Paper for the 4th Nordic Pragmatism Conference
Fallibilism, as an aspect of pragmatist epistemology that Charles Peirce assigned to scientific inquiry, can be illuminated by a study of law. Before he became a judge, Oliver Wendell Holmes, Jr., along with Peirce, Chauncey Wright and William James, associated as members of the Metaphysical Club of Cambridge in the 1870s, recalled as the birthplace of pragmatism. In early essays, Holmes advanced a concept of legal fallibilism as incremental social inquiry. Holmes treats common law cases more like scientific experiments than as deductive applications of already clear rules. Common law rules may be seen as a product of 1) the conflicts that occur in society, 2) the channeling of conflicts into legal disputes and decisions, 3) the gradual classification of judicial decisions into groups, and 4) the development of consensual understanding, expressed in rules and principles, as to how future cases should be classified and decided.
This does not involve only lawyers and judges. Especially in controversial cases, it may indirectly involve an entire community or society. The legal process is seen as an extended intergenerational process of inquiry. It illuminates the relation of thought, expression, and conduct among a community of both expert and lay inquirers, applied to the problems of social ordering. It thus illuminates the view of inquiry as commonplace problem-solving and its relation to professional investigation and reflection.
The study of legal fallibilism also addresses the connection between natural and normative inquiry. Both legal and scientific inquiry are fundamentally social and part of the general study of the sociology of knowledge. Both focus on the primacy of cases and exemplars in the process of intersubjective classification, and on the dual role of concepts in guiding the conduct of professional inquirers and framing and maintaining the coherence and consistency of both expert and general belief.
I am grateful for the opportunity to explore with you an aspect of pragmatism that I believe to be underappreciated: its philosophy of law. In the brief time available I would like to elucidate legal fallibilism as a distinctive theory of law, focusing on its roots in the early essays of Oliver Wendell Holmes Jr. Then I will address its contribution to a more general social theory of inquiry (as a sociology of legal knowledge), and finally touch on the merits of this as a research program for philosophy.
Legal inquiry connects everyday problems with professional and expert knowledge. It explores the function and interaction of discrete communities of inquiry, both expert and lay. Moreover, it provides insight into the relation of natural and normative inquiry. For a joint understanding of these two traditionally distinct areas of knowledge, we may look to Charles Peirce and Oliver Wendell Holmes Jr., respectively. As my time is limited I will discuss Holmes, and then make a comparison with Peirce that would surely have horrified both of them. The two men were as distant in their personal relations as they were close in their philosophical radicalism.
First, this requires a break with conventional Holmes scholarship. To borrow a phrase from Peirce, Holmes has been kidnapped—by the conventional paradigm of analytical legal positivism, long dominant in western legal philosophy. He is commonly associated with twentieth century legal realism, and you are surely aware of the Scandinavian legal realists. Early twentieth century legal realism was influenced by then contemporary social and behavioral science. It emphasized legal reform, motivated by a reaction against the false certainty of “formalist” and “mechanical” jurisprudence. In the previous century, Holmes had (1872a, 92) defined law as prediction of what courts will enforce, which was later interpreted as judicial behaviorism or instrumentalism. In essence, it was entirely different.
The evidence is found in two formative papers that Holmes wrote in the 1870s. In the first (1870), he notes that Anglo-American common law “decides the case first and determines the principle afterwards,” in a process of gradual cumulative classification and generalization that he called “successive approximation.” He cautions against judges giving premature reasons in deciding unfamiliar cases, and advocates highly particularized decisions in the early stage of inquiry into new classes of dispute. The judge, assisted by the jury, should simply apply a standard of prudence, or the foreseeability of harm under novel conditions.
Holmes drew on John Stuart Mill’s criticism of the syllogism, which Mill saw as reasoning not from general to particular but from “particulars to particulars.” In his famous System of Logic (1843), which Holmes read in 1867, Mill declared that the general is simply used as a guide. But for Holmes in his 1870 essay, the relevant general cannot be used as a guide for new particulars, because itdoesnot yet exist. How does it come to be? As new problems arise and new disputes are decided, gradually a pattern emerges. “It is only after a series of determinations on the same subject-matter, that it becomes necessary to ‘reconcile the cases,’ as it is called, that is, by a true induction to state the principle which has until then been obscurely felt. And this statement is often modified more than once by new decisions before the abstracted general rule takes its final shape. A well settled legal doctrine embodies the work of many minds, and has been tested in form as well as substance by trained critics whose practical interest it is to resist it at every step.” (1870, 77). He would later emphasize that such generalization is influenced by feedback and adjustment within society.
I. Normative inquiry, and normative knowledge, begins in law with disturbances in the social fabric channeled into systematized and participatory dispute resolution.
To illustrate, let’s take an imaginary visit to Copenhagen harbor back in the days of sail. With the vagaries of wind and tide there must have been constant collisions bringing ship owners into the courts claiming money for loss of cargo and damage. Imagine a detailed account unfolding in the courtroom of how two ships collided, perhaps at night. When did one crew see the other ship, what did the crew do then? Sailing is tricky and complicated, and in the absence of a clear error, we’ll assume a fair hearing so that the judgment goes against the vessel that was least prudent under the circumstances. Can we fairly say the case was decided by a rule of law?
Over time similar collisions occur and prudent practices develop to the point where the courts can and will say, yes, this ship or that was burdened and failed to display a certain light or post a lookout or douse a certain sail to avoid the collision. An illustration of this might be the display of all those colored lights on ships at night, identifying sailboats, anchored boats, tugboats, barges. Their first introduction led to cautionary rules and thence to decisions and legal standards and rules. Thus does the class of “collision cases” develop into general standards and rules, over time.
II. Legal normativity is a web or network of standards emerging from disparate practices and woven together by professionals whose mission it is to impose coherence, predictability, and consistency.
This is not, I hasten to say, a precise historical account. It is a loose simulation drawn from the 1870 essay by Holmes, which in turn is the product of several influences: a close study of 19th century English and American cases, broad reading in philosophy as well as law, an attitude toward knowledge shared with his friends of the Metaphysical Club, and the influences on them from the Scottish Enlightenment, applied to their readings of Kant, Hegel, and Darwin (Kellogg, 2007). This mix of influences has been said to have led pragmatism toward a radically naturalized reading of Kant and Hegel.” (Margolis, 2010)
You may see elements of a Darwinized Hegel in Holmes’s approach to rule-making, perhaps influenced by his only admitted mentor Chauncey Wright, who in 1873 published an influential essay “The Evolution of Self-Consciousness,” written at the encouragement of no less than Charles Darwin himself. Holmes appears to have absorbed Wright’s attitude, and he took it in a different direction, toward the development (he avoids the term “evolution”) of legal intelligence, as part of a socialized ordering process. And now you see what he implied in defining law as prediction of what courts will do. Law is not a set system of rules with a preexisting answer for every new case. It is a constantly developing system of classification.
Two years later Holmes writes another important essay (1872b). Here he addresses a more difficult issue. Most legal cases do not really match the simulation I just gave. They come into a context of preexisting law. Difficult cases often seem to be enmeshed between two (or more) opposing precedents or generals. The example he uses is the conflict of nuisance with property rights, like the battle between neighboring landowners over the placement and height of a wall. Upon repeated instances, in the absence of legislation, the courts will eventually work out a formula for placement and height. Thus are opposing generals reconciled over time, again through fallibilist inquiry.
However, this is hardly the conventional view. The dominant analytical approach to jurisprudence views law as an authoritative and comprehensive body of doctrine. The assumption that it always contains an answer fails to explain the persistence of difficulty and uncertainty—dare I say novelty. This attitude gives rise to skepticism, and many legal realists went to the opposite extreme in seeing uncertain cases as “legally indeterminate.” This opens the door to judicial behaviorism or instrumentalism—law is the sum of subjective influences on judges, or their immediate sense of the “best” consequences.
III. In the problem of the doubtful case we find the advantage of pragmatism as a theory of law. The analytical model leads to an all-or-nothing dualism. The putative certainty of analytical fundamentalism is opposed by a cynical subjectivism, even relativism. Pragmatism sees the doubtful case as a stage of inquiry and classification.1
In his 1873 essay, Holmes proposes an alternative to the analytical model. In the doubtful case, opposing generals are not reconciled either by analytical logic or judicial behaviorism or instrumentalism, but again by a social process of experimental, successive approximation. He applies the earlier cumulative model of 1870 to the problem of resolution of conflicts among rules and precedents. Again, his approach is, “particularize first, generalize later.” Here is the key passage:
The growth of the law is very apt to take place in this way: Two widely different cases suggest a general distinction, which is a clear one when stated broadly. But as new cases cluster around the opposite poles, and begin to approach each other, the distinction becomes more difficult to trace; the determinations are made one way or the other on a very slight preponderance of feeling, rather than on articulate reason; and at last a mathematical line is arrived at by the contact of contrary decisions, which is so far arbitrary that it might equally well have been drawn a little further to the one side or the other. (1872b, 119)
Holmes suggests a process whereby new experience falls into a grey area between existing generals, eventually revealing a new pattern which first appears as a “line,” ultimately redefining the generals themselves.
In a moment I will ask whether this normative model has any parallels to natural inquiry. Can the two learn from each other? The influential “Edinburgh School” of the sociology of scientific knowledge (SSK) has advanced a remarkably similar classification model of scientific inquiry, emphasizing that scientific knowledge proceeds from particulars to particulars, and that “every act of classification has the form of a judgement, every act changes the basis for the next act, every act is defeasible and revisable. . . .” (Barnes et al., 1996, ix). I have found that there is much to be learned from a comparison here of the interaction between particulars and generals, and the role of communities of inquiry, in the cumulative growth of both legal and scientific knowledge.
But first I should address whether Holmes’s model is relevant for contemporary law. Does law really follow patterned judgments by communities of inquiry, as Holmes suggested 140 years ago, or is everything handled by legislation and administrative rulemaking?
There is a danger today of being kidnapped by the dominant analytical model of law dating at least from Jean Radin and Thomas Hobbes and reinforced by Hans Kelsen and John Austin (not to mention Napoleon Bonaparte!) Law is the creation of the sovereign, or a code, or a set of texts or other authorities—and embodied in a static analytical matrix. This view is so entrenched that Holmes is conventionally viewed as within the analytical positivist tradition. I have argued for 30 years that he rejected the analytical model. I have to concede that he made comments that may sound sympathetic to the analytical model—he was, after all, a judge. But careful examination reveals that he consistently returned to the 1873 line-drawing analogy—the classification model of law—throughout 50 years on the bench, applying it even to legislation and constitutional law (Kellogg 2007).
Can we find a classification model in contemporary law? How about a new problem like assisted suicide? This class of dispute started out as a series of criminal prosecutions of doctors for murder, until the opposing claim of personal autonomy got some traction, from constitutional language, applied to changing medical circumstances. The problem soon found its way to the appellate courts. In 1999 Professor Cass Sunstein wrote a book called A Case at a time: Judicial Minimalism on the Supreme Court, in which he cautioned the same thing as Holmes did in 1870—decide the cases one at a time, it’s often premature to lay down a sweeping general rule. Ultimately, we may need legislation, but even that can’t come too soon, before the experimental stage, which includes a process of feedback and adjustment. Legislation is part of the process of inquiry.
One implication of this is to undermine the classical model of democratic social choice, famously criticized by Kenneth Arrow. Social choice is constantly ongoing outside the ballot box, for good or ill, in the process of conflict resolution, influenced by feedback from various relevant communities. As with assisted suicide, each successive decision responds to feedback from diverse communities of interest, which may include medical, legal, and academic professionals, senior citizens, lobbying groups, and so on—even philosophers! Every decision is influenced by social adjustment and the adoption of new practices, like new medical procedures and living wills. The classical democratic model falls short in ignoring this continuing process of adjustment.
What are the key elements here? 1. We are looking at cases not singly, as raising a question of existing law against a synchronic analytical background, but as stages of inquiry into social problems, and against a diachronic background. 2. Notwithstanding the role of “great judges,” the guiding intelligence is not individual but social—hence it implies a socialized epistemology. 3. Inquiry itself is generated not by dispassionate curiosity but by conflict, and is far messier than any ideal model of dialogue. 4. Inquiry takes place in a context of preexisting generals to which we look backwards even while plotting new cases in relation to them. 5. The judicial role of comparing and contrasting is best understood as an incremental and cumulative line-drawing. 6. Judges are best seen as members of a community of inquiry, but acting within a network of other communities, both expert and lay. 7. The interaction between disparate communities operates as a “feedback loop” from judicial decisions to their effects, which feed new experience back into the judicial system.
IV. Pragmatism replaces the analytical problem of “legal indeterminacy” with a study of inquiry into uncertainty, leading to classification, as an aspect of the sociology of (legal) knowledge.
These are some of the principal insights of legal fallibilism for the pragmatist theory of inquiry. Can this normative study of law enlighten our understanding of natural inquiry, or of inquiry in general?
Peirce and Holmes: the Real and the Right as Organizing Concepts
Kenneth Stikkers puts Peirce in the forefront of the history of sociology of knowledge (2009). He notes that Peirce had already suggested, prior to Dilthey and Durkheim and without any apparent benefit from the insights of Marx, that the forms of human knowing are fundamentally forms of social life, without reducing the latter to the forms of economic life. It is unfortunate that Holmes and Peirce were not more compatible. If they had been more inclined to engage each other, perhaps the combination of Peirce’s breadth with Holmes’s empirical and historical focus might have illuminated parallels between natural and normative inquiry and the role of “umbrella concepts” such as the “real” and the “right.”
Let us examine a famous passage in which Peirce addresses the role of inquiry in constituting what we understand as the “real”:
The real, then, is that which, sooner or later, information and reasoning would finally result in, and which is therefore independent of the vagaries of me and you. Thus the very origin of the conception involves the notion of a COMMUNITY, without definite limits, and capable of complete information, so that reality depends on the ultimate decision of the community. . . .Reality consists in the agreement that the whole community would eventually come to. (Peirce 1984, 239, 241, 252)
Peirce held that the social impulse, the desire to reconcile our personal habits with those of our neighbors, leads us to believe in the “independently Real.” The real, then, is for Peirce an organizing concept. Inevitably we find that others hold views different from ours, and sooner or later the strength of our tenacity is worn away. Unless we make ourselves hermits, we shall necessarily influence each other’s opinion, so that the problem becomes how to fixate belief, not in the individual merely, but in the community as a whole. (Peirce 1986, 250)
Peirce’s model of knowledge as inquiry is radically naturalist. It is rooted in the doubt-belief formula, which Dewey later elucidated as a theory of social inference.2 Peirce extends this naturalist formula all the way up, to metaphysical questions concerning the “real.” It also famously implies the element of “construction” of the real as a social phenomenon. But it takes naturalism only so far, and leaves the discussion of inquiry and social construction still uncomfortably abstract.
We should be wary of assuming that “inquiry” is a natural or endemic condition. It is certainly not rooted in pure curiosity or always done in the antiseptic context of a library or a laboratory. Peirce emphasizes that inquiry is prompted by doubt. What is doubt? How does it arise and operate? How does inquiry then occur? The doubt-belief formula needs to acknowledge that doubt must have its own history and physiology. Legal doubt is driven by the problem of disputes flowing into the courts. This doubt is rooted, then, in conflict. Holmes, who fought in the American Civil War, was acutely aware that conflictual doubt could take many forms, including the desire for revenge, and be resolved in other ways, including by violence.
Since Peirce and Dewey, the cutting edge of fallibilism studies has been in the history and philosophy of science, influenced by more recent texts such as those by Thomas Kuhn, Imre Lakatos, Paul Feyerabend, Larry Laudan, and others. Empirical studies of fallibilism are to be found in the burgeoning studies of actual research programs, including the “strong programme” of the University of Edinburgh, where I held a MacCormick Fellowship in 2009.
I was struck there by the remarkable similarities of the “strong programme” of science studies to the Holmes model of legal inquiry. I have already quoted passages from the signal Edinburgh text, Scientific Knowledge: A Sociological Analysis (Barnes et al., 1996). There also, the model of inquiry is reasoning from particulars to particulars, a social and interactive process of classification, with an emphasis on conceptual transformation. Comparing the Holmes model of law with the Edinburgh model of science opens up the connection of two hitherto separate areas of knowledge, and heralds an expanded relevance of sociology of knowledge for philosophy.
Turning to normative inquiry, even more apparent is the role of the “right” as an organizing concept. The notion of right directly influences action, even as the notion of the real does so less directly. The literature on the right is distinctly structured in a hierarchical manner, indeed governed by the concept of justice, itself a highly structural idea. But there are two separate bodies of literature on the “right,” the legal and the moral. Differentiating the legal from the moral right is a vast body of literature addressing such notions as authority and process. In contrast to the analytical approach, which has dominated this body of work at least since Hobbes, Holmes shows how the legal right is in constant transformation.
V. Analytical positivism maps legal knowledge as a fixed matrix. While a formalist positivism has been challenged in science, it remains dominant in law. Holmes’s early research is the first clear model of law in transition as a system of social classification.
What is missing from the contemporary body of legal literature is found in the literature of science studies: a detailed critique of the social system of inquiry and classification itself. Analytical jurisprudence is blind to this, because it is rooted in an individualist and static epistemology. John Austin focused on the nature of legal rights as fixed commands. H.L.A. Hart turned instead to the nature of law as a body of rules, revealed in a close study of legal language. Ronald Dworkin took exception to the model of rules, making the case for the operation of “principles.” Rather than undermining the positivist model, Dworkin effectively preserved the notion that there is always an a priori correct answer to any doubtful case.
Missing from the analytical model is any detailed concern with methods of research, the interaction of professional and lay communities of inquiry, actor-network theory, the social nature of research traditions, and many other aspects of the sociology of knowledge commonly found in contemporary science studies. Here lies rich potential for a new generation of legal research.
There are other important aspects of Holmes’s early research that have been widely ignored and deserve more extended attention than I have time for. Soon after the two essays just quoted, he delved into legal history and anthropology, leading to his 1881 treatise The Common Law, incorporating several further insights.
For example, contemporary normative ideas and practices, even the modern rules of liability for the ships colliding in Copenhagen Harbor, have roots in distant past, in primitive notions, which are given new reasons even while the original practices remain. For example, the limitation of liability in admiralty law is a product of the ancient desire for vengeance, which was mitigated by the “primitive” practice of surrender of the offending instrument of harm. For Holmes, legal inquiry into personal injury was a historical replacement of the ancient blood feud, retaining survivals of its ancient past.
The process of replacement of primitive notions with putatively “rational” models has been messy, chaotic, uneven, utterly incompatible with Rawlsian or Habermasian ideal conditions of dialogue. But it has also produced important prophylactic notions such as “due process.” It is not “self-correcting” in the sense often given to scientific fallibilism, but is voluntaristic and malleable. Unlike abstract and idealized models of dialogue, its actual nature can be examined in historical detail, warts and all.
Having reached the limit of my time, I will end with a provocative late comment by Holmes in 1899, suggesting that his diachronic map is applicable to the moral realm, indeed to the realm of ideas in general:
It is perfectly proper to regard and study the law simply as a great anthropological document. It is proper to resort to it to discover what ideals of society have been strong enough to reach that final form of expression, or what have been the changes in dominant ideals from century to century. It is proper to study it as an exercise in the morphology and transformation of human ideas. (1899, 212)
Holmes brings to the general theory of inquiry a more distinct focus on history and morphology. He suggests in this remarkable passage that the bones of human metaphysics lie barely concealed in the historical record, and that their connection with human conduct and social change is palpable.
Barnes, Barry, David Bloor, & John Henry. 1996. Scientific Knowledge: A Sociological Analysis. Chicago: Chicago University Press.
Holmes, Oliver Wendell. 1870. “Codes, and the Arrangement of the Law,” in Kellogg (1984), 77.
____1872a “Book Notice” reprinted in Kellogg (1984), 91
____1872b “The Theory of Torts” reprinted in Kellogg (1984), 95
____1881. The Common Law. Boston, Little Brown & Co.
____1899 “Law in Science and Science in Law” Collected Papers at 212. Harcourt, Brace and Company.
Kellogg, Frederic R. 1984. The Formative Essays of Justice Holmes: the Making of an American Legal Philosophy. Westport & London: Greenwood Press.
_____. 2007. Oliver Wendell Holmes Jr., Legal Theory, and Judicial Restraint. Cambridge University Press.
_______. 2009. “The Construction of Legal Positivism and the Myth of Legal Indeterminacy.” ssrn.com/abstract=1536124
Margolis, Joseph. 2010. Pragmatism’s Advantage. Stanford: Stanford University Press.
Mill, J.S. 1843. A System of Logic, Ratiocinative and Inductive.
Peirce, Charles. 1984. Writings of Charles S. Peirce: A Chronological Edition, vol. 2, 1867-1871. Bloomington: Indiana University Press.
_____. 1986. Writings of Charles S. Peirce: A Chronological Edition, vol. 3, 1872-1878. Bloomington: Indiana University Press.
Sleeper, Ralph. 1986. The Necessity of Pragmatism: John Dewey‘s Conception of Philosophy. Yale University Press.
Stikkers, Kenneth, et al. 2009. John Dewey Between Pragmatism and Constructivism. New York: Fordham University Press.
Sunstein, Cass R. 1999. One Case at a Time: Judicial Minimalism on the Supreme Court, Cambridge and London: Harvard University Press
Wright, Chauncey. 1873. “The Evolution of Self Consciousness.” In Philosophical Discussions. New York: Henry Holt & Co., 1877.
1 Both untenable positions stem from the positivist, analytical model of law, viewed as having a fixed boundary with definable contents. See Kellogg, 2009.
2 For Ralph Sleeper, the key to Dewey’s logic was understanding inference “as a real event of transformational force and power, causally real in the emergence of new features of things ‘entering the inferential function.’ It takes inference as action, as behavior that causes changes in reality through interaction with things.” (1986, 83).