The present study provides a replication of Malloy and Levis's (1988) finding of persistent human avoidance responding during extinction. Forty-seven percent of the subjects who learned failed to extinguish by Trial 500. This result was of particular interest given that the extreme avoidance persistence was obtained in many cases with only one CS-UCS (shock) pairing and with a relatively mild UCS (2.5-mA shock level). A comparison between extinguishers and nonextinguishers suggested that the traditional laws of extinction were operating in both groups. These data parallel those found with infrahumans suggesting the governance of comparable conditioning principles. The theoretical importance of these findings are discussed.Attempts by learning theorists to extend established laboratory principles of fear-avoidance conditioning and conflict learning to explain the development, maintenance, and unlearning of human psychological symptoms have resulted in important and influential contributions (e. g. , Dollard & Miller, 1950;Eysenck, 1979;Levis, 1985;Mowrer, 1950;Reiss & Bootzin, 1985;Stampfl & Levis, 1967;Wolpe, 1958). In the 1970s, a number of critics challenged the viability of a traditional learning approach to account for the persistence of human symptoms, which sometimes last for years (e.g., Gray, 1971;Kimmel, 1971;Rachman, 1976;Seligman & Johnston, 1973). The main argument advanced was that laboratory demonstrations of extreme persistence of avoidance and fear responding in extinction were rare. Mackintosh (1974), in reviewing the avoidance area, confirmed the critics observations that in most cases infrahuman avoidance extinction, as well as fear extinction, is relatively short lived. Extinction of both responses typically occurs within 20 to 40 trials; rarely does this process require more than 100 trials.Although Mackintosh's (1974) conclusions accurately reflected the literature as a whole, more studies than he reported demonstrating persistent avoidance responding in extinction existed at that time (Baker, 1968;Denny, 1971;Kostanek & Sawrey, 1965;Levis, 1966; Levis, Bouska, Eron, & McIlhon, 1970;Oliverio, 1967;Seligman & Campbell, 1965;Solomon, Kamin, & Wynne, 1953). Since that time additional studies can be added to the list (e.g., Levis & Boyd, 1979; McAllister, McAllister, Scoles, & Hampton, 1986;Stampfl, 1987). To date, infrahuman examples of extreme resistance to extinction of both fear and avoidance responding cannot be considered to be rare with the procedural methods and principles responsible for producing the effect clearly delineated (see Levis, 1989).
125However, this issue is less clear when the human avoidance conditioning literature is reviewed. Prior to the publication of the Malloy and Levis (1988) study, we could not find one human study that showed substantial resistance to extinction of avoidance responding. Bank (1965) reported that after 80 acquisition trials the most persistent group avoided for a mean of only 36 trials. Meyer (1970) and Drake and Meyer (1972) found that, afte...