Why Prakrit before Sanskrit?

A Deep Dive into Indian Epigraphy’s Most Enduring Paradox

India has an incomprehensively long history and so does its linguistic tradition. So far, we can only speculate about the language spoken by the Harappans, or for that matter those spoken by the aborigines deeper in the subcontinent. There is no speech specimen to work on and there’s little in the way of inscriptions. We were fortunate with Ancient Egyptian, as its continuity through Demotic and later Coptic allowed us to trace sound changes back to antiquity. This continuity has enabled us to develop a fairly reliable model of how the earliest Pharaohs may have sounded. With the Harappan tongue (or tongues), we have no such continuum to work with. The handful of inscriptions we’ve uncovered so far too remains to be deciphered. All we have is a sharp, abrupt jump from that to Sanskrit, roughly around, by most reckoning, 3500 BC. There’s most certainly plenty of cultural continuity, but we’re talking linguistic continuity here. That gap remains to be filled.

We know that Sanskrit is the oldest Indic language, older than the entire Dravidian family. This debate has been extensively studied and settled in an earlier article on the subject.

---

---

The traditional view is that most non-Dravidian languages spawned from small set of nonstandard ancestors collectively called Prakrit. These Prakrit tongues are typically contrasted with Sanskrit in prestige, status, and semantic standard. Whatever be one’s position on this contentious subject, learned consensus largely remains that Prakrit follows Sanskrit and not the other way around. In which case, we must address a nagging question—Why doesn’t epigraphy support this order? The oldest specimens of Indic writing, outside of Indus seals and aboriginal cave scribbles, known to us come from the imperial edicts of Aśoka. These edicts are in Pali, Aramaic, and even Greek. But not Sanskrit. This is roughly between 260 and 232 BC. But the first specimen to exhibit Sanskrit, as traditionally held, wouldn’t come until around 150 AD with Rudradaman’s rock inscription at Junagadh. That’s almost four hundred years between written Prakrit and written Sanskrit.

Fig. 1: Aśoka’s edicts come in a variety of languages including not only Aśokan Prakrit but also foreign tongues like Greek and Aramaic, but no Sanskrit. [courtesy World Imaging in “Greek and Aramaic inscriptions by king Ashoka,” World History Encyclopedia, https://www.worldhistory.org/image/259/greek-and-aramaic-inscriptions-by-king-ashoka/. Accessed 6 October 2024.]

If Sanskrit emerged before Prakrit, how come Prakrit inscriptions predate Sanskrit inscriptions? This is a linguistic paradox that has been extensively studied and investigated for almost a century now with few clear answers. In this article, we’ll attempt to follow this study and arrive at a working theory that stands the test of reason and logic. In light of available historical facts, of course. In the course of this investigation, we will also tangentially touch upon the history of writing in the Indian context. The epigraphic gap between the Indus collapse around 1800 BC and Aśokan edicts around 250 BC is troubling. We will appreciate this gap, if not fully decode it. The first order of business in that direction is to clearly establish the distinction between Sanskrit and Prakrit. And address some misconceptions along the way.

The Concept of Diglossia

A proper understanding of the subject at hand warrants a degree of familiarity with a couple of linguistic concepts. One such concept is diglossia but even before that, we ought to understand code-switching. While the term itself might sound unfamiliar, the phenomenon it describes isn’t. Unaware of its technical appellation, we engage in code-switching almost instinctively at all times. At least those of us who speak more than one language. So, what is it?

Code-switching is the practice of alternating between two or more languages or language varieties within a single conversation, sentence, or even phrase. It typically occurs in multilingual or bilingual communities where speakers are fluent in more than one language. Code-switching can happen for various reasons, such as expressing identity, conveying specific meanings, accommodating different listeners, or adhering to social or cultural norms. Or just plain old habit. An excellent example in the Indian context is Hinglish, a seamless nonstandard blend of Hindi and English:

“Kal meeting hai, so don’t be late.”

In this example, the same sentence is rendered partly in Hindi and partly in English. In fact, even the Hindi part comes embedded with an English word, meeting. This practice of instinctively switching between two or languages in the same context, often in the same sentence, is common to all multilingual societies. In some cases, also a marker of social status. Hinglish is widely used in urban India, among younger generations, and reflects the linguistic and cultural interaction between Hindi and English. The phenomenon is often studied in sociolinguistics because it reveals how language interacts with social contexts, helping speakers negotiate different social identities or roles. It also shows how languages influence each other in multilingual settings.

Since code-switching involves multiple languages or dialects, one language or dialect typically stands out as the dominant substrate. It serves as the primary grammatical framework within which elements from other languages are inserted during conversation. This dominant tongue is called the matrix language, the second concept fundamental to the subject at hand. The non-dominant components are called embedded languages. The matrix language controls the sentence structure, word order, and other grammatical rules, while the embedded language provides individual words or phrases. In the case of Hinglish, for instance, Hindi is the matrix language because it dictates the word order and the sentence structure, and English, which contributes little more than rudimentary vocabulary, serves as the embedded.

The matrix language concept is particularly important in studying code-switching because it helps linguists understand how two or more languages interact in bilingual speech, with one language, the matrix, setting the structural norms and the other, the embedded language, contributing lexical items or short phrases.

And now we can finally discuss diglossia. It’s a subset of the above sociolinguistic situation, where two distinct varieties of the same language coexist within a speech community. The varieties can be classified as High and Low, and they often differ significantly in terms of grammar, vocabulary, and usage contexts. The High variety is typically used in formal settings, such as in literature, education, government, and religious contexts. It is usually regarded as the prestigious or “standard” form of the language. The Low variety, on the other hand, serves informal settings, such as everyday conversation, family interactions, and social gatherings. It may be considered colloquial or nonstandard.

Speakers generally do not blend High and Low varieties within the same context. For example, one might use the High variety during a speech or a religious ceremony, while the Low variety is employed in casual conversations with friends or family. Consequently, code-switching is virtually nonexistent between these varieties. Diglossia often mirrors societal attitudes toward different language forms, with the High variety typically associated with education and prestige, whereas the Low variety may be linked to regional identity or informal speech. A noteworthy example in the Indian context is the distinction between the Hindi spoken in Delhi and that in Mumbai or Bihar. The former represents standard, educated speech, while the latter varieties are frequently dismissed or even derided as uneducated vernaculars. Beyond India, this distinction is observed in Greece with Katharevousa and Demotic, though Demotic has now become the standard. In the Arab world, Standard Arabic contrasts with dialects like Egyptian Arabic. Similarly, in German-speaking regions of Europe, Hochdeutsch (High German) is distinguished from Plattdeutsch or Niederdeutsch (Low German). Examples also exist in English, including within subsets like American English, where the East Coast dialect enjoys more prestige than Ebonics or Texan English. Similarly, in British English, there are distinct class differences between Queen’s English and vernaculars such as Geordie or Cockney.

While code-switching is not a characteristic of diglossia, the concept of matrix language remains relevant. In this context, the matrix language refers to the more dominant language in everyday speech—the register most commonly used in informal communication. Typically, this is the Low variety specific to a region, such as Bihari Hindi in Bihar or Cockney in certain parts of London. The Low or matrix dialect is the one most speech communities default to unless in formal settings that require the High or “educated” variant. With these fundamentals in place, we are now sufficiently equipped to discuss Sanskrit and its relationship with Prakrit.