The recent experimental literature on the acquisition of English antonyms is reviewed, with special reference to the position of Eve Clark and the particular ontogenetic assumptions her position entails. The assumptions examined are: (1) in a hierarchically organized lexical domain, the order of acquisition appears to be from the top of the hierarchy downward; (2) linguistically unmarked or semantically positive terms are acquired before their marked or negative counterparts; and (3) systematic misinterpretation of the marked or negative term indicates confusion between it and the unmarked term. It is shown that many recent findings are at variance in some major way with predictions from Clark's theory. A recurrent problem has been that children's comprehension task performance is confounded with their behavioral biases specific to the linguistic and nonlinguistic contexts in which the lexical terms are presented. (Author/AMH)