This study was concerned with the sources of variation in the attitudes of manual workers toward social class and related matters. A review of the literature suggested that, for workers to adopt a perspective that challenges the existing distribution of rewards and privileges in society, they have to be involved in a series of relationships, in work and in the wider community, that create barriers to the influence of the values of dominant groups in society. Evidence from studies in several societies suggests that involvement in predominantly working-class milieux, at work, in the community, and in kinship and friendship networks, leads to the worker adopting a perspective that is "deviant" from the dominant value system. Exposure to the influence ) of people from other classes, on the other hand, 'Would weaken the' barriers and increase the likelihood that the worker would adopt a perspective akin to that of groups higher in the social hierarchy. This study involved the application of the perspective outlined above to differences in the attitudes of manual workers in four Ontario communities. A questionnaire survey was administered in the four committees, which were selected because, in terms of class structure, they differed from each other in ways that were believed to be associated -with differences in working-class attitudes. The indicators of class attitudes examined in this thesis are class identification, choice of models describing the bases of the stratification system, an index of militancy, and support for the New Democratic Party. Residence in a predominantly working-class community was expected to influence the worker to adopt a "deviant" set of attitudes. But community differences in class identification and choice of class models among workers did not follow the expected pattern. Militancy and support for the NDP were found to be related to differences in the class compositions of the conm.mities studied, but, over-all, community differences were not as large as evidence from studies in other societies would lead one to expect. It was argued that the extensive geographical mobility among workers, and the ethnic diversity in three of the comm.mities, attenuated sanewhat the effect that differences in canmunity class stnicture were expected to have on workers' attitudes. Elements in the work situation did, however, show a more substantial relationship with differences in workers' attitudes. Membership in a trade union was found to be the most consistently strong influence on the workers' adoption of "deviant" attitudes. White-collar kinship and friendship affiliations, on the other hand, particularly marriage to a woman who had, at one time or another, been employed in white-collar work, were found to be related to the worker adopting attitudes more similar to those held by people higher in the social hierarchy. It was argued that the barriers to the influence of the dominant culture are nrutua.lly reinforcing, so that removal of one weakens the power of the others. Geographical mobility among workers, combined with ethnic and religious diversity, were seen as weakening even the solidarity ties that might exist in a community predominantly working-class in social composition. And the large number of workers who-were married to women with experience of non-manual work was seen as a further factor weakening the barriers to the influence of the dominant atlture. It was argued that, even though the trade union appears to be the most effective barrier to the dominant culture, the absence of other strung barriers lessens the power the union has to provide a set of alternative definitions of social reality for workers. It was concluded that the weakness of the barriers to the influence of the dominant culture may be suggested as one reason why Ontario workers do not provide the kind of support to the New Detoocratic Party that is provided to parties of the left by workers in other western societies. / Thesis / Doctor of Philosophy (PhD)