Experimental work on the concept of linguistic difficulty is summarized. Inherent linguistic difficulty is distinguished from contrastive linguistic difficulty. Studies of phonological acquisition are cited which tend to support the notion of an ordered acquisition of language features, and it is recommended that we look to cross-linguistic evidence, to linguistic universals, for guidance in establishing measures of pronounceability for particular sounds and sequences of sounds. In discussing semantic difficulty, the author focuses on the need for the language learner to discover the semantic correlates of the various parts of speech. The need for more study in the area of contrastive difficulty is noted, and existing work in contrastive phonology is discussed. (PM)