Power and Insight in Jain Discourse V-VI

— V —

Any empirical analysis of conversational implicatures requires a careful description of the discursive field and conventional ways of speaking in different 'speech situations' which form part of the 'communicative competence' (HYMES (1972b))^[90] and 'repertoire' of the members of a particular 'speech community' (GUMPERZ (1964), (1972)).^[91] From an observer's point of view, there are four formal contexts, or 'speech situations'^[92] and associated 'speech events' and 'genres', which are relevant in the case of Jain discourse: (a) the religious debate (vivāda^[93] or prayoga),^[94] for instance at a royal court or modern court of law, (b) the public sermon (pravacana),^[95] (c) the interaction between ascetics and non-ascetics in informal settings (bhāṣā), and (d) the interaction between ascetics (vinaya).^[96] Each of these settings involves a different set of social and stylistic conventions, norms and expectations.^[97] The agonistic debate at a royal court or a court of law for instance is a highly pragmatic affair whose outcome might decide the fate of a particular monastic group. Here, the contextual rules are defined by the king or the state, and classical philosophical and political rhetorical skills are in demand. The public sermon, on the other hand, is comparatively informal, despite its conventional setting and asymmetrical structure. It conveys religious content, inspiration and insight in an entertaining manner, and is often followed by casual discussions at the end. The fundamental norms of comportment are defined by the monastic community itself. The interactions between ascetics and non-ascetics ideally take place within a hierarchical context of worship of ascetic purity, and are sometimes highly ritualised.^[98] However, there are also informal interactions, which require greater discriminatory skills on the part of the ascetics for not transgressing the norms of non-violent interaction. The code of polite interaction within the monastic order itself is derived from the principles of ascetic seniority (dīkṣā-paryāya) and group leadership, and operates in a similar manner.

I have ordered these four ritualised types of discourse sequentially in terms of their increasing coincidence with Jain ideals, starting with political rituals, then conversion rituals, interaction rituals, and finally monastic rituals. Other forms of classification are possible; for instance in terms of degrees of formality, social setting (assembly or individual contact),^[99] or regarding the relative importance of discourse in ritual.^[100] However, for our primarily sociological purpose, the crucial variable is 'interactional competence', that is, the knowledge and ability to use rules and moral judgement for interaction in specific situations. Interactional competence presupposes 'cognitive competence' (PIAGET (1937)), 'linguistic competence' (CHOMSKY (1965)),^[101] 'communicative competence' (HYMES (1972a)), including 'oratory skill' (BLOCH (1975)). HABERMAS (1984: 187 ff., 231 f.) defines interactional competence more specifically as the universal pragmatic ability to regulate social conflict consensually through rational agreement, even in situations of severe discord, in accordance with the three universal pragmatic validity claims. He distinguishes further between the intuitive or conscious knowledge of specific grammatical and socio-cultural rules and contexts, and the mastery of the universal pragmatic conditions of the possibility of mutual understanding (Verständigung).^[102] Because Habermas' notion of 'communicative competence', a composite of cognitive, linguistic and interactional competencies (in a more narrow sense),^[103] is more abstract and formal than Hymes' sociolinguistic notion of 'communicative competence', it circumvents the problem of analytical reification of 'speakers' or of 'speech communities'. Building upon HABERMAS' (1991: 111–3) distinction between two levels of competencies, also reflected in his differentiation between universal 'moral' and cultural specific 'ethical' principles, and HYMES' (1972b: 57 f.) notion that knowledge of grammatical rules and of determinate ways of speaking together form the communicative competence of a particular speech community or individual, I propose to differentiate further between general 'interactional competence' and specific 'Jain interactional competence', that is, the formal procedural knowledge necessary for constituting discourse in agreement with Jain principles.^[104]

Jain interactional competence is informed by general 'Jain religious competence', which can be defined as the sum total of Jain religious knowledge learned by an individual. In addition to general 'Jain interactional competence', religious rules and contextual knowledge necessary for navigating specific contexts can be studied, and the 'Jain religious repertoire' of a specific population or individual analysed (as part of a particular 'socio-cultural repertoire').

General 'interactional competence' refers to the ability of a flexible reflective regulation of total conduct in terms of universal moral presuppositions, contrasting with institutionalised settings, which, through their formal properties, relieve the individual to some extent from the task of defining the situation and from difficult moral choices. 'Jain interactional competence' varies with the degree of knowledge and the internalisation of general and cultural specific (sectarian etc.) It is necessary to distinguish Jain principles and their contextual rules of use; that is, the cognitive and noncognitive (motivational, emotional etc.) know how to act in accordance with abstract and specific doctrinal principles in different socio-cultural circumstances. Jain principles and rules of application, of varying degrees of specificity.^[105] General Jain principles are characterised by constitutive rules, generative of the field of Jain discourse as a whole. They are exemplified by ideal Jain speech events. Jain interactional competence is context-sensitive. The corresponding rules are deliberately constructed to enable a performer to react flexibly to emergent properties of a specific speech situation, and to shape it in accordance with Jain principles. Since principles and rules only negatively determine the field of pragmatic possibilities, speaking the Jain way, or avoiding interaction altogether, are creative acts, performances, which ideally transform any speech event into a Jain discourse. Actual speech performances, or Jain 'speech styles' in the terms of HYMES (1972b: 57), I would argue, can not be sufficiently understood in terms of statistical features of overt behaviour, or as deviations from a general norm or ideal, but only as constitutive acts dependent on 'qualitative judgements of appropriateness' of selected features in a given situation.^[106]

Theoretically, the required interactional competence of the ideal 'pure ascetic' (śuddha-sādhu) increases with the degree of informality of the interactional situation, because ascetic self-control, in accordance with Jain principles, is the price for freedom from external (social and ritual) control. In practice, Jain ascetics try to prevent rule-transgressions due to insufficient procedural knowledge with various methods: for example, by teaching context-sensitive rules not in abstract, but through exemplary cases, or by keeping inexperienced ascetics away from morally overcomplex situations through strict supervision. Similar competencies are outlined for the Jain laity. An ideal lay person is characterised, for instance by Hemacandra (YŚ 1.47–56), as one who is listening (śravaṇa), memorising (dhāraṇā), showing respect (vinaya), being an expert in sacred lore (pravacana-kuśala), and also avoiding religiously unprofitable speech (sat-kathā) etc. (WILLIAMS (1983: 265– 72)). Thus, it is always the Jain interactional competence, the religious and cultural repertoire of an individual or group, which generates the capacity of defining or judging situations in terms of Jain principles, of imposing or inferring implicatures, and determining the relevant type of rhetoric or hermeneutic procedure.^[107]

Once the settings and the required competencies are considered, the social implications of Jain discourse can be analysed in terms of the conditions and presuppositions of the acceptability of particular validity claims in Jain intersubjective hermeneutics. These can be discovered by negating speech acts as a whole.^[108] The question is always: What are the social implications of a religious claim in this particular context? 'What makes a religious statement acceptable?'^[109] Ultimately, such a question can only be answered by the participants themselves, because only they can determine on which grounds to accept or to reject a particular claim. In other words, the explication of implications can never be achieved objectively, for instance by explaining the 'cogency of ideas' through a description of 'the structure of a problem situation' alone, as GOMBRICH (1988: 11) suggests, but is always an out come of specific intersubjective processes of socio-cultural self-interpretation.^[110]

Classical performative approaches to ritual or ritual language, for instance, which made use of Grice's or Searle's analysis failed to investigate this (like Grice and Searle themselves). In fact, they neglected the role of discourse in ritual altogether, in favour of either cultural determinism (Tambiah) or mental (Humphrey and Laidlaw) or material reductionism (Bloch). Following Searle's and Habermas' interpretation of the dual (illocutionary and propositional) structure of the speech act, TAMBIAH (1985: 153) argues that, because ritual action appears to mediate between (cultural) ideas and (political) practice, it can be interpreted as a performative matrix of 'indexical symbols (and indexical icons) as duplex structures carrying semantic and pragmatic meanings':

'The duality thus points in two directions at once—in the semantic direction of cultural presuppositions and conventional understandings and in the pragmatic direction of the social and interpersonal context of ritual action, the line-up of the participants and the process by which they establish or infer meanings. We note that the sense in which I imagine actors to infer indexical meaning is similar to Grice's formulation of "conversational implicature", in that by saying or enacting something a certain meaning is implicated, which can be readily understood (conventional implicature) or is capable of being "worked out" (non-conventional implicature), given certain contextual features and certain common understandings' (TAMBIAH (1985: 154)).

Although he treats ritual acts as speech acts (thus identifying action and his own action description), speech itself is curiously absent in Tambiah's conception (and where he talks about it, he returns to the traditional 'power of words', śabda-śakti, theory). Moreover, the propositional component of the ritual act, its symbolism, is not treated in terms of specific validity claims whose debatable normative implications may or may not be acceptable, but identified with timeless and incontestable cultural paradigms. Tambiah's performative analysis is therefore still oriented towards the theological model of textual hermeneutics, that is, the dogmatic explication (phronesis) of contextual implications of mythic paradigms or sacred texts. His approach assumes that the knowledge embodied in texts is superior to the interpreter's, and has to be accepted.^[111] Because his performative analysis does not clearly distinguish between ritual acts and ritually mediated forms of social interaction, it must assume socially binding effects emerge automatically from the experience of ritual acts themselves, and that the authority of traditional ritual form should always be accepted, without question. The underlying notion of ritual enactment, a variation of the Neo-Kantian model of value realisation, cannot account for processes of intersubjective negotiation of conditions of acceptability and for the contextually varying social implications of ritually communicated validity claims. The presumed timelessness of mythical paradigms re-produces the characteristic pseudo-concreteness of Durkheimian functionalist studies of ritual performance, positing social integration both as a presupposition and result of ritual, without investigating the constitutive role of discourse in and about ritual itself.

Tambiah's theory does not throw light on Jain religious discourse, which appeals to reason and insight, instead of formulaic repetition, and whose contextually specific normative claims are constantly challenged and revised by the Jain ācāryas, under the threat of failure to appeal. In sum, because of the lacking theory of social constitution and cultural change, and the absence of sociolinguistic interest, Tambiah's approach cannot assist analysing the problem of contextual ambiguity of indexical meaning, which I deal with later in this essay. As an alternative, I suggest investigating Jain ritual discourse through the analysis of the Jain discourse ethics and traditional interpretative procedures for negotiating social meaning; to avoid falling into the same trap as Bloch, who identifies an increase of formality generally with a loss of meaning, which Tambiah rightly criticises.

Even if the idealist limitations of universal pragmatic or culturalist discourse analysis could be overcome, there is still the problem of how to analytically reconstruct implicatures in specific cases in a methodical way. This question is particularly prominent in theological, juridical and moral-philosophical debates concerning the status of rule application, that is, of interpretative processes that are involved in the specification of principles and rules through the explication of the contextual implications of given principles and rules.^[112] As language in general does,^[113] explicit normative principles and rules always entail an anonymous indexical element. That is, they implicitly refer to a range of possible situations as their field of applicability, which we must know and accept, if we wish to act accordingly. To explicate general conditions of applicability, one has to refer to paradigmatic situations. Paradoxically, the range of applicability diminishes with the increasing specification of the conditions, and increases with the diminishing specification.^[114] In other words, the problem is the principal vagueness and ambiguity of the indexical meaning of symbolic forms, which, as GLUCKMAN (1955: 293) and TURNER (1986: 44) argued, is in fact a positive functional feature of abstraction, because it secures the adaptability of general principles to changing contexts.^[115] An exhaustive characterisation of merely negatively determined situations, therefore, cannot and, indeed, should not be achieved once and for all by participants, even if a principle itself is accepted.^[116]

The contextual meanings and social implications of symbolic forms (and the forms themselves) are always underdetermined and hence negotiable.^[117] They can only be established temporarily through use and agonistic intersubjective interpretative processes, whose outcome is always open, even if methods of argumentation and discursive procedures are conventional,^[118] and predicated on idealised universal pragmatic presuppositions. I therefore suggest supplementing standard descriptions of ritual, or action in general, by an analysis of the discursive procedures which define situations and establish relevance and contextually acceptable social meaning. The discursive field surrounding rituals and other social settings should be the main focus for an investigation of social dimensions of ritual efficacy. Conventional analyses of ritual focus almost exclusively on the given ceremonial setting and on the performative experience itself, and tend to give undue weight to the procedural knowledge of ritual functionaries, textualised cultural ideas, and articulated subjective experience. The resulting meaning reductionism can be avoided by investigating discursive negotiations of validity claims in situ,^[119] including possibly the relationship between general, for instance moral, language games and institutionalised forms of reflexivity, such as theology and the social sciences. The analysis^[120] of implicit judgements invoked in everyday practices of moral justification, moral utterances and moral disputes, can also demonstrate how cultural resources are used as potential reasons for claims of validity. From this perspective, social solidarity does not only emerge exclusively as a consequence of partaking in 'effervescent' rituals, to use Durkheim's expression, but primarily—and increasingly so—as the result of the participation in socially constitutive discourses of norm legitimation and shared definitions of situations.

The paradigmatic example of practical hermeneutical explications of implied indexical meaning, are intentionally multivocal forms of communication within 'master-slave' relationships; often discussed either in terms of formalisation of discourse, and of understanding implicit directives (e.g. commands), or in terms of the instrumentalisation of interpretation as such by the rhetoric of spiritual power in contexts of traditional authority. Building on the work of Brown and Levinson and Strecker on the influence of social structure on discourse, DRECHSEL (1994: 50), for instance, draws an analogy between the relationship of off-recordness and on-recordness in politeness theory, and the relationship of the ambiguous symbolism of 'sacred' kingship and the unequivocal social and material implications of royal pomp, which, he suggests, have to be worked out by the lower strata of society qua unspoken command. Although mere compliance or non-compliance with implicit directives in no way shows that they are legitimate or illegitimate,^[121] the explanatory problem of this institutional theory is the same as in Habermas' analysis of latent strategic action in terms of universal normative presuppositions. The king's dual socially constitutive role, as a living symbol of society and as an individual with its own specific interests, involves his subjects in a paradoxical doublebind situation.^[122] Because the king is the condition of socio-cosmic existence, royal power appears not primarily 'exploitative', but 'attractive', to the extent that the subjects try to read the king's mind and, ideally, to fulfil his unsaid wishes. Because the king is the condition of social order everyone is forced to identify with him. In fulfilling his wishes, the subject reproduces the conditions of its own social existence as a social being. A good example is the relationship between the king and his officials in the context of the 'polite society' of the court, and their characteristic mutual strategic orientation: the characteristic ambiguity of the king's role forces the officials to work out the 'implications' of royal gestures, and to fulfil his unspeakable material needs in form of service, presents etc.^[123] Even a cursory glance at the vast literature on South Asian kingship shows the empirical significance of such phenomena, first theorised by HEGEL (1981: 145–77) in his analysis of the dialectic of recognition, and investigated by ELIAS (1983) in his sociological study of the French court and KANTOROWICZ (1957) on the ideology of European kingship.^[124] BURGHART (1996: 308) found similar mechanisms within the political discourse under the Hindu monarchy in Nepal of the 1980s, combining arguments from Habermas and Bloch: 'The king cannot speak informally to the body politic: rather he must speak formally to everyone, for he represents everyone. Little that he utters, therefore, can be taken at face value, and one treats Governmental speech with some suspicion. It is, as a matter of course, something distorted.'

The relationship between a king and his subjects is structurally homologous to the relationship between senior and junior Jain ascetics, and to the relationship between ascetics and laity (śrāvakas, 'listeners', or upāsakas, 'servants'). In Jain scriptures, the junior partners in both relationships are characterised by the combined virtues of politeness (vinaya) and serving the guru (vaiyāvṛttya) (Uttar 30.30–7). From the point of view of Jain laity, even the relationship between Jains and Non-Jains should be analogous to the relationship between ascetics and Jain-laity (ascetics: laity:: laity: non-Jains). Similar (on / off record) strategic orientations are therefore to be expected. From an anthropological perspective, the crux of politeness theory is that functional contributions for the maintenance of public self-image are not coerced but received through processes of deliberate 'fulfilment' of face-wants of speakers of superior status by 'hearers' of inferior status; particularly the 'faces' of legitimate public figures, which condense the sense of identity in wider social circles. Social power operates here indirectly through the medium of free will and consent, if only under the implicit threat of social sanctions in the case of non-satisfaction of facewants. In the words of TURNER (1986: 30), ritual in general is a mechanism that 'converts the obligatory into the desirable.'^[125]

— VI —

It is essential for my argument to investigate Jain conceptions of pragmatic language usage, and to compare them with the theories outlined in the previous sections. The central question concerns the applicability of the theory of communicative action and its forms of operationalisation to the Jain context; particularly the role of the 'universal' validity claims and of the Gricean maxims. I have argued that the Jain theory of speech plays a similar role within Indian social philosophy as Habermas' theory of communicative action does within Western philosophy, because both proclaim the primacy of morality and ethics in language usage. Underlying Habermas' theory of communicative intent (verständigungsorientierter Sprachgebrauch) is not only the notion that, by definition, in the immediate context of communication all participants must have a common interest in the maintenance of linguistic exchange, and hence implicitly presuppose the ideal of non-violence, at least of the avoidance of physical violence, but also that they are implicitly oriented toward reaching agreement.^[126] In this regard, Habermas' notion of infinite consensus represents a functional equivalent of the Jain notion of infinite knowledge leading to salvation (see infra, p. 129 f.). Moreover, universal pragmatic validity claims can be usefully compared and contrasted with the Jain vows of ahiṁsā and satya. On lower levels of abstraction, the Jain doctrine of the ways of speaking (bhāṣā-jāta), as exposed already in the older texts of the Śvetāmbara scriptures, can be usefully compared and contrasted with the conversational maxims of Grice; and conversational implicatures in Jain discourse, theorised in Jain literature, can be analysed as forms of latent strategic action in terms of Brown and Levinson's typology of FTAs. I do not argue that there is an exact correspondence between these modern and ancient Jain schemes of interpretation. Nor do I believe that there is only one way of reconstructing the social implications of Jain theories of language from the point of view of modern social philosophy and logic. But the comparison generates new perspectives on both Jain and Western philosophy, in particular on the ways in which fundamental ontological distinctions or questions can function as codes for the constitution of different discourses.

The principal difficulty of this enterprise is the apparent incompatibility of dialogical and monological perspectives; of the—quasi-legal—intersubjective and egalitarian normative underpinnings of the theory of communicative action and the self-centred and hierarchical normative foundations of the Jain theory of speech, that is, their different ontological commitments.^[127] Habermas' differentiation between universal 'moral' and culturally specific 'ethical' presuppositions undermines from the outset the claim to universality of any religious ethical system. Since the legitimacy of Jain religious speech is rooted in the traditional authority of the speaker (āpta), that is, his/her Jain religious competence, rather than the procedural form of communicative action (śabdānuśāsana),^[128] from the perspective of the theory of communicative action, its claims to universality can only gain acceptability in an open field of discourse to the extent that it articulates phenomena for which no other language is available.^[129] Jain authors have also formulated a principles of discourse intended to transcend cultural boundaries. These principles can, however, only claim universality to the extent that they fulfil the condition of universal acceptability. Yet, this criterion is not foregrounded in Jain texts. A rational defence of the universality of Jain ethics will need to reconstruct the presuppositions of the egological Jain ethical perspective from the point of view of general interest. Conversely, discourse ethics rooted in formal pragmatics has only weak regulative force and needs to be supplemented by obligatory norms of action. The reduction of religion to morality and ethics and the epistemic approach, which Habermas inherited from Kant, are however shared, to some extent, by Jain doctrine. The following comparative study of Jain discourse ethics begins with the question of the status of religious language in Jain philosophy, especially intentionally multivocal language, which, at first sight, appears to violate the fundamental norms of universal pragmatics.

The main difficulty in understanding Jain concepts of religious language derives from the paradoxical, direct and indirect, nature of Jain religious knowledge (āgama). The problem is that, doctrinally, every utterance can be interpreted both from the transcendental or ultimate point of view (niścaya-naya or pāramārthikanaya) (henceforth PN), and from the practical point of view (vyavahāra-naya) (henceforth VN).^[130] For a participant, ultimate and relative points of view are not necessarily absolutely distinct, but complementary, hierarchically related modes of orientation, which mutually implicate one another (like noesis and noema in Husserl's notion of intentionality). We are thus confronted with a conception of stages of religious insight and corresponding ways of using and interpreting language, measured in terms of degrees of insight (jñāna) and restraint (saṁyama), i.e. practice of non-violence (ahiṁsā).^[131]

The Jain tradition uses the term āgama ambiguously.^[132] It refers primarily to the ultimate truth, which Mahāvīra, the last Tīrtha(ṅ)kara, had experienced directly and preached to the world, and in a second sense to 'canonical' Jain scriptures, which still form the doctrinal basis for sermons and writings of present day ascetics (MĀLVAṆIYĀ (1968: 10–13).^[133] In classical Jainism, knowledge of ultimate religious truth and reality can only be gained through meditation (bhāvanā). However, in the present 'unhappy' time cycle (duṣamā avasarpinī) direct insight cannot be achieved anymore. Instruction through the words or testimony (śabda or āgama) of Mahāvīra, handed down by the religious authorities (ācārya), is the only way of gaining religious knowledge at all. The doctrinal primacy of cognition over language explains why Jains (and Buddhists) 'have tried to preserve the meaning of the words concerned and not (like the Brahmins—P.F.) words themselves' (MĀLVAṆIYĀ (1968: 1)). Jainism teaches that, ultimately, all words—and indeed: doctrines—are neutral; it is the attitude of the knower alone that turns them into means of valid cognition (pramāṇa).^[134] Conversely, for the knower, each word of the Jain scriptures appears to condense the teachings of Mahāvīra in a summary fashion. This is part of its evocative power:

'The Jainas maintain that the meanings of one single sentence, grasped by different hearers, are innumerable in accordance with their innumerable capacities … a sentence in the āgama has the power to suggest in various ways innumerable meanings' (MĀLVAṆIYĀ (1968: 12)).^[135]

It is crucial for any understanding of Jain attitudes to language that, from the ultimate point of view (PN), sensuous cognition (mati) and words (śruta or āgama)^[136] are both forms of indirect, mediated knowledge (parokṣa) (TS 1.11).^[137] This is why Digambara Jains insist that the sermons of a Jina take the form of a miraculous sound (divya-dhvani), which radiates the meaning (artha) of his teachings instantaneously, not mediated through words. As there is no language for the unspeakable (avaktavya) ultimate truth, any language can be used to express it.^[138] As a consequence, 'from the absolute standpoint, the validity of a word or sentence is not intrinsic but extrinsic. That is, it depends on the merit of the speaker as well as of the listener' (MĀLVAṆIYĀ (1996: 2)). The conventional perception that 'there is no possibility whatsoever of any faults or defects in the content of the Agamas' thus rests entirely on the dogmatic belief that 'the speakers of the Jain Agamas were selfrealized (i.e. omniscient, P.F.) persons' (MĀLVAṆIYĀ (1968: 3)). The normative basis of traditional hermeneutics euphemistically stated: 'In substance, this means that the absolute standpoint mainly keeps the listener in view while determining the validity of the scriptures, and the empirical standpoint mainly keeps the speaker in view while determining the same' (MĀLVAṆIYĀ (1968: 2)).^[139]

Although kevala-jñāna (omniscience), the highest form of pāramārthika knowledge according to Jain doctrine, cannot be achieved in the present age, even today's ascetics are supposed to have mind reading (manaḥ-paryāya) and clairvoyance (avadhi) capacities. Through their training in non-identification they are expected to be capable of switching perspectives between pāramārthika-naya and vyavahāranaya, and, thus, to look at the same phenomenon both directly and indirectly (note that from PN, VN appears as a form of indirect cognition). This has significant consequences for the apprehension of verbal utterances. From the religious point of view (PN), language appears merely as a superficial ornament, which disguises the essential truth and beauty of the inner soul, of pure consciousness, which can be experienced through meditation only.^[140] In other words, the ambiguity of the concept of āgama is itself understood in terms of the 'absolute' and 'practical' points of view, which permit distinguishing clearly between ultimate religious meaning and linguistic means.^[141] The main consequence of this perspective is the cognitive desubstantialisation of everyday conceptions of language. From the worldly point of view (VN), the propositional content of a word (śabda) depends on convention (and on its power of signification).^[142] The semantic, and pragmatic, implications of a grammatically correct verbal utterance vary according to context. Speaker's intention (abhiprāya) is one of the contextual conditions of meaning. Yet, it is rarely discussed in the context of objectivist Jain (and Nyāya) semantics.

My general argument is that the doctrinal distinction between practical and transcendental orientations, and the learned ability of reversing perspectives, informs both the discernment and the systematic generation of plural, multivocal meanings in Jain discourse.^[143] In my view, the cognitive competence of perspective alternation is a fundamental presupposition of Jain hermeneutics and rhetoric, and can be used both for the pursuit of insight and power.^[144] Rooted in Jain ontology, the distinction between empirical and transcendental perspectives was an implicit feature of Jain doctrine even before its philosophical conceptualisation. The fundamental religious problematic of Jainism, framed by the asymmetrical soul / body, non-violence / violence codes, creates paradoxes and pragmatic ambiguities which require interpretation and contextual specification / conditioning with the help of additional coding.^[145] How to translate these asymmetrical codes into practice? Many Jain texts address this problem. According to classical Jain philosophy of standpoints (anekāntavāda), the epistemic tension created by the basic soul-body dualism of Jain doctrine, can only be resolved, if at all, by a process of perpetual alternation of PN-VN perspectives.^[146] In Kundakunda's Samaya-sāra (SSā), this epistemic necessity is endowed with a soteriological function. The technical Jain syād-vāda philosophy, the conditionalisation of assertions, by contrast, is based on the conventional point of view (VN).^[147] How do these two types of Jain perspectivism deal with the problem of multivocality? The Jain religious point of view (PN) infuses all phenomena with new meaning by re-coding them in terms of the doctrinal body / soul distinction. I shall designate this type of systematic distortion or appropriation of conventional meaning which at the same time opens up new semantic space, 'ontological ambiguity', in contrast to 'linguistic ambiguity',^[148] in a wider sense, including expressions whose implications are semantically and / or pragmatic underdetermined (vagueness, indexicality, indirectness, presupposition etc.). Ontological ambiguity is not the same as linguistic ambiguity. But it can only be expressed in linguistic form. Therefore, both can be easily confused. Systematically distorted communication qua doctrinal re-coding and latent strategic communication are both predicated on the intentional creation of multiple meanings. Like latent strategic communication, ontological codes imply disguised generalised indexicality in form of the asymmetry of the binary code. Hence, from a practical point of view, both are social practices of symbolisation. Only their purposes, and perspectives, are different. From the point of view of Jain philosophy, the estranged view of the everyday world, which is intentionally created by the transcendental Jain perspective, is a product of discriminating knowledge, not of deliberate deception. This does not mean that it is unambiguous. LUHMANN (1990) argues, in my view convincingly, that every binary distinction creates at the same time clarity on one level, and ambiguity on another, ad infinitum. Different types of ambiguity become visible in the light of different distinctions. For instance, for someone applying the dogmatically prescribed Jain philosophical standpoints (naya) (TS 1.34), and the corresponding procedures for disambiguating contextual meaning, plurivocality becomes problematic in the light of conventional univocality, etymology etc.

More research is required to delineate different types and functions of multivocality in Jain discourse. However, one aspect of the Jain analysis of ambiguity deserves to be mentioned at this point. According to Jain logic, linguistic ambiguity^[149] should not be confused with vague or incomplete description, which Jain naya philosophy contrasts with the epistemic ideal of definite description.^[150] The problem of definite description has recently been discussed by HARE (1981: 40–3) regarding the media tion of general principle and specific situation, e.g. moral conflict.^[151] Hare advocates a variant of the two-level theory of ethics. He argues that an exhaustive description of a particular situation from the point of view of intuitive prima facie principles is impossible, because they must be very general and simple. To inform moral choices, prima facie principles need to be supplemented by exceptions, or by critical principles, which 'can be of unlimited specificity'.^[152] Both prima facie principles and critical principles are universal prescriptions. However, critical principles can only be universalised on condition of unlimited processes of specification by an infinite intelligence, that is, an omniscient being. Because a superhuman 'ideal observer'^[153] can only serve as a theoretical vanishing point for finite cognition, the need for prima facie principles and exceptions arises (HARE (1981: 44 ff.)).^[154] HARE's (1981: 58) characterisation of the cognitive preconditions of 'the rigor of pure prescriptive universality' enables us to recognise the similarities between the Jain ideal of definite description, as a standard against which incomplete or vague propositions can be measured, and Habermas' ideal consensus of an infinite community of interpretation, against which both systematically distorted and latent strategic communication can be assessed. The subject-philosophical role of 'omniscience' and the universal pragmatic role of 'infinite discourse' are functionally equivalent.^[155] The comparison suggests also that the Jain ideal of omniscience is a logical consequence of Jain epistemology, and not a mere product of fanciful religious speculation.^[156] The question whether in specific contexts vague utterances are ambiguous or merely semantically underspecified^[157] cannot be rationally determined from a nonomniscient perspective. From a participant's point of view, under given normative conditions, any utterance can be created or perceived as an intentionally plurivocal symbol, even if the implicit indexical meaning or normative command can be inferred unequivocally. Everything depends on the pragmatic context.

A peculiarity of the Jain theory of speech, derived from the Jain karman theory, is that utterances are seen as material substances which stick to the impure soul but not to a pure one. In fact, speech as a technique of social influence is not only an important subject for proselytising ascetics, trying to attract followers (systematic distorted and latent strategic action are identical in this case),^[158] but also an explicit topic of the Jain doctrine of karmic bondage, that is, the mechanism of binding (bandha) the soul of a listener, through the medium of his / her passion (kaṣāya) and desire (rāga), via an influx (āsrava) of insight-generating pure matter (śuddhapudgala) (cf. TS 5.22). It is not necessary to recall the details and history of this doctrine, because SCHUBRING's (2000: 174, § 84) authoritative depiction of Jain cosmology implicitly operates with a theory of Jain rhetorical influence. Schubring shows, for instance, how processes of possession, which are considered to be the reverse of insight, are explained in certain texts in terms of the sending of inauspicious atoms (aśubha-pudgala) (SCHUBRING (2000: 151, § 69)); and how individuality (which increases with social standing) and thus, ideally, the degree to which karman is felt, generates the power of acting upon others, and to influence and bind people by imprinting (dhāraṇā) certain karmic perceptions onto their soul via the ejection of pudgala (SCHUBRING (2000: 181–7, § 87–91)):

'This process [of speech] is, to put it briefly, the ejaculation (nisarai, nisṛjati) of substances (davva) taken in (geṇhai) previously (now being ready either for use or in store). They consist of ∞ atoms (aṇantapaesiya), occupy the space of ¿ units, last for 1 sam. and own all qualities possible with reference to colour, smell, taste and sensation. Their reception is meant to bring about a close contact (puṭṭha, ogāḍha), i.e. with the units of the soul (ātma-pradeśa, Prajn.), and to it both fine and coarse particles (aṇu and bāyara) are subjected, which is discussed circumstantially. The reception takes place either with or without interruption (antara) in that either reception or ejaculation, or both reception and ejaculation occur within 1 sam., and then, by the way (267a), they will belong to the same content of speech (true, false etc.) for which they were taken in. Their destiny depends on the intensity of speech. As we learn from Prajn. on Pannav. 262b. and from Vy. on Vij. 621b, when speaking [in a] low [voice] the particles of speech leave the mouth in coarse portions (abhinna), but they do not reach far and will perish soon, whereas when speaking loud they are finely divided (bhinna), and in this case speech will increase infinitely and reach the boundaries of the world' (SCHUBRING (2000: 149, § 68).

The strategic production of religious influence via karmic binding is of course not the primary concern of Jain karman theory, but certainly one of the possible consequences of the application of Jain categories to worldly problems. From the conventional point of view (VN), Jain scriptures describe speech as a potential weapon (duppauttamaṇo-vāyā ) (Ṭhāṇ 10.93): 'Speech, so Pannav. 255b says, originates in the soul, while it becomes manifest in the body in the shape of a thunderbolt … [and] exists only the moment when being spoken' (SCHUBRING (2000: 148 f., § 68).^[159] Yet, fundamentally (PN) it is described only as an external ornament of the soul without any intrinsic power, except the power of signification:

'[1] Speech is different from the self (no āyā bhāsā, annā bhāsā). It is concrete (lit. fashioned, rūviṁ), devoid of consciousness (acitta) and inanimate (ajīva) though peculiar to living beings (jīvāṇaṁ bhāsā, no ajīvāṇaṁ bh.). [2] Speech exists and is divided' (bhijjai) neither before nor after but only during actual speaking. ….' (Viy 621a = 13.7.1ab, summarised by DELEU (1970: 199 f.)).^[160]

If the hearer is aware of this s/he cannot be bound. Both interpretations derive from the dualistic doctrinal distinction between soul and body, and their asymmetrical hierarchical relationship.^[161]

We can conclude from the foregoing discussion that, from the point of view of Jain doctrine, both linguistic conventions and intentional language are vyavahāra— worldly oriented—and ultimately of no religious value. Any pragmatic use of language, even for purposes of religious instruction, has only relative value, because the speaker is forced to orientate him / herself toward the external world, and to turn away from the ideal meditative focus on the potentialities of the soul. From this perspective, one can hardly speak of a Jain concept of 'religious language' at all. Jains never fully acknowledged the religious significance of mantras and other types of mystical utterances, which could be distinguish as 'religious' from 'non-religious' language. This applies even to Ardhamāgadhī, the liturgical language of the oldest scriptures. The only religious 'languages' which Jains clearly recognise are silence and meditative sounds, like Mahāvīra's legendary divya-dhvani.^[162]

Because Jain (and most Buddhist) philosophers recognise that language is rooted in convention and mainly used for everyday communication, and other practical purposes, discourse theoretical perspectives promise a useful new angle on the neglected question of the socially constitutive function of Jain discourse. Studies of the Jain philosophy of language to date focus almost exclusively on Jain semantics, in particular on the seven nayas and the sapta-bhaṅgī, that is, pre-defined (but in principle infinite) perspectives for the analysis of the semantics of words or sentences under different contextual conditions. However, MATILAL (1981: 60 f.), JOHNSON (1995a: 253), and GANERI (2001: 137), (2002: 275, 279 f.), following earlier 20th century authors, interpret the Jain naya-schemes as models of 'discourse pluralism', intended to integrate different perspectives in a syncretic and ultimately, that is, from the perspective of unlimited perspectives, 'complete account'. The underlying theory of 'intellectual ahiṁsā' is doubted by CORT (2000: 341) with reference to (a) scriptural examples of the 'history of Jain struggles with non-Jains', and (b) criticism of the assumption that omniscience is 'the sum total of all possible nayas': 'The Jains posit that there is an absolutely true perspective' (CORT (2000: 332)). The argument could be strengthened by a slight modification. Mistaking the 'partial truth' of a one-sided (ekānta) statement for the 'whole truth' is not the only way of being wrong (durnaya) (CORT (2000: 331)) according to Jain philosophy. Jain perspectivism is only concerned with the relationship between partially and wholly true testimony (pramāṇa), not with entirely false testimony (apramāṇa), which is excluded at the outset.^[163] Hence, it is not admissible to say that, by definition, ALL statements contain an element of truth. Strictly speaking, Jain naya doctrines are not theories of discourse, since they are focused exclusively on semantics. Yet, the Jain scriptures and commentaries dealing with the practicalities of ascetic life contain a general theory of the pragmatics of language usage, based on Jain discourse ethics, and show how Jain principles applied to worldly problems generate unique analytical possibilities of perceiving and manipulating speech. In contrast to Jain semantics, which privileges the perspective of the listener, Jain pragmatics privileges the perspective of the speaker.