Large scale data for analyzing sentiments of sub-cultures

A long research tradition has been started on affective connotations of language since Osgood and colleagues provided good evidence of the universality of three affective dimensions – evaluation, potency and activity. Primary cross cultural research has shown that affective meanings depend on the cultural and social background of individuals. In this tradition, sentiments have been shown to be quite stable within societies, thereby ignoring possible differences within culture.. Nevertheless, in his research approach David Heise takes into account potential influences of sub-cultures. Yet due to lack of appropriate data his assumptions haven’t been be explored in greater detail.

In our Project we want to explore such differences on affective meaning of words between sub-cultures with unique data we recently conducted. A Germany wide sample of 3000 participants rated 910 words in total on three affective dimensions. Additionally we asked for various social characteristics of the participants. The quoted sample captures the social diversity in Germany very welltherefore it is highly appropriate to explore the varieties of affective connotations within the German culture.

In this contribution, we introduce our large scale data set as a promising basis for sociological investigations on the affective foundations of sociality. We further suggest different strategies for analyzing these data. One strategy is to identify word fields whose affective meanings vary with social environments like milieus or lifestyles. A second strategy is to relate affective meanings to attitudes, norms and action tendencies. As a third strategy, we try to locate the individual’s social position in terms of affective meanings.