Mary B. Hesse’s Philosophy of Science and Her Legacy

As part of her seminal work on models and analogies in science, Mary Hesse develops a general 'network theory of meaning' in which she posits that all language, including scientific language, is fundamentally metaphorical. For this, she takes her cue from Ludwig Wittgenstein's theory of family resemblance (cf. PI, § 67), according to which general terms are applied to their diverse instances not on the basis of essential features common to them all, but rather on the basis of complex similarity relations (cf. PI, § 66). In this view, Hesse claims, any application of a general term to a novel case amounts to a metaphorical transfer of that term from its familiar domain of application (the source, as it were) to the new case (the target) – a process which always alters, however slightly, the meaning of the term (cf. Hesse 1988, 2). Whether the reapplication continues to be judged metaphorical or whether the new case is ultimately incorporated into the literal meaning of the term is more a matter of pragmatics than semantics, according to Hesse (cf. ibid., 3). Thus, in a sense, "all language is metaphorical" (ibid., 1). 
As Hesse explains, the metaphorical transfer of terms is based on the recognition of antecedent similarities between their old and the new cases of application. But these similarities are not, and cannot be, stated in literal (stable, univocal) language; for, there is, in Hesse's view, ultimately no such language. As she says, "[…] similarity and difference are irreducible primary relations, prior even to application of the simplest predicate: they are shown and not said" (Hesse 1988, 7). In this way, the network theory of meaning manages to preserve a nonparaphrasability and hence indispensability of metaphors while providing objective criteria of adequacy for them and hence a sense in which metaphors can be true or false. 
In this talk, however, it will be argued that Hesse's cognitive theory of metaphor relapses into a constructivist antirealism. For Hesse ultimately regards the similarities based on which all descriptive concepts are applied not as objective features of the world, but as a matter of psychological or social agreement (cf. Hesse 1985, 40). Thus, the truth values of the resulting classifications remain culturally relative and mind-dependent. In contrast, it will be shown that Wittgenstein's original theory of meaning already contains a theory of perception, developed from his private language argument and the famous rule-following paradox, that guarantees genuinely epistemic access to objective resemblances in the world. This access is established through a direct acquaintance – what Wittgenstein calls "familiarity" (PG, § 37) – of the conceptual agent with such resemblances, which is manifested in their competent (also linguistic) interaction with the objects of the corresponding resemblance classes and is neither conceptually nor phenomenally mediated. Wittgenstein's overall theory thus proves to be more realist than Hesse appreciates in her own interpretation and development of it, and can indeed help to redeem the realist program of Hesse's approach.

Contemporary debates on scientific realism take the 1999 standard version by Stathis Psillos as a starting point. His proposal delves with the long-discussed topic of realism from three stances: ontological, semantic and epistemic. In this paper I suggest to use Mary Hesse's insights on history of scientific change, especially in her 1961 book Forces and Fields (Hesse 1961), to counteract some of the criticisms that standard realism has received in the last two decades. Specifically, the focus will be on responding to the Pessimistic Meta-Induction (PMI) Argument, since PMI is the main argument against Scientific Realism. 


In order to understand Hesse's contributions to the Scientific Realism debate, I will first present Hesse's approach to explanation in science; Hesse introduces an account of science from what she calls 'satisfactory models', and presents it as an alternative to the Positivist and Realist's accounts of that time. Key characteristics of satisfactory models à la Hesse are; a) these models should have 'open texture' and b) they are characterized as intelligible thus, they are meaningful, testable and extendable. Within this context, giving an account of history of scientific change translates into giving an account of how models 'transition'.


The main challenges Hesse's faces are a) explaining continuity in science when a model has to be abandoned; b) explaining what happens with concepts that have been disregarded by current theories. To the first, Hesse affirms that there are 'traces' left by the theory and that those 'traces' account for continuity in science; Regarding the second, she posits that there is not such a thing as a 'death metaphor', since most key theoretical concepts are still relevant because of the aforementioned 'traces'. Both answers to these challenges have something in common: none of them make use of neither the term 'success' nor the term 'truth'. This is noteworthy because current challenges to Scientific Realism, especially PMI, hinge on the use of these terms by scientific realists in their explanations and definitions. For instance, Anjan Chakravartty's (2017) epistemic stance on Scientific Realism depends on the concept of truth. 


It is also worth noting that PMI is a challenge that comes from the History of Science, and it is from the same realm that Hesse's response originates. Larry Laudan first articulated PMI in 1981. To counter this argument, Psillos employs the 'non-miracle' argument to support realism. According to this perspective the success of a theory would, from this point of view, legitimize some sort of realism. However, critics argue, using pessimistic induction arguments, that the 'non-miracle' argument may be a fallacy. This is because history reveals that scientific theories, once considered valid, true, or proven, are often rejected, falsified, or abandoned. Consequently, this casts doubt on the validity of previous and possibly current theories (Levitt, 2005; Wray, 2018). In her analysis of the history of scientific change, Mary Hesse challenges this notion. Her ability to do so arises from her avoidance of explaining continuity in science using the terms 'success' or 'truth'.

From the 1960's onwards, there are significant critical voices in the philosophy of sciences against the dominance of positivism and traditional realism. These introduced new elements to consider, such as the role of the history in science or the relevance of the scientific community. Mary Hesse's work is part of this reconsideration, whose new realist approach departs from the more traditional conceptions, although she rarely appears in general histories of philosophy of science or even in handbooks. 
In this talk we will address her moderate approach to realism, named as modified realism. With this proposal Hesse distances herself from logical positivism, from Popper's critical realism and also from Kuhn's conception. However, in spite of this critical attitude, she does not suggest an absolute rejection either of the traditional theses, or of those derived from positivism, or of those proper of the 1960's in philosophy of science. Among the problems she considers, there is the idea whether, within science, a disruptive or a cumulative view should be chosen in order to account for scientific activity (Hesse, 2022: 283).
We will therefore first look at her realism as compared to traditional views. This will lead us to consider whether there is a middle ground between the extremes of scientific realism and relativism in her philosophy. To tackle this issue, we will thus approach her notions of truth (Hesse, 1970), and of scientific progress (Hesse, 1992). In our approach, we will consider a differential element transversal to all of Hesse's theses: the role of language. Through her network model it will be possible to establish both the influence of Duhem's and Quine's holism as well as the resonance of her ideas in later authors. 
Thus, the analysis of modified realism will allow us to situate her philosophical conception within the different possible considerations of science: either as a process or as a product. Hesse's proposal confronts naive positions which caricature or ignore scientific practice. We consider that in the light of her modified realism, a dynamic view within the different elements that conform scientific practice can be grasped. The purpose is to devise an account of science different from the incomplete views provided by some of the classical protagonists of 20th century philosophy of science. 
Hesse's views encompass both traditional scientific procedures and the new features of the philosophical turn of the 1960s, hence, her position is particularly relevant within 20th century philosophy of science. But also, given that she treats science as historical, social, and pragmatic, without forgetting its empirical character, it is a position worth of serious consideration for the development of 21st century philosophy of science.


References
Hesse, M. B. (1980). Revolutions and Reconstructions in the Philosophy of Science. Bloomington: Indiana University Press.Hesse, M. B. (1992). 'Science beyond Realism and Relativism' in Cognitive Relativism and Social Science (Edited by D. Raven; L. van Vucht Tijssen; J. de Wolf). New York: Routledge. Hesse, M. B. (1974). The structure of scientific inference. California: University of California Press.

In her Revolutions and Reconstructions in the Philosophy of Science (1980), Mary Hesse argues that natural science should be integrated into a wider epistemological framework embracing the philosophy of social science, hermeneutics, and the sociology of knowledge, in order to counter the traditional deductivist account of scientific explanation. In fact, she defends a post-empiricist view of natural science that closely resembles the hermeneutic analysis of the human sciences as expressed e.g. by Jürgen Habermas. That account is based on an instrumental rather than theoretical account of scientific knowledge, which Hesse conceives of as continuous with the hermeneutical model of understanding that, for Habermas, is employed in the cultural studies. According to that model, the structure of an object of study can never be analyzed to the point of eliminating all contingency (Habermas 1971: 161). Similarly, Hesse defends that in the natural sciences theoretical facts are only imperfect translations or interpretations of practical facts. Furthermore, she maintains that the deductivist account of scientific explanation should be replaced by the network model developed by Duhem and, later, by Quine, that is, the view that what is primarily significant in science is the interpretive expression we gave to what is observed.
In presenting these views, Hesse elaborates on an issue that has also been raised by Thomas Kuhn in a series of papers published after The Structure of Scientific Revolutions. In these papers, Kuhn deals with translation and interpretation in the history and philosophy of science, and argues that "no more in the natural than in the human sciences is there some neutral, culture-independent, set of categories within which the population – whether of objects or of actions – can be described. … Discovery is required in both cases, and hermeneutic interpretation is how discovery is done" (Kuhn 1989: 220). A similar conception (with the due differences, of course) can be found in Hesse. For her, interpretation of a theory should be understood as involving holistic stories about the concepts one aims to interpret. Moreover, Hesse famously maintains that to interpret foreign concepts implies the introduction of these concepts in terms of analogies and metaphors and, consequently, she argues that natural science cannot dispense with linguistic techniques of metaphors and modelling (cf. Hesse 1982: 707 and 709).
The proposed paper aims to deal with the aformentioned issues and to address Hesse's account of scientific explanation from a less explored perspective. Special emphasis will be given to Hesse's view that "society interprets itself to itself partly by means of its view of nature. But nature is informed by human meanings and is subject, in its theoretical aspects, to hermeneutic methodology" (Hesse 1980: 186). 
References
Hesse 1980, Revolutions and Reconstructions in the Philosophy of Science, Indiana University Press
Hesse 1982, "Comment on Kuhn's 'Commensurability, Comparability, Communicability'," Philosophy of Science 2, pp. 704-711.
Kuhn 1989, "The Natural and the Human Sciences", in T. Kuhn, The Road Since Structure, University of Chicago.
Kuhn 1982, "Commensurability, Comparability, Communicability", Philosophy of Science 2, pp. 669-688.
Habermas 1971, Knowledge and Human Interests.