First International Conference on Spoken Language Processing (ICSLP 90)

Kobe, Japan
November 18-22, 1990

Distinctive, Redundant, Predictable, Neotssary, Sufficffint Accounting for English /bdg/-/ptk/

Leigh Lisker

Haskins Laboratories and University of Pennsylvania, New Haven, CT, USA; Philadelphia, PA, USA

Traditionally phonologists have appealed to three independent phonetic features in separating English /bdg/ and /ptk/: V(oice), A(spiration), and F(orce of articulation). Moreover, these features were unequal,- one "distinctive," the others "predictable" and/or "redundant." Once F was distinctive; more recently V has been assigned this role. Moreover, where a distinctive feature once was the principal category marker, now it is not clear that a "distinctive" feature has special perceptual status. Experimental evidence suggests that no single property can be isolated as the perceptually distinctive one, nor does any seem so devoid of "cue potential" that it maintains its neutrality under all the conditions that experimental ingenuity and modern technical facilities can contrive.

Full Paper

Bibliographic reference.  Lisker, Leigh (1990): "Distinctive, redundant, predictable, neotssary, sufficffint accounting for English /bdg/-/ptk/", In ICSLP-1990, 933-936.