How do textual practices function to make sense out of large data sets, when that “data” is not “given,” but rather narrated in extended prose form? Known as “cases,” “observations,” or “facts,” extended textual chunks formed a key currency of the human sciences into the twentieth century. Just how to structure such particularised units in order to produce generalised knowledge has been the subject of much recent scholarly enquiry. Medical cases might be synthesised into some ideal disease profile, or psychiatric observations ordered through paper technologies like the table. But generalisations of this kind were usually accompanied by (some) of their constituent cases, and often enough, collections of observations almost stood alone—like those added by Hippolyte Bernheim to his major works on (hypnotic) suggestion (1888, 1891). Focusing on emerging scientific psychology in nineteenth-century France, this paper considers individual observations as textual knowledge-making entities in their own right. It explores how narrative form and literary techniques articulate with the way psychological cases were generalised—as series, collections, syntheses. It traces conditions under which narrative elements—range of narrative voices, level of detail, temporal organisation—could work to destabilise the categorisation of a given observation. Conversely, some literary techniques generated larger narrative arcs out of groups of observations, thereby reinforcing the epistemic or ontological weight of their sequencing, while diversity in narrative form could also hold observations apart. Ultimately, mapping textual practices against knowledge-making functions provides insight into the ways psychological forms of case-writing evolved out of older disciplinary traditions.

My talk starts with some general definition of the term “replication” and remarks on the historiography about experimentation, recalling the discussion about this feature of scientific practice in the works of Shapin & Schaffer; Collins, Galison, etc. But in these works, no mention is made on how psychologists became concerned and used replication. Already Kant, who examined the possibility of psychology becoming a real science, had mentioned as one obstacle the non-replicability of psychological introspection. Therefore, for Wundt and his colleagues the demonstration of psychological experiments being replicated was crucial. In my talk, I will a) expose views on replication and replicability expressed by leading psychologists such as Wundt, Titchener, Watson and Dunlap; b) offer examples and compare how this tool was used in some empirical studies in connection with psychological methods (e.g. introspection, association experiments and mental testing), and c) examine the role replication played in controversies of the time. The examination of historical sources shows what scientists in the past understood by “replication” and why they thought that this should be part of the process of knowledge construction. I conclude that replication gained historical prominence, as soon as the main objective of psychological research was to identify laws. In general, it seems that the debates about replication deepened consciousness of psychologists about problematic aspects of psychological experimentation (in general), influencing the standards of scientific research.

In 1935, Stanley Smith Stevens published two articles in which he urged for a revolution in psychology. Building on P.W. Bridgman’s methodological prescriptions for physicists, Stevens argued that all psychological concepts need to be strictly defined in terms of public operations. If psychology is to be taken seriously as a rigorous science, Stevens argued, psychologists have to make sure that they are not talking at cross purposes when they are discussing their theories about ‘experience’, ‘consciousness’, and ‘sensation’, they have to make sure that their concepts are ‘operationally defined’. In view of Stevens’ success in spreading the operationist message, it is small wonder that he is widely viewed as the intellectual father of psychological operationism. In this paper, however, I argue that Stevens was not the first scholar to translate Bridgman’s conceptual strictures to psychology. I show (1) that Gary Boring and B.F. Skinner had already been applying Bridgman’s approach to psychology when Stevens was still an undergraduate student and (2) that Douglas McGregor coined the term ‘operationism’ before Stevens. Since Boring, Bridgman,McGregor, Skinner, and Stevens were all affiliated with Harvard in the early 1930s, the question arises what role these scholars played in the development of psychological operationism. In this paper, I answer this question by examining a large set of documents from the Harvard University Archives. Instead of taking Stevens’ papers as the *starting point* of the operationist turn, I reconstruct the intellectual climate at Harvard in the years leading up to the publication of the operationist manifestos.