Ludwik Fleck and Thought Collectives
The various forms of positivism that developed from the mid-nineteenth to the mid-twentieth century were all concerned with method, paying less attention to the people who were pursuing that method or the institutional structures within which they operated. Popper paid some heed to the character of the individual scientist, insofar as he stressed the importance of a critical investigative attitude. But Popper’s epistemology (like his political theory) was individualistic; he vested the advance of science in the actions of the bold individual who doubted an existing claim and found a means to refute it. Popper paid less attention to the institutions of science, and was actively hostile to suggestions of collectivism, redolent as they were of the Marxist philosophy and Communist politics that he loathed.33
The recognition of science as a collective activity thus laid the grounds for a radical challenge to received views of science that would flourish in the second half of the twentieth century. Whether one had read Comte or Ayer or Popper, one could have come away with the impression that scientists, like Descartes in his room staring at melting wax, lived, worked, and thought alone. Yet anyone who studied science in action—as Comte instructed us to do—or who participated in scientific research knew that wasn’t so. Yet somehow this had escaped sustained scholarly attention.
Ludwik Fleck (1896–1961) changed that. A microbiologist who made the social interactions of scientific life a centerpiece of analysis, in hindsight he is credited with developing the first modern sociological account of scientific method. In his 1935 work, The Genesis and Development of a Scientific Fact: An Introduction to the Theory of Thought Style and Thought Collective, Fleck shifted attention from the individual scientist to the activities of communities of scientists, and proposed that scientific facts are the collective accomplishment of communities. In doing so, he pioneered the analysis of the social interactions that yield scientific facts.
Fleck was aware of the logical positivists’ work; he sent his work to the Viennese positivist Moritz Schlick seeking help to get it published.34 He was also in contact with historians and philosophers of medicine and mathematics in Poland at that time. But scholars have mostly concluded that his work was primarily influenced by his experience as a researcher and his attention to developments in science, particularly the rise of quantum mechanics in physics, which (he believed) had led to the emergence of new styles of thinking.
Fleck’s key point was that scientists worked in communities in which styles of thought became shared resources for future work, including the interpretation of observations. He labeled these communities “thought collectives.” Groups of scientists within any particular discipline—biology, physics, geology—constituted thought collectives whose common ways of thinking made it possible for them to work together, share information, and interpret that information in meaningful ways. Without a thought collective, science could not exist. He wrote:
A truly isolated investigator is impossible … Thinking is a collective activity.… Its product is a certain picture, which is visible only to anybody who takes part in this social activity, or a thought which is also clear to the members of the collective only. What we do think and how we do see depends on the thought-collective to which we belong.35
The term “thought collective” may invoke the specter of thought police, and Fleck recognized that collectives could be conservative or even reactionary—as he believed religious thought collectives were. But a thought collective could also be democratic and progressive, and this was the key to understanding science. Science (unlike most European religion) has a democratic character: all researchers can participate in an equitable way, and through their interactions with each other, refine and change the views of the whole.
Fleck had a radical view of how far such change could go, stressing that over time changes could be so great that the meanings of terms changed, that problems that were previously seen as central could now be dismissed as irrelevant or even illusory, and new issues would emerge that previously went unrecognized. While the increments of change were small—the pathways of change more evolutionary than revolutionary—eventually the thought style may have changed so much that the old view is essentially unrecognizable, even indecipherable.
Thoughts pass from one individual to another, each time a little transformed, for each individual can attach to them somewhat different associations. Strictly speaking, the receiver never understands the thought exactly in the way that the transmitter intended it to be understood. After a series of such encounters, practically nothing is left of the original content.36
Scientific ideas, like evolution itself, may change dramatically over time, but they do so by the accumulation of small transformations and differing interpretations.
“Whose thought is it that continues to circulate?” Fleck asks. His answer: “It is one that obviously belongs not to any single individual but to the collective.”37 As Helen Longino would later put it in a slightly different context, “Of course, Galileo and Newton and Darwin and Einstein were individuals of extraordinary intellect, but what made their brilliant ideas knowledge were the processes of critical reception.” Fleck would say: of reception and transformation.38 Newtonian mechanics is not equivalent to the contents of the Principia, nor is evolutionary biology coincident with the contents of the Origin of Species. The ultimate outcome is the result of Newton and Darwin’s work and the diverse ways in which over time it has been interpreted, adjusted, and altered.
Scientific progress in this view is inextricably connected with the institutions of science such as conferences and workshops, books and peer-reviewed journals, and scientific societies through which scientists share data, assess evidence, grapple with criticisms, and adjust their views. Scientific research is organized, it is cooperative and interactive, it creates shared worldviews, and observations are interpreted in accordance with these worldviews. Progress, Fleck holds, consists of the revision and adjustment of worldviews as the community deems appropriate, and over time these adjustments may be so great as to constitute a new worldview, a new style of thought, even a new reality.39 What the thought collective previously recognized as physical reality may no longer be viewed as reality. Fleck is unambiguously anti-realist on this point: what members of a collective call truth is merely what the thought collective has settled upon at that point. He is also unambiguously anti-individualist and anti-methodological: the agency of scientific progress is located not in the individual but in the group, and the core of science lies not in a particular method but in the diverse interactions of that group.
Under-determination: Pierre Duhem
Fleck’s work received some attention when first published, but became much more famous in later years when it came to be viewed as anticipating and influencing the work of Thomas Kuhn. Something similar may be said about Pierre Duhem (1861–1916), whose work was recognized by the Vienna Circle but is now seen as influential primarily because of its uptake by the American philosopher W.V.O. Quine (1908–2000).
To scientists, Duhem is known as a founder of chemical thermodynamics, but he was also a sedulous historian and acute philosopher of science.40 To philosophers and historians of science today, he is best known for his 1906 book, The Aim and Structure of Physical Theory, with its refutation of the notion of a critical experiment and its articulation of what has come to be known as the principle of under-determination.41
Duhem’s central argument was simple: The Baconian idea of a crucial experiment is mistaken, because if an experiment fails there are many reasons why that might be, so we don’t necessarily know what has gone wrong. Conversely, if an experimental test of a theory succeeds, other consequences of the theory may yet be shown to be incorrect. The support for a theory must in principle include all the potential tests of it, and its refutation must be considered in light of all the possible elements that were necessary to perform the experiment in the first place. As the physicist Louis de Broglie put it in 1953
in the preface to the English edition:
According to Duhem, there are no genuine crucial experiments because it is the ensemble of a theory forming an individual whole which has to be compared to experiment. The experimental confirmation of one of its consequences, even when selected among the most characteristic ones, cannot bring a crucial proof to theory, for … nothing permits us to assert that other consequences of the theory will not yet be contradicted by experiment, or that another theory yet to be discovered will not be able to interpret as well as the preceding one the observed facts.42
Put simply: any test of a hypothesis is simultaneously a test of the specific hypothesis under consideration and of the experimental setup, auxiliary hypotheses, and background assumptions. A failed experiment does not necessarily reveal where the failure lies, and a successful experiment does not preclude that a different experimental arrangement or other auxiliary hypotheses would have revealed some difficulty. Duhem wrote: “Any experimental test [in physics] puts into play the most diverse parts of physics and appeals to innumerable hypotheses; it never tests a given hypothesis by isolating it from the others.”43
Nor does experimental evidence exhaust the range of possible theoretical options open to us: Duhem was explicit that hypotheses are not simply inductions from observation. It is impossible, he asserted without equivocation, to “construct a theory by a purely inductive method.”44 Both theory and experiment have a role in science, and it is mistaken to view experiments as more crucial than theory, mistaken to view them as the source of theory, and above all, mistaken to view them as the final arbiter of theory.
Duhem was not rejecting experimentation. On the contrary, he argued that “the sole purpose of physical theory is to provide a representation and classification of experimental laws.”45 Experiment is essential both to discovering those laws in the first place and to testing the general physical theories that we develop to account for them. The “only test permitting us to judge a physical theory and pronounce it good or bad is the comparison between the consequences of this theory and the experimental laws it has to represent and classify.” This view is essentially probabilistic: an experiment can neither verify nor refute a theory; rather it simply tells us whether a theory is “confirmed or weakened by the facts.”46
De Broglie suggested that a key to Duhem’s thought was his interpretation of Léon Foucault’s famous experiment in which he demonstrated that the speed of light in water is less than its speed in a vacuum, taken by many as a crucial experiment validating the wave (as opposed to particle) theory of light. Duhem disagreed. Even if Foucault’s experiment contradicted Newton’s corpuscular theory, other forms of corpuscular theory might yet be consistent with the result.47
Yet Duhem did not adopt the radical holism with which his name later became associated. (Holism is the idea that theories stand or fall in their entirety and that a challenge to any one component is potentially a challenge to the entire intellectual fabric.) In places, it may appear that he is on the verge of radical holism, as when he writes of the “radical impossibility [of separating] physical theories from the experimental procedures appropriate for testing these theories,” or that an “experiment in physics can never condemn an isolated hypothesis but only a whole theoretical group.”48 But elsewhere he makes clear that he believes some elements of our belief structure are so well established that we are unlikely to doubt them, and rightly so. Some elements of our work are well confirmed through other sources, or strongly linked to principles that we have little doubt are correct. Basic instruments such as thermometers and manometers, for example, are unlikely to be distrusted, as are the concepts that accompany them, such as temperature and pressure. Indeed, he insists that in testing the accuracy of a proposition, a physicist must make use of a whole group of theories that are accepted by him as “beyond dispute.” Otherwise he would be paralyzed; it would be impossible for him to proceed. (One may suppose that basic principles of thermodynamics, such as conservation of mass and of energy, are in his mind.) Likewise if an experimental test fails, it does not tell us where the failure lies. It tells us only that somewhere in the system “there is at least one error.”49
In sum, the physicist can never subject an isolated hypothesis to experimental test, but only a whole group of hypotheses; when the experiment is in disagreement with his predictions, what he learns is that at least one of the hypotheses constituting this group is unacceptable and ought to be modified; but the experiment does not designate which one should be changed.50
Duhem did not conclude that for this reason we should be radically skeptical. Rather he argued that we should adopt an attitude of reasonable humility toward intellectual commitments. Following Claude Bernard, he reminds us to be anti-dogmatic, to maintain an openness to the prospect that our theories may need revision, and to preserve an essential “freedom of mind.”51 Hypothesis, theories, and ideas in general are essential for stimulating our work, but we should not have “excessive faith” in them.52 We should not be too pleased with our own accomplishments. As Americans at that time might have put it, we should not become “auto-intoxicated.”53
In the face of an apparent refutation, how does a scientist decide which element(s) of the relevant nexus of theory, instruments, experimental setup, and auxiliary hypotheses should be revised? On this point, Duhem is not entirely satisfactory, invoking Pascal that there are “reasons which reason does not know.” In the end, he concludes that these decisions ultimately are matters of judgment and “good sense.”54 Duhem uses history to underscore this point:
We must really guard ourselves against believing forever warranted those hypotheses which have become universally adopted conventions, and whose certainty seems to break through experimental contradictions by throwing the latter back on more doubtful assumptions. The history of physics shows us that very often the human mind has been led to overthrow such principles completely, though they have been regarded by common consent for centuries as inviolable axioms, and to rebuild its physical theories on new hypotheses.55
Yet at the same time, he makes equally clear his conviction that history gives us grounds for confidence in the processes of scientific investigation, so long as we do not become dogmatic. He concludes with the following passage:
The history of science alone can keep the [scientist] from the mad ambitions of dogmatism as well as the despair of … skepticism. By retracing for him the long series of errors and hesitations preceding the discovery of each principle, it puts him on guard against false evidence; by recalling to him the vicissitudes of the cosmological schools and by exhuming doctrines once triumphant from the oblivion in which they lie, it reminds him that the most attractive systems are only provisional representations, and not definitive explanations. And, on the other hand, by unrolling before him the continuous tradition through which the science of each epoch is nourished by the systems of past centuries … it creates and fortifies in him that conviction that physical theory is not merely an artificial system, suitable today and useless tomorrow, but that it is an increasingly more natural classification and an increasingly clearer reflection of realities which experimental method cannot contemplate directly.56
W.V.O. Quine and the Duhem-Quine Thesis
Duhem’s views became known to American audiences primarily through the Harvard philosopher Willard Van Orman Quine, and in the process came to be viewed as more radical than they arguably were. Quine took the problem of refutation and reformulated it under the rubric of what has come to be known as “under-determination.” If theories are tested not in isolation but in whole theoretical groups, then how do we know which piece of the group is in need of revision when something goes awry? Duhem’s answer was: We rely on judgment. Quine’s answer is: We don’t know. Knowledge, he insists, is a web of belief. When we encounter a refutation, there is a universe of potential adjustments we can make, a universe of threads that can be tightened or loosened to sustain the fabric or reweave it. In Quine’s words: “our sta
tements about the external world face the tribunal of sense experience not individually but only as a corporate body.”57
Duhem would have agreed with that, but he also believed that evidence could lead us to reexamine and adjust parts of that corporate body appropriately. This is one of his two key purposes of experimentation—to strengthen or weaken the support for particular elements in physical theory. If saving the phenomena required us to abandon something that is very strongly held—such as conservation of energy—we would be unlikely to do it. We would conclude that the experiment revealed a problem somewhere else or that there was a problem with our instrumentation. For Duhem, the various parts of the whole theoretical group are not created equal and not equally up for grabs. But Quine thinks that they are, concluding, famously: “any statement can be held true, come what may, if we make drastic enough adjustments elsewhere in the system.”58
Quine’s radical holism came to be known as the Duhem-Quine thesis and is taken by many scholars to weaken the grip of evidence on theory, because if theories are under-determined by experiment—and we have a world of choices in how to respond to experimental failure—then what is the basis for our belief?59 It appears that some additional component is necessary to explain how scientists come to the conclusions that they do. This became the foundation of a great deal of what followed: some scholars have argued that the concept of under-determination underpins the entire set of challenges to empiricist philosophy that developed in the second half of the twentieth century, including the work of Thomas Kuhn and emergence of the field of science studies.60
Why Trust Science? Page 4