Groups - Primary and Secondary

Statuses and roles describe how an individual is positioned (2.2.1–2.2.3). But people are not arranged one at a time. They cluster — and the clusters have properties that no description of individuals can supply.

Mr Deshpande retires on a Friday after thirty-one years at the same company.

There is a function. The general manager, who has known him for four years, gives a speech containing two accurate facts and one anecdote supplied by someone else. He is given a clock and a shawl. Forty people applaud. Several of them have worked beside him for a decade and know his wife's name, his son's marks, and the year of his father's death.

On Monday he does not go to the office. Nobody expects him to.

On Tuesday nobody calls. This is not cruelty. The people at that office were connected to him through a position, and the position is now occupied by someone else. Within three weeks the daily texture of his life there — who brought whose tea, the joke about the third-floor lift, the man who covered for him the week his father was ill — has simply stopped, and the machine has closed over the gap.

Meanwhile, on Wednesday, one of four men he has known since he was eleven telephones for no reason. They speak for forty minutes about nothing. They have been doing this for fifty years. None of them can remember what any of the others does for a living without thinking about it.

Two kinds of group. One shaped who he is. The other shaped what he did all day for three decades.

First: what is a group?

Before the distinction, a clarification that beginners routinely miss. Not every collection of people is a group.

The distinction is analytically important because these three behave completely differently. An aggregate cannot act. A category cannot act either — but it can become a group, and the process by which a category acquires consciousness, organisation and the capacity to act collectively is one of the central questions in sociology.

You have already met it twice under other names: Marx's class-in-itself becoming class-for-itself (4.1), and Mills' translation of troubles into issues (1.1.1). It is also what every social movement attempts (11.2). A category is a statistical fact; a group is a social force; the transformation is the whole game.

Cooley's distinction

Charles Horton Cooley (1909) — whom you met in 1.1.3 as the author of the looking-glass self — gave the classic contrast.

Primary groups are small, enduring, face-to-face associations characterised by intimate cooperation and a strong sense of "we." Their key features:

Small enough for everyone to know everyone.
Enduring, often lifelong.
Face-to-face (in Cooley's era; see the discussion below).
The whole person is involved — not a segment. Your family knows you tired, ill, wrong and unimpressive.
The relationship is an end in itself, not a means to something else.
Members are not interchangeable. You cannot replace a friend with an equally qualified friend.

The family, the childhood friendship circle, a village neighbourhood, a close work team over many years.

Secondary groups are larger, more impersonal, and organised around a specific purpose. Their features are the mirror image:

Larger, often too large for everyone to know everyone.
Time-limited, lasting as long as the purpose does.
Segmental — only the relevant part of the person is engaged.
Instrumental — the relationship is a means to an end.
Members are interchangeable, which is precisely what makes organisations work: the post outlives the holder (2.2.1).

A company, a university, a political party, a professional association, a trade union, a queue at a government office.

Cooley's larger claim, and the reason he called them primary, was developmental: primary groups are where human nature is formed. "They are the nursery of human nature," he wrote — the site where a self, a conscience and a capacity for sympathy are produced. That is 1.1.3's internalisation and Part 2.3's socialisation, seen from the group's side.

And note the macro version you already have: Tönnies' contrast between Gemeinschaft (community — relationships as ends, bound by kinship, place and shared understanding) and Gesellschaft (association — relationships as means, bound by contract and calculation), which frames whole societies the way Cooley frames groups (4.6, 11.4).

The most instructive finding about groups in the industrial-sociology literature came by accident.

Between 1924 and 1932, researchers at the Western Electric Hawthorne Works near Chicago ran a long series of studies on productivity. One of them, the bank wiring observation room, simply watched fourteen men wiring telephone equipment, with an observer present for months.

The company had a wage scheme designed to reward higher output. It did not work as intended, and the reason was that the fourteen men had formed a primary group inside the secondary organisation, with its own norms, its own sanctions, and its own idea of a fair day's work.

They had settled on an informal output quota — well below what management expected and what the men could produce. The norms were explicit within the group and unwritten anywhere. Someone who exceeded the quota was a rate-buster. Someone who fell short was a chiseler, free-riding on others. Someone who reported a colleague to a supervisor was a squealer. Enforcement was by ridicule, exclusion, and a practice they called binging — a sharp flick on the upper arm, delivered half in play.

Every element of 2.1.3 and 2.1.4 is present: norms, argot, informal sanctions, and an enforcement system that outperformed the company's financial incentives.

Two cautions, because this study is often taught badly. The broader "Hawthorne effect" — the claim that being observed by itself raises productivity — has been substantially challenged by re-analyses of the original data, and should be treated as contested rather than established. And the researchers' own managerial framing has been criticised for treating workers' restriction of output as irrational sentiment rather than as a rational response to piece-rate systems that had historically been re-cut once output rose.

What survives, and survives well, is the core finding: secondary organisations are riddled with primary groups, and those informal groups frequently determine what actually happens. Anyone who has worked anywhere knows this. It is also why organisational reform that ignores informal structure reliably fails.

Go deeper

Why primary groups matter more than they look

Three findings worth carrying.

They do the socialising. Cooley's point (2.3): the self is formed in intimate groups, and no secondary group can substitute. This is why the family remains, in every society studied, the most consequential institution for life outcomes.

They carry people through hardship. Research on soldiers in combat — Shils and Janowitz's classic wartime study of the German army, and much subsequent work — found that what sustained men under conditions where ideology and fear both predicted collapse was the small primary group: the squad, the eight or ten men who would notice. This finding has been refined and argued over since, but its core has held: primary bonds, not abstract commitments, are what people endure for.

They are where informal power lives. Hawthorne is one instance. The general principle: in any large organisation, the formal chart tells you the official structure and the primary groups tell you what will actually happen. Which is 1.5.2's debunking, applied to organisations.

Not a historical replacement

A correction to a common misreading, including of Tönnies.

The story usually told is: traditional societies were made of primary groups; modernity replaced them with secondary ones; hence anonymity, loneliness and the loss of community. There is something real in this — urbanisation and mobility genuinely change the density of primary ties (11.4) — and as stated it is wrong in three ways.

Primary groups have not disappeared; they have relocated and multiplied. Families, friendship circles, work teams and congregations are everywhere. What changed is that people now choose many of them rather than inheriting them by place — which is a change in the basis of primary ties, not their abolition.

Modern institutions are dense with primary groups. See Hawthorne. The secondary organisation is the container, not the content.

And "loss of community" is a claim that keeps being made in every generation and needs checking each time rather than assuming (11.4 examines the evidence, which is more mixed than the nostalgia suggests).

Are online groups primary?

A live question, and worth handling carefully rather than with a reflex.

Cooley's criterion was face-to-face interaction, which digital groups by definition lack. But the functional criteria — endurance, whole-person involvement, non-interchangeability, being an end in itself, the strong "we" — can plainly be met online. Long-standing friendships maintained by messaging, communities formed around rare illnesses, diasporic groups holding a family together across three countries, groups of people who have never met and would grieve each other.

Three honest observations rather than a verdict.

Digital ties often maintain primary groups formed offline, which is a different thing from creating them. Some online groups clearly do meet the functional criteria and it would be dogmatic to deny it. And the platform shapes what is possible: a group whose interaction is public and permanent behaves differently from one that is private and ephemeral, because 2.1.4's sanctions operate differently when there is a permanent record.

The sensible position is to treat "primary" as a set of properties a relationship may have to a greater or lesser degree, rather than a category defined by physical co-presence. Part 9.7 and 12.3 take this up properly.

Neither type is inherently good

The most important corrective in this lesson, and it is easy to miss because the vocabulary is warm.

Primary groups are not automatically benign. The most intimate group most people belong to is also the setting in which they are most likely to be hit, controlled, coerced into marriage, or prevented from leaving. Families enforce caste endogamy, gendered obligation and silence about abuse — precisely because they are primary: the sanctions available (love, shame, exclusion, belonging) are the strongest ones there are (2.1.4). Feminist sociology's insistence on studying the household as a site of power (1.4.4, 9.1) rests on exactly this.

Secondary groups are not automatically cold. Impersonality has real virtues. A hospital that treats you as a case rather than as a relative treats everyone by the same rule. Weber's account of bureaucracy (4.3) makes this explicitly: impersonality is what makes fairness at scale possible, and its absence is what we call nepotism. A world of only primary relations would be a world in which everything depended on whom you knew — which is, in fact, the complaint people make about institutions where primary ties dominate.

The honest formulation: primary groups provide belonging and identity and are the most effective machinery of control ever devised; secondary groups provide fairness at scale and cannot supply meaning. Every society needs both, and the interesting question is which does what.

"Any collection of people is a group." Aggregates and categories are not groups. The transformation of a category into an acting group is one of the discipline's central questions.

"Primary means important, secondary means unimportant." Primary means formative — first in the making of a self. You may spend far more hours in secondary groups.

"Modernity destroyed primary groups." They relocated and multiplied; they also fill the inside of every large organisation.

"Primary groups are warm and supportive." They are also where most violence, control and coercion occur, and for the same structural reason: their sanctions are the most powerful available.

"Secondary groups are alienating." Impersonality is what makes equal treatment at scale possible. Weber saw both the cost and the achievement.

"Online groups can't be primary." Judge by properties — endurance, whole-person involvement, non-interchangeability — rather than by physical co-presence.

Why it matters

It tells you which sanctions are available in a setting. In a primary group, the enforcement instruments are belonging, shame and exclusion — devastating and unappealable (2.1.4). In a secondary group they are formal and limited. This single observation explains a great deal about where people conform most.

It explains why organisations behave unpredictably. The formal structure is the plan; the primary groups inside it are the reality. Any reform designed only against the chart will be absorbed by the informal group and quietly neutralised.

It locates where a self comes from. Cooley's nursery, and Part 2.3's whole argument. If you want to know how someone became who they are, ask about their primary groups, not their institutions.

And it prepares the next two lessons, which take the internal machinery of groups apart: how membership itself creates us-and-them (2.2.5), and how the sheer number of members transforms what a group can be (2.2.6).

Not every collection is a group. An aggregate shares a place; a category shares a characteristic; a group interacts, shares identity, and has mutual expectations. Turning a category into an acting group is a central sociological question (class-for-itself; social movements).

Cooley: primary groups are small, enduring, intimate, whole-person, non-interchangeable, and ends in themselves — "the nursery of human nature." Secondary groups are larger, purposive, segmental, instrumental and interchangeable.

Tönnies' Gemeinschaft/Gesellschaft is the same contrast at the scale of whole societies.

The Hawthorne bank wiring room shows primary groups operating inside a secondary organisation, with their own norms (rate-buster, chiseler, squealer) and sanctions (ridicule, binging) that overrode the company's incentives. The general "Hawthorne effect" is contested; this finding is not.

Primary groups do the socialising, carry people through hardship (Shils and Janowitz on combat units), and are where informal organisational power lives.

Modernity relocated primary groups rather than abolishing them; online ties are better judged by properties than by co-presence.

Neither type is inherently good: primary groups possess the most powerful sanctions in existence, and secondary impersonality is what makes fairness at scale possible.

Social group — two or more people who interact, share identity, and hold expectations of one another.

Aggregate — people co-present without ongoing interaction or shared identity.

Social category — people sharing a characteristic without necessarily interacting.

Primary group — a small, enduring, intimate group involving the whole person, valued as an end in itself (Cooley).

Secondary group — a larger, purposive, impersonal group whose members are functionally interchangeable.

Gemeinschaft / Gesellschaft — Tönnies' contrast between community bound by shared understanding and association bound by contract.

Informal group — a primary group arising spontaneously within a formal organisation.

Rate-buster / chiseler / squealer — the Hawthorne workers' argot for those who exceeded, fell short of, or informed on the group's output norm.

Class-in-itself / class-for-itself — Marx's distinction between a shared structural position and organised collective consciousness (4.1).

Unit cohesion — the primary-group bonding found to sustain soldiers in combat (Shils and Janowitz).