Cultural Building Blocks

Most frameworks for comparing cultures are catalogs. Hofstede gives you six dimensions, Schwartz seven, the GLOBE study nine. Each is a useful coordinate system for plotting where societies sit relative to one another. This is a different bet: that beneath the descriptive dimensions sit a smaller number of generative ones, and that much of what we call "a culture" is what those parameters produce when they run.

Six blocks, below. None of them is new. Each maps onto at least one well-developed research tradition, and I name those traditions throughout, because the receipts are the point. What's new is the compression: work from economics (Fukuyama, Greif, Mokyr), anthropology (Foster, Henrich), and cross-cultural psychology (Rotter, Nisbett, Schwartz, Gelfand), folded into six coupled parameters. The aim is parsimony: the smallest set of knobs that throws off the largest amount of surface culture.

One reflex this framework refuses up front: the insistence, drilled into cross-cultural research for fifty years, that cultures are "different, not better or worse." Banfield said otherwise in 1958 and got pilloried for it, and the field learned the lesson. But the data kept piling up on his side. Wide trust tracks prosperity and narrow trust tracks corruption; zero-sum thinking taxes everything a society could otherwise build; tightness shows up in innovation rates. Some settings of these knobs produce flourishing and some produce stagnation. Pretending otherwise was never rigor. It was ideology with a methods section.

Building blocks vs. dimensions

The word "dimensions" implies independent axes you measure and plot. "Building blocks" implies something that combines. A handful of underlying settings interact to generate the complex, idiosyncratic patterns we live as national or regional cultures, the way a genotype underdetermines but shapes a phenotype. That generative claim is the speculative part of this framework.

The six building blocks

1. Trust Radius

How far the default expectation of trustworthiness extends beyond the immediate family: to extended kin, then neighbours, co-ethnics, co-religionists, and at the far end strangers and abstract institutions.

This one parameter shapes whether exchange is impersonal or relationship-bound, whether institutions are formal or run on personal ties, the scale at which voluntary organizations and firms can form, the prevalence of corruption, and how freely information moves. Wide-radius societies can build large organizations and markets among strangers. Narrow-radius societies route economic and social life through family and known relationships.

Lineage. One of the most worked-over ideas in the social sciences. Banfield (1958) named "amoral familism" (the inability to cooperate beyond the nuclear family) the engine of a poor Italian village's stagnation. Fukuyama (1995) built his account of comparative prosperity on high- vs. low-trust societies, and used the phrase "radius of trust" directly. The construct sits inside the larger social-capital literature (Coleman 1988; Putnam 1993, 2000). Greif (1994) showed how collectivist vs. individualist trust orders produced different trading institutions among medieval Maghribi and Genoese merchants. Most strikingly, Henrich (2020) and Schulz et al. (2019) argue that the Western Church's centuries-long ban on cousin marriage dissolved intensive kinship in Europe and widened trust toward strangers. That's a concrete historical mechanism for moving this exact knob. And it has been operationalized: Delhey, Newton & Welzel (2011) measure trust radius straight from World Values Survey data.

2. Agency Location

Where a society's default worldview locates cause and control: in the individual, in the group, or in external forces (fate, gods, nature, "the system").

This shapes how people explain success and failure, the texture of religious life, the form of political organization, the framing of mental distress, and the perceived point of trying to change one's circumstances.

Lineage. Rotter's (1966) internal-vs-external locus of control is the psychometric ancestor. Rothbaum, Weisz & Snyder (1982) distinguished primary control (change the world) from secondary control (adjust the self to it) and argued cultures weight them differently. The cross-cultural attribution literature is the empirical core: Miller (1984) found Indians offer more situational and Americans more dispositional explanations for the same behaviour; Morris & Peng (1994) found the same split in how Chinese and American newspapers explained murders. Nisbett (Nisbett et al. 2001; The Geography of Thought, 2003) folds this into a broader contrast between holistic and analytic cognition, and Markus & Kitayama (1991) into independent vs. interdependent self-construal. Further back, Weber's account of how Calvinist ideas about agency and predestination reshaped economic conduct is a study of this parameter changing.

3. Resource Model

Whether value is felt to be fixed, so one person's gain is another's loss, or expandable, so the pie can grow.

This shapes the baseline toward cooperation or rivalry, attitudes to commerce and profit, risk and time preference, and how people read inequality, trade, and immigration. It's also the most live of the six. Zero-sum thinking is a quiet tax on everything a society could otherwise build, and that fault line runs straight through today's politics.

Lineage. Foster's (1965) "image of limited good" is the classic anthropological statement: peasant societies treat all desirable things (land, love, honour, health) as finite, so any individual's advance is presumed to come at the community's expense. The negotiation literature calls the cognitive version the "fixed-pie bias." Recent empirical work nails it down: Chinoy, Nunn, Sequeira & Stantcheva (2023; AER 2026) built a survey measure of zero-sum thinking across 20,000 Americans and traced it to ancestral mobility, immigration, and the legacy of enslavement, and found it predicts political views within a single society. On the positive-sum side, Mokyr (2016) and McCloskey (2010) argue that sustained growth needed a cultural shift first: people had to start treating wealth-creation as legitimate and open-ended, something a society makes.

4. Status Currency

What a society actually pays social esteem for: achievement, wealth, honour, piety, age and wisdom, harmony, lineage, or contribution to the group.

Whatever earns status is what people compete to accumulate, so this parameter quietly steers individual behaviour and the institutions that grow up to allocate the prize.

Lineage. Schwartz's (1992) theory of basic human values sorts much of this into achievement, power, benevolence, and tradition, and has been measured across dozens of societies. Henrich & Gil-White (2001) distinguish two universal routes to status, dominance (taken by force or intimidation) and prestige (freely conferred for valued skill), which cuts across the question of what the status is for. Nisbett & Cohen (1996) showed honour functioning as the dominant currency in specific ecologies (herding economies, the US South) and the violence that defends it. Weber's distinction between class (market position) and status group (social honour) sits underneath the whole question.

5. Change Cost

The perceived cost of deviating from established patterns: how tightly social norms bind, and how heavily a society punishes departure.

This sets the innovation rate, the room for social mobility and reinvention, the speed of adaptation to new conditions, and the intensity of generational conflict.

Lineage. The most thoroughly mapped of the six. Pelto (1968) first contrasted "tight" and "loose" societies. Gelfand et al. (2011, Science) operationalized tightness–looseness across 33 nations, linking tightness to histories of ecological and human threat (density, scarcity, conflict, disease). Schwartz's conservation-vs-openness-to-change axis and Hofstede's uncertainty avoidance and long-term orientation measure overlapping ground. Cultural-evolution models of conformist transmission (Boyd & Richerson 1985) supply the mechanism: how strongly a population copies the common variant.

6. Power Distance

How steeply power is stratified, and how legitimate that steepness feels. Is subordination accepted as the natural order, or does authority have to keep earning it?

This shapes the form of organizations and families, how openly juniors challenge seniors, the felt distance between rulers and ruled, and whether getting ahead is imagined as climbing a ladder or levelling a field. It's distinct from Status Currency: that block says what earns esteem; this one says how far apart the top and bottom sit, and how stable the gap is.

Lineage. The most-cited of Hofstede's (1980) dimensions, the power distance index. Schwartz's hierarchy-vs-egalitarianism orientation measures the same tension from the values side, as does Triandis's vertical vs. horizontal individualism and collectivism. The deeper root is the old question of the legitimacy of authority (Weber's typology of domination), and the GLOBE study (House et al. 2004) carries the construct into organizational life.

How the blocks combine

Calling these building blocks rather than dimensions is a claim that they interact to produce recognizable cultural types.

Consider a configuration: narrow trust radius, external/fatalistic agency, a zero-sum resource model, honour as the dominant currency, high change cost, steep and taken-for-granted hierarchy. Those settings travel together, and together they sketch the classic honour-based agrarian pattern that Banfield, Nisbett & Cohen, and Henrich each described from different starting points. Flip the knobs the other way (wide trust, internal agency, positive-sum resources, achievement currency, loose norms, a flatter hierarchy where authority is answerable) and you get something close to the self-image of a modern commercial democracy. These two don't exhaust the space. The point is that the labels we reach for, "honour culture," "high-trust society," are often shorthand for particular settings of a few shared parameters. The instrument above is just this idea made tactile.

That these configurations recur is the interesting part, and also the methodological problem. The six parameters are almost certainly not orthogonal. Henrich's work suggests a single historical cause, the dissolution of intensive kinship, can move trust radius, agency location, and several others at once. So I won't pretend the six are independent axes you can set freely. The honest claim is weaker and more interesting: these are coupled parameters whose common configurations explain a lot, and whose correlations are themselves a thing the framework should try to explain.

Relationship to existing frameworks

This framework is downstream of all of the following. What it adds is compression: a smaller, generative set.

And the broader comparative-culture frameworks this sits beside:

• Hofstede's cultural dimensions. The most-cited descriptive taxonomy (power distance, individualism, uncertainty avoidance, and so on). Broad coverage, with no claim that the dimensions are generative or minimal. Several of his dimensions feed the blocks above.
• Schwartz's value theory. Arguably the most rigorous of the lot, with basic individual values and three bipolar cultural orientations (embeddedness/autonomy, hierarchy/egalitarianism, mastery/harmony). Heavy overlap with Status Currency and Change Cost. This framework is partly an attempt to compress Schwartz further.
• GLOBE study (House et al. 2004) and Trompenaars & Hampden-Turner. Further dimension sets in the management-research tradition.
• Inglehart–Welzel cultural map. Two empirical axes (traditional/secular-rational, survival/self-expression) from World Values Survey data. Change Cost and Status Currency load onto these.
• Moral Foundations Theory (Haidt). Describes the moral intuitions (care, fairness, loyalty, authority, sanctity) that vary across groups. Largely orthogonal, and complementary.
• Grid–group cultural theory (Mary Douglas). Worth singling out, because its two axes map unusually cleanly: "grid" (how much rules and roles constrain a person) is close to Change Cost, and "group" (how strongly people are bound into a bounded collective) is close to Trust Radius.
• Cultural evolution theory (Boyd & Richerson; Henrich). Supplies the transmission mechanisms by which any of these are inherited and change. It explains how the knobs move, and leaves what the knobs are to a framework like this one.
• Edward Hall. High/low-context communication and monochronic/polychronic time, the kind of texture this framework deliberately leaves out.

What's deliberately left out

A minimalist framework earns its keep by what it excludes. Time orientation (Hall), gender-role differentiation (Hofstede's masculinity/femininity), and communication context were all considered and dropped, either as too contested or as derivable from the six above.

For a while the hardest omission was power distance: how steeply power is stratified, and how legitimate that feels. It kept refusing to reduce to the others. Status Currency says what status is awarded for; it stays silent on how far apart the top and bottom sit, or how much subordination people swallow as legitimate. So it's been promoted from footnote to the sixth block above. That the set grew once is worth remembering. It means the boundary is a judgement call, and some later reader may well find a seventh that earns its keep.

Measurement

Wherever possible the move is to reuse existing instruments. Another reason the lineage matters.

• Trust Radius. WVS generalized-trust items, refined by the radius method of Delhey, Newton & Welzel (2011); plus economic indicators (firm size, contract enforcement, corruption indices).
• Agency Location. Rotter's locus-of-control scale and attribution tasks; corpus analysis of how success and failure get explained in media and proverbs.
• Resource Model. The zero-sum-thinking battery from Chinoy et al. (2023); behavioural economic games (public-goods, ultimatum).
• Status Currency. The Schwartz Value Survey and Portrait Values Questionnaire; analysis of who a society lionizes (obituaries, honours, folk heroes).
• Change Cost. Gelfand's tightness scale; demographic and legal proxies for tolerance of deviance.

To these, add two cross-cutting sources. Text-as-data: embeddings of proverbs, religious texts, and contemporary corpora, to estimate the parameters from language. Behavioural and anthropological evidence: child-rearing, ritual, conflict resolution, gift-giving. An open empirical question for all six is temporal stability: how fast do these move, and in response to what?

Open questions

1. Orthogonality and clustering. How correlated are the six? Do they fall into a small number of stable configurations, and if so, what explains the clusters: shared ecology, a common historical cause (à la kinship intensity), or feedback among the parameters themselves?
2. Is "generative" more than a metaphor? The central bet is that surface cultural traits can be predicted from the six settings. That is testable and largely untested. If the blocks don't out-predict the existing dimension sets, the compression isn't buying anything.
3. Is six the right number? The set grew once already. The live question is whether it's still missing something, or whether Status Currency and Power Distance sit close enough to merge.
4. Causal direction and change. What actually moves these parameters (ecology, institutions, religion, economic structure), and over what timescales?
5. Within-society variation. None of these is monolithic. Class, region, urban/rural, and generation can differ as much within a society as between societies. A framework pitched at "cultures" has to say how it handles that variance.

Building from here

The proposal stands or falls on prediction. If knowing a society's six settings tells you more about its surface culture than the dimension catalogs do, the framework earns its keep; if not, it's one more taxonomy. The instruments exist, most of the data exists, and question 2 above is sitting there waiting for someone to run it. I'd like help finding out.

Thoughts or feedback? Send me a message on X.

Sources