---
title: "The Lewis Model: why the Finn stays silent, the Italian interrupts and the German asks for a spec"
date: 2026-09-28T12:00
tags: [management, psychology, culture, hiring]
description: "Richard Lewis splits cultures into linear-active, multi-active and reactive. Where those differences come from (terrain, crops, religion), where the model lies, and how to use it in hiring, sales and code review."
embedding:
  version: 1
  model: "text-embedding-3-small"
  dimensions: 512
  hash: "sha256:84584e0b6bdf54c5a3ad2f9aa47dd054e0a052e7affb1680c443c93a08e5b773"
  vector: "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"
  scale: 0.00143334
---

Six people in a meeting. The German asks where the spec and the acceptance criteria are. The Italian cuts in on the third word and spends a minute explaining how this was already solved in a previous project. The Finn says nothing. After four seconds of silence someone decides the Finn did not understand the question and starts explaining again. The Finn stays silent a bit longer. Then says one sentence that closes half the discussion.

None of them is rude, slow or passive. They just hold different beliefs about **why people open their mouths at all**.

Richard Lewis - a British linguist who spent two years in Finland and five in Japan - packed that observation into a three-category model. It is crude, consultants love it, academics attack it, and it still explains the meeting above better than most of the more careful theories.

<!--truncate-->

## Who Lewis is and where the model comes from

[Richard Donald Lewis](https://en.wikipedia.org/wiki/Richard_D._Lewis) (b. 1930) comes from a Lancashire mining family and studied modern languages at Nottingham. He came to the 1952 Helsinki Olympics, stayed in Finland for two years and learned Finnish. In 1955 he opened a Berlitz school there, then five more, then schools in Norway and Portugal. In 1966 he opened Berlitz in Tokyo, lived in Japan for five years and tutored [Empress Michiko](https://en.wikipedia.org/wiki/Empress_Michiko). By 2017 he claimed ten languages.

That matters for reading the model: it did not grow out of questionnaires but out of thirty years of work as an interpreter and trainer who watched negotiations break down over nothing.

The root of the model is [Edward Hall](https://en.wikipedia.org/wiki/Edward_T._Hall): **monochronic** cultures (one thing at a time, time is a line) and **polychronic** ones (many things at once, time is a cloud). Lewis widened these into *linear-active* and *multi-active*, and after Japan [added a third category](https://www.crossculture.com/about-us/the-model/) - *reactive* - which did not fit Hall's scheme at all.

The main book is [When Cultures Collide](https://www.amazon.com/When-Cultures-Collide-Leading-Across/dp/1473684822) (1996, latest edition 2018). For teams there is *When Teams Collide* (2012), for corporate strategy *Fish Can't See Water* (2013, management book of the year in Denmark).

## Three corners

The model is drawn as a triangle. Lewis picked the colours deliberately: **blue** for cool factual planners, **red** for warmth, emotion and talkativeness, **yellow** for the soft harmony of courteous listeners.

![The Lewis Model triangle: linear-active, multi-active, reactive](img/lewis-model-triangle-en.png)

*My own redraw of the triangle based on Lewis' descriptions. Country positions are approximate.*

### Linear-active (blue)

Germany, Switzerland, Austria, Scandinavia, the Netherlands, Britain, the US, Canada, Australia. Nearby: Czechia, Estonia, Latvia, Lithuania.

- Task before relationship. One thing at a time, a linear agenda.
- Speech is information exchange. They talk and listen in roughly equal proportion.
- Truthful rather than diplomatic. Not afraid of confrontation, stick to logic over emotion.
- Believe in rules, contracts and official channels. Dislike "getting it done through a friend".
- Punctuality, quality, reliable delivery dates. They pay on time.
- Status comes from achievement. The boss is low key.
- "A good product sells itself" - and that is also their blind spot: they often fail to see that in other cultures sales rest on relationships.
- [Rationalism and science dominate their thinking more than religion](https://www.crossculture.com/about-us/the-model/linear-active/).

### Multi-active (red)

Southern Europe (Italy, Spain, Portugal, Greece), Latin America, the Arab world, Africa, the Balkans, Poland, Hungary, Romania.

- Relationships and feelings before the task. Many things at once, an agenda feels like a cage.
- Conversation is roundabout and animated. Frequent interruptions, almost no tolerance for pauses.
- Information comes **from people**, not from a document. Rumour and gossip are a working channel.
- Less respect for official announcements and rules, more for their own place in the hierarchy. They like a strong boss who also protects them.
- Late with deadlines and payments. Plan in less detail, but improvise well and handle chaos.
- In negotiation: charisma, rhetoric, negotiated truth, generous hospitality and gifts.
- [They value privacy less than company](https://www.crossculture.com/about-us/the-model/multi-active/), and adhere far less to strict Protestant values.

### Reactive (yellow)

Japan, China, Korea, Taiwan, Vietnam, Singapore, Hong Kong, Thailand, Indonesia, Malaysia, the Philippines. In Europe: Finland.

This is the most underrated category, and Lewis describes it in the most detail.

- **They listen before they leap.** First establish the other's position, then form their own.
- They do not interrupt. After you finish - **a pause**. The pause does not mean "nothing to say", it means "your words carry weight and are being considered".
- They reply with clarifying questions rather than opinions.
- Communication mode: monologue - pause - reflection - monologue. Blues and reds run a dialogue with constant interjections instead.
- They rely on non-verbal signals and speak in half-utterances, assuming the listener will fill in the rest. That is a form of compliment.
- They use impersonal and passive constructions ("one of the machines seems to have been tampered with") to avoid assigning blame.
- Self-disparagement is a tactic. Westerners wrongly read it as a weak position.
- [Small talk does not come easily](https://www.crossculture.com/about-us/the-model/reactive/). "Well, how is it going?" may get an honest complaint, because it is a direct question.

Lewis quotes a Bank of Finland delegation asking: "how can we make an impact?" at international meetings. That is not shyness. It is a protocol collision in which whoever respects the pause always loses to whoever cannot stand it.

## It is a spectrum, not three boxes

The thing that gets lost in retellings: countries sit **on the edges** of the triangle, not in its corners.

- **Finland** sits on the linear-active ↔ reactive edge. German punctuality plus Japanese pause. Lewis wrote a whole book about it - *Finland, Cultural Lone Wolf*.
- **Russia** is on the multi-active ↔ reactive edge: relationships and informal connections like the reds, but with strong patience, opacity and "let us see what the other side says" like the yellows.
- **Estonia** is closer to the blue corner with a shift towards yellow, from the Finnish neighbourhood. Hence the local reputation: everything by the rules and on time, but small talk and "sell yourself in the meeting" fall flat.
- **France and Belgium** sit between blue and red. **India and Turkey** between red and yellow.

Different editions of Lewis' diagram place countries slightly differently, and that is itself a signal: the model has no exact coordinates.

## Where it comes from: terrain, crops, plague, faith

Lewis describes **what** happens but barely explains **why**. The "why" belongs to a different literature - and there the link to geography and tradition becomes tangible.

### Rice and wheat

The cleanest study on this is [Talhelm et al., Science, 2014](https://www.science.org/doi/10.1126/science.1246850). The authors did not compare East with West, they compared **China with China**: rice-growing southern regions against wheat-growing northern ones. Rice needs shared irrigation systems and about twice the labour per unit of land - neighbours are forced to coordinate planting and share water. Wheat needs rain and your own field.

The result: people in rice regions show more interdependent thinking, avoid conflict more, divorce less and patent less. Same country, same language, same government.

That is literally "reactive versus linear-active", explained by irrigation rather than national character.

### The plough

[Alesina, Giuliano, Nunn, QJE, 2013](https://scholar.harvard.edu/sites/scholar.harvard.edu/files/nunn/files/alesina_giuliano_nunn_qje_2013.pdf) tested another hypothesis: where historical agriculture required a **plough** (heavy soil, upper-body strength, control of a large animal), the work went to men and women stayed in the home. Where shifting hoe cultivation was practised, women worked the fields alongside them.

Centuries later, and across the migration of descendants to other countries, this still predicts attitudes towards working women. In other words, **the shape of a farming tool predicts the composition of your team today**.

### Herders and the culture of honour

[Nisbett & Cohen, Culture of Honor](https://books.google.com/books/about/Culture_Of_Honor.html?id=vBI4DgAAQBAJ) (1996): a farmer's crop cannot be stolen overnight, a herder's flock can. In mountains and dry pastures, where the state is absent, the only protection is a reputation for answering an insult immediately. Hence cultures of honour: the Caucasus, the Balkans, Sicily, the Middle East, the American South.

In business terms that means **blunt public criticism is not "honesty" there, it is an attack on status** - and what you get is not a discussion but a vendetta. Blues usually find this out too late.

### Pathogens and density

[Fincher, Thornhill et al., Proc. Royal Society B, 2008](https://royalsocietypublishing.org/doi/abs/10.1098/rspb.2008.0094): historical pathogen prevalence predicts collectivism. The logic is simple - where infection is common, xenophobia and strict ritual norms work as a behavioural immune system: avoid the stranger, prepare food only this way, follow the purity rules.

[Gelfand et al., Science, 2011](https://www.science.org/doi/10.1126/science.1197754), a 33-nation study, generalises this into **tightness-looseness**: the more historical threats (population density, scarce resources, disasters, epidemics, invasions), the "tighter" the culture - many strong norms and low tolerance of deviance. Japan and Singapore are tight, the Netherlands and Brazil are loose.

### Religion

Religion here is less a cause than a **preserved form** of economic history - but a very durable form.

- Weber and the Protestant ethic: work as calling, contract as obligation, modesty in consumption, distrust of any intermediary between you and the truth. Lewis writes explicitly that multi-actives hold little Protestant strictness.
- Catholic and Orthodox societies produce a denser family and community structure: confession, forgiveness, a patron. Truth there is negotiated rather than absolute.
- Confucianism: hierarchy, face, harmony and a long horizon. A direct "no" breaks face on both sides, so it is replaced by a pause and a clarifying question.
- Islam: a strong role for the spoken word and personal vouching, relationship before contract.

There is a sharper test too. [Putnam](https://press.princeton.edu/books/paperback/9780691037387/making-democracy-work), comparing northern and southern Italy, showed that the gap in institutional quality persists for centuries and traces back to medieval city communes against the Norman feudal south. One country, one religion, one language - and completely different trust in strangers.

Lewis himself said something softer in the same direction: [there are no universal values](https://www.businet.org.uk/assets/media/RichardLewis_UniversalValues.pdf), even "truth" is built differently. For a German or a Finn the truth is the truth. In Britain or Japan it is fine as long as it does not rock the boat.

## Where the model lies

You should use the Lewis Model knowing its weak points. Otherwise you get polite racism with a diagram instead of a tool.

**1. No published methodology.** Hofstede and GLOBE have questionnaires, samples and data you can attack. The Lewis triangle has neither - it is a consultant's generalised experience. Academic citations exist (for example [Ursula Ott's work on negotiation](http://irep.ntu.ac.uk/id/eprint/34938/1/12504_Ott.pdf)), but that is application of the model, not validation of it.

**2. The ecological fallacy.** [Brewer & Venaik, 2014](https://doi.org/10.1177/0170840613517602) lay out the core methodological problem of all national "dimensions": a score computed at country level **cannot be applied to a person from that country**. The average temperature of a country tells you nothing about the temperature in your room. The same authors [go further](https://doi.org/10.1177/1350507616629356) and argue such dimensions breed cultural ignorance rather than cure it.

**3. Within-country variance beats between-country variance.** A Tartu professor and a Tallinn salesperson differ from each other more than the "average Estonian" differs from the "average Latvian". Talhelm's work is exactly about this: there is more variation inside China than between China and the West.

**4. The model is static.** It describes the world Lewis observed in the 1960s-90s. Remote work, English as the working language, international education and a shared professional culture sand down the corners. A developer from Tokyo in an international team often behaves "bluer" than his own engineer father.

**5. National culture is only one layer.** There is also corporate (Amazon vs SAP), professional (developers vs sales - each with their own linear and multi), regional, family and personal. The professional layer is often stronger than the national one.

**6. The triangle can flatter.** The "rational planners" corner conveniently holds the countries the consultants and their clients come from; everyone else gets "emotional", "late", "gives gifts". Postcolonial critics [went after Hofstede](https://doi.org/10.2167/md051.0) for precisely this. When reading the multi-active descriptions, keep in mind who wrote them and who they were sold to.

And to be fair: [Lewis himself acknowledges](https://en.wikipedia.org/wiki/Richard_D._Lewis) these are stereotypes, and defends the model as a practical frame rather than a truth.

## What to do with it in business

The model is useful in exactly one mode: as a **list of hypotheses about what might go wrong**, plus a set of ready corrections.

### Clients and sales

- A blue client is sold by a document: spec, benchmark, SLA, written references. Lunch is a waste of their time.
- A red client is sold by a person. Until there is personal contact, the spec goes unread. Meet in person, call rather than write.
- A yellow client is sold by time. The first meetings are not about the deal but about whether you will still exist in five years. The absence of a quick yes is not a no.
- A contract is the end of negotiation for a blue and the start of a renegotiable relationship for a red. Most mutual grievances about "they do not keep their word" are born here.
- Payment discipline is a cultural variable, not a measure of decency. Plan cash flow accordingly.

### Hiring and interviews

- A pause before an answer is not a knowledge gap. Give the candidate those five seconds and do not fill them yourself.
- "Sell yourself" and "tell me about your achievements" is a filter that selects multi-actives and cuts reactives and northern linear-actives. If that was not your intent, you did it anyway.
- On the format-under-pressure problem specifically I wrote in ["Why live coding is a bad way to test knowledge"](/ru/blog/управление/pochemu-layv-koding-plohoy-sposob-proverki-znaniy/) (in Russian). Cultural differences in tolerance for public silence amplify that effect.
- Ask about behaviour in specific past situations rather than about self-description. Self-description depends far too much on the cultural norm of modesty.

### Feedback and code review

This is where the differences bite daily. "This is wrong, rewrite it" is a normal engineering sentence to a blue, public humiliation to a red, and a loss of face to a yellow. And the reverse: "maybe we could consider another approach, if you have time" reads to a blue as "no changes required".

What works in practice: agree **as a team** on explicit levels ("blocking / suggestion / taste"), automate everything that can be automated, and take humans out of style arguments. More in ["Code review and conflict in team dynamics"](/ru/blog/управление/code-review-i-konflikt-v-dinamike-komandy/) (in Russian).

### Meetings and distributed teams

- A live meeting with an open mic is a multi-active arena. Whoever speaks fastest defines the decision.
- Async writing is a linear-active arena. Whoever writes more structurally wins.
- Reactives lose in both formats, and the most careful risk assessment in the team leaves with them.
- The corrections are simple: a pre-read with the proposed decision, an explicit round of every participant, a written round after the meeting, and a "no decision recorded before the next day" rule for big things.

### Roles, composition and team stages

The Lewis Model maps well onto the **stage** of a team. Differences peak at forming and storming, because shared norms do not exist yet and everyone runs their home protocol. I have an old university write-up on stages and role allocation (RUP roles, layered and mixed schemes) - ["Estimating project complexity, quality and team"](/ru/blog/универ/otsenka-slozhnosti-kachestva-i-kollektiva-proekta/) (in Russian). It also covers why groups of 3-7 work best.

An important caveat to "let us hire a diverse team": diversity by itself does not produce results, it produces more options and more friction. It only works together with explicit norms - who speaks when, what counts as a decision, how disagreement is expressed. Without that, a diverse team simply adopts the norms of the loudest member.

Another lens on the same thing is [spiral dynamics](/ru/blog/психология/spiralnaya-dinamika-greyvza/) (in Russian): useful as a spectrum of values, harmful when used as a ladder someone must be pulled up. The Lewis Model has exactly the same trap.

### Goals

All of the above collapses if the team has no aligned goal - then cultural differences stop being noise and become the explanation for why "they" do not work. On how goals cascade and where they tear, see ["Business goals and alignment"](/en/blog/management/biznes-celi-i-soglasovannost/).

## How to use it without doing harm

1. **A hypothesis, not a label.** "His silence may be respect rather than doubt" is a working thought. "He is Finnish, so he is quiet" is not.
2. **Start with yourself.** The title *Fish Can't See Water* is about this: your own protocol is invisible because it feels like "just normal". You are someone else's stereotype too.
3. **Ask the person.** A direct "how do you prefer to get criticism - in review, in DM, or by voice?" beats any diagram.
4. **Change the process, not the people.** You will not rewrite someone's culture, but you can rewrite a meeting format in five minutes.
5. **Drop the model once you know the person.** A stereotype is low resolution. As soon as you have data about an actual colleague, it overrides the country.

The Lewis Model does not describe people. It describes the **default interface** a person brings into a conversation. Interfaces can be negotiated - you just have to know they differ.

## Sources

- Richard D. Lewis, *When Cultures Collide: Leading Across Cultures*, 4th ed., 2018
- Richard D. Lewis, *When Teams Collide: Managing the International Team Successfully*, 2012
- [The Lewis Model](https://www.crossculture.com/about-us/the-model/), Richard Lewis Communications
- [Talhelm et al., "Large-Scale Psychological Differences Within China Explained by Rice Versus Wheat Agriculture"](https://www.science.org/doi/10.1126/science.1246850), Science, 2014
- [Gelfand et al., "Differences Between Tight and Loose Cultures: A 33-Nation Study"](https://www.science.org/doi/10.1126/science.1197754), Science, 2011
- [Alesina, Giuliano, Nunn, "On the Origins of Gender Roles: Women and the Plough"](https://scholar.harvard.edu/sites/scholar.harvard.edu/files/nunn/files/alesina_giuliano_nunn_qje_2013.pdf), QJE, 2013
- [Fincher, Thornhill, Murray, Schaller, "Pathogen prevalence predicts human cross-cultural variability in individualism/collectivism"](https://royalsocietypublishing.org/doi/abs/10.1098/rspb.2008.0094), Proc. R. Soc. B, 2008
- Nisbett & Cohen, *Culture of Honor*, 1996
- [Brewer & Venaik, "The Ecological Fallacy in National Culture Research"](https://doi.org/10.1177/0170840613517602), Organization Studies, 2014
- [McSweeney, "A triumph of faith - a failure of analysis"](https://doi.org/10.1177/0018726702551004), Human Relations, 2002