| HPS 2103 | History and Philosophy of Science Core Seminar | Spring 2022 |
The modern discipline of history and philosophy of science has, at its core, the idea of a fertile interaction between the history of science and the philosophy of science. There are firm guides that can be given as to what is good history of science and what is good philosophy of science. They are linked here. There do not seem to be corresponding guides for what is good integrated history and philosophy of science. That is, there are none beyond the requirement that all such work must adhere to the standards of good history of science and good philosophy of science. As long as those standards are respected, work in integrated HPS should be academically responsible.
That the work is academically responsible does not entail that it is productive scholarship, however. The purpose of this document is to attempt a catalog of productive ways that the H and P have been integrated in existing work. The goal is not an exhaustive catalog. Rather the goal is to provide examples that suggest productive avenues of research
All efforts to integrate history of science and philosophy of science should be subject to critical scrutiny. Examples of the sorts of assessments are included below.
The catalog is divided into two main parts: when the H informs the P; and when the P informs the H. They are not exclusive. In each case there will be a reverse reaction as well. The examples have been divided according to which of the interactions is more evident or stronger.
First, however, there is a delicate matter of the definition:
While there is much talk of "integrated history and philosophy of science," the discussion has not provided a precise definition of the term that delimits it boundaries. As a first step in this direction and stimulus to discussion, I offer the following formula:
A work in integrated HPS contains both history of science and philosophy of science with both contributing to the main results of the work. These results may be in the history of science, in the philosophy of science or in both.
This formula includes two assumptions:
1. There is no measure of how integrated the two components must be in order to count. Rather the presumption is that any integration at all is sufficient to earn the title of integrated HPS. Of course if the integration is trivial, then the title is only earned trivially.
2. The modifier "integrated" is not intended to be restrictive. Integrated HPS just is HPS. The addition of "integrated" is merely a reminder that there should be some interaction between the historical and philosophical components.
The modern tradition in history and philosophy of science centered around the idea that paying proper attention to the history of science is essential if we are to have a responsible philosophy of science. The idea is expressed in the introduction paragraph of the first chapter of Kuhn's Structure.
"History, if viewed as a repository for more than anecdote or chronology, could produce a decisive transformation in the image of science by which we are now possessed. That image has previously been drawn, even by scientists themselves, mainly from the study of finished scientific achievements as these are recorded in the classics and, more recently, in the textbooks from which each new scientific generation learns to practice its trade. Inevitably, however, the aim of such books is persuasive and pedagogic; a concept of science drawn from them is no more likely to fit the enterprise that produced them than an image of a national culture drawn from a tourist brochure or a language text. This essay attempts to show that we have been misled by them in fundamental ways. Its aim is a sketch of the quite different concept of science that can emerge from the historical record of the research activity itself."
The suggestion was that we could consult the history of science and read off from it a better account of the nature of science. This approach is implemented in the grand narratives of the nature of scientific change of the latter part of the twentieth century by such figures as Kuhn, Lakatos, Feyerabend and Laudan.
The approach can be more narrowly targeted. If we are to provide ideas and principles responsive to the actual practice of science, that same consultation of the history can serve us.
(Norton) My monograph, The Material Theory of Induction, employs this approach. It draws heavily on historical examples to support the analysis of the various manifestation of inductive inference. The most prominent case concerns inference to the best explanation. There I was uncomfortable with the dominance of misleading non-scientific examples as underpinnings of present accounts of inference to the best explanation. As a corrective, I assembled a roster of prominent cases from that literature and investigated each to arrive at the simple model offered in Chapters 8 and 9. The model is, as a matter of logic, independent of the history. it was discovered by examining the history and, once identified, can be independently assessed.
(Norton) This approach was also used heavily in my second
monograph, The
Large-Scale Structure of Inductive Inference.
there were many problems concerning the large-scale structure
of relations of inductive support that I could not solve merely by
reflecting on the problems ahistorically. Is the large scale structure
unique, for a given evidence base? Are the circularities of inductive
support that arise in such structures harmful? Each of these problems was
solved by consulting examples in the history of science. The resulting
solutions are displayed in several chapters. These solutions are once
again, as a matter of logic, independent of the history. They were
discovered by examining the history and, once identified, can be
independently assessed.
Problems
1. Selection Effects
To serve as evidence for theses in philosophy of science, the history must be representative of the science overall that is the subject of the philosophy. The sample of historical case studies chosen is almost inevitably unrepresentative.
If one believes the experiments are the decisive factor in science, then one's history will collect great experiments. For example, see Rom Harré, Great Scientific Experiments: Twenty Experiments that Changed our View of the World. The resulting inductively recovered philosophy of science will favor experiments.
Corresponding selection effects arise if one privileges theory of experiment, or if one believes that science advances through the achievements of a few great men, or the communal effects of scientists jointly, or if one believes that accepted sciences contain considerable socially constructed content.
2. Formulations of the History
A related effect is that all histories, barring the most banal, must be described using notions and language intelligible to the modern reader. That requires a reformulation of the history in ways that may prejudice the philosophy recovered. This concern is prominent in feminist critiques of philosophy of science.
3. Instabilities in Science
Science evolves and, as it evolves, its methods change. The science of the 20th century was often produced by large research teams unknown in earlier centuries. The science of the 21st century is now employing machine learning techniques unknown in past centuries. In the face of such changes, the import of history is diluted, for it reports what happened prior to the latest changes.
Responses
1. Many Case Studies
An obvious but weak response to selection effects is to seek as wide a
range of examples as possible. Since the past of science contains
enormously many possibilities, the sampling can only be imperfect.
2. Refutation of a Generalization
History of science, informative to philosophy of science, need not employ a simple inductive model that says "These past cases have been X, so all future ones will too." The history can be used to refute a strong generalization in present philosophy of science. All that is needed is one counter-example in the history.
(Norton) This is the logic of my paper,"The Determination of Theory by Evidence: The Case for Quantum Discontinuity 1900-1915," Synthese, 97(1993), 1-31. I display an important case in which clever analysis by Poincaré, Ehrenfest and others were able to demonstrate that, on the evidence, quantum discontinuity was unavoidable. It is a counterexample to the thesis that evidence always leaves theories underdetermined.
3. Instantiation
A simple inductive generalization from a few cases will always be risky. However in practical HPS work, there are other possibilities. We may arrive at some thesis on the basis of present, philosophical analysis. The result, we may fear, may merely be a possibility, uninstantiated in science. The display of examples of instances in the history of science can reassure us that the situation is otherwise.
(Norton) My monograph, The Material Theory of Induction, employs this strategy. The principal claim is that inductive inferences are warranted by background facts with local applicability. This thesis is sustained by an analysis independent of the history. The monograph then supplies many instances of successful inductive inferences in the history of science in which the material inductive warrant is identifiable.
It is easy to image that present science has extracted all that is useful and relevant in the past work of scientists. However time and again, we find that useful and even important work in the past is lost. It can be recovered by historical investigations.
One of the most important instances of this pertains to the evidential basis of a scientific theory. Often enough--but definitely not always--the evidential case amassed for a theory at the time of its creation is diluted in subsequent textbook renderings of the theory. Once the theory is accepted and has become a canonical part of the science, the textbook writers seek to speed on to the newer material where advanced work is to be done. What can result is a reduced version of the case of the for the science.
(Norton) At the time of its creation, Einstein explored many possibilities other than special relativity to account for the phenomena covered by the theory. A notable part was the exploration of emission theories of light that are compatible with the celebrated null result of the Michelson-Morley experiment. These explorations are still useful today when the find the simplified textbook accounts of the evidence for the theory facile. See "Einstein's Investigations of Galilean Covariant Electrodynamics prior to 1905," Archive for History of Exact Sciences, 59 (2004), pp. 45-105.
(Norton) We are now commonly told that a quantum discontinuity in energy spectrum saves the phenomena of the Planck distribution of energy in the black body spectrum. Many correctly find that an incomplete basis for accepting the discontinuity. Might there not be other unconceived alternatives? My paper,"The Determination of Theory by Evidence: The Case for Quantum Discontinuity 1900-1915,: Synthese, 97(1993), 1-31, shows how Poincaré, Ehrenfest and others were able to demonstrate in 1911-12 that, on the evidence, quantum discontinuity was unavoidable. Their beautiful analysis disappeared from the textbooks after about 1930.
(Norton) The "hole argument" has had a powerful presence in philosophy of space and time since the mid 1980's. It could in principle have been discovered independently of its place in the history of science. That is not what happened. Rather, the three John's--Stachel, Earman and Norton--were intrigued by the place of the argument in Einstein's work of 1913 on general relativity and recognized that it raised issues of present interest in philosophy of space and time.
The present state of science is rife with anomalies whose nature can be hard to understand if one knows none of the history. Their presence may even not be evident. Historical research can reveal these anomalies and, once their tenuous historical origins are understood, appropriate responses to them can be formulated.
(Norton) Readers of Einstein and of the writings of those who work in his tradition will be assured that the general theory of relativity generalizes the relativity of motion from inertial motion to accelerated motion. If the theory is examined independently, it soon becomes opaque as to how the theory implements the extension. What we learn from the history is that seeking this extension was of enormous heuristic value to Einstein while he worked on general relativity. However it was a heuristic that was never met by the ensuing theory. Einstein, nonetheless, continued to report otherwise. For details see my "General Covariance and the Foundations of General Relativity: Eight Decades of Dispute," Reports on Progress in Physics, 56(1993), pp.791-858.
(Norton) A related problem concerns "Mach's Principle." According to it, the inertia of bodies is due to an interaction with all the other masses of the universe. Considerable effort is expended in present physics to devise a theory that conforms with it. What is overlooked is that the principle itself has scant basis. Mach himself likely never espoused it and his contemporaries--correctly in my view--regarded the principle as an attempt at a priori physics. Einstein was its chief proponent, but he abandoned it in his later years. See my "Mach's Principle before Einstein." in J. Barbour and H. Pfister, eds., Mach's Principle: From Newton's Bucket to Quantum Gravity: Einstein Studies, Vol. 6. Boston: Birkhäuser, 1995, pp.9-57.
(Norton) Present physics takes seriously the idea that information and computational notions are essentially connected with thermodynamic entropy and that connection is exploited routinely in speculative physics. Closer examination of the supposed connection, however, finds it to be dubious and weakly supported by sketchy and implausible argumentation. An historical examination of the evolution of the idea, starting with Maxwell's 19th century demon, shows that it evolved from dubious origins. Its history is an accumulation of incomplete and implausible speculation that has never succeeded in securing its results. I have investigated this history extensively and the work is linked in this compendium.
Problem
The judgments in the examples above have required a present day assessment of the cogency of the argumentation of figures in the past. Someone who does not find the present state of the science anomalous might well write a more celebratory history.
(Norton) Such is definitely the case for much of the history written of the dubious relationship of information and entropy. See for example, Harvey Leff, Andrew F. Rex, eds., Maxwell's Demon 2: Entropy, Classical and Quantum Information, Computing.
One the most important functions of philosophy of science is to provide a compass for the writing of the history of science. To write the history of science, a historian needs to have some conception of what constitutes science. The conception requires answers to many questions:
• Where does science end and some other human activity take over?
• What outside of the science is still an essential part of the historical narrative? Here "outside" includes activities temporally preceding the science and those happening at the same time.
Philosophy of science provides answers to these questions.
(Norton) Much of my writing in the history of Einstein's
work on relativity is interested in the rationality or otherwise of
Einstein's investigations. Here I need some standard of rationality. In
principle those standards might differ in different time periods. Even if
that is the case, it is a project for philosophical analysis to determine
what those standards are. Much of my work on inductive inference is
designed to be useful in these sorts of historical analyses. (There
is surprising stability in these standards over time. It may not seem
that way since the subject matter can change. There are exceptions to
this stability. Most notable is the introduction over the past few
centuries of probabilistic techniques of statistical analysis.)
Problem
The literature in philosophy of science has presented a wide range of conceptions of the nature of science. They vary from extreme positivist pictures, in which science is merely compact descriptions of past regularities, to skeptical accounts that attribute much of science to conventional agreements amongst scientists.
There are also multiple accounts of the many individual notions that might appear in historical analysis. If we are interested in the rationality of science, do we take Bayesian analysis as the standard; or abductive analysis; or (Norton's) material approach?
If philosophers of science cannot agree on the right conception, what are historians to do? How can they responsibly write a history of some piece of science if the philosophical literature cannot agree on what constitutes the science of that piece?
A part of the last question that deserves separate discussion is the question of what is important in science.
At any moment, there is a lot of scientific activity. Some parts of it are of lesser importance; others are of great importance. Deciding which is which is a question for philosophy of science. The resulting judgments of relative importance then may guide which problems attract the attention of historians.
Work in relativity theory and in quantum theory in the first quarter of the twentieth century has attracted outsized attention because of the importance it is judged to have for the physics that follows. The experiments of Lummer and Pringsheim around 1900 in Berlin on the distribution of energy over the various frequencies of heat radiation is, in itself, no different from a thousand other painstaking efforts in physics. However because of their importance in guiding Planck's work that led to quantum theory, their experiments have attracted considerable historical interest.
(Norton) Sadi Carnot's 1824 Reflections on the Motive Power of Fire is an extraordinary work. It introduces a new mode of analysis that forms the basis of thermodynamics. His innovations may appear to fall from nowhere until we compare them with the work of his father, Lazare, on the efficiency of ordinary machines. I join others in holding that Sadi was working by analogy with his father's work. That analogy plausibly led Sadi to two extraordinary features of his work. First, he had the audacity to propose a general theory of heat engines that would cover all possible heat engines. Second, he based his account on the notion of a thermodynamically reversible process. Such a process is inconsistent in its basic conception: it is a process in which a system changes, but is, at the same time, always in equilibrium, so it does not change. I relied on my philosophical analyses of thermodynamics to pick out these two features as especially important. See "How Analogy Helped Create the New Science of Thermodynamics."
While philosophers of science may well balk at the idea of discerning which are the proper truths of modern science, such judgments do play an important role in the writing of the history of science. Historians are generally more eager to write the history of a science that has a lineage that leads up to our best modern science. The history of astronomy has proven far more attractive to historians than the history of astrology, for example. This is not to say that the resulting history is necessarily Whiggish. One can choose to write on Newton, as opposed to some other of his contemporaries, and still write in a way that is historically responsible.
Problem
What is Correct Science?
Here differences of opinion can lead to divergent histories. Which is the right history?
A prominent case concerns a dissident strand in the physics community of the early 20th century. These dissidents--most notably Einstein--were uncomfortable with the received "Copenhagen" interpretation of quantum theory. It was initially traditional in historical writing to dismiss these dissidents. They were simply supposed to be too inflexible to grasp the profound novelty of the new theory. Starting in the 1960s and 1970s, the realization grew that the dissidents had a point. A new and energetic literature emerged on the foundational problems of quantum theory. With it came a new historical literature. It puzzled over how the mainstream of physics had been bluffed, largely by Bohr, into overlooking profound imperfections in the foundations of quantum theory. See James T. Cushing, Quantum Mechanics: Historical Contingency and the Copenhagen Hegemony.
February 6, 12, 2022