“T
alks too much” was the comment on the first report card of Oliver Wendell Holmes, Jr., the future Supreme Court justice and founder of the constitutional law of free speech.
1
Wendy Holmes (as he was known to his intimates) was six. He lived for another eighty-seven years, and no one ever accused him of keeping his views to himself. He opined regularly to dozens of correspondents, among them the English jurist Sir Frederick Pollock (their letters are collected in two volumes), the political theorist Harold Laski (two volumes), the Washington lawyer Felix Frankfurter (one volume), the diplomat Lewis Einstein (one volume), the Irish priest Patrick Augustine Sheehan (one volume), the progressive journalist and eccentric Franklin Ford (one volume), the philosopher Morris R. Cohen, the Chinese jurist John Wu, the Japanese nobleman Kentaro Kaneko, and a stable of female confidants that included Nina Gray, Alice Stopford Green, Baroness Charlotte Moncheur, and Holmes’s Anglo-Irish
paramour, Lady Clare Castletown, of Granston Manor, Upper Ossory, Ireland.
When Holmes’s brothers on the bench—he served for twenty years on the Supreme Judicial Court of Massachusetts and for thirty years, beginning in 1902, on the Supreme Court of the United States—fell behind in the production of opinions, he begged his chiefs to reassign their cases to him. In the end, he wrote over two thousand opinions, believed to be a record for judges sitting in courts of last resort. It was his habit, every few years, to refresh himself by traveling to England, leaving his wife behind, where he sought out the company of the leading philosophical and literary lights of the land and where he flirted aggressively with most of the women he met, once going so far in his banter with the twelve-year-old daughter of one of his hosts, Tom Hughes, the author of
Tom Brown’s School Days
, that Hughes sent a letter afterward demanding to know Holmes’s intentions. He had none, except to experience once again the pleasure of his own flamboyance. “He would catch a subject, toss it in the air, make it dance and play a hundred tricks, and bring it to solid earth again,” an English acquaintance described his social manner. “He liked to have the ball caught and tossed back to him, so that he could send it spinning away again with a fresh twist.”
2
Holmes’s father, Dr. Oliver Wendell Holmes, Sr., a pioneer of the germ theory of disease, the author of the patriotic poem “Old Ironsides,” cofounder of the
Atlantic Monthly
(whose name he came up with), coiner of the term “Boston Brahmin,” the first person to refer to Boston as “the Hub of the Solar System,” and dean of the Harvard Medical School, was said to have been the greatest talker of his day. He thought the talent worth acquiring, and made it a rule that any child who uttered a clever remark at the dinner table be given extra marmalade. Holmes Sr. was five feet three and round; his son, fully grown, was six-three and lean, and their relations were notoriously fractious. But this was one trait that got passed along.
The law provided Holmes with a steady supply of occasions to exercise his gift for turning a phrase, but he did not allow the engine
to idle. Early in his career, when he was in practice in Boston, he would walk into the office in the morning and announce to one of his firm’s junior associates, “Mr. Evans, I am ready to contradict any statement you will make.”
3
Mr. Evans evidently felt it his duty, each morning, to oblige. Fifty years later, Holmes liked to tell his colleagues on the Supreme Court, when they were conferring about a case, that he would admit any general principle of law they proposed, and then use it to decide the case under discussion either way. It was not persiflage, or it was not only persiflage. Holmes had a profound appreciation for the malleability of words, and that appreciation is the demon that sits at the bottom of his thought. His challenge to his fellow justices was not an excuse to show off his forensic dexterity; he was making a point about the nature of language. There have been hundreds of efforts since Holmes published
The Common Law,
in 1881, when he was thirty-nine, to sew a political label on him. Commentators have tried to prove that he was a progressive, a liberal, a civil libertarian, a democrat, an aristocrat, a reactionary, a Social Darwinist, and a fascist. But none of these efforts has gotten to the core of what Holmes did, because all of them have mistaken the implications for the premises. They have focused on the accidents of his thought.
Calling Holmes a progressive or a reactionary is like calling, say, Wittgenstein a progressive or a reactionary. It assumes that he was interested in the political consequences of his ideas. But one thing that can be said with certainty about Holmes as a judge is that he almost never cared, in the cases he decided, about outcomes. He didn’t read the newspaper, and he was utterly, sometimes fantastically, indifferent to the real-world effects of his decisions. He saw the law as a series of minutely varied, frequently boring, but sometimes delightfully nasty intellectual problems. These took the form, in his view, of concrete conflicts, each with some slight but potentially crucial point of difference from the rest, submitted for resolution by the ultimately absurd process of reasoning out a result from abstract principles. From the very beginning of his career, Holmes regarded the reasoning part as factitious and, in its relation to the result, logically ex post facto. “It is the merit of the common law that
it decides the case first and determines the principle afterwards,”
4
is the first sentence of the first legal article he ever published, “Codes, and the Arrangement of the Law” (1870), and he spent the next sixty-two years as a jurist and a judge trying to be faithful to this insight. That it was an impossible task by definition—as though one were to make it a principle never to rely on principles—seems only to have whetted his sense of sport.
Holmes repeated his view of the relation between reasons and results in the opinion that, after his opinions in the free speech cases, is his most celebrated, the dissent in
Lochner vNew York
. (1905): “General propositions,” he said there, “do not decide concrete cases.”
5
This is a radical assertion; for it leaves us with the question, If principles don’t decide cases, what does? Holmes’s effort to answer that question led him, after many years, and very improbably, to the free speech opinions which capped his career, and on which his reputation now rests.
In 1995, the University of Chicago Press published the first three volumes of
The Collected Works of Justice Holmes
, which include all the nonjudicial writings, and is the first such comprehensive edition. It was edited by Sheldon M. Novick, whose
Honorable Justice
(1989) was the first full-length biography of Holmes to make use of all the archival material—a fact that is testimony chiefly to the speed with which Novick was able to take advantage of the opening, in 1984, of the Holmes Papers to biographers after fifty years of failed attempts to get an official life completed. Mark DeWolfe Howe, a professor at Harvard Law School who had once served as Holmes’s secretary, managed to finish two detailed and elegant volumes,
The Shaping Years
(1957) and
The Proving Years
(1963); but these take the story only to 1882, the mere brink of Holmes’s fifty-year career as a judge, and Howe died in 1967. Novick’s biography, though serviceable as a narrative of Holmes’s life, was notable for its relegation of virtually all discussion of Holmes’s jurisprudential work to the footnotes, which sometimes assumed a rather disputatious
tone. In his main text, Novick suggested that Holmes espoused “a kind of fascist ideology”; he described Holmes as “a violent, combative, womanizing aristocrat whose contribution to the development of law was surprisingly difficult to define”; and he announced it to be his conclusion that Holmes was, in his thought, a structuralist—a judgment that can be called, among the many that have been offered about Holmes’s work, unique.
6
How this came to be the person chosen to put together the official edition of Holmes’s writings by the trustees of the Holmes Devise (the fund established from Holmes’s bequest of a portion of his estate to the government of the United States) is a mystery, as is the decision of the University of Chicago Press to publish the volumes in the form in which they appear. The first volume contains 333 pages of text; 198 are by Oliver Wendell Holmes, and 135 are by Sheldon M. Novick. Most of these are devoted to a disquisition on “Holmes’s Philosophy and Jurisprudence,” with an ensuing “Critical Appraisal.” This is a subject which has received continuous attention, at a pretty high level of critical inquiry, since about 1914. A bibliography of Holmes scholarship (none is provided) might have served the purpose more economically and, in what is presented as a memorial edition, more tastefully. Still, though it took sixty years, this is how the job was done.
7
Novick’s idea that Holmes was a structuralist belongs to a history that runs parallel to the history of efforts to define Holmes’s politics. This is the history of efforts to define Holmes’s philosophy. Holmes has been called a formalist, a positivist, a utilitarian, a realist, a historicist, and a pragmatist (not to mention a nihilist). Commentators who cleave to one of these terms usually find themselves spending a good deal of time explaining why commentators who favor one of the other terms cannot possibly be right. This is generally an easier business than defending the term they prefer; and the reason is that none of these terms can possibly be right, because each singles out one aspect of the law as the essential aspect, and it was Holmes’s genius as a philosopher to see that the law has no essential aspect.
A case comes to court as a unique fact situation. It immediately
enters a kind of vortex of discursive imperatives. There is the imperative to find the just result in the particular case. There is the imperative to find the result that will be consistent with results reached in analogous cases in the past. There is the imperative to find the result that, generalized across many similar cases, will be most beneficial to society as a whole—the result that will send the most useful behavioral message. There are also, though less explicitly acknowledged, the desire to secure the outcome most congenial to the judge’s own politics; the desire to use the case to bend legal doctrine so that it will conform better with changes in social standards and conditions; and the desire to punish the wicked and excuse the good, and to redistribute costs from parties who can’t afford them (like car accident victims) to parties who can (like car manufacturers and insurance companies). Hovering over this whole unpredictable weather pattern—all of which is already in motion, as it were, before the particular case at hand ever arises—is a single meta-imperative. This is the imperative not to let it appear as though any one of the lesser imperatives has decided the case at the blatant expense of the others. A result that seems just intuitively but is admittedly incompatible with legal precedent is taboo; so is a result that is formally consistent with precedent but appears unjust on its face. The court does not want to seem to excuse reckless behavior (like operating a railroad too close to a heavily populated area), but it does not want to raise too high a liability barrier to activities society wants to encourage (like building railroads). It wants the law to run in a politically desirable direction, but it does not want to be caught appearing to bend an anachronistic legal doctrine in order to compel a politically correct result.
There is also (to put the final spin on the system), within each of these competing imperatives, the problematic that Holmes’s dictum about the nature of reasoning identified, which is the business of deciding what counts as relevant within that particular discourse and what does not. This series of problems begins with the question of what the legally relevant “facts” in the case really are; it runs through the questions of what counts as an analogous case, what counts as an applicable general legal principle, what counts as a
benefit to society, and so on; and it ends with the question of what counts as a “just result.” Holmes thought that there were no hard-and-fast distinctions in any of these areas; he believed that the answer always boils down to a matter of degree. And when he said that general propositions do not decide concrete cases, he was saying that even if we were to select one imperative to trump all the others in our approach to judicial decision-making, we would still find that the consequences for any particular case were indeterminate. “Cost-benefit analysis” is as malleable as “rights talk.” When there are no bones, as T. S. Eliot once said, anybody can carve a goose.
There are bones, though. For cases get decided and verdicts get returned and opinions get written, and by a process that does not seem arbitrary or self-interested to the people who do the deciding, returning, and explaining. If the various discourses of fairness, policy, precedent, and so forth are simply being manipulated rather than applied, they are being manipulated to justify an outcome which has been reached in obedience to some standard. When Holmes said that the common law decided the result first and figured out a plausible account of how it got there afterward, the implication was not that the result was chosen randomly, but that it was dictated by something other than the formal legal rationale later adduced to support it. The purpose of
The Common
Law
was to discover what that something was.
The book originated as a series of twelve lectures at the Lowell Institute, in Boston, which Holmes delivered, to a packed hall, extemporaneously. It was the product of prodigious research—Holmes was regarded by his contemporaries as a frighteningly disciplined worker—and it has seemed to many competent commentators a kind of thicket of contradictory approaches, all bristling with pedantic details of legal antiquarianism and technical analyses of legal doctrine.
The impression of methodological confusion arises from Holmes’s unwillingness to see the law under just one aspect. He
was suspicious of formalism—the extraction of abstract legal doctrines from the analysis of sequences of cases—which was the prevailing academic approach to the law at the time: he once referred to Christopher Columbus Langdell, the dean of the Harvard Law School and founder of the case method of legal instruction, as one of “the powers of darkness.”
8
(Langdell was probably in the audience for the Lowell Lectures, and he offered Holmes a job afterward anyway.) But Holmes had, in fact, praised Langdell’s casebooks, and he used the case method himself when he taught at Harvard; and if there are formalist elements in
The Common Law,
it is because there are formalist elements in the law. Doctrinal continuity, after all, is something highly valued by judges; it is, at a minimum, the best hedge against reversal. At the same time, Holmes believed that the law was susceptible to a utilitarian analysis, since the law is also an instrument of social policy; a moral analysis, since the law is a record of the conduct a society sees fit to penalize; and a historical analysis, since the law has historical roots and evolves in response to changing social conditions.
Thus the celebrated sentence in the opening paragraph of
The Common Law
, “The life of the law has not been logic; it has been experience,”
9
does not say that there is no logic in the law. It only says that logic is not responsible for what is living in the law. The active ingredient in the compound, what puts the bones in the goose, is the thing called “experience.” Holmes was using that word in a particular sense. He meant it as the name for everything that arises out of the interaction of the human organism with its environment: beliefs, sentiments, customs, values, policies, prejudices—what he called “the felt necessities of the time.”
10
Our word for it (in many ways less satisfactory) is “culture.”
Understanding Holmes’s conception of “experience” is the key to understanding almost everything that is distinctive about his view of the law. Three features seem especially significant. The first is that experience is not, in Holmes’s view, reducible to propositions, even though human beings spend a lot of time so reducing it. “All the pleasure of life is in general ideas,” Holmes wrote to a correspondent in 1899. “But all the use of life is in specific solutions—
which cannot be reached through generalities any more than a picture can be painted by knowing some rules of method. They are reached by insight, tact and specific knowledge.”
11
Even people who think their thinking is guided by general principles, in other words, even people who think thought is deductive, actually think the way everyone else does—by the seat of their pants. First they decide, then they deduce.
This is obviously the idea that stands behind the assertion, in
Lochner
and many other places in Holmes’s writing, that general propositions do not decide concrete cases. Logical reasoning from a prioris is just not the way people make practical choices. Holmes thought that learning the abstract legal doctrines on which judicial decisions are expressly based—what used to be called “black letter law”—was therefore poor training for a lawyer. Judges do invoke these doctrines when they are explaining their decision, but (as Holmes was pointing out when he volunteered to use the same principle to decide a given case either way) the doctrines are never sufficient to account for the result reached. The hole always has a different shape from the arrow sticking out of it. So that anything that might operate as a motive for a judge’s decision—a moral conviction, a political preference, even (as he put it) “the blandishments of the emperor’s wife”
12
—was, according to Holmes, legally material if it helped lawyers guess the result correctly.
This is the essence of the so-called prediction theory of the law, expressed in Holmes’s most famous essay, “The Path of the Law” (1897), by the sentence: “The prophecies of what the courts will do in fact, and nothing more pretentious, are what I mean by the law.”
13
Holmes was fifty-six when he wrote “The Path of the Law,” and the starkness of its realism, the thoroughness with which it dismisses the notion that the law can be understood by reference to some higher system of morality or rationality, has seemed to some scholars—notably Morton Horwitz—to represent a loss of faith, on Holmes’s part, in the moral cohesion, what Horwitz calls “the power of custom,”
14
of late-nineteenth-century American society. But the prediction theory, at least, was not new to Holmes in 1897. It appears in one of his earliest essays, published in the
American Law
Review
in 1872, where Holmes argued that it is not the law that determines the outcome in a particular case, but what judges
say
is the law. For “a precedent may not be followed; a statute may be emptied of its contents by construction … . The only question for the lawyer is, how will the judges act?”
15
From the very beginning, Holmes’s view of the law was premised on the assumption that law is simply and empirically judicial behavior. A rule may be written down, it may express the will of the sovereign, it may be justified by logic or approved by custom; but if courts will not enforce it, it is not the law, and lawyers who bet their cases on it will lose.
A second distinctive feature of Holmes’s conception of experience is that it is not individual and internal but collective and consensual; it is social, not psychological. This is the feature responsible for his most important contribution to the civil law, which is the invention of the reasonable man. The reasonable man is the fictional protagonist of modern liability theory. If you are injured as a result of an act of mine, what triggers civil liability? There are, traditionally, three ways to answer this question. The first is to say that it is enough merely to prove causation: I act at my peril, and I am therefore liable for any costs my actions incur, whether I could have foreseen them or not. The legal term for this is “strict liability.” The second way of answering it is to say that I am liable for your injury if I wickedly intended it, but I cannot fairly be held liable for injuries I never contemplated. This is the theory of
mens rea
—“the guilty mind.” And the third is to say that even if I neither wished for nor anticipated the possibility of your injury, I am liable to you anyway if my act was careless or imprudent. When I act without exercising due care, then I do act at my peril. This is the theory of negligence.
Holmes treated the problem of liability for injury in a series of dense and intense texts: “The Theory of Torts” (1873), “Trespass and Negligence” (1880), Lectures III and IV of
The Common Law
(1881), and “Privilege, Malice, and Intent” (1894), which is possibly his most brilliant essay. Just what Holmes is saying in these discussions is a matter of apparently endless dispute.
16
The subject seems fraught in part because it is often hard to distinguish, in Holmes’s writing, between the descriptive and the prescriptive—between
what Holmes believed the law was in practice and what he thought the law ought to be. Holmes didn’t do a lot to help his readers make this distinction, but the reason is that his favorite method of argument was to show that what the law ought to be is what it pretty much already is, only under a wrong description. In the case of tort law (the common law governing civil liability for an injury not arising out of a contract), for example, Holmes argued for answer three—liability ought to be triggered by a finding of negligence—but he did so by attempting to demonstrate that negligence was already, and more or less always had been, the rough basis for tort liability. He argued, in other words, that if in our analysis of tort cases we dropped terms like “guilt” and “fault” and replaced them with terms like “carelessness” and “recklessness,” we would find that we generally got the same results. The advantage of replacing the moral language of sin with the economic language of risk was not to punish a different class of wrongdoers or a different category of wrongs. It was simply to make explicit what moral language tends to disguise, which is that (in the words of
The Common Law
): “The substance of the law at any given time pretty nearly corresponds, so far as it goes, with what is then understood to be convenient”
17
—a sentence in which, as the context makes clear, we are meant to understand “convenient” in the widest possible sense.
The problem for Holmes wasn’t, therefore, what the basis for tort liability ought to be. The problem was what the basis for deciding that a particular act was negligent ought to be. Assuming that we want to make persons who act carelessly pay the cost of cleaning up their tortious messes, how do we determine what sort of behavior counts as careless? How do we distinguish a tort from an accident, or from the permissible by-product of a socially desirable activity? One way of doing this would be to devise a series of general rules for conduct, violation of which would ipso facto constitute negligence; but this solution was obviously ruled out for Holmes by his contempt for the malleability of general rules. His alternative proposal was that we should do judicially what we all do anyway when we are confronted with a judgment call, which is to evaluate the conduct at issue by the lessons of experience. “Experience is the test,” as
he put it in
The Common Law,
“by which it is decided whether the degree of danger attending given conduct under certain known circumstances is sufficient to throw the risk upon the party pursuing it.”
18
Whose experience? The experience, Holmes said, of “an intelligent and prudent member of the community.”
19
He didn’t mean by this a
particularly
intelligent and prudent person—a judge, for instance. He meant, precisely, a person who is neither particularly prudent nor particularly imprudent, an “average member of the community”—in other words, a jury. “When men live in society,” he explained in
The Common Law
, “a certain average of conduct, a sacrifice of individual peculiarities going beyond a certain point, is necessary to the general welfare. If … a man is born hasty and awkward, is always having accidents and hurting himself or his neighbors … his slips are no less troublesome to his neighbors than if they spring from guilty neglect. His neighbors accordingly require him, at his proper peril, to come up to their standard, and the courts which they establish decline to take his personal equation into account.”
20
Putting it this way made blameworthiness, Holmes thought, into what he approvingly called “an external standard,” a standard before which the defendant’s state of mind (a legal imponderable anyway) becomes irrelevant. “A man may have as bad a heart as he chooses,” as he put it in “Trespass and Negligence,” “if his conduct is within the rules.”
21
“The reasonable man” is the phrase commonly associated with this theory of liability. Holmes didn’t coin it—it began appearing in American and English opinions around 1850—but, along with his English friend Frederick Pollock, he probably did as much as anyone to define and establish it. What makes the concept work (on Holmes’s theory) is that it represents a composite. It is a collective noun, a statistical fiction, an averaging out across the whole population. The “reasonable man” knows, because “experience” tells him, that a given behavior in a given circumstance—say, taking target practice in a populated area—carries the risk of injuring another person. Of course, any action in any circumstance carries some risk, however remote, of injuring another person; and reasonable people
know this. But this knowledge is not what reasonableness consists in. What reasonableness consists in is the knowledge of the greater or lesser
probability
of an injury being caused by such and such an action in such and such circumstances. “Even in the domain of knowledge,” as Holmes put it, “the law applies its principle of averages.”
22
By putting negligence at the center of tort liability, Holmes got accused, by some twentieth-century commentators, of making it easier for industry to escape liability for injuries, to workers or customers, incidental to its enterprise—injuries for which it would have had to pay under a theory of strict liability. But Holmes did accept a principle of strict liability for what he called “extra-hazardous” activity. Under this principle, a company that uses dynamite in the normal course of its business, for example, can be held accountable for any injuries it causes, even if it has taken reasonable precautions to avoid them. But Holmes didn’t regard strict liability as inconsistent with the concept of reasonableness, or even with the concept of negligence; for the activities society labels extra-hazardous are just activities experience has led the reasonable man to believe to be risky per se. “Negligence” is infected a little by the kind of moral coloration Holmes deprecated in legal language: it suggests a personal failing on the part of the defendant. But all Holmes meant by it was acting in the face of foreseeable risk. We may have perfectly honorable reasons for doing so, but we also have to be willing to take our legal chances.