Kenneth Spence
Spence was a Hull student and his theories follow the same general form as Hull’s theories. In particular, Spence wants to express his ideas mathematically whenever possible and to present a comprehensive theory of behavior. He also uses most of the same variables that Hull used in his behavior theory: Drive, Habit, Inhibition, Incentive, internal stimuli, and the fractional antedating goal response (AKA the factional anticipatory goal response). However, the arrangement of these concepts within the theory is different.
The main idea of Hull’s theory was the multiplicative relationship between Drive (D) and Habit Strength (SHR or just H). So, Reaction Potential (Performance) = DxH. Drive refers to states like hunger and thirst which give rise to internal stimuli (Drive Stimuli). These Drive Stimuli can be conditioned to behavior just as external stimuli can. Drives provide the basis for reinforcement--- a behavior is reinforced if it reduces a Drive (a biological need). H is the variable that represents learning and its main influence is the number of reinforcements that the target behavior receives. So, in the contiguity theory of learning vs the reinforcement theory of learning, Hull falls on the reinforcement side. In his later theorizing Hull added incentive (K) as another variable. Incentive is the value a subject places on an outcome. For example, an increase in the amount of food (Crespi Effect) will cause a hungry rat to run faster to reach the goal box. Hull believed that K interacts with D and H multiplicatively. So, Reaction Potential = DxHxK. Finally, Hull thought that inhibition (I) of various sorts decreased responding. So, ignoring behavioral oscillation, Reaction Potential = DxHxK-I.
Spence had many of the same variables as Hull, but they were given slightly different composition, and interacted with each other differently. In particular, Spence is a contiguity S-R theorist. Reinforcement doesn’t lead to learning (H). Reinforcement, in his view, affects incentive (K). Spence more fully develops K than Hull did. In Spence’s view K depends upon the fractional anticipatory goal response (rg-sg). Hull gets credit for inventing rg-sg, but he thought that it contributed to H, not K. Spence thought that rg-sg led to increases or decreases in K depending upon how strong the rg was. Thus, K was, in a sense, a result of learning (classical conditioning). The rg-sg learning needs to be distinguished from H which is the amount of learning for the relevant instrumental/operant behavior. Recall that sg is an internal movement-produced stimulus, so it in addition to drive stimuli is an internal stimulus (as with Guthrie and Hull). Furthermore, in fleshing out his notion of Drive, Spence explains that in his theory, D, is the result of an internal response which he calls re. The e-subscript refers to emotion. So the basis for drive (for Spence) is an internal emotional response--- demonstrating, once again, that the notion that behaviorism only was willing to speak about the relationship between observable stimuli and overt responses is false (unless we are speaking about later Skinner). Anyway, Spence, too, gives us an equation meant to explain the probability of a response (reaction potential): reaction potential = Hx(D+K)-I. The same variables are used as with Hull, but D+K is a single quantity that is multiplied by H. The relationship between D and K is additive not multiplicative. This means that either D or K could be zero without the expression “Hx(D+K)” itself being zero (unless both D and K are zero). For Hull, all of D, H, and K must be non-zero in order for the expression “DxHxK” to avoid being zero.