How Language Shapes Culture: The Science

By Joe Newsum

The previous page made the argument: language creates culture, because language shapes thinking, thinking shapes behavior, and behavior repeated at scale is what culture is. This page is the evidence. Not consulting frameworks. Peer-reviewed research, linked so you can read it yourself, organized into the four bodies of work that carry the argument: what language does to perception, to judgment, to brains, and to behavior, and what all of it means for anyone trying to change an organization.

Evidence 1: Language Shapes What People Perceive

Start with the most basic question: does the language you speak change what you can literally see and remember?

The Russian blues study is the cleanest demonstration. Russian has no single word for blue. It obligatorily distinguishes goluboy, the lighter blues, from siniy, the darker blues. Winawer and colleagues had Russian and English speakers do a speeded color discrimination task, and Russian speakers were reliably faster at telling two blues apart when the colors straddled that linguistic boundary. English speakers, with one word for the whole range, showed no such advantage. The critical detail: when Russian speakers had to silently rehearse a string of digits while doing the task, the advantage disappeared. Language was not describing the perception after the fact. It was participating in the perception itself.

The pattern extends far beyond color. The Kuuk Thaayorre, an Aboriginal community in northern Australia studied by cognitive scientist Lera Boroditsky, have no words for left and right. Everything is expressed in cardinal directions: there is an ant on your southeast leg. Because the language demands it, speakers maintain constant compass-level orientation that most English speakers cannot approach. The vocabulary requirement built a cognitive capability.

And it extends to memory and judgment. Fausey and Boroditsky compared English and Spanish speakers watching the same accidental events. English favors agentive phrasing even for accidents: she broke the vase. Spanish commonly uses a non-agentive construction: se rompió el florero, the vase broke itself. The grammar showed up in memory. English speakers remembered who was involved in accidents significantly better than Spanish speakers, while both groups remembered intentional acts equally well. The same research program showed that describing an identical incident as “ripped the costume” versus “the costume ripped” changed how much people blamed the person involved and how much financial liability they assigned. Same event. Different grammar. Different judgment.

Evidence 2: Framing Changes Decisions Without Changing Facts

Tversky and Kahneman’s classic framing experiment gave people a choice between two responses to a disease outbreak. When the options were framed in terms of lives saved, 72% chose the certain program. When the mathematically identical options were framed in terms of lives lost, only 22% did. Nothing about the world changed. Only the words changed. The words were the decision.

The mechanism underneath is loss aversion: losses loom roughly twice as large as equivalent gains. Which means the linguistic frame a leader chooses, protecting share versus capturing growth, fixing a weakness versus building a strength, is not a communication nicety. It shifts what the organization is willing to decide and risk. Every metric name, every initiative name, every way of describing the quarter is a frame, and the frame does cognitive work before anyone consciously deliberates.

Think about the last big decision your team debated. How was it framed, and who chose the frame?

Evidence 3: Shared Language Literally Synchronizes Brains

Uri Hasson’s neural coupling research at Princeton put a speaker and listeners in fMRI scanners. When they shared a language, the listener’s brain activity mirrored the speaker’s, region by region, with a short delay, and in the listeners who comprehended best, some regions actually fired in anticipation of the speaker. The tighter the coupling, the greater the comprehension. When the listener did not speak the language, the coupling was simply absent. Shared language is not a metaphor for alignment. It is the mechanism by which two brains align.

And language changes the individual brain too. In Lieberman’s affect labeling research at UCLA, simply putting a feeling into words measurably reduced activity in the amygdala, the brain’s threat center, while engaging the prefrontal control regions. Naming an experience changes the brain’s response to it, which is part of why a team with shared language for hard conversations can actually have them: the language itself lowers the threat.

Evidence 4: The Language of Commitments Changes Behavior

Perception, judgment, and brains matter because they drive behavior. The behavioral research is the most directly relevant to culture, because it studies the exact linguistic form that Culture by Commitment is built on: the commitment.

Vanessa Patrick and Henrik Hagtvedt’s empowered refusal research tested two ways of declining a temptation: “I can’t” versus “I don’t.” I can’t frames the refusal as an external constraint. I don’t frames it as an identity: this is who I am. Across four studies, including a longitudinal field study, the I don’t framing produced substantially more resistance to temptation and more follow-through on goals, mediated by a measurable sense of psychological empowerment. Two words. Different identity. Different behavior. It is the smallest possible demonstration that commitment language works by becoming who you are, which is exactly how Southwest’s Living the Southwest Way and McKinsey’s Obligation to Dissent operated at scale.

Then there is the largest body of behavioral evidence: Peter Gollwitzer’s implementation intentions. An implementation intention is a commitment in a specific linguistic form, when situation X arises, I will do Y. Gollwitzer and Sheeran’s meta-analysis across more than 90 studies found that this if-then language reliably and substantially increases the odds that an intention becomes a behavior, compared to general intentions of equal motivation. In one well-known study by Milne, Orbell, and Sheeran, 91% of participants who formed an implementation intention exercised that week, versus 35% of the group given the motivational intervention alone. Same motivation. Different language. Nearly triple the behavior.

What the Science Does and Does Not Say

A moment of honesty that most language-shapes-thought writing skips. These effects are real, replicated, and measurable. They are also bounded. Language biases perception and judgment; it does not imprison them. The Russian speakers’ advantage was about a tenth of a second, and it could be disrupted. Framing shifts decisions; it does not fully determine them. Anyone who tells you language is destiny is overselling the science.

But notice what the boundedness implies for organizations. If a single word choice can shift perception by a tenth of a second, a single frame can move a decision by fifty points, and a single sentence form can double follow-through, then imagine the cumulative effect of an entire organization’s vocabulary, operating on every perception, every judgment, and every intention, thousands of times a day, for years. Culture is that cumulative effect. And the boundedness is exactly why the methodology works the way it does: precise commitments rather than vague values, repeated relentlessly rather than launched once, concentrated in one domain past the threshold where the new language becomes the default. You do not need language to be destiny. You need it to be a bias, applied consistently, at scale. That is what a culture is.

It is also what McKinsey built without a single training course: five questions and a shared vocabulary, biasing tens of thousands of people’s thinking toward the same problem-solving approach, every day, for decades. The science did not exist as a package when the firm started doing it. The mechanism did.

If your organization’s vocabulary is biasing thousands of judgments a day, what is it biasing them toward?

The behavioral research points directly at the next question: if commitment language changes behavior, what makes some commitments stick while others fade? Read next: Why Commitments Work.