Douglas proposed two concepts to frame her cross-cultural inquiry into societies: group and grid. The concepts are simple:
“Group is obvious—the experience of a bounded social unit.
Grid refers to rules which relate one person to others on an egocentred basis” (1970:viii). Group and grid are independent variables. Group and grid also are continuous variables; for
example, one could imagine a “group” as a sliding scale, varying
from “no sense of a bounded social unit” to “some sense of
bounded social unit” to “well-developed sense of bounded social unit.” In Douglas’s initial explanations, however, grid and
group are simplified as nominal variables—you live in either a
“weak-group” or “strong-group” society—and their relationships can be presented in a simple 2 x 2 table:
GROUP
When a society experiences grid and group (Cell C), “the
quality of relations is ordered and clearly bounded. If group is
found by itself (Cell D), or grid is found without group (Cell B),
the quality of relations is different” (Douglas 1970:viii). Douglas
further explained the two categories:
The group itself is defined in terms of the claims it makes over
its constituent members, the boundary it draws around them,
the rights it confers on them to use its name and other protec-
| B | C |
| A | D |
GRID
tions, and the levies and constraints it applies. . . . The term
grid suggests the cross-hatch of rules to which individuals are
subject in the course of their interaction. As a dimension, it
shows a progressive change in the mode of control. At the
strong end there are visible rules about space and time related
to social roles; at the other end, near zero, the formal classifications fade, and finally vanish. (1982a:191–192)
As an example of a strong-grid society, we can recall (see pp.
113–14) the traditional society of Bajoeng Gede, Bali, as described
by Bateson and Mead (1942). In Bali, Mead wrote about the complex rules concerning personal interactions, in which a “fixed
and complicated set of regulations, obligations, and privileges”
produces a grid in which “space and time and social status form
an orderly whole” (1942:10). In contrast to the Balinese stronggrid society, in weak-grid societies social “boundaries begin to be
arbitrated. Individuals, deciding to transact across them, weaken
the classifications” (Douglas 1982a:192). Paradoxically, as grid
weakens and as individuals more freely engage in social transactions, the rules governing such transactions may become more
explicit, legalistic, and increase in number. For example, one
might argue that modern American society is a weak-grid society since we elevate the individual and celebrate free choice in
economic and social realms, yet we regulate the behaviors and
expectations of interactions through our legal system.
Up to this point in her analysis, Douglas’s ideas are interesting but not really unique: numerous schemes have been used to
classify societies. What is significant is the way Douglas links
grid and group to other dimensions of culture, such as the economic and political expressions of differing social contexts; symbolic structures relating to the human body and society; and
cosmological statements regarding nature, time, human nature,
and social behavior (1970, 1982a, 1982b).
In situations with strong-grid/strong-group (Cell C), the individual’s social experience is defined first by the social boundary maintained between the group and outsiders and second by
the clear rules of behavior between group members. Individual
behavior is regulated in the name of the group. Within the group,
clearly defined social sectors exist (such as classes, castes, agegrades, and so on) that may have specialized roles and unequal
access to resources, and a variety of solutions will be used to resolve conflicts between those groups. Strong grid/strong group
societies will tend to be larger than other societies and last longer
without internal fissioning. Since their members perceive the
possibility of “persisting as a group into the future,” strong
group/strong grid societies can make levies of their members
(taxes, corvée, military service) to invest in their continued existence.
Strong group/weak grid societies (Cell D) also emphasize the
definition and authority of the group, but because of their lowgrid conditions they will lack formalized internal divisions or
segregated social sectors. Consequently, relationships between
individuals are ambiguous and the resolution of conflicts more
difficult. In the face of conflicts, the only penalty for internal disputes is expulsion from the group or the fissioning of the group.
This has several consequences: “disagreement is driven underground” since mechanisms for resolution are ill-developed.
Covert factions develop, and group members committed to the
maintenance of society will argue for stronger group boundaries
to control admission and strengthen the group. Consequently,
strong group/weak grid societies will tend to be small and subject to internal divisions (Douglas 1982a:205–206).
The two remaining possibilities, strong-grid/weak-group
and weak-grid/weak-group, have very different social outcomes.
In extreme cases of strong-grid/weak-group societies (Cell B), the
individual is tightly regulated by social rules and individual autonomy is minimized, but so is the individual’s affiliation with
any social groups, since those groups—by definition—do not exist. The individual’s role and behavior are unambiguously defined by powers that are remote, impersonal, and insulating
(Douglas 1982a:207).
In weak-grid/weak-group societies, “the social experience
of the individual is not constrained by any external boundary”
and is unregulated by ascribed status classifications (Douglas
1982a:207). All social classifications are potentially negotiable;
the relationships between individuals are ambiguous and their
mutual obligations implicit. Individuals can transact freely.
However, there will be an opposing tendency to regulate individual behaviors that violate social contracts, such as laws pro-
hibiting breach of contracts, rules protecting the free exercise of
rights, and legislation defending religious minorities. In such
societies, rewards go to innovators, economic activities specialize and expand, and the market is controlled by alliances. Nothing succeeds like success in weak-grid/weak-group societies,
and success is measured by the size of one’s following, whether
these are customers, political supporters, or fans (Douglas
1982a:207–208).
An analysis of grid and group is linked to broader issues of
symbolism. In essence Douglas asked, “If symbols mark systems
of classifications, and if classification systems are reflections of
social systems, then what types of classifications are associated
with which types of societies?” She argued that
the most important determinant of ritualism is the experience
of closed social groups. The man who has that experience associates boundaries with power and danger. The better defined and the more significant the social boundaries, the more
bias I would expect in favour of ritual. If the social groups are
weakly structured and their membership weak and fluctuating, then I would expect low value to be set on symbolic performance. (1970:14)
In strong-group/weak-grid societies, humanity is divided
into insiders and outsiders, and nature is similarly classified into
two classes: the lovable, vulnerable, cuddly part of nature and
the threatening, dangerous, untamable part of nature. The parallel dualities in humanity and nature are more than the extension
of metaphors but reflect “the use made of nature in moral justifications” (Douglas 1982a:210). Outside the village edge, beyond
the border of society, evil forces lurk waiting to penetrate the human realm. Within the community, one must find the contaminated agent of evil, drive him out, and purify the realm of
humanity again.
In strong-group/strong-grid societies, the same emphasis
on society’s borders is present, but it is modified by the presence of strong grid. The explicit rules about the interrelations
between group members are justified by a “transcendental
metaphysics which seeks to make an explicit match between
civilization and the purposes of God and nature” (Douglas
1982a:210). When social relations are confirmed by ritual, the
laws of nature are used for moral justification. In strong
group/strong grid societies, theoretical models are elaborated,
divine sacrifice is highly developed, and social contexts are justified by appeals to “natural law” and cosmological analyses
(Douglas 1982a:210–211).
In weak-group/strong-grid societies, theoretical notions of
society and the cosmos are undeveloped, and cosmologies are
incoherent and eclectic “things of shreds and patches” (Douglas
1982a:211). Finally, in weak-group/weak-grid societies, the unrelenting competition and selection exercised by individuals is
expressed in cosmologies that convey the “excitement and rewards of competition” (Douglas 1982a:212).
In addition, since success is measured by the size of one’s following in weak-group/weak-grid cases, society is a source of
constant concern because approval, once granted, also may be
withdrawn. Because following is based on performance, there is
intense demand for high standards: our movie stars can never
make a flop, our presidents cannot be adulterers.
Nature stands in contrast to the competition of weak-grid/
weak-group society. Outside the realm of humanity and an oasis
of innocence, nature is not cited to justify social relations since
nature and society are separate realms. Yet “a wistful sense of
alienation from nature never wins against the excitement and rewards of competition” (Douglas 1982a:212).
Building on the framework shaped by group and grid, Douglas offered an amazing list of predictions about such different
topics as cookery (“In D the classification of edible foods is
likely to assert the prohibition against eating carnivores”), gardening (“D will not be specially interested in gardening as a cultural activity, whereas C is likely to use this medium to justify
and expand its view of society as hierarchized, trained and compartmentalized”), monuments (“though it has nothing much to
raise public monuments to, B respects the burial places of its
private dead”), and attitudes about youth and old age, death,
personal abnormalities and handicaps, punishment, and justice
(1982a:214–226).
But if we accept Douglas’s predictions as plausible, how do
we decide where Society X fits on the grid/group diagram?
First, grid/group analysis assumes that society is the product of
interactions between the individual and her/his environment,
the environment consisting of “all the other interacting individuals and their choices.” A group does not exist “independently
of the volition of its independent members. Their investment of
time and energy quickens its life and marks its boundaries”
(Douglas 1982a:198). Alternately, as appeals are made to “an
ethic of individual value,” grid diminishes in force.
Second, the analysis of grid/group occurs at the level “at
which people find it necessary to explain to each other why they
behave as they do” (Douglas 1982a:201). This level—“the social
accounting level, the level of justification and explanation”—
can be observed and investigated by the anthropologist (Douglas 1982a:201). For example, if an individual conducts the
varying facets of life—residence, work, shared resources, marriage, and kinship—within a common social set, then group is
strong. If I spend the early morning as a member of my family,
my day as an employee, then return to my family most evenings
except for the one night a month I join a poker party and the
other evenings I spend as a member of a homeowners’ association, as a seminar participant, as a PTA member, and other socially disconnected networks, then group is weak. Similarly, if I
justify my actions by appeals to the rights of the individual or,
alternately, to my expected roles as a member of a particular social unit, then grid is varyingly weak or strong. None of this requires anthropologists to restructure their basic methods, and it
produces tantalizing lines of inquiry.
Fundamentally, Douglas outlined an approach in which
“most values and beliefs can be analyzed as part of society instead of as a separate cultural sphere” (1982b:7). Group/grid
analysis, Douglas wrote,
is a method of identifying cultural bias, of finding an array of
beliefs locked together into relational patterns. The beliefs
must be treated as part of the action, and not separated from it
as in so many theories of social action. The action or social context, is placed on a two-dimensional map with moral judgements, excuses, complaints and shifts of interest reckoned as
the spoken justifications by individuals of the action they feel
required to take. As their subjective perception of the scene
and its moral implications emanates from each of them individually, it constitutes a collective moral consciousness about
man and his place in the universe. (1982a:199–200)