1. Jayarāśi Bhaṭṭa (800–840),Footnote 1 as the author of The Lion [Destroying] the Delusion of CategoriesFootnote 2 (Tattvôpaplava-siṁha, TUS), the only surviving work considered to be loosely related to Indian tradition of materialism (cārvāka, lokāyata), with which he was formally associated, is widely accepted to represent a case of an Indian sceptic. In a recent paperFootnote 3, I examine the anumāna chapter of his magnum opus, and argue, on the basis of the nature Jayarāśi’s criticism, that we have no reasons to consider him a sceptic, but instead a (highly) critical materialist. In the same paper, I propose to call anumāna ‘debational inference’. Despite the widely held opinion that Jayarāśi’s views are exclusively negative, typical of a sceptic, there is some evidence to the contrary. In this paper, I wish to adduce some additional support to my thesis and collect a number of clearly distinguished positive views present in the Tattvôpaplava-siṁha, in particular in its anumāna section, to which my analysis is mostly confined. By ‘positive’ I mean views which assume an assertoric form of either ‘x exists’ or ‘x does not exist’, which is different from what can be assosciate with a sceptic parlance, namely ‘we have no reasons to accept that x exists’ or ‘we have no reasons to accept that x does not exist’, which does not entail the rejection of the main clause (‘… that x exists’ and ‘… that x does not exist’).

2. Rejection of universals. On a number of occasions Jayarāśi plainly rejects the existence of universals which he considers impossible to exist, a selection of examples is provided below:

*2a. ‘… because the universal is not explicable’Footnote 4, in the discussion with the Naiyāyikas;

*2b. ‘So such universal does not exist; and how it cannot exist has already been explained’,Footnote 5 in the discussion with the Mīmāṁsakas;

*2c. ‘… because the universal is impossible’,Footnote 6 in the discussion with the Buddhist idealists;

*2d. ‘If [one argues] that the scope [of inference] is the universal as a real thing, [we respond that] it is not the case, because the [universal] does not exist’,Footnote 7 in the discussion with the Buddhist idealists;

*2e. ‘[One may argue that] the scope [of inference] is the universal as an unreal (conceptual) thing, [because] it was said: “the universal is conceptual, with its form presented in the cognitive awareness”. But this also is not proper, because how could the non-existent universal become something which is the scope of the cognitive awareness?’,Footnote 8 in the discussion with the Buddhist idealists;

*2f. ‘If [it is aruged that the object of anumāna (‘debational inference’)] is non-different, then it would result in anumāna having the unique particular as its scope. One should not say that [in anumāna] a superimposed universal is cognised, [because] the undesired consequence would be that the universal would be something really existent’,Footnote 9 in the discussion with the Buddhist idealists;

*2g. ‘If [it is argued that] a combination of a unique particular and the universal can be made [in anumāna], that is not true, because a combination of the existent and the non-existent is not possible’,Footnote 10 in the discussion with the Buddhist idealists;

*2h. ‘It is not explicable that a universal which is void of all designation could be cognised’,Footnote 11 in the discussion with the Buddhist idealists;

*2i. ‘Suppose that the universal of “the cow” (go) and of “the gayal” (gavaya) is denoted by the term “similarity”, but that is inexplicable. It has alread been explained before why [the universal] is not possible’,Footnote 12 in the discussion with the Naiyāyikas.

2.1. As we can see, on several occassions (e.g. examples *2a, *2b, *2c, *2f, *2g, *2i) Jayarāśi takes the non-existence, or the impossibility of universals for granted, as something which does not require any further proof (or as something the refutal of which has already been adequately substantiated). His refusal to accept universals extends not only to universals as really existent things (vastu) on par with standard macroscopic objects such as trees, pots etc., but also applies to their existence as pure concepts (examples *2e and *2f). The reason for this is that, following the premiss of the correspondence theory of truth, even a concept purely existing in mind would have to have some kind of its denotatum, or objective substratum, that is, it would call for a really existent thing which exists either in the mind itself of outside, inasmuch as terms have to name, and concepts have to refer to, reals. In fact, in his rejection of the existence of universals he goes back to the refutal detailed earlier (TUS1 4.5–7.11 = TUS2 78–90, §1.13a2.a = TUS3 9–17) in a section to prove the thesis that ‘the universal cannot exist’ (na ca jāteḥ sambhavo ’sti). The section beging with stating possible alternatives of how the universal (jāti, sāmānya) could relate to the particular: ‘this [universal water] is either different from water individuals, or not different, or both different and non-different (TUS1 4.6: sā udaka-vyaktibhyo ’bhinnā bhinnā bhinnâbhinnā vêti). Neither in the initial thesis nor in the subsequent argumentation does he anywhere question the existence of individuals. In fact, throughout his work he consistently takes their existence for granted and, to my knowledge, there is not a single argument in the all his work to question or put to doubt the existence of individuals (vyakti), or particulars (viśeṣa). The assertion that universals do not exist occurs in the discussions both with those who accept universals (e.g. Nyāya, Mīmāṁsā) and with those who reject their existence (the Buddhists).

2.2. He further claims (example *2h) that universals are nor nameable, hence cannot be cognised and conceptualised. Being void of all designation means that an object cannot be meaningfully named, referred to and designated verbally in a way which does not lead do aporias or contradictions. The background for the unnameability of universals could theoretically be their non-existence in the first place, but then the argument would be circular:

… → universals are not nameable → universals are not cognised → universals do not exist → universals are not nameable → …

figure a

Such circularity wouldn’t have escaped Jayarāśi’s attention, and what is meant as an argumentative structure is a direct correspondence between nameability and cognizability as a premiss: all that can be cognised can also be named and vice versa, hence anything that cannot be named cannot be cognised either. Since universals can neither be named nor cognized (because their analysis leads to paradoxes and contradictions), therefore they cannot exist.

figure b

The thesis of the non-existence of universals is therefore derived from the premiss of their unnameability/unknowability. Interestingly, the underpinnings of Jayarāśi’s reasoning may resemble the Vaiśeṣika realist claim that all that exists is nameable and cognisable, which entails the claim that anything that is unnameable, is at the same time not cognisable and does not exist.Footnote 13

2.3. Most importantly, Jayarāśi does not counterbalance his rejection of the universal with a parallel refutal of the particular, as one would well be justified to expect a sceptic to do, namely to suspend his opinion with respect to both, or criticise both in the same measure. What may seem to be a passage which could possibly undermine my contention and provide such a sceptic’s counterbalancing argument against the existence of the particular carries quite different implications and, on a close reading, strenthens my thesis. The passage in question is a part of a longer discourse on the impossibility of universals:

[Just as water universal cannot be demonstrated to exist, so] can the plurality of water individuals not be demonstrated. Water [as such as well as a water individual] is differentiated from [anything else] which does not have the water form as something having the water form, but how is [such a water individual] differentiated from [another water individual] which [also] has the water form? Is it differentiated [from other water individuals] as something having the water form or by means of [having] some other form? If [on the one hand] it is differentiated [from other water individuals] as something having the water form, then [all] other water individuals would turn out to be something not having the water form, like sap etc. If [on the other hand] it is differentiated from [other] water individuals as something not having the water form, then an undesired consequence [would follow, namely that] it is not water, in the same way as fire etc. [is not water]. If it is objected: “Even though there is no difference [among water individuals] as regards their [universal] water form, [nonetheless] one should assume that there is] difference among [different] forms related to respective sub-classes [of water individuals] (intermediate water universals)”, [we say:] that is indeed true. [However,] is a form related to a respective sub-class (intermediate universal) [of a given water individual] established through its identity with [universal] water or established through its non-identity with (difference from) [universal] water? … [Consequently,] there is no other universal (intermediate universal) which could perform the function of differentiation [between numerous water individuals]. Or, if [such intermediate universals] are accepted, then the immediate consequence is the undesired dissolution [of the idea of the universal]. Therefore, [each and every water individual] is mutally differentiated from each and every [other water individual] by its own form (itself) alone, it is not differentiated by the universal etc., because of an undesired consequence that the water universal etc. would not be differentiated [from any other thing]. Hence the following is established, namely that the plurality of water individuals [as subsumed under one abstract class or one universal] cannot be demonstrated; since this [plurality] cannot be demonstrated, there is no universal which could be called the water universal, just as there is no [universal] “itself-ness”.Footnote 14

In the passage dealing with water as an instantiation of a general principle, what Jayarāśi rejects is not the plurality of individuals (vyaktīnāṁ nānātvam), which would indeed present such a counterbalancing act of a sceptic on par with his rejection of universals, but the idea that there are no univerals at all which could allow us to differentiate between particulars and conceptually group them into such a group. He takes recourse to the idea of multi-layered heierarchy of universals, a theory well-known from the Nyāya-Vaiśeṣika system, as laid down for instance by Praśastapāda.Footnote 15 At the ontological summit there is the primary, or main universal (para- / mahā-sāmānya), which is omnipresent (sarva-gata) and inheres in all that substantially exists (aneka-vr̥ttin). In practical terms, it is the universal existence (sattā), possessed by the first three ontological categories, namely by sub-stances (dravya), qualities (guṇa) and movements (karman), to be distinguished from existentiality (astitva) predicated of all that is there either substantially or eidetically, i.e. of all ontological categories.Footnote 16 It is the most general universal. Below it, there is a range of lower, secondary, or subordinate universals (apara-sāmānya), also known as intermediate universals (avāntara-sāmānya), distinguishable in layers. All the lower, subordinate universals have a dual character: that of the universal, i.e. repeatability (anuvr̥tti), or their repeatable instantiation in individuals, and that of the particular (viśeṣa) due to their uniqueness and exclusiveness (vyāvr̥tti), as something that is excluded from dissimilar entities. First come three main sub-universals (as the second layer of universals), viz. substantiality (dravyatva), qualitativeness (guṇatva) and mobility (karmatva), each of which is instantiated (anuvr̥tta) in all substances, all qualities and all movements, respectively, but also each of which is exluded (vyāvr̥tta) from the ranges of the two remaining universals, e.g. substantiality (dravyatva) is exluded from, or can never be instantiated in, any quality or movement. Below, as indicated by ‘etc.’,Footnote 17 there are further layers of sub-universals of the third layer. For instance, under substantiality (dravyatva) of the second layer of universals, there is a range of further nine sub-universals, such as earthness, or earth universal (pr̥thivītva), that are instantiated in, and embrace nine substances: earth (pr̥thivī), water (āp), fire (tejas), air (vāyu), aether (ākāśa), time (kāla), space (diś), soul (ātman) and mental organ (manas). Similarly, qualitativeness (guṇatva), as the second layer of universals, subsumes third-layer univerals of twenty four varities, such as colourness, or colour universal (rūpatva), tasteness, or taste universal (rasatva), etc., instantiated in twenty-four kinds of qualities. The third second-layer universal of mobility (karmatva) subsumes five sub-universals of the third level, such as the universal of upward motion (utkṣepaṇatva), etc. The third-layer universals differentiate between, say, different classes of substances, qualities and movements, and thus help distinguish earth (pr̥thivī) from water (āp) etc., colour (rāsa) from taste (rasa) etc., upward motion (utkṣepaṇa) from downward motion (avakṣepaṇa) etc. The next, fourth universal layer is characterised by innumerable universals, such as cowness (gotva), potness (ghaṭatva), clothness (paṭatva) etc., that inhere in actual individuals grouped in respective universal-based classes, such as cows (go), pots (ghaṭa), cloths (paṭa) etc. These help distinguish various kinds of objects that instatiate the substance of, say, earth, and help differentiate between cows, pots, cloths etc., or various kinds of colour patches that are instantiations of particular colours, e.g. to distinguish between red, yellow, green etc. Each and every layer of these universals involves the same spectrum of contradictions which Jayarāśi points out, which renders these universals impossible.

However, as Jayarāśi points out, one would further necessarily require a fifth layer of universals by dint of which one could differentiate between particular cows or, in his case, particular water individuals. And this is the crux of his argument, delineated in the above-quoted passage, against the notion of the universal. Specific singularly instantiated universal, called ‘itselfness’ (svatva), or the universal of itself (the individual object), would be required in order to differentiate it (e.g. a particular water individual) from all other similar objects belonging to the same class (e.g. water objects) conceived of on the basis of a fourth layer universal (e.g. waterness). Such a singular-instantiated universal, or the ‘itself-ness’ (svatva), is a reference to the expression ‘by its own form (itself) alone’ (svenâiva rūpeṇa), and would be an abstraction of a singular object’s own form. However, such universals of inifnite number corresponding to infinite singular entities, individuals, would involve the same logical problems and inconsistencies as all the previous layers of universals. Consequently, an attempt to differentiate between singular objects on the basis of their singular fifth layer universals would render the very existence of such individuals impossible. The tacit presupposition for the validity of Jayarāśi’s argument is clearly the actual, indefeatable existence of individuals, singular entities, each of which is simply there on its own, or ‘by its own form (itself) alone’ (svenâiva rūpeṇa), and not by dint of some abstract singular-instantiated universal (svatva). The conclusion therefore is not that the plurality of singular individuals is not there, but rather that if one relied on universals to distinguish between classes and individuals, the very notion of the individual, and the correlated notion of the plurality of the individuals, would collapse. Therefore, his argument has to be treated as a strong argument of the nominalist thesis: all that exist are individuals alone.

2.4. One may respond that Jayarāśi does counterbalance his refutation of universals elsewhere, namely in the critique of Buddhist nominalism, a section beginning with: ‘Who says the following, namely that the universal is not comprehended? On the contrary, it is indeed comprehended [in the following manner]: one recollects that “this [object x1 of class X] is similar to that [object x2 of the same class X]” because, on observing the second physical object [x2] and subsequent [objects x3xn that belong to the same class class X], one observes the recollection of the previous [object x1].’Footnote 18 Accordingly, we would have – in the person of Jayarāśi – a philosopher who, on some occasions, dismisses universals, but on other occasions, defends their existence. This whole argument in favour of the universals should be taken with caution though. As a closer examination of the whole argument reveals, Jayarāśi uses the above statement in favour of the existence of the universals only provisionally in order to demonstrate that Buddhist criteria for perception are not exhaustive and they allow space to subsume other entities, apparently non-existent but accepted as existent by Buddhist opponents (e.g. the Naiyāyikas), under perceptibility as defined by the Buddhists. In other words, the criteria for the non-perception of perceptible particulars (adr̥śyânupalabdhi), such as pots, in order to infer their non-existence in a particular place, under the condition that they exist elsewhere, allow also for an inclusion of universals. Accordingly, the Buddhist argument against the universal based on the idea of the meeting of the conditions of apprehension (upalabdhi-lakṣaṇa-prāpti) of a perceptible object is misapplied, and the conditions of perceptibility are defined improperly; if a macroscopic particular such as a pot is perceived, but not the universal, then another reason for (and definition of) the perceptibility should be sought,Footnote 19 and the Buddhist definition of perception and perceptibility has to be redefined.

Further, the above argument should be placed in its proper context. In fact, it is a part of a longer discourse on the universals as related to macroscopic wholes, which begins a little earlier with the Buddhist thesis against the existence of the universals that ‘the universal of the pot does not exist, because it is not grasped when [the pot] is not grasped’.Footnote 20 The Buddhist claim is that since macroscopic objects, such as pots (ghaṭa), do not exist, inasmuch as they are reducible to its parts that ultimately exist, also the universals that correspond to the macroscopic objects, such as ‘the universal of pot’ (ghaṭa-sāmānya), cannot exist, because they would have no substratum at all. In other words, the Buddhists maintain that the universals are as inexistent as macroscopic wholes. The purpose of Jayarāśi’s entire discourse is therefore to demonstrate that the Buddhist rejection of macroscopic wholes does not solve the problem of the existence of the universals, neither of which (the wholes and the universals) can meet the conditions of apprehension (upalabdhi-lakṣaṇa-prāpti) of an object, because the latter may for instance be related to the ultimately existent parts (colours, shapes etc.) into which the conceptual constructions (the wholes) such as pots can be analysed: a universal is believed by the Naiyāyikas to be perceived even when only one object is perceived in which the universal subsides and the object functions as the locus for the universal; if one can still perceive really existent colours, shapes etc. of the ultimately non-existent wholes, these parts can function as the locus for the universal equally well, and therefore the Buddhists cannot dismiss their existence.

Consequently, the discourse in which Jayarāśi allegedly argues in favour of universals has a much more complex structure: it is a discourse on universals as related to macroscopic wholes rejected by the Buddhists. And as such it does not provide any positive instance for the claim that, on occasions, Jayarāśi argues against universals, and in other cases, he argues in their favour, the way a genuine sceptic is expected to do.

3. Nominalism. Here, I clearly differentiate two positions in this account: Jayarāśi’s plain rejection of universals, which would normally be tantamount to nominalism, but not in the case when a possibility of scepticism is at stake, and his outright acceptance of the existence of individuals, or singular entities, i.e. nominalism per se. Jayarāśi’s nominalism to the effect that each and every singular entity exists ‘by its own form (itself) alone’ (svenâiva rūpeṇa), and not by dint of some abstract singular-instantiated universal (svatva), is reinstated by him at least on one more occasion, in a section being a criticism against the Buddhist (Diṅnāga-Dharmakīrti tradition) theory of inference, and causality-based inference (kāryânumāna) in particular. There, he says:

If [it is claimed that] body is not compatible with the possession of cognitions (lit. does not belong to the same class of objects that possess cognitions), [we respond that] fire also is not compatible with smoke (lit. does not belong to the same class of objects that possess smoke), [and there cannot be any causal connection between smoke and fire as a basis for causality-based inference]. All entities are differentiated in their essence from each other as having their forms produced by their [respective] specific causes and as having their individual essences existing in specific place and time (i.e. as existing within their individually specific spatial and temporal coordinates). And there is neither any mutual repeatability of form among them nor any repeatability of one universal [in them all]. This idea has been expressed [by Dharmakīrti himself]: “All entities, naturally (lit. through their individual essences) – inasmuch as they are established through their individual essences – are characterised by differentiation from [all things that have] similar essences and dissimilar essences; therefore…”Footnote 21

The nominalist emphasis in the above passage is not tentative, entwined in the sceptic’s argumentative structure in which it serves to refute one idea and waits to be refuted itself in the sequence. Jayarāśi seems to take it for granted that all that exist are only individual entites which are both (1) absolutely differentiated from each other, and there is no mutual repeatability of form among them (na ca teṣām anyonyâkārânugamo ’sti), and (2) cannot be subsumed under one heading of class on the basis of an abstract idea of the universal or class, viz. ‘there is no repeatability of one universal in them’ (nâpy eka-jāty-anugamaḥ). The first exposition of the nominalist claim, referred to above in § 2.3., occurs in a section criticising the school of Nyāya, and the second such exposition is mentioned in the critique of the Buddhists, two completely divergent, incompatible metaphysics. Jayarāśi therefore does not endorse the real existence of the particulars simply because that would well serve as a rhetorical context-dependent device with the goal to refute a universalist position, or because their tentative acceptance features as an logical consequence (prasaṅga) undesired by the opponent. The existence of the particulars is his own position.

Of note is that Jayarāśi quotes Dharmakīrti’s own famous verse from the Pramāṇa-vārttika (PV 3.40) that highlights the nominalist aspect of the reality as a groundwork for his theory of exclusion (apoha).

Interestingly, his strong nominalist position does not stand in contradiction with his initial ciritcism, expressed at the very outset of his treatise, of all categories (tattva), including materialist reals (tattva): ‘The reals (ontological categories) such as earth etc. are well-known among [all] people. When examined, [even] these [basic reals] cannot be established, so what about all other [categories]?’Footnote 22 His nominalist stance allows us rather to read his seeming rejection of all the reals such as earth etc. accepted by the standard Cārvākas differently: the category of tattva (real) presupposes the idea of universal (sāmānya) or class (jāti), the existence of which he denies. To say that the categories, or reals (tattva), such as earth etc. are dubitable entails neither a sceptical position vis-à-vis the existence of particulars nor a rejection of the materialist position but rather an acknowledgement that what exists are actual particulars which can be classified under conceptual classes of earth, water, fire and air on an arbitrary basis, as our daily use of the particulars may require.

4. Consciousness as a product of matter. A clearly materialist thesis Jayarāśi advances on at least three occasions is that consciousness is product of matter.

*4.1. ‘If [fire] leads to the origination of (lit. produces) [smoke] as [its] material cause, then how could it be possible that [fire] which belongs to a different category [than smoke] could be the material cause [of smoke]? Or, if one accepts that, then [one has also to accept] that consciousness will arise in the foetus only from the combination of the body and sense organs. So enough of this talk of consciousness beyond death (lit. in the other world).’Footnote 23

*4.2. ‘If there is no rise of consciousness without a consciousness which belongs to the same category, then how could there possibly be the rise – from fire – also of smoke, which does not belong to the same category [as fire]? If [it is argued that] in the case of fire there is the case of the same category [with smoke] thanks to the form of visual aspect, then also the body etc. belongs to the same category [as consciousness] thanks to having the form of unique particulars. With this in mind, it was said: “[consciousness arises] only from the body – so [says] Br̥haspati [the Cārvāka]”.’Footnote 24

*4.3. ‘And in the same manner, the heterogenous body etc. will cause the production of consciousness in the foetus etc. Therefore, the other world is not established.’Footnote 25

In all three cases Jayarāśi evokes the well-known materialist claim that consciousness is a product of a combination of four basic elements: earth, water, fire and earth none of which is conscious. This well-known thesis of the Cārvāka-Lokāyata tradition is mentioned on a number of occasions, for instance in The Disourse on the Fruits of Asceticism (Sāmañña-phala-sutta): ‘The person is made from four elements, … earth … water … fire … wind…’,Footnote 26 or in Haribhadra-sūri’s doxographic Compendium of the Six Systems (Ṣaḍ-darśana-samuccaya): ‘As a result of the combination of [four] elements such as earth etc. as well as due to the transformation of the body consciousness [arises] in the self (the body) in the same manner as the intoxicating power [arises] from the ingredients of a liquour.’Footnote 27 It features also in quotations directly ascribed to the CārvākasFootnote 28 and in what was the *Br̥haspati-sūtra / *Bārhaspatya-sūtra: ‘(1.2) Earth, water, fire and air are the reals. (1.3) The terms “body”, “sense” and “object” [refer] to their combination. (1.4) Consciousness [arises] from them.’Footnote 29 In this view, consciousness is thus a completely new feature that is not present in any of the ingredients, or elements (mahā-bhūta), but occurs when these are combined in right proportions and under adequate conditions, in a way similar to – as it is phrased in some accounts – the production of alcohol from ingredients, such as ferment starter etc. (kiṇvâdi), that do not have the intoxicating power,Footnote 30 or the occurrence of the red colour in paan.Footnote 31

Argument *4.1. might be interpreted as a sceptic’s hypothetical retort against the Buddhist but for its conclusion: ‘enough of this talk of consciousness beyond death’ (alaṁ para-loka-vijñāna-kalpanayā).Footnote 32 The whole passage occurs in Jayarāśi’s criticism of the Buddhist theory of causality-based inference (kāryânumāna), in the context of the relation between smoke and fire. Bringing the idea of ‘consciousness beyond death’ is simply out of context here, and it plays no role because the idea of life after death, or metempsychosis, is not the subject of the debate. It is merely an incidental, casual remark that there is no life after death, and the remark reinforces my interpretation that Jayarāśi’s response ‘[one has to accept] that consciousness will arise in the foetus only from the combination of the body and sense organs’ is not merely hypothetical, but it is his actual belief.

This conclusion is further bolstered by argument *4.2., which ends with a quotation from the Br̥haspati-sūtra / Bārhaspatya-sūtra: ‘“[consciousness arises] only from the body – so [says] Br̥haspati [the Cārvāka]”’, a fragment identified as sūtra 1.7.Footnote 33 Again, the citation of Br̥haspati’s sūtra is clearly not a hypothetical position evoked only as an uttara-pakṣa against the Buddhists, but a genuine position Jayarāśi acknowledges. As a rule, Jayarāśi does generally not quote his opponents, rather he paraphrases their thoughts, albeit there are some exceptions (e.g. AK 6.4, PV 3.40, MŚV 2.184, TSa 2895). The above reference to Br̥haspati would be one of such relatively few passages explicitely cited by Jayarāśi apparently in support of his own position, and all of them are related to the Cārvāka literature, primarily to Br̥haspati.

Similarly, in argument *4.3., Jayarāśi draws a straightforward conclusion, which is not a hypothetical one, expressed in the optative (e.g. prasiddhyet) from the claim that the body made of a combination of material elements produces consciousness in the foetus: ‘Therefore, the other world is not established’ (tataś ca na para-lokaḥ prasiddhyati).

To accept that consciousness is a product of material elements that concur in the physical body is tantamount to two co-related claims: first that on the disintegration of the physical body also consciousness has to disappear for ever, and therefore consciousness has to be recognised to be by nature of transitory and impermanent character, and second, that there is no soul (ātman, jīva), either as a non-material substratum of consciousness or as non-corporeal consciousness itself that may continue to exist independent of the physical body.

5. Rejection of the soul. Jayarāśi devotes a whole section (TUS1 74.11–83.7 = TUS3 149–162) to demonstrate that one can present no valid proof for the existence of the soul (ātman) in any established system (Nyāya-Vaiśeṣika, Jaina, Sāṁkhya-Yoga, Mīmāṁsā and Vedānta), which constitutes a section of the chapter of his refutation of inference. Repeatedly, he declares that ‘similarly, the inference of the soul from pleasure, pain, cognition etc. is not possible’,Footnote 34 ‘one cannot know that the soul exists, being a substratum of its effects such as pleasure etc.’,Footnote 35 or ‘accordingly, also the inference of soul does not work within the framework of the Mīmāṃsakas’,Footnote 36 etc. He eliminates a range of inferences – probably all he knew or considered exemplary – of the existence of the soul, such as for instance those based on the model of the soul functioning as a subtratum for or a seat of qualities (cognition, pleasure, pain, memory, agency, etc.), or the soul being in its essence cognitive faculties, etc. He demonstrates that all these inferences fail.

In another excursus (TUS1 55.8–58.7 = TUS2 262–272 = TUS3 116–121), he closely examines Buddhist inferences meant to establish a personal series (santāna), as a Buddhist equivalent of the soul in other systems. He reiterates that no valid proof of such a personal series can ever be formulated:

*5.1: ‘Also for the following reason a proof of a personal series is impossible, because consciousness is one, and its oneness [is established], because it is not possible that it could have another form [than its own], or if it had [another form beside its own], there would be an undesired consquence follow that it would lose its [own] form of consciousness. Since such an undesired consquence follows, there is no proof of a personal series’.Footnote 37

*5.2: ‘Also for the following reason a personal series is impossible: consciousness is distinguished from something of unreal composition as having real essence; however how can it be distinguished from something of another nature? … When it is deprived [of its own nature as consciousnes], then a series of cognitions is not possible. Since it is impossible, rituals such as veneration of reliquary mounds etc. are meaningless / useless’.Footnote 38

*5.3: ‘Also for the following reason a personal series of acts of consciousness is impossible, because [particular] consciousness x is not different in nature from other acts of consciousness which occur before, after or at the same time [with it]’.Footnote 39

*5.4: ‘Also for the following reason a proof of a personal series of acts of consciousness is impossible, because the cause-effect relation between two acts of consciousness is impossible, when they occur at the same time’.Footnote 40

*5.5: ‘And therefore there can be no personal series, nor is there any dychotomy of non-conceptual and conceptual cognitions, nor is any dychotomy of deviant and non-deviant (true) [cognitive acts] possible in the system of the Buddhists’.Footnote 41

In his analysis of the proofs of the soul and of the personal series, Jayarāśi largely follows the same method of reductio arguments (prasaṅga) to demonstrate that no such proof can be produced. How to interpret this? On the one hand, this could indicate the sceptic’s stance who merely witholds the final conclusion and suspends his judgement, without drawing the final conclusion to the effect that ‘therefore there is no soul or personal series’ (*tasmān nâsty ātmā santāno vā). Such a hypothesis finds support in numerous qualified statements that no valid proof is available, which are restricted to a particular case, e.g. ‘there is no proof of the personal series’ (na santāna-siddhiḥ), ‘one cannot know that the soul exists, being a substratum…’, ‘the inference of soul from … is not possible’, ‘the inference of soul does not work within the framework of…’, etc. On the other hand, Jyarāśi is occasionally more explicit and seems to rule out a possibility of a personal series and soul in general, instead of merely exluding, sceptic-like, a possibility of a proof, e.g. ‘a personal series is impossible’ (santānânupapattiḥ).

To say that ‘there is no proof of x’ allows for a sceptic interpretation, whereas to assert that ‘there is no x’ is unequivocal and compelling, unless we take it as a shorthand for the former (e.g.,*santāna-siddhy-anupapattiḥ). Since Jayarāśi does occasionally speaks of the impossibility of the personal series etc., the cumulative evidence would tilt the judgement towards the endorsement of the soul’s non-existence as his belief. I will return to this important issue in the conclusions (§17).

6. Rejection of afterlife. The rejection of the existence of a matter-independent and body-independent consciousness and of an eternal soul leads to another tenet which Jayarāśi apparently admits, namely that there is no afterlife, since there is no conscious substratum, or soul, that would transmigrate. This is what he expresses on a few occasions, for instance: ‘enough of this talk of consciousness beyond death’ (*4.1.) and ‘the other world is not established’ (*4.3).

In addition, there is an interlude inserted in Jayarāśi’s criticism of the Buddhist theory of perception, and the passage is not an immediate response to the Buddhists’ arguments and to their account of erroneous cognition (vyabhicāri-jñāna), which just precedes it, but rather presents an independent discussion which is not prompted directly by any thesis of the opponent, and is not related to the issue of erroneous cognition, which has in fact been definitely resolved.Footnote 42 Its initial portion reads as follows:

Besides, just as a visible form (perceptible data) is produced by [an earlier] visible form which is [its] material cause, in the same way also [its] cognition is produced precisely the same [visible form] which is [its] material cause. The nature of this [visible form] that [is operational] with respect to the production of [another subsequent] visible form is precisely the same as that which [is operational] with respect to the production of [its] cognition, inasmuch as this [visible form] does not have any other nature in order to produce [its] cognition.

If it is argued that cognition [of the visible form] is generated by [this visible form] as its efficient cause, and [another subsequent] visible form [is generated by the same visible form] as [the latter’s] material cause, then [the question arises]: how is it possible that one and the same thing involves multiple forms (i.e. has multiple natures)? For it is not possible that [one and the same thing, such as the visible form, becomes] something different such as the cognoscible (object of cognition), once its name is different, [i.e. once its name changes from “the material cause” to “the efficient cause”]. Also [in such case] – like [another subsequent] visible form [produced by the previous visible form] – [its] cognition will assume the nature of the visible form [and will cease to be cognition as distinguished from matter]. And when [cognition] assumes this [material visible form], then there can be no soul which [transmigrates] to the other world. {Conclusion *6.1} And if there is no such a soul [that transmigrates], then there cannot be the other world (afterlife). Having precisely this in mind, the Venerable Br̥haspati said the following: “Since there is no one [transmigrating] to the other world, there is no other world (afterlife)”.Footnote 43

[The above argument can be reverted.] If [one argues that cognition of the visible form] indeed [assumes] the nature of the visible form, even though it is produced by a material cause which is the visible form, then also the visible form can equally assume the form of [its] cognition, because it is produced by [its] material cause which is the [same] visible form [which produces its cognition], just like the cognition [produced by it]. If [it is argued that] cognition is produced by [an earlier] cognition which is [its] material cause, in the same way also the visible form [perceived by the cognition] is produced by precisely the same [cognition] which is [its] material cause, inasmuch as this [cognition] does not have any other nature in order to produce the visible form. And this idea [has been expressed by Dharmakīrti himself in the Pramāṇa-vārttika]: “Entities have this or that form, [since] they are produced by causes which have this or that form. How should this visible form etc.Footnote 44 be non-cognition, [being] produced by a cause which is the same [as the cause which produces its] cognition?” [{Conclusion *6.2} That being the case, the material visible form and cognition have the same nature.]’Footnote 45

The discussion of the causal relation between a material visible form (rūpa) and its cognition (jñāna) leads Jayarāśi to draw at least possible five following conclusions:

Conclusion *6.1: There is no such soul that transmigrates, ergo there can be no afterlife.

Conclusion *6.2: Cognition has the same nature as the material visible form, which is a conclusion expressed through the citation of Dharmakīrti’s verse.

Conclusion *6.3: ‘Cognition would have no nature at all’.Footnote 46

Conclusion *6.4: ‘Cognition has a nature consisting of a multitude of forms’.Footnote 47

Conclusion *6.5: ‘The combination [of the causes] of the cognition would not be possible’,Footnote 48ergo cognition would occur without being produced.

The latter three are not admissible at all on purely logical grounds, because cognition has to have some nature (at least that reflecting its object), cannot be multi-form (because it presents its own object, namely the visible form) and cannot be completely uncaused. What remains logically permissible are Conclusions *6.1 and *6.2, which would clearly be rejected by the Buddhist opponent only on (dogmatic) grounds other than logical inconsistencies they involve. Strikingly, they both point to the materialist claims that there is no soul and afterlife and that cognition has the same nature as its objects, namely it is of material nature. And both are endorsed by the Cārāvakas. Accordingly, the purport of the complex argument is that whichever way the Buddhist would attempt to explain the causal relation between a visible form and its cognition, he either ends up with logical aporias or with the materialist claim. The discussion can therefore provide additional evidence that both Conclusions *6.1 and *6.2 are acceptable to and endorsed by Jayarāśi. Especially the first of these conclusions is also supported by his reference to *Br̥haspati-sūtra / *Bārhaspatya-sūtra. Strikingly, nowhere in the discussion and in the possible conclusions does he develop arguments towards an undesired consequence (prasaṅga) to the effect that one would have to accept a soul or a permanent substratum of consciousness similar to the pudgala or ātman, which would be even more inadmissible to the Buddhists.

Still another excursus directly concerns the common belief that ‘the very first cognition [of the newborn] which occurs immediately after the exit from the mother’s womb is preceded by another [earlier] cognition, because it is a cognition, just like the second cognition [of the newborn is preceded by the previous cognition occurring immediately after the exit from the womb]’,Footnote 49 which requires that there has to be a series of cognitions that have existed prior to the birth and that go back to the previous rebirths. This argument is rejected by Jayarāśi who succinctly reiterates the afore-mentioned point (Conclusion *4.1.): ‘… enough of this talk of consciousness beyond death’.Footnote 50

All the above provides cumulative evidence that the impossibility of afterlife, which is not even questioned anywhere in the whole text of the Tattvôpaplava-siṁha, is another materialist thesis Jayarāśi himself endorses. This is also tantamount to Jayarāśi’s rejection of transmigration, or saṁsāra.

7. Rejection of karmic retribution, heavens and hells. An explicit denial of the moral link between one’s past deeds and one’s future life and of the fruition of past deeds (karman) is the very first thesis of Jayarāśi’s work: ‘“There is no fruit of these [past deeds], such as heaven etc.”. That is true indeed, because the accumulated deed (karman) [is not possible].’Footnote 51 Both ends of the first folio of the single preserved manuscript of the Tattvôpaplava-siṁha are damaged, so one may only conjecture the missing reading of sambhā{*vâbhāvāt(?)}, but there can hardly be any doubt that Jayarāśi emphatically agrees (satyaṃ tāvad) with the proposition that heavens and hells do not exist and there is no karman or its respecive fruition. Even if one treats the thesis ‘there is no fruit of these [past deeds], such as heaven etc.’ as a pūrva-pakṣa, there can be no doubt that Jayarāśi concedes it by saying ‘that is true indeed’.

8. Rejection of liberation. In Jayarāśi’s analysis, the issues of the soul (ātman) and liberation (mokṣa) are closely interrelated, and the question of liberation features also in his refutal of some kind of body-independent substratum of consciousness or soul discussed above (TUS1 74.11–83.7 = TUS3 149–162) as well as on a few other occasions.

For instance, while examining the Sāṁkhya view of the soul, he demonstrates that the ideal of liberation is without basis: ‘Since there can be no experience [of the results of one’s actions (bhoga)], also liberation is without any justification’,Footnote 52 as well as ‘And therefore, there follows the undesired consequence of non-liberation. Accordingly, liberation is not possible, because that which is experienced and that which experiences cannot be eliminated’.Footnote 53 He draws a similar conclusion in the case of Vedānta: ‘And therefore the efforts [undertaken] with the purpose of liberation are futile’.Footnote 54

Last but not least, the discussion of innately momentary character of the being, accepted by the Buddhists, and the related idea that there is no destruction per se, leads Jayarāśi to draw the following conclusion concerning the impossibility of liberation:

However, [since there cannot be any destruction of qualities], the undersired consequence of there being no liberation would not follow [for you], because the extinction of qualities would not be possible. This does not appear among contradictions for us (sc. does not occur among what we consider contradictions), because [we] do not entertain the belief that there is liberation.Footnote 55

This is an unequivocal rejection on Jayarāśi’s side of the ideal of liberation and afterlife, and cannot be treated as a rhetoric prasaṇga-like device, tentatively valid for the sake of the discussion, directed against someone who does not accept liberation (‘if you accept x, what would follow is that there is liberation, something which you do not admit’), because Jayarāśi directs his arguments against the Buddhists who do accept the ideal.

9. Rejection of ritualism, religious revelation and Dharma. Since Jayarāśi rejects the existence of a permanent body-independent, non-corporeal agent, or soul, karmic retribution, heavens and hells as well as the ideal of liberation, it would be quite natural to assume that he must have dismissed the purposefulness and meaningfulness of religious practice and rituals. In fact, we not only have hints to this effect but also find at least one direct case of evidence that he denounces religious ritualism: ‘Since it is impossible, rituals such as veneration of reliquary mounds etc. are meaningless / useless’.Footnote 56 This should not come as a suprise because the meanignlessness of rites and religious path is a logical corollary of a rejection of such concepts as the soul and karmic retribution and values such as heaven and liberation.

A whole section of Jayarāśi’s work is devoted to the rejection of testimony, either as teachings expressed by an authority or the authorless Veda (TUS1 115.2–119.26 = TUS1 213–220). I will not sumbit this particular section to a closer scrutiny here, but only present general conclusions drawn from this section. Jayarāśi does not accept the concept of the authority (āpta) in any form, either human or non-human. ‘The authorities’ are defined by him as ‘those whose nature is shaped by direct experience’.Footnote 57 And this rejection of the authority must apparently extend also to the validity of what is generally considered the contents of a religious authority’s teaching, including Dharma, the moral law, albeit Jayarāśi does not refute the concept itself directly (at least I have failed to notice any direct criticism of Dharma). But this would be a clear conclusion to draw from Jayarāśi’s critique of the authority: Since one could know Dharma (which is by definition beyond the purview of perception, inference and other cognitive criteria) only from the testimony either of those who have a direct experience of it or of an authorless account of it (the Veda), and since there is no valid authority whatsoever, there cannot be any source of knowledge of Dharma (see also §15 below).

10. Doxastic nexus. The above analysis reveals that an interrelated range of beliefs were entertained, or at least sanctioned, by Jayarāśi, all of which form a rather consistent model: (idea 1) consciousness as a product of matter (§4), → (idea 2) transitory character of consciousness (§4), → (idea 3) rejection of the soul (§5), → (idea 4) rejection of transmigration and saṁsāra (§6), → (idea 5) rejection of afterlife (§6), → (idea 6) rejection of heavens and hells (§7), → (idea 7) rejection of karmic retribution (§7), → (idea 8) rejection of meanigfulness or religious rites and acts (§8), → (idea 9) rejection of final liberation (§9).

All these beliefs, including the rejection of some kind of consciousness which would continue to exist beyond death, either as a personal series (santāna) accepted by the Buddhists or as some kind of the soul (ātman, jīva) adopted by most other systems, alongside corollary beliefs, strikes a familiar chord with a number of evidence related to the Cārvākas, including Br̥haspati’s sūtra that deny the possibility of a transmigrating agent and aferlife. It is well reproduced by Haribhadra the doxographer:

The Lokāyatas say: there is no soul, there is no final beatitude, righteousness and unrighteousness do not exist, there is no fruit of virtue and vice (good and bad deeds / karman).Footnote 58

All these ideas are preserved as early as in The Disourse on the Fruits of Asceticism (Sāmañña-phala-sutta):

‘Oh King, there is no [gain in] donations, no [gain in] sacrifice, no [gain in] ritual (idea 8). There is no result or fruition of good deeds or bad deeds (idea 7). There is no this world [as a place for karmic retribution] (idea 7), no the other world (afterlife) (idea 5). There is no mother and there is no father (i.e. there is no gain from the respect towards them) (idea 8), there are no spontaneously originated beings (i.e. divine and hellish beings) (idea 4, 6). There are no ascetics and brahmins who, having trodden the right path and having gained their access, have have fully experienced this world and the other world (afterlife) themselves and have seen it with their eyes directly and have born witness to this (ideas 5, 9).

This person is made of four elements (idea 1); when death comes, earth goes into or returns to the mass of earth, water goes into or returns to the mass of water, fire goes into or returns to the mass of fire, air goes into or returns to the mass of air (idea 2, 3). … This [talk of] alms giving [that bear results in the afterlife] is the teaching of fools (idea 8); empty and false blabber is their teaching that there something exists [after death] (idea 5). Both the ignorant and the wise decompose and disappear after the disintegration of their bodies (idea 2), [and] they no longer exist after death (idea 4).’Footnote 59

All the above beliefs which Jayarāśi seems to accept, or at least does not contest, provide evidence that he should be classified as a materialist, or a qualified materialist, not a sceptic.

11. Rejection of the supernatural and imperceptible. On a few occasions he refers to such entities such as demons (piśaca), atoms (paramâṇu) and god (mahêśvara). Does he mention them (1) because they are instantiations of merely sensorily imperceptible entities or (2) because they are absolutely beyond our cognition, like dharma according to the Mīmāṁsakas, and one cannot predicate anything of them, or (3) because one can argue for their existence with the same force as for their non-existence, or (4) simply because they are non-existent? Only the third interpretation would make Jayarāśi a sceptic, whereas a positive response to first two would be inconclusive (anaikāntika). To mention such entities in a prasaṅga-type of argument as an undesired consequence for the interlocutor is meaningful granted the opponent does accept the existence of such entities but may consider them either imperceptible or beyond our ordinary cognition. In another case, such a mention to an opponent who rejects their existence may be simply irrelevant and is an indication that the author himself considers such entities as absurd or inexistent as his opponent does. Below are the four instances I have found:

*11.1. ‘If [the sense-object contact] is known through a cognition of a pot etc. which arises from such a [contact], this is not correct, because when this [contact] is not known, then the fact that [this cognition] arises from this [contact] cannot be known [either]. This contact is [then] like something [of the sort of] demons, atoms and god’.Footnote 60

*11.2. ‘If [lucid character of perception (pratyakṣa-spaṣṭatā)] means that it is produced by the unique particular that is not cognised, how can this be known? It can neither be known through perception …, nor through inference [as such] … . Inference based on essential nature [cannot demonstrate it] …, inference based on causality cannot [demonstrate it] either, because an effect of something like this [unique particular which is not cognised] is not perceived. An effect produced by something similar to demons, atoms or god is never perceived in this world.’Footnote 61

*11.3. ‘[For the Buddhist idealist, a cognitive act] is immersed only in itself as consciousness alone, not being dependend on anything else [, i.e. on something objective, different from consciousness], because that which is different from it (i.e. something objectively existent) would be similar to demons and god’.Footnote 62

*11.4. ‘And [in the case of causality-based inference of fire from smoke, if the essence of a particular smoke] merged with the [universal] form of smoke, the knowledge of fire could not be inferred from the awareness (sc. perception) of this [particular smoke], because the relation between [a particular smoke and particular fire] is unknown [to us], like it is [equally unknown] to a resident of the Coconut Island [far away]. And the impossibility to know such a [relation follows], because the fire to be inferred is like a ghost or god.’Footnote 63

Instances *11.1. and *11.4. occur in discussions with the Naiyāyikas who do accept the existence of such entities, but would argue that these are imperceptible. Accordingly, Jayarāśi’s refence to them may only entail that they are mentioned as imperceptible without any implication that he himself rejects their existence. Instances *11.2. and *11.3. however occur in the discussion with the Buddhists of the Dignāga/Dharmakīrti tradition who would certainly reject the concept of god (mahêśvara), demons and atoms (see e.g. the Ālambana-parīkṣā). Accordingly to mention them points to their inexistent character rather than to their mere imperceptibility. His reference to the entities considered either supernatural or imperceptible may therefore provide some evindence that he did reject their existence and took their fictitious character for granted. Consequently, this could be interpreted as an indication that he apparently rejected invisible reality which is intrinsically beyond our senses.

12. Existence of macroscopic, non-momentary objects. What Jayarāśi takes for granted throughout his work is existence of composite wholes (avayavin) and macroscopic objects, and there are numerous illustrations to be found. For instance, a whole section (TUS1 §§ 9.6–16 94.2–106.15 = TUS3 179–199) is devoted to the refutation of Buddhist arguments against the existence of non-momentary macroscopic wholes (akṣaṇika avayavin), whereupon Jayarāśi proceeds to refute the idea of momentariness and critically analyses arguments against non-momentary wholes (TUS1 §§ 9.17–18 106.18–108.3 = TUS3 199–201). These two, namely the idea of indivisible spatial units, or parts (avayava), out of which an apparent macroscopic whole is composed, and the idea of unsplittable temporal units, or moments (kṣaṇa), which jointly contribute to an impression of a continuous entity, are closely related in his analysis, inasmuch as both are structurally identical. Since there is no single passage in Jayarāśi’s work which would refute the existence of non-momentary macroscopic wholes, we may assume that the section in question speaks in favour of the sole existence of momentary unique particulars (kṣaṇika svalakṣaṇa). If he were a sceptic refuting various possitions and phrasing new arguments adopted to a particular context, then one would expect him to formulate a refutation of such non-momentary macroscopic wholes in the context of his discussion, e.g., with the Nyāya-Vaiśeṣika, Mīmāṁsā or the Jainas (which all admit such wholes). But he does not seem to ever do it. Here are some selected instances of Jayarāśi’s detailed arguments and his positive views on the existence of macroscopic, non-momentary objects.

*12.1. ‘The grasping of the whole is not preceded by the grasping of all the parts, or the grasping of the whole is not preceded by the grasping of several parts. On the contrary, there is the apprehension [of the whole], when there is the completeness of [the conditions of perception] such as the contact of the [object’s] body and the sense organ, the light etc. For otherwise, in the case when there is the grasping [of the whole] accepted as preceded by the grasping of its parts, there would be no apprehension of the whole etc.’Footnote 64

*12.2. ‘It is argued [by the Buddhists] that “what is the [momentary cause] being of the nature to produce several effects which exist simultaneously [such as cognition, the blue, etc. (vijñāna-nīlâdikaṁ kāryaṁ)] is such a numerically one [momentary unique particular (ekaṁ nīla-svalakṣaṇaṁ)] which arises from its own causes”. If this is the case, then also my own [macroscopic whole (avayavin)] is produced from its own casues as having the nautre to produce several effects which exist consecutively.’Footnote 65

Conspicuously, nowhere else does he explicitly speak of an idea of his own (mama / mamâpi), and this can hardly be a rhetoric figure of a sceptic.

*12.3. ‘[It may be argued that] “If destruction is not possible, it follows that what has been produced is permanent”. [We respond:] the existence with a limit is impermanent, however the existence without a limit is permanent. So how could possibly something which has been produced be permanent, if there is no destruction [of it]? Or let us assume that it is permanent, [we find] no fault with it.’Footnote 66

He thus, in a sequence of steps, argues that there is no proof to demonstrate that composite, macroscopic, non-momentary wholes cannot exist and that what exists instead are their spatial and temporal parts alone, which is a typical Buddhist nominalist position. This alone could be treated as a sceptic’s arugment that it is merely not possible to prove such a position upheld by the Buddhists, not that the position is entirely unsound. However, what is significant is his statement: ‘let us assume that is it permanent, [we find] no fault with it’, which may be treated as an indication that this is precisely the position he may subscribe to, namely that non-momentary, composite wholes exist. Further, this stance – compounded with a total absence of any criticism on his part against a counterposition, namely that composite, macroscopic, non-momentary wholes do not exist – makes him not a sceptic who counterbalances contrary positions, but rather a thinker who clearly has a positive view.

12.2. Such a contention is further supported by the following. A powerful, well-known argument, adopted by Jayarāśi from Buddhist tradition and employed on a number of occasions, is based on the paradox of the indivisible whole and its parts. It is used by Jayarāśi for instance against the Buddhist concept of the particular visible form (rūpa) which triggers perception, in the following passage:

Accordingly, would the visible form produce the [perceptual] cognition [of it] (1) with one part of it or would it be complete as a cause [producing its perception] (2) with its entirety? (Ad 1) If this [visible form] produces [the perception of it] with one part of it, this is not correct, because it is not accepted that that which is indivisible can have [even] one part. (Ad 2) If [the visible form] produces [the perception of it] with its entirety, then the visible form is complete in being the cause of [its percepual] cognition [producing it] with its entirety, and does not proceed to produce another visible form [which replaces it in a momentary series]. Etc.Footnote 67

The above is just an instance of the whole–part argument. Structurally, precisely the same argument can be successfully used to refute various kinds of concepts: universals, atoms, wholes, causes, etc., i.e. all entities considered homogenous, unitary and indivisible. The structure of it is as follows, albeit it may have some variants:

Premiss 1: The whole unit, in order to be a whole unit, has to be homogenous, unitary and indivisible (at least conceptually).

Premiss 2: The whole unit performs its (ontological, epistemological) function as an indivisible whole.

Premiss 3: The whole unit is related to all its sections (parts, space divisions, time units, etc.) either (2a) with its entirety or (2b) with its parts.

If (2a), then the whole is complete in one of its parts in which it resides (ergo it is identical with it), and it cannot occur in any other parts of it at the same time → contradiction with Premiss 2.

If (2b), then the whole unit has parts → contradiction with Premiss 1.

Conclusion 3: Therefore, the whole unit, being self-contradictory, cannot exist.

This argument, apparently of Buddhist origin, is recorded as early as in the Nyāya-sūtra:

[7]. No whole can exist, because [each of its] parts cannot occur in the entirety [of the whole] or in a part [of the whole]. [8] No whole can exist, because it cannot occur [in its parts with its entirety or with its part]. [9] And [no whole can exist,] because it cannot occur as separate from its parts. [10] And the whole is not the same as its parts.Footnote 68

It was famously employed by Vasubandhu in his Twenty Verses being The Proof of the Sole Existence of the Contents of Consciousness (Viṁśatikā – Vijñāpti-matratā-siddhi):

[11] Such an [objective basis for consciousness] – either as homogeneous one or as a heterogenous complex – cannot become the contents [of perception] due to atoms [that are said to constitute it], and also these atoms accumulated collectively [cannot become] the contents of perception], because the indivisible atom cannot be proved [for the following reasons:] [12] (a) due to [its] simultaneous connection with the sextet [of the cardinal directions], the indivisible atom would have six parts; (b) because six [indivisible atoms] would occupy the same space unit, the physical object (sc. an aggregate composed of the atoms) would have a dimension of [one] indivisible atom, [ergo would be as invisible as the atom]; [13] (c) if their accumulation [in space] is not a conjunction of indivisible atoms, then what is this [conjuction] of? And it is not the case that – because [ indivisible atoms] have no parts – their conjuction cannot be proved. [14ab] It cannot be consistently assumed that that in which a division [into sections which correspond to] spatial sections of cardinal directions is not possible, is a homogenous one.Footnote 69

Jayarāśi successfully applies the above argumentative structure to refuting a number of concepts, such as the universals (TUS1 4.5–7.11), a (momentary) cause which is expected to produce both its perceptual cognition and a subsequent suchlike entity in the next moment (TUS1 46.7–14), inherence (TUS1 75.1–7), etc. Obviously, exactly the same argumentative structure could successfully be also employed against external, macroscopic, non-momentary wholes, for instance: the macroscopic, non-momentary whole – being an objective basis for a conceptual unit ‘the whole’, which is something beyond and above its parts and which cannot be reduced to a mere assambledge of the parts) – resides (is related to) its parts either (2a) through its entirety or (2b) through its parts. If (2a), then it is reduced to one of its parts only. If (2b), then there is no whole at all, viz. there is nothing beyond and above the particulars thought of as ‘parts’. Such an argument could have a number of variants applicable against, for instance, the concepts of the aggregate whole (avayavin) developed, e.g., by the Nyāya-Vaiśeṣika, the Jainism, Sāṁkhya-Yoga or Mīmāṁsā, as the criterion of objectivity and veracity of perception and other cognitive criteria (pramāṇa). These two concepts – universals and wholes – were analysed jointly in India as two sides of the same coin: just as the whole exists through / inheres in its parts, in the very same way also the universal exists through / inheres in its particulars, etc.

Such a possibility to apply the same argumentative structure against the macroscopic wholes couldn’t have escaped his brilliant philosopher’s acumen. Strikingly, however, Jayarāśi never employs this argument against macroscopic, non-momentary object, something one should expect him of, granted that he was a sceptic. And there must have been a reason for that. He never links these two issues, precisely because, and this seems to be the only explanation, he did admit the existence of external, macroscopic, non-momentary objects of our experience, i.e. aggregate, composite wholes, as genuine objects of our experience and daily dealings, whereas he rejected the existence of universals.

Interestingly, I have failed to find an instance when Jayarāśi employs the whole–part argument against atoms as well, which he seems to reject on other occasions.

12.3. In fact, in a longer passage (TUS1 95.20–100.21 = TUS3 182–190) he refutes a number arguments formulated by Buddhist idealists against the existence of macroscopic, non-momentary wholes. He calls these arguments ‘arguments meant to refute the whole’ (TUS1 98.16: avayavi-nirākaraṇa-paraṁ sādhanam) or ‘meant to deny the external object’ (TUS1 100.23: bāhyârthâpahnave). All these arguments are based on the idea that wholes cannot exist because they could have various properties at the same time, some of which would be contradictory. The line of arguments begins with his own contention, being in itself a conclusion of his refutation of Buddhist non-apprehension: ‘In the same way [as the non-apprehension of the whole], also the whole – even though it does not produce its effect – will not abandon its intrinsic nature, because it arises from nothing but its own cause, the nature of which is a non-producer of effects.’Footnote 70 First, he examines an argument that the whole cannot exist because it cannot be multicoloured, i.e. both coloured and non-coloured at the same time (raktârakta-prasaṅga). This section concludes with: ‘If the fact that a blue substance and a blue cloth are coloured means that they originated in a place directly adjacent [to the colour], even then this does not prove that [such wholes as a blue substance and a blue cloth] do not exist, because it is accepted that compounded [entities] do exist.’Footnote 71

The next argument formulated by the Buddhist idealists against the whole is based on the idea that it can be both covered and not covered at the same time (āvr̥ttânāvr̥tta-prasaṅga), and therefore is self-contradictory: ‘it is covered, since one part of it is covered, and it is not covered, since [another part of it] is not covered’ (ekasminn avayave āvr̥tte āvr̥ttaḥ, anāvr̥tte anāvr̥ttaḥ). This is likewise rejected by Jayarāśi who concludes: ‘this does not lead to any differentation in [the whole’s] instrinsic nature’ (na ca svarūpa-bhedam āsādayati).

He proceeds to analyse another argument against the wholes based on simultaneous mobility-immobility of the object (calâcalatva-sādhana), and concludes that ‘one does not observe any differentation in [the whole] when it is moving [in one part] and not moving [in another part]’ (tathâvayavi-calâcalatvena na bhedo dr̥śyate), such as for instance an immobile person moving his/her hand.

Similarly, he rejects other arguments against the existence of the whole based on the fact that one part of it is perceived and another part is not, and that wholes cannot exist because they are no longer perceived once they disintegrate or are conceptually analysed into its parts, quoting here as the pūrva-pakṣa a (metrically corrupted) verse which in its correct form is either a distorted version of a verse from the Abhidharma-kośa or goes back to an unknown common source:

When something is fragmented, there is no [longer] a cognition of it. And there is no [longer a cognition of it when it is fragmented] by means of the mind. However, the conventionally real is something [in the form of] “this pot”. The ultimately real is [that which exists] in a different manner, [viz. is neither spatially nor conceptually analysable].Footnote 72

According to this important Abhidharmic principle phrased by Vasubandhu, anything which does not withstand the ordeal of analysis, either physical fragmentation into parts and pieces or mental reduction to more fundamental conceptual components, does not ultimately exist and can merely be considered an empirically true, or an ultimate fiction. This analysis was employed by the whole Abhidharmic tradition, including Vasubandhu, to an analysis of the whole, such as the famous dialogue between Monk Nāgasena and King Milinda on the identity of the person (Mil 26: ‘What is then this Nāgasena? – Is the hair Nāgasena? nails, … teeth, … skin, … flesh, … sinews, … bones, etc.’) or the charriot (Mil 27: ‘What is then this charriot? – Is the pole the charriot? the axle, … wheels, … the framework, … the ropes, … the yoke, etc.’). Abhidharmic analysis leads to the rejection of the essence (svabhāva) of the compounded objects such as macroscopic wholes or persons (anātman, pudgala-nairātmya), and instead postulates the ultimate existence in the form of indivisible, insoluble entities (dharma). Jayarāśi examines precisely this Abhidharmic critique of the wholes and demonstrates that is without basis in the final resort.

Jayarāśi’s all above refutations of arguments formulated against the macroscopic, non-momentary wholes existent independently of the mind (avayavi-nirākaraṇa, bāhyârthāpahnava) should not be treated as a mere rhetoric strategy of a sceptic who is keen to demonstrate that no solid argument can be formulated in favour of a claim P because an equally strong rejoinder against it can be phrased, or an alternative argument backing claim non-P. He means what he says: external, macroscopic, non-momentary objects do exist objectively.

13. Direct experience as the ultimate resort. On a number of occasions, we can notice clear indications that allow us to draw an approximate picture of Jayarāśi’s positive views on epistemology. As it is well known, he found the standard system of epistemology based on cognitive criteria (pramāṇa) as problematic, pointing out that without properly defining cognitive criteria one cannot effectively use them:

The establishment (validity) of cognitive criteria (pramāṇa) depends on [their] proper definition, whereas the establishment of the cognoscibles depends [in turn] on the cognitive criteria. When this [definition of the cognitive criteria] is absent, how can one [admit] … that the other two, [sc. cognitive criteria and cognoscibles], become the objects of real everyday practice / of a discussion on what exists? If a discussion is held [on these two, namely cognitive criteria and cognoscibles], even though such a [definition of cognitive criteria] is not established, then one could [equally] engage in a discussion on the existence of a colour in the soul or in a discussion on the existence of pleasure in a pot.Footnote 73

However, the existing definitions of cognitive criteria are flawed, and to demonstrate this faultiness and inadequacy is the main purpose of his work. This seems the main reason for modern researchers to interpret Jayarāśi as a sceptic: since no proper and consistent definition of a cognitive criterion can be offered, therefore one has no reason to accept any cognitively valid procedure as an adequate source of knowledge about the world, and consequently no categories and reals (tattva) can be established. Does it mean that, according to Jayarāśi, we as philosophers are left empty handed? It seems that afterall the main purpose of philosophising for Jayarāśi, as well as for Br̥haspati, is that of reflecting the world (TUS1 1.12 = TUS2 68 = TUS3 1: pratibimbanârtham), albeit both pursue their task in different ways. The concluding verse of the Tattvôpaplava-siṁha famously states:

Thus, when all categories/reals are completely dissolved in the above manner, all practical actions (which entails thought, speech and bodily activity) can be enjoyed as something which does not [have to be] reflected upon.Footnote 74

Would it support Franco’s (1987, p. 44) contention that ‘The wisdom in this case would consist in the understanding that all principles are annihilated. When this stage is reached (by the examination of the definitions of the means of knowledge), the worldly path is not only something which has to be followed by lack of choice (as in the above quotation from Sextus), but it becomes something delightful’? This is something I would contend by demonstrating that Jayarāśi did accept also some cognitive procedures as means to know the world. Below, I provide some instances.

*13.1. ‘It is established that the existence of a [real thing] is [proved] through the apprehension of the real thing alone, and [in such a case] its non-existence is without any counterproof’.Footnote 75

It is the apprehension of an object which establishes that the object is there, and it neither requires any additional substantiation nor can be negated by any means, due to its obvious character.

*13.2. On a number of other occasions, Jayarāśi resorts to a direct observation of facts as the final adjudication, such as in the discussion on the validity and correctness of word forms:

Suppose the following argument concerning the validity of the sacred and eternal language of the Vedas:] “Even when there is no definition of [the Sanskrit word of the Vedas], it is the fact that they are proper.” Well, that being the case, also Apabhraṁśa words, such as gāvī, goṇī, goputtalikā, [all denoting “cow”], will turn out to be [equally] proper, despite the absence of definition. If, due to the absence of their definition, [the Apabhraṁśa words such as] gāvī etc. are not taken as proper, then also the [Sanskrit] words of the sūtras would become improper, due the absence of their definition]. Moreover, if the word is devoid of definition, what happens then? Will [that lead to] the disfiguration of the pronouncer’s face, or would the word have no meaning? … Suppose, to begin with, [that it leads to] the disfiguration of the pronouncer’s face, when one pronounces [such Apabhraṁśa] words as gāvī etc. [Well,] such pronouncers are observed who frequently pronounce the word gāvī, but no disfiguration of the pronouncer’s face is ever observed.’Footnote 76

*13.3. ‘It is not proper to reject what one directly experiences, because the undesired consequence would follow that one could deny everything’.Footnote 77

Direct experience brings obvious, irrefutable knowledge, because otherwise nothing could be trusted and everything could be denied. And this is not what one would expect to hear from a sceptic, namely that the denial of and mistrust in everything might present an undesired consequence. It does not seem likely that Jayarāśi would accept the conclusion that one could deny everything (sarvâpalāpa). As we can see throughout his work, Jayarāśi occasionally does refer to direct experience (anubhava), or to ‘the apprehension of the real thing alone’ (vastu-mātrôpalabdhi), which are apparently treated by him as the final instance to adjucate the veracity of a claim.

A pertinent question therefore arises how direct experience relates to perception (pratyakṣa), which is the first and foremost of the cognitive criteria (pramāṇa), the validity of which he rejects. And another one, what exactly is ‘direct experience’ (anubhava) and how is it defined? It does not come as a surprise that Jayarāśi nowhere attempts to provide a definition (lakṣaṇa) of it, which would allow him to establish and define what ultimately exists, following his own method laid down at the outset of his work,Footnote 78 as referred to above (§13):

proper definition (lakṣaṇa) → cognitive criteria (pramāṇa) → the cognoscibles (prameya) / categories, reals (tattva).

To deliberately avoid defining what counts as direct experience may indeed position him as a sceptic to some extent, but perhaps this may allow for another rational, non-sceptical reading of it (see §15 below). He also seems to avoid any direct equation of direct experience and perception, though, albeit the context would suggest that direct experience, as a cognition of what is presently there in front of us and derived primarily from sense organs, is very much like what most other philosophers would term ‘perception’. His reluctance to use the term ‘perception’ (pratyakṣa) may be due to the fact that once we apply this technical term well established in Indian epistemological tradition, it comes ‘with the benefit of the inventory’ of the pramāṇa model, which he does not subscribe to.Footnote 79 In other words, pratyakṣa brings in the problem of a proper and complete definition of it as a pramāṇa, whereas anubhava does not, because it remains outside of the pramāṇa model.

*13.4. He further asserts that direct experience (anubhava) and memory (smr̥ti) exclude each other:

How can possibly the [memory], produced by [a cognitive act] in the form “there is water here”, assume the form “I saw [water there]”? … As regards these [two alternatives], if [the memory of water] means the appropriation of the [original] direct experience, then it is not possible that [a cognitive act assumes] a form of the memory [of water]. If [it does assume] the form of the memory [of water], then the appropriation of the [original] direct experience is not possible, because [cognitive acts] in the form of memory and direct experience are defined as positioned in mutual exclusion.Footnote 80

A conclusion to draw would be that memory cannot play the same role of a reliable source of knowledge as direct experience does. And nowhere in his work does he obviously resort to memory as a final instance. In addition, this passage supports my contention that he accepts direct experience, opposed to memory, as a genuine source of knowledge of the world.

14. The role of recognition. In his refutation of the Buddhist concept of momentariness, Jayarāśi seems to admit, in rather unusual terms, still another means of knowing the reality, in addition to direct experience (anubhava), or perception (pratyakṣa), namely recognition (pratyabhijñāna):

The [non-momentary] existence of [numerically] one [thing] – i.e. its being the agent [producing] all subsequent effects – of which one becomes aware by means of experience arisen previously, is internalised (sc. apprehended) [due to joint operation] through perception and by means of recognition. The very same existence [of the numerically one thing] is [later] internalised (sc. apprehended) by means of recognition. If [the Buddhist asks me:] “How do you know?”, [I reply:] we know it through nothing but recognition [of the form:] the very same existence [of the numerically one thing now] occurs. Also in [our] very first experience [of the thing] this existence occurs. [Therefore it is the same thing].Footnote 81

This is an extremely important passage which reveals that Jayarāśi did hold some positive beliefs also concerning epistemology. He clearly places himself directly within the debate with the Buddhist idealist in a twofold linguistic manner: by making the latter ask him personally a direct question in the second person singular: ‘How do you know?’ (kathaṁ punar vetsi), and by responding in the first person plural: ‘We know it through nothing but recognition’ (pratyabhijñānād eva janīmaḥ). Whether we take the plural in a general way to refer to the common experience of all people (‘we, the people…’), or to be a kind of pluralis majestatis traditionally and frequently used in texts with reference to the singular author of the text, in either way Jararāśi, who speaks of his own personal experience as a source of knowledge, is included in the collective. With this contention ‘we know it through nothing but recognition’, his argument ends: he no longer presents any prasaṅga-type of justfication which would point to a range of further undesired consequences.

Recollection is a corelation of two acts of direct experience (anubhava), an original one, and a subsequent one, which leads to the conclusion that we experience one and the same non-momentary object, and not a series of momentary events. Also in this case we can observe that, for Jayarāśi, the reference to direct experience (anubhava) triggered by perception (pratyakṣa) and recollection (pratyabhijñāna) is the highest court of appeal. Ath the same time, to accept direct experience (perception) and recognition should not be taken as tantamout to Jayarāśi’s recognising both/either of them as cognitive criteria (pramāṇa), the very idea of which is perforce the object of Jayarāśi’s criticism.

15. Defective character of senses. Let us return to the crucial question why Jayarāśi does not engage in any kind of philosophical system building, for instance, by providing a definition of what he vaguely calls ‘direct experience’ (anubhava), instead of using a well established technical term ‘perception’ (pratyakṣa). A part of the answer can be found in the following two examples, in which Jayarāśi refutes also the cognitive reliability of testimony of an authority (āpta).

*15.1. ‘Furthermore, since sense organs are the substrata of qualities (correct functioning) and defects, one cannot get over the doubt concerning the defective character of such a cognition which arises through these [defective sense organs], just like with respect to cognition [derived from] words generated by an activity of a man.’Footnote 82

*15.2. ‘Suppose the following: “Since the man is the seat of defects (sc. is liable to commit mistakes), when [something] is produced by him, one suspects that also [that thing] is produced by defects [related to the man].” [And this is why ordinary texts, which have authors, are liable to suspicion, whereas the Veda is not, being authorless]. [We respond:] That being the case, then since also sense organs happen to be the seats of defects, one can entertain a doubt that acts of cognition arisen through them have no cognitive validity (aprāmāṇya). And then one could not rely on cognitive validity (prāmāṇya) of anything at all.’Footnote 83

On both occasions, Jayarāśi rejects an authoritative character of verbal testimony, but what is most important here is that he points to the same problem with both verbal testimony and cognition derived from sense organs, or perception, which is the potentially defective character by default.

It is this defective nature of sense organs that apperently prevents him from admitting perception, or direct experience, as a reliable warrant, or a cognitive criterion (pramāṇa), whereas all definitions of perception presented by other schools and analysed by him (and known to us) presuppose their reliability. There is no single definition in the stock of definitions collected by him which would seriously take this deficiency of sense organs into account. On the contrary, all definitions of pramāṇas pressupose full reliability of respective cognitive criteria (not even the Mīmāṁsā definition). In fact, and this is what he discusses at the outset of his work, in a section devoted to the Nyāya definition of perception, such reliability, or non-deviance / non-erroneousness (avyabhicāritva) of perception (and other cognitive criteria) may be due to the fact that it is produced by a complex of non-defective causes, or because it is free from sublation, i.e. falsification in the future, or on account of efficacy of subsequent activity based on it, or otherwise.Footnote 84 None of these criteria can be met in practice, even in the case of perception, and the requirement that a cognitive criterion is expected to be 100 per cent error- or defect-proof cannot be ever achieved. This may therefore be one of the reasons why Jayarāśi dismisses the pramāṇa model of epistemology as such. At the same time, as he notices, any cognitive subject has to rely on direct experience (anubhava), which is the only direct source of knowledge that remains at one’s disposal and on which all other cognitive criteria rest (barring perhaps verbal testimony). Despite this, the conclusion he draws is not necessarily that of the sceptic, namely the rejection of all knowledge claims and distrust of the veracity of all kinds sources of knowledge. Instead, he decides to chose what he calls ‘the worldly path’, or the truth of everyday direct experience, with all its limitations taken into account:

The worldly path (laukiko mārgaḥ) should be followed… / With respect to everyday practice of the world (loka-vyavahāra), both the ignorant and the wise are similar.Footnote 85

Incidentally, the second hemistich may echo an early materialist idea preserved in the Sāmañña-phala-sutta: ‘Both the ignorant and the wise decompose and disappear after the disintegration of their bodies’.Footnote 86

There may be one more important reason why Jayarāśi declines to subscribe to the pramāṇa model, namely that the model is generally expected to warrant complete reliability of what we take for cognitively valid procedures. This is the other side of the same coin: On the one hand, being defective, sense organs occasionally and not consistently delude us, therefefore we can not always rely on them. On the other hand, a pramāṇa system is expected to be complete in two senses: in the sense of providing accurate definitions that successfully eliminate all defective cognitive procedures and in the sense that whichever act of cognition is derived from its sources, following the definitions, is always true with no exception. And this cannot be achieved, according to Jayarāśi. We could call such a situation metaphorically ‘Jayarāśi’s incompleteness theorem’.Footnote 87 First, no complete, exhaustive definition of a cognitive criterion can be formulated so that any cognitve procedure that is covered by it is indeed reliable and uniformly yields truth. Second, within any definition of a given cognitive criterion, we can find cases that a cognitive procedure meets the terms of the definition but at the same time its result is doubtful, i.e. it cannot be decided whether it is true of false. A pramāṇa model understood as a system of correct definitions that warrant reliability of cognitive procedures based on them is a fiction.

His solution to the problem is what we could term a compromised non-erroneousness requirement, or limited trust in the reliability of knowledge derived from senses, which occasionally happen to be defective. We do not expect absolutely all our sense-derived cognitions to be true and reliable, but generally we trust and rely on most of them. This is what we do on everyday basis, and we cannot even hope for any other epistemic instruments which are always non-defective and doubt-tight. All our direct experiences taken jointly present a system of checks and balances that guarantee limited reliability, if not perfect non-erroneousness, of cognitions derived from senses. This compromised non-erroneousness requirement is an effect of pratical exigency for ‘both the ignorant and the wise’, whose sense organs can be defective to the same degree.

This problem concerns not only perceptions (pratyakṣa) but ‘debational inferences’ (anumāna) also verbal communication. Jayarāśi draws our attention to the impossibility to provide a complete definition of words (śabda), not only in the context of religious (Vedic) revelation but also with respect to daily communication. He concludes that ‘words so defined do not exist’,Footnote 88 but we still manage to communicate, and we do it through defective verbal means.

The problem therefore does not necessarily lie with the nature of our direct experience (anubhava) or words (śabda) we use to communicate meanings but rather with our expectation to arrive at a complete model of definitions. Definitions present ideal situations, in a sense they refer to universals (sāmānya), but what ultimately exist are the particulars (viśeṣa), not covered by definitions.

There remains an issue to decide but to which I find no answer. In examples *15.1. and *15.2. above, Jayarāśi points to the defective character of both senses accepted as reliable within the pramāṇa system and speakers likewise believed to be absolutely reliable (āpta). He seems to accept the limited reliability of direct experience (anubhava), but would that mean that one should therefore accept testimony (āptôkta) as likewise reliable, but on a limited scale? Perhaps. After all he himself trusted that what Dharmakīrti had expressed in writing was precisely what he had genuinely meant, and that when people see the sun in the sky, the sun is usually indeed there.

16. Rejection of omniscience. From sources external to Jayarāśi’s own work, we may infer that he also explicitly rejected the belief in omniscience and omniscient beings. We know this from Jaina sources, and we have no reason to to dismiss their authenticity in this case.Footnote 89 In his refutation of the Laukāyatikas (AṣS1 35.2–42.116 = AṣS2 29.20–36.6) found in the Commentary in Eight Thousand Lines (Aṣṭa-sahasrī), Vidyānanda (AṣS1 37.9 ff. = AṣS2 31.2 ff.) draws attention to a subgroup of the Laukāyatikas or Cārvākas (he uses these two terms interchangeably), namely to ‘those who propound the dissolution of [all] categories’ (tattvôpaplava-vādin), which is an univocal reference to Jayarāśi and his circle. Traditional Cārvākas reject the existence of the omniscient because this can neither be proved through perception (pratyakṣa), which is the primary cognitive criterion, nor through debational inference (anumāna), which is secondary to and dependent on perception, whereas omniscience is beyond the direct grasp of senses, and therefore not amenable to perception, which cannot prove it (AṣS2 30.8–29). Jayarāśi is distinguished from the traditional Cārvākas in not accepting any cognitive criterion and any categories, and therefore also not accepting omniscience and the existence of an omniscient being. In Vidyānanda’s opinion, such an epistemological stance does not allow Jayarāśi to differentiate, in terms of a successful proof, between rejection and acceptance of omniscience:

Some, namely those who propound the dissolution of [all] categories, accept that all categories of cognitive criteria such as perception etc. as well as all categories of the cognoscible are dissolved. [However], their stance, which is absolutely void of [any backing through] cognitive criteria cannot be differentiated from another stance that all [categories of cognitive criteria and of the cognoscible] are not dissolved.Footnote 90

Vidyānanda does not report Jayarāśi’s views as expressing doubt whether an omniscient being can be there or not, but as asserting such a being’s non-existence due to deficiency of any proof.

17. Conclusions. With all the above evidence in view, it now remains to decide whether to interpret Jayarāśi as a materialist, a sceptic, both or neither? Jayarāśi does not, as a rule, proceeds to demonstrate the principle of equipollence (isostheneia) of two opposing views, the well-known trademark of the sceptic (e.g. Pyrrhonist): he never claims that one can argue both in favour of x and agaist x equally satisfactorily. In fact, positive cases when he supports claim P on one occasion and then claim non-P on another are quite difficult to find. Similarly, he never suspends judgement on an isue, maintaining that we cannot be sure about x. In this sense, he does not fit a standard description of a sceptic.

There is one aspect which could potentially make him a sceptic, though. His whole treatise is famously devoted to the destructive critique of all (noteworthy) definitions of all cognitive criteria, and instead he presents no alternative, no positive solution. His strategy appears to be solely negative, and aimed at demonstrating that one is not justified in any manner to any knowledge claims. At the same time, he suggest to follow the standard practice of the world (loka-vyavahāra), which entails a provisional entertainment of ordinary beliefs, including a commonsensical belief that the world is out there the way we see it. This may indeed appear like a sceptic’s stance, as argued by Julia Annas and Jonathan Barnes, who provide two slightly different accounts of the sceptic:

The sceptic argues that, under pressure from his arguments, we must abandon many, or indeed all, of our claims to knowledge and confess that in truth we know very little. At first, that conclusion may seem heady and sparkling; but on repetition it may come to appear flat. For the sceptical challenge leaves all our beliefs intact: provided only that we do not claim to know anything, we may continue with our usual assertions and persist in our usual beliefs. [emphasis mine] (Annas–Barnes (1985, p. 7)).

The ancient sceptics did not attack knowledge: they attacked belief. They argued that, under sceptical pressure, our beliefs turn out to be groundless and that we have no more reason to believe than to disbelieve. As a result, they supposed, our beliefs would vanish. (Annas–Barnes 1985, p. 8).

Both these accounts might be adequately applied to some of what Jayarāśi says. He challenges beliefs, and the title of is work intimates that it is about the dissolution of the belief in categories and reals. He also seem to urge that one should follow the worldly path (laukiko mārgaḥ) and continue with our usual assertions and everyday beliefs, even though they may be groundless. With this interpretation, Jayarāśi’s attitude might resemble, for instance, that of David Hume, who did not call to suspend all judgment the way a Pyrrhonian sceptic would do because of lack of any objective grounding of knowledge, but who could be classified as an Academic sceptic, prone to follow the reasonable judgments one continues to commonsensically make in ordinary life, regardless of lack of any objective ground for his/her knowledge.

However, attractive as it may be to find a genuine case of Indian scepticism, such reading would, as I argue in this paper, misrepresent Jayarāśi, given all the evidence we have. First and foremost, it can be demonstrated that Jayarāśi does entertain positive views on a number of issues, at least the following:

17.1. he rejects universals as objectively existent entities (§2);

17.2. he is a nominalist: he accepts only particulars which for him are external, macroscopic, non-momentary objects (§3), see below 17.13;

17.3. he takes consciousness to be a product of matter (§4);

17.4. he rejects the existence of the soul or any kind of conscious entity (such a a personal series, santāna) that could exist independently of the material body (§5);

17.5. he rejects afterlife and transmigration (§6);

17.6. he rejects karmic retribution (karman), including virtue (puṇya) and vice (pāpa)(§7);

17.7. he rejects heavens and hells (§7);

17.8. he rejects liberation (§8);

17.9. he rejects ritualism and and the purposefulness of religious cult (§9);

17.10. he rejects religious revelation and religious authority (āpta) (§9);

17.11. he rejects Dharma (§9);

17.12. he rejects the supernatural and imperceptible, such as demons (piśaca), atoms (paramâṇu) and god (mahêśvara) (§11);

17.13. he accepts external, macroscopic, non-momentary, material objects as only existent (§12);

17.14. despite rejecting the pramāṇa model of epistemology, he accepts direct experience (anubhava) as the ultimate resort (§13);

17.15. he most probably also accepts recognition (pratyabhijñāna) as an additional instrument to know (recognise) the world (§14);

17.16. he rejects omniscience and the existence of omniscient beings (§16);

17.17. further, as I demonstrated in §10, his views consistently fall into the body of beliefs commonly associated with materialists in India from the fifth/fourth centuries BCE.

We should bear in mind that also to reject claim P can be a positive view, i.e. the claim that it is the case that non-P.

We have now strong reasons to believe that Jayarāśi can be classified as a representative of the Cārvāka/Lokāyata tradition, since a number of his own beliefs overlap with those known to be represented by Indian materialists, in addition to the fact that he is classified as a ‘Laukāyatika’ by Vidānanada. Further, his quotations from Br̥haspati should be treated quite differently than the references to other Indian thinkers whom he critised. Br̥haspati is referred to by Jayarāśi because he himself saw himself as a representative of Br̥haspati’s tradition.

What follows is that ascription of Jayarāśi to a sceptical tradition is most problematic, and both his method and statements that make a sceptic appearance should be reconsidered.

What might appear as an approach to a problem typical of a sceptic, turns out to be a different method of critical examination on the part of a rationalist. Quite instructive in this respect is a neat account of Jayarāśi’s views which Vidyānanda provides in the Aṣṭa-sahasrī (AṣS1 37.9–42.116 = AṣS2 31.2–36.6), based also on direct citations, and Jayarāśi’s method is faithfully recapitulated as follows:

Accordingly, since in this way, a definition of cognitive criteria in general is not possible, and also a definition of [respective] cognitive criteria such as perception etc. in particular is not possible, therefore the category of cognitive criteria, when closely examined, cannot be established. Since it is not established, how could one possibly establish the category of the cognoscible? On the basis of such a reasoning, the dissolution of [all] cagetories is established.Footnote 91

In other words, with respect to a particular problem P, if there is no available proof of P, we simply reject P, instead of being contended with a suspention of judgement whether P or not-P is the case. This is a basic method a positive/dogmatic philosopher follows, as distinguished from the sceptic. For instance, with respect to god, since no valid proof is at hand, one may either reject his existence, which would be a position of a positive/dogmatic philosopher, or accept that his existence is unknowable in principle, as the agnostic does, or withhold our view and remain comfortable with the fact that we can never know whether god exists or not, being a sceptic’s position. What Jayarāśi effectively does both in this particular case of the validity of cognitive criteria and on many other occasions is that, since no proof of P can be demonstrated, therefore non-P follows. This is also how Vidyānanda interprets his method, for instance in the case of omnicience (§16). It is impossible to find any evidence or prove that an omniscient being exists, therefore one asserts that there is no omniscient being. This methodological principle of economy is that of Ockham’s Razor, which stipulates ontological and epistemological parsimony. Not surprisingly, both William of Ockham and Jayarāśi refused to admit universals, for instance. As long as one has no rational reason, or proof, one does not allow entities beyond necessity. And this is an approach very different from that of the sceptic. As a philosopher, Jayarāśi seeks parsimony, and demands a proof for any belief to be accepted as true.

But how to interpret an overall strategy of Jayarāśi who in most cases is merely keen on demonstrating that there is no proof of x, but formulates no explicit conclusion? No doubt, we would expect of a ‘dogmatic’ philosopher that he rounds up a discussion in unequivocally assertoric terms that either ‘x is the case’ or ‘x is not the case’. But, similarly, this is precisely what should also be anticipated in the case of the sceptic, namely to articulate at least some kind of conclusion to the effect that ‘therefore x is doubtful’, or ‘therefore x cannot be known for certain’, or ‘we have no good reasons to adjudicate between ‘x exists’ and “x does not exist”’. It seems however that it is a general strategy of Jayarāśi of not expressing what he considers an obvious conclusion of his examination of a particular topic: since there is no evidence to adopt the existence of x, therefore the existence of x is to be rejected, following the principle of economy. It is apparently his peculiar style that has contributed to the generally accepted belief that he is a sceptic.

He does so on many occasions, for instance in a section quoted above (§14), he develops his line of argumentation in favour of the existence of a non-momentary object based on recognition, but he does not draw any conclusion, which is clearly there to be supplemented (below in square brackets):

…we know it (i.e. that this is the same thing) through nothing but recognition [of the form:] the very same existence [of the numerically one thing now] occurs. Also in [our] very first experience [of the thing] this existence occurs. [Therefore it is the same thing].Footnote 92

We have therefore to reverse our thinking of Jayarāśi as a sceptic and admit that he did entertain a range of positive beliefs, most of which were in sync with the standard doctrine known to be represented by Indian materialists, albeit he did develop his most original and individual way of philosophical enquiry, which did prove destructive to his opponents to the degree that they simply ignored him en mass instead of engaging with his critique.

An implication of my discussion of Jayarāśi’s actual systemic affiliation is far reaching. Since it is rather problematic to classify both Nāgārjuna and Śrī Harṣa as sceptics (except perhaps for their methods),Footnote 93 the edifice of the genuinely sceptic tradition of India – with the three pillars of scepticism removed – also crumbles. As a result, we may in fact end up with no proper Indian sceptic in the true sense known to us except for Sañjaya Belatthiputta (Sanskrit: Sañjayin Vailasthaputra).