There is something we take for granted when we talk about endangered languages: that the language in question exists. Not in the trivial sense that someone, somewhere, speaks it, but in the deeper sense that it has been made into a thing: an object with boundaries, a history, a name, native speakers, and the capacity to be lost, but also the capacity to be consciously learnt. This may seem obvious, but it is not. It is the result of a long and specific historical process that coalesced in the seventeenth century. And understanding that process changes everything about how we think about language death, language revival, and why people fight for languages they have never spoken.
This post develops a claim that is simple but counter-intuitive: a language, in the modern political and social sense, exists only when it has been instituted as such. Not in the narrow sense of legal recognition, but in a deeper sense that involves materiality, and, crucially, affect.
A Language That Could Not Be Lost
Consider Gaulish, the Celtic language formerly spoken in what is now France. We do not know when it disappeared, and it may be impossible to know for sure. In the Latin sources of antiquity, we find references to people speaking gallice — “in Gaulish”, or more accurately, “in the manner of the Gauls.” In the second century AD, the jurist Ulpian mentions that certain legal documents may be written in Gaulish alongside Latin, Greek, and Punic. In the sixth century, Gregory of Tours refers to a sanctuary known by its Gaulish name. As late as the ninth century, the author of the Gesta Karola Magni describes small dogs called veltres “in the Gaulish language.”
But what does gallice actually mean? It could refer to a Celtic language distinct from Latin. It could mean the Latin of Gaul, with its local colour. It could be a rhetorical style. The term is radically underdetermined. This is the point: there was no grammar of Gaulish, no dictionary, no metalinguistic commentary on its decline, no apparatus of memory. Nothing that would allow this way of speaking to be conceptualized of as an object or as any form of autonomous entity, capable of being lost, and therefore in need of saving.
From the perspective of those we would now call its speakers, Gaulish does not, cannot disappear. At each generation, people speak gallice, but that manner of speaking shifts, absorbs more Latin, transforms. This does not register as a loss, because the very conditions for there to be a loss — the prior identification of a bounded object called “a language” — do not exist. No shared affect could crystallize around the preservation of something that had not yet been made into a thing. There was, as far as we can tell, no mourning, no regret, no mobilisation: nobody noticed, so there could be no uptake. What does not exist as a language in the modern sense cannot “die,” and therefore cannot be saved.
A Language That Became Thinkable
Now consider Cornish. Unlike Gaulish, Cornish underwent a progressive process of objectification through a series of institutional devices: grammars, dictionaries, literary corpora, but also monuments, narratives about last speakers, and learning manuals (see Kensa Broadhurst’s wonderful new book on Cornish in the nineteenth century here).
The founding moment of this transformation is the encounter between the antiquarian Daines Barrington and Dolly Pentreath in 1775. Barrington, a Fellow of the Society of Antiquaries in London, travelled to the far west of Cornwall to determine whether anyone still spoke the language. He found Pentreath, an elderly fishwife in Mousehole, and wrote up his findings for the journal Archaeologia. The article, “On the Expiration of the Cornish Language,” is a remarkable document — and not only for what it says about Pentreath.
What Barrington produces in this encounter is something that had never existed before: the figure of the “last speaker.” Dolly Pentreath becomes, as far as I can determine, the first person in history to be publicly identified as the last native speaker of a language. And in the same gesture, Barrington invents (or, rather, consecrates) something else: the idea that the end of a language constitutes a loss. Before this article, the fading of a way of speaking was simply what happened. Afterwards, it became something one could lament, or mourn.
This is not a minor rhetorical shift. It is the creation of an affective object. By “affect” I mean something precise, drawn from the philosophy of seventeenth century Dutch philosopher Baruch Spinoza: an affect is what happens when something external to us strikes our bodies and modifies our capacity to act — increasing it or diminishing it. Affects are not private emotions. They are social, they circulate, they accumulate, and they can be captured and redirected by institutions. Barrington’s article, by inventing the category of the last speaker and framing her existence as a loss, creates the conditions for an entire chain of affects (sadness, nostalgia, obligation) to attach themselves to the Cornish language. The encounter produces the void as a social fact, and that void becomes capable of affecting other bodies across time.
The Book That Made the Absent Present
Sad affects alone do not produce action. This is a point on which Spinoza is unambiguous: sadness diminishes our power of acting. Mourning, nostalgia, regret, are all forces that constrain. By themselves, they can produce heritage movements, but they cannot generate a revival movement. Something else is needed: an institution capable of converting those sad affects, of relaying them into something that increases the collective power to act.
This is where Henry Jenner’s Handbook of the Cornish Language (1904) becomes decisive. The Handbook is not just a description of Cornish. It is a device that renders the absent language present. Spinoza writes, in his Ethics (IV, 9), that an affect whose cause we imagine as present is stronger than one whose cause we imagine as absent. Jenner’s book takes a language that existed only as a memory, a loss, a trace in manuscripts and books long out of print, and incarnates it in a material, manipulable, appropriable form. For the first time, Cornish can be learnt. It can be held in one’s hands in affordable book form. The absent cause becomes present, and the affect that attaches to it is intensified.
But the Handbook does something more than make the language available. It provides a reason. Jenner’s most famous sentence reads: “Why should Cornishmen learn Cornish? There is no money in it, it serves no practical purpose, and the literature is scanty and of no great originality or value. The question is a fair one, the answer is simple. Because they are Cornishmen.”
This sentence is almost universally read as a claim about identity: you are Cornish, therefore you should express that identity by learning the language. But — and this is the crucial but — look at the structure of the argument carefully. Jenner does not say: “By learning Cornish, you will become Cornish.” He says: “Because you are Cornish, you should learn Cornish.” Cornishness is assumed to already exist, prior to and independent of the language. The language is not what makes you Cornish. It is what you owe to your Cornishness.
And here is what most analyses miss: what follows immediately after. Jenner writes that Cornwall, “but for a few survivals of Duchy jurisdictions, is legally and practically a county of England.” The Cornish language is described as the last remaining trace of what made Cornwall a distinct entity, after the dissolution of all the institutions of the Duchy. The language becomes, in Jenner’s account, the continuation of a political institution by other means. It is not an expression of identity. It is the minimal institutional form, the last vessel capable of carrying the weight of collective difference in a territory stripped of every other form of autonomous institutional life.
What Institutions Do to Affects
The contrast between Gaulish and Cornish is not a contrast between a language that was forgotten and one that was saved. It is a contrast between a language that was never made into a losable object and one that was rendered available for loss over several centuries, and therefore, more than a century later, available for recovery.
Gaulish produces no revival movement, because it is never transformed into an institution, in a political and ideological framework where such a transformation is not yet thinkable. Cornish, by contrast, ceases to be spoken as a collective practice at precisely the moment when the institutionalisation of languages becomes a norm — for languages like French and English first, and then, by extension, for languages that come to be seen as “minority” or “minoritized.” This movement only happens from the late sixteenth and seventeenth centuries onwards, and then again only in western Europe at first. What this movement does is to turn spoken languages into what had been the linguistic model until then: Latin or Greek — i.e. languages that were only used in specialized settings.
This has consequences for how we think about language revitalisation today. Too often, language policies focus on producing objects (grammars, festivals, school curricula) without attending to how those objects affect (or fail to affect) the people they are meant to serve. A revitalisation effort fails when it remains trapped in a logic of heritage and melancholy: when the institutional apparatus produces only sad affects, only backward-looking commemoration, without converting that sadness into a desire that increases the collective power to act. This, I believe, is one of the deep causes of the relative failure of many revitalisation movements: the wager on identity as an endpoint, rather than on the political project of reshaping the affective power of language that holds people together.
The question, then, is not (just) “Why save a language?” It becomes: why does it suddenly become thinkable, desirable, and possible to do so — for whom, through what mechanisms, and with what consequences? That is the question this research project is designed to answer.