|Mukesh Singh's illustration of the Kauravas from 18 Days.|
After arguing against an objector who argues that motivating speech (arthavādas) are not meaningful (anarthakya), Kumārila turns away from explanation of motivating speech in the Vedas to passages in the Mahābhārata and the Purāṇas.
He says (and here I follow Jha 25-26 with some changes, Sanskrit is from SARIT):
The above explanations apply also to passages occurring in the Mahābhārata and Purāṇas, etc. For, as well, in these cases, we have the express injunction: "One should recite these before persons of the four castes," which shows that they are the means of accomplishing certain desirable ends; and when we proceed to seek for this desirable end, we do not accept the mere recitation of the words as bringing about any result; and find that the true result lies in a full comprehension of the causes of dharma, artha, kāma, and mokṣa, along with their opposites, adharma, anartha, duḥkha, and saṃsāra, for the purpose of acquiring those of the former set and avoiding those of the latter.He has argued above that, through postulation (arthāpatti) which functions as a last interpretive recourse when other epistemic instruments have failed, we are to understand motivating speech as praising or denigrating certain methods, so as to compel people to act according to Vedic commands. But here, he applies this approach to a command given by Vyāsa in the MBh, in the Śāntiparvan (12.314.45), where he commands that the story he has just narrated should be recited in front of all four castes. This is a command, and as such, it, like all other commands, has a result desirable to people. But that result isn't just hearing the recitation (an observation which will be important in the next section on mantras), it is understanding the meaning of the text itself.
एवं भारतादिवाक्यानि व्याख्येयानि । तेषामपि हि श्रावयेच्चतुरो वर्णानित्येवमादिविध्यनुसारेण पुरुषार्थत्वान्वेषणादक्षरादि व्यतिक्रम्य धर्मार्थकाममोक्षाधर्मानर्थदुःखसंसारसाध्यसाधनप्रतिपत्तिरुपादानपरित्यागाङ्गभूता फलम् ।
And in certain portions of these works, as in the chapters on dāna, rājadharma, and mokṣadharma, of the Śāntiparva of the Mahābhārata, we meet even with direct injunctions, while in others there are arthavāda-s, in two forms, what has been done for someone else (parakṛta) and stories of the distant past (purākalpa) And for the other portions, those containing descriptions of events and stories, if we accept these in their literal sense (tātparya), then, with reference to these at least, the injunction of reciting them would become useless, and hence we take these to indirectly imply praise or denigration. And as these descriptions have been inserted with the sole purpose (tātparya) of such praise or denigration, one should not be primarily focused on their truth (tattva).
तत्रापि तु दानराजमोक्षधर्मादिषु केचित्साक्षाद्विधयः केचित्पुनः परकृतिपुराकल्परूपेणार्थवादाः । सर्वोपाख्यानेषु च तात्पर्ये सति श्रावयेदिति विधेरानर्थक्यात्कथंचिद्गम्यमानस्तुतिनिन्दापरिग्रहः । तत्परत्वाच्च नातीवोपाख्यानेषु तत्त्वाभिनिवेशः कार्यः ।In this section, Kumārila argues that if we look at the Mahābhārata, we can find that it contains the same types of expressions that he's been discussing in the context of the Vedas. There are direct injunctions (sākṣāt vidhi-s) as well as motivating speech (arthavāda-s). What motivates us in the Mahābhārata? Well, the fact that other people, (it is implied they are good people), have done certain things, and actions also from good people which happened a long time ago. We are inspired by good and heroic deeds of ancestors. However, these deeds do not actually have to have happened for us to be inspired, according to Kumārila.
In a statement which might scandalize some today, Kumārila says that some parts of the text involves stories whose literal sense or ordinary purpose (tātparya) would be useless (anarthakya) and so we should understand them as indirectly praising or denigrating some action. He doesn't identify which parts of the text, but he specifies that these are cases where the descriptions of what is going on do not need to be entirely correct or at least we shouldn't be primarily focused on their truth (tattva). In other words, it is no harm to the ethical motivating ability of the Mahābhārata that some parts of it are fiction! (This echoes what he has said elsewhere about the Vedas, too--not that they are fiction, but that they can employ hyperbole and metaphor to motivate.)
Why would upright people like Vālmīki (the author of the Mahabhārata) and Vyāsa (its narrator) use such means? Why not write completely true stories to inspire people? Kumārila says,
Since they were guided by their study of the Veda (vedaprasthānābhyāsena), Vālmīki, Vyāsa, and others composed their works on the same lines as the Veda. And, as those for whom these works were intended were persons of varying mental abilities, it was only proper for them to insert every kind of matter in their works. Now, some people are taught just by injunctions. But for others, they are taught by injunctions mixed with motivating speech, some which are short and some which are lengthy. This approach is so that everyone's minds would be gripped.
वेदप्रस्थानाभ्यासेन हि वाल्मीकिद्वैपायनप्रभृति भिस्तथैव स्ववाक्यानि प्रणीतानि । प्रतिपाद्यानां च विचित्रबुद्धित्वाद्युक्तमेवैतत् । इह केचिद्विधिमात्रेण प्रतिपद्यन्ते । अपरे सार्थवादेनापरेऽल्पेनार्थवादेनापरे महता । सर्वेषां च चित्तं ग्रहीतव्यमित्येवमारम्भः ।He does not say it explicitly, but we can surmise that maybe he thinks the people who have lower intelligence who might be attracted to the fictional parts. (Edited to add: This is of course speculation, and is only based on the strongly rational tendencies of Mīmāṃsā. Jha, perhaps to blunt the force of this implication, translates vicitrabuddhi as "varying degrees of intelligence and diverse tastes," but I don't see that the idea of "taste" is necessarily present here.) Now, the author of the Mahābhārata is working with a Vedic template, but is also drawing on their ordinary experiences.
Some of these injunctions and prohibitions are grounded in the Veda (śruti), some are about wealth and happiness, based upon considerations of ordinary pleasure and pain, grounded in ordinary experience (loka). Similarly, among the motivating speech, too, some are Vedic (vaidika), some are ordinary (laukika), while some are made-up (racita) by the author himself in accordance with the logic of poetry (kāvyanyāya). However, all of these are authoritative, since they are for the purpose of praise.
तत्र तु केचिद्विधिप्रतिषेधाः श्रुतिमूलाः केचिदर्थसुखादिषु लोकमूलास्तथाऽर्थवादाः केचिद्वैदिका एव केचिल्लौकिका एव केचित्तु स्वयमेव काव्यन्यायेन रचिताः । सर्वे च स्तुत्यर्थेन प्रमाणम् ।Kumārila identifies a variety of grounds for the composition of the Mahābhārata, which includes the Vedas, but does not exclude the possibility of something made-up or constructed (racita) by the author himself. Again, the status of something as fiction, or at least as not the whole and complete truth does not mean it cannot be authoritative (pramāṇa) in the realm of ethics. He concludes with some remarks about the parts of them which can't be related meaningfully to any Vedic injunction:
As for those portions however, which are not capable of being taken along with any injunction, some of them bring about pleasure just by themselves, in their being heard, like the the descriptions of the Gandhamādana mountain range and so on. And some, such as the descriptions of wars, serve to embolden everyone, the brave as well as the coward, and thereby serve a distinctly useful purpose for the kings of men. In those cases, however, where none of these is possible, such for instance as the hymns to deities, which we do not find capable of bringing about any perceptible result, we assume (kalpanīyam) an unseen result (apūrva).
ये तु वाक्यशेषत्वं न प्रतिपद्यन्ते तेऽपि केचित्स्वयमेव श्रूयमाणगन्धमादनादिवर्णकप्रभृतयः प्रीतिं जनयन्ति । ये तु युद्धवर्णकास्ते सर्वेषां शूराणां भीरूणां चोत्साहकराः पार्थिवानामुपयुज्यन्ते । यत्र तु न किंचिद्दृष्टमुपलभ्यते तत्र विशिष्टदेवतादिस्तुतिद्वारमदृष्टं कल्पनीयमित्येषा दिक् ।Jha, perhaps relying on a later commentator, takes him to be changing topics to the Purāṇas here, but it seems that his remarks could apply to both texts. Not all parts are, like the Vedas, aiming at the knowledge of dharma. Because their authors are following the Vedas, naturally, these texts will have ethical impact and serve purposes. But Kumārila admits that some parts are there just for being pleasant, like descriptions of nature. Some purposes are purely military or political. And he leaves room for the possibility of unseen result (apūrva) arising from hymns, for instance, which do not immediately lead to visible effects, but which we might postulate need to have some effect, since they are not there for any of the other reasons.
I should note in closing that as I was finishing this blog post, I encountered Anand Venkatkrishnan's dissertation which discusses this passage in some detail. He has his own translation of the text which readers should look at (80-81) in a discussion of itihāsapurāṇa literature more broadly in the Mīmāṃsā context.