A study analyzed the distribution of words in a three-million-word corpus of text from the "Wall Street Journal," in order to test a theory of the acquisition of word categories. The theory, an alternative to the semantic bootstrapping hypothesis, proposes that the child exploits multiple sources of cues (distributional, semantic, or other) that are readily available in an information-rich environment, to acquire syntactic categories of words. This study investigates the quantity of information that is available in a specific language environment. The Lexical Space approach to categorization, which looks at the position of each word in lexical space and the structure of the resulting information, was used. Results indicate that co-occurrence and adjacency of words is highly indicative of their syntactic category, and that the high number of possibly contradictory context features poses no serious problem for bootstrapping. Contains 15 references. (MSE)