Dear fellow readers,
Hi. For my second blog post as the guest host for March 2021, I want to provide a digest version of my recent article in JHS titled: Commands and the Doctrine of the Apūrva in Prābhākara Mīmāṃsā, so that one doesn’t have to slog through a long article written for specialists to access and evaluate my interpretations. I herein condense 30 pages to 5.
In the article’s first section, I argue that Prabhākara (fl. c. 660-680 CE?) subordinately incorporates Śabara’s earlier model of the command’s computation, and thereby subverts Śabara’s theory of human motivational behaviour whereupon motivation exclusively derives from a calculation of self-interest (i.e., from desire for a result and knowledge of the means to the result). I explore two accounts that Prabhākara offers for how Śabara’s theory is subordinately incorporated: the command’s entailment function (upādāna) and Prabhākara’s theory of linguistic signification (anvitābhidhāna).
The framework within which this debate occurs is a perception model of language (i.e., how the language-listener computes meaning), specifically of the archetypal Vedic injunction: svargakāmo yajeta (‘the man desiring heaven should sacrifice’).
I’ll first briefly explain Śabara’s model, so we can then turn to how and why Prabhākara subverts it. Connected with Śabara’s theory of human motivational behaviour is his position that the phonetic form: svargakāmo yajeta has the semantic structure: yāgena svargam bhāvayet (‘one should bring heaven into being through sacrifice’). All Mīmāṃsakas agree that verbal suffixes are the structurally predominant items in relation to the verbal roots, but for Śabara specifically, the finite verbal suffix –eta in yajeta directly signifies what he calls the bhāvanā (lit: ‘the bringing into being,’ best understood as intentional action to produce an outcome). This bhāvanā as the governing linguistic item generates syntactico-semantic expectancies (ākāṅkṣās) for its necessary relata/thematic roles, first for an outcome, and then for an instrument. Heaven (expressed by the nominal stem svarga-) and sacrifice (expressed by the verbal root yaj-) construe as the outcome and instrument respectively, even though we don’t get case-suffixes attached to the linguistic items svarga- and yaj- to express these thematic roles (no svargam [accusative case marker] and no yāgena [instrumental case marker] in svargakāmo yajeta). Rather, the syntactico-semantic expectancies of the governing bhāvanā lead the language-listener to compute the phonetic form as having this semantic structure as its directly signified meaning (for Śabara there is no sentence-meaning of svargakāmo yajeta other than this: yāgena svargam bhāvayet is how the term-referents floating around in mind construe). The payout of Śabara’s model is that the language-listener first computes a cause-and-effect relationship between sacrifice and heaven, and human motivation occurs subsequently from self-interest for a result when one has knowledge of the means (sacrifice) to the result (heaven). See Cummins 2021 pp. 292-295 for further elaboration and translations of relevant passages.
intervention is tied up in his introduction of the new the category of the commandee
(niyojya). Prabhākara uses this category to create a two-layer/two-moment
computational model whereupon the initial impulse to act occurs at the
first step of the computational process, prior to knowledge of sacrifice as
a means to heaven which is computed at layer/moment two. This intervention
allows Prabhākara to move Śabara’s semantic structure: ‘one should bring heaven
into being via sacrifice’ (yāgena svargam bhāvayet) to a subsequent
second moment & second implicit layer of meaning within the listener’s computational
process, and thereby subvert Śabara’s position that humans are solely motivated
to act out of self-interest.
We can begin to unpack Prabhākara’s intervention by way of his justification for the computational primacy of the relationship between command (niyoga) and commandee (niyojya). Prabhākara’s justification is a circularity problem between action and agent in the context of the command: the idea is that the very existence of an action (i.e., its performance) depends upon an agent to perform the action, but no one can ever be an agent as such prior to the action’s actual commencement. Thus, were the command to pick out and command an ‘agent’ to do a sacrifice, commands would fail to function: no one could recognize oneself as the person to whom the command pertains as they’re not yet an agent, hence the need for a linguistic category beyond that of the agent (kartṛ), the commandee (niyojya), to account for how one recognizes oneself as the person to whom the command pertains prior to the action happening.
At this point, I want to pause and highlight the difference between justification and motivation. These Sanskrit intellectuals are never motivated to hold complex positions because of something such as a circularity problem. Rather these are their justifications, and this needs to be differentiated from their motivations, i.e., the payout of their interventions, about whichthey are frequently silent or opaque. The overarching goal of Prabhākara’s sub-commentary on Śabara’s Bhāṣya is to square the topics of Mīmāṃsā with his countertheory of the command (niyoga). To think of Prabhākara’s introduction of this countertheory as motivated by a circularity problem would be deeply mistaken: the real motivation is a divergent theory of human motivational behaviour. With this in sight, let us turn back to unpacking Prabhākara’s theory of the command (niyoga).
For Prabhākara, the term ‘man-desiring-heaven’ (svargakāmo) communicates the commandee (niyojya). Recall that for Śabara, ‘heaven’ (svarga-) construes as the outcome of the causal action (the bhāvanā). Not so for Prabhākara. For Prabhākara, the verbal suffix –eta signifies the command (niyoga), and the command needs its necessary relatum, the commandee. The term ‘man-desiring-heaven’ (svargakāmo) fills this conceptual slot. The crucial detail to understanding Prabhākara’s intervention here is that the term: ‘man-desiring-heaven,’ whereby one recognizes oneself as the commandee, does not specify an outcome from the fulfillment of the command. The term ‘man-desiring-heaven’ functions exactly the same way as the terms ‘short’ and ‘tall’ if I say: ‘short people go to the left; tall people go to the right’ à no outcome or reward is explicitly specified. On Prabhākara’s model, all that happens at this first moment of the computational process is that one recognizes oneself as someone to whom the command to perform a sacrifice pertains, by way of recognizing oneself as someone who has a desire for heaven, and feels the initial impulse to act.
Prabhākara’s introduction of the circularity problem (explained above) necessitates the computational primacy of this layer of meaning between command (niyoga) and commandee (niyojya). Prabhākara’s rationale is that humans would never recognize commands as pertaining to themselveswithout the linguistic category of commandee as distinct from the category of agent (kartṛ), and the introduction of this category comes part-n-parcel with the removal of an outcome being explicitly specified from fulfilling the command. This is the real payout of Prabhākara’s intervention.
Thus, for Prabhākara, humans have a propensity to obey. We’re not quite Pavlov’s dogs, but we are still somewhere on the spectrum. Yoshimizu is right: this is not a Kantian categorical imperative. Put slightly differently, Prabhākara’s intervention is not about putting morality back into the performance of the Vedic sacrifices (even though for Śabara the model surrounding the performance of the Vedic sacrifices is mechanistic and amoral), but it still appears that for Prabhākara there is a want for the Vedic rites & Veda to be something more than purely mechanistic. This is, perhaps, part of the reason behind why being a Prābhākara became so attractive and why the textual tradition gained such momentum, despite the messiness and difficulty of Prabhākara’s own treatise(s).
With Prabhākara’s intervention in sight, I want to turn to how Prabhākara subordinately incorporates Śabara’s theory. Recall that for Prabhākara, the language-listener’s computation of svargakāmo yajeta is a two-step sequential process: the listener first computes the relationship between command and commandee as the explicit meaning (and explicit meaning always comes first), and then on this two-step model, Śabara’s one-step model – of cause-and-effect between sacrifice and heaven – is incorporated as the subsequent and implicit layer of meaning.
Prabhākara offers two accounts for how this is done via both the command’s entailment function (upādāna) and via his new theory of linguistic signification (anvitābhidhāna). Both accounts hinge upon the logic of the Veda as non-misleading. For Prabhākara, the Veda is explicit: ‘the man desiring heaven must sacrifice,’ and if sacrifice were not the cause of heaven this would be misleading the man desiring heaven, and so we necessarily have the implicit computation of sacrifice as the cause of heaven.
Prabhākara’s apparatus of upādāna was first studied in monograph-length treatment by Yoshimizu (1997) in German. Yoshimizu translates upādāna as ‘organismus’ (= incorporation), and this is correct. The reason I translate upādāna as: ‘the command’s entailment function’ is that this is the definition Prabhākara himself offers (niyogākṣepa upādānam: incorporation is the command’s entailment), and I think this translation offers insight for the non-specialist into how the apparatus functions. My interpretation differs from Yoshimizu’s in that I believe Prabhākara takes conscious steps to subordinately incorporate Śabara’s computational model and thereby subvert Śabara’s theory of human motivational behaviour (i.e., I am offering a new account of Prabhākara’s engagement with Śabara, the predecessor whose treatise Prabhākara writes his own treatise as a commentary upon). I believe this is all crucial to pin precisely, as on my read Prabhākara’s engagement with Śabara constitutes the first major point of engagement across the aisle in what becomes a major debate on human motivational behaviour within the framework of a perception model of commands (Maṇḍanamiśra then inverts Prabhākara’s two-step model in the Bhāvanāviveka).
Prabhākara’s account of how Śabara’s theory is subordinated via the command’s entailment function (upādāna) consists of two connected moves:
i) The rejection that the Veda enjoins sacrifice as a means to heaven via an explicit subordinate injunction (viniyoga).
ii) The position that sacrifice as the means to heaven is enjoined implicitly through the command’s entailment function.
Prabhākara’s two positions here on subordinate injunctions (viniyoga) and the command’s entailment function (upādāna) in relation to the computation of sacrifice as the cause of heaven are two sides of the same coin. Suppose you believe X is a duck, and I believe X is a goose. I then make two arguments, one to reject why X cannot be a duck and one to argue why X is a goose. This is what Prabhākara does: on 2.1.1-4 he says sacrifice as a means to heaven cannot be explicit meaning (as per viniyoga)and on 4.1.2 he says it is implicit meaning (as per upādāna).
i) In commenting on MS 2.1.1-4 – the exact same place where Śabara outlines his theory explored earlier – Prabhākara makes the antithetical argument that a cause-and-effect relationship between sacrifice and heaven is not explicit meaning in the archetypal Vedic command.
Prabhākara carefully side-steps direct engagement with Śabara’s core concept of ‘bringing into being’ (bhāvanā) and the key cognitive function through which it operates, syntactico-semantic expectancies (ākāṅkṣās), and introduces to the topic the question of a subordinate injunction (viniyoga). The key point for our purposes of understanding Prabhākara’s motivation for doing this is that all the subordinate injunctive elements constituting the hierarchical list, from explicit statement (śruti) to context (prakaraṇa), are considered explicit meaning. Thus, by making the argument that there is no subordinate injunction (viniyoga) to perform sacrifice for the sake of heaven, Prabhākara finds a compelling way to reject Śabara’s position that the Veda explicitly communicates sacrifice as a means to heaven (and thereby open up a space for human motivation here to not be preceded by knowledge of a means to an end), while side-stepping direct engagement with Śabara’s conceptual apparatus. See Cummins 2021 pp. 298-301 for further explanation and a translation of Prabhākara on the relevant passage.
ii) In discussing the command’s entailment function (upādāna), Prabhākara is crystal clear that this is the mechanism by which the language-listener implicitly and subsequently understands sacrifice as a means to heaven (see Cummins 2021 p. 304 for a translation of the relevant passage). Prabhākara speaks very minimally on this point, perhaps because of what he is doing to Śabara’s model.
For a monograph-length study of upādāna and surrounding matters in German, see Yoshimizu 1997.
For a much briefer general discussion of upādāna in English, see Cummins 2021 pp. 302-4.
Let us now turn to Prabhākara’s second account for how the cause-n-effect relationship between sacrifice and heaven is computed, as per his theory of linguistic signification, anvitābhidhāna. This requires a brief detour to first table Prabhākara’s linguistic theory (the following ‘detour’ corresponds to Cummins 2021 pp. 306-7):
The earlier and standard Mīmāṃsā doctrine of Śabara (early 500s CE?) was that terms function exclusively to present to the listener’s mind the word-meaning, not the sentence-meaning (Śabara puts the burden for the latter on the cognitive functions of the listener). Śabara toes this line to defend the position that terms only directly signify eternal universals such as class-properties, behind which lies the defense of the Veda as eternal, while also accounting for language’s infinite productivity.
Enter Bhartṛhari, the grammarian-philosopher for whom ultimate reality consists in non-duality. An axiom of Bhartṛhari’s thought is that language parallels reality, and in the way that phenomenological duality is ultimately unreal, in the same way there is no reality to the division of the utterance into words and word-meanings.
Prabhākara articulates his Mīmāṃsaka theory of ‘the direct signification function [of terms and term-referents] as syntactically co-construed’ (anvitābhidhāna) as a concessionary position to Bhartṛhari (see Cummins 2021 p. 306 for evidence that Prabhākara is engaging Bhartṛhari specifically). On Prabhākara’s theory, terms have no signification function to communicate term-referents in isolation, but only as syntactically construed with other term-referents. As Prabhākara says, you can’t even articulate a term in the nominative case in Sanskrit (e.g., ghaṭaḥ ‘a pot’) without it entailing a copulative verb ‘exists’ (asti). In taking this position, Prabhākara cedes Śabara’s earlier position that words only present to mind word-meaning (i.e., Prabhākara concedes to Bhartṛhari that terms and term-referents have no function/reality in isolation), but against Bhartṛhari toes the line of compositionality à that the complex expression (the sentence) is determined by the meaning of its real constituent parts (the words), even though you can never use the constituent parts (words) apart from as sentences (Prabhākara toes this line to uphold the deeper doctrine of the Veda’s authority). Prabhākara grounds his defense of compositionality in a quasi-behaviourist model of language acquisition, where the child observes variable responses to variable stimuli (e.g., ‘bring the horse’ vs. ‘bring the cow’).
With his new linguistic theory in sight, let us now turn back to how Prabhākara accounts for the subordinate incorporation of Śabara’s model as per his new linguistic theory. I translate the key passage of Prabhākara on 6.1.1-3. Prabhākara’s constructed interlocutor has now conceded that the commandee (niyojya) is linguistically encoded in svargakāmo yajeta, but raises the problem of how to then account for sacrifice as the means to heaven:
[Objection:] What term is the grounds for [understanding] sacrifice as the means to the accomplishment of the desire on the part of the man being commanded to action? The root √yaj in yajeta, first off, merely communicates sacrifice, but not [that sacrifice] possesses a relation with something else. The verbal suffix too [only] communicates the agent [or] the agent’s number, but not that sacrifice [is] for the sake of something else.
On this we say:
[Therefore], the term (pada) is able to be called the grounds [for understanding sacrifice as a means to heaven, and] as for what you’ve said that there wouldn’t be ascertainment as to which term communicates [sacrifice as a means to heaven], that is entirely rejected by anvitābhidhāna.
[Objection:] If so, then the [topic at hand] has been resolved entirely on 1.1.24-25.
[No.] All that’s been rejected [there] is that the understanding term and term-referent is not fixed. That topic opposes an illusory or exclusionary on the substitution [principle]. However, it hasn’t [been rejected on 1.1.24-25] that terms [produce] only the awareness of [their corresponding] term-referents. Therefore, it is right: ‘one should bring heaven into being by sacrifice’ (yāgena svargaṃ sādhayet).
Prabhākara’s answer here can be unpacked to a point. Prabhākara here says that on 1.1.24-25, he defended a principle of compositionality, but that now extra non-compositional meaning can also be accounted for as per his linguistic theory of anvitābhidhāna.
Alas, this is all Prabhākara says.
However, in saying this, Prabhākara is clear that this ‘extra’ meaning of sacrifice as the means to heaven is not explicit and is computed after the explicit first step of the computational process whereby the listener understands the relationship between command and commandee (as Prabhākara says, there is no specific term to communicate this extra meaning of sacrifice as, but since we do understand it, it must be built in there … somehow).
This is perfectly consistent with what we already saw Prabhākara say about the command’s entailment function à the two positions are coherent, in that they both subordinately incorporate Śabara’s layer of meaning as the second moment/layer of meaning within the computation process.
On the one hand, Prabhākara does give a very compelling account of what svargakāmo yajeta actually means. I would translate this literally as: ‘the man who desires heaven should sacrifice.’ This sentence does not seem to actually explicitly encode what Śabara wants it to, that sacrifice is the means to heaven (yāgena svargam bhāvayet), but rather implies it.
However, it is also worth noting that Prabhākara’s two-step model requires either:
i) multiple moments or layers within the computational process.
ii) a pre-syntax function of terms to present a term-meaning to mind.
And this is in tension with his position on 1.1.24-25. His concessionary position to Bhartṛhari, of anvitābhidhāna – the direct signification function of terms and term-referents as syntactically co-construed – requires syntactic construal of all term-referents, i.e., sentence-meaning, to be a one-step simultaneous process.
This tension is perhaps why Prabhākara himself never says anything about syntactic construal as a simultaneous one-step process when talking about his new linguistic theory.
Śālikanātha, in fact, is the first Prābhākara to explicitly say syntactic construal of all term-referents is a simultaneous one-step process in the Vākyārthamātṛkā I, but he has a solution.
Maybe more on Śālikanātha next week …
Apologies for the long blog. I condense 30 pages of article to 5.