engleski

Śaṅkara’s Authorship of the Chāndogyopaniṣad- Bhāṣya

This presentation will discuss the application of modern authorship verification methods on Sanskrit text. An appropriate example that can be used for explicating this problem can be the commentary on the Chāndogya-Upaniṣad traditionally attributed to Śaṅkara. Paul Hacker’s traditional philological methods of determining Śaṅkara’s authorship will be addressed only briefly as the presentation will be focused on a computational experiment based on new stylometric developments in authorship verification, recently introduced General Imposters (GI) framework. The presentation will explain how the Sanskrit text should be prepared and how the GI procedure works. The problem with Sanskrit texts is that sandhi has to be manually resolved in order to disjoint the words ; the GI algorithm is namely based on the frequencies of some text constituents, usually words. However, the algorithm also works with n-grams for which one must not go through time- consuming manual resolving of sandhi. Afterwards it will be reported which way of text preparation yields best results. In the next step, GI procedure will be applied to two Śaṅkara’s texts: the commentary on Śvetāśvatara-Upaniṣad for which Śaṅkara’s authorship is seriously disputed and then to the commentary on Chāndogya-Upaniṣad that has never been a subject of critical examination. Both texts will be compared to a set of “imposters”, texts that do not belong to Śaṅkara, and furthermore to a set of his genuine works. GI algorithm will determine whether the two texts in question are closer to the set of Śaṅkara’s genuine works than to the set of “imposters”.

stylometry, advaita, authorship, imposters, authorship verification, Upaniṣads, commentaries

nije evidentirano