«This study shows how the scientific analysis of choice behavior requires several methodological issues to be clarified. Above all, the meaning of theoretical terms is of the utmost relevance.»
«Theoretical concepts are formulated by isolating some fundamental predicates that identify the variables characterizing an experiment (independent, dependent, and constrained variables). It means that an experiment is determined by theoretical predicates, even when the latter have the minimum level of generalization.»
«A proper explanation of phenomena must rest on (a) the identification of experimental predicates exactly defined, (b) the determination of methodological criteria for introducing new predicates, and (c) the exact identification of the connections between predicates in terms of functional relation, which characterizes the scientific language at the syntactic level. If we were to pretend to introduce predicates such as “choice,” “maximization,” and “utility” into experimental language prescinding from the identification of the methodological criteria to connect these predicates to strictly defined preexisting experimental predicates, we would risk constructing a pseudo-explanatory (and therefore non-scientific) language or arbitrarily expand (always in pseudo-explanatory terms) a preexisting scientific language.» IT
(L’analisi scientifica del comportamento di scelta [The scientific analysis of choice behavior], III, p. 149; II, pp. 25, 10-11)
L’analisi scientifica del comportamento di scelta
[The scientific analysis of choice behavior]
[The scientific analysis of choice behavior]
- L’analisi scientifica del comportamento.
[The scientific analysis of behavior]
- Il processo di discriminazione come processo di rinforzo.
[Discrimination as reinforcement process]
- Il concetto di interesse. Interesse finale e interesse strumentale.
[The concept of interest. Final and instrumental interests]
- Il concetto di intensità dell’interesse.
[The concept of interest intensity]
- Esplicazione di alcuni esperimenti mediante il concetto di intensità dell’interesse.
[Explication of some experiments using the concept of interest intensity]
- Scelta, utilità, massimizzazione, come concetti pseudo-esplicativi.
[Choice, utility, maximization as pseudo-explicative concepts]
- La struttura del discorso scientifico: la relazione funzionale.
[The structure of scientific discourse: the functional relation]
- La relazione funzionale nella psicologia.
[The functional relation in psychology]
- La relazione funzionale nella fisica.
[The functional relation in physics]
- Termini teorici e termini osservativi. I problemi della generalizzazione e dell’astrazione.
[Theoretical and observational terms. Generalization and abstraction issues]
- Le variabili intermedie in psicologia.
[Intervening variables in psychology]
- Il significato dei termini teorici nella fisica. L’esempio del campo elettromagnetico.
[The meaning of theoretical terms in physics. The example of the electromagnetic field]
- Il livello dei predicati. Relazioni tra predicati di diverso livello.
[The level of predicates. Relations between different-level predicates]
- La distinzione tra leggi sperimentali e teorie.
[The distinction between experimental laws and theories]
- Le regole di corrispondenza.
[The correspondence rules]
- Analisi critica del concetto di regola di corrispondenza.
[Critical examination of the concept of correspondence rule]
- I postulati limitativi.
[The restrictive postulates]
Carnap R., Meaning and necessity, Chicago, University of Chicago Press, 1947
Hempel C.G., La formazione dei concetti e delle teorie nella scienza empirica, Lilano, Feltrinelli, 1961
Herrick R.M., Lever displacement during continuous reinforcement and during a discrimination, in: “Journal of Comparative and Physiological Psychology”, 1963, vol. LVI
Mendelssohn K., Sulla via dello zero assoluto. Sviluppo e significato della fisica delle basse temperature, Milano, Il saggiatore, 1966
Millenson J.R., Principles of behavioral analysis, N.Y., Macmillan, 1967
Mowrer O.H., Lerning theory and behavior, N.Y., Wiley, 1960
Mowrer O.H., Jones H.M., Extinction and behavior variability as functions of effortfulness of task, in: “Journal of Experimental Psychology”, 1943, vol. XXXIII
Nagel E., La struttura della scienza, Milano, Feltrinelli, 1961
Rethlingshafer D., Motivation as related to personality, N.Y., McGraw-Hill, 1963
Rethlingshafer D., Eschenbach A., Stone J.T., Combined drives in learning, in “Journal of Experimental Psychology”, 1951, vol. XLI
Rogers E.M., Physics for the inquiring mind, Princeton (N.J.), Princeton University Press, 1960
Skinner B.F., Scienza e comportamento, Milano, Angeli, 1971
Whiting J.W.M., Mowrer O.H., Habitat progression and regression: a laboratory investigation of some factors relevant to human socialization, in: “Journal of Comparative Psychology”, 1943, vol. XXXVI
1/13«From a methodological perspective, behaviorism states that the difference between human and animal organisms is not qualitative; it is a matter of degrees of complexity. Indeed, considering human organisms qualitatively different entails resorting to metaphysical judgment factors, whereas the latter must be detected and eliminated from scientific analysis. Once acknowledged that the difference is a matter of degree, animals appear to be the most suited organisms for the experimental study of the structure of psychological processes. That allows psychology to turn from a metaphysical analysis of mental processes into the scientific analysis of behavior; it becomes an experimental science, as well as physics and biology.» IT
2/13«We introduce the term “interest” as a nominal definition for the final element of an operation sequence and not as a concept explicating operant behavior. If one stated that a subject engages in a particular behavior because it has a particular interest in carrying out that behavior, then the concept of behavior would be explained in mentalistic terms. Namely, a hypothetical internal variable (interest) would be assumed to be the cause of behaviors.
Therefore, when we talk about a subject’s interests, we refer only to the final elements of given operation sequences, followed by particular reinforcers.» IT
3/13«Based on the above observations, the concept of drive can be expressed in operational terms as the result of certain operations on an organism, notably deprivation operations through which a given stimulus S becomes a reinforcer S+. It is possible to find the appropriate operations for each primary reinforcer.
The term drive is defined thereby as a relation between a deprivation state and some reinforced behavior. This definition excludes that the drive can be considered as the cause of behaviors.» IT
4/13«We note, again, that the analysis of reinforcement laws always presupposes a deprivation operation. Therefore, the relation between the behavior modifications and the reinforcement contingencies (given a particular deprivation state) is only one aspect of the complex reinforcement process. The second aspect is the relation between the behavior modifications and the changes in the deprivation state (given the reinforcement contingencies, which are kept constant). Hence, the reinforcement process must be explicated in terms of changes in reinforcement contingencies, but it must also be explicated considering the variations in the deprivation state.» IT
5/13«We have defined the intensity of an interest based on the compactness of the operation sequences. The latter can, in turn, be considered a function of the deprivation state. We may expect that a subject with a high deprivation always carries out an operation sequence (leading to a particular reinforcer) while excluding any interference related to other operation sequences.» IT
6/13«Suppose that the intensity of the interest I1 is much higher than any others. In this event, it will take much longer for interest I1 to decrease to a level lower than some other interests. Therefore, while the alternation of operation sequences occurs straightaway in the previous case, here, where the difference in intensity between I1 and any other interests is very high, there is no alternation of operation sequences but repetition of identical operation sequences. In this event, the subject behaves as if it had virtually only one interest, and its behavior is predictable with certainty.
In the first case, we deal with operation sequences that intersect, as the differences between the intensities of interests are tiny. In the second case, there is no intersection between different operation sequences but a single operation sequence, as the difference in intensities between one interest and any others is very large.» IT
7/13«Saying that an organism “focuses and filters at once the potential stimuli from the external environment” is to formulate a pseudo-explanation about the fact that some interests of the subject have a higher intensity than others.
This pseudo-explanation results from the fact that starting from an analysis of phenomena showing that a subject may engage in alternating behaviors (and, hence, may have interests with different intensity levels), one wonders why organisms engage in these alternating behaviors. Thus, the problem is no longer to explicate alternating behaviors but to identify the (internal) causes that would determine the onset of such alternating behaviors. In the hypothesis under consideration, alternating behavior would be attributable to an internal process named choice. The rules of the scientific method are violated thereby: a factor that cannot be controlled is introduced in the explanation, based on a pre-scientific concept of causality according to which, given two factors, one factor should determine or give rise to the other factor (causality as a genetic relation).» IT
8/13«Even though one wanted to assume the relation between an alternate behavior and a subject’s choice process as a functional relation, namely a relation between two variables, one dependent and one independent (causality as functional dependence), this would not be possible. The reason why it is impossible to establish such a relation between the two factors in question is that the choice is hypothesized as an internal variable and, hence, methodologically different from the variable represented by the behavior. In short, it is a useless duplication that leads us to conclude that the terms “choice” and “alternate behavior” have the same meaning.» IT
9/13«The analysis of choice behavior is usually deepened by referring to the concept of expectation. In short, it is stated that choosing is based on learned prospects, that is, expectations.
Even expectation, meant as anticipation of the satisfaction of a given interest, represents a pseudo-explanation. […] To state that forecasting an event is based on how often that event occurred in the past (its frequency) is to refer to the past reinforcements received by the subject. In this hypothesis, the concept of expectation is tantamount to saying that a subject has an interest related to a reinforcement process that took place throughout a sequence of past experiences. Therefore, the statement that the subject engages in a given behavior because it expects the satisfaction of a given interest from that behavior is mentalistic and pre-scientific. It is an expression referring to the fact that the subject carries out a given behavior because, based on its past history, that behavior will be reinforced with a certain probability.» IT
10/13«The approach to scientific problems in structural terms eliminates the metaphysical issue of the conventionality of postulates. The methodological meaning that can be attributed to the concept of conventionality of postulates is the possibility of always inserting new postulates within the scientific discourse. However, this fact does not question the validity of the preexisting postulates within the scientific discourse; it only requires that the latter are characterized according to higher or lower abstraction levels than the new ones.
It follows that the scientific discourse is an open discourse, meaning that the postulates and, in broader terms, the predicates that constitute it do not exclude the possibility of expanding it; such enlargement never alters the validity of preexisting structures but places these structures in broader explanatory contexts.» IT
11/13«The metaphysical problem of the transition from sensory data to theoretical terms, typical of Hempel’s position, must therefore be replaced by the methodological problem of identifying different levels of generalization (which entail, syntactically, the insertion of restrictive predicates) and different abstraction levels of the predicates belonging to scientific language.
This argument can be repeated about the correspondence rules, which many scholars of the scientific method have introduced to explain the alleged odd property of the theoretical language, which would make no reference to observables and, at the same time, would explicate observable phenomena. Namely, to explain the problem of the empirical nature of theoretical terms.» IT
12/13«In psychology, the explanatory usefulness of intervening variables is denied by Skinner, who claims that inner conditions are useless to control behavior unless they can be manipulated. [...]
If the intervening variables have no explanatory role, then the question remains whether the same can be said of all theoretical concepts, even those of physics. On the other hand, if theoretical concepts have to be considered essential for the explication of phenomena, then it is necessary to explain why intervening variables cannot be considered theoretical concepts.» IT
13/13«The statement that experimental laws would have a life of their own derives from the assumption about their observability. If experimental laws differ from theoretical laws in terms of observability, then the fact that an experimental law can be incorporated into one theory or another may appear as a confirmation of the difference between the two types of laws. However, once the metaphysical presupposition of observability has been eliminated, that an experimental law can be incorporated into more than one theoretical schema does not allow for concluding that the experimental law is autonomous from the theory. It is just a matter of a change in postulates of the theory within which the experimental law is included. However, the sole effect of such a change in postulates is a reformulation of the law in terms of the new theory.
Therefore, it cannot be maintained that an experimental law has a life of its own compared to the theory because an experimental law is always expressed in theoretical terms. That is true even though, at a given stage of scientific knowledge, these theoretical terms can have a very low abstraction level compared to the higher abstraction level of a theoretical schema with a broader domain of application.» IT