A Word-Level Digital Reader of the Prasthānatrayī with Śaṅkara’s Bhāṣya: Corpus, Method, and an Open Offline Reading Aid for Advaita Vedānta Canon

Jul 16, 2026

Upload by Asher

Prasthānatrayī—the ten principal Upaniṣads, the Brahmasūtra, and the Bhagavadgītā—together with Śaṅkara’s bhāṣya forms the core corpus of Advaita Vedānta, yet it is difficult for learners because sandhi, long samāsa compounds, and dense scholastic prose obscure word boundaries and grammatical meanings. This work provides an open, fully offline, word-level HTML reader for the entire canon, enabling clickable word analyses (padaccheda, morphology, syntax, gloss) across both root text and commentary. With lemma-indexed concordance search, an evaluation and expert review loop validate accuracy at 99%+ agreement for high-confidence forms, suitable as a freely redistributable teaching aid.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages12

Reading Time30 min

Abstract

Introduction

arXiv :2607 .07282v 1 [ cs .CL] 8 Jul 2026 A Word-Level Digital Reader of the Prasthānatrayī with Śaṅkara’s Bhāṣya: Corpus, Method, and an Open, Offline Reading Aid for the Advaita Vedānta Canon Tamal Maharaj* Department of Computer Science Ramakrishna Mission Vivekananda Educational and Research Institute Belur Math, Howrah, West Bengal, India July 2026 Abstract The Prasthānatrayī—the ten principal Upaniṣads, the Brahmasūtra, and the Bhagavadgītā, together with Śaṅkara’s commentaries (bhāṣya)—is the foundational corpus of Advaita Vedānta. For the non-specialist, and even for the student, the text is difficult to read at the word level: continuous euphonic combination (sandhi), long nominal compounds (samāsa), and dense scholastic prose obscure where one word ends and the next begins, and what each word means grammatically. We present an open, fully offline, word-level digital reader of the entire Prasthānatrayī with Śaṅkara’s bhāṣya. In the reader, every word—both of the root text (mūla) and of the commentary—is clickable and resolves to a pop-up giving its euphonic split (padaccheda), morphological and syntactic analysis, and gloss. Because every word carries a lemma, the reader doubles as a concordance: a search on a dictionary headword retrieves all of that word’s inflected and sandhi-hidden occurrences—and its occurrences inside compounds—across both layers. The resource covers thirteen commentarial units (2,971 verses, sūtras, and prose sections; 36,881 analysed word-occurrences of root text) and a global dictionary of 95,587 distinct commentarial surface forms. We describe the corpus, the hybrid analysis pipeline—a rule-based sandhi-viccheda engine layered over an authoritative inflected-form lexicon and attested-corpus look-ups, with large-language-model (LLM)-assisted word analysis subjected to an adversarial two-pass verification protocol—and a durable human-expert review loop that lets a Sanskritist correct machine output once and have the correction survive all subsequent regenerations. An intrinsic evaluation against independent Sanskrit resources finds that high-confidence analyses agree with an authoritative inflectional lexicon on over 99 % of attested forms, and an independent, band-blind adjudication confirms that analysis quality degrades predictably across the confidence bands, with errors concentrated in the low-confidence tier that thereview loop targets. The reader is a single, self-contained HTML file requiring no server, no installation, and no network access, and is offered as a freely redistributable teaching and reading aid. Keywords: Advaita Vedānta; Śaṅkara; Prasthānatrayī; Upaniṣad; digital humanities; computational Sanskrit; sandhi-viccheda; morphological analysis; digital philology; open educational resources. * Corresponding author: [tamal@gm.rkmvu.ac.in](tamal@gm.rkmvu.ac.in. A live)[. A live](tamal@gm.rkmvu.ac.in. A live) instance of the reader is publicly available at [https:](https:)//[cs.rkmvu.ac.in/](cs.rkmvu.ac.in/) ~tamal/prasthanatrayi/prasthanatrayi-sankara-bhasya-detail.html 1 Introduction The three prasthānas or “points of departure” of Vedānta—the śruti-prasthāna (the principal Upaniṣads), the nyāya-prasthāna (the Brahmasūtra of Bādarāyaṇa), and the smṛti-prasthāna (the Bhagavadgītā)—constitute the canon on which every classical Vedāntic school grounds its authority. Śaṅkara’s commentaries (bhāṣya) on these three are, for the Advaita tradition, the paradigmatic exposition; they are read, memorised, and debated to this day. Yet the corpus is hard to read at the granularity a learner needs. Classical Sanskrit is written in continuous euphonic combination: adjacent words fuse at their boundaries by regular phonological rules (sandhi), and nominal compounds (samāsa) can run to many members without internal spacing. A single orthographic “word” on the page, e.g. धूमेनाɟव्रयते, may in fact be two or three words (dhūmena āvriyate), and Śaṅkara’s prose—syntac

FAQs

What problem does the word-level digital reader address in the Prasthānatrayī and Śaṅkara’s bhāṣya?

It tackles the learner difficulty caused by sandhi, long nominal compounds, and dense prose that hides word boundaries and grammatical roles. The reader makes each apparent “word” clickable and recoverable at true word granularity.

What happens when a user clicks a word in the reader?

A pop-up provides its euphonic split (padaccheda), morphological and syntactic analysis, and an English gloss. This applies to both the root text (mūla) and the commentary.

How does the reader function as a concordance?

Every word is linked to a lemma, so searching a headword retrieves all inflected and sandhi-hidden occurrences. A split-aware search also finds occurrences inside compounds across the root text and the full commentary.

Does the resource require installation or internet access?

No. The reader is packaged as a single self-contained HTML file that runs offline without installation and without network access.

Jul 16, 2026

Upload by Asher