I am currently working with some amazing colleagues at the Vienna University of Technology on the formalisation of Mīmāṃsā deontic logic (for further information, read this post). One of the problems we are facing is that duties prescribed in Vedic prescriptions appear to be interpreted as regarding only specific eligible people, the adhikārins. For instance, one needs to perform a Kārīrī sacrifice if one desires rain, so that the duty to perform it does not apply generally to all. Even in the case of a sacrifice one has to perform throughout one’s life, such as the Agnihotra, the same restriction applies, since Mīmāṃsā authors interpret it as meaning that one has to perform it if one desires happiness, i.e., throughout one’s life, since one always desires happiness.

This led us (in fact, Björn) to apply to our case a binary version of the “Ought” operator, e.g.

O (A/B)

which can be read as “One ought to do A, if one is in the situation B”.

Now, the problem is: How to interpret the “/B” part?
As far as I can see, there are at least three possible solutions:

• Read it as meaning “provided you desire that…”. This will in fact cover most cases. However, it will not cover instances such as the distinction between the application of, e.g., “Don’t tell lies” (nānṛtaṃ vadet) to the person of the sacrificer, or to the ritual context of the Darśapūrṇamāsa only (see ŚBh ad 3.4.12–13). One might try to construe also these cases as meaning, respectively, “provided any possible desire” and “provided that you desire x” (with x=the same desire prompting the Darśapūrṇamāsa). But this forces a little bit the framework of desire, since desiring that x, does not immediately mean performing the corresponding sacrifice. Moreover, in this way the distinction between prescriptions applying to the person and to the action is blurred.
• Read it as meaning “provided the situation x”, with this situation being in the case of the Kārīrī “the desire of rain”, in the case of prescriptions regarding the sacrificer as person “T” (=any possible situation) and in yet the case of the prohibition to tell lies “during the DPM”. A problem with this interpretation is that it is difficult not to interpret
  

  O (not to tell lies/DPM)

  as a subset of

  O (not to tell lies/T)

  which means that we could never have a case in which something is prohibited (or prescribed) for T but there are particular prescriptions enjoining exceptions (one can immediately recall the problem of violence and the fact that it is generally prohibited, but prescribed in the case of the Agnīṣomīya). A further problem is that the different applicability of the prohibition to tell lies (etc.) are not described by Śabara as referring to a more or less extended portion of the same range of applicability, but rather as referring to two different sets (in one case: the person of the sacrificer, in the other: the action itself).

• Read it as meaning different things. In one case “/B” would be interpreted as meaning one’s desire, in others as referring to either a specific ritual context or the generality of one’s being a person. But is such a different interpretation of the same operator justifiable?

For further information on our project on deontic logic and on its participants can be read here. A slight change has occurred, however, insofar as Francesco Genco has joined the group, and we are not working on linguistic issues for the time being.
Thanks to Bama Srinivasan, who signalled me a missing sentence in the opening paragraph.