Aesthetics of Sanskrit Poetry from the Perspective of Computational Linguistics: A Case Study Analysis on Siksastaka

Jivnesh Sandhan, Amruta Barbadikar, Malay Maity, Pavankumar Satuluri, Tushar Sandhan, Ravi M. Gupta, Pawan Goyal, Laxmidhar BeheraarXiv 2023 · DOI · arXiv

Abstract

This research presents a novel, interpretable framework for the computational analysis of Sanskrit poetic aesthetics. By integrating the ancient principles of Kāvyasāstra with modern natural language processing (NLP) techniques, the authors develop a hierarchical ensembling approach. This framework is applied to a case study of the Śikṣāṣṭaka, demonstrating how traditional aesthetic schools can be formalized into computable modules. The results highlight the potential for automated systems to evaluate poetic quality through a structured, multi-layered analysis of sentiment, style, and suggestion.

1. Introduction

The intersection of Natural Language Processing (NLP) and literary analysis has given rise to the field of computational aesthetics. While machines have become proficient at generating text, understanding the underlying reasons for a poem's beauty remains a significant challenge. This is particularly true for Sanskrit poetry, which is governed by the complex and sophisticated theoretical framework of Kāvyasāstra.

The authors identify four primary obstacles to the computational study of Sanskrit poetics: the inherent subjectivity of aesthetic value, the extensive use of figurative language, the deep cultural and religious context required for interpretation, and the scarcity of large-scale, annotated datasets. This paper proposes a roadmap to address these hurdles by formalizing ancient poetic theories into an interpretable computational framework.

2. Theoretical Foundations of Sanskrit Poetics

Sanskrit literary theory is built upon several major schools of thought, each identifying a different core element as the "soul" of poetry. These schools provide a structured taxonomy for what might otherwise be considered purely subjective experiences of beauty.

Major schools of Sanskrit poetics (Kāvyasāstra) and their foundational treatises.

As detailed in Figure 1, these schools include Rasa (emotional sentiment), Alaṅkāra (figurative ornaments), Rīti (poetic style), Dhvani (suggestion), Vakrokti (oblique expression), and Aucitya (propriety). The proposed framework leverages these categories to create a modular analysis system.

3. Methodology: Hierarchical Ensembling Framework

The researchers introduce a seven-module hierarchical ensembling framework. These modules are divided based on their computational complexity:

3.1 Deterministic Modules

These modules handle objective, rule-based features of the text.

  • Meter (Chandas): Analyzes syllable patterns (short vs. long) to identify poetic meters.
  • Style (Rīti): Evaluates poetic style based on compound density and phonetic complexity.

3.2 Learning-Based and Expert Modules

These modules address deeper semantic and aesthetic layers, often requiring expert annotation for training.

  • Alaṅkāra: Identifies sound-based and meaning-based figures of speech.
  • Rasa: Determines the dominant emotional mood or sentiment.
  • Vakrokti: Analyzes indirect or "crooked" speech that contributes to poetic flavor.
  • Dhvani: Evaluates the "suggestive" meaning that transcends the literal text.
  • Aucitya: Serves as a meta-classifier or "balancer" that learns the appropriate weights for the other modules to ensure harmonious integration.

4. Case Study: The Śikṣāṣṭaka

To validate the framework, the authors conducted a deep analysis of the Śikṣāṣṭaka, an eight-stanza Sanskrit poem. The study utilized a human-in-the-loop approach, where four experts with advanced degrees in Sanskrit provided detailed annotations. These annotations included compound types, dependency parsing, prose order (anvaya), and specific aesthetic features like rasa and alaṅkāra.

5. Results and Discussion

The framework successfully categorized the Śikṣāṣṭaka as Uttama Kāvya (fine or superior poetry). Specifically, the analysis identified the poem's effective use of Vipralambha-śṛṅgāra (the mood of separation) supported by the Vaidarbhī style, which favors soft consonants and short compounds.

Key insights from the study include:

  • Formalization of Aesthetics: The research demonstrates that "beauty" in Sanskrit poetry can be mapped to a structured taxonomy, such as the Dhvani school's classification of poetry into Uttama (high), Madhyama (medium), and Adhama (low) levels.
  • Computational Feasibility: The authors distinguish between currently computable tasks (like meter identification) and ongoing semantic challenges (like metaphor detection), providing a clear research trajectory.
  • Interactive Tools: A web-based application was developed to visualize these poetic layers, making complex traditions accessible to a broader audience.

6. Limitations and Future Work

The primary limitation of the current work is its scale, as it focuses on a single case study. Expanding the framework to larger corpora will require significant efforts in data labeling. Furthermore, modules like Dhvani and Vakrokti still rely heavily on human expertise. Future research directions include the use of cross-lingual models and cognitive studies (e.g., EEG) to align physiological human responses with the framework's aesthetic classifications.

Conclusion

By bridging ancient Indian poetics with modern computational linguistics, this research provides a robust foundation for the automated analysis of Sanskrit literature. The hierarchical ensembling framework offers a path toward understanding the "why" behind poetic beauty, moving the field closer to a truly interpretable computational aesthetics.

Summary updated

Abstract

Sanskrit poetry has played a significant role in shaping the literary and cultural landscape of the Indian subcontinent for centuries. However, not much attention has been devoted to uncovering the hidden beauty of Sanskrit poetry in computational linguistics. This article explores the intersection of Sanskrit poetry and computational linguistics by proposing a roadmap of an interpretable framework to analyze and classify the qualities and characteristics of fine Sanskrit poetry. We discuss the rich tradition of Sanskrit poetry and the significance of computational linguistics in automatically identifying the characteristics of fine poetry. The proposed framework involves a human-in-the-loop approach that combines deterministic aspects delegated to machines and deep semantics left to human experts. We provide a deep analysis of Siksastaka, a Sanskrit poem, from the perspective of 6 prominent kavyashastra schools, to illustrate the proposed framework. Additionally, we provide compound, dependency, anvaya (prose order linearised form), meter, rasa (mood), alankar (figure of speech), and riti (writing style) annotations for Siksastaka and a web application to illustrate the poem's analysis and annotations. Our key contributions include the proposed framework, the analysis of Siksastaka, the annotations and the web application for future research. Link for interactive analysis: https://sanskritshala.github.io/shikshastakam/
BibTeX
@misc{sandhan2023aesthetics,
  author = {Jivnesh Sandhan and Amruta Barbadikar and Malay Maity and Pavankumar Satuluri and Tushar Sandhan and Ravi M. Gupta and Pawan Goyal and Laxmidhar Behera},
  title = {{Aesthetics of Sanskrit Poetry from the Perspective of Computational Linguistics: A Case Study Analysis on Siksastaka}},
  year = {2023},
  eprint = {2308.07081},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2308.07081},
}
Figure