Hesse and Ramsey and Model Pluralism

Models B Feb 2020 Hesse and Ramsey and Model Pluralism Peter P. Kirschenmann (Emeritus) Faculty of Philosophy and Faculty of Exact Sciences Vrije Universiteit, Amsterdam, Netherlands Borssenburg 6 NL-1181NV Amstelveen, Netherlands Tel.: **31-20-6453336, E-mail: jorakir@hotmail.com Outline: Should one creatively proliferate models, in science and theology, amongst other areas? Now, models come in many types and roles. In view of the question raised, I consider cognitive or epistemic models, which include theoretical and explanatory models or models fostering insight and understanding. I take the accounts of two pioneering thinkers about models in the philosophy of science and theology, Mary Hesse and Ian Ramsey, to find out what they might imply for the proliferation question, which will always also involve the question of how to test models, how to choose between possibly incompatible models. I present Hesse’s analysis of scientific models in terms of analogies, and point out that her idea that what she called ‘neutral analogies’ should be further examined implies that there will be successive improved models. Being aware of the many models considered and used in science, she also discussed the issue of choosing between competing models; she concluded that there was no unfailing way of justifying such choices, although eventually there usually are decisive tests. Ramsey distinguished between “picturing” and “disclosure” models. Picturing models are purely descriptive, also in theology, and therefore leave to place for mystery. Disclosure models, which for Ramsey have also become significant in science, are born in moments of insight when the universe discloses itself. Theological disclosure models are metaphoric rather than analogical, with metaphors rooted in disclosures. These models are not explanatory, but attempts at articulating, and reaching some understanding of, the divine mystery. They do not entail testable consequences, but are judged by their “empirical fit”, their capacity of harmonizing the most diverse events. To articulate what a cosmic disclosure discloses, he asserted, we shall need a never-ending succession of theological metaphors; he supported this assertion with his further analysis of theological metaphors into experiential model and logical qualifiers. Ramsey, thus, gave explicit reasons for a necessary proliferation of models. Whether his account, with its liberally conceived criterion of empirical fit can avoid relativism and subjectivism, remains a question for discussion. I have a few comments on models in psychology and the social sciences, on Ramsey’s discussion of them, and on whether there are incompatible models, say, of human behavior. Keywords: analogy, cosmic disclosure, empirical fit, Hesse, metaphors, mystery, models, proliferation, Ramsey, science, theology 1. Introduction Should one creatively proliferate models, in science and theology, amongst other areas? Would such proliferations be useful, even necessary, or possibly harmful, confusing, relativizing? Now, in general, models come in many kinds and roles. There a nitrogen emission models, space models for buildings, models of religious practices, the Dutch polder model for decision-making, pedagogical models, the classical model of science, religions as models of and for the whole of reality. Many models are practical, many are theoretical. I shall only consider cognitive or epistemic models, which include theoretical and explanatory models and models generating understanding and insight. I take the accounts of two pioneering thinkers about models in the philosophy of science and theology, Mary Hesse and Ian Ramsey, to find out what their accounts might imply for the proliferation question. 2. Hesse and Analogies In science, cognitive models can have various functions or roles. One is in applications of a theory, e.g., the use of a model of a pendulum for applying the laws of classical mechanics. Another is the use of a model, e.g. that of the atom, for the (further) development of a theory. Mary Hesse (1963) analyzed models mainly in their last-mentioned role. She used the word ‘model’ in two senses, distinguished in terms of the analogies involved (Hesse 1963, 9ff.). Take the billiard-ball model of the dynamical theory of ideal gases. Gas molecules are taken to be analogous to billiard balls. The properties they share she called “the positive analogy”; the properties of the billiard balls which we know do not belong to gas molecules she called “the negative analogy” of the model; and the remaining properties of the model, not yet known whether they are positive or negative analogies she called the “neutral analogy”. Now, by “model1” she meant the way we are imagining the gas phenomena directly without the negative analogy. “Model2”, then, is the model considered as a second object including the negative analogy. The analogies concerned here she referred to as ‘material analogies’ (as opposed to ‘formal analogies’ between two interpretations of the same theory). Model1 plays its main role in the theoretical explanation of phenomena. In using a model2, the neutral analogy, she stressed, is especially interesting, as it makes new empirically testable predictions or hypothetical explanations possible. And it is in this way that models and theories may be further developed. The predictions, of course, may also turn out to be wrong, which may lead to a total rejection of a model. Hesse (1963, 112f.) stated that one can argue that it is more reasonable to choose an hypothetical explanation based on a model for the explicandum over one that is not; that in the case of two models for an explicandum, one should select the hypothesis based on a model which is more similar to the explicandum, also for predicting not yet observed characters. She investigated at great length whether the analogies and such choices can be justified logically - inductively, probabilistically or otherwise. She pointed out that there was no foolproof general logical justification of such choices. (Inductive justifications, e.g., fail altogether if our universe is a chance universe (Hesse 1963, 130)). Hesse said nothing explicit about a proliferation of models in science. She certainly did not advocate an unbridled proliferation of models. What she certainly would have advocated, I think, is that one investigated the neutral analogies of models, imaginatively developing further and better models. Yet, advocacy was not her goal, but rather the analysis of actual scientific procedures and developments. These involve, as she knew and plentifully referred to, a great variety of often competing models. We just saw that she considered such competing models and discussed choices between them. One can add here that there will always be decisive empirical tests of such hypothetical explanations and predictions, which can lead to definite choices between hypotheses. This may often take quite some time. Just think of the astronomical models of Ptolemy, Copernicus, and Tycho, which were competing with each other for a couple of centuries, or think of the corpuscular and the wave theory of light. With her appreciation of the theoretical importance of scientific models she was, as she pointed out, a follower of Campbell, and not of Duhem, for whom models might be psychological aids in suggesting theories, but not essential to an eventual ideally mathematized theory. It is appropriate in our context to mention that Hesse (1963, 95f.) considered one theological analogy, which can be regarded as basis for the theological model of God as sculptor. The analogy is between the relationships sculptor-statue and God-world. For Hesse, this is not an analogy in her (material) sense, since “the causal relation asserted between God and the world cannot, in principle, be identical with that between sculptor and statue.” 3. Ramsey and Disclosure Models Ian Ramsey was Professor of the Philosophy of the Christian Religion. His specific aim was to determine the special character of models in theological language. In all disciplines, he said (Ramsey 1964, 1ff.) – in his favorite, to me somewhat strange phrasing – models enable us to be “articulate about” their subject, about the electric field, about the universe, or about God. He followed Max Black’s (1962, 219ff.) distinction between scale and analogue models, which he preferred to call ‘picturing’ and ‘disclosure’ models. Scale models or picturing models, like model trains, he characterized as being purely, though reliably and genuinely, descriptive. This holds not only for science, but also for theology, namely for models like God as judge or the Father-Son relationship, which leave no place for mystery, wonder and worship. Analogue or disclosure models, Ramsey stated with Black, have become more and more significant in science. Typically, they involve a change of medium, like Maxwell’s fluid model of the electric field. Such disclosure models, for Ramsey (1964, 13f.), are born in a moment of insight when the universe discloses itself in the meeting points of the phenomena and the model. It gives us reliable scientific understanding when it “chimes in with and echoes” the phenomena, and when it is prolific in generating experimentally testable deductions. It then “fits the universe and enables us to be reliably articulate about it.” The situation it paralleled in theology by disclosure models, such as God as Strong Tower or the Logos model of Jesus. Only, they differ from scientific models in that they are not used to generate experimentally testable deductions. Rather (ibid., 16f.), they are judged by their success in incorporating and harmonizing the most diverse phenomena and events at hand, by what Ramsey calls their ‘empirical fit’. So, theological models are assessed in a much looser way. Though as a commonality, he stressed that both scientific and theological disclosure models “point to mystery, to the need for us to live as best as we can with theological and scientific uncertainties.” (ibid., 21) Disclosure models, especially theological ones, for Ramsey are metaphoric rather than analogical. “… metaphors, like models, are rooted in disclosures and born in insight … enabling us to be articulate” (ibid., 50f.), e.g., as when we speak of a religious leader as a light (ibid., 48). Theology, for Ramsey, is founded in what he calls “cosmic disclosures” (58ff.), “occasions of insight … when the universe declares itself in a particular way around some group of events” (ibid., 58f.) or situation, for instance, around the parable of the loving father, which then can yield a model for God. But, to speak more accurately of God, we should unite the phrases of various metaphors. This illustrates the principle “that to talk about what a cosmic disclosure discloses we shall need a never-ending succession of phrases which can combine in pairs to give a never-ending series of metaphorical inroads” (ibid., 60). Ramsey underlines this diversification with his analysis of theological metaphors further into two parts, an experiential model and a “logical qualifier”. In expressions like ‘Heavenly Father’ or ‘infinite love’, the adjectives qualify the respective nouns, indicating that these should not be taken literally, which also shows “the inadequacy of all models” (ibid., 60). Such qualifiers stimulate “the never-ending development of any model” (ibid.), in which way theology points to mystery. He compared this development with a convergent series, as if this development resulted in God as infinite love, infinitely infinite love, infinitely infinitely infinite love, etc., “in this way witnessing to the fact that the heart of theology is permanent mystery.” (ibid.) Yet, what is disclosed in this process? Ramsey himself raised the question of the “ontological commitment” of a model: “What is its reference?” (ibid., 13) However, his general answer is rather tautological: “… the model, whether in science or theology, provides us with its own understanding of … what the disclosure discloses.” (ibid., 20) Such an answer, it would seem, does not help much in distinguishing between a disclosure of something real and a fanciful disclosure. Ramsey, as we have seen, clearly and explicitly called for a proliferation of theological models, both in the sense of different models and in the sense of further developments of existing models. He wrote: “… theology demands and thrives on a diversity of models …” (ibid., 60) and he stated specifically that qualifiers provide for “the endless construction” of metaphorical models (ibid., 61). Both ways witness to theology’s “grounding in permanent mystery” (ibid.). We get here “the problem of verification and the problem of reasonable preferences”, which he also dealt with in his later book (Ramsey 1973, 62), namely, with the question: “… how do we test between those who talk about these cosmic disclosures in different and incompatible ways, who articulate about these cosmic disclosures in terms of vastly different models?” (ibid.) His answer was: “we shall reasonably prefer that discourse which (a) formally is the most simple, coherent, comprehensive and consistent; (b) materially establishes the best empirical fit – and by empirical fit I mean the kind of fit presupposed by hymns” (ibid.) which characterize the models concerned. It is not clear whether his account and this answer, with its broadly conceived criterion of “empirical fit”, can do the job and avoid subjectivism and relativism. What is remarkable is that he did not touch upon the diversity of (non-Christian) religions, their own ways of probing cosmic mystery, and the sort of relativism involved in this diversity. 4. Social-Scientific Coda Ramsey (1964, 22ff.) also discussed models in psychology and the social sciences. He stressed that, aside from scientific models with observable predictions, disclosure models which do justice to humans as persons are indispensable. They, as ‘a loves b’, can only be verified in terms of what he called ‘empirical fit’ (ibid., 38). Another example, he gave, is a psychiatrist getting to understand a patients’s condition; a disclosure may occur and he might say ‘mild depressive’ (ibid., 39). This model of ‘mild depressive’ will have to be tested against the patient’s whole life. Clearly, human behavior displays so many aspects that the use of various explanatory models is inevitable. Helen Longino (2013) distinguished and discussed major approaches to human aggression and sexuality, ranging from behavior genetics to social-environmental approaches. She considers them to be incommensurable, even incompatible. I would side with Ramsey who, as we just saw, stated that if there should be incompatible models we would have to decide which to prefer. However, I do not consider Longino’s models as really incompatible (Kirschenmann 2015). 5. Conclusions Should models be multiplied or even proliferated, in science and theology? For Mary Hesse, we saw, an essential characteristic of models, especially scientific models, was the use of analogies. Her idea that neutral analogies should be further examined implies that there will be successive improved models. She was aware of the many models that have been considered and used in science and also of the issue of choosing between competing models, pointing out that there was no fool-proof way of justifying such choices. For Ian Ramsey, picturing models certainly have their many good uses. Yet in some contrast with Hesse, he stressed a necessary proliferation of disclosure models in science and especially in theology, namely models characterized by the use of metaphors. In theology, these models are not explanatory, but attempts at articulating, and reaching some understanding of, the divine mystery. With his view that this task is many-faceted and never-ending, Ramsey gave an explicit reason for the proliferation of models. Yet also in science, even though models explain how and why the processes in the world occur, there always remain mysteries, such as the mystery that the world is. That is why for Ramsey – and I would agree – models in science at the same time are articulations of cosmic disclosures. References Black, Max (1962) Models and Metaphors: Studies in Language and Philosophy, Ithaca, New York: Cornell University Press. Freudenthal, Hans (1961) The Concept and Role of the Model in Mathematics and Natural and Social Sciences, Dordrecht-Holland: D. Reidel Publishing Company. Hesse, Mary B. (1963) Models and Analogies in Science, London and New York: Sheed and Ward. Kirschenmann, Peter P. (2015) “Scientific Pluralism Between Realism and Social Constructivism? Some Critical Comments”, in Christian Kanzian et al. (eds.) Realism – Relativism – Constructivism, Contributions of the Austrian Ludwig Wittgenstein Society, Vol. XXIII, 157-159. Longino, Helen E. (2013) Studying Human Behavior: How Scientists Investigate Aggression & Sexuality, Chicago and London: The University of Chicago Press. Ramsey, Ian T. (1964) Models and Mystery, London: Oxford University Press. Ramsey, Ian T. (1973) Models for Divine Activity, London: SCM Press Ltd. CV:  Peter P. Kirschenmann is professor-emeritus in the philosophy of the exact natural sciences and philosophical ethics, Vrije Universiteit, Amsterdam, also a long-time member of the study group “Natural Science and Theology”. Published articles can be found in his Science, Nature and Ethics: Critical Philosophical Studies, Delft: Eburon 2001, and, more recent ones, on academia.edu and researchgate.net. 1