Peoples and individuals around the globe continuously develop their own communicative habits. With each generation, adjustments to changing circumstances are made-economic circumstances, natural circumstances, and, for instance, mobility circumstances. The outcome of such transitions is cultural variation, which is visible in hierarchical social systems, belief systems, legal systems, traditions, attire, and all kinds of rituals. Communicative systems are part of culture, and they deserve a role in research focussing on language and communication. However, applying culture as a variable is a challenge, not only because of the cultural variation between peoples and individuals but also because the effects of culture on actual language utterances are hard to measure. Another issue is the dominance of Anglowestern cultural patterns in many analyses. This paper explains these issues and critically reviews the various criteria that well-known cultural models-like the one by Hofstede (1980), Lewis (1969), and Hall (1959, 1976)-use to categorise cultures. Examples of such criteria are: region, relationship with uncertainty, femininity/masculinity, and power relations. The paper concludes by giving a number of practical solutions to the challenge of treating culture as a variable in sociolinguistic research. These solutions are related to, amongst others, the reviewing process for journals, widespread norms of 'good academic language', author/editor selection, cross-cultural academic cooperation, and sharing of funds.