Navigation – Plan du site
Comptes rendus de lecture

Leonardo PARRI, Explanation in the Social Sciences. A Theoretical and Empirical Introduction

Nicola Negri
p. 266-269
Référence(s) :

Leonardo PARRI, 2014, Explanation in the Social Sciences. A Theoretical and Empirical Introduction, Soveria Mannelli, Rubbettino, 254 p.

Texte intégral

1The aim of Leonardo Parri’s book is “the revitalization” (p.10) of the concept of social law. However, his support of the indispensability of social laws in order to explain social facts doesn’t imply that the standard model should be that of the hard natural sciences (ch.6). One has instead to turn towards soft natural sciences, like geology, climatology, etc., which deal statistically with high complexity phenomena. Let me consider the author’s arguments.

2In the first three chapters, Parri sums up the role of theories and laws in scientific inferences and conciliates philosophical realism with the impossibility of theory-free empirical observations. Following a path linking Durkheim with John Searle, he then admits the possibility of a social ontology, i.e. the conception of the social as an objective emergent reality, in spite of its undeniable social constructive character. On these bases, in ch. 4 the author is more precise about his revitalization of the concept of social law. With stringent references to Carl G. Hempel and Paul Oppenheim, as well as to Karl Popper, Parri situates the so-called covering law model, i.e. the core of nomological explanation, within the domain of inductive-statistical explanation (p.71). The revitalized social laws are, thus, probabilistic ones. Consequently, following the arguments of authors like Charles-Henry Cuin, Parri can relax the caveat which led Raymond Boudon, Jon Elster and John H. Goldthorpe to discriminate between laws and models, in order to account for the differences between natural and social knowledge. According to Parri, types and models are basically nomological products and implicitly presuppose the existence of probabilistic laws (p. 127).

3The covering law model in the domain of the inductive-statistical explanation infers its explanandum—e.g. the free riding level in an association which supplies public goods—from preceding conditions— the association size, the amount of its selective incentives—by means of a probabilistic covering law, Mancur Olson’s law (p.73-74). Thus, the why-questions concerning a “what?”, the free riding level, are answered by the establishment of causal probabilistic connections between macro phenomena. In ch. 5 and 7 Parri clarifies that the probabilistic laws which answer why-questions aren’t self-sufficient, they in fact “only partially open” (p.82, a vague statement, indeed) the black box of the causal processes which produce the macro-connections. Thus, in agreement with Hempel and Oppenheim, but also with Wesley C. Salmon and Mario Bunge, the author states that, in order to enter the black box, the answers to the why-questions have to be complemented with the analysis of micro mechanisms (that are themselves also ruled by laws), which can answer the how-questions (p.82-84). For example, in the inferences based on Olson’s law, one has to look into the individual decisions and actions of both the associational leaders and potential members (p.85). It is thus possible to integrate the probabilistic covering laws with methodological individualism and analytical sociology. This becomes clearer in the final chapter, where the Weber-Boudon’s model connecting micro and macro and the well-known James S. Coleman’s boat are discussed. Microfoundations are further inquired in ch. 8, devoted to intentional explanation, which rests on probabilistic covering laws derived from the theory of instrumental or value-rational action (p.180).

4The enlightenment of the unescapable role of nomological knowledge in all kinds of inferences and in intentional explanation, allows Parri in ch. 9 to criticize the historicist divide between history and social sciences. It is in fact now easy to show the epistemic impossibility of the radical inductivism cum empiricism of historicism, as well as the groundlessness of its pretense to dispense with nomological explanations, due to the fact that historians deal with the intentions of real actors (p.184-189). Parri shows that history is a social science, even though a sui generis one. It in fact resorts to theories and laws, explicitly in structural history and tacitly in more descriptive analytical narrations (p.193). On these bases, I think it is possible to look for synergies between history and analytical sociology. The latter fosters an evidence-based nomological knowledge, which can be used by history to achieve its goal: the connection of the relevant events in a complete and intelligible detailed description. On the other hand, such circumstantial historical descriptions, built on theories and laws, give analytical sociology useful elements for the contextualization of its models and the establishment of the appropriate ceteris paribus conditions. This would avoid the possibility that sociological formalization ends up, as Goldthorpe fears, in a sterile “sociological dandyism” without substantial relevance.

5All in all, the book provides an Arianna’s thread for facing the labyrinth of the abstract questions concerning the ontological and epistemological foundations of the social sciences. The scope of this achievement stands out if one considers that in contemporary sociology there is an inclination to appreciate extremely subjective sociological narrations, because of their critical significance, i.e. their capability of making public opinion aware of major social problems. Parri himself observes that, in the last quarter of the 20th century, sociology, together with anthropology, fell prey to this post-modernist trend more than economics and political science (p.53). Scientifically-inclined sociologists are thus pushed, by the rhetoric prevailing in their own field, to justify their scientific orientation. However, their effort to argue in a systematic way in favor of a scientific approach implies the risk of being entrapped in complex questions of philosophy of science. Paradoxically, the scientific-inclined sociologist is thus pushed away from the basic messages of Robert K. Merton’s Social Theory and Social Structure, which spurs to link theory and research, not to focus solely on theories detached from empirical observations, not to read only research books, but to maintain a practical research engagement. From this point of view, Parri’s book is an excellent contribution to the philosophy of social sciences. Moreover, some of its properties make the book useful for the social scientist who wants to move away from the dilemmas of the philosophy of social science in a conscious and non-skeptical way, in order to practice empirical research. In this respect, the fact that Parri is a sociologist helps. His competences help us link, thanks to several examples taken from empirical research, the postulates of the tradition of analytical philosophy (from Hume to Popper, Hempel, Searle, Hare, Bunge, etc.) to the models of the social processes of Boudon’s La Place du désordre and Coleman’s Foundations of Social Theory. The implications of the examined epistemological and ontological questions become thus transparent for the sociologist and the advanced student of sociology.

6Still, for a reader engaged in the problems of sociological theory and research, the book leaves some relevant questions open. Let’s begin with the concept of uncertainty. Parri’s central thesis is that the empirical laws of the social sciences can only be probabilistic, given the impossibility of establishing strict ceteris paribus conditions, able to halt the influence of “other relevant causes and disturbing factors” (p.99, p.127). Thus, uncertainty seems first of all due to the sociologist’s difficulty of having a synoptic vision of social reality (on its own ontologically certain), given the high number of dimensions which constitute it. If this were the problem of uncertainty, the progress of nomological knowledge in sociology would mainly be a question of improving survey devices and data bases, of dealing with the problems of undetected heterogeneity (e.g. by means of longitudinal data), of taking contextual differences more into account (e.g. thanks to multilevel analysis). However, this is not the case if one assumes that social reality is ontologically uncertain, as Parri himself seems to admit in his “La sociologia economica tra le mura dell’ordine e le scosse dell’incertezza” (Rassegna italiana di sociologia, v. 41/1, 2000). In this case, the progress of the social scientific nomological knowledge depends on the capability to model uncertainty, that is to account for the mechanisms which generate it. This would imply and deserve different kinds of methodological implications.

7The issue of modeling uncertainty is linked to a second question left open by Parri: value rationality. The author asserts that an actor aiming at a value performs, in a new or changed situation, a sort of rational calculation in order to identify the best means to realize it (p.172-173). However, this neglects the fact that a “value stance” refers to the “others” and can be conceived like a situation which Randall Collins defines as a “ritual”. In this situation, the sacred character of the pursued value-aim permeates the process which realizes it and, thereby, the involved means. In value-oriented actions, the resort to certain means can thus be considered like following a constitutive rule of the actor’s identity. In such a symbolically structured domain of action, it is improbable that the adoption of another rule in a changed situation could be considered like a neutral “technical” choice, which doesn’t in any way devalue the excluded alternatives. Therefore, the existence of a symbolic pattern, which connects a value with its means of realization, risks giving a problem of adaptation the meaning of a “reconversion”. This could engender identity crises and open areas of uncertainty for the actors, similar e.g. to “Buridan’s ass” situations. Value rationality could thus work as a mechanism generating ontological uncertainty in the social reality emerging from the interactions among actors, independently of the problems linked to scientific observation. The modeling of this process should thus include Collins’ theory of rituals and be considered by Parri’s theses on value rationality.

8The third question left open in the book concerns the techniques used for modeling social processes. Parri says that social experiments are impossible (p. 92, p.127). The social sciences would be in fact unable to create the artificial conditions which isolate the relevant relation from confusing factors, to define reliable counterfactual circumstances, to show in vitro that theoretical laws can generate the desired empirical configurations. Parri seems skeptical vis-à-vis the various experimental techniques used by the social sciences. Although he treats models like the Schelling’s residential segregation one (p.136), he doesn’t consider the agent-based models used in analytical sociology. Still, resorting to these models and experiments is important; in fact, the analysis of individual choices and their consequences requires distancing oneself from the atomistic modeling of rational choice. Furthermore, as showed by Peter Abell et al. (“Microfoundations of Social Theory”, Sociologica, 2014/2), the difficulty in resorting to rational choice models reduces the explanatory power of probabilistic macro laws. Hence, it seems there is a tight relation among: i) the microfoundation of probabilistic covering laws, stressed by Parri himself; ii) the difficulties in using the standard rational choice theory for these microfoundations; iii) the resort to agent-based models and experiments. This relation should be more closely inquired.

9To sum up, the need to go deeper into the three “open questions” I pointed out doesn’t challenge the framework of Explanation in the Social Sciences. On the contrary, it is thanks to the clarity of the book, that the three questions emerge and can easily be linked to the epistemological and ontological bases of the social sciences.

Haut de page

Pour citer cet article

Référence électronique

Nicola Negri, « Leonardo PARRI, Explanation in the Social Sciences. A Theoretical and Empirical Introduction », Revue européenne des sciences sociales [En ligne], 53-2 | 2015, mis en ligne le 15 novembre 2015, consulté le 22 avril 2018. URL :

Haut de page


Nicola Negri

University of Turin, DCPS

Haut de page

Droits d’auteur

© Librairie Droz

Haut de page