Can Language Still Be Taught?
Writing, meaning and authorship in the age of AI
The disappearance of the blank page
The blank page has disappeared. What once marked the beginning of writing—empty, uncertain, resistant—has been replaced by immediate suggestion. Structure appears instantly, sentences follow, arguments take shape before the first real question has been fully formed. There is no longer a moment of not knowing where to begin. And yet, that moment was never empty. It was the space in which thinking emerged.
Writing has always been more than expression. It was a process of discovery, a slow construction of meaning shaped by hesitation, revision and doubt. Now, writing increasingly begins with generation and that subtle shift changes the role of the student in ways that are only just becoming visible.
Writing as thinking
For centuries, language education rested on a quiet assumption: that writing reflects thought. Students were asked to articulate what they understood, to structure arguments, to refine ideas through language. The act of writing was inseparable from the act of thinking. That relationship is now shifting.
“I don’t know what I think until I write it down.”
Joan Didion, Writer & Essayist
This insight captures what is at stake. Writing is not merely a way of expressing thought; it is a way of producing it. When writing is partially outsourced, thinking itself risks becoming externalised. The process no longer moves naturally from thought to language, but instead begins with generated output that must be interpreted, filtered and, at best, understood afterwards.
The sequence changes—and with it, the depth of engagement.
From writer to curator
Students are no longer only writers; they are increasingly becoming curators of language. They work not from a blank page, but from an abundance of possible formulations. Text appears, options emerge and the task shifts towards selection and refinement.
What matters is no longer simply the ability to produce language, but the ability to discriminate between what is meaningful and what is merely plausible. The skill lies in recognising which sentence carries weight and which only appears to do so.
Writing, in that sense, becomes less an act of creation and more an act of judgment.
The crisis of authorship
This shift introduces a question that is difficult to answer with precision: what is still an “own text”? A student may generate a draft, reshape it, refine tone and structure and integrate fragments of original thinking. The result may be coherent, persuasive and technically strong. But its origin is no longer singular.
“Authorship is no longer a singular act, but a distributed process.”
N. Katherine Hayles, Distinguished Research Professor, UCLA
This is not plagiarism in the traditional sense. It is something more subtle: a diffusion of authorship in which the boundary between human and machine contribution becomes increasingly opaque. And when authorship becomes unclear, the meaning of writing itself begins to shift with it.
Language as the house of meaning
Language is not a neutral instrument. It does not simply carry meaning—it shapes it. The words we choose determine what can be thought, how ideas are structured and how reality is interpreted.
“Language is the house of being.”
Martin Heidegger, Philosopher
If language is increasingly generated by systems trained on probability, then something changes at a deeper level. Not only how we write, but how we think. The structure of language becomes less individual, less situated and less connected to lived experience.
What is at stake is not only expression, but interpretation itself.
The risk of fluent emptiness
Artificial intelligence produces language that is increasingly fluent. It is grammatically correct, stylistically consistent and structurally coherent. It performs exactly as language is expected to perform.
But fluency is not the same as meaning.
A text can be persuasive without being grounded, structured without being understood. It can resemble thought without emerging from it. The result is not incorrect language, but something more elusive: language that functions, yet does not fully signify.
“Non-conscious systems can process information, but they do not understand it.”
N. Katherine Hayles
The risk, then, is not that students write poorly, but that they write convincingly without fully understanding what they are saying.
The paradox of the synthetic voice
Because artificial intelligence is trained on vast corpora of existing language, it tends towards the most probable formulation—the most likely next sentence, the most statistically plausible phrase. What emerges is a particular kind of voice: smooth, balanced and increasingly uniform.
Over time, this has consequences.
The more language is generated, the more it begins to converge. Expression moves towards the average. The edges—where originality, dissent and creativity reside—become less visible. What remains is a shared, synthetic style in which everything is well written, but little stands out.
The return of voice
In such an environment, the human voice becomes more valuable precisely because it is not optimised. It carries context, experience and perspective. It reflects not only what is said, but where it comes from.
Artificial intelligence can simulate tone, but it cannot inhabit it. In a world of generated language, voice becomes the last form of authenticity.
Embodied language
This becomes even more apparent when language touches experience. A text about grief, written by a system, may be structurally convincing. It may even feel emotionally accurate. But it does not emerge from lived reality. It is not embodied.
Human language is. It carries traces of memory, history and experience. It reflects not only meaning, but presence. And that distinction cannot be replicated through generation alone.
Implications for language education
All of this has consequences for how language can—and should—be taught. The goal can no longer be limited to producing grammatically correct or structurally sound text. Those capabilities are increasingly available by default.
Instead, the emphasis must shift towards understanding. Students must learn to interpret language, to recognise tone and intent, to identify assumptions and to position themselves in relation to what is being said.
“Language learning is fundamentally about meaning-making, not sentence production.”
Council of Europe, CEFR Framework
Writing remains important, but not as an isolated skill. It becomes part of a broader process of thinking, interpreting and taking responsibility for meaning.
The deeper shift
This leads to a paradox that sits at the heart of this transformation.
The easier it becomes to produce language, the harder it becomes to say something meaningful. The more fluent the output, the less certain the understanding. The more accessible writing becomes, the more fragile authorship appears.
Education is no longer constrained by a lack of language. It is challenged by an excess of it.
Europe’s position
In this context, Europe is not without direction. Its intellectual traditions have long treated language not as a tool, but as a medium of thought. From rhetoric to philosophy, from literature to law, language has been understood as the space in which meaning is formed.
That tradition now becomes relevant again. Not as a form of resistance, but as a foundation for adaptation. Europe may not lead in the speed of language generation, but it can lead in the depth of interpretation.
What remains
So what remains of writing?
Not everything, but something essential. The ability to understand meaning, to construct arguments and to take ownership of words. The capacity to recognise what is being said, what is implied and what is missing.
Because in a world where text can always be generated, the value of education lies elsewhere. Not in producing language. But in understanding it.
Final line
If anyone can generate a text, the value of education lies in knowing what it means—and whether it should have been written at all.
This article is part of the series The University After AI, published in the Culture & Education section of Altair Media.
Photo by Lilartsy on Unsplash
A pen above the page—capturing the moment where writing once began with thought and now increasingly begins with suggestion.
