Understanding Perspectivism

Michela Massimi and Casey D McCoy

Reviewed by Jordi Cat

Understanding Perspectivism: Scientific Challenges and Methodological Prospects
Michela Massimi and Casey D McCoy (eds)
London: Routledge, 2019, £120.00 / Open Access
ISBN 9781138503069

In memory of Ron Giere

This volume, edited by Michela Massimi and Casey D. McCoy, consists of ten essays whose goal, according to the title, is understanding perspectivism. The essays address the themes of pragmatism, pluralism, and realism that characterize Ronald Giere’s well-know account of perspectivism (Giere [2006]). The editors are forthcoming about the fact that different authors use the term ‘perspective’ differently, as if introducing different perspectives on perspectivism. Do considerations of perspectives serve, in some capacity, the purpose of addressing other themes or problems? Do considerations of such themes or problems in turn serve the announced purpose of addressing the problems and prospects of perspectivism? Both? Neither? The different essays illustrate different designs and uses of what one may call a PIPO (perspective in, perspective out) machine.

But what’s in a word? One guiding issue in this review will be the various uses to which the term ‘perspective’ is put in philosophy. Philosophical terms, or scientific ones for that matter, are typically introduced as either philosophical metaphors, technical terms, or a hybrid of both. Whether open-but-centred metaphorical uses or more strictly bounded and precise technical ones, philosophical notions and their associated notations are expected to serve a function, either in the conceptual articulation of a view or argument, or in their presentation.

Philosophy doesn’t differ from other disciplines in terms of the social role played by linguistic and other symbolic trends. Philosophers too may place untold credit on trend-setting newly coined terms, even when they don’t track any relevant or novel concepts. Sometimes a trend is just a trend, and a label just a label, and herrings are red. Is perspective talk ever useful—and if so, how?—or ever dispensable?

To flesh out the philosophical significance of the themes of realism, pluralism, and pragmatism in relation to perspectivism, let me first introduce how they emerged in Giere’s naturalistic account of scientific perspectivism. In his ([2006]) book, Giere introduced the expression ‘scientific perspectivism’. He traced the unqualified philosophical view he called perspectivism to earlier philosophers such as Leibniz, Kant, and especially Nietzsche.

Talk of ‘perspectives’ and the introduction of perspectivism in science was, according to Giere, in response to a specific problem: the conflict between realism and constructivism in the studies of science, which was emblematic of the more widespread—and more rhetorical—science wars. It is also an alternative to a universalist and a priori approach to the world and its scientific study. And the systematic, metaphorical use of the term ‘perspective’ serves a heuristic function, connecting and emphasizing relevant features of scientific practices and products. For Giere, it is grounded in a methodological tool of naturalistic standards, or, in his terms, a naturalistic perspective. Ultimately, his attempt to bring about reconciliation between the warring viewpoints and projects was what lead him to the position he called perspectival realism.

In Giere’s view of scientific inquiry, scientists can achieve varying degrees of scope and accuracy in the representational—and instrumental—value of constructed models. But they can never have complete and exact knowledge, only imperfect and partial (incomplete) fit. According to Giere, perspectives are constituted by the construction of scientific models and especially by the role of general principles. The issue of realism concerns what this kind of knowledge can be knowledge of, and the commitment to a corresponding ontology.

According to Giere, multiplicity is the distinctive mark of perspectival knowledge and inquiry. The relevant multiplicity, or plurality, is rooted in the unavoidably incomplete character of maps or models, instruments, theories, doctrines, disciplines, and so on aiming at representation. Plurality also provides possibilities both for a cognitive division of labour and for cooperation. Giere considers them as the social epistemology of theoretical and experimental modelling and, specifically, as examples of distributive cognition (another bridge between realism and social constructivism).

Lastly, besides naturalism, realism, social cognition, and pluralism, Giere linked his scientific perspectivism to pragmatism. This link stems from his rejection of a priori, universal foundations and his emphasis on scientific methods and goals, with their own practical rationality (which in turn relies for inference, planning, and action on realism about causal relations).

Much of this perspective talk is interchangeable with other terms that Giere and this volume’s contributors have used in the same cases and contexts, for instance, in talk of a theoretical, methodological, naturalistic, and pragmatist stances. For decades, post- or neo-Kantian talk of frameworks, conceptual schemes, even paradigms has done similar work in philosophy of science and epistemology more broadly (see, for instance, Michael Lynch’s ([1996]) work on truth in context, pluralism, and objectivity). ‘Stance’ talk has been distinguished from talk of frameworks, most notably by Bas van Fraassen ([2002]). Stances are taken to include values, preferences, emotions, even policies (see, for instance, recent work by Martin Kusch ([2017], [2019]) on relativism). For purposes other than those mentioned above, the term ‘perspective’ seems ostensibly dispensable.

Giere’s ([2006]) views, as well those of contributors to the Massimi–McCoy volume, resonate with an abundance of broad and detailed epistemological work. This should enrich ongoing discussions in the literature, for example, that on relativism and disagreement (see, most recently, Kusch 2017], [2019]), on approximate truth (see Niiniluoto [1987]; Boyd [1990]; Kuipers [2019]), the latter also a defender of what he calls critical scientific realism), and on pluralism (see, for instance, Rescher [1993]; Lynch [1996]). And there is also the considerably older but still growing literature on naturalism and pragmatism.

Now, back to the volume. In the opening essay, Hasok Chang joins the pragmatist tradition in pointing to the possibility of a humanist account of science, by which he means the business of human agents. Over the last few decades, pragmatism, as well as metaphysics, has sought to pursue a critical philosophy away from the so-called linguistic turn. In philosophy of science, the corresponding efforts—again, alongside metaphysics—have become known and totemically advocated as the practice turn. Like Giere and others, Chang finds pragmatism conceptually entwined with considerations of society, plurality, historicity, and realism. He concludes the chapter defending on those grounds the methodological value of integrating the disciplines of history and philosophy of science—especially their interdependent, critical, and normative functions.

Along the way, we get a discussion of perspectivism—which while part of Chang’s pragmatist argument, may come across as a side polemic—criticizing Massimi’s less radical account of perspectival realism. Chang seeks to out-Kant her by pointing to the cognitively, as well as historically and culturally, situated character of knowledge claims. This is taken to stand in contrast to Massimi’s account, which points just to the truth conditions of knowledge claims, alongside some aperspectivally representable ontological reality to satisfy them.

In her essay, Melinda Bonnie Fagan is concerned with a problem in the social epistemology of scientific modelling: how is effective interdisciplinary collaboration possible when it seeks to explain a complex phenomenon while holding incompatible standards of explanation? Rather than posing a philosophical challenge for perspectivism, she sensibly identifies the source problem as a scientific one, with broad epistemic and social dimensions. Fagan brings in perspectivism, looking to Giere and the philosophical debates from the last decade for an analogy that will play a heuristic role in addressing this problem. She garners an array of eight broad types of possible relations between perspectives that she organizes into a taxonomy applicable to explanatory models. She uses two distinctions: direct/indirect and similarity/difference. Other chapters in the volume act as case studies for Fagan’s taxonomy, allowing it to be tested against somewhat different but related problems (or perspectives?), and with different conclusions.

In his essay, Paul Teller follows Giere on the path from the central role of models in scientific representation to the unavoidable perspectivism and partial and inexact knowledge. What sort of realism may philosophers find room for? Teller laments that Giere has said too little. To advance the analysis, he offers a critical, negative account. If the characterization of model-based knowledge is to be (approximately) trusted, the same inexactness will characterize any scientific realism. But this conclusion, he argues, is incompatible with the tenets of standard realism, specifically its referential element: exact existential claims that may be reformulated as the assumption that scientific theories provide approximately true claims about things included in the non-empty extensions of the important terms in their formulations. Yet such extensions cannot be fixed. For a positive account, Teller might have mentioned or had recourse to any of the sources I have mentioned above on approximate truth and realism.

By contrast, Juha Saatsi puts forward a critique of perspectival realism in relation to the problem of plurality of incompatible perspectives, and to two other assumptions of standard scientific realism: explanation and progress. In this sense, his is one of the essays that complement Fagan’s. He distinguishes between explanation and explanatory understanding, and draws attention to the pragmatic and contextual, non-factive elements that contribute to understanding. In particular, he wants to include the valuable role of counterfactual understanding in the form of answers to what-if questions. This helps address the issue of the integration of putatively incompatible explanations, such as semi-classical explanatory models and approximations that combine models of, for instance, light rays and surface waves. The distinction between factive and non-factive elements helps circumvent the typically factive incompatibility that challenges realism.

What about perspectives? Saatsi uses expressions such as ‘perspective’ for frameworks, ‘explanatory perspectives’ for explanatory models, and ‘theoretical perspectives’ for theories. This seems to do little more than help emphasize the source of conflict between explanatory models and the inclusion of pragmatic, contextual elements. He then uses the occasional ‘perspective’ expression to translate his conclusions about realism into more specific conclusions about perspectival realism.

Against the background of realist commitments, Collin Rice and J. E. Wolff also tackle the problem of conflicting explanatory models—in his essay, Rice frequently calls them inconsistent (while Giere and others have warned against this narrower characterization, applicable only to a logical relation between propositions).

Unlike Saatsi and others, Rice wants to uphold a standard of genuine explanatory understanding restricted to the so-called factive content of models. Saatsi’s solution to the problem, then, will not do. Rice’s move, however, involves extending the scope of typically idealized and inaccurate explanatory factive information to the modal or counterfactual, which he considers more exact and thus in line with tenets of standard realism. Rice and Staasi simply place counterfactual information in different parts of their respective account.

Where lies the reality component of explanation? Rice’s alternative to perspectival realism borrows from physics the technical concept of universality class. The concept is applied in cases of multiple scale-dependent characterizations of physical phenomena (notably in condensed-matter physics). Accordingly, the reality captured by integrated explanatory models would reside in a more abstract pattern of shared similarities. It is a case of integration relying on similarities, as identified by Massimi, Fagan, and others. However, the strategy might not work in cases where the universal pattern is not what scientists find genuinely and usefully explanatory of the phenomenon at hand; but then we might be witnessing a conflict of standards. Rice’s examples, from physics and biology, illustrate the problem, but aren’t really examples of the proposed solution. And the solution qualifies as a solution to a problem of perspectivism only to the extent that he identifies a version of the problem as the one perspectival realism meant to solve.

Wolff also examines an alternative to perspectival realism, namely, structural realism. But although the proposal is in important respects similar to Rice’s, she’s quick to point out that she intends hers to be complementary to perspectival realism. As she points out, while perspectival realism focuses on differences between models or representations, structural realism focuses on commonalities (as does Rice’s universality solution). The purpose of her essay, however, is not to present structural realism as a novel account of realism, but to present it as a solution, complementary to perspectival realism, to the same problem of plurality and potential incompatibility. And by way of demonstration, she examines the case of the representational theory of measurement.

Wolff provides an exemplary brief presentation identifying what she calls the additivity paradigm (she might as well have called it perspective): the assignment of quantities to empirical attributes in such a way that relations between numbers bear the relevant similarity (mapping) to relations between empirical attributes. From this relation, she concludes that the theory can be understood as a form of structural realism about representation. Some readers might have qualms here about her standard of realism in relation to other discussions of the problem.

In the end, however, she still manages to identify a type of perspectivism that may be called perspectival representationalism. Measurement theory’s structural realism lacks, she argues, the resources to justify the attribution of particular metrical structures to particular empirical properties. What makes this perspectival? She calls a perspective the theory-dependent choice from a plurality of possible metrical structures. And if representationalism is a kind of realism, is this perspectival realism? It counts as perspectival realism provided that the different enabling theories themselves are each considered both mutually incompatible and complementary. The essay is a contribution to the broader debate over realism, yet effectively served by discussion of perspectivism. Insofar as structural realism also complements the element of perspectival realism, the complementarity relation is a critique of both or, at least, of their exclusive and unrestricted versions.

David Danks also argues in his essay that scientific perspectivism poses a challenge to standard realism. His criticism is that even when not called by that name, perspectivism risks being either banal or ‘unsafe’—portraying science as contingent, ‘hyperlocal’, and individualized, and implying radical relativism. Ultimately, he aims to defend an alternative he calls a safe and significant kind of perspective, characterized by the concepts and goals that filter and transform information for individual scientists and their communities. And it is at the individual level that scientific perspectivism proves to be (naturalistically) continuous with the perspectival nature that psychology research claims for ordinary cognition, in both its limitations and its successes.

The final three essays in the volume feature case studies that draw different conclusions about pluralities of perspectives. Mazviita Chirimuuta looks to different models of the motor cortex to offer yet another critique of perspective realism. For a taxonomy of possible relations between different models—and perspectives—of the same system, she settles for a distinction between competition and complementarity. The two perspectives at stake, which Chirimuuta is also satisfied to call approaches, are the intentional and dynamical perspectives linked to each kind of cortex model.

Her characterization of the contrast relies on a rich notion of perspective that includes theoretical frameworks—including physiological and mathematical models—disciplinary backgrounds, experimental protocols, data analysis standards, explanatory standards, mathematical formalisms, and simplification techniques. Next she introduces a distinction between the mathematical models and their philosophical interpretations to note that both perspectives share key mathematical models and clash on their broader interpretations. But this leads to a critical dilemma for the significance of pluralism and realism: either the difference between the mathematical models is instrumental, with no implications for realism, or the competition between rival interpretations has implications for realism, but none for pluralism. Faced with either pluralism without realism or realism without pluralism, perspectivism—at least in this case and in these terms—simply fails to support perspectival realism.

Anya Plutynski looks to cancer models to defend theoretical pluralism in cancer research. She wants to contrast thinking in terms of perspectives or cognates, with connotations of partiality and plurality, to the monism and dogmatism of thinking in terms of paradigms. On the premise that cancer is intrinsically a complex phenomenon in a problem-driven field of research, she concludes that its understanding and treatment are best served by different yet complementary kinds of models, each treating a different kind of cause, often on a different spatial or temporal scale. Historically, the same way of thinking suggests a piecemeal model of progress rather than the revolutions suggested by paradigm talk. Thinking in terms of perspectives plays a heuristic role here.

The collection closes with an essay by Sandra Mitchell in which she looks to research on protein structure and the plurality of models of protein folding. Combining her own familiar defence of integrative pluralism and Giere’s presentation of perspectivism, she argues that perspectivism provides the intellectual resources to identify and understand the epistemic virtues of integrating compatible models in actual scientific practice. The sort of plurality that perspectives can identify stems from distinctive epistemic characteristics of modelling practices: selection or inclusion of features, and partiality (incompleteness) or exclusion. Both involve the intentions of model users and their context, as well as constraints from any relevant method and medium.

The notion of perspective that can do this plurality justice is open-ended and rich. She includes considerations of driving problems, assumptions, methods, instruments, experimental arrangements, different kinds of representations, and explanatory and predictive standards. Differences in goals and standards can already account for the epistemic value of plurality. For instance, physics, chemistry, and biology provide broad perspectives on protein structure. For models in a relation that is neither one of reduction nor one of incompatibility, she points to the additional epistemic value of integration. One example is the role of top-down constraints in the application of models from different perspectives. Chemical factors may constrain thermodynamical processes and biological factors may constrain chemical processes. The role of each kind of model enhances the understanding of the situation captured by the other. Another, and perhaps the more striking, example of epistemic value is methodological: the important role of joint refinement methods. According to Mitchell, besides heuristic value, thinking in terms of perspectives—suitably understood—provides a framework for the explication, guidance, and assessment of actual modelling practices, something that cannot be provided by empirical description alone.

To conclude, the informative and nuanced essays in this volume will help the reader consider the philosophical scope and value of perspectivism and of talk of perspectives. Many of the proposals seem complementary in that they identify different kinds of cases to which perspective talk can apply, and the different uses and different kinds of claims about those cases. Different authors struggle to stick to perspective talk without shifting into using familiar technical terms such as ‘approach’, ‘paradigm’, and especially ‘framework’, all thinly abstract and used analogically or metaphorically. In fact, perspective talk seems as unstable as the conceptual content it is meant to track, explore, and exploit. But it shows how philosophical language and its concepts so often rely—yes, usefully, but unavoidably?—on an ambiguous hybrid of the metaphorical and the (ideally) technical. As Nietzsche might have pointed out, ‘perspective’ might have a more valuable history than definition.

Jordi Cat
Indiana University
jrecat@gmail.com

References

Boyd, R. [1990]: ‘Realism, Approximate Truth, and Philosophical Method’, in W. Savage (ed.), Scientific Theories, Minneapolis, MN: University of Minnesota Press, pp. 355–91.

Giere, R. [2006]: Scientific Perspectivism, Chicago, IL: University of Chicago Press.

Kuipers, T. [1990]: Nomic Truth Approximation Revisited, Cham: Springer.

Kusch, M. [2017]: ‘Epistemic Relativism, Scepticism, Pluralism’, Synthese, 194, pp. 4687–703.

Kusch, M. [2019]: ‘II-Relativist Stances, Virtues, and Vices’, Proceedings of the Aristotelian Society, 93, pp. S271–91.

Lynch, M. [1996]: Truth in Context: An Essay on Pluralism and Objectivity, Cambridge, MA: MIT Press.

Niiniluoto, I. [1987]: Truthlikeness, Dordrecht: Reidel.

Rescher, N. [1993]: Pluralism: Against the Demand for Consensus, Oxford: Clarendon Press.

Van Fraasen, B. [2002]: The Empirical Stance, New Haven, CT: Yale University Press.