Search

New and powerful methods for text analysis potentially facilitate advances in theorizing and studying a key aspect of cultural change: the themes that actors encounter in the representations with which they come into contact. We employ topic models, an increasingly popular means of extracting themes from large bodies of texts, to analyze change in press representations of U.S. government arts funding - a controversy that threatened the survival of the National Endowment for the Arts - between 1986 and 1997. Our theoretical contribution is to identify three key analytic tasks in the study of culture change: identifying analytic objects (themes of frames) to be studied; plotting the emergence of these themes; plotting their differentiation from other themes; identifying changing prevalence of themes over time; and identifying change in the valuation of these themes (and of key entities within them) over time as well. Our methodological contribution is to develop and employ measures of emergence, differentiation, prevalence and evaluative change, with which we analyze more than eight thousand articles from five U.S. newspapers related to our case. Our substantive contribution is to demonstrate that what initially was depicted as a controversy over decisions by a single government agency came to be portrayed as one of many fronts in a broader "culture war" between social conservatives and secular progressives and to identify changes in the political opportunity structure that facilitated this evolution. [ABSTRACT FROM AUTHOR]

Science and religion have long competed over world¬views, empirical beliefs, and moral authority. In the U.S., after years of quiescence, competition flared in the late 20th cen¬t¬ury, with conflict over abortion, euthanasia, and the teaching of evol¬ution. We ask to what extent this competition is re¬flected in U.S. public opinion, as tapped by items in the 1988 General Social Sur¬vey. We com¬plicate conventional analyses in two ways. We challenge the conceptual op¬position of science vs. relig¬ion, by introducing a third dimen¬sion of belief and cognit¬ive auth¬or¬ity, "enchantment" (the be¬lief that unseen agents other than the Deity exert agency in the everyday world); and we explore the possibility that enchanted belief systems may com¬pete with scientific and religious belief systems or, un¬der some cir¬cum¬stances, be¬come integrated with one or the other. Second, we explore population heterogeneity in at¬titudes towards science, religion and the supernatural, ex¬amining two different types of structure employing two different methods. We use Lat¬ent Class Analy¬sis to identify five classes of respondents: those committed, respectively, to relig¬ious, scientific and en¬chant¬ed worldviews; those who endorse none of these, and those committed to both science and religion. With a new method, Relational Class An¬aly¬sis, we identify two classes of re¬spondents whose answers covary in similar ways: Op¬inions of one are differentiated along the conventional divide: science vs. religion (and enchantment); those of the other oppose in¬stitutionalized cultural authority (both scient¬ific and religion) to an enchanted under¬standing of the world. The diversity of opinion among highly religious Americans, and the presence of a previously overlooked political and religious constituency of sci¬ence supporters, may have important implications for sci¬ence policy and science ad¬vo¬cacy. [ABSTRACT FROM AUTHOR]

Students of white Americans' attitudes towards African-Americans have argued about the meanings of the former's responses to attitude surveys, especially about whether certain responses are socially appropriate displacements of conventional racism or, instead, reflections of political, moral, or economic beliefs independent of racial attitudes. We argue that such attitudes cannot be assessed one at a time, but instead must be viewed as related to schemata or associational networks within which each element derives its meaning from other elements with which it is associated. We further argue that interpreting survey responses requires the analyst to make inferences about the network of attitudes within which they are embedded; that schematic heterogeneity is typical of survey populations; and that understanding racial attitudes requires a partitioning of respondent samples into subsets of schematically similar agents. ..PAT.-Unpublished Manuscript [ABSTRACT FROM AUTHOR]

Much research of the "digital divide" has presumed that adults on the wrong side of the divide may be economically disadvantaged, yet little if any research has tested this premise. After discussing a range of mechanisms - both real and spurious - that could account for differences in income growth between workers who do and do not use the Internet, we use data from the CPS and from a special re-interview of respondents to the GSS to examine the impact of Internet use on change in income over periods of 12 to 18 months. Our analyses document strongly significant positive associations between web surfing and income growth, and we find some evidence that is inconsistent with the hypothesis that the association are spurious. It appears that skills and behaviors associated with certain kinds of Internet use were rewarded by the labor market at the beginning of the 21st century. Although there is not reason to believe that such differentials will persist indefinitely, our findings do indicate that technological divides, especially during periods in which new technologies are rapidly diffusing, are a valid concern for students of inequality, and that technological change can reproduce and, at times, challenge social inequality. These results suggest that research on inequality in access to and mastery of technology deserves more systematic attention from students of social stratification. ..PAT.-Unpublished Manuscript [ABSTRACT FROM AUTHOR]

In Republic.Com, Cass Sunstein argued that the Internet posed a threat to American democracy. Although the Internet vastly expands the scope of the political opinions to which citizens have access, it also increases users' ability to find information sources with viewpoints that match their own. Because the politically attentive tend to seek out information consistent with their existing views, Sunstein argued, the Internet threatens to polarize political attention and discourse. This paper uses data from the 2000 General Social Survey "Internet module" to test the Sunstein thesis. Among the minority of Americans who sought political information on the Internet, we find little evidence for the Sunstein thesis. As many politically attentive users went to sites that challenged their views as to sites that reinforced them, even more went to sites they perceived as politically neutral, and a plurality visited a combination of sites. Fewer than 10 percent focus primarily on opinion-confirming sites. Although few respondents aver that it is "very true" that Internet political sites lead them to change their minds about public issues, many report that such sites help them form initial opinions or lead them to view issues as more complex. Although caution is appropriate in interpreting self-report data, these results suggest that Americans who go on-line for political information are relatively open to discrepant political information and relatively disinclined to limit themselves to views similar to their own. Thus far, then, the Internet has not contributed to a balkanization of political discourse. [ABSTRACT FROM AUTHOR]