When describing 'a culture', anthropologists are trying to capture what is shared, the code of shared rules and common meanings. 'A culture' is always a composite, an abstraction created as an analytical simplification. In the real world, the knowledge described as cultural is always distributed among individuals in communities.

There is a danger of taking this abstraction as having a concreteness, an existence as an entity and a causal agent 'it' cannot have. Although both specialists and non-specialists alike are prone to talk about 'a culture' as if it could be a conscious being (e.g. "American culture values individuality") or to talk as if 'a culture' were like a group, i.e. something one could 'belong to' (e.g. "a member of another culture"), we need to guard against the temptation to reify and falsely concretise culture as a 'thing' and remember that 'it' is a strategically useful abstraction from the distributed knowledge of individuals in communities.

Hofstede's cultural dimensions lend themselves to the drawing of comparisons between different countries and/or cultures and categorising them accordingly. Unfortunately, they are often seen as a reason for intercultural behavioural differences and even have come to be applied as independent variables. A knowledge of cultural dimensions can create the illusion of 'knowing' a culture, without providing any deeper understanding of that culture or a system of orientation for concrete action.

Most anthropologists have come to espouse what can be described as a 'social constructionist approach'. They tend to regard culture as based on shared or partly shared patterns of meaning and interpretation as opposed to sharing broadly consistent, yet complex, patterns of behaviour and modes of existence. Thus, people's identifications with, and affiliation to, a multiplicity of different cultures, e.g. national, organisational, professional, gender and generational cultures, are subject to change, and boundaries between cultural communities become fluid and contingent. This implies that national, corporate or professional cultures are seen as symbolic practices that only come into existence in relation to, and in contrast with, other cultural communities. People's cultural identity constructions and their social organisations of meaning, are, in other words, contextual. This relational approach to culture suggests that every individual embodies a unique combination of personal, cultural and social experiences, and thus any communication and negotiation is ultimately intercultural.

Another issue in anthropology relevant in this context is that of culture as a system of public meanings versus private codes in the minds of individual members. Advocates of the former, such as Clifford Geertz point to the way 'Japanese culture' exists prior to (and irrespective of) the of the birth of any individual Japanese. They state that – like the Japanese language – it consists of rules and meanings that transcend individual minds. The counter-arguments in favour of what can be called a 'distributive model of culture' are equally compelling. Such a view takes as fundamental the distribution of partial versions of a cultural tradition among members of a society. Thus, it can take into account the different perspectives on a way of life held by men or women, specialists or non-specialists. 'A culture' is seen as a pool of knowledge to which individuals contribute in different ways and degrees. This attitude is very much in line with Holden's approach to the concept of culture discussed below.

In any case, a debate about what culture 'really is' is not likely to be fruitful. For some purposes, it is useful to view a peoples' cultural knowledge as if it were a single coherent system; for other purposes, it is necessary to take into account its distribution within a community.