Abstract
D.Phil. (Psychology)
In the first part of the thesis the need for theory development in psychology is sketched by a review of the relevant literature, thus justifying the necessity of theoretical research in psychology. The appropriate method for theory development - comparative metatheoretical analysis - is delineated, and situated against the background of two
major approaches to theory development in psychology: the transcendental and the empirical. The key terms that form the basis of most of the analyses are defined, and the aim and scope of the thesis is briefly stated. The second part consists of a review of the competing philosophical views regarding theorizing and scientific inquiry in the social sciences, with the goal of developing an adequate metatheory for psychology. It starts with a historical overview of developments in the philosophy of science,
emphasizing issues that are relevant to psychology. The development of scientific realism is reviewed, with a specific focus upon McMullin's empirical argument in favour of realism, and Bhaskar's transcendental justification of realism. It is also shown how realism was adapted by Bhaskar in order to suit the social sciences. Social constructionism the diametric opposite of realism - is then reviewed, as well as the
intermediate positions of Manicas and Rosenberg, and Layder. These positions are all critically evaluated, and a choice in favour of a realism as the most appropriate metatheory for social science and psychology is made. In the third part of the thesis the principles of the realist position are developed further and applied to psychology. In this regard, a framework is developed to assess the epistemic status of psychology as a distinctive and legitimate social science, when seen in relation to the other social sciences. The implications of a stratified reality for psychological explanation is examined, and these implications are also linked to psychology' 5 distinctive subject matter and domain of investigation. The epistemic status of psychological explanation in relation to ordinary language accounts is examined, and the influence of
stratification in psychological explanation is analyzed. A substantive realist position for psychology is then developed, which starts with a brief review of Manicas and Secord's realist position. The realist position developed has some differences in emphasis when compared to that of Bhaskar, most notably as far as the notion of structural
explanation is concerned. Four theoretical case studies the theoretical debate between Guilford and Eysenck, Gustaffson 's structure-of-intellect theory, Campbell and Fiske's validational model,and cross-cultural ability research in Africa - are used to illustrate
the thesis that psychological theories are in fact realist explanations aimed at approximating the causes of observed effects. This development of a substantive realist position is followed by a comprehensive analysis of the epistemic power of psychological theories. This analysis is initiated by distinguishing between various levels of causal explanation in psychology, and a framework of levels of causal
explanation is developed and linked to the notion of explanatory power. The realist view of psychological theories also entails that psychological theories are seen as having certain epistemic characteristics, and these characteristics can also be seen as evaluative criteria. A model for theory evaluation in psychology is developed in which these criteria are subsumed, whereby theories can be evaluated according to two major axes of evaluation: an axis assessing the epistemic gains effected by any given theory; and an axis which assesses the influence of external factors (or the social domain) upon the epistemic status of a theory. The utility of the model is demonstrated
by two applications: a preliminary assessment of Sternberg's theory of human reasoning, and an extended case study which charts the development of Festinger' s theory of cognitive dissonance, and which culminates in the evaluation of the theory.
The epistemic gains achieved by the model for theory evaluation are discussed, and the thesis concludes with an afterword where the benefits and limitations of the investigation are discussed.