Theory & Psychology

Criticisms of the very idea of experimentation in social psychology are longstanding; a focal claim recently has been that social psychological hypotheses are non-empirical. We contest this claim, but argue that many experiments in social psychology are pointless nonetheless because they are fundamentally circular. Testing hypotheses requires operationalization; operationalization requires assumptions; and in social psychology, we argue, the necessary assumptions often already imply that the hypotheses can be confirmed. Confirmability of the hypotheses of a number of experiments recently reported in the Journal of Experimental Social Psychology is shown to be implied by two illustrative truistic principles central to theories assumed in any tests of these hypotheses. We suggest that research aimed at finding specifically social psychological laws may only yield unfalsifiable truisms, while useful social psychological research aims elsewhere.

As Lise and Michael Wallach (2001) demonstrate, much social psychological research is gratuitous because, so often, experiments in that discipline merely confirm a set of unfalsifiable presuppositions. In spite of that major flaw, however, social psychological research may make important contributions, by revealing the effectiveness of manipulations that may be either stronger or weaker than common sense suggests, by helping us understand real-life phenomena, and by showing how desired ends may be achieved by practical means. My only quibble with the paper concerns the Wallachs' ideas about the status and functioning of psychological concepts. If they assume, as they sometimes seem to, that concepts are identities with independent existence apart from observation, I take exception. Social psychological phenomena should be treated as expressions of more general principles of a very general psychology that begins with empirical evidence.

We chart our agreements and disagreements with Wallach and Wallach's (2001) arguments. After listing some areas of agreement, we then make four arguments that undermine the Wallachs' project. First, we argue the discovery of `laws' is a goal neither of scientists nor of the discipline. Second, we suggest the Wallachs' claim that social psychological theories are infused with unfalsifiable theories about the nature of psychological variables misses the logical distinction between measurement models and conceptual hypotheses. Third, we look at the role that circularity plays in deductive logic and conclude that circularity is rarely a serious problem in practice. Fourth, we examine what effect bits of research have on scientific progress, and argue that `non-progressive' papers do not hinder progress. The best response to the Wallachs' concerns about social psychology is provided by a glance at progress within the field today. The state is very good; circularities, tautologies and untested assumptions do not pose a significant threat to progress in the science of social psychology.

While appreciating Kimble's (2001) praise of our paper (Wallach & Wallach, 2001), we argue against his reduction of mental states to behavioral dispositions. Crandall and Schaller (2001) make five points in criticism of our argument that much research in social psychology is plagued by circularity. We show that their first point is based on a misunderstanding, that their third point is irrelevant, and that their last two points do not attempt to answer our argument. This leaves their second point as Crandall and Schaller's only actual counterargument—namely that confidence in a measure for a construct is independent of confidence in the construct's relationships to other constructs. We show this claim of independence to be false.

This article examines a portion of the theory about and research on `race' in the field of psychology. Specifically, we examine whether the construct of `race' can be used efficaciously as a natural kind variable that explains data in psychological inquiry. The foregoing consideration involves an exploration of the so-called `biological/genetic' bases of `race' from the conceptual paradigms of (a) mutual exclusivity and inalterablity and (b) gradations on a continuum of genetic data. Our critique of these positions suggests an abdication of `race' as a natural kind variable because both of the above arguments for the concept are ultimately incoherent given the genetic data that they seek to explain. Consequently, `race' as a natural kind variable cannot explain psychological data. However, because the idea of `race' has become diffused into societies, `race' should be studied as a purely cognitive concept that influences social perception.

Discourse analysis has come to represent something of a `growth industry' in both research and critical psychology. Despite the apparent indebtedness of many such methods of discourse analysis to Foucault, there exists no strictly Foucauldian method of analysing discourse. Through a close reading of Foucault's `The Order of Discourse' (1981a), this paper re-characterizes the concept of discourse from a firmly Foucauldian perspective. Whilst not arguing against discourse analysis per se, the author indirectly takes issue with erroneous applications of Foucault's conceptualization by clarifying his perspective on what discourse is, and on what `discursive analysis' should entail. This critical presentation of the Foucauldian notion of the discursive will be contrasted with two prominent approaches to discourse analysis in psychology, namely those of Parker (1992) and Potter and Wetherell (1987). Key issues in this regard revolve around the themes of knowledge, materiality and history. By outlining the core components of what Foucault (1981a) terms `the order of discourse', and through the exposition of a four-step `method' of discursive critique, the author propounds an image of what a Foucauldian discursive analytic method may have looked like, should it have ever existed, before specifying exactly why one never did.

In this contribution we propose a comparative reconstruction of Kohlberg's and Freud's theories of moral development. We concentrate not on the developmental processes and their driving forces, but on the emerging conceptions of morality according to both theories. Each of the developmental stages is characterized in terms of the different sets of criteria typically constituting the changing (implicit) rationale for moral judgement and action. These criteria are represented according to the `logic of development' published earlier in this journal (van Haaften, 1998), which makes it possible precisely to describe the relations between the criteria within and through the successive stages in both theories. We also make some suggestions about stages preceding and following those that Freud explicitly mentions. Our reconstruction may serve as a first step towards a more exact comparison, and eventually perhaps partial integration, of different theories of moral development.

This review essay considers recent contributions to the Jung literature by Pietikäinen, Stevens and Yates. All three books comment on the interplay between Jung's life and psychology, but interpret his writings very differently from each other. Stevens looks to biology for corroboration of archetype theory, whereas Pietikäainen extrapolates from Cassirer' s philosophy a conceptual framework for redefining archetypes as cultural products. The anthology edited by Yates is quintessentially about symbolic manifestations of universal existential concerns with death and survival of the soul, though she opts for a more literal interpretation of Jung's concerns. Comparison of the books reveals the status quo of Jung scholarship, but hints on directions in which a study of Jung's polyphonic works could contribute to current debates and inquiries in academic psychology.