This paper argues that Chomsky's and Halle's restriction on the features [+high, +low] are too severe; that this restriction is inconsistent with the generative treatment of affricates, laterally-released stops, and prenasalized stops; and that the restriction is inconsistent with the notion that linguistic descriptions are abstract theories about cognitive structures. Research was supported in part by a grant from the Office of Education, U.S. Department of Health, Education, and Welfare to the Pacific and Asian Linguistic Institute (PALI) of the University of Hawaii. A brief bibliography is included. (DD)