canonical and Abhidharma vijnana theory, journal of Indian Philosophy, 23, pp. 38-39, notes 177 and 178.
 Glossed by Yasomitra as `duravabodha, 'difficult to understand'. See Abhidharma-Kosa-bhasyavyakhya IV.4.
 ANACKER, S. (1984) (Seven Works of Vasubhandhu, p. 159n) (Delhi: Motilal Banarsidass) argues that Vasubhandhu accepts the necessity of external stimuli in his Mahaydnasamgraha-bhasya (Peking/Tokyo ed. Tibetan canon, vol. 112, p. 275,4,3). Anacker also suggests that Vadavidhi 32 understands alaya to be little more than a metaphor for the idea of karmic seeds as a collective (ibid., p. 183n). In Mahayanasamgraha-bhasya (vol. 112. p. 277,5,1) 'seeds' are a metaphor for 'a special force within the consciousness series'. See also ANAKER, ibid., p. 71, where he argues that in the Pancaskandhaka-prakarana 'series' is nothing other than a metaphor for the genetic relationship between aggregate moments.
 See AUNG DAVIDS & RHYS-DAVIDS, trans., Points of Controversy, passim., Dowling, op. cit., note 4, pp. 68-70. MCGOVERN (1923) Manual of Buddhist Philosophy, I, pp. 128-129, translates avijnapti as 'not-manifested', and its opposite `vijnapti' as 'manifested'. Following this one might want to render the Yogacara term `vijnaptimatra' as 'manifestation-only' or 'appearance-only'.
 See VERDU, A. (1985) Early Buddhist Philosophy in the Light of the Four Noble Truths (Delhi, Motilal Banarsidass), pp. 35-37.
 mid., p.36.
 Ibid., p. 37.
 TAKAKUSU, J. Essentials of Buddhist Philosophy, p. 69.
 Note that Vasubhandhu places 'subject' before 'object' in the scheme of things in his Trimsika. Vasubhandu suggests that there are three types of vijnana-parinama, which occur upon the fruition of karman. They are: (i) vipaka--different fruition; (ii) manas--deliberating fruition; (iii) visaya-vijnapti--object-manifestation see Yamada Isshi (1977) Vijnaptimctrata of Vasubhandhu in Journal of Royal Asiatic Studies, pp. 162-163.
 quoted in Waldron, (1994), op. cit., note 2, p. 211.
 For a discussion of this Abhidharma context, see for instance KING, R. (1995) Early Advaita Vedanta and Buddhism: the Mahayana context of the Gaudapadiya-Karika (Albany, State University Of New York Press), pp. 108-118.
 citta-matropalambbena jneyarthanupalambhata, jneyarthanupalambbena syac cittanupalambhata.
 upalabdhim samastitya nopalabdhih prajayate, nopalabdhim samasrit a nopalabdhih prajayate.
 See OSAKI, A. (1977-1978) What is meant by destroying the alayavijnana?, Journal of Indian and Buddhist Studies, 26, pp. 1064-1069. In Mahayana-samgraha III.9, Asanga points put that one enters the perfected nature (parinispanna-svabhava) upon the complete cessation (nirakarana) of the notion of 'Cognitive-Representation-Only' vijnaptimtrasamjna), there being no object to be so represented.
 LAMOTTE, E. (1938) Mahayana-samgraha: la Somme du Grand Vehicule d'Asanga (Louvain, Bibliotheque du Museon), tome II, p. 85.
 See Anacker, (1984), ibid., p. 183n.
 Abbidharmakosa-vyakhya ad. II. 36c d. See Waldron, op. cit., note 2, p. 213 and p. 252, note 147.
 artha-sattvatma-vinapti-pratibhasam prajayate, vijnanam nasti casyarthas tad abhavat tad apy asat.
 For the basis of this view in the Pali canon see RAHULA, W. (1974), p.28. For the Abhidharma version see Basubandhu's Abbidharmakosa-bhasya 5.25.