Shabdabodha and the Machine: Mimamsa, Nyaya and Vyakarana Theories of Sentence Meaning as a Critical Framework for Large Language Models
Abstract: The rapid rise of large language models has revived an old question in a new form: what does it mean to understand a sentence? The Indian Knowledge System (IKS) possesses one of the most sustained traditions of reflection on this question, developed over nearly two millennia by the Mimamsa, Nyaya and Vyakarana schools under the name shabdabodha, the cognition that arises from words. This paper undertakes a textual and comparative study of these theories and examines their relevance for evaluating machine-generated language. The Mimamsa definition of the sentence, the debate between abhihitanvayavada and anvitabhidhanavada, the holistic doctrine of Bhartrhari, the Nyaya conditions of akanksha, yogyata, asatti and tatparya, and the classical account of how word-meaning is acquired are analysed from primary Sanskrit sources. It is argued that the contrast between static word embeddings and contextual transformer representations closely parallels the contrast between the Bhatta and Prabhakara theories of sentence meaning, and that the distributional learning of language models corresponds to only two of the eight classical means of grasping word-meaning. The central thesis of the paper is that machine-generated text can satisfy the formal conditions of shabdabodha while failing the conditions of shabda-pramana, because the Nyaya criterion of the apta, the trustworthy speaker, is not met. On this basis a four-level diagnostic framework and the concept of superimposed intention (aropita-tatparya) are proposed. The study concludes that IKS offers not a nostalgic parallel but a rigorous evaluative vocabulary for the age of artificial intelligence.
Keywords: Indian Knowledge System, Shabdabodha, Mimamsa, Nyaya, Bhartrhari, Testimony, Large Language Models, Sanskrit Semantics
How to Cite?: Dr. Devendra Singh, "Shabdabodha and the Machine: Mimamsa, Nyaya and Vyakarana Theories of Sentence Meaning as a Critical Framework for Large Language Models", Volume 15 Issue 9, September 2026, International Journal of Science and Research (IJSR), Pages: 1621-1626, https://www.ijsr.net/getabstract.php?paperid=SR26923215908, DOI: https://dx.doi.org/10.21275/SR26923215908