Metacontingency Terminology, Philosophical Assumptions, and the Scientific Dialogue: A Response to Ardila-Sánchez and Hayes (2023)

Abstract

Ardila-Sánchez and Hayes (2023, this issue) discuss how differing philosophical assumptions might reflect on differing metacontingency terminology and debate some of the arguments presented by Sampaio and Haydu (2023). We respond by restating the urgent need to clarify the definition of cultural milieu, which is illustrated by an argument about discriminative stimuli as components of the cultural milieu. We clarify that the differences in metacontingency terminology that we did not emphasize were related to interlocking behavioral contingencies (IBCs) and cultural consequences, and not to cultural milieu and group-rule generation. We question any rigid separations of “Skinnerian” and “Kantorian views” of cultural phenomena, insisting that we focus on a unified culturo-behavior science. We elucidate that verbal responses and stimuli may participate in IBCs, cultural antecendents, or selecting environment variables; answer some questions about the latter two concepts; and clarify that the we presented a COVID-19 psychological support project not to empirically validate the concept, but to illustrate and test the conceptual coherence of the terms and theory.

Cultural Milieu: A Concept in Need of Clarification

Institutions and organizations are affected by the prevailing beliefs, resources, cultural traditions, etc., that compose their context. This statement—expressed in these terms—would not lead to any major disagreement. However, the vagueness and polysemy of the lay vocabulary hide complexities and pitfalls that scientific terminology should avoid. In that direction, behavior analysts could instead say that culturants are affected by a cultural milieu. But despite advancing our scientific terminology, the concept of cultural milieu still involves ambiguities. In our target article (Sampaio & Haydu, 2023), we analyzed previous work by Houmanfar and colleagues (Ardila Sánchez et al., 2019; Houmanfar & Rodrigues, 2006; Houmanfar et al., 2009, 2010, 2020) to show why that was the case. The reply by Ardila-Sánchez and Hayes (2023, this issue) debates some of our arguments, but presents further evidence on the urgent need to clarify what the cultural milieu actually is.

Ardila-Sánchez and Hayes (2023, this issue) tried to clarify “three different stimulus functions and stimulus objects assigned to the cultural milieu.” One of them was “factors comprising the setting in which the interlocking behavioral contingencies [IBCs] are generating products.” That clearly relates to previous presentations of the concept by Houmanfar and colleagues and is compatible with our presentation of the concept of cultural antecedents. According to Ardila-Sánchez and Hayes, another component of the cultural milieu would be “institutional stimulus functions with respect to the shared responding of individuals in interlocking behavioral contingencies.” As the authors precisely state, the relations of institutional stimulus functions with shared responses and cultural behavior do not necessarily participate in IBCs and metacontingencies. We could say that Kantor’s (1982) “cultural psychology” does not refer specifically to culturant selection. In sum, we agree that institutional practices and metacontingencies are related but relatively autonomous, and that this relation should be further explored. We contend, nevertheless, that this can be accomplished by any behavior analyst interested in culturo-behavior science, regardless of their preferred philosophical assumptions (see next section).

Finally, another component attributed to the cultural milieu by the authors was as “discriminative stimuli that set the occasion for the aggregate products.” However, we must contend that discriminative stimuli are behavioral events—antecedent environmental events affecting the response classes of an individual (e.g., Pierce & Cheney, 2017)—while aggregate products are cultural events, describing environmental events produced by responses of two or more individuals (e.g., Glenn et al., 2016). Indeed, Houmanfar and Rodrigues (2006, p. 23), in the very article cited by the Ardila-Sánchez and Hayes (2023, this issue), clarify that they are proposing the concept of cultural milieu as an antecedent term in the metacontingency that is parallel—not a synonym—to an antecedent in the behavioral contingency. Furthermore, Houmanfar et al. (2010), in the other article cited by Ardila-Sánchez and Hayes, simply do not employ the expression “discriminative stimuli.” Therefore, affirming that the cultural milieu functions as or is composed of “discriminative stimuli that set the occasion for the aggregate products” not only confuses different levels of analysis, exemplifying a category error, but also misrepresents how the concept has been previously presented. The need for clarification is evidenced here once more.

Different Terminologies and Philosophical Assumptions

For Ardila-Sánchez and Hayes (2023, this issue), we “correctly identified the philosophical origins of the differences between the two-term and the five-term metacontingencies” but mischaracterized this difference as jargon. However, we did specify that the differences we alluded to were the ones regarding (1) “IBCs,” “interlocked behaviors,” and “socio-interlocked behaviors”; and (2) “cultural consequences or selecting environmental condition,” and “consumer practices”—the two terms common to both proposals. We did not refer to the whole proposal of the elaborated account of the metacontingency (“five-term metacontingencies”). Actually, it is not clear to us that the two terms added in that account (cultural milieu and group rules) were consequences of that philosophical difference. At least, this was not made clear in the works employing these terms. Ardila Sánchez et al. (2019), for instance, did not explicitly state this and only cite one work by Kantor (1982) dealing with individual behavior and not cultural-level analysis. As for the term “jargon”—which we used only once—we acknowledge that it can carry some negative connotations that we did not intend to imply. The more appropriate term is “terminology”—which we employed more frequently in our target article.

We also acknowledge that different terminologies are often (though not always) related to different philosophical assumptions. However, from our point of view, those who adopt different philosophical assumptions should not be restrained to discuss or investigate other proposals. That would hinder conceptual and empirical advances. In the present case, we do not feel it necessary to rigidly distinguish a “Skinnerian” and a “Kantorian” view of cultural phenomena. One does not have to choose one perspective to adopt and be limited by that choice.

Are we to abandon the ideal of a unified culturo-behavior science (Cihon & Mattaini, 2020) upon which much of the work in the Think Tanks on Cultural Studies, for instance, has been developed (Vasconcelos et al., 2022)? Should we divide our science along rigid philosophical lines drawn some decades ago by the authors who first proposed them (Kantor, Skinner, Glenn, etc.)? Should we give up on the possibility that “Skinnerians” and “Kantorians” can find common ground to discuss culturo-behavior science? We do not think so. And we think the proposal of the elaborated account of the metacontingency itself points in that unified direction.

Science demands dialogue. And dialogue demands efforts to establish common terminology. Philosophical perspectives in science should ultimately inform better description, prevision, and control of the subject matter, and be confronted on this ground. They should not be immutable, but evolve. Mainly, they should “bend” to empirical evidence, being changed and discarded as needed. If that is the case, some mixture of (compatible) assumptions might not be a problem, but rather a variation that could be selected by better description, prevision, and control. Our target article’s aim was to facilitate such a dialogue and evolution.

Why Do We Need the New Terms Cultural Antecedents and Selecting Environment Variables?

Responding to the need for clarification, Sampaio and Haydu (2023) differentiated two functions of the cultural milieu by referring to the behaviors being affected: cultural antecedents affect the culturant (IBCs) and selecting environment variables affect the selecting environment (behaviors of members external to the culturant). First, we did not intend that distinction to account for all the roles of verbal stimuli in IBCs. Actually, we dealt with some of the roles of verbal behavior in metacontingencies by discussing group-rule generation. We contended that group-rule generation “is not an element of a metacontingency distinct from the culturant, but that it describes verbal components of the interlocking behavioral contingencies.” So, we acknowledged that verbal stimuli are part of IBCs. Second, these verbal components also include verbal responses composing the culturant—responses that produce verbal stimuli to other individuals. In sum, verbal responses and stimuli may participate in cultural antecedents, IBCs or selecting environment variables.

As for cultural antecedents, they are surely based on an analogy to the contingency of reinforcement. However, behavioral antecedents are not limited to discriminative stimuli, including, at least, motivating operations and contextual stimuli (Pierce & Cheney, 2017). Furthermore, behavioral antecedents are not limited to “discrete stimulus events with well-defined quantitative and qualitative properties.” Temporally extended events, relations among events, fuzzy categories, etc., can also have those functions (Pierce & Cheney, 2017). Analogously, cultural antecedents also encompass temporally extended events, relations among events, fuzzy categories, etc. What is fundamental is never to confuse the behavioral and cultural levels of analysis. Cultural antecedents (or cultural milieu, in this regard) cannot be discriminative stimuli as they pertain to a cultural—not behavioral—unit of analysis.

As for Ardila-Sánchez and Hayes’ (2023, this issue) concerns about selecting environment variables, we concur that this is a more generic term than cultural antecedents. But, as a science, we gain with the use of this term a helpful distinction between factors influencing the culturant and factors influencing behaviors involved in cultural consequences. That distinction was not clear with the use of only one term (cultural milieu). More distinctions should be proposed in the future. One hope this is only an initial step.

On the same theme: “selecting environment” differs from “selecting environment events or conditions.” The former encompass a set of individuals who produce the later. The selecting environment is a fundamental variable for the culturant. However, because it is composed of cultural and/or behavioral events, it is also affected by other environmental variables. In other words, some variables affect the selecting environment, which in turn affect the culturant. What we tried to do was to clarify some of the complexities involved in those relations.

Illustration, Implementation, and Interpretation

To illustrate the distinction between cultural antecedents and selecting environment variables, and our argument about verbal components of IBCs, our target article also described a COVID-19 Psychological Support Project undertaken at a public university. A previous version of the manuscript, reviewed by Ardila-Sánchez or Hayes, included a title and small parts that could imply that the COVID-19 project was directly informed by our proposal, that we analyzed some specific data about it, or that the project involved applied research directly relevant to our proposal. Neither was the case. The COVID-19 project was meant only to illustrate how the proposed new terms could be employed in the description of a cultural intervention. One of the authors of the target article was directly involved in the project’s coordination from its beginning, but we did not present an applied research—and neither intended to suggest that. Our focus was on a conceptual analysis.

On that note, we disagree that “one cannot say that a model is applied if there was no implementation during the assessment or at any point in the course of the intervention” (Ardila-Sánchez & Hayes, 2023, this issue). Behavior analysts sometimes say this; they use these words to describe this kind of work. What we cannot do, of course, is attribute the same importance and implications to (a) the incorporation of the model to plan or implement the intervention and (b) the kind of post hoc descriptive analysis we presented. The former is more productive and innovative. The latter, on the other hand, can be instructive while introducing new terms and concepts. One may liken this to an interpretative analysis or the presentation of examples. The main goal here is not to empirically validate the concept, but to logically clarify and test the conceptual coherence of the terms and theory.

Concluding Remarks

We thank Ardila-Sánchez and Hayes (2023, this issue) for the opportunity to debate and further clarify some of the ideas we presented in the target article. This dialogue exemplifies the importance of elucidating the elaborated account of metacontingencies and the possibility of researchers coming from somewhat different philosophical backgrounds to productively do science together. Of course, another way of doing so must be by conducting empirical research. We hope this dialogue also proves fruitful in spurring ideas for experiments, quasi-experiments, and other data-based investigations that come to light in the near future.

Data Availability

Data sharing is not applicable to this article as no datasets were generated or analyzed during the current study.

Declarations

Conflict of Interests

All authors state that there is no conflict of interest.