How Language Shapes Thought, According to Lera

There's a version of this story that goes: language is just plumbing. Thoughts happen, then words carry them from one brain to another. The pipe doesn't change the water.

Lera Boroditsky has spent her career running experiments that make that picture look increasingly wrong.

In a recent conversation with the Institute of Art and Ideas' How the Light Gets In, the cognitive scientist laid out her research program with the clarity of someone who has had to defend it many times—because, not long ago, she did. The field didn't always want to hear it.

The hypothesis that kept getting buried

Linguistic relativity—the idea that the language you speak shapes how you think—has had a complicated century. The strong version of the hypothesis, associated with Benjamin Lee Whorf, posited that language determines thought: speakers of different languages literally cannot think certain thoughts. That version overreached, got shredded, and dragged the whole research program into disrepute.

The backlash had a particular champion. Chomskyan linguistics, which dominated the second half of the 20th century, held that all languages share a universal deep structure—and that surface differences between them, however spectacular, don't penetrate cognition. If languages are fundamentally the same underneath, speakers of all languages will think fundamentally the same way.

Boroditsky's account of why that framework stalled research is worth sitting with. "No one had discovered what that core is," she told the interviewer. "It was a hypothetical core. But it was just a claim." A claim that, because it preemptively closed the question, made the experimental program seem pointless before it started.

What changed? Two things, she argues. First, fieldwork. As linguists actually went out and documented the world's languages in detail, the structural differences between them turned out to be striking and real—not superficial noise on top of a shared grammar, but genuinely different systems for carving up space, time, causality, and color. Second, cognitive psychology accumulated decades of evidence that cognition is highly malleable. "Once you have those two puzzle pieces—languages really differ and cognition is really malleable—you put those together, that creates an environment where you can find these incredibly profound differences between speakers of different languages."

Color as a test case

The paradigm Boroditsky returns to most often involves color, and for good reason: it's clean, it's measurable, and it's the kind of finding that hits you somewhere pre-verbal.

Russian makes a categorical distinction between light blue (goluboy) and dark blue (siniy) that English doesn't. English speakers have one word—blue—that covers the whole range. The question is whether that linguistic difference is merely a labeling convenience, or whether it actually changes how speakers perceive the color spectrum.

Boroditsky's experiments show it does. Russian speakers are faster to distinguish between colors that fall on opposite sides of the goluboy/siniy boundary than they are to distinguish between colors within the same category—even when the physical distance between hues is identical. English speakers don't show that asymmetry. The boundary, for them, doesn't exist. At the neural level, electrophysiological research published in Frontiers in Human Neuroscience documents ERP (event-related potential) components that register categorical color boundaries in speakers whose language names them—responses that are absent or attenuated in speakers whose language doesn't draw the same line.

What makes this finding interesting isn't just that Russian speakers perform differently on a color task. It's the implication about attention. Boroditsky frames it this way: "Language can guide our attention to some elements and not others in the world. The world is very complicated, and there's an infinite set of things we could potentially pay attention to." Language doesn't show you the world. It shows you a selection of it—and the selection differs depending on which linguistic system you've internalized.

Trapped, or just unexamined?

The obvious follow-up is whether this amounts to being trapped by your language—whether speakers of a language with coarser vocabulary are cognitively impoverished in some domain.

Boroditsky doesn't go there. The problem, in her framing, isn't limitation so much as unexamined assumption. "I think people believe the structures of their language way too much. People think that the language they speak reflects reality or is a true window onto reality. When in fact, each language is just a human-made artifact that was constructed over thousands of years under circumstances that may no longer apply to our lives anymore."

The trap isn't the language itself—it's monolingualism's version of the fish-in-water problem. If you've only ever swum in one linguistic medium, you have no external vantage point from which to notice that it's a medium at all. Learning a second or third language doesn't just expand vocabulary; it gives you access to a different perspective, and in doing so, it relativizes the first one. The structures stop feeling like reality and start feeling like choices.

This is a more modest and more defensible claim than Whorf's. It doesn't say your language locks you into a cognitive cell. It says your language is a set of defaults that feel invisible until you encounter a different set—and that encountering different defaults is, cognitively speaking, genuinely useful.

Personal stakes, research stakes

Boroditsky's own story sits underneath all of this in ways she acknowledges directly. She grew up in the former Soviet Union, moved to the United States at twelve, and found herself in arguments where the combatants were using the same English words to mean fundamentally different things. "You and I might be talking about freedom, we think we're talking about the same thing and we're disagreeing, but we're in fact using the word 'freedom' in different ways."

That experience—of discovering that shared vocabulary can paper over completely different conceptual frameworks—is exactly the phenomenon her research subsequently documented at the experimental level. It's a rare and useful thing when a researcher's biography and methodology point in the same direction.

What disappears with a language

The last third of the conversation turns toward language loss, and the stakes get harder to sit with.

There are roughly 7,000 languages spoken in the world today. Boroditsky describes the rate of attrition as alarming—though the specific figures are contested enough in the literature that treating any single number as authoritative would be a mistake. What's less contested is the direction of travel.

What disappears when a language dies isn't just vocabulary or grammar. It's the accumulated cognitive and cultural labor of every generation that spoke it—every decision about how to parse space and time and causality, every refinement of a conceptual framework, every way of relating to a landscape that a word encodes. "When the last fluent speakers of the language pass away, all of that intellectual labor is gone," Boroditsky says. "And then the loss that the descendants of those people feel is extremely strong because they lose the connection to their past, to their ancestors, to the ways of life that brought them up."

She makes the point that many of the languages currently disappearing are poorly documented—which means there's no archive, no reconstruction path, no recovery. It's not even a matter of translating the library. The library is the people, and when they go, the library goes.

Diversity as baseline, not aberration

Boroditsky's closing argument is about where cognitive science sets its default. For decades, the standard human subject in psychological research was American undergraduates—a sampling choice that, as anyone following the WEIRD (Western, Educated, Industrialized, Rich, Democratic) critique of psychology knows, has systematically distorted what the field thought it knew about "universal" human cognition.

Boroditsky wants to invert that. Rather than treating the WEIRD subject as the template and everyone else as a variation to be explained, start from the assumption that linguistic and cognitive diversity is the norm. "The human mind has created not one way of perceiving reality, but thousands of them," she argues—and that's a testament to flexibility, not fragmentation.

That reframe changes what you're looking for. Instead of mapping deviations from a presumed universal, you're documenting the full range of what human minds can do. And that range, it turns out, includes things that speakers of any single language might not be able to readily imagine.

When asked which language she'd most want to learn, Boroditsky doesn't reach for Mandarin or Spanish. She names Kumeyaay—the language of the people whose land she's lived on in San Diego for a decade. "Sometimes it's good to do something that's just good for the soul."

That answer tells you something about how she thinks the whole question should be oriented: not toward utility, not toward reach, but toward connection—to place, to history, to the particular way of seeing the world that only that language carries.

By Amelia Nwofor, Science Desk Editor