FLUENT IN NONE: LANGUAGE, AI, AND THE QUIET UNMAKING OF THE AYURVEDIC SCHOLAR

Dr. Aakash Kembhavi

MD (Ayu-Shalya), PGDMLS, MS (Counseling & Psychotherapy) Chief Editor, International Journal of Ayurveda

Disclaimer: The views expressed in this article are my personal opinions and reflections. They do not represent the position of any institution, organisation, or body with which I am or have been affiliated. This article is intended for academic and informational discussion only. It is not medical advice and should not be used as a substitute for consultation with a qualified healthcare professional. In the interest of transparency, integrity, and honesty, I disclose that this article was developed in collaboration with AI tools, which I used to support research, structuring, and drafting. The central idea, the arguments, and the final views and conclusions are entirely my own, and I take full responsibility for the content.

THE QUESTION BEHIND THE QUESTION

Writing in Bloomberg Opinion, Andy Mukherjee took aim at India’s decision to push a third language onto its ninth graders from this July, two of which must be indigenous. His argument was not really about languages at all. It was that a school system whose graduates already cannot communicate competently in the languages they have, and whose returns on an extra year of schooling now trail sub-Saharan Africa, has chosen to spend its political energy on a nationalist language project rather than on the unglamorous work of fixing teachers, curricula and learning itself.

He asked the sharpest question of all: in an age being reshaped by artificial intelligence, an age in which India is already falling behind, why are we making teenagers memorise the grammar of a classical tongue they will abandon at sixteen?

Read that critique, and then look at the BAMS graduate.

The parallel is not approximate. It is exact.

We have produced a scholar who has memorised Charaka in Devanagari, sat examinations written in English, and will counsel patients in Kannada, Tamil, Marathi or Hindi — and who is, as an academic instrument, fluent in none of these.

The school debate Mukherjee describes is, in Ayurveda education, not a looming policy error. It is our settled, unexamined reality.

THE THREE-LANGUAGE TRAP IN AYURVEDA EDUCATION

The BAMS student does not face a choice between languages. The student is silently expected to master three at once, with genuine scaffolding for none.

Sanskrit is the language of the inheritance — the medium of Charaka Samhita, Sushruta Samhita, Ashtanga Hridayam and the Nighantus, and a compulsory subject in the early years of the degree.

English is the mandatory medium of instruction, the language in which every clinical and para-clinical subject is taught, the language of university examinations at RGUHS and its counterparts, and the only language in which research, dissertations and indexed publication are permitted.

The regional language — the mother tongue in which the student actually reasons, argues, doubts and feels — is the medium of every real clinical encounter that student will ever have.

Three languages, three universes, and no bridge built between any of them.

The school debate Mukherjee criticises at least argues openly about which languages a child should carry.

Ayurveda education makes no such argument. It simply assumes a trilingual competence it has never taught, examined or supported, and then quietly blames the student when that competence fails to appear.

THE ENGLISH PROBLEM: MANDATORY MEDIUM, MISSING FOUNDATION

This is the empirical heart of the matter, and it is where honesty is most uncomfortable.

Every BAMS subject is taught in English.

University examinations are set and answered in English. Synopsis, dissertation and any aspiration to publish are English-only enterprises.

Yet the majority of students entering BAMS in Karnataka and across much of India come from Kannada or other regional-medium schooling, often from semi-urban and rural backgrounds where functional academic English was never built.

They arrive having studied their sciences in one language and are abruptly required to learn anatomy, physiology, dravyaguna and roganidana in another — with not a single hour of formal bridging support.

The consequences are visible to anyone who has graded a script honestly.

Students reproduce paragraphs rather than understand them.

They write answers by pattern-matching to remembered phrases rather than by reasoning.

Dissertations arrive grammatically opaque, their arguments buried under sentences the author could not have constructed and frequently cannot explain.

We then read these scripts and quietly conclude that the student is weak. The student is not weak.

The student has been denied a language of thought for academic engagement and is being examined in it regardless.

This is not a failure of intelligence. It is a failure of design.

THE SANSKRIT PROBLEM: RITUAL WITHOUT COMPREHENSION

If English is taught as a medium without a foundation, Sanskrit is taught as a ritual without a meaning.

Sanskrit in BAMS is, too often, a recitation exercise.

It is delivered as a compulsory subject — frequently by faculty who are not themselves scholars of Ayurvedic Sanskrit — through methods that produce transliteration but not interpretation.

The student emerges able to chant a shloka and unable to unpack it: unable to trace its commentarial tradition, unable to situate a verse within its Darshana context, unable to tell where Charaka is making an epistemological claim and where he is recording an observation.

The classical inheritance becomes a performance, a thing to be displayed at convocations, rather than a living resource to be reasoned with.

This is the deepest loss of all, because genuine Sanskrit literacy is the one thing that would let an Ayurvedic scholar return to the primary sources as primary sources — not as translated excerpts filtered through a chain of editors, but as text to be engaged, contested and built upon.

We hold the key to our own library and have taught a generation only to admire the lock.

THE REGIONAL LANGUAGE PROBLEM: THE INVISIBLE MEDIUM

The mother tongue is the one language the student fully possesses, and it is the one language the system pretends does not exist.

It is not the medium of instruction.

It is not the language of examination.

It is not permitted in research.

And yet it is the only language in which the student can take a history, explain a prognosis, counsel a frightened patient, or translate a classical concept into something a person in Bidar or Belagavi can act upon.

Clinical communication competence in the regional language is a core professional skill — arguably the core skill of a physician — and it appears nowhere as a thing to be taught or assessed.

We have built a graduate who cannot connect the classical text, the scientific literature and the patient’s lived suffering, because each of these lives in a different linguistic world, and we have never asked anyone to build the bridge.

THE INSTITUTIONAL PARALYSIS: WHY THIS HAS NOT BEEN FIXED

The reason this persists is not malice. It is a collective looking-away.

NCISM prescribes curricula, credit hours and competencies, but no language support.

State universities set examination standards in English but operate no foundational English programme.

Individual colleges have neither the mandate nor the resources to repair what the national and university frameworks ignore.

And the faculty who might lead such a repair are themselves products of the same broken pipeline — many of them, in candour, not confident academic writers in English either.

So everyone defers to everyone else, and the deficit is inherited intact by the next cohort.

Mukherjee’s charge against Indian school policy lands here with full force. We, too, have spent our reforming energy on the wrong layer. We have tinkered with subject lists, revised credit allocations, rolled out CBDC modules and debated competency mapping — important work, but work that sits on top of a foundation we have refused to inspect. The student cannot read the classics, cannot write the examinations, and cannot be assessed in the language of the clinic. No amount of curricular rearrangement reaches that.

THE FOURTH LANGUAGE: AI, AND THE ABSENCE OF ANY POLICY

And now a fourth language has entered the classroom, and it is the only one in which our students have suddenly become fluent — by proxy.

Artificial intelligence tools now write the assignment, draft the synopsis, compose the answer, polish the dissertation and generate the seminar.

A student who could never construct an English paragraph can now submit flawless ones.

On the surface this looks like a solution to the very language crisis described above. It is, in fact, its most dangerous mutation. The foundational deficit has not been closed; it has been hidden.

The student who could not reason in English still cannot reason in English — but the evidence of that incapacity has been laundered away, and we, as examiners, can no longer see what we are failing to teach.

The specific harm to Ayurveda is acute.

These tools confidently fabricate.

They invent shloka numbers, attribute verses to the wrong Samhita, hallucinate references that were never written, and produce plausible-sounding classical claims with no textual basis whatsoever.

A discipline whose authority rests on fidelity to primary sources is now absorbing, at scale, a stream of fluent and undetectable fabrication — embedded in dissertations that will be archived, cited and built upon. The student cannot catch the error, because the student lacks the Sanskrit and the reasoning to catch it. That is precisely the gap that drove them to the tool in the first place.

What is most striking is the silence around it.

There is, at present, no AI policy in Ayurveda education worth the name.

NCISM has issued no framework. Universities have no declaration norms, no viva protocols designed to test authorship, no guidance on permissible assistance versus academic misconduct.

Most administrators, many principals, and a large share of faculty do not fully grasp either the capabilities or the reach of what their students are already using daily.

We are, as a sector, presiding over a transformation we have not named — neither harnessing these tools deliberately for what they genuinely can do, nor guarding against what they demonstrably do wrong.

Used well, AI could be a tutor that finally helps a struggling student bridge into academic English and reason through a classical argument.

Used in the policy vacuum we now inhabit, it is simply a more elegant way of not knowing.

WHAT REFORM WOULD ACTUALLY LOOK LIKE

The remedies are neither exotic nor expensive. They are a matter of will.

First, a mandatory bridging course in academic and clinical English writing at the point of entry to BAMS — not conversational English, but the specific genre of reasoning, structuring and writing that the degree silently demands.

Second, Sanskrit taught as a hermeneutic and interpretive discipline rather than a recitation exercise, delivered by faculty trained in Ayurvedic textual scholarship, with the explicit goal of returning students to the primary sources as readers and not reciters.

Third, formal recognition of regional-language clinical communication as an assessed clinical competency, because it is one.

Fourth, examination design that rewards demonstrated reasoning over memorised prose — a shift that lowers the language barrier directly, by valuing understanding over fluency of reproduction.

Fifth, a serious national conversation on whether qualifying-level papers might be answered in regional languages, with English reserved deliberately for research and postgraduate work where it is genuinely indispensable.

And sixth, urgently: a clear, sector-wide policy on the use of AI tools in Ayurveda education.

Not a ban, which is unenforceable and unwise, but a framework — declaration norms for assistance, viva and process-based assessment that tests authentic authorship, faculty literacy programmes so that those who examine understand the tools their students use, and a positive doctrine for deploying these tools to teach rather than to disguise. A discipline that does not decide how it will use AI will have AI decide for it.

THE STUDENT WE ARE FAILING

Return, at the end, to one student — the first-year from Bidar or Bellary or any other city in any state of India, who has memorised the Tridosha in a Sanskrit she cannot interpret, studied biochemistry in an English she cannot yet reason in, will one day treat patients in a Kannada or other language, the syllabus refuses to acknowledge, and now submits work written in a fourth language that is not hers at all. She is not failing Ayurveda. Ayurveda education is failing her.

Mukherjee’s verdict on India’s schools was that the system needs substantive reform, not a language project that burdens students, widens inequality and misses the real challenges of the AI era.

The same verdict, almost word for word, fits us.

Ayurveda education does not need another nationalist gesture toward Sanskrit, nor another defensive flourish about heritage. It needs to teach its students to read, to reason and to write — in the languages they actually carry — and it needs to decide, deliberately, what role intelligent machines will play before those machines quietly finish deciding for us.

The elite will, as always, find their exits; it is the student from the small town, stuck with whatever we hand her, who will pay for our inattention.

A tradition that survived twenty-five centuries did so because, in every generation, it found the language to transmit itself.

We are now at genuine risk of losing that transmission — not to hostility from outside, but to a linguistic confusion within that we have refused to name, and to a technology that offers, with great fluency, to do our forgetting for us.

Share your thoughts in the comments below.