Acta Comportamentalia

Apresentação deste número

Why aversive control is not an option in clinical setting

As relações comportamentais estabelecidas entre as pessoas compõem o ambiente especial que nos distingue das outras espécies. Embora muito do sofrimento humano decorrente das relações com a natureza tenha sido amenizado com os avanços tecnológicos, aversividades advindas de relações entre indivíduos são ainda frequentes. Nas relações sociais, por emparelhamento, um indivíduo pode ser um reforçador condicionado e generalizado, porém, ao infligir estimulação dolorosa incondicional a outro, este pode, também, tornar-se um estímulo aversivo condicionado. Os efeitos do uso de estímulos aversivos já são bastante conhecidos: eliciação de respostas emocionais negativas, predisposição para fuga e ataque à fonte estimuladora, etc. Entretanto, tais efeitos não impedem a ocorrência do comportamento punido e dispõem o indivíduo a agir de qualquer modo que amenize os estados corporais indesejáveis produzidos, com alta probabilidade de interrupção da interação com o indivíduo controlador. O objetivo do presente texto não foi uma apresentação rigorosa de novos dados sobre controle aversivo, mas, uma reflexão sobre os efeitos já conhecidos. Muitos dos efeitos desastrosos do controle aversivo já foram descritos, assim como algumas pistas para entender porque abusamos dele. Porém, dado o grande desafio em promover o abandono de tais práticas, pesquisas sobre controle aversivo ainda são bem-vindas e necessárias.

The behavior relations established among people compose the special environment which distinguishes us from other species (the social environment or culture). Although much of the human suffering caused by the direct relations with the nature have been attenuated due to technological advancements, the suffering generated in the aversive relations among individuals is still frequent and an equivalent progress in the reduction of its effects hasn't taken place yet. As a function of the evolutionary history of the specie antecedent events contingent to the response-reinforcement relation acquire properties similar the consequent stimulus. In the social relations, as well as one can be a conditioned and generalized reinforcer, by inflicting painful unconditional stimulation to another, one can also become a conditioned aversive stimulus. Defining aversive stimuli as the one which selects the response that removes it, the damage to the social relations that the aversive control can exert can be estimated. The effects of the aversive stimuli use are already widely known among behavior analysts and it entails: negative emotional responses elicitation, predisposition to escape, attack against the stimulus source or acting in an incompatible way to the demanded, etc. However, such effects besides not preventing the occurrence of the punished behavior, predispose the individual to act in any manner that relieves the undesired corporal states produced, be it through the ingestion of substances, engaging itself in bizarre, dangerous or illogical behaviors or even interrupting the social interaction with the controller. In the clinical setting, these effects are frequently observed and aversive behavior interactions are the base of many of the diagnosed problems. Thus, it is contended that the use of aversive strategies in the therapy must be abolished not only because they don't solve the problem but also because they produce others. The objective of the present text was not a thorough display of new data about aversive control, but a reflection about the already known effects instead. Many of the hazardous effects of the aversive control have already been described, as much as some clues to understand why we use and abuse it. Nevertheless, hence the great challenge in promoting the abandonment of such practices, research on the aversive control is still welcomed and needed.

Two behavioral formulations of punishment: definition, explanation and some implications

As principais formulações teórico-conceituais de punição são as de Skinner (1953/2003) e de Azrin e Holz (1966/1975). Este artigo relata os resultados de um estudo que teve como objetivo explicitar as concepções teóricas e algumas das implicações dessas duas formulações. Para Skinner, punição é um "procedimento" definido pela apresentação de um estímulo reforçador negativo ou pela retirada de um reforçador positivo, cujo efeito supressor se deve à ocorrência de respostas competitivas. Skinner focaliza suas análises nos subprodutos indesejáveis produzidos por essa técnica. Azrin e Holz consideram punição como procedimento e processo. Seu foco de análise é o processo de redução comportamental, visto como efeito direto de certos estímulos contingentes ao responder. Entre as variadas implicações das duas formulações na área aplicada, encontram-se os posicionamentos distintos de autores incondicionalmente contrários e de autores parcialmente favoráveis ao uso da punição. No campo teórico, foram identificadas diferentes linhas de pesquisa e de interpretações sobre os mecanismos responsáveis pela supressão do responder. Dadas as diferenças constatadas entre essas concepções, discutem-se possíveis desdobramentos, para a Análise do Comportamento, de se utilizar uma mesma denominação para duas concepções distintas. Espera-se ter contribuído com o esclarecimento das particularidades que distinguem cada formulação.

In behavior analysis, the main current formulations of punishment are Skinner's (1953/2003) and Azrin e Holz's (1966/1975). This article describes a theoretical-conceptual research aimed at explaining the theoretical conceptions and some practical implications of these two formulations. It was conducted a study that consisted in selecting and examining bibliographies that presented discussions on theory, conceptions and applications (uses) of punishment. In Skinner's (1953/2003) formulation, punishment is understood as a "procedure", defined by him as the presentation of a negative reinforcer stimulus, or the withdrawal of a positive one. Skinner focuses the analysis of punishment on two aspects: the transitory nature of its effects and on its unfortunate emotional byproducts for the punished individual and for the society at large. To explain the behavioral suppression observed in punishment, Skinner turns to the process of negative reinforcement, consisted by the strengthening of competitive responses, and to the respondent conditioning through the elicitation of emotional responses incompatible with the punished response. On Azrin and Holz's (1966/1975) formulation, punishment is understood as a procedure and also as behavioral process, characterized by the reduction of the behavior probability. Thus, their study focus is to describe and explain the process of suppression or weakening of the punished responses. Authors in favor of this definition usually contend that the emotional byproducts of punishment, pointed by Skinner, are due to the type of stimulation used or to methodological issues and not to idiosyncrasy of the procedure itself. In this conception of punishment the efficacy in suppressing responses is considered as a result of the proper contingencies arrangement, and the explanation of its suppressive effects is a function of the direct relation between the behavior and the consequence, the same way as supposed for the positive reinforcement, but in the opposite direction. On the applied field, these distinct conceptions lead to a polarity among authors that are: unconditionally contrary to the use of punishment; and those who defend its use on critical situations, with the use of moderate aversive stimulus and in controlled settings. On the theoretical field, each definition tends to promote distinct research directions: on researching punishment, Skinnerians are more interested on the study of the effects of the aversive stimulation use; but the supporters of Azrin and Holz seem to seek the proper stimulus and conditions to suppress a given response class. It was verified yet that because of these different theoretical positions, it is also discussed the role of punishment in the operant theory. That means, it is questioned if punishment should be conceived only as a procedure or also as a behavioral process. Due to the great differences observed between these two conceptions of punishment, some possible inconveniences, for the Behavior Analysis as a field, are discussed in what relates to the use of only one term to relate to both. With the dada here presented and discussed, it is expected to contribute to a better comprehension of different positions behavior analysts have about this complex theme: punishment.

Early learning involving aversive stimuli: Initial considerations

O objetivo deste artigo foi discutir as questões: (1) Quando começam as primeiras aprendizagens envolvendo estímulos aversivos?; e (2) Como são essas aprendizagens? Para tanto, foram analisados estudos experimentais com diferentes espécies, incluindo-se humanos, que se debruçaram sobre aprendizagens pré-natais e aprendizagens durante os primeiros dias após o nascimento. Apontam-se evidências de que a exposição pré-natal a determinados estímulos possa alterar a sua eficácia em um período pós-natal - mesmo com estímulos considerados aversivos incondicionados. Além disso, verifica-se a formação pré-natal de pareamentos respondentes aversivos. Com relação aos primeiros dias de vida pós-natal, apontam-se dados sobre a dificuldade de formação de pareamentos respondentes aversivos quando estes concorrem com a aprendizagem de relações de vinculação com o cuidador. De maneira ampla, sugere-se a importância de determinadas aprendizagens pré e pós-natais na formação e fortalecimento do vínculo entre cuidador e filhote de forma que estas possam retardar o aparecimento de algumas associações com estímulos aversivas. Por fim, conclui-se que considerar o nascimento como o começo de uma história de aprendizagem é um equívoco. Sendo a aprendizagem um fenômeno cumulativo, olhar para o período pré-natal pode somar na compreensão de comportamentos complexos que surjam e/ou que se mantenham a partir do contato com estímulos aversivos.

From a radical behaviorist perspective, this article had as objective to look into the early learnings that involve aversive stimuli, selecting and analyzing an experimental literature facing answer two questions: (1) When do the early learnings involving aversive stimuli start?; and (2) How are these learnings? In order to do that, we analyzed experimental studies with different species that looked into learning that occurred in prenatal period and during first days after birth. We indicate evidences that learning begin before birth: the direct observation of fetus rats suggests that the formation of respondent relations with aversive stimuli already within 19 days of gestation in rats (Smotherman & Robinson, 1985); the induction of aversive association during gestation produces the non-choice of the conditioned stimuli (CS) after birth (Gruest, Richer, & Hars, 2004); and the prenatal exposition to stimuli which are normally avoided in early days of life increases the preference for these in the post-birth period, altering its function (Sneddon, Hadden, & Hepper, 1998). We also highlight the importance of attachment in the early days after birth, collecting data that indicate the neutralization of some aversive associations' formation, aversive associations that could harm the bond between the caregiver and the younglet in periods in which such attachment is crucial for the infant's survival: aversive association formation using milk as CS only happens before the 20 first days if the milk is not obtained by the suction of a female lactating rat's nipple (Martin & Alberts, 1979); and the association of neutral stimuli to electric shocks in rats' early days of life increases the preference for such stimuli (Roth & Sullivan, 2005). One of the possible explanations is that if the aversive association occurred with the same ease in the early days of life as they occur in adult age, the mother could become an aversive conditioned stimulus for having behaviors such as biting and kicking the younglets. Nevertheless, the neutralization of some aversive association in early life can facilitate the attachment, and its maintenance, to the caregiver. Therefore, we suggest that considering birth as the beginning of a history of learning with aversive stimuli is a misconception. In addition, the unconditioned aversive characteristic of certain stimuli cannot be justified by the presence of the aversive quality already in birth, since its function may have been altered by experiences prior to parturition. Being learning a cumulative phenomenon, considering prenatal period could count in comprehension of complex behaviors that arise and/or last from the contact with aversive stimuli.

Skinner and the asymmetry between reinforcement and punishment

O comportamento operante é definido como aquele que é afetado pelas suas consequências. Poderia se esperar, então, que enquanto alguns eventos ambientais posteriores ao responder teriam função fortalecedora, outras teriam uma função inversa, enfraquecendo-o. Entretanto, Skinner, ao discutir os mecanismos subjacentes ao fenômeno da punição, recorre a um modo explicativo alternativo, não estritamente selecionista. Adicionalmente, Skinner apresenta um conjunto de críticas ao uso da punição, enfatizando os seus efeitos negativos. Em tese, o reforçamento não teria os mesmos problemas e por isso seria uma melhor alternativa de controle. Reforçamento e punição seriam, nesse contexto, assimétricos. No presente ensaio foram identifi cadas e examinadas criticamente as oito principais objeções skinnerianas à punição. O objetivo foi avaliar se a caracterização que Skinner faz da punição se aplicaria ou não ao reforçamento, discutindo a partir daí a simetria ou assimetria entre os dois fenômenos. Observou-se que todas as operações e efeitos apresentados como típicos da punição (seus problemas intrínsecos) existiriam também no reforçamento. Portanto, usando a própria caracterização skinneriana, conclui-se que os dois fenômenos comportamentais seriam simétricos.

The cornerstone of the operant theory is the evidence that the responding is affected by its consequences. The consequences which increase the probability of responding are termed reinforcing, little divergences are found in what concerns to it. However, when the issue is about decrease in responding, different perspectives are confronted. When discussing the punishment's underlying mechanisms, Skinner turns to an alternative account, not strictly selectionistic, explaining the suppression of responding through the elicitation of incompatible emotional responses and the negative strengthening of responses that eliminate or reduce the aversive stimulation (be it unconditional or conditional). Additionally, Skinner presents a set of critiques to the use of punishment emphasizing its negative effects. Supposedly, reinforcement wouldn't have the same problems and, due to it, would be a better alternative for the control of behavior. Reinforcement and punishment, in this sense, would be asymmetric behavior phenomena. In another theoretical perspective it's stated that while some consequences strengthen the responding, others weaken it, contending a symmetrical relation between both phenomena. Extensive debates and studies concerning this issue have taken place, none has ever definitely refuted the other though. The aim of the present essay was to evaluate if Skinner's characterization of punishment would or not apply to reinforcement, discussing from this perspective the symmetry or asymmetry between the two phenomena. The eight main Skinnerian objections to punishment have been identified and critically examined. It was observed that the eight arguments proposed by Skinner as peculiar to punishment can also be extended to reinforcement, once in both relations: 1) continued exposure to the same events (aversive or reinforcing) produces habituation; 2) the effects over responding are transitory when the contingency is discontinued ; 3) there is elicitation of respondents which can interfere with the occurrence of the response; 4) conflicting relations may emerge when the same stimulus signalize two or more possible consequences (reinforcing or aversive); 5) other responses, not directly subjected to the consequences, are affected (induction); 6) stimuli, present during the history of selection, can acquire conditional functions (aversive or reinforcing); 7) the removal of the consequent stimulus (reinforcer or punisher) produces an automatic relation of control differing only in the direction of the effect (strengthening or suppressing); and 8) there might be an increase in the probability of aggressive responses. Thus, for all the operations and effects presented by Skinner as typical of the punishment (its intrinsic problems) there is also a counterpart in the reinforcement. So, making use of Skinner's own categorization, it was concluded that these behavior phenomena are symmetrical.

After all, what is aversive control?

Embora a literatura relativa à análise do comportamento descreva as relações que envolvem controle aversivo, esse termo não é, em si, definido claramente. O presente texto visa analisar conceitualmente o que significa "controle" e que fatores o caracterizam como "aversivo". O termo "controle" é analisado funcionalmente, considerando-se as probabilidades relacionais envolvidas nas relações operantes e respondentes. Destaca-se a característica de bi-direcionalidade do controle na interação organismo/ambiente, o que caracteriza essa relação como dinâmica e responsável pela renovação contínua do comportamento e pela sua individualidade. A análise dos fatores que caracterizam o controle é feita a partir das operações e seus efeitos comportamentais, bem como da natureza dos estímulos envolvidos nas relações nomeadas como aversivas. Conclui-se que nenhum desses fatores permite a caracterização objetiva do agrupamento das relações aversivas. Sugere-se que o maior desenvolvimento de estudos experimentais sobre respondentes encobertos (respostas "emocionais") pode ser um fator que ajude a tornar mais objetiva essa classificação.

Aversive control is present in most of the relations between an organism and the environment. However, this predominance has not generated a commensurate volume of studies among behavior analysts: in the majority of research in this area, only positive reinforcement in operant contingencies is arranged. This text assumes that a major development in studies of aversive control is a necessary condition for conducting behavior analysis in a more complete and balanced way, both in basic research and in applied conditions. The scientific meaning of the term "control" is discussed within Behavior Analysis, as well as the factors that characterize this control as "aversive." The analysis of the concept is functional, considering the relational probabilities between responses and stimuli that occur in a continuous interaction between organisms and the environment. We conclude that, regardless of their qualitative classification, the term "control" means that the probability of occurrence of an event (the controlled one) has changed by the occurrence of another event (the controller). This definition encompasses both operant and respondent relations. The bi-directionality of the control (between organism and environment) is considered the essential feature for the behavior to be understood as a dynamic process with continuous changes that produce individuality. Aversive control is then characterized by the "operations" (addition and subtraction of the stimulus), their "effects"(increase or decrease of the response probability) and the "nature of the stimulus" involved (aversive or appetitive). We conclude that the joint analysis of the operation and the effect is sufficient to define each behavioral relation, but does not help identifying the common factor that justifies their grouping under the name of aversive control. The development of experimental analysis on covert respondents (related to feelings or emotional responses) is suggested as an alternative criterion for the characterization of the aversiveness of the control to be defined more clearly without the use of circular reasoning or subjective considerations.

Who is afraid of aversive control?

Experimentos sobre fuga, esquiva e punição usando estimulação aversiva praticamente pararam nos últimos 30 anos devido a diversas razões, que incluem controle ético mais estrito. No lado aplicado, a análise do comportamento é conhecida pela forte preferência pelo uso de contingências de reforço positivo. Aqui discutimos a necessidade de se continuar a pesquisa básica que envolve controle aversivo, assim como estudos não experimentais em situações aplicadas, e o uso de dados obtidos por outras abordagens em psicologia e por qualquer outra ciência do comportamento. Controle aversivo faz parte da vida; nem sempre é ruim, nem sempre é evitável, e nem sempre é estressante.

Experiments on escape, avoidance, and punishment using aversive stimulation have practically stopped for the last 30 years due to several reasons, including more strict ethical guidelines. On the applied side, behavior analysis is known for its strong preference for the use of contingencies of positive reinforcement. Here we discuss the need for continuing basic research involving aversive control as well as non experimental studies in applied situations, and the use of data collected by other approaches in psychology and any other behavioral science. Aversive control is part of life; it is not always bad, not always avoidable.