Abstract
Many different techniques can group synonymous words together, but it is much harder to untangle the nuance and emphasis borne by individual members of a synonym set. The distributional hypothesis holds that you may understand a word by the contexts in which it appears. This thesis applies a sort of transposition of the distributional hypothesis to groups of synonyms. I postulate that, by aggregating the differences among the contexts of similar words, one may discover the range of semantic factors borne by a synonym group as a whole, as well as the implications of choosing an individual member of a synset. By applying an analytical framework along these lines to extract semantic factors for set of synonyms, I initially find factors that are of a different sort than the factors used in most decompositional analyses, and see promise in developing the approach further with varying methods of feature extraction.