The challenge of giving a satisfactory answer to the question ‘what is an explanation’ has been a major concern in the philosophy of science since the days of Plato and Aristotle. This endeavor became particularly popular in the last sixty years when Hempel and Oppenheim wrote their classic “Studies in the logic of explanation” in 1948; however, we still lack an account that may be satisfactory representative of the heterogeneity of explanatory practices in science. 

From the first chapters of “Explanatory pluralism”, Mantzavinos argues against the main premises of that challenge and invites us to think out of the box with an innovative and clear proposal. According to the author, the explanatory enterprise should be approached as a plurality of “games” that are governed by different rules and values. Under this conception, explanations are part of a social process and, as such, are permanently unfolding, contextually and historically constrained. Therefore, the attempt to offer a single ideal structure that represents any adequate explanation in science seems to be misleading. Instead, Mantzavinos -faithful to his economic background- suggest an ambitious framework for understanding explainer’s behavior and choices: Systematically analyzing each explanatory games, their rules, and their normative features. These issues are carefully approached through the eleven chapters of the book, always with an eye on scientific practices, and numerous erudite references to examples from the history of science, mythology, and religion.

The book begins with a concise introduction to the problems that a philosophical theory of explanation should solve, which may be summarized in two main concerns: (1) to describe accurately the different facets of the explanatory practice; and (2) to provide a framework that account for the normative values that rule the assessment of explanatory activities and their outcomes. Shedding light on these issues constitutes the central aim of the book, but before addressing them, Mantzavinos dedicates some pages to critically discuss unitary perspectives about explanation (chapters two and three) and to present the general principles of his pluralist account (chapters four and five).

As already anticipated, the author objects the analytic effort that has been made toward developing a unitary model of what is supposed to be a scientific explanation, which is to lose the forest for the trees. Specifically, Mantzavinos points out that this kind of approach focuses only on the abstract concept of explanation, missing numerous features that characterize the explanatory enterprise in science. At the same time, unitary accounts seem to have only a limited range of application. For making this last point evident, on chapter three, Mantzavinos advances in the analysis of three examples of explanations in different fields of social science. Through these examples, he tries to show that none of the “three more influential models of explanation” is representative of all the examples but only fits one of them.

Although the arguments that Mantzavinos presents against unitary perspectives are highly compelling, the review about classic perspectives of explanation, and the analysis through the examples have some points that are not so accurate. In the first place, it is noticeable that Wesley Salmon’s perspective is qualified as a unitary account, considering that in later works of this author we find statements that are not really distant from the pluralistic proposal developed in the present book. Consider the following statement an example: “it might be better to list various explanatory virtues that scientific theories might possess, and to evaluate scientific theories in terms of them. Some theories might get high scores on some dimensions, but low scores on others (…) I have been discussing two virtues, one in terms of unification, the other in terms of exposing underlying mechanisms. Perhaps there are others that I have not considered.” (Salmon, 1990, p.20). This quote far from defending a single explanatory ideal seems to anticipate Mantzavinos’s proposal.

In the second place, considering the mentioned analysis of the examples, we also find some unclear interpretations. A sociological explanation, for example, is interpreted as being a causal mechanistic kind. The argument in support of this idea is that a social “mechanism” is identified for acquainting the fact that education does not increase social mobility. However, this “mechanism” does not rely on any intervention or manipulation, and consequently, does not fit neither with Salmon causal-mechanical ideas about explanation nor with the new mechanicist proposal. Similarly, the use of structural equation models in econometrics is interpreted as a case of Woodward’s causal manipulationist explanation. But then again, the case presented is not an explanation but a kind of practice and this kind of practice could be perfectly used also for identifying mechanism’s components. The problem here is that the author does not recognize the importance of interventions and manipulation for mechanistic explanations, neither the influence of Woodward’s perspective on new mechanicist (see Craver, 2007, Chapter 3).

These oversights, however, are only an exception in the methodical style exhibited throughout the book. Furthermore, they do not lessen the importance of Mantzavinos’s remark about the existence of a plurality of explanatory styles and the need for a framework that can do them justice. In this direction, the author proposes a naturalistic approach according to which explanations are conceptualized as products of problem-solving activities developed in a specific context. As anticipated, the different contexts and the set of rules that constrain them are defined as explanatory games.

Due to the naturalistic character of this approach, Mantzavinos warns that it does not involve metaphysical commitments to the ideas of causality or reductionism. This means that both causal and non-causal explanations could be considered equally adequate and that there is no privileged level of explanation. In addition, this perspective does not involve an apologetic stance toward a previously defined value but instead provides a structure for acquainting the different values present in each explanatory game.

In the same vein, Mantzavinos emphasizes the historical character of the explanatory enterprise as a social process that “consists of the attempt of the participants to this process to provide answers to puzzles and solutions to theoretical problems” (p.31). Then, explanations are defined as permanent unfinished works that involve constant changes in the ways of representing the phenomena and evaluating the plausibility of the proposed solutions. With these ideas in mind, Mantzavinos moves to the challenge of providing an abstract description of the explanatory games and their changes (chapters six to nine) as well as a framework for assessing their adequacy (chapters ten and eleven).

In this direction, the author firstly addresses and exhaustively disentangles the rules of the games. These rules are classified into: (a) Constitutive rules, that involve the background knowledge, metaphysical assumptions and conceptions of what counts as an “explanandum” for each game; (b) Rules of representation,  that define what kinds of representational bearers are considered adequate, in virtue of what these bearers represent the phenomena, and how they are connected to it; (c) Rules of inference, the inferential strategies that are used and considered adequate; and finally, (d) Rules of scope, that specify the range of phenomena to which the explanatory game can be applied. All these rules constitute an abstract framework for describing the features of any explanatory game being faithful to its own particular rules.

Following, the kind of analysis that can be done is illustrated through two historical examples: A case from economics concerning the value of the commodities, and a case of medicine about the functioning of the heart and cardiovascular system. This example’s analysis ends up being more focused on historical details than in a systematic examination of the rules. However, it exemplifies well the changes of rules that give place to new games over time and the plurality of games that may be found unfolding in parallel. These two dimensions for appreciating the plurality of games (over time and in parallel) are called by the author vertical and horizontal dimension correspondingly.

The emergence and change of rules take place in the process of learning of the explainers who engage in problem-solving activities. In front a specific problem to solve, explainers may interpret it as a known problem, and apply a solution that has been successful in the past. If the solution works well it is strengthened. Conversely, if the solution does not work any longer, or if the problem cannot be classified under any familiar kind, the problem is interpreted as a new one. Mantzavinos describes then all the strategies that are employed for solving new problems which involve inferential strategies of a heuristic nature, the use of cognitive and creative resources, imagination, among others. All this process is framed in a specific social context that gives the feedback that leads to a routinization and adoption of a problem solution, or to its rejection and searching for alternatives.

From this perspective, there is a continuum between commonsensical and scientific explanations, being the specificity of the institutional context in which these explanatory games are embedded what distinguish them. In this regard, Mantzavinos examines the features of three different kinds of explanatory context -children’s explanations, religion, and science- showing the particularities that these contexts imprint in explanations. In this sense, the possibility of criticism and revision are highlighted as two of the most outstanding features of scientific context.

After examining the features of explanatory games, Mantzavinos finally addresses the task of characterizing the normative features of the explanatory enterprise. With this aim, he remarks that the notion of explanatory pluralism should be distinguished from an “anything goes” perspective, and proposes a multidimensional critical appraisal for assessing explanations. This appraisal involves acknowledging the validity of multiple epistemic and non-epistemic values, as well as recognizing these values may be discussed and reformulated. Therefore, the evaluation of each explanation is comparative, not to a perfect ideal, but against a set of values that are contextually determined. These particularities lead the author to suggest that explanatory methodology is governed by hypothetical rather than categorical imperatives. In this sense, solutions are to be creatively discovered “provisionally accepting a series of values and normative criteria that have emerged and then inquire into how the different rules help to achieve them” (p.167). This perspective involves recognizing that principles and rules are prone to error, fallible, and might be constantly reviewed, where the institutionalization of the possibility of criticism is “the best means to facilitate the correction of errors when choices of rules are involved.” (p.177)

With these ideas, Mantzavinos concludes this highly ambitious and creative piece. I am sure any careful reader will appreciate the author’s committed effort in providing new strategies for solving old philosophical problems, moreover, for approaching explanatory practices sensitively to its particularities.  In this respect, it is worth highlighting his contribution of a clear framework for analyzing the rules of explanatory games, identifying the different features of explanatory dynamics, and assessing the normative features of explanations. All these issues give place to a systematic and pragmatic proposal that stands out among other pluralistic approaches. There are still, however, some problems to be pointed out.

In the first place, it is interesting to note that while this perspective conveys some tools for acquainting the particularities of different explanatory games, it does not address some key questions that have challenged pluralistic proposals. Specifically, how can we relate different kinds of explanations about a certain phenomenon, and how do we define the better kind of explanation’s approach. Following Mantzavinos, we may be able to analytically disentangle each explanatory game, admit a plurality of them, and even analyze the change of games through time. The problem is the implications that the notion of ‘explanatory games’ has for approaching the multiplicity of explanations occurring at a certain point in time, called by the author ‘horizontal dimension’. Interpreting these explanations as different explanatory games unfolding in parallel hampers the possibility of relating them. Since different explanatory games are thought to be constrained by a specific context, values, and rules, then we cannot relate them to a unified understanding of a phenomenon. In the context of scientific explanation, where it is common to find different kinds of explanations for acquainting the features of the same phenomenon, not taking into account how to relate those approaches seems to be a real flaw.

In the second place, as we anticipated, the richness of the Mantzavinos’s ideas excels their poor illustration in the examples selected. These examples leave the most interesting questions aside. Just to mention some cases, when the analysis of rules of games is exemplified through a case from economics and other from cardiology, the rules of those games are characterized very broadly. Those examples focus more on the dynamics of change of explanations rather than in the explanations themselves and do not provide an exhaustive analysis of the structure of explanations. Similarly, the exemplification of the horizontal dimension is illustrated in a very simplistic way, appealing to religious and scientific explanations of the origin of the universe. The most interesting thing about the horizontal dimension is to see how we can deal with a plurality of games in the same domain, and specifically in the scientific domain. To simply acknowledge that we may have different kinds of explanation in so different domains seems to be useless. These problems about the exemplification of ideas could be considered minor, but are not so innocuous in front a naturalized proposal that bases its principal arguments in the importance providing an approach that is sensitive to the particularities of explanatory practices.

Finally, and clearly less significant, the background discussions related to the ideas presented are not always clarified and many classical problems are passed over, which make it a risky book for beginners. Nevertheless, I am sure experts in explanation will welcome the new air these ideas bring and also will make the most of this proposal by applying it to the analysis of specific scientific contexts.  There is no doubt that this book opens numerous issues and questions to continue thinking a so complex and unanswered topic.

 

REFERENCES

Craver, C. (2007) Explaining the Brain: Mechanisms and the Mosaic Unity of Neuroscience. Clarendon Press: Oxford.

Salmon, W. C. (1990). Scientific Explanation: Causation and Unification. Critica, 22(66), 3–23.

 

© 2017 Itatí Branca

 

Itatí Branca is an assistant professor at the Faculty of Psychology, National University of Córdoba, Argentina. She is affiliated to the Research Centre of the Faculty of Philosophy and Humanities of National University of Córdoba and is finishing her Ph.D. in psychology.