Trust in Numbers

The Pursuit of Objectivity in Science and Public Life

Theodore Porter

• Editorial: Princeton University Press
• Año: 1996
• ISBN:
  9781400821617
• Categoría: Matematica
• Categoría: Storia e filosofia della matematica

This investigation of the overwhelming appeal of quantification in the modern world discusses the development of cultural meanings of objectivity over two centuries. How are we to account for the current prestige and power of quantitative methods? The usual answer is that quantification is seen as desirable in social and economic investigation as a result of its successes in the study of nature. Theodore Porter is not content with this. Why should the kind of success achieved in the study of stars, molecules, or cells be an attractive model for research on human societies? he asks. And, indeed, how should we understand the pervasiveness of quantification in the sciences of nature? In his view, we should look in the reverse direction: comprehending the attractions of quantification in business, government, and social research will teach us something new about its role in psychology, physics, and medicine.

Drawing on a wide range of examples from the laboratory and from the worlds of accounting, insurance, cost-benefit analysis, and civil engineering, Porter shows that it is "exactly wrong" to interpret the drive for quantitative rigor as inherent somehow in the activity of science except where political and social pressures force compromise. Instead, quantification grows from attempts to develop a strategy of impersonality in response to pressures from outside. Objectivity derives its impetus from cultural contexts, quantification becoming most important where elites are weak, where private negotiation is suspect, and where trust is in short supply.

______________________ Part I ______________________

POWER IN NUMBERS

Now it must here be understood that ink is the great missive weapon, in all battles of the learned, which, conveyed through a sort of engine, called a quill, infinite Numbers of these are darted at the enemy, by the valiant on each side, with equal skill and violence, as if it were an engagement of porcupines.
(Jonathan Swift, “The Battle … between the Ancient and Modern Books,” 1710)

______________________ CHAPTER ONE ______________________

A World of Artifice

I thought it was the task of the natural sciences to discover the facts of nature, not to create them.
(Erwin Chargaff, 1963)

MAKING KNOWLEDGE IMPERSONAL

The credibility of numbers, or indeed of knowledge in any form, is a social and moral problem. This has not yet been adequately appreciated. Since the 1970s, debates about objectivity between philosophical and sociological camps have been polarized mainly over the question of realism. The claim that science is socially constructed has too often been read as an attack on its validity or truth. I consider this a mistake, as well as a diversion from more important issues. Perhaps there is something to be accomplished by arguing whether science can get at the real nature of things. But the answer can scarcely be peculiar to science, unless we are to suppose that systematic research is incapable in principle of identifying real entities, even though we can do so as if by instinct in our everyday lives. I find this and the opposite doctrine equally implausible. This book does not presuppose and will not defend any position on the much-vexed philosophical issue of realism.

If a declaration of faith is called for at the outset, I would say that interested human actors make science, but they cannot make it however they choose. They are constrained, though not absolutely, by what can be seen in nature or can be made to happen in the laboratory. Experimental interventions, guided but not dominated by theoretical claims, have often been remarkably effective. There remain subtle questions about what should count as truth. I am content to invoke Ian Hacking’s modest but elegant formulation, “It is no metaphysics that makes the word ‘true’ so handy, but wit, whose soul is brevity.”¹ Let us suppose for the sake of argument that scientific investigation is able to yield true knowledge about objects and processes in the world. It must nonetheless do so through social processes. There is no other way.

To accept this point is only to fix the terms for discussing a problem, not to solve one. Through what specific social processes is scientific knowledge made? How wide a circle of inquirers and judges is involved in the process of deciding what is true? The standard view has long held that in mature sciences, the truth is worked out or negotiated by a community of disciplinary specialists whose institutions are strong enough to screen out social ideologies and political demands. I will try to show toward the end of the book that the effectiveness of this segregation has been exaggerated—that the sciences have been compelled to redefine their proper domain in order to monopolize it, and that much of what passes for scientific method is a contrivance of weak communities, partly in response to the vulnerability of science to pressures from outside. But for the moment it is enough to think about processes of constructing knowledge that are internal to disciplines.

According to the individualist form of rhetoric about science, still much used for certain purposes, discoveries are made in laboratories. They are the product of inspired patience, of skilled hands and an inquiring but unbiased mind. Moreover, they speak for themselves, or at least they speak too powerfully and too insistently for prejudiced humans to silence them. It would be wrong to suppose that such beliefs are not sincerely held, yet almost nobody thinks they can provide a basis for action in public contexts. Any scientist who announces a so-called discovery at a press conference without first permitting expert reviewers to examine his or her claims is automatically castigated as a publicity seeker. The norms of scientific communication presuppose that nature does not speak unambiguously, and that knowledge isn’t knowledge unless it has been authorized by disciplinary specialists. A scientific truth has little standing until it becomes a collective product. What happens in somebody’s laboratory is only one stage in its construction.

In recent times, peer review has achieved an almost mythical status as a mark of scientific respectability.² It rivals statistical inference as the preeminent mechanism for certifying a finding as impersonal and, in that important sense, objective. It is by no means sufficient in itself to establish the validity and importance of a claim, however. Indeed, it is a mistake to speak as if the validity of truth claims were the principal outcome of experimental researches. Experimental success is reflected in the instruments and methods as well as the factual assumptions of other laboratories. Day-to-day science is at least as much about the transmission of skills and practices as about the establishment of theoretical doctrines.³ Experimental truth claims depend above all on the ability of researchers in other laboratories to produce results sufficiently similar, and to be convinced that the similarity is indeed sufficient.

Just how this transmission of skills, practices, and beliefs takes place is among the crucial issues in contemporary studies of science. Significantly, the problem has arisen in the context of the new interest in laboratories and experiments. Already in the 1950s, Michael Polanyi argued that science involved a crucial element of “tacit knowledge,” knowledge that could not be articulated or reduced to rules. In practice, this meant that books and journal articles must necessarily be inadequate vehicles for the communication of such knowledge, since what matters most cannot be conveyed by words. Following his reasoning, the crucial institution for the transmission of science is an apprenticeship undertaken by a student with a master scientist.⁴

To argue this way is to diminish the importance of the published paper or textbook, to locate knowledge first of all in the laboratory and not the library. It is to doubt the universality of science, to confine it to particular spaces. In principle, of course, the barriers around those spaces are easily breached. Nature, we suppose, is uniform: another researcher carrying out the same procedures, even on another continent or in another century, should obtain the same results. Such a principle, though, counts for little unless it can be instantiated in practices. In practice, replication is anything but easy. This insight has been developed most fully by Harry Collins, who considers that independent replication is effectively impossible. Those who try to build their own copy of a new instrument or experimental setup, on the basis of printed information alone, normally fail. Detailed reports and private communications make it easier to reproduce an experiment, but also compromise any claims to independence. The usual way of learning to use a new instrument or technique is to experience it directly. This, argues Collins in a case study that is now widely regarded as paradigmatic, is the only way that the TEA laser was ever reproduced.⁵ He may exaggerate the point, but this is a phenomenon that practicing scientists have long understood. Ernest Lawrence warned in the 1930s, for example, that it would be foolhardy to attempt to build a cyclotron without sending someone to work with one in his Berkeley laboratory. “It is rather ticklish in operation,” he explained, “and a certain amount of experience is necessary to get it to work properly.”⁶

This line of argument may have important implications for our understanding of claims to scientific truth. If experimental setups are really so ticklish, and the phenomena so difficult to produce reliably; if experimental findings are almost never independently replicated, but instead are always reproduced using instruments that have been calibrated against the original: then experimental regularities should perhaps be interpreted in terms of human skill rather than of stable underlying entities and the operation of general laws of nature. Or if these alternatives are not incompatible, then at least the problem of transporting skills beyond the confines of a single laboratory must be seen as a critical one. Without such communication there could be nothing like objectivity, since every laboratory would have its own science. To recur again to Polanyi’s language, science would be nothing but “personal knowledge.”

Polanyi himself didn’t think that it was: “Whenever connoisseurship is found operating within science or technology we may assume that it persists only because it has not been possible to replace it by a measurable grading. For a measurement has the advantage of greater objectivity, as shown by the fact that measurements give consistent results in the hands of observers all over the world.”⁷ Here, though, he attributed to the very nature of measurement what had in fact been accomplished within certain domains through heroic efforts. The construction of measurement systems that could claim general validity was not simply a matter of patience and care, but equally of organization and discipline. Administrative achievements of this kind lie at the heart of most experimental and observational knowledge. Mathematics and logic were less intractable from this standpoint.

Theoretical reasoning is of course not beyond criticism. It is, for example, vulnerable to the charge that it has been spun out by a fevered brain, and bears no relation to any actual world. On the other hand, it adapts very nicely to the printed page, which in retrospect seems its natural medium. Thus it can be communicated far more easily than anything depending on special experience. And rigorous deduction can almost compel assent. In the extreme case of pure mathematics, those who accept, even as useful fictions, the axioms, should be led ineluctably to the conclusions. To be sure, mathematized theory in science is rarely so pellucid or so rigorous that its significance and bearing can be grasped immediately by distant readers. Appreciation of this kind of science, too, is easier for those who share an intellectual community with the author. As Polanyi observed, even inference according to formulas remains an art: “There exist rules which are useful only within the operation of our personal knowing.” Collins argues similarly about mathematical deduction and artificial intelligence.⁸ Still, distance is much less of an obstacle for purely theoretical sciences than for sciences of experience, and the problem of reproduction is correspondingly attenuated. Little wonder that the term “science,” meaning demonstrated knowledge, was applied to logic, theology, and astronomy long before there were communities of experimental researchers.⁹

In the seventeenth century, experimentation was still associated with practices like alchemy, with all its connotations of mystery and secrecy.¹⁰ How was this private knowledge transformed into fit material for a culture of objectivity? The historical literature has only just begun to deal with this question. Sociologists have taken it more seriously. At least two lines of response are being developed. One focuses on how experimental results, which can normally be witnessed by only a few people, came to be accepted as truthful by nearly everyone. This was above all a triumph of rhetoric—of what I call here technologies of trust—and also of discipline. Parts 1 and 3 of this book are centrally concerned with these issues, though not mainly with respect to laboratories.

The other broad explanation for the objectification of experiment emphasizes the spread of laboratory practices. Independent replication may be rare, but the reproduction of methods is not. By the eighteenth century, experimental knowledge had to a large degree come to be defined in terms of potential reproducibility. Seventeenth-century experimental philosophers, such as Robert Boyle, exhibited a great fondness for the odd happening, whose intractability was taken as testimony to the advantage of experience over vain theorizing. But singular events provided a poor basis for making communities of researchers, since those who were not present could do little with them but hope they had been faithfully reported. Lorraine Daston instances Charles Dufay, a French researcher of the 1720s and 1730s, to epitomize a different experimental ideal. Whereas Boyle was famously prolix, Dufay was austere, informing his readers only of what was essential for producing an effect. And he considered that the effect should not be reported until it had been brought under good experimental control.¹¹ Such practices enhanced the lawlikeness of nature, since well-behaved laboratory phenomena would thereafter have a more secure ontological status than mere events. They also promoted a spirit of public knowledge, at least within the specialist community, since close laboratory control offered the best chance for reproducing work at other sites.

Still, the obstacles to the replication even of what seem to us the most basic of experiments, like Newton’s separation of colors using a prism, could be formidable.¹² Personal contact, often involving extended visits to other laboratories, was and remains invaluable for the sharing of methods and results. Boyle’s contemporaries seized every opportunity to view his air pump in operation, and to witness the results he claimed to have produced.¹³ In our own time the spread of instruments and techniques through direct contact has been institutionalized in a variety of ways. Most involve brief or lengthy visits. Those who want to master a new instrument or technique travel to a laboratory where it is already working if they are young, or import a graduate student or postdoctoral researcher from such a laboratory if they are well established. Knowledge, then, does not diffuse uniformly outward from the place of its discovery. It travels along networks to new nodes, and what appears as universal validity is in practice a triumph of social cloning.¹⁴

In the early life of a new technique, when it is still on the cutting edge, personal contact will most often be crucial for its spread to other laboratories. Indeed, this may be just what “cutting edge” means in experimental science. But experiments that succeed, again perhaps by definition, will not long remain in the domain of intricate craft skill and personal apprenticeship. The air pump may again be taken as emblematic. Boyle required the most prodigious efforts of glass-blowing and the most adept handlers of leather and sealing wax, as well as a large personal fortune, to build a pump that worked some of the time. Already in Boyle’s day, though, there were shops specializing in scientific instruments, and they soon added air pumps to their repertoire. Any air pumps that were incapable of producing the experimental phenomena associated with a vacuum would be sold at first to unhappy customers, and then not at all. As the pumps were improved and standardized, the phenomena became more easily reproducible.¹⁵ In recent times, such technologies have proliferated. Not only have instruments been standardized; nature has too. Chemists buy purified reagents from catalogs—and they would be quite helpless if they had to extract them from the soil. Cancer researchers depend on patented strains of mice and would not know how to interpret results derived from ordinary field mice.

The growth of science has to a large degree involved the replacement of nature by human technologies. Ian Hacking has made this insight into the basis for an important general book on philosophy of science. Experiments succeed, he observes, when they permit the reliable manipulation of objects. At least some of these objects, such as lasers, may never exist outside the laboratory. Most or all cannot be found in anything like a pure form, except when they are created by human interventions. But as these artificial or purified objects come to be more reliably manipulated, they begin to be incorporated into other experiments, and perhaps also into processes outside the laboratory. This is perhaps the most crucial sense in which laboratories are self-vindicating.¹⁶

Bruno Latour argues that science is now inseparable from technology, and uses the term “technoscience” to symbolize their merger. Both, he suggests, aim to construct black boxes, artificial entities that are treated as units and that nobody is able to take apart. The black boxes of the scientist may be laws or causal claims as well as material technologies, but these depend on instruments and reagents for their production, just as instruments cannot be built, operated, or interpreted without the benefit of scientific knowledge. Our interventions have become too powerful for us to talk usefully any more about science in terms of learning what happens in nature, independent of human activity. Every scientific claim succeeds by mobilizing a network of allies: reagents, microbes, instruments, citations, and people. If the network is strong, a new fact is created. It is an artifact, but it is nonetheless real, for it can be enliste...

Índice

• Cover Page
• Title Page
• Copyright Page
• Contents
• Preface
• Acknowledgments
• Introduction: Cultures of Objectivity
• Part I: Power in Numbers
• Part II: Technologies of Trust
• Part III: Political and Scientific Communities
• Notes
• Bibliography
• Index