Watts, the 1965 Los Angeles Riots, and the Communicative Construction of the Fear Epicenter of Los Angeles

Abstract

215 mental maps collected in 1998 indicate that Watts is Los Angeles’ epicenter of fear. Spatial cluster analysis provides evidence that this fear is most closely associated with the 1965 “Watts” riots. The role played by television in constructing spatial fear is proposed as a possible explanation for the persistence of this collective memory. Analysis of phone survey data and of a set of mental maps indicates that the greatest fear of Watts is found among respondents with stronger dependency relations with television.

Keywords:

• Social Fear
• Mental Maps
• Media Dependency
• GIS
• Spatial Effects

Notes

A previous version of this paper was presented at the American Sociological Annual Conference, Chicago, 2001

1. The following narrative is based on Cohen & Murphy (1966) and on the McCone Report (1965).

2. Watts is a very small section of Central Los Angeles. South-Central, considered to be the traditional “Black” area in Los Angeles, is far larger, spanning over more than 50 square miles.

3. We use the whole map-maker sample as baseline to emphasize that we are interested in deviation from the “typical” mental image (as we capture it through our sampling procedure described in the methods section) of Los Angeles.

4. For example, television compared to newspaper connected respondents are slightly older (average age 63 vs. 60), somewhat less affluent (62% make more than $35,000/year vs. 78%), less educated (40% are college educated vs. 53%), and more likely to be female (60% are female vs. 47%). In addition, 66% of the television- (compared with 75% of the newspaper-) dependent respondents also use interpersonal communication for satisfying basic social needs.

5. The units (cases) that are paired are pixel locations. Within the one mile radius area are 104 pixel locations. For each location we compare the value for the pixel in the two maps (i.e., television- vs. newspaper-inflected maps). T-ratios are very large due to low variance. To ensure that the significance was not artificially boosted by the number of pixels or by low variance, two supplementary analyses were performed. First, we changed the map resolution, from 104 to 36 pixels. An analysis performed at this rougher level of resolution, produced similar results, although, as expected, at relatively lower t-values. Second, to check the robustness of our t-tests we repeated the pair-wise analyses by using non-parametric tests (Wilcoxon), which provided similar results in terms of significance and directionality.

Additional information

Notes on contributors

Sorin Adam Matei

Sorin A. Matei (Ph.D., USC, Annenberg School) is Assistant Professor at Purdue University and leader of several research projects dealing with community life and media effects in the US and in Europe

Sandra Ball-Rokeach

Professor Sandra J. Ball-Rokeach (Ph.D., University of Washington) is the director and principal investigator of the Communication Technology and Community Program at the Annenberg School for Communication, University of Southern California and 1st 5 LA, which were the main funding resources for the research reported in this paper