This article uses Mīmāṃsā tools for a non-exegetical purpose, i.e., in order to build a system of representations of imperatives. The purpose of the system is even further away, since it regards Artificial Intelligence. In fact, the authors start with the (evident, but often overlooked) observation that we communicate with computers mostly through imperatives and not assertive statements (do this, then if x, then do z…). I am not at all interested in Artificial Intelligence, but it is easy to see how computer programming has forced logicians to take into account variables which had been neglected for centuries, such as the temporal value of truth (a≠b and a=b can both be true, if only one adds the temporal variable t1 and t2). In this case, it has stimulated Bama Srinivasan and Ranjani Parthasarathi to look for a logical solution to the representation of imperatives.
This goal also entails that Mīmāṃsā is used instrumentally. The authors have most probably not dealt with Sanskrit sources and do not seem to be very familiar with secondary sources, either, Sanskrit words are used, but often imprecisely (be it because of typos, mukya for mukhya or pravrithi for pravṛtti and so on, or because of wrong translations, such as the rendering of bhāvanā as “goal” or “motivation”). More importantly, Mīmāṃsā is dealt with as if there were no differences among Mīmāṃsakas and as if all its exegetical rules had been expressed on the same level and in the same form as Srinivasan and Parthasarathi reproduce them (i.e., as no exegesis had been needed). In other words, (although the authors might not be aware of it) the historical Mīmāṃsā is a source of inspiration for a re-shaped Mīmāṃsā1, which they then use as the foundation of their representation of imperatives.
At this point, one might ask why I am reading it and what good such articles can be for scholars of Sanskrit. Well, the article is more than intriguing, insofar as it reduces the Mīmāṃsā deontics to a few essential basics. It calls us to action, insofar as we can either accept this representation or attempt a better one, but cannot just refute it (as this would be unfair). Further, the idea of formalising the representations of imperatives in Mīmāṃsā could be quite useful, also for exegetical purposes. For instance, it can be used to point out why Mīmāṃsakas cared about a certain conflict among rules or why they accepted a certain passage as an imperative or not. Yes, life is more complicated than one hopes and Srinivasan and Parthasarathi have made Mīmāṃsā look like an open reservoir of rules just ready to be used. But more historically-aware scholars can take advantage of their work, even of its bold generalisations, and try to look through the Mīmāṃsā texts for their underlying logical structure.
Have you ever attempted to rewrite a Sanskrit text using logical notation?
(Friday is my day for broad readings. You can find my weekly and monthly blogging plan here).
the paper looks like it has very little to do with mīmāṃsā, as it seems to be a modification to an earlier framework proposed by vranas. while i definitely appreciate the attempt to bring indian theories into contemporary discussions of the philosophy of language and its practical implications, and the idea that mīmāṃsā’s principles are in fact capable of formalization, i have to agree with you: this doesn’t get scholars of mīmāṃsā (or philosophers, for that matter) very far in understanding the major themes and controversies in the system. more specifically, there are lots of things that their formalization can’t capture: the way that mīmāṃsā evaluates injunctions, for instance, whether cooccurence counts as conjunction or exclusive disjunction, often depends on the position that the injunction occupies within a hierarchy of injunctions, or on the actual content of the injunction (which the calculus cannot access, since it treats injunctions in the matter of propositional logic rather than predicate logic).
Andrew, thanks for this comment. I agree that I did not learn anything about Mīmāṃsā through this article. I also wonder whether the same article could help logicians to understand deontic issues better (although I am sure that it helped its authors looking at deontics in a fresh way). Still, I like the fact that they tried, and I see that it is a hard path. In this sense, may I ask you what do you think that a predicate logic approach could achieve in this connection?
well, i am just spitballing, but say we represented an injunction as a ternary function B(g,m,p) [bhāvanā with a goal, means, and procedure]. then we might be able to relate, for instance, the goal of one injunction to the means or procedure of another injunction (thus forming relations of implications between injunctions), or we might be able to interpret B1 B2 differently according to whether they share g or not, etc.
Andrew, that’s a nice idea. Let us start with the idea that —according to Pūrva Mīmāṃsā— Vedic prescriptions require a goal which MUST be something desirable. Vedic prescriptions also require an instrument and a procedure (which is defined as what is instrumental for the instrument). Accordingly, hierarchies of prescriptions can be created through the preceding elements. For instance, prescription A can relate directly to the desired goal, whereas prescription B may have as goal the realisation of the instrument needed in A, prescription C may have as goal the realisation of the instrument of B and so on. It follows the need to take into account multiple variables to account for this chains of prescription (since the prescription B would violate the principe of the need of a desirable goal) if it were not for its link with A.
Anyway, I am still at the very beginning, as you see…
Thanks Dr.Freshci for the comments of our work. As rightly pointed, what we have taken from Mīmāṃsā is just tip of the iceberg. And in that paper, not much have been mentioned about Mīmāṃsā, since the focus was directed towards computational aspects. We are presently working on other concepts, particularly from injunctions and shall post an article soon expanding our core ideas, realting to the same work on MIRA.
Andrew’s idea is interesting! But the predicate logic holds for the values of true or false. For the ternary definition of bhāvanā B(g,m,p), are means (m) and procedures (p) injunctions? Is `g’, a `goal’ or `an intention to achieve the goal’? If so, as per predicate logic, we cannot give the values of true or false for bhāvanā. An example from Mīmāṃsā would help to understand better and we could take the formalization further.
By the way, I found the title of the blog interesting 🙂 I’ve started studying sanskrit very recently, but have always felt that Indian philosophy has lots to offer. There are some good translations available in English and other Indian languages. So, IMHO sanskrit is not a pre-requisite to read and understand Indian Philosophy. Having said that, my Ph.D supervisor and colleague Dr. Ranjani Parthasarathi is well versed in Sanskrit and she often corrects me with the way I understand the translations.
Thanks a lot for this interesting comment. I am not sure I agree that Sanskrit is not a pre-requisite, although I agree that it should not. The translations I have been dealing with are often either word-to-word translations which cannot be understood independently of the Sanskrit text (say, R.N. Sarma’s translations of Śalikanātha), or loose commentaries (say, Sandal’s translation of the Mīmāṃsā Sūtras). In both cases, much is lost of the depth of the original —furthermore, only a very minor segment of Sanskrit texts have been at all translated and even less with a philosophical approach. A good reason to work more in teams!
I’ve uploaded my dissertation on Mīmāṃsā Inspired Representation of Actions (MIRA) in https://annauniversity.academia.edu/BamaSrinivasan/Dissertation. Chapter 3 would give an idea of Mīmāṃsā concepts that have been looked at from the logical/computational perspective. Comments and suggestions are most welcome.
Thank you! I will let you know as soon as I manage to read it (and hopefully understand it).