Despite a near-universal acknowledgement of the Prophet Muḥammad’s supreme eloquence in the Islamic tradition, ḥadīth (his recorded speech and deeds) is largely absent from the earliest discussions of Arabic grammar. The standard explanation for this absence, forwarded by the 8th/14th century grammarian Abū Ḥayyān al-Gharnaṭī, is that early grammarians were concerned about ḥadīth’s linguistic integrity since it was transmitted non-verbatim, often by non-Arabs. In this dissertation, I show that this explanation is insufficient and anachronistic; it back projects critical considerations of ḥadīth transmission that crystallized centuries after the earliest works on Arabic grammar were written. I offer an alternative thesis: that ḥadīth’s association with urban scholars kept it out of discussions of Arabic grammar in the 2nd/8th century, which revolved around the natural speech of semi-nomadic Arab tribes. To support this, I first show that concerns expressed about the philology of ḥadīth transmission applied equally to poetry and the recitations (qirāʾāt) of the Qurʾān. I then show that the primary goal of Arabic philologists at the time was to record the speech of those semi-nomadic Arabs whose language had not been corrupted by contact with non-Arabs because they believed their speech best exemplified the linguistic environment of the Qurʾān’s revelation. Finally, I show that the process of legitimation, based on contact with other scholars and the nature of data collection, differed in the fields of Arabic philology and ḥadīth transmission/Islamic jurisprudence to such an extent that one could not readily accept the data of the other based on the other’s terms alone. Further, we find that concerns about linguistic fidelity in oral transmission were not as significant as they were in later centuries, that the linguistic reality of this transmitted material was more variable than previously assumed, and that there was a concerted effort beginning at the end of the 2nd/8th century to legitimate the Arabic of the Prophet and put it to use in exegetical and legal contexts.