Between Words, Learning: Towards a Language Ecology in the Age of AI

Language under strain, attention under pressure

We begin with what we see: the fatigue, the distraction, the faces blank and bodies slumped in front of screens. And then there’s what we sense, without quite being able to name it: a kind of shift, a quieter unraveling. Not something that breaks, but something that fades, an erosion of some sort.

In places of learning, even those with the most sophisticated instructional designs, a fundamental ingredient seems to be thinning out: language. And yet, it’s language that holds thought together, structures memory, gives shape to emotion, and weaves the threads that create relationships.

And it’s language, now, that’s being mechanized, standardized and outsourced to machines and their artificial intelligence. As if speaking, understanding and transmitting has become a simple mechanism.

So maybe, before designing the next training module, we need to pause and ask ourselves: what language are we still learning in? And more urgently, what language do we want to learn in tomorrow?

Language, a common good under extraction

A few decades ago, we were taught to speak “well.” Today, we’re taught to speak “correctly,” And maybe there’s something else creeping in: a culture where we’d rather sound polished than sincere. Where vulnerability risks ridicule, and earnestness feels dangerous. In a world perpetually observed, “trying” has become cringe, and cynicism a kind of way to speak without ever being exposed. But learning demands the opposite. It asks us to stumble, to reach, to say the not-quite-right thing before we find the right one.

Language is becoming an interface, not an experience. This shift mirrors what sociolinguists call linguistic insecurity, a phenomenon where speakers internalize a standard usage norm and view their own speech as inferior.

Meanwhile, our words feed statistical models. Every message sent, every comment posted, every hesitation picked up by a microphone becomes data, raw material for the language industry. The phone in our pocket has become a language pump, a silent siphon turning the living into the predictable.

In the learning space, this creates a strange effect: the more we formalize, the more we lose. The more we align content, the more voices go silent!

What learners have to say (in their own tone, with all their clumsiness, bursts, repetitions) disappears in favor of a modeled eloquence. Language that “gets the job done,” but leaves no room for reply, or for thought.

AI Newspeak: a language without relation

There are connections, and there are relationships. The former are technical, fast, efficient. The latter take time, pass through discomfort, and demand presence. They’re the ones that transform people, their potential and their development.

Generative AI gives us sentences with no silence, answers with no waiting, scripts without hesitation. A disembodied language.

As Orwell wrote in 1984, “The purpose of Newspeak was not only to provide a medium of expression… but to make all other modes of thought impossible.” A language designed for compliance leaves no space for divergence or emergence.

When an onboarding journey is fully automated, everything flows… on paper. And yet, something’s missing! The weight of a glance, the friction of a misplaced word.

Learning isn’t just access to content. It’s stepping into dissonance, into otherness.

Learning without speaking is learning half-way

There’s no living pedagogy without speech, no transmission without dialogue.

Vygotsky (and lots of recent researchers) theorized it, but any attentive facilitator already knows from experience: what turns content into understanding isn’t precision, it’s circulation.

In an AI-first module (where content flows fast, where prompts produce scripts, where questions are anticipated before they’re even asked) what’s missing is often the essential: the surprise of an unexpected word, the tension of a disagreement, the complicity of a half-suppressed laugh.

All these micro-events give knowledge its texture, its memory, its grounding.

Language doesn’t simply deliver a lesson; it is the thing that makes learning possible, anchoring it in our memories.

What if our words became unreadable tomorrow?

In some companies, algorithms filter applications based on “neutral” criteria, all wrapped in perfectly inclusive language (yet biases persist…).

There’s danger in confusing language with communication. In believing that a message that lands is a message that resonates. In forgetting that words aren’t made to be flawless, but to carry: trouble, an idea, a world. Communication moves information where language moves meaning, it delivers when language reveals. In a world obsessed with clarity and correctness, we risk treating words as perfectly efficient containers, when in reality they are living vessels.

Taking care of language again

At WDHB, we still believe competence can’t be taught. It has to be lived! It has to be told and shared. Sharing always passes through words and voices, through tone and breath.

Our responsibility is here: to keep language’s transformative power alive.

To resist its drift toward pure utility. To treat it like a garden: cultivated, diversified, made livable.

Three gestures, among others, that we make in designing our solutions, that our facilitators embody, that our people do believe in:

— favor dialogic and sensorial formats, where speech flows in all its dimensions;

— train leaders not just to pitch, but to hold space for language;

— question, every time, the linguistic effects of the tools we implement.

Toward a living language ecology

Language is the water of learning. Invisible, but essential. If it dries up in the pipelines of automation, everything we’re trying to grow (desire, knowledge, connection) wilts.

And maybe it’s time to listen to those who’ve long been weaving other stories.

Donna Haraway reminds us: “It matters what stories make worlds, what worlds make stories.” (Staying with the Trouble, 2016)

So let’s pay attention. To words that take their time, to silences that hold.

To stories that don’t explain everything, but open paths.

And maybe each learning space could once again become a place where language takes root.