What do sociologists mean when they describe culture as founded on "shared understandings"? Sharing an understanding does not necessarily imply having the same opinions but rather agreeing on the structures of relevance and opposition that make symbols and actions meaningful. Because meaning is contextual, different people might interpret the same reality in different ways. Yet standard quantitative sociological methods are not designed to take such heterogeneity into account. In this article, I introduce a new method -- relational class analysis -- that uses attitudinal data to identify groups of individuals that share distinctive ways of understanding the same domain of social activity. To demonstrate its utility I use it to reexamine the cultural omnivore thesis. I find that Americans' understandings of the social symbolism of musical taste are shaped by three competing logics of cultural distinction, in a manner that complicates contemporary sociological accounts of artistic taste. Adapted from the source document.