Conditioned reinforcement: Neglected or outmoded explanatory construct?

Abstract

The construct of conditioned reinforcement has been challenged over the past two decades, both as an adequate explanation of behavior in various situations (e.g., behavioral chains) in which it was previously regarded as crucial and in terms of the validity of the concept of conditioned value on which it is based. Recent research is reviewed that answers these criticisms and vindicates the importance ofthe construct. Also discussed are unanswered empirical issues regarding the concept and its implications for behavioral phenomena of major importance for general-process learning theory, such as autoshaping and imprinting.In reaching our goal [of understanding human behavior] the principleof secondary reinforcement will be of great analytical assistance.... It gives a powerful and indispensable tool for the solutionof many vexing and absorbing problems of human action. (Keller & Schoenfeld, 1950, p. 260) Although it is characteristic of human behaviorthat primary reinforcers may be effective after a long delay, this is presumably only because intervening events become conditioned reinforcers. (Skinner, 1953, p. 76) There is probably no concept in all of psychology that is in sucha stateoftheoretical disarray as the conceptof secondary reinforcement. (Bolles, 1967) During the period from 1940 to 1960, when behavior theory was the dominant explanatory framework for psychology, the concept of conditioned (secondary) reinforcement-the idea that stimuli paired with primary reinforcers acquire reinforcement properties in their own right-played a central role in extrapolating to human behavior the concepts of reinforcement developed in the animal laboratory. Human motives and values were regarded as only minimally determined by biological drives and thus were primarily the result of learning.

Conditioned Reinforcement Dynamics in Three‐link Chained Schedules

“…That is, the conditioned reinforcement properties of the middle-link stimuli could be due in part to the time to food signaled by their onset and by the time to the terminal-link stimulus. Such a dual-process view is consistent with a variety of other data (e.g., Williams, 1994b), including those presented in Experiment 1.…”

“…The question posed is how the stimuli that intervene between the choice and outcome modulate this control by the delayed discrimination contingencies. Several different, and sometimes opposing, effects (e.g., conditioned reinforcement, blocking, marking, bridging) have been postulated in different situations (see Williams, 1994b, for a discussion).…”

“…All had prior experimental experience in both a responseacquisition study using delayed reinforcement (see Williams, 1994b, for details) and a successive discrimination problem (Williams, 1994a), both in experimental chambers different from that used here, and with different stimuli. Food deprivation was maintained by allowing 90-min access to laboratory chow approximately 5 min after the end of the experimental sessions (see Hurwitz & Davis, 1983, for the justification of this procedure).…”

Operant Conditioning

“…If the same stimulus is present in all links (tandem schedule), or if the six different stimuli are presented in random order (scrambled-stimuli chains), performance is maintained in all links and the overall reinforcement rate is close to the maximum possible (6I, where I is the interval length). Other studies have reported very weak responding in early components of a simple chain fixed-interval schedule (e.g., Catania et al 1980, Davison 1974, Williams 1994review in Kelleher & Gollub 1962). These studies found that chains with as few as three fixed-interval 60-s links (Kelleher & Fry 1962) occasionally produce extreme pausing in the first link.…”

“…Many experimental and theoretical papers on conditioned reinforcement in pigeons and rats have been published since the early 1960s using some version of the concurrent chains procedure of Autor (1960Autor ( , 1969. These studies have demonstrated a number of functional relations between rate measures and have led to several closely related theoretical proposals such as a version of the matching law, incentive theory, delay-reduction theory, and hyperbolic value-addition (e.g., Fantino 1969a,b;Grace 1994;Herrnstein 1964;Killeen 1982;Killeen & Fantino 1990;Mazur 1997Mazur , 2001Williams 1988Williams , 1994Williams , 1997. Nevertheless, there is as yet no theoretical consensus on how best to describe choice between sources of conditioned reinforcement, and no one has proposed an integrated theoretical account of simple chain and concurrent chain schedules.…”