In this dissertation I have attempted to document and analyze the interaction of three domains within Sanskrit literature: grammar, poetry, and their exegesis. Each of these played an important role in determining what correct Sanskrit was, although no final and long-lasting consensus was ever reached, and multiple varying opinions are represented in the corpus I have examined. Unlike in the European grammatical tradition, which emerged in Hellenistic Alexandria (3rd–2nd cent. BCE) with a focus on editing and explaining a relatively fixed body of (poetic) texts, the first extant and most important Sanskrit grammar, Pāṇini’s Aṣṭādhyāyī (The Treatise in Eight Lessons), predates the start of classical Sanskrit poetry (2nd–4th cent. CE) by some six centuries. Moreover, Pāṇini’s grammar mainly sought to describe the idealized spoken language of an elite class of Brahmans, at least according to Patañjali (turn of the CE), an early and highly revered commentator on the Aṣṭādhyāyī. Although classical Sanskrit poetry, broadly speaking, adheres to the linguistic standards given by Pāṇini, a number of apparently aberrant usages can be found, and later commentators on both grammatical as well as poetic texts sought to evaluate them, taking, in most cases, Pāṇini’s grammar as the definitive linguistic authority. This may, from the Western vantage point, appear to be a fairly black-and-white process: if the grammar provides for a form, it is correct; if the grammar forbids a form, it is incorrect. In Sanskrit, however, there was much gray area because the exact interpretation of the Aṣṭādhyāyī, i.e., what it taught, was continually debated, and different exegetical standards came into, and fell out of, vogue over the centuries. The reason for this variability lies in the grammar’s style of composition. Pāṇini gives some four thousand rules, by means of which one can derive correct forms from stems and suffixes; the correct forms themselves are usually not specified. Already in the first grammatical commentaries, the exact meaning of the rules were disputed, and a complex system of interpretation developed, which included the discovery of “hidden hints” in Pāṇini’s text. These, in turn, were to inform us of further metarules, which, in many cases, accounted for forms apparently neglected by Pāṇini but considered part of the target language. Consequently, when later commentators came to explain linguistic forms in poetry, they had to reach a number of interpretive decisions with several forks along the way, including even how to analyze a particular word, i.e., what its constituent parts are. The entire edifice of exegesis gains further stories when we take into account the non-Pāṇinian grammars and their accompanying commentarial literature.,To tackle this rather large body of literature spanning close to two millennia, I have focused on two works of alaṅkāraśāstra, “poetics,” that deal with the topic of linguistic purity in poetry, Bhāmaha’s Kāvyālaṅkāra (Ornamentation of Poetry) and Vāmana’s āvyālaṅkārasūtra (Aphorisms on the Ornamentation of Poetry). In each work, we find a number of specific grammatical problems addressed, but in different ways and representative of different attitudes toward poets and the interpretation of Pāṇini’s grammar. In preparation for studying the application of grammar to poetry, I have dedicated a chapter to the more general role of śāstra, roughly “science” or “knowledge system,” in the earliest extant works on poetics so as to more precisely understand their nuanced and delimited domain with regard to poetic composition (ch. 2). In addition to a general overview of the chapters on linguistic purity (ch. 3 and ch. 5), I have presented two longer case studies, one from Bhāmaha’s work (cf. 4) and one from Vāmana’s (ch. 6), in which I trace the development of two remarks on poetic usage throughout the grammatical tradition and in a range of commentaries on poetry. In each instance, I have been able to demonstrate not only how long-lasting and varied these debates were, but also how the authorities on correct Sanskrit evolved over time.,My general findings point toward a much more vibrant and multifaceted exegetical tradition than has generally been assumed in secondary literature. Some commentators were driven by the traditional association of grammar with the Vedas and held the opinions of Patañjali as sacrosanct — a trend that gained momentum over time — whereas others felt free to turn to different grammars and interpretations of the Aṣṭādhyāyī not supported by, or even in contradiction to, Patañjali. Furthermore, the non-Pāṇinian grammarians contributed more to the history of Sanskrit than has generally been assumed, and we must continue to read and publish new texts in order to obtain a more representative picture.