TL;DR
The standard account of linguistic pride in India runs in two halves. The first — that strong subnational identity produces better social provision — is a genuine research finding: Prerna Singh’s How Solidarity Works for Welfare builds an entire comparative framework around “the subnationalist motivation for social development,” and treats the differential strength of subnationalism across Indian states as the thing to be explained.1 The second half — that the same pride reliably curdles into hostility toward outsiders — is repeated constantly and, in the sources I could actually read, is mostly asserted rather than measured. A systematic review of the literature on Dravidian political communication frames identity, language and cultural symbols as instruments of mobilisation and welfare delivery, alongside populism and digital media described as “double-edged.”2 I went looking for the chauvinism half with the same standard and did not find it. That asymmetry is the honest finding, and it is more useful than a verdict.
The feeling that builds a movement is the same one that makes it dangerous to stand near. Photo: Fleur on Unsplash.3
The Half That Is Actually Established
Start with what has been done properly, because it is genuinely counter-intuitive and it usually gets skipped in favour of the argument.
Prerna Singh’s How Solidarity Works for Welfare: Subnationalism and Social Development in India is a book-length comparative treatment of a specific causal claim: that solidarity at the level of a linguistic region motivates elites and publics to invest in shared social goods. Its chapters set out “the subnationalist motivation for social development,” then ask where the differential strength of subnationalism across Indian states comes from in the first place.1
That ordering matters. The argument is not that some states happen to be nicer. It is that a shared identity below the level of the nation creates a bounded public whose members will accept redistribution toward each other — and that this is why some Indian states built schooling and health systems far ahead of their income level, while others did not.
If that is right, then “linguistic pride” is not decoration on top of politics. It is part of the machinery that produced Tamil Nadu’s and Kerala’s social indicators.
What the Literature on Tamil Nadu Actually Says
A systematic review of the academic literature on Dravidian political communication, published as an ISEC working paper, describes the machinery in some detail.2
Its framing is about mobilisation. The paper analyses “the dynamic reciprocity existing between populism, welfare programs, activism and political communication,” and highlights “how and why regional identity, linguistic politics, and cultural symbols have gained such traction in voter mobilization and party rhetoric.”2 It treats activism as “a communicative instrument, deeply ingrained in social justice and anti-Brahmin rhetoric,” and reads the legacies of Periyar and Annadurai as showing “how symbolic communication fosters Tamil identity.” Welfare initiatives — free laptops, houses — are analysed “as strategic means of mobilizing marginalized populations.”2
Note what that is and is not. It is a claim about how identity is used to build electoral coalitions and deliver programmes. It is not a claim that the identity is directed against anyone. The review’s one flagged risk is about the medium rather than the sentiment: the move to digital platforms is “a double-edged phenomenon,” offering engagement while posing problems of manipulation.2
The identity itself, in the account a Frontline interview gives, is explicitly acquired rather than inherited — “Being Tamil is an identity you acquire by doing, by speaking, by belonging” — describing a journey “from anti-Brahmin reform to a mass political movement” in which “a century-old call for equality evolved into a lasting model of power, identity, and governance.”4
An identity you join by speaking the language is, definitionally, an open one. That does not make hostility impossible. It does mean the mechanism is not ethnic in the way the chauvinism argument usually assumes.
The Half I Could Not Substantiate
Now the part where I have to report a failure rather than a finding.
The second half of the standard argument holds that linguistic pride predictably hardens into hostility — against migrants, against Hindi, against whoever is nearest. Everyone knows the examples. The anti-Hindi agitations are real historical events.5 Language-based street violence in Indian cities is a real, recurring phenomenon.
What I could not find, in any source I was able to read, was rigorous work establishing the general claim: that stronger subnational linguistic identity causes more hostility to outsiders, measured, across states, controlling for the obvious confounders. Not “here are incidents” — incidents are easy — but the comparative treatment that Singh’s work gives the welfare side.
It may exist. My search was not exhaustive and I was working from a fixed set of citations. But I want to be precise about the state of my own evidence: I am able to cite scholarship for the solidarity-produces-welfare direction and not for the pride-produces-chauvinism direction, and I am not going to paper over that with anecdotes.
This is the same problem as confusing agreement with correctness. That a claim is widely repeated, including by people I would expect to be careful, is not evidence for it. It is evidence that it is repeatable.
Why the Asymmetry Might Be Real
Three possibilities, and I cannot distinguish between them from here.
The claim might be true but unstudied. Hostility is harder to measure than school enrolment. There is no clean cross-state index of linguistic chauvinism, and building one would require decisions about what counts that are themselves contested. Absence of measurement is not absence of phenomenon.
It might be true only in specific conditions. The interesting version of the hypothesis is not “pride causes hostility” but “pride causes hostility when the group perceives its language losing ground in employment, administration or migration patterns” — a conditional claim that would explain why the same identity looks generous in one decade and defensive in the next. That is testable and, as far as I can tell, untested.
Or the two things might be less connected than the pairing implies. Solidarity toward an in-group and hostility toward an out-group are psychologically distinct, and a large literature outside this subject finds they do not travel together as reliably as intuition suggests. If that holds here, the “two faces of the same coin” framing is simply wrong, and the coin has one face plus a separate phenomenon that sometimes coincides with it.
I lean toward the second, on no evidence I can show you, which is why it is in this section and not the argument.
What This Changes
Mostly it changes who has the burden.
If you want to argue that Tamil, Kannada, Marathi or Bengali linguistic assertion is dangerous, you are making an empirical claim, and the comparative evidence currently sitting on the table points the other way — toward identity as an engine of social provision. That does not settle it. It does mean the argument needs to be made rather than assumed, and that a list of incidents is not the same as a rate.
And if you want to argue that linguistic pride is purely benign, the honest position is that nobody has measured the downside carefully enough for you to claim that either.
The useful stance is the uncomfortable one: one half of a familiar two-part story has research behind it, the other half is in the position astrology-style claims usually occupy — intuitive, widely believed, and not yet checked.
Where I’d Hold This Loosely
Five limits, and the first two are about this post rather than the subject.
An earlier version of this post presented four phrases in quotation marks that appear in none of the sources cited for them — including a Bengali sentence attributed to a Cambridge University Press book page, and a rhetorical question of the post’s own formatted as though someone had said it. Invented illustrative phrasing dressed as quotation is the worst defect in the whole archive audit, and all four are gone.
Second, fourteen of its claims rested on a single partisan outlet, cited for characterisations of language politics in states the post did not otherwise source. Those claims and that citation are removed rather than re-supported. Nineteen citations are down to five, all verified.
Third, I am citing Singh’s book from its publisher’s table of contents, not from the text. I can see the argument’s shape and chapter structure; I have not read the evidence chapters, so I have described the thesis and deliberately quoted no findings from it.
Fourth, the ISEC paper is a working paper and a review of other people’s literature rather than primary research. It is a reasonable guide to what the field says and a weak basis for any specific causal claim, which is why I have used it for framing only.
Fifth, and against my own framing: “I could not find the evidence” is a statement about my search, not about the world, and readers should weight it accordingly. I worked from the post’s existing citation set plus what those pointed to. Someone who knows this literature may be able to hand me the comparative study I am saying does not exist, and if so the post is wrong in an interesting way rather than a boring one. I would rather be corrected than keep repeating the half I cannot support.
Footnotes
-
https://www.cambridge.org/core/books/how-solidarity-works-for-welfare/subnationalism-and-social-development-across-indian-states/5A3842A9EB30BAAB5714369068B8555F ↩ ↩2
-
https://www.isec.ac.in/wp-content/uploads/2025/10/WP-596-Thirupati-and-Anil-Kumar-Final.pdf ↩ ↩2 ↩3 ↩4 ↩5
-
https://unsplash.com/photos/a-large-crowd-of-people-holding-flags-and-flags-eADQs40WywY ↩
-
https://frontline.thehindu.com/books/dravidian-pathway-tamil-nadu-dmk-periyar-annadurai-identity/article70262501.ece ↩
-
https://en.wikipedia.org/wiki/Anti-Hindi_agitations_of_Tamil_Nadu ↩
