Panini’s analysis and innovations may therefore be seen as the foundation of all high-level programming languages. But the Ashtadhyayi also had an indelible effect on Sanskrit, the language he was describing: it gave this spoken tongue the stability of formal languages — like programming languages — in which a set of rules precisely constrains the symbols, syntax and usages. Natural languages have the tendency to change over time, but Sanskrit has remained astonishingly unchanged in the two and a half millennia since Panini. There have been strong trends toward certain usages, such as the use of compound words, but in general “the stress on refinement and correctness, the overwhelming anxiety to live up to a felt Paninian ideal, kept the language formal for everyone, and channelled creativity towards involution, elaboration, and increasing precision.”²³

So one of the problems of working with Sanskrit texts is that internal linguistic usages give you very little evidence, if any, of provenance and dating. If you had a pandit in contemporary Varanasi write a letter in Sanskrit and time-machined it back 2,000 years, his ancestors would be able to read it with perfect ease. In Sanskrit, therefore, the usual distinction between normal and formal language is collapsed, and the original derivation of the language’s name from the root samskrta, “constructed, finished, well or completely formed,” carries precise denotative value. It is only later that “Sanskrit” comes to suggest refined speech, to refer to the language of the shishta, the educated, the superior, the polite.

Then, as now, Indians spoke more than one language in daily use. Sanskrit was the eternal language of the cosmopolis, of the marga, the path; the Prakrits were the “natural, normal, ordinary” regional languages, the languages of desha, of place. Because Prakrits were subject to change, the stricter grammarians regarded them as apabhramsa, “degenerate languages” that had sloughed off from the eternal Sanskrit through careless usage. People spoke both Sanskrit and Prakrits, and they were not — as elsewhere in the world — speaking two registers of the same language. They existed in a condition that is better described as “hyperglossia” than “diglossia”: “What we encounter is not an internal split (di-) in registers and norms, typically between literary and colloquial usage,” Sheldon Pollock tells us, “but a relationship of extreme superposition (hyper-) between two languages that local actors knew to be entirely different.”²⁴

The Kamasutra, a Sanskrit text addressed to urban sophisticates, advises that “by having one’s conversations in the assemblies neither too much in Sanskrit nor too much in the local language a person should become highly esteemed in the world.”²⁵ Too much eternal language and you revealed yourself as a total marga snob; too much Prakrit, on the other hand, marked you as a desi bumpkin. Sanskrit drama reproduced these social dynamics as a convention: the upper-class males — kings, ministers, educated Brahmins — spoke Sanskrit; everyone else spoke Prakrit: merchants and bankers, women (with the exception of courtesans), and of course the lower classes.

Sanskrit — once the language of liturgy — was officially available only to the “twice-born” of the caste system, to its top three tiers of Brahmins, Kshatriyas (warriors), and Vaishyas (traders). Sudras — the manual laborers, the people who provided services — were forbidden to learn or speak Sanskrit, as were those who fell outside the caste system altogether, such as tribal peoples. Thus those who spoke against the Brahminical system — the Buddha, for instance — often used Prakrit languages because Sanskrit was marked as the language of the elite.

Whatever its social value, Sanskrit’s stability and emphasis on precision meant that it was regarded as an ideal language for the sciences and philosophy. Since it was believed to be an eternal language, the fifty phonemes of Sanskrit, the matrka, were regarded as the root vibrations from which the universe had emerged, and were sometimes worshipped as the “little mothers.”²⁶ Sanskrit phonemes and words therefore had an elemental, essential link to reality that other languages lacked. This view was challenged by Buddhist thinkers who argued that “the signifier is related to the signified as a matter of pure convention,” but the notion that the truth could only be spoken in Sanskrit, and grammatically correct Sanskrit at that, was immensely persuasive.²⁷ Two monks were said to have argued that the Buddha’s words should be translated into Vedic Sanskrit, since people were corrupting them by repeating them in local dialects. The Buddha himself rebuked them, “Deluded men! This will not lead to the conversion of the unconverted,” and he commanded all monks, “You are not to put the Buddha’s words into [Vedic-Sanskrit] verse. To do this would be to commit an infraction. I authorize you, monks, to learn the Buddha’s words each in his own dialect.”²⁸ This injunction was obeyed, and Pali, “a new and parallel sacred language,” was created by Buddhists, and yet, by the second century CE, “a vast Buddhist canon in Sanskrit” had been created.²⁹ The rebels against Vedic authority — the Buddhists, the Jains, the Tantrics — had to speak in Sanskrit after all.

It was perhaps the multitude of viewpoints and ideologies attempting to speak to each other and against each other in Sanskrit that intensified its grammarians’ search for even more exactness. Beginning in the fourth century BCE and culminating in the eighteenth century, an effort was made to create a shastric or scientific Sanskrit that could “formulate logical relations with scientific precision.”³⁰ In this specialized, condensed Sanskrit, the sentence “Caitra goes to the village” would be rephrased as “There is an activity which leads to a connection-activity which has as Agent no one other than Caitra, specified by singularity, [which] is taking place in the present and which has as Object something not different from ‘village,’”

The sentence “Out of friendship, Maitra cooks rice for Devadatta in a pot, over a fire” would be broken down into:

(1) An Agent represented by the person Maitra;

(2) An Object by the “rice”

(3) An Instrument by the “fire”

(4) A Recipient by the person Devadatta

(5) A Point of Departure (which includes the causal relationship) by the “friendship” (which is between Maitra and Devadatta);

(6) The Locality by the “pot”³¹

The shastric version of the original sentence would therefore be something like:

There is an activity conducive to a softening which is a change residing in something not different from rice, and which takes place in the present, and resides in an agent not different from Maitra, who is specified by singularity and has a Recipient not different from Devadatta, an Instrument not different from … ³²

and so on.

So the shastric thinkers tried to create a low-level version of Sanskrit, a counterpart to assembly code. In fact, Rick Briggs, a NASA specialist in artificial intelligence, points out that this decomposition of natural language is very similar to what computer programmers do when they attempt to represent knowledge in semantic nets, which use “triples” to embody logical relations: “cause, event, friendship; friendship, object1, Devadatta; friendship, object2, Maitra; cause, result, cook; cook, agent, Maitra …” and so on. Briggs writes:

It is interesting to speculate as to why the Indians found it worthwhile to pursue studies into unambiguous coding of natural language into semantic elements. It is tempting to think of them as computer scientists without the hardware, but a possible explanation is that a search for clear, unambiguous understanding is inherent in the human being.³³