Freedom because of duty: the problem ofagency in Mīmāṃsā

Elisa Freschi

May 23, 2012

Contents

1 What is Mīmāṃsā? 1

2 Is there anything like free will in Indian philosophy? 3

3 Agent’s causation 43.1 Epistemological basis of compatibilism in Mīmāṃsā . . . . . . . . . 43.2 Vedic-exegetical basis of the concept of agenthood in Mīmāṃsā:

consciousness and ability . . . . . . . . . . . . . . . . . . . . . . . . 53.3 Bhāṭṭa and Prābhākara Mīmāṃsā vs. Nyāya . . . . . . . . . . . . . 73.4 The role of commands . . . . . . . . . . . . . . . . . . . . . . . . . 83.5 Agent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

4 Causality and determinism 124.1 Is there determinism in Mīmāṃsā? . . . . . . . . . . . . . . . . . . 124.2 Causality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134.3 Training desires . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

5 Source of ethics 155.1 Immoral sacrifices prescribed by the Veda . . . . . . . . . . . . . . 16

6 Conclusions 17

References 18

1 What is Mīmāṃsā?

Mīmāṃsā (lit. ‘desire of thinking’, hence “reflection”) is one of the six tradition-ally recognised Indian philosophical systems (darśana). It was born out of anancient tradition of exegesis of Sacred Texts and keeps as its primary focus theVeda (‘knowledge’, Indian sacred texts, not accepted as such by Buddhist and Jainaschools). Thus, like all other philosophical systems generally look at Vaiśeṣika for

1

natural philosophy and at Nyāya for logic, so they look at Mīmāṃsā as a reser-voir for exegetic rules, making it possibly the main source for the Indian approachto hermeneutics in general. This influence is particularly evident in the case ofVedānta schools, where Mīmāṃsā rules, adjusted to the Vedānta view of the Vedaor other Sacred Texts, have been systematically applied and constitute the back-ground of most theological discussions.

The bulk of the system is based (as usual in India) on a collection of sūtra ‘apho-risms’, Jaimini’s Mīmāṃsāsūtra (hence MS) which would be quite obscure withoutŚabara’ Bhāṣya (‘commentary’, hence ŚBh). There is no direct evidence about thedate of Śabara, who knows some sort ofMahāyāna and is aware of a theory of sphoṭa(which seems more primitive than Bhartṛhari’s one1), but does not refer to anyknown author after Patañjali (c. 2nd c. b.C.). Some centuries later, perhaps in the7th or 8th c., Kumārila Bhaṭṭa and Prabhākara Miśra wrote philosophical engagedcommentaries on the ŚBh. These commentaries have been again commented uponby later Mīmāṃsā authors (Pārthasārathi, e.g., wrote a line-to-line commentary onKumārila’s Ślokavārttika and Śālikanātha a similar gloss on Prabhākara’s Bṛhatī ).According to the different tenets of these two main thinkers, Mīmāṃsā is tradition-ally distinguished in two schools, the Bhāṭṭa Mīmāṃsā, which follows Kumārilaand the Prābhākara Mīmāṃsā, which follows Prabhākara. Mīmāṃsā may also bereferred to as PūrvaMīmāṃsā, to distinguish it fromUttaraMīmāṃsā (or Vedānta).For brevity’s sake, I shall restrict the use of “Mīmāṃsā” to Pūrva Mīmāṃsā only.A thinker belonging to the Mīmāṃsā school is called Mīmāṃsaka ‘follower of theMīmāṃsā’.

The mainMīmāṃsā tenets originated out of issues connected with Vedic exege-sis. Mīmāṃsā authors departing from Śabara uphold the a priori validity of cogni-tions (and, consequently, the general reliability of our cognitions), accept linguisticcommunication as instrument of knowledge and focus on action and exhortation.The validity a priori of cognitions (called svataḥ prāmāṇya by Kumārila) is the basisof an apologetics of the Veda. Kumārila explains at length that our sense facultiesonly regard what is (directly or indirectly) perceptible, that common human beinghave no access to morality (defined as kārya ‘what has to be done’), and that there isno reason to believe that there have ever been exceptional human beings who couldhave access to it. Thus, the Veda is the only instrument of knowledge regardingwhat has to be done instead of what there is. To sum up, cognitions are valid untiland unless they are falsified by subsequent evidences, the realm of what has to bedone can only be known through the Veda, and hence the Veda cannot be falsified.Therefore, given that all cognitions are to be considered true unless and until theopposite is proved, the Veda is a genuine instrument of knowledge. By the sametoken, the same principle means that the Veda is only an instrument of knowledgein its unique field, i.e., what has to be done, whereas as for the field of what thereis, sense perception and the other instruments of knowledge based on it (inference,analogy , cogent evidence and absence) are the only valid source of information.Consequently, unless and until contrary evidence arises, the world is the way weperceive it.

1Bhartṛhari’s date is itself controversial, but scholars tend to agree on the 5th c. AD.

2

Linguistic communication (śabda or śāstra, chiefly equivalent to the Veda) isnecessarily an instrument of knowledge out of the same reason, namely, that unlesswe accept it, we would have no way to access the realm of deontic, i.e., of whathas to be done. Last, the Mīmāṃsā habit to investigate the Brāhmaṇa part of theVeda, which mainly consists of exhortations, made them aware of the importanceof this topic and led them to investigate the mechanisms, within language, whichmake one undertake an action. More importantly for the present topic, it also madeMīmāṃsakas consider the issue of agency from the viewpoint of how one is madeinto an agent by a (Vedic) injunction.

2 Is there anything like free will in Indian philoso-phy?

In the intention of its editors, this volume deals with Free Will, Agency and Self-hood. However, a search for a precise synonym of free will in the South Asiancontext is most probably going to be vain, insofar as “free will” is, in Western cul-ture, a concept overload with history, especially Christian history.

In fact, although arguments against or in favour of determinism are very an-cient, the usage of “ free will” (translating liberum arbitrium) is a relatively recentlabel (possibly 13th c. AD) and might seem partly redundant, insofar as its op-posite (‘bound will’, or servum arbitrium) entails paradoxical elements (how canone’s will be bound? Would not it cease to be a will if it were bound?) and onlymakes sense within certain perspectives, such as Martin Luther’s soteriology. Thedisproportion of God’s omnipotency and one’s free will has turned, in contempo-rary philosophy, into that between the evidences of natural scientists, which seemto weigh in favour of determinism (or at least this is their reception among lay peo-ple), and the psychological feeling of freedom. Classical Indian philosophy, it goeswithout saying, predates these developments.

However, although the definition and the semantic of the syntagm “ free will”are historically loaded, one can more easily determine the set of questions or sce-narios this is meant to deal with in Western philosophy and theology. In this way,one can find not a precise synonym, but a functional equivalent2 of this term inIndia. These sets of questions or scenarios are, roughly:

1. The problem of the agent’s causation: how does the process of volition leadto an action?

2. Moral responsibility vs. determinism

3. One’s psychological experience of freedom

2One might suggest, for instance, that the authority of the Veda is for Prābhākaras the functionalequivalent of the categorical imperative in Immanuel Kant, since both fulfil the same role, namely, pro-viding a fix term of reference for what ought to be done. The concept of a “functional equivalent” hasbeen developed by Raimon Panikkar in the context of his comparative approach to Christian, Buddhistand “Hindu” terms.

3

4. Religious discourse

We will see (section 3.1) that the psychological experience of freedom, thoughnot thematised, is often the implicit departure point for Mīmāṃsā authors, who de-scribe the phenomenology of will without taking into account its absolute freedom.3

Thus, in the following, the first two scenarios will be examined in some moredetail. No much space is needed for the fourth one, since Mīmāṃsā authors denyany role to God as a philosophical entity. They may subjectively adore a personalGod, but tend to be quite strict in denying to Him/Her the role of ontological foun-dation. In other words, God has no place as a justification for the system and thereis, consequently, no need to discuss one’s freedom in respect to his omnipotence.The absence of any comparison with God also entails that Mīmāṃsā authors donot need to specify in which sense one can be said to be free, although one is notas free as God is —since, for instance, God can assume all possible forms and wecannot. Pratyabhijñā authors, in this connection, suggest that the limited subjectsonly enjoy a fraction of the God’s infinite power of freedom.

3 Agent’s causation

3.1 Epistemological basis of compatibilism in Mīmāṃsā

In addition to the general points brought forward above (§1), it is worth remem-bering that Mīmāṃsā authors in general favour accounts which mirror one’s ex-perience. Let us consider, for instance, the cognition one has upon entering in adark room. At first, one hardly sees anything at all, but in a few seconds, one startsdistinguishing pieces of furniture, etc. Is this later experience still a case of sense-perception (pratyakṣa)? Buddhist epistemologists deny it this status, since it is theresult of conceptual thinking, insofar as one only distinguishes, e.g., a table insofaras she expects to see one and, hence, conceptually organises her vague sensory datain order to fit her expectations. By contrast, Mīmāṃsā authors contend that, sinceone experiences it as a case of sense-perception, one could only refute it this statusif one had very strong reasons for doing it (see Taber 2005). The Buddhists statethat the notion of a table ought to be due to conceptual thinking, since perceptionhas been a priori defined as only grasping the ultimate particular. But this claimseems to prefer the consistency of the Buddhist epistemological system over that

3This turn has occurred only recently in Western philosophy, where, following P. F. Strawson 1962,Brian Earp welcomed the shift of focus on the question “what it is like to feel free?” rather than on theolder one “are we free?”:“It is worth saluting this shift in emphasis from the millennia-old, seeminglyintractable debates about free will to current philosophical and psychological work on the empiricalquestions surrounding belief in free will, and how it may be influenced by context, motivation, andother factors.Earp 2011, p. 25. Johannes Bronkhorst, though not explicitly making this poinst, whilediscussing Indian conceptions of free will favours an “experiential” approach, which takes into accountour decisions within the process of undertaking actions, independently of whether they are absolutelyfree (i.e., if determinism is wrong) or not (Bronkhorst forthcoming). This favour is probably determinedby the similar attitude of Indian author.

4

of our experience of the world.4 Similarly, since one obviously feels that her willhas a role in the process of undertaking an action, Mīmāṃsakas do not dispute this.Unlike their Western colleagues, Mīmāṃsā philosophers do not question the degreeof freedom of the decisions one experiences as free. They do not, e.g., argue for thefact that our experience of freedommight just be an epiphenomenon accompanyingthe process of undertaking an action,5 or that our experience of freedom might bein fact a fake, since our decisions are completely determined by who we are, some-thing which is a priori determined by facts we cannot interfere with, such as genesand early education (cf., by contrast, G. Strawson 1998). The fact that decisionsare experienced as free is enough for Mīmāṃsā authors to treat them accordingly.Within this framework (which basically takes our intuitions about freedom at face-value), one could also just speak of “will”, since from the point of view of theway a single action is caused, nothing changes if the general laws of the universeallow freedom or not —one would nonetheless believe he has freely decided todo X. Thus, and given that Mīmāṃsakas accept the laws of karman and the ritualcausality linked to apūrva, Mīmāṃsakas adopt what Western authors would label acompatibilist stance.6

3.2 Vedic-exegetical basis of the concept of agenthood inMīmāṃsā:consciousness and ability

Furthermore, Mīmāṃsā authors dealing with agenthood (kartṛtva) primarily focuson the sacrificial agent and that the lack —in the Indian context— of an ongoingpolemics on free will allows for them to discuss technical problems (such as towhom does agenthood pertain in case there are several ritual performers, MS book3, or several sacrificers, MS book 6) without having to question the very conceptof agenthood. Of chief importance, in this case, is that the agent is the one whoconsciously undertakes actions (for an antithetic position, see Cardona’s contribu-tion in this volume). Thus, to be an agent, one needs to be aware of what one is

4This is, e.g., also Kumārila’s answer to the Buddhist Pramāṇavādin who contend that it is illogical toclaim that a single perceptual act can grasp both the individual and the universal inhering in it. Kumārilamaintains that this must be the case, because so runs our experience, see Taber 2005.

5Conscious will, writes for instance the psychologist Daniel M. Wegner, “is not a direct perceptionof relation [between thought and action] but rather a feeling based on the causal inference one makesabout the data that do become available to consciousness —the thought and the observed act.” (Wegner2002, p. 67). And, even more clearly: “The experience of will […] is the way our minds portray theiroperations to us, not their actual operation. Because we have thoughts of what we will do, we candevelop causal theories relating those thoughts to our actions on the basis of priority, consistency, andexclusivity. We come to think of these prior thoughts as intentions, and we develop the sense that theintentions have causal force even though they are actually just previews of what we may do” (Wegner2002, p. 96). Both passages are quoted in Bronkhorst forthcoming.

6The theory of karman introduced a loose determinism as a pre-condition for all Indian schools, sothat the only option left was some sort of compatibilism and everyone automatically adjusted. MostIndian authors are in fact by default compatibilists. They assume automatically that we are determinedby karman, but also that karman is not an inextricable chain. If it were one, no liberation would bepossible (or, according to how strict one’s interpretation of karman is: no liberation would be possiblewithout a Divine intervention. Predestination is, accordingly, only admitted by the Theistic school ofDvaita Vedānta, founded by Madhva (see “Dvaita Vedānta” in this volume. See also, e.g., Zydenbos1991 and Pandurangi’s review of Zydenbos’ article, Pandurangi 2012.

5

doing. For instance, in a rather late compendium of Mīmāṃsā rules, the MNS, itis said that —notwithstanding a possible interpretation of a Vedic statement— “thebones [of a dead sacrificer] are not the agents, either on their own or through anarrangement, only living persons, who are (indirectly) indicated by them, are thesacrificers”. 7 Consequently, automatic movements are not enough to be an agent,whereas the same movements, if performed on purpose are the activity of an agent(on movement as different from action, see Freschi 2010). An agent may also notperform any movement at all and yet be called an agent, if she instigates anotherto undertake a certain action (see ŚV, Ātmavāda 75-87). Accordingly, both thesacrificer (yajamāna) and the officiating priest are said to be agents. 8

However, being conscious is not enough, insofar as one must also be able toperform the action, even if one does not actually perform it. This requisite hasan evident ritual origin, since lame and blind people, or animals cannot performaccurately the sacrifice and thus they cannot be ritual agents.9 It is, however, ap-plied also outside the ritual sphere, for instance when Kumārila argues against thepossibility of a God by saying that omnipotence (sarva-kartṛtva, literally ‘omni-agenthood’, i.e. agenthood directed to all possible items) would only be possiblefor an embodied agent, since without a body, one is not able to undertake all possi-ble actions.10

7MNS ad MS 10.2.17, Translation in MNS, p. 657. The Sanskrit text runs as follows: asthnāṃ svataḥsaṃvidhānena vākartṛtvāt tallakṣitā jīvanta eva yaṣṭāraḥ. A parallel passage of Pārthasārathi Miśra’sŚāstradīpikā has “It is never possible, either directly or through an arrangement, that a dead person isan agent” (na kathaṃ cid api mṛtasya sākṣād vā saṃvidhānād vā kartṛtvaṃ sambhavati, quoted in MNS,p. 239).

8See MNS 3.7.8, which depicts an objection claiming that the sacrificer should perform the sacrificehimself and a reply explaining that he might be an agent also if has priests officiate instead of him. InBenson’s translation: “[Obj.:] Because the agent and the enjoyer (of the result) are in syntactic agree-ment in the statement, ‘One desirous of heaven should sacrifice (yajeta)’, because the suffix of the middlevoice (i.e.m in the verb ‘yajeta’ (let him sacrifice)) is taught by smṛti (i.e., grammar) to be employedwhenthe result of an action accrues to the agent, and because the rite, together with its subsidiaries, producesthe result, the actions of sacrifice etc., together with their subsidiaries, should (all) be performed by thesacrificer. [Reply:] No. In as much as the action of hiring (the priests) would otherwise be inapplicable,the meaning of the text (śāstra [the Veda]) is only that at a rite which has subsidiaries, the condition of be-ing an agent is common to the direct agent and the instigating agent”, (MNS, pp. 446-7, square bracketsare mine). The Sanskrit text runs as follows: svargakāmo yajeteti kartur bhoktuś ca sāmānādhikaraṇyāt,kartṛgāmiphala ātmanepadasmaraṇāt, sāṅgasya phalajanakatvāc ca sāńgaṃ yāgādi yajamānena kāryam.na. parikrayānyathānupapattyā sāṅge sākṣātprayojakasādhāraṇakartṛtvasyaiva śāstrārthatvāt (the tran-scription has been slightly changed to fit the conventions of the present chapter). See also ŚV, Ātmavāda74. For the underlying Grammatical conception of agenthood, see Cardona’s contribution in this volume.

9The sūtra expressing this point is MS 6.1.2. The hint at animals is found in commentaries andMīmāṃsā textbooks, such as the MNS on 6.1.2.10But the notion of an embodied God is inherently contradictory (how could he be revered by different

people in different places simultaneously, if he were linked to a body?). On Kumārila’s arguments againstthe existence of God, see Krasser 1999.

6

3.3 Bhāṭṭa and Prābhākara Mīmāṃsā vs. Nyāya

The philosophical school Mīmāṃsā authors mostly refer to is the Nyāya school.11

Thus, a comparison of Mīmāṃsā and Nyāya views usually throws additional lighton both. Naiyāyika authors usually detect a sequence leading from cognition (e.g.,of the pleasant flavour of an apple) to volition (icchā) to the action (see Dasti’scontribution in this volume). Mīmāṃsā authors of the Bhāṭṭa school usually startwith desire (e.g. of eating something fresh and healthy), which leads to the effort(prayatna or pravṛtti, i.e., the fact that one undertakes the action of eating an apple).Prābhākara Mīmāṃsākas complicate the picture by adding the role of injunctions,so that one desires something, one consequently identifies as the one addressed by acertain injunction (e.g. “if you want to live long and healthy, eat apples every day”)and the fact that one feels enjoined leads to one’s effort.

Both schools of Mīmāṃsā tend not to focus on whatever happens after theeffort, i.e. on the actual performance of the action. In this sense, just like pre-contemporary Christian thinkers, they focus on “free will” and disregard the actualperformance of the action. Thus, Mīmāṃsakas would define free-will as just thecapacity to do what one wishes, independent of whether one actually does it or not.What one needs in order to be an agent is just the capacity to fulfil the action (seeabove, p. 6).

As for the cognitive element, previous cognition is indeed mentioned in theŚBh as preceding desire and, therefore, as an evidence of a self enduring from themoment of cognition to that of desire:

Through desire we grasp a self.

[Obj.:] How?

[Reply:] Because desire regards something coveted which has beenknown before. For instance, we do not desire those sweet tree-fruitswhich no human being has ever experienced, since they are [only] foundon the left side of the Mount Meru. Nor does one knower desire whathas been grasped by another person. […] Therefore we understand thatthis [desire] has the same agent as that grasping.12

Unlike Nyāya authors, however, Mīmāṃsā ones do not always stress the con-nection of desire with cognition. Furthermore, unlike in Nyāya, Prābhākaras deemdesire to be important only insofar as the subject identifies herself as the addresseeof a prescription insofar as her specific desire is mentioned in it:

It has been explained that the one who desires heaven, etc., has to beunderstood as the enjoined one, since duty is connected to himself [insentences such as “The one who desires heaven should sacrifice”]. And

11On the parallel and contrastive development of the two schools’ views on the key issue of LinguisticCommunication/Verbal Testimony, see Freschi and Graheli 2005.12icchayā ātmānam upalabhāmahe. katham iti. upalabdhapūrve hi abhiprete bhavatīcchā. yathā

merum uttareṇa yāni asmajjātīyaiḥ anupalabhapūrvāṇu svādūni vṛkṣaphalāni, na tāni prati asmākamicchā bhavati, […] tenopalambhanena samānakartṛkā sā ity avagacchāmaḥ (ŚBh ad 1.1.5, pp. 60-72).

7

it has been explained at the beginning of the eleventh [book of the MS]that the desire for heaven is a specification of the enjoined person.13

Thus, desire is only an indirect cause of action (the direct cause being the prescrip-tion).

To sum up, an agent is identified as one who undertakes an action, who could beable to perform it in full andwho does it consciously. At least in ordinary experience(outside it, see §5), she also needs to have a goal in view while undertaking theaction, as expressed in a well-known verse by Kumārila:

Without a goal, even the fool does not undertake any action.14

In short, according to Nyāya:

cognition → desire → undertaking of the action → action

According to the Bhāṭṭa Mīmāṃsā:

desire → undertaking of the action

According to the Prābhākara Mīmāṃsā:

desire → recognition of oneself as the addressee by the prescription

prescription → undertaking of the action

3.4 The role of commands

A border line problem, crossing over the boundaries described in section2 is that ofthe role of commands inMīmāṃsā, which can be seen as regarding one’s psycholog-ical experience of freedom and at the same time the agent’s causation. Commandsare usually neglected in all fields of Western philosophy,15 just like exhortative lan-guage is usually neglected in Western linguistics (see Freschi 2008, Freschi forth-coming a). By contrast, they play a fundamental role in Mīmāṃsā (see Freschi2007, Freschi 2012a, Freschi 2012b).

A command, Mīmāṃsā authors maintain, only provides for one to feel enjoined(niyujya). It does not make one necessarily perform what the command enjoinedthem. For an action to be undertaken and then performed, one still needs the wholeprocess described above, section 3.1, to take place. However, people who have beentrained (see below, section 4.3) to hold in high esteem a certain authority (notably,

13kāryasya svasambandhitayā bodhyaḥ svargakāmādir niyojya iti vyutpāditam. svargakāmanā ca niyo-jyaviśeṣaṇam ity ekādaśādye vyutpāditam (VM II ad 23, PrP, p. 440, Sarma 1990, p. 44). I could not findany passage at the beginning of ŚBh ad MS 11 which could be the one Śālikanātha has in mind. Thus,he probably refers to Prabhākara’s lost commentary thereon. See also TR IV, sections 10.2–10.3.2 inFreschi 2012b.14prayojanam anuddiśya na mando ’pi pravartate | (ŚV, sambandhākṣepaparihāra 55ab).15A notable exception is Bocheński 1974.

8

the one of the Veda), will rejoice at undertaking the actions enjoined (see TR §10.11and cf. the notion of ātmatuṣṭi, see fn. 32) and are thus very likely to undertakethem. (People for whom a certain text or person is not authoritative just do not feelenjoined by it or her and thus are irrelevant.)

Can one still speak of free will if one undertakes an action due to a prescription?Is there room left for free will, after having heard a Vedic command? Before an-swering, it is worth remembering that prescriptions only regard specific people, i.e.,the ones identified by the desire mentioned in the prescriptive sentence. Hence, aspecific prescription applies to one according to what she desires. Does it mean thatit applies according to who she is, in a deterministic way? The case of the Śyena,where people who desire to harm their enemies are blamed hints at the opposite.Such people are blamed because one can train oneself to desire something betterthan harming one’s enemy.16

Once one has been enjoined to do something, one is thus free not to perform theaction enjoined, but one is not free not to feel enjoined. Furthermore, one is freeto not to desire a certain goal, so that the prescription will not apply to them. Thisapplies to optional goals, such as sons, rain, cattle, villages. In exceptional cases(see below, §5), it seems to apply also to the general goal of happiness.17

3.5 Agent

Naiyāyikas tend to speculate on how agenthood can refer to an underlying self(called ātman). By contrast, Mīmāṃsakas generally discuss empirical agents andrefrain from discussions on the ātman, unless they need to polemically address theirBuddhist opponents (like in the ŚBh passage quoted in the preceding section). Evenwhen this is the case, most notably in Kumārila’s works, Mīmāṃsakas are keen toprove the existence of something closer to the Western common-sense notion of“subject” or “person” than to that of an underlying ātman. This means that whatthey envision is a subject which is intrinsically an agent, rather than a changelessātman to which agenthood is later attached as an adventitious quality. Kumārilaopposes the two aspects with the terms puruṣa ‘person’ and ātman ‘self’ in ŚV, Āt-mavāda 29.18 Kumārila is even ready to go as far as to risk denying the fixedness(nitya) of the self, in order to make room for the possibility of changes:

It is not prohibited to say that the self is not fixed (nitya) |

16One might suggest that these people are blamed although they cannot help, because of their nature,desiring to harm. However, this paradoxical interpretation (one can be blamed although one could nothave done otherwise) would not suit the legalistic attitude of Mīmāmṣā (about which, see for instanceLingat 1973). Similar claims have beenmade only within a strong theistic context (e.g. in some branchesof Lutheranism emphasising responsibility even over things one cannot control).17It is probably worth remembering that this degree of freedom only regards male Brahmins. Śūdras,

women, etc. undergo many more constraints due to their own nature, so that it is possible that theycannot freely train their desires in whatever direction. I am not overstressing this point here becauseit seems to be rather a socio-political issue, inherently linked with the history of classical India, ratherthan a specific feature of Mīmāṃsā philosophy.18For a similar terminological opposition in Prābhākara Mīmāṃsā, see Freschi 2012a.

9

if what one means is just that it can evolve (vikriyā), [since] there is nocessation of it by that ||

If there were an absolute destruction of it, there would be destructionof the actions performed and accrual of actions non-performed [hence,random connection of actions and results] |

but not in the case it reaches a different stage, like in the ordinary ex-perience of childhood, youth, [adulthood and old age] (when the sameperson is said to be a child, then a young person, then an adult and thenan elder person) ||19

His commentator Pārthasārathi explains further:

If only because it evolves [youwant to say that it is] not fixed,then let it be so (as you prefer)! In fact, there is no cessationof one’s own nature, since one recognises [one as the sameperson]. 20

Thus, writes Kumārila, if one were to deny the continuity of the self, one would beforced to conclude that the actions performed by a certain self would lead to a resultenjoined by a different one, and this is paradoxical and runs against common sense.By contrast, the self can evolve, just like a person changes from being a child toan adult. A Buddhist Pramāṇavādin might reply that it is exactly the case that theenjoyer is different than the doer. Against this view Kumārila replies that the selfis both constant and non-constant:

According to me, even when the person (nara) reaches a stage of plea-sure or pain |

it never relinquishes its nature of being consciousness, a substance andsomething which exists ||

[…]

Therefore, since both options (i.e., total fixedness and total destruc-tion) have been eliminated, a person must be accepted as consisting ofsomething which disappears and something which recurs, like gold inan ear-ring, etc. || 21

He further adds (possibly against Sāṅkhya and what came to be known as Ad-vaita Vedānta) that there is no intermediate entity that changes instead of the subject:

19(nānityaśabdavācyatvam ātmano vinivāryate | vikriyāmātravācitve na hy ucchedo ’sya tāvatā || syātāmatyantanāśe ’sya kṛtanāśākṛtāgamau | na tv avasthāntaraprāptau loke bālayuvādivat || ŚV, Ātmavāda22cd-23cd)20yadi vikāramātram anityaṃ tad astu, na hi vikāramātreṇa svarūpocchedo bhavati pratyabhijñānāt.

(Nyāyaratnākara ad loc.)21sukhaduḥkhādyavasthāś ca gacchann api naro mama | caitanyadravyasattādirūpaṃ naiva vimuñcati

|| […] tasmād ubhayahānena vyāvṛttyanugamātmakaḥ | puruṣo ’bhyupagantavyaḥ kuṇḍalādiṣu svarṇavat|| (ŚV, Ātmavāda, 26 and 28)

10

Nor is it the case that the condition of being an agent and that of beingan experiencer have as substrate the stages of a person [rather, theyhave as substrate the person himself/herself] |

therefore, the agent himself reaches his fruit, since he is the real pos-sessor of all stages ||22

Thus, Kumārila’s treatment suggests that he presupposes a notion of subjectas a knower and an agent, who is able to change through time. Then, one mightargue, what enables one to recognise her as a single subject? To this general prob-lem, the Nyāya answers by delineating an underlying character which does notchange, namely the ātman. This would remain stable, allowing accidental qual-ities to change without altering it. By contrast, Kumārila seems to imply that nosuch distinction makes sense, and that, on the contrary, there is no need to postulatea subtle, changeless ātman. The subject itself guarantees continuity, though in flux,and one can recognise the continuity as constituted by the changes, just in the caseof the subsequent stages of a river. The agent is, in other words, constant throughhis changes.23

Furthermore, Kumārila states that the subject is knowable through ahampratyaya‘notion of an I’, that is, the notion one has of oneself when one refers to oneself insentences like “I am going”, or “I am doing an effort”. This stance has been ad-versed exactly by the schools who consider the ātman to be an underlying entity, notto be confused with the empirical “I”. The fact that Kumārila and his commentatorsstick nonetheless at this evidence for its existence most probably means that theyimplicitly deny any sharp distinction between the empirical I and an underlyingātman.

Prābhākaras dispute this view out of a different standpoint, namely because theydeem it impossible for the subject to be at the same time the knower and the knownentity. They maintain instead that the subject is knowable in each act of cognitionsince all cognitive acts throw simultaneously light on themselves, on their contentand on their knower qua knower, i.e., as a subject and not as an object (see TR,chapter 1, partly translated in Freschi forthcoming b). Thus, Prābhākaras are evenstricter in underlying the active nature of the subject and in both sub-schools, onegrasps the subject while it is “doing” something. For this reason, other psychicorgans (such as buddhi or antaḥkaraṇa) are hardly mentioned at all (see Freschiforthcoming b). Although I am not aware of any explicit discussions about this inMīmāṃsā texts, the general attitude of Mīmāṃsā authors is to deal with a complexsubject, which acts (knowledge is included in action), desires and experiences. Itspossible subdivisions are irrelevant because the only entity one actually encountersis such a subject (and because, once again, Mīmāṃsakas are more interested inaccounting for our experience than in creating an alternative explanation of it).

22na ca kartṛtvabhoktṛtve puṃso ’vasthāsam āśrite | tenāvasthāvatas tattvāt karttaivāpnoti tatphalam ||(ŚV, Ātmavāda 29)23One is reminded of the Mīmāṃsā concept of kuṭastha- and pravāhanityatva. The first one is im-

mutable fixedness, the second one is fixedness through change. A typical example is that of word-referents, which remain the same though their individual tokens change.

11

Accordingly, discussions about the self in itself (i.e., independent of any actionit might be performing) are strikingly rare in Mīmāṃsā. So rare that a reader mightsuspect that Mīmāṃsā authors were not interested in investigating about an hypo-thetical ātman underlying all experiences, given that their focus was much more onthe active subject of all these experiences. In general, Mīmāṃsā arguments aboutthe ātman as a substance derive much from their Nyāya antecedents and are hardlyoriginal (see Freschi 2012a).24 A case in which the notion of ātman is discussedin Mīmāṃsā texts is the context of dharma and adharma (merit and demerit) ofeach individual. Mīmāṃsā authors are absolutely strict in refuting the Advaita po-sition of a single ātman, also because this would lead to a confusion of dharmaand adharma (see PrP, Tattvāloka, p.345). In one of the few extensive ontologicaldescriptions of the self I am aware of, i.e., the one found in Śālikanātha Miśra’sTattvāloka (PrP, p. 343), the ātman is said to be inseparably linked to its attributes(guṇa). These arise only when the ātman is in contact with the inner organ (manas).This, in turn, is only present in a body. Thus, although an ātman is theoreticallyconceivable also without a body, its attributes, the ones through which one graspsit, are only possible within a body. Mīmāṃsakas are quite clear in distinguishingthe subject from its external body,25 nonetheless the body understood as an or-ganism is said to be the only possible substrate of activities which are typical ofthe subject, such as being an agent (see above, section 3.2) and enjoying (see PrPTattvāloka p. 331). In TR 2 it is said that plants do not have bodies (śarīra) becausethey are not able to experience, and bodies are exactly defined as the instrument forrealising experience (bhogasādhana).26 Thus, the subject and the body are mostprobably mutually dependent: although a bodiless subject might be theoreticallyconceivable, in ordinary experience subjects can only be such insofar as they areembodied. Conversely, bodies are defined insofar as they enable one to experiencereality. Corpses are not bodies and dead subjects will probably need some newform of (subtle?) body.

4 Causality and determinism

4.1 Is there determinism in Mīmāṃsā?

From the literal point of view, one’s will can only be defined as “free” if there is thechance for it to be bound. Is this possibility ever mentioned by Indian authors? Asimilar hint may be detected in some theistic schools, where it is said that God aloneis free (he is, e.g., īśvara and svatantra), whereas we are all bound (i.e., paratantra

24For instance, a later Bhāṭṭa author, Gāgābhaṭṭa (16th c.), mentions the ātman in connection with hisargument that pleasure (sukha), cognition (jñāna) and desire (icchā) do not reside in the body, nor inthe sense organ, nor in the inner organ (manaḥ). Therefore, they must reside in the ātman, which isthus established (tasmāt tadāśrayatayā ātmasiddhiḥ) (Bhāṭṭacintāmaṇi, Gāgābhaṭṭa 1933 (repr.) Pp. 53-54.) Further psychic entities are not mentioned at all and the whole argumentation repeats a Naiyāyikascheme (see the Naiyāyika commentaries on NS 1.1.10).25See the discussion in ŚBh ad 1.1.5, Frauwallner 1968, p. 34 and then PrP Tattvāloka, p. 320 and

327-329.26This passage is discussed and partly translated in Freschi 2012a.

12

‘eteronomous’ or like ‘cattle’ paśu). Furthermore, onemight suggest that the karmanmight be thought to obstruct one’s will (this view can be found in dramas, cf. theattitude of minor characters in the Śākuntalā, see Nuckolls 1987). However, inphilosophical schools one does not encounter a deterministic view of karman beingexplicitly endorsed.27

Does this non-determinism interact with the Mīmāṃsā account of causality?Or does the non-determinism only refer to the moral level, whereas accounts ofcausality regard the ontological realm?

4.2 Causality

WithinMīmāṃsā, different accounts of causation intersect. One is the (almost) pan-Indian one based on karman. Another is the ritual paradigm, linked to the apūrvaone accumulates during rituals. This is a force ‘which was not there before [theritual]’, generated by the ritual and reaching until its result, thus explaining howthe ritual one performs today can be the cause of the village one will conquer insome weeks, etc. It is not clear whether the two paradigms might clash, for instance,if a person whose karman does not allow to have sons may perform a putrakāmeṣṭi(a sacrifice meant for people who want to have sons) and eventually get a son. Thetopic is not explicitly dealt with, as far as I know, in Mīmāṃsā.28 The general ideaseems to be that both paradigms express rules about the world and that the karman-paradigm rules the “natural” world whereas the apūrva-one rules the human world.In case of conflict, the apūrva one might be thought to overcome the other just likehuman causation overrides natural forces. But it could also be suggested that ifone can be the adhikārin ‘responsible agent’ for a certain sacrifice (i.e., if one hasenough wealth to perform it, is not disabled in any limb, etc.), this must mean thatone’s karman has lead him or her to this goal.29

The question as to whether deterministic causation (be it through karman orthrough apūrva or both) could ever hinder one’s free will seems to be resolved inMīmāṃsā through a compatibilistic approach. Compatibilism in a non-technicalsense is in fact one of the distinctive marks of Mīmāṃsā authors, who seem toprefer to adhere to our complex experience rather than to reduce it to some basicand consistent principles (see above, section 3.1). Mīmāṃsā thinkers describe theprocess of the agent’s causation as a complex process, inwhich a role is played by theagent’s initial desires and one by the exhortations they receive and accept because ofthese desires. That the latter role does not rule out the former is assumed as if it wereself-understood and is not the object of a separate treatment. The key point whichdoes not make the laws of karman end up in a rigid determinism probably lies in

27I am very much inclined to think that also the Buddhist Pramāṇavāda (notwithstanding what seems amechanicistic account of causality) is not deterministic, as shown by the fact that Dharmakīrti refutes thepossibility of inferring a result from its causes, so that even a karmic cause cannot be said to invariablylead to a certain result.28On causality within and outside Mīmāṃsā, see Halbfass 1980.29Furthermore, it might be argued that the theory of karman derives from some form of ritual causation

or has been influenced by it (Göhler 2011, p. 135), and this historical connection could partly explainwhy the two paradigms seem to be seen as compatible by Mīmāṃsā authors.

13

the fact that Mīmāṃsā authors seem to presuppose that one can influence her likesand dislikes. Since one can decide, for instance, not to want more cattle, or at leasttrain himself to want just a little bit more, he will not be liable to the authority of theprescriptions concerning the Citrā sacrifice. Similarly, onemight infer that the sameapplies in ordinary experience. Karman, most probably, inclines one towards doingsomething, but does not completely determine her to do it. Compatibilism wouldthen mean, within Mīmāṃsā, that the authority of the Veda over a ritual agent (and,similarly, of an adult over a child, a landlord over his servants, etc.) together withthe ritual agent’s past karman are not incompatible with her freedom to adhere tothe Veda (etc.), with adherence being the preferred option, the one one should longfor, since the Vedas tell us it is the best one. This precinct of freedom is stressed byPrabhākara in sentences such as

The [Vedic] injunction tells one what to do, it does not tell one that hehas to do it.30

4.3 Training desires

This again leads to a wider question, i.e., Are we free only when we act indepen-dently of desire? Or can one speak of free will also in the case of acts determinedby desire? In other words, is eating while hungry an instance of free will? Or isonly the whimsical movement of one’s arm with no exact reason a free act? Indianschools such as Mīmāṃsā have naturally acknowledged the role of desire. Accord-ingly, the implicit answer seems to be a stoic one, i.e., that one can train oneselfto desire the right things, since desire is part of one’s rational behaviour and notoutside it.More in detail, for Mīmāṃsā authors, desire is part of the natural worldwhich is controlled by the laws of karman and restricted by the Vedic injunctionsand prohibitions. An evidence of this attitude is found in discussions concerningthe parisaṅkhyāvidhi pañca pañcanākhā bhaktavyāḥ “the five five-nailed ones mustbe eaten”. The parisaṅkhyāvidhi is a prescription restricting something else. In thiscase, Mīmāṃsā authors explain, it restricts one’s natural appetite —which would bedirected to everything— to these five animals only:

Here it is not the eating which is prescribed, since it is already obtainedthrough one’s longing (rāga) for it, and an activity which has been ini-tiated prior [to the injunction] because it has been obtained due to de-sire (rāga) cannot be prescribed (since prescription only cause to knowsomething previously unknown).31

Thus, if we try to get a coherent picture out of these elements, karman lies at theorigin of the natural desires one experiences. Vedic prescriptions and prohibitions

30kartavyatāviṣayo niyogaḥ. na punaḥ kartavyatām āha (Bṛ, Bṛ, I/38,8f. For an insightful discussion ofthis topic, see Yoshimizu 1994.)31atra ca na bhakṣaṇaṃ vidheyam; rāgataḥ prāptatvāt. nāpi rāgaprāpteḥ pūrvapravṛttyā vidheyatvam

(Śambhubhaṭṭa’s Prabhāvalī on Khaṇḍadeva’s Bhāṭṭadīpikā 1.2.4., Khaṇḍadeva 1987, p. 33). The text hasbeen written in mid-17th c.

14

rule them and if one is addressed by a Vedic prescription or prohibition, she cannotbut feel enjoined by them. Consequently, she will restrict her natural desire to onlythe things which are permitted by the Veda. Here lies the precinct of application ofher free will, which is, however, lead by the Vedic word. Once again, free will inthis view is not tantamount to one’s whims, but rather to one’s faculty to consent ordissent to a Vedic word.

5 Source of ethics

Mīmāṃsā authors uphold the externality of moral authority. There is no innersource for morality, they maintain, because all human instruments of knowledgeultimately rest on sense perception and sense perception cannot fill in the gap be-tween the is and the ought (in Sanskrit terms, between what is siddha ‘already es-tablished’ and what is sādhya ‘to be brought about’). The only source for moralityis, hence, the Veda as instrument of knowledge. The Veda is the only way to knowabout what one ought to do. Other sources, such as smṛti (‘traditional texts’, such asthe Mānavadharmaśāstra and the other Dharmaśāstras) and sadācāra (‘behaviourof good people’) are valid insofar as they are based on the Veda. People who areversed in the Veda and who have studied the Mīmāṃsā rules to interpret it may endup developing an inner feeling for what ought to be done, as suggested in the Dhar-maśāstras32 and in Kumārila’s TV,33 but this is nothing other than an analogicalextension of the hermeneutic rules they learnt and hence no independent source.

The Veda is an instrument of knowledge about what one ought to do, becauseit prescribes something to be done instead of describing a state of affairs. It is thusa collection of commands. The Vedic commands are of two sorts. On the one handthere are commands which regard compulsory sacrifices, to be performed regularlythroughout one’s life (nitya) or on certain occasions (naimittika), such as the birthof a son. On the other, there are optional sacrifices (kāmya), dependent on one’swishes. One needs to perform an Agnihotra every day, whereas a Citrā is onlyperformed if one wants cattle. In the Mīmāṃsā classification, however, all sorts ofsacrifices refer to one’s desires. The first group applies throughout one’s life becausesuch sacrifices identify as the person who has to perform them “the one who desireshappiness”, and the desire for happiness accompanies one throughout one’s life.34

So farwith the “pure” (i.e., Vedic-exegetical)Mīmāṃsā core. However, Kumārila35

and most later Mīmāṃsakas seem to admit that it is theoretically possible to annihi-32See, e.g., Mānavadharmaśāstra, 2.6: vedo ’khilo dharmamūlaṃ smṛtiśīle ca tadvidām | ācāraś caiva

sādhūnām ātmanas tuṣṭir eva ca || 6 || Olivelle translates as follows: “The root of the Law is the entireVeda; the tradition and practice of those who know the Veda; the conduct of good people; and what ispleasing to oneself” (Olivelle 2004), but the context and the position of sādhūnāṃ suggests that also“what is pleasing to oneself” only regards good people who know the Veda.33TV ad 1.3.7, passim. On the topic of ātmatuṣṭi, the whole Davis 2007 is insightful and intriguing.

See also Francavilla 2006, pp. 165–176, which dissents with Davis as for the importance of ātmatuṣṭi,but agrees as for its being dependent on the authority of the Veda.34On the equivalence of svarga and happiness, see ŚBh ad 6.1.2. The goal of happiness identifies the

one who has to perform the sacrifice, and the MS and the ŚBh explain that happiness is everyone’s goal,see MS 4.3.15.35See Mesquita 1994 and the discussion in Taber 2007.

15

late one’s desires. At that point, one would no longer be under the influence of Vedicand non-Vedic commands, since commands depend on the identification of a desireof the person who should fulfil them (see Freschi 2007). One would be left with thepossibility to act focusing only on the actions, since one would no longer long fortheir results. This could be labelled freedom, although it is a very abstract degree offreedom, in which no room for acting whimsically is left. Later Mīmāṃsakas tendto interpret this stage as that in which one dedicates to God all actions, so that oneno longer performs them for the sake of their results, but rather for the supreme pur-pose of pleasing God (see the closing verses of Āpadeva’s Mīmāṃsānyāyaprakāśa,MNP). This would be a relative freedom (freedom from the common desires, butdependence on a higher-order one).36 It is difficult to establish whether the ac-ceptance of a desire-free action is a move of Kumārila, influenced by Vedānta (asmaintained in Mesquita 1994), or whether this possibility has always been heldin account by Mīmāṃsā authors. This issue has to do with the relation betweenMīmāṃsā and Vedānta and with the problem of their historical connection. It mightin fact be suggested (see Parpola 1981 and Parpola 1994) that the fact that these twoschools were known as Pūrva- and Uttara-Mīmāṃsā (see above, §1) points to a re-mote time, in which their basic texts formed a single sūtra, initially commentedupon as a whole and only later parted into two. If this were the case, then onecould easily imagine that the “Vedānta development” was in fact already part ofthe background of Mīmāṃsā authors, who however left its discussion to commen-tators on the Brahmasūtra (against the view of a single Mīmāṃsāsūtra, later splitinto Pūrva-Mīmāṃsāsūtra and Uttara-Mīmāṃsāsūtra, see Bronkhorst 2007).

5.1 Immoral sacrifices prescribed by the Veda

Given that the Veda is the only source of morality, one might expect the Veda toprescribe only rituals having legitimate outputs, such as sons or cattle, and to avoidprescribing rituals whose results run against the very Vedic rules. In fact, sincethe Veda (see above, section 5) is the only source of ethical norms, how can itprescribe to do something which it prohibits to do elsewhere? How can one decidethat what is prescribed by the Veda should not be performed, unless because a Vedicprescription tells us not to perform it? Andwould not this secondVedic prescriptioninvalidate the first? And if so, would not this mean that the Veda on its whole isno longer absolutely valid? Furthermore, the fact that undertaking a ritual actuallyprescribed entails a punishment leads one to question the origin of morality. If theVeda is the only source of morality, how could it be that something it prescribed isimmoral?

The problem is dealt with in regard to the Śyena sacrifice, which is performed inorder to annihilate one’s enemy (and about which, see Kataoka 2012). Now, if oneperforms it, one will obtain the annihilation of their enemy, while at the same timebeing guilty of violence, since one would have transgressed another Vedic com-mand, namely ‘one should not perform violence on any living being’ (na hiṃsyāt36The issue of theism in later Mīmāṃsā is quite complicated. Suffice it here to say that after the 13th

c. and especially in South India, Mīmāṃsā moves very close to Śrī Vaiṣṇavism (for further references,see Freschi 2012b, chapter 1.1.2.

16

sarvā bhūtāni). Thus, they will end up in hell. So far, the apūrva-causality seemsto work in a rather mechanical way. If for one the slaughter of her enemy is soimportant that she is ready to endure the consequences, she might perform it (andliterally go to hell as a consequence). Basically, the Veda tells one how to obtain theslaughter of her enemy, it is up to her whether to perform it or not. In each case, shewill endure the corresponding consequence. Thus, the Veda does not enjoin one toperform the Śyena (see Prabhākara’s quote in section 4.2). It only tells one how toperform it, in case he had freely chosen to perform an illicit (qua violent) act. Infact, only ‘one who is desirous to harm [one’s enemy]’ (hiṃsākāma) is a suitablecandidate for the performance of the sacrifice. She alone would be the addresseeof the prescription and have the status (adhikāra) required to perform the sacrifice.However, the condition of being desirous to harm one’s enemy is itself forbiddenby the prohibition to perform any sort of violence.

Consequently, and in accordance with the principle expressed in section 4.3,one might speculate that one’s general appetite (i.e., unspecified desire) to harmhis enemies falls within the restrictive power of the prohibition to perform anyviolence, so that one ought never get to the stage of becoming “one who desires toharm her enemy”. This also means that if one becomes one who desires to harmone’s enemy, one has already violated a Vedic prohibition and is therefore to beethically blamed.

6 Conclusions

The basic Mīmāṃsā approach to the issue of agency and free will is compatibilist,namely, the psychological experience of one’s freedom of action is asumed to bevalid, since one experiences one’s actions as free and since the karman- or apūrva-based causalities cannot be ascertained to eliminate all precincts of application offree will. In fact, human beings are lead to act, according to Bhāṭṭa authors, bytheir desires, and, according to Prābhākara authors, by injunctions which, in turn,identify them through their desires. Consequently, their precinct of free will seemsexactly to lie in one’s faculty to train their desires. Even from the point of view ofPrābhākaras, who stress the role of Vedic commands, free will is presupposed bythe claim that, although the Veda tells one what to do, it does not make one do it.

Agency does not accrue to an underlying ātman, but rather seems to constituteone of the subject’s essential characters. Accordingly, the agent subject is said notto be immutable and does instead change through time.

Abbreviations

Bṛ Prabhākara (1934-1967). Bṛhatī ad Śābarabhāṣya. Ed. by S.K. Rā-manātha Śāstrī and S. Subrahmanya Sastri. Madras: University ofMadras.

17

MNP Franklin Edgerton (1929). Mīmāṃsānyāyaprakāśa of Āpadeva. In-troduction, Sanskrit Text, English Translation and Notes. NewHaven: Yale University Press.

MNS Mahādeva Vedāntin (2010). Mīmāṃsānyāyasaṅgraha. A Com-pendium on the Principles of Mīmāṃsā, edited and translated byJames Benson. Ed. by James Benson. Wiesbaden: Harrassowitz.

MS, ŚBh, TV Kashinath Vasudev Abhyankar and Ganesasastri AmbadasaJośī, eds. (1971–1980). Śrīmajjaiminipraṇite Mīmāṃsā-darśane: Mīmāṃsakakaṇṭhīrava-Kumārilabhaṭṭapraṇita-Tantravārtikasahita-Śābarabhāsyopetaḥ. 2nd (1st ed. 1929-1934).Ānandāśramasaṃskṛtagranthāvaliḥ 97. Poona: Anandasrama.

PrP Śālikanātha Miśra (1961). Prakaraṇa Pañcikā of ŚālikanāthaMiśrawith the Nyāya-Siddhi of Jaipuri Nārāyaṇa Bhaṭṭa. Ed. by A. Sub-rahmanya Sastri. Darśana Series 4. Benares: Benares Hindu Uni-versity.

TR Rāmānujācārya (1956). Tantrarahasya. A Primer of PrābhākaraMīmāṃsā, critically ed. with Introduction and Appendices. Ed. byK.S. Rāmaswami Śāstri Śiromaṇi. Baroda: Oriental Institute.

TS Annambhaṭṭa (1979). Tarkasaṅgrahaḥ Annambhaṭṭaviracitaḥbālapriyāsahitaḥ [Bālapriyā commentary on Annambhaṭṭa’sTarkasaṅgrahadīpikā]. Ed. by N.S. Ramanuja Tatachariar. Hydear-abad: Prāptisthāna.

ŚV Kumārila and Pārthasārathi (1978). Ślokavārttika of Śrī KumārilaBhaṭṭa with the Commentary Nyāyaratnākara of Śrī PārthasārathiMiśra. Ed. by Dvārikādāsa Śāstrī. Prāchyabhārati Series 10.Varanasi.

ŚV Uṃveka Bhaṭṭa (1971). Ślokavārttikavyākhyatātparyaṭīkā. Ed. byS.K. Rāmanātha Śāstrī, rev. by K.Kunjunni Raja, and R.Thangaswamy. 2nd revised. Madras University Sanskrit Series 13.Madras.

References

Abhyankar, Kashinath Vasudev and Ganesasastri Ambadasa Jośī, eds.(1971–1980). Śrīmajjaiminipraṇite Mīmāṃsādarśane: Mīmāṃsakakaṇṭhīrava-Kumārilabhaṭṭapraṇita-Tantravārtikasahita-Śābarabhāsyopetaḥ. 2nd (1st ed.1929-1934). Ānandāśramasaṃskṛtagranthāvaliḥ 97. Poona: Anandasrama.

Annambhaṭṭa (1979). Tarkasaṅgrahaḥ Annambhaṭṭaviracitaḥ bālapriyāsahitaḥ[Bālapriyā commentary on Annambhaṭṭa’s Tarkasaṅgrahadīpikā]. Ed. by N.S.Ramanuja Tatachariar. Hydearabad: Prāptisthāna.

Bocheński, J.M. (1974). Was ist Autorität? Freiburg, Basel, Wien: Herder.Bronkhorst, Johannes (2007). “Vedānta as Mīmāṃsā”. In: Mīmāṃsā and Vedānta:

Interaction and Continuity. Ed. by Johannes Bronkhorst. Delhi: Motilal Banar-sidass, pp. 1–91.

18

— (forthcoming). “Free will and Indian philosophy”. In: Antiquorum Philosophia.Davis Donald R., Jr. (2007). “On ātmatuṣṭi as a source of dharma”. In: Journal of

the American Oriental Society 127.3, pp. 279–296.Earp, Brian D. (2011). “Do I Have More Free Will Than You Do? An Unexpected

Asymmetry in Intuitions About Personal Freedom”. In: The New School Psy-chology Bulletin 9.1, pp. 21–27.

Edgerton, Franklin (1929). Mīmāṃsānyāyaprakāśa of Āpadeva. Introduction, San-skrit Text, English Translation and Notes. New Haven: Yale University Press.

Francavilla, Domenico (2006). The Roots of Hindu Jurisprudence. Sources ofDharma and Interpretation in Mīmāṃsā and Dharmaśāstra. Vol. VII. CorpusIuris Sanscriticum. Torino: Cesmeo.

Frauwallner, Erich, ed. (1968). Materialien zur ältesten Erkenntnislehre der Kar-mamīmāṃsā. Kommission für Sprachen und Kulturen Süd- und Ostasiens;Philosophisch-Historische Klasse; Sitzungsberichte. Wien: ÖsterreichischeAkademie der Wissenschaften.

Freschi, Elisa (2007). “Desidero Ergo Sum: The Subject as the Desirous One inMīmāṃsā”. In: Rivista di Studi Orientali 80, pp. 51–61.

— (2008). “How do exhortative expressions work? Bhāvanā and vidhi in Rāmānu-jācārya and other Mīmāṃsā authors”. In: Rivista di Studi Orientali 81, pp. 149–185.

— (2010). “Indian Philosophers”. In: A Companion to the Philosophy of Action. Ed.by Timothy O’Connor and Constantine Sandis. Wiley-Blackwell, pp. 419–428.

— (2012a). “Action, Desire and Subjectivity in PrābhākaraMīmāṃsā”. In: Self andNo-Self: Hindu and Buddhist Ideas in Dialogue. Ed. by Irina Kuznetsova, Jonar-don Ganeri, and Chakravarti Ram-Prasad. Dialogues in South Asian Traditions:Religion, Philosophy, Literature, and History. Farnham: Ashgate.

— (2012b). Duty, language and exegesis in Prābhākara Mīmāṃsā: Includingan edition and translation of Rāmānujācārya’s Tantrarahasya, Śāstraprameya-pariccheda. Jerusalem Studies on the History of Religion 17. Leiden: Brill.

— (forthcoming a). “The Study of Indian Linguistics. Prescriptive Function of Lan-guage in the Nyāyamañjarī and in the Speech Act Theory”. In: Open Pages inSouth Asian Studies. Proceedings of the Workshop held in Moscow. 27–28th April2011. Ed. by Alexander A. Stolyarov. Leiden.

— (forthcoming b). “Does the subject have desires? The Prābhākara Mīmāṃsā an-swer”. In: Pushpika: Tracing Ancient India Through Text and Traditions. Con-tributions to Current Research in Indology. Number 2. Ed. by Giovanni Ciotti,Paolo Visigalli, and Alastair Gornall. Oxford: Oxbow Books Press.

Freschi, Elisa and Alessandro Graheli (2005). “Bhāṭṭamīmāṃsā and Nyāya on Vedaand Tradition”. In: Boundaries, Dynamics and Construction of Traditions inSouth Asia. Ed. by Federico Squarcini. Kykéion Studies and Texts I.3. Firenzeand New Delhi: Firenze University Press and Munshiram Manoharlal, pp. 287–323.

Gāgābhaṭṭa (1933 (repr.)). Bhāṭṭacintāmaṇi. Ed. by Sūryanārāyaṇa Śarmaśukla.Chowkhamba Sanskrit Series 25 and 27. Benares: Vidya Vilas Press.

19

Göhler, Lars (2011). Reflexion und Ritual in der Pūrvamīmāṃsā. Studie zur frühenGeschichte der Philosophie in Indien. Beiträge zur Indologie 44. Wiesbaden:Harrassowitz.

Halbfass, Wilhelm (1980). “Karma, apūrva and “natural” causes: Observations onthe growth and limits of the theory of saṃsāra”. In: Karma and Rebirth in Clas-sical Indian Traditions. Ed. by Wendy Doniger O’Flaherty. Berkeley, Los An-geles, London: University of California Press.

Kataoka, Kei (2012). “Is Killing Bad? Dispute on Animal Sacrifices between Bud-dhism and Mīmāṃsā”. In: Samskṛta-sādhutā ‘Goodness of Sanskrit’. Studies inHonour of Professor Ashok N. Aklujkar. Ed. by Yohichika Honda, Michele Des-marais, and Chikafumi Watanabe. New Delhi: D.K. Printworld, pp. 349–367.

Khaṇḍadeva (1987). Bhāṭṭadīpikā, with the commentary Prabhāvalī of ŚambhuBhaṭṭa. Ed. by Ananta Krishna Sastri. 2nd (1st ed. Bombay 1921-22). Sri GaribDass Oriental Series, no. 50. Delhi: Sri Satguru.

Krasser, Helmut (1999). “Dharmakīrti’s and Kumārila’s Refutations of the Exis-tence of God: A Consideration of Their Chronological Order”. In:Dharmakīrti’sThought and Its Impact on Indian and Tibetan Philosophy. Ed. by ShoryuKatsura. Wien: Verlag der Österreichischen Akademie der Wissenschaften,pp. 215–223.

Kumārila and Pārthasārathi (1978). Ślokavārttika of Śrī Kumārila Bhaṭṭa with theCommentary Nyāyaratnākara of Śrī Pārthasārathi Miśra. Ed. by DvārikādāsaŚāstrī. Prāchyabhārati Series 10. Varanasi.

Lingat, Robert (1973). The classical law of India (translated with additions by J.Duncan M. Derrett). Berkeley: University of California Press.

Mahādeva Vedāntin (2010). Mīmāṃsānyāyasaṅgraha. A Compendium on the Prin-ciples of Mīmāṃsā, edited and translated by James Benson. Ed. by James Ben-son. Wiesbaden: Harrassowitz.

Mesquita, Roque (1994). “Die Idee der Erlösung bei Kumārilabhaṭṭa”. In: WienerZeitschrift für die Kunde Süd-Asiens 38, pp. 451–484.

Nuckolls, Charles W. (1987). “Causal Thinking in Śākuntala: A schema-theoreticapproach to a classical Sanskrit drama”. In: Philosophy East and West 37.3,pp. 286–305.

Olivelle, Patrick (2004). Manu’s Code of Law. A Critical Edition and Translation ofthe Mānava Dharmaśāstra. Oxford; New York: Oxford University Press.

Pandurangi, Veeranarayana (2012). “A Review of “Jaina Background of DvaitaVedānta” by Robert Zydenbos”. In: Saṃskṛta Vimarśaḥ 6 Special issue releasedduring the WSC, New Delhi, pp. 280–318.

Parpola, Asko (1981). “On the Formation of the Mīmāṃsā and the Problems Con-cerning Jaimini with particular reference to the teacher quotations and the Vedicschools (Part I)”. In:Wiener Zeitschrift für die Kunde Südasiens 25, pp. 145–177.

— (1994). “On the Formation of the Mīmāṃsā and the Problems Concerning Jai-mini with particular reference to the teacher quotations and the Vedic schools(Part II)”. In: Wiener Zeitschrift für die Kunde Südasiens XXXVIII, pp. 293–308.

Prabhākara (1934-1967). Bṛhatī ad Śābarabhāṣya. Ed. by S.K. Rāmanātha Śāstrīand S. Subrahmanya Sastri. Madras: University of Madras.

20

Rāmānujācārya (1956). Tantrarahasya. A Primer of PrābhākaraMīmāṃsā, criticallyed. with Introduction and Appendices. Ed. by K.S. Rāmaswami Śāstri Śiromaṇi.Baroda: Oriental Institute.

Śālikanātha Miśra (1961). Prakaraṇa Pañcikā of ŚālikanāthaMiśra with the Nyāya-Siddhi of Jaipuri Nārāyaṇa Bhaṭṭa. Ed. by A. Subrahmanya Sastri. Darśana Se-ries 4. Benares: Benares Hindu University.

Sarma, Rajendra Nath (1990). Verbal Knowledge in Prābhākara Mīmāṃsā. SriGarib Das oriental series 60. Including the text of Śālikanātha Miśra’sVākyārthamātṛkā. Delhi: Sri Satguru.

Strawson, Galen (1998). “Free will”. In: Routledge Encyclopedia of Philosophy. Ed.by Edward Craig. Vol. 3. London: Routledge, pp. 743–753.

Strawson, Peter Frederick (1962). “Freedom and Resentment”. In: Proceedings ofthe British Academy 48, pp. 1–25.

Taber, John (2005). A Hindū critique of Buddhist epistemology: Kumārila onperception: the ”Determinatin of perception” chapter of Kumārila Bhaṭṭa’sŚlokavārttika / translation and commentary. RoutledgeCurzon Hindū StudiesSeries. London and New York: RoutledgeCurzon.

— (2007). “Kumārila the Vedāntin?” In: Vedānta and Mīmāṃsā: Interaction andContinuity. Ed. by Johannes Bronkhorst. Delhi: Motilal Banarsidass.

Uṃveka Bhaṭṭa (1971). Ślokavārttikavyākhyatātparyaṭīkā. Ed. by S.K. RāmanāthaŚāstrī, rev. by K.Kunjunni Raja, and R. Thangaswamy. 2nd revised. MadrasUniversity Sanskrit Series 13. Madras.

Wegner, Daniel M. (2002). The Illusion of Conscious Will. Bradford Books. Cam-bridge MA: Massachusetts Institute of Technology.

Yoshimizu, Kiyotaka (1994). “Der ‘Organismus’ des geheiligten Veda. Eine Studieder Niyoga-Lehre Prabhākaras mit ausgewählten Übersetzungen der Bṛhatī”.PhD thesis. Universität Wien.

Zydenbos, Robert (1991). “On the Jaina background of Dvaita Vedānta”. In: Journalof Indian Philosophy 19, pp. 249–271.

21