Holmes, Langdell and Formalism


Holmes, Langdell and
Formalism

PATRICK J. KELLEY

Abstract. Both Holmes and Langdell believed that science was the model for all
human inquiry and the source of all human progress. Langdell was influenced by an
unsophisticated scientism, which led him to attempt to identify the true meaning of
legal doctrines. Holmes was influenced by the sophisticated positivism of John Stuart
Mill, which led him to attempt to reduce legal rules and doctrines to scientific laws
of antecedence and consequence, justified only by their social consequences. Both
Holmes and Langdell concluded that judges ought to decide a case by applying the
rules established by precedent, without appeal to any special claims of justice and
without appeal to any higher-order normative principle.

1. “Formalism” in Late Nineteenth Century Judicial Opinions

Karl Llewellyn’s The Common Law Tradition: Deciding Appeals is the locus
classicus for the thesis that the common-law courts in the United States changed
their opinion-writing style from an early nineteenth-century “grand style,”
involving broadly stated policy rationales, to a formal style, involving an
established rule or doctrine to be applied in mechanical fashion to the 
facts (Llewellyn 1960, 62–75). Llewellyn supported this thesis by reporting
his reactions to opinions from different decades in the New York,
Massachusetts, Pennsylvania and Ohio reports (Llewellyn 1960, 64–157).

Llewellyn’s examples and his impressionist methodology seem persua-
sive but not conclusive in support of his thesis of a general historical trend
in opinion-writing styles over the course of the nineteenth century. Two
questions arise. First, was there such a change and, if so, what kind of a
change was it? Second, what was it that influenced the judges on these
common law courts to change their style of opinion-writing?

One ought to approach these questions with a certain degree of scepticism,
for “formalist” is a derogatory term employed by early legal realists like
Llewellyn to characterize methods of legal reasoning they rejected. We do
not find any self-described legal formalists in the nineteenth century.

© Blackwell Publishers Ltd 2002, 108 Cowley Road, Oxford OX4 1JF, UK and 350 Main Street, Malden, MA 02148, USA.

Ratio Juris. Vol. 15 No. 1 March 2002 (26–51)



Nevertheless, my own studies in the history of torts and in the doctrine of
proximate cause have convinced me that there was indeed a change in the
nature of appellate court opinions in the United States during the nineteenth
century (Kelley 1990, 357–61; 1991, 57–82). Broadly described, the change
may be said to be this: Earlier in the nineteenth century, courts were more
likely to justify their decisions by reference to specific prior cases and to
general principles; later in the nineteenth century, courts were more likely 
to justify their decisions by reference to settled rules or settled doctrines.

What can account for this change? At least four developments internal to
the law pushed judicial opinions in the same direction: (1) the breakdown 
of the forms of action and the rise of substantive law, (2) the maturation of
each state’s common law as it moved beyond an early foundational period,
(3) the attempt to retain a unified common law in the face of the splintering
of the common law into many different common laws in the different states,
and (4) the influence of nineteenth-century scientism on legal thought. Each
development had separate and substantively distinguishable influences on
legal scholars, lawyers, and judges. The influences may have overlapped 
in some, but not necessarily all, cases. I will discuss these developments in an
order that roughly reflects both their relative importance and their chronology.

The Breakdown of the Forms of Action

The traditional way of categorizing and thinking about the common law—
the forms of action—broke down. This proceeded rapidly throughout the
first half of the nineteenth century in both England and the United States,
culminating in the procedural reforms abolishing the forms of action around
the middle of the nineteenth century in the New York Field Code and the
English Procedural Reform Acts.

The common law under the forms of action was basically a law of pleading:
procedural law based on formulaic pleadings. The pleading reforms abolished
the formulaic pleadings and adopted fact-based pleading. This moved the
common law at its core from a law of procedure and pleading formulas
toward a substantive law, which led to the flowering of the substantive law
treatises, using substantive categories like contract and tort, in the latter half
of the nineteenth century. This whole process is well-explained by Milson
(1981) and his successor J. H. Baker (1979, 74–82).

A. W. B. Simpson (1981, 651–68) has written thoughtfully on the rise of the
treatises.

In moving from a formulaic law of pleading to a substantive law, legal
doctrine obviously becomes more and more important. The change in the
nature of judicial opinions toward more reliance on settled doctrines 
may reflect, in part at least, this sea-change in the very nature of the
common law.

Holmes, Langdell and Formalism 27

© Blackwell Publishers Ltd 2002.



The Maturation of Each State’s Common Law

Although we often seem to ignore the fact when we talk about “American
Legal History,” what we have is a collection of separate histories of the com-
mon law in each of the states. The common law in each state can be said to
have started with a foundational period, in which the common law courts of
that state did two things. First, under the reception statutes, the courts sorted
out which parts of the English law they would adopt and which they would
reject. Second, they recognized basic general principles to guide courts in
subsequent cases. During this period, the common law in that state could 
be said to be more open-textured and less certain than it would be later on,
when the legal system in that state became more settled. It would make
sense that judges in the foundational period would write opinions that were
primarily based on general principles, with the explicit or implicit under-
standing that they were making basic choices for a relatively open-textured
legal system.

The Splintering of the Common Law

Toward the middle of the nineteenth century, it became obvious that each
state was developing its own peculiar common law, and that the unity and
coherence of “the common law” was thereby threatened. In practical terms,
this became a problem for lawyers who might want to practice in more than
one state and for law teachers and law schools who professed to be able to
prepare law students from many different states for law practice in the states
from which they had come. More importantly, this was a problem for those
supporting the nationalist ideal of a united people, sharing common political
commitments, a common language, and a common law.

These problems led to a drive to identify and retain a core common law
applicable for the most part in all the states. One way to promote consistency
in the common law over a number of states is to focus primarily on doctrine
and avoid arguably peripheral arguments about broader normative justifica-
tions. The drive toward a single common law gave a special twist and a
special urgency to the treatise movement in the United States. This drive was
also expressed in the collections of “Leading Cases,” later replaced by American
Law Reports (ALR), and the various “Restatements” of the law.

The Rise of Scientism

From the middle to the end of the nineteenth century, most educated,
sophisticated people were enamored with one form or another of scientism
—the belief that science was the model for all human inquiry and the source
of all human progress (Sorell 1991). This general trend affected late nineteenth-
century writers on the common law.

28 Patrick J. Kelley

© Blackwell Publishers Ltd 2002.



The effects of scientism on legal thought can be explored most fruitfully in
the work of Christopher Columbus Langdell and Oliver Wendell Holmes. In
the rest of this paper, I will argue that both were greatly influenced by the
prevailing scientism: Langdell by a crude popular understanding of science
and Holmes by the highly sophisticated positivism of Auguste Comte. I will
further argue that the effect of scientism on both Holmes and Langdell was
to emphasize doctrine and to exclude the broader normative principles that
common law courts had traditionally invoked to invent, develop, and limit
doctrine.

2. The Old Story about Holmes and Langdell

You have probably heard the old story about the relationship between Oliver
Wendell Holmes, Jr. and Christopher Columbus Langdell. In the story,
Langdell was the prime example and most enthusiastic cheerleader for the
legal formalism that infected American judges and lawyers in the latter half
of the nineteenth century. The formalists banished policy and principle from
the law and decided cases by applying technical legal doctrine with syllogistic
logic to the facts. In the story, Holmes was the arch-foe of legal formalism,
emphasizing the necessary role of public policy in judicial decisionmaking
and insisting in his celebrated quote that “the life of the law has not been
logic; it has been experience” (Holmes 1963, 5).

The most compelling evidence for the truth of the old story is the history
of that very quote. The line became famous as part of Holmes’s general
critique of legal formalism in the first paragraph of The Common Law but it
began life earlier as part of Holmes’s critical review (Holmes 1880, 234) of
Langdell’s methodology in the second edition of Langdell’s Cases on the Law
of Contracts (Langdell 1879). The conclusion seems obvious: Holmes developed
his devastating critique of legal formalism by generalizing his criticism of
Langdell’s methodology. That conclusion in turn seems to confirm the story,
which pits Holmes the chief anti-formalist against Langdell the archetypal
legal formalist.

The story is simple and powerful. It has a hero in Holmes and a villain 
in Langdell. To be sure, since it’s a modern story, the villain is not so much
evil as benighted. It has a happy ending: Legal formalism is vanquished and
clear-eyed realism prevails. Finally, it has a moral: We should all avoid the
sterile formalism of Langdell and his ilk.

But is the story true?

3. Holmes on Langdell

If we look at other things Holmes wrote about Langdell, the story grows
more complicated.

Holmes, Langdell and Formalism 29

© Blackwell Publishers Ltd 2002.



As an editor of the American Law Review, Holmes published highly favor-
able reviews (1871, 1872) for each of the two volumes of the first edition of
Langdell’s pathbreaking casebook on contracts, as they came out. Comment-
ing on the first volume, Holmes (1871, 540) advised “every student of the
law to buy and study the book,” and praised the organization of the case-
book, which arranged the cases first according to significant legal doctrines
and second, for each doctrine, presenting the leading cases developing that
doctrine chronologically and without comment. Holmes (1871, 540) noted
that “[t]racing the growth of a doctrine in this way not only fixes it in the
mind, but shows its meaning, extent, and limits as nothing else can.” Comment-
ing on the second volume, Holmes (1872, 353) again praised the organization
of the casebook, focusing this time on the index, which referred to the cases
under “the general principles of the law of contracts” instead of the subject
matter of the contract (e.g., coal), or the popular name of the particular kind
of contract (e.g., charter-party or insurance).

The way Langdell organized his casebook was about the only thing to
react to, for Langdell provided no commentary on the cases or the principles
they stood for. But the analytical organization of the law was a peculiarly
important issue in the 1870s, when the law was stuck in the immediate
aftermath of the collapse of the forms of action as organizing categories. And
Holmes in the early 1870s was primarily concerned with that issue. In his
first major article, Holmes (1870) had set himself the task of providing a
thorough philosophical arrangement of the whole body of law. Holmes
evidently saw as complementary to his project both Langdell’s overall arrange-
ment of the law of contracts in terms of legal doctrines and legal principles
as subcategories, and Langdell’s chronological organization of leading cases
within each subcategory. Holmes and Langdell evidently stood united against
the mindless attempts in the currently popular practitioner treatises to
organize the law around the different kinds of situations to which the law
could be applied or, rather less mindlessly, around the different kinds of
contracts as popularly understood.

In the second edition of his casebook on contracts, Langdell (1879) expanded
his index to include short essays setting forth his analysis of the relevant
legal doctrines. There were thus separate essays on Offer, Acceptance, and
Consideration, among other doctrines. This essay-index, later published sep-
arately as A Summary of the Law of Contracts (Langdell 1880), revealed for the
first time Langdell’s substantive theories. In his review of the essay-index,
Holmes (1880, 234) criticized Langdell’s purely logical approach to the deter-
mination and application of legal doctrine. In the concluding sentence of that
criticism, Holmes suggested that Langdell’s purely logical methodology was
unscientific:

As a branch of anthropology, law is an object of science; the theory of legislation is a
scientific study; but the effort to reduce the concrete details of an existing system to

30 Patrick J. Kelley

© Blackwell Publishers Ltd 2002.



the merely logical consequences of simple postulates is always in danger of becom-
ing unscientific, and of leading to a misapprehension of the nature of the problem
and the data. (Holmes 1880, 234)

Holmes’s review of Landgell’s second edition was not, however, uniformly
negative. Holmes’s vigorous critique of Langdell’s methodology was sand-
wiched between almost equally vigorous praise both for Langdell’s power-
ful doctrinal analysis and for Langdell’s pedagogically helpful organization
of contract doctrines. Thus, at the start, Holmes said:

No man competent to judge can read a page of [Langdell’s Appendix] without at
once recognizing the hand of a great master. Every line is compact of ingenious and
original thought … It may be said without exaggeration that there cannot be found
in the legal literature of this country, such a tour de force of patient and profound intel-
lect working out original theory through a mass of detail, and evolving consistency
out of what seemed a chaos of conflicting atoms. (Holmes 1880, 233–4)

And, in closing his discussion, Holmes said:

[T]he book is published for use at a law school, and … for that purpose dogmatic
teaching is a necessity … A professor must start with a system as an arbitrary fact,
and the most which can be hoped for is to make the student see how it hangs
together, and thus to send him into practice with something more than a rag-bag of
details. For this purpose it is believed that Mr. Langdell’s teachings, published and
unpublished, have been of unequaled value. (Holmes 1880, 234)

The picture of Holmes’s relationship to Langdell that we get from this book
review is more subtle than the old story would have it. True, Holmes
attacked Langdell’s purely logical methodology as unscientific, but Holmes
also praised the originality and power of Langdell’s doctrinal analysis, com-
paring Langdell favorably to the more conventional, less original treatise on
contracts just published by Sir William Anson (Holmes 1880, 233). Holmes
(1880, 234) also praised Langdell for providing his students with a useful
guide to the doctrinal structure of the law of contracts. In light of his earlier
enthusiastic reviews of Langdell’s work, this review reads like an attempt 
to correct Langdell’s methodological errors and thereby possibly reclaim a
former ally.

The same curious mixture of admiration and rejection is evident in
Holmes’s comments (1946) on Langdell in a letter to Frederick Pollock dated
April 10, 1881. Holmes had mailed a copy of The Common Law to Pollock 
a little over a month before and had, in the interim, received a letter, now
missing, from Pollock. Evidently Pollock had not commented on The
Common Law in his letter, probably because he had not yet received it. So in
the second paragraph of his letter Holmes went on and on about his book,
an effusion evidently triggered by Pollock’s discussion of certain points 
of contract doctrine either in Pollock’s missing letter or the advance sheets

Holmes, Langdell and Formalism 31

© Blackwell Publishers Ltd 2002.



Pollock had sent earlier. Holmes (1946, 17) said he had discussed the
question of bilateral contract accepted by letter in his Eighth Lecture in 
The Common Law, as well as “one or two other things of which you speak.”
After briefly summarizing his position on acceptance by letter, Holmes went
on, without breaking for a new paragraph, to urge Pollock to read Langdell.

Here is Holmes (1946, 17) on Langdell to Pollock: “I should like you to see
the Appendix to the 2nd ed. of Langdell’s Cases, also published separately in
a small book called (I think) Elements of Contract. A more misspent piece of
marvelous ingenuity I never read, yet it is most suggestive and instructive.”
Holmes (1946, 17) went on in the letter to criticize Langdell’s excessive devo-
tion to logic and to note that Langdell’s explanations and reconciliations of
the cases “would have astonished the judges who decided them.” But what
is truly astonishing is that, in the middle of a discussion of Holmes’s and
Pollock’s writing on contracts, Holmes urged Pollock to read Langdell’s
essay-index, which he condemns as a “misspent piece of marvellous ingen-
uity,” yet praises as “most suggestive and instructive.”

4. Holmes’s Lectures on Contract in The Common Law

Holmes’s letter to Pollock and his curious recommendation that Pollock read
Langdell points us back to Holmes’s three lectures on contract (Holmes 1963,
195–264) in The Common Law. A careful analysis of those lectures supports
another conclusion we might not have expected: The primary target of
Holmes’s contract lectures was not Langdell’s doctrinalism but Frederick
Pollock’s attempt (1876) to explain the common law in terms of Savigny’s
will theory of contract (1848) and Henry Sumner Maine’s (1970, 295–328)
theory that the law of contract has evolved from recognizing only formal or
objective contracts to recognizing consensual or subjective contracts.

Each of Holmes’s three lectures on contract attacked one of Pollock’s
central theoretical claims and each lecture developed an alternative theory
radically at odds with Pollock’s.

Thus, Holmes’s first lecture (1963, 194–226), on the history of the common
law of contracts, was a careful, brilliantly argued brief against Pollock’s ap-
plication of Maine’s evolutionary thesis to the common law of contract. The
primary focus of the lecture was a detailed history of the earliest appearance
and subsequent development of consideration in the common law, tending
to show that the requirement of consideration had developed internally and
was not borrowed from the Roman law causa. This was all a direct attack on
Pollock’s speculation (1876, 147–53), based on Maine’s evolutionary theory,
that the common-law doctrine of consideration ultimately derived from 
the Roman law causa, which entered English law first in the equity courts,
which were the first to start enforcing informal, consensual contracts. Pollock
had carefully labeled his theory as speculation based on what you would
expect if Maine’s theory were true (Pollock 1876, 152). Pollock’s theory was,

32 Patrick J. Kelley

© Blackwell Publishers Ltd 2002.



therefore, on its own terms, a test case to determine whether Maine’s evolution-
ary theory applied to the common law.

Holmes’s second lecture (1963, 227–40) developed a descriptive theory 
of the elements of contract in deliberate opposition to Pollock’s theoretical
explanation of the formation of a contract based on Savigny’s mutual common-
intention version of the will theory of contracts. Pollock’s sophisticated
discussion of the requisites for a valid contract (1876, 1–8) began with the
generally accepted doctrine that a legal agreement formed by a proposal 
and an acceptance was necessary for a valid contract and went on to make
explicit the seemingly undeniable connection between offer and acceptance
and the mutual common-intention will theory expounded by Savigny. In 
the teeth of case law adopting offer and acceptance as requisites for a valid
contract and in the teeth of Pollock’s persuasive argument that offer and
acceptance embodied the mutual common-intention form of the will theory,
Holmes, in his analysis of the elements of contract, dispensed with offer and
acceptance altogether. Instead of offer, acceptance, and consideration as the
elements, Holmes has promise and consideration. Although Holmes (1963,
235) defines promise as an “accepted assurance that [an] event … shall come
to pass,” it turns out that the assurance is accepted not by the expression of
consent but simply by the act of providing the consideration called for by the
promisor. Moreover, the test of consideration is also purely external. Con-
sideration is what by the terms of the agreement is given and accepted as the
motive or inducement of the promise (Holmes 1963, 230).

Holmes’s third lecture (1963, 241–64), on void and voidable contracts, was
a subtle, somewhat roundabout response to the entire last half of Pollock’s
treatise (1876, 356), in which Pollock discussed the doctrines of mutual mis-
take, fraud, misrepresentation, coercion, and undue influence as “subjective
conditions” for a valid contract (Pollock 1876, 355). Pollock (1876, 356) had
again invoked the will theory of contract: If a party’s consent to an agree-
ment is not true, full, and free, the contract is not binding on that party.
Coercion or undue influence means the consent was not free; mistake,
misrepresentation, or fraud means that consent given in ignorance of a
material fact was not true or full. Pollock’s thorough, sophisticated analysis
seemed to make of these topics a stronghold for the mutual common
intention theory.

Holmes’s third lecture on contract can be seen as an ingenious response to
the challenge posed by Pollock’s analysis of these “subjective condition”
doctrines. Holmes shifted the initial focus away from the individual doc-
trines. He focused instead on the judicial remedies and categorized Pollock’s
“subjective condition” cases as void contracts or voidable contracts. He
attempted to show that a contract was void only when one of the objective
elements of a contract was missing (Holmes 1963, 241). He attempted to show
that a contract was voidable only for failure of an express or an objectively
implied condition in the contract (Holmes 1963, 246).

Holmes, Langdell and Formalism 33

© Blackwell Publishers Ltd 2002.



Throughout his discussion of contract, Holmes persistently attacked Pollock’s
three major theoretical positions and turned those attacks into critically
important structural elements in each of his three lectures. By contrast, when
Holmes attacked Langdell’s positions on acceptance by mail (Holmes 1963,
239–40), conditions precedent and subsequent (Holmes 1963, 247–8), and
dependent promises (Holmes 1963, 260–64), those attacks were not central to
Holmes’s theoretical arguments. In each instance, Holmes used Langdell’s
analysis simply as a starting point against which he developed his own
analysis of these subsidiary issues.

Moreover, there are notable similarities between some of Langdell’s other
discussions and Holmes’s treatments of the same topics. These similarities
suggest that Holmes borrowed without attribution the doctrinal insights of
Langdell that were consistent with Holmes’s positivism. There are at least
four of these similarities. First, Langdell (1880, 129–30) questioned the ap-
plicability of Maine’s category of consensual contract to common-law con-
tracts whose breach is redressible in assumpsit, for in each case there must
be consideration for the contract to be enforceable and consideration is a
formal requirement. Holmes (1963, 215) said that “consideration was a form
as much as the seal.” Second, Langdell (1880, 2) pointed out that acceptance
as an independent element of an enforceable contract is not really important
in either unilateral or bilateral contracts, for those contracts are not enforce-
able until the consideration called for by the offer is provided, and providing
the consideration in and of itself supplies sufficient evidence of acceptance.
Holmes (1963, 238) left out acceptance as an essential element of contract
because giving consideration counted as evidence of acceptance, and accept-
ance without consideration did not make a contract. Third, Langdell (1880,
234–44) relegated the will theory of contract formation to a legal fiction and
insisted that the contract was made not by the subjective wills of the parties
but by their physical acts. Holmes (1963, 240) insisted that contracts were
made by the overt acts of the parties, not their subjective intent. Fourth, in
his discussion of consideration, Langdell (1880, 82) argued that detriment 
to the promisee was sufficient, even without a benefit to the promisor; he
argued (1880, 82) that the consideration must be the inducement for the
promise only “in contemplation of law,” and need not be the inducement in
fact; and he argued (1880, 83) that the initial offer conclusively designates the
consideration that was called for. Holmes (1963, 230) defined consideration
as what is designated by the parties as the conventional inducement for the
promise.

Altogether, these similarities suggest that Holmes may have borrowed
heavily from Langdell on these four issues. Holmes evidently spoke from 
his own experience when he wrote to Pollock that Langdell’s summary was
“most suggestive and instructive.”

34 Patrick J. Kelley

© Blackwell Publishers Ltd 2002.



5. Holmes, Langdell and Legal Science: A Thesis

When we take into account Holmes’s other writings relevant to Langdell,
then Holmes’s relationship to Langdell seems very complicated indeed.
How are we to understand that relationship? My thesis is this. The key to
understanding the relationship between Holmes and Langdell lies in their
different understandings of legal science, which nevertheless lead them to
similar prescriptions for judicial decisionmaking.

Although Langdell was fifteen years older than Holmes, they both began
publishing for a legal readership around 1870. That was a time when one
form or another of scientism was all the rage. Darwin’s Origin of Species was
published in 1859 and almost immediately exerted a significant influence.
Modern developments, such as railroads, textile mills, and steamships, all
attributable to scientific advances, seemed to be transforming everyday life.
The study of man and his institutions, long thought to be the province of
philosophers, moralists, and political theorists, was invaded by those march-
ing under the banner of science: anthropologists, sociologists, and economists.
Not surprisingly, serious writers about the law claimed the mantle of science
as well. For example, Frederick Pollock, in his treatise on contract (1876, vii),
and Christopher Columbus Langdell, in his pioneering Casebook on Contracts
(1871, vi–vii), both characterized their work as scientific. Given the popularity
of science in general, one might expect that authors with widely differing
aims and methodologies would all claim to be doing legal science.

6. Langdell and Unsophisticated Scientism

Christopher C. Langdell invented the case study method of teaching law and
applied it to his teaching of contracts. The case method was based on
Langdell’s belief that “law is a science, and … all the available materials of
that science are contained in printed books” (Langdell 1887, 124). Students
could learn the law of contracts scientifically, by studying original sources:
decided cases that Langdell had selected for his pioneering casebook in
contracts. What did Langdell mean when he said that “law is a science”? The
answer is not completely clear. Over the last twenty-five years, we have seen
vigorous and sophisticated academic debate over Langdell’s legal science.1

Holmes, Langdell and Formalism 35

© Blackwell Publishers Ltd 2002.

1 See: Gilmore 1974, 12–14; 1977, 42–8 (Langdell’s science a simple-minded formalism); Speziale
1980 (contra Gilmore: Langdell’s science empirical, dedicated to formulating working hypo-
theses from analysis of original sources); Grey 1983 (contra Gilmore and Speziale: Langdell’s
science a curious combination of induction from the cases of principles then used to deduce
subsequent conclusions, analogous to John Stuart Mill’s inductive explanation of geometry);
Reimann 1992 (Langdell’s science similar to but in important ways different from Kantian
formalism of the German pandectists, so Langdell halfway between Kantian formalism and 
a rigorously empirical, inductive, historicist legal science); LaPiana 1994, 55–78 (Langdell’s
science an amalgam of technical legal analysis of cases, a coherent classification of the law 
based on fundamental but technical substantive doctrines, and a loosely understood Austinian
positivism separating law and morality).



To answer the question, we must look carefully both at what he said and at
what he did.

Langdell said precious little about what he meant when he said “law is a
science.” He wrote about it first in the Introduction to the first edition of 
his Casebook on Contracts (Langdell 1870, vi–vii). He talked about it again in
his speech celebrating the 250th anniversary of the founding of Harvard
University (Langdell 1887, 123–4). From these two sources, we can identify
the essential ideas that Langdell thought were entailed by his notion that 
law was a science. First, law must be studied scientifically, by studying 
the original sources—the decided cases. Second, law as a science consists of
certain fundamental doctrines, which are relatively few in number. Third,
those fundamental doctrines grew to their current form slowly, and that
growth can be traced through a series of cases. Fourth, an appropriate classi-
fication and arrangement of those fundamental legal doctrines would facilitate
the scientific study of the law.

A private letter from Langdell to Theodore Dwight Woolsey of Yale,
recently discovered by William LaPiana in Yale’s manuscript collection
(LaPiana 1994, 77), adds an additional element to Langdell’s notion of law 
as science. Writing about the study of jurisprudence, Langdell expressed 
his understanding of a distinct line between the study of law as it is and the
study of law as it ought to be. He concluded that lawyers and law professors
ought only to study the law as it is. Based on the views expressed in 
this letter, Langdell’s notion of law as a science probably included John
Austin’s rigid distinction between law and morality.

Since Langdell said so little about what he meant by law as a science, we
are forced to look at the details of what Langdell actually did in his Summary
of the Law of Contracts (1880) in order to understand what he meant. In the
Summary, we can see Langdell at work as a legal scientist. This work followed
a consistent pattern. By examining decided cases, Langdell isolated the funda-
mental doctrines applied by the courts in contract cases, including the basic
requirements of offer, acceptance, and consideration. By careful, critical analysis,
Langdell then penetrated to the true meaning of each doctrine. For example,
that method led him to the following statement of the true meaning of “offer”:

An offer, as an element of a contract, is a proposal to make a promise. It must be made
by the person who is to make the promise, and it must be made to the person to
whom the promise is to be made. It may be made either by words or by signs, either
orally or in writing, and either personally or by a messenger; but in whatever way it
is made, it is not in law an offer until it comes to the knowledge of the person to
whom it is made. (Langdell 1880, 12)

He then used those meanings to elaborate a coherent, logically consistent
blueprint for the law—here, the law of contract.

Langdell thus seemed to work with the dispassionate objectivity of a
scientist: Isolating the true meaning of the doctrinal terms was a matter of

36 Patrick J. Kelley

© Blackwell Publishers Ltd 2002.



critical analysis; coherently applying those meanings over a broad range of
questions called for analytical discrimination and logical consistency; the
resulting doctrinal conclusions were reached without injecting any nor-
mative views into the analysis. In fact, because the fundamental doctrines
themselves were stated conceptually, without reference to any moral or policy
base, and the application of those doctrines was determined by criteria of
logical coherence, the resulting purified doctrinal structure seemed purely
conceptual. It thus offered a self-proclaimed scientist like Langdell the oppor-
tunity to be completely objective. And the process of deciding legal issues
under Langdell’s theory could then be purely analytical.

For example, Langdell recognized that in bilateral contracts the consider-
ation for the first promise is a counter-promise by the offeree. From this,
Langdell argued purely analytically that notice of acceptance must actually
be communicated to the original offerer for a valid bilateral contract:

When the contract is to be bilateral, … the offer [still] requires an acceptance and the
giving of the consideration to convert it into a binding promise; but as the consider-
ation consists of a counter-promise, so the giving of the consideration consists in
making this counter-promise. (Langdell 1880, 12)

[T]he acceptance of the original offer, in the case of a bilateral contract, must be
expressed, i.e., must be made by words or signs; … the reason for this is, that the
acceptance contains a counter-offer. Moreover, the reason why the counter-offer
makes it necessary that the acceptance should be expressed is, that communication to
the offeree is of the essence of every offer. The acceptance, therefore, must be com-
municated to the original offerer, and until such communication the contract is not
made. (Langdell 1880, 15)

Langdell’s legal science, based on the objective meaning of legal doctrines,
was radically different from traditional legal reasoning, which focused on
the normative rationale for prior decisions. Appeals to higher-order normative
principles seemed to be banished from Langdell’s legal science altogether.
For example, in defending his position that the effective date of a bilateral con-
tract by mail was the date the letter of acceptance comes to the knowledge of
the offerer, not the date it was mailed, Langdell recognized that some have
argued that “the purposes of substantial justice, and the interests of the con-
tracting parties as understood by themselves,” (Langdell 1880, 20) would be
served by the date-of-mailing rule. Langdell responded: “The true answer to
this argument is, that it is irrelevant”2 (Langdell 1880, 21).

Holmes, Langdell and Formalism 37

© Blackwell Publishers Ltd 2002.

2 “Assuming it to be relevant,” Langdell went on to make three arguments for his position,
based on the understanding of the parties, substantial justice, and the need for certainty about
legal obligations. Langdell’s argument is a model of brevity: “The only cases of real hardship
are where there is a miscarriage of the letter of acceptance, and in those cases a hardship to one
of the parties is inevitable. Adopting one view, the hardship consists in making one liable on a
contract which he is ignorant of having made; adopting the other view, it consists in depriving
one of the benefit of a contract which he supposes he has made. Between these two evils the
choice would seem to be clear: the former is positive, the latter merely negative; the former 



Langdell didn’t pause to explain why arguments from substantial justice
and the self-understood interests of the parties are irrelevant, but the structure
of his analysis here and elsewhere in the Summary suggests the following
two reasons.

The first reason is this. Langdell had previously established, by critical,
scientific analysis of the legal meaning of acceptance in a bilateral contract,
that acceptance necessarily contains a counter-offer. He had further estab-
lished, by his critical, scientific analysis of the legal meaning of offer, that an
offer must actually be communicated, as “communication to the offeree is
the essence of every offer” (Langdell 1880, 15). It follows necessarily, then,
that the acceptance must actually be communicated to make a bilateral contract.
Indeed, according to Langdell, this is so fundamental that the original
offeror could not change it by the terms of his offer. If he said the offer would
be accepted at the time of posting the acceptance, that “declaration would be
wholly inoperative” (Langdell 1880, 20).

The second reason is this. Judicial decision based on the objectively dis-
coverable meaning of legal doctrines is, for Langdell, the only appropriate
way to decide these questions. As Langdell showed, equally persuasive
arguments based on “substantial justice and the self-understood interests of
the parties” can be made for the opposite position, so there is no objective,
disinterested way of using these kinds of arguments to choose between the
opposing positions. A judge who accepted one argument over another, then,
would simply either be mistaken that one argument is stronger than the
other or would be choosing arbitrarily or for some undisclosed reason
between two positions equally supportable by “substantial justice” argu-
ments. Langdell’s concern for scientific objectivity, therefore, seemed to
support a legal analysis that stopped at the level of legal doctrine.

We can now better understand Langdell’s notion that law is a science. The
legal scientist focuses on the decided cases. The scientist extracts, by the
ordinary techniques of legal analysis, the current fundamental legal doc-
trines in any particular field of law. The scientist makes no judgments about
what the law ought to be, so he separates the fundamental legal doctrines
from the normative justifications given by judges and others, and “purifies”
those doctrines by technical, objective analysis of their true meaning. This
purification is simply a rational determination of what the law is. The sci-
entist may then recommend that future cases be decided by logical applica-
tion of the purified doctrines because that will simply be an application of
the law as it is.

38 Patrick J. Kelley

© Blackwell Publishers Ltd 2002.

imposes a liability to which no limit can be placed, the latter leaves everything in statu quo. [See
Vassar v. Camp, 1 Kern. 441, Cas. on Contr. 110, and compare par. 14.] As to making provision
for the contingency of the miscarriage of a letter, this is easy for the person who sends it, while
it is practically impossible for the person to whom it is sent. [See Br. & Am. Tel. Co. v. Colson,
L.R. 6 Exch. 108, 118, Cas. on Contr. 45, 51, per Bramwell, B.]” (Langdell 1880, 21).



It is tempting to criticize Langdell’s legal science as unscientific or inco-
herent or both. A number of commentators (Phelps 1892; Tiedeman 1892;
Gilmore 1974, 1977; Grey 1983), in addition to Holmes, have succumbed to
the temptation. Langdell’s legal science seems incoherent because it is in-
ternally inconsistent. The historical contingency of current rules and doctrines
recognized by Langdell’s choice of cases illustrating the development of
current case law would be erased by courts deciding cases the way Langdell
proposed they do, because continued logical application of the true meaning
of current doctrine would freeze the law at that stage in its development.
Langdell’s methodology seems unscientific for several reasons. Langdell did
not include in his base for induction all the decided cases on a particular
topic. The subsequent induction of the true rule or the true meaning of a
legal doctrine was not a scientific induction at all. It just reflected Langdell’s
preconceived notions, which led him to include some cases in the base and
to exclude others. Moreover, Langdell on some questions, such as the effect-
ive date of acceptance by mail, included cases reaching diametrically differ-
ent results. One cannot scientifically derive by induction a single rule from
diametrically opposed cases. Even within the set of cases arbitrarily chosen,
what Langdell purported to determine scientifically from the cases was not
a scientific law, or a generalized description of common characteristics of the
set, but a concept—the true meaning of a legal doctrine. Because the purportedly
scientifically-determined concept is not itself normative, Langdell did not
purport to derive an “ought” from an “is.” But if it was not normative, there 
was then no apparent reason to apply the concept to decide subsequent cases.

Langdell included in the domain of cases for his scientific analysis only
those illustrating or developing a particular legal doctrine; he excluded cases
that recognize that the doctrine is defeasible when an overriding normative
principle applicable to the facts suggests that the ordinary application of 
the legal concept would be unjust. Langdell thus set up a non-normative
conceptual system of law radically at odds with the underlying phenomena.
This points to the most fundamental problem with Langdell’s methodology.
It was inconsistent with the way common law judges decided cases, then
and now. Judges do not seek to discover the true meaning of the legal terms
prior judges used in explaining their decisions in prior cases; judges do not,
then, try to decide cases before them by applying those legal meanings to
achieve a purified doctrinal consistency over a range of cases, regardless of
the specific claims of justice by the parties in an individual case. Instead,
legal doctrines are created, controlled, and limited by judicial recognition of
a set of broader normative principles, each of which may have different weight
in different factual situations. None but the most narrowly legalistic judge,
insensitive to concerns to do justice in the case before him, would even
consider deciding a case simply by applying the legal doctrine without con-
sidering whether, in light of broader normative principles, that application
would lead to a just result on those facts.

Holmes, Langdell and Formalism 39

© Blackwell Publishers Ltd 2002.



At first glance, this last critique seems to be what Holmes, Langdell’s first
critic, wrote in his review (Holmes 1880) of Langdell’s essay-index. Holmes
addressed his criticism “to the ideal of the final methods of legal reasoning
which [Langdell’s] Summary seems to disclose.” Holmes characterized
Langdell’s ideal in the law as the logical integrity of the legal system. In pur-
suing this ideal, Holmes noted (1880, 234), Langdell seemed “less concerned
with his postulates than that the conclusions from them hang together.” To
support this criticism, Holmes quoted Langdell’s dismissing as irrelevant an
argument from substantial justice and the self-understood interests of the
contracting parties, discussed above. Holmes went on to argue (1880, 234) that
logical consistency in the law is less important in shaping the substance of
the law than other influences: “experience,” “human feelings,” “the justice and
reasonableness of a decision,” and “history and the nature of human needs.”

On its face, then, Holmes’s critique of Langdell’s methodology seems to
endorse traditional methods of legal reasoning from higher-order normative
principles. That conclusion might be premature, as the following discussion
of Holmes’s scientific positivism and his theory of judicial decisionmaking
suggests.

7. Holmes and Sophisticated, Scientific Positivism

Although Holmes rarely labeled his work “scientific,” a critical analysis of
his theories, read against the background of the works he had read, suggests
strongly that Holmes’s aim and ultimate accomplishment was to apply the
scientific positivism of August Comte and John Stuart Mill to the common law.

Neither in The Common Law nor in any of his other articles or speeches does
Holmes admit an allegiance to any particular school or tradition. He scrupu-
lously followed the advice (Holmes 1953, 20–21) he gave to others not to
attach a “fighting tag” to one’s work. We do, however, have one invaluable
aid in placing Holmes in the appropriate philosophical context. From 1865 until
his death in 1935, Holmes kept a list of books he had read, by year and some-
times by month. These book lists survived, and the list for the period from
1865 through 1880 has been published (Little 1954). We have the opportunity,
then, to read Holmes’s work against the background of the books he himself
had read. By comparing Holmes’s writing with Holmes’s reading, we may be
able to trace more precisely the specific influences on Holmes’s thought.

Holmes was a prodigious reader; the search for specific influences in the
mass of books on his reading list seems a daunting task. Fortunately, three
separate indicators suggest that the works of John Stuart Mill appearing 
on Holmes’s reading list were particularly important to his intellectual
development. First, Holmes’s sympathetic and knowledgeable biographer,
Mark DeWolfe Howe (1957, 210, 212–17), singled out the period from 1865 
to 1867 as critically important for Holmes’s intellectual development, and
emphasized the strong influence of John Stuart Mill’s work on Holmes in

40 Patrick J. Kelley

© Blackwell Publishers Ltd 2002.



that period. Second, significant events in Holmes’s life point in this direction
as well. In the summer of 1866 Holmes traveled to England, where he met
John Stuart Mill himself (Howe 1957, 223–44). Howe (1957, 208) characterizes
this trip as “the pilgrimage of a maturing mind which had already found its
tendencies.” Third, the number and weight of Mill’s books appearing on
Holmes’s reading list suggest their importance: From 1865 to 1867 Holmes
read seven works by John Stuart Mill,3 including his two formidable tech-
nical works in philosophy, A System of Logic and An Examination of William
Hamilton’s Philosophy.

The works of Mill that Holmes read all reflect Mill’s commitment to cer-
tain basic tenets of Auguste Comte’s positivism (Simon 1963, 172–201). Mill
made clear the extent of his agreement with Comte in his 1865 essay Auguste
Comte and Positivism (Mill 1969), which Holmes read in 1865 or 1866 (Little
1954, 171). Mill explicitly approved Comte’s evolutionary theory of the three
stages of human thought and Comte’s views on the limitations on human
knowledge. A brief review of those basic Comtean positions, as explained by
Mill, may thus be helpful.

Comte claimed to have discovered an invariable law of three successive
stages in the evolution of human thought about phenomena. In the theo-
logical mode of thought, phenomena are attributed to the will(s) of living
beings, either natural or supernatural. In the metaphysical mode of thought,
phenomena are explained by abstract metaphysical entities, such as the
“nature” or “efficient cause” of a thing. In the final, positive mode of thought,
the futile search for the essential nature and ultimate cause of an event 
is given up, and phenomena are explained by their relationships to other
phenomena. As Mill explained:

We know not the essence, nor the real mode of production, of any fact, but only its
relation to other facts in the way of succession or of similitude. These relations are
constant; that is, always the same in the same circumstances. The constant resem-
blances which link phænomena together, and the constant sequences which unite
them as antecedent and consequent, are termed their laws. The laws of phænomena
are all we know respecting them. (Mill 1969, 265)

These laws of phenomena are all men have ever wanted or needed to know,
however, since “the knowledge which mankind, even in the earliest ages, chiefly

Holmes, Langdell and Formalism 41

© Blackwell Publishers Ltd 2002.

3 The seven works are Considerations on Representative Government; Examination of Sir William
Hamilton’s Philosophy; Utilitarianism; Auguste Comte and Positivism; A System of Logic; Principles 
of Political Economy; and Dissertations and Discussions: Political Philosophical and Historical: See
Little 1954, 169, 171, 173–4. In addition, Holmes read extensively in the then current critical
commentary on Mill, including David Masson, Recent British Philosophy; John Henry Bridges,
The Unity of Comte’s Life and Doctrine: A Reply to Strictures on Comte’s Later Writings Addressed to
J.S. Mill; James McCosh, An Examination of Mr. J.S. Mill’s Philosophy; Herbert Spencer, The Test
of Truth [1865, Fortnightly Review 1: 531ff.] (discussion of the debate between Hamilton and Mill)
(Little 1954, 171–4). Holmes’s reading of Mill was accompanied by extensive reading of other
English and American positivists, including George Henry Lewes’s Biographical History of
Philosophy and Aristotle, and Chauncey Wright’s devastating review of Herbert Spencer’s work:
“The Philosophy of Herbert Spencer” [1865, North American Review 100: 423ff.] (Little, 1954, 170).



pursued, being that which they most needed, was foreknowledge … When they
sought for the [metaphysical] cause, it was mainly in order to control the effect, or if
it was uncontrollable, to foreknow and adapt their conduct to it. Now, all foresight of
phænomena, and power over them, depends on knowledge of their sequences ….”
(Mill 1969, 292–4)

The heavy emphasis on positivism in Holmes’s reading during his formative
years yields the following hypothesis: In his own later work in legal theory,
Holmes attempted to apply the positivism of Mill and Comte to the law.
Reading Holmes’s work in light of Holmes’s early reading seems to confirm
that hypothesis.

A number of elements in Holmes’s famous theory in The Common Law
(Holmes 1963, 34–129) that tort and criminal law liability standards evolved
from subjective to objective standards suggest that it was strongly influ-
enced by the positivist philosophy of Comte and Mill. Holmes adhered to
two commitments in this evolutionary theory that seem to be derived from
the positivist insistence on the significance of foreknowledge, based on know-
ledge of the scientific laws of antecedence and consequence. These commit-
ments were to the external purpose of the law, which operates by the threat
of sanctions, giving men incentives to avoid certain acts (Holmes 1963, 42,
88); and to the consequent need for the law to be knowable if it is to be
effective in achieving its external purpose (Holmes 1963, 88–90).

In addition, Holmes seemed to subscribe to and apply the positivist theory
that human thought always progresses through three stages. In his second
“Primitive Notions” article (Holmes 1877, 654) preceding The Common Law,
Holmes had explicitly characterized the evolution from early liability rules,
based on revenge against the causing agent, to later liability rules, based on
the attenuated fault of the owner of the causing agent, as a progression
from Theological to Metaphysical thinking. Holmes did not explicitly charac-
terize the three-stage evolution in liability standards sketched in The Common
Law as a progression from Theological to Metaphysical to Positive, but the
structure of the theory tracks with these three stages. In the first stage,
liability standards reflect a primitive desire for revenge against the causing
agent, based on its ascribed evil will (Holmes 1963, 34–5). In the second
stage, liability standards are based on an imaginary thing—the moral
blameworthiness of the average member of the community (Holmes 1963,
86–8). In the final stage, that general standard becomes specified into fixed,
definite and certain rules (Holmes 1963, 88–92, 98–103). Those liability rules
are in the form of scientific laws of antecedence and consequence identifying
particular behavior which, under particular circumstances, will be followed
by particular legal sanctions. Moreover, those liability rules are based on
scientific laws about the danger of particular behaviors under particular
circumstances, discovered by experience (Holmes 1963, 89, 98–9, 126–9).

Consistent with positivist reductive epistemology, Holmes continually
reduced morally-freighted terms in the law, such as intent, malice, and

42 Patrick J. Kelley

© Blackwell Publishers Ltd 2002.



negligence, to descriptions of voluntary action with knowledge of circum-
stances enabling a reasonable man to foresee danger (Holmes 1963, 43–7,
104–14). These reductions purge the law of metaphysical notions and relate
liability standards to scientific laws of antecedence and consequence.

Finally, Holmes’s theory of common law development in Lecture I of 
The Common Law (Holmes 1963, 5–33) can be seen as thoroughly positivist. If
the only things we can know are our experiences of phenomena and the
scientific laws of similitude and of antecedence and consequence which 
we derive from those experiences, it follows that the only basis for a judicial
decision we can know must be the consequences of that decision—its
legislative policy (Holmes 1963, 31–2). Thus, it follows from the positivist
epistemology that, whatever judges think the reason for their decision may
be, the only knowable basis for the decision is its legislative policy—its
consequences (Holmes 1963, 31–2).

Holmes’s last major statement of his legal theory, “The Path of the Law”
(Holmes 1897), exhibits in stark form his positivist methodology and com-
mitments. The law is reduced to predictions of what courts will do in fact
(Holmes 1897, 457); legal duty is reduced to a prediction of certain conse-
quences from certain conduct under certain conditions (Holmes 1897, 458).
The true ground of any legal decision lies in the consequences of that
decision (Holmes 1897, 466–7). The law is progressing toward an ideal, in
which the content of the law will be based on rational choices, informed by
scientific knowledge of the social consequences of different choices (Holmes
1897, 468–9). The study of legal history shows us the origins of current laws,
and in so doing frees us to reconsider the law from the standpoint of its social
consequences (Holmes 1897, 469–74).

8. The Problem with Holmes’s Judging

After writing The Common Law as a young man, Holmes spent fifty years of
his professional life as an appellate court judge. From 1882 to 1902 he served
as a justice of the Massachusetts Supreme Judicial Court; from 1902 to 1932
he served as a Justice on the United States Supreme Court. Recently, scholars
studying Holmes’s judicial career have been puzzled by the seeming dis-
crepancy between Holmes’s apparent theories and his performance as a
judge (Grey 1989; Tushnet 1977). Given Holmes’s seemingly radical critique
of traditional legal reasoning, his seemingly radical call for explicitly policy-
based judicial decision making, and his championing of experience over
logic as the basis for judicial decisionmaking, they expected to find that his
judicial opinions would be marked by independent, critical, policy-based
rejection of precedent, and that his decisions would be based on a sensitive
exploration of the human experiences to which the law applied. Instead,
they found that Holmes as a common-law judge hardly ever dissented, de-
ferred obsequiously to precedent, and was almost obsessively concerned

Holmes, Langdell and Formalism 43

© Blackwell Publishers Ltd 2002.



with drawing distinct lines between competing strands of precedent. They
were disturbed to find in Holmes’s opinions not a sensitive exploration of
the full range of human experience but a harsh disregard for human weak-
ness and natural human emotion. And they were puzzled that in cases in
which the court had to resolve policy questions, Holmes ordinarily refused
to make an independent policy choice, deferring instead to the legislature 
or to public opinion or to precedent. Thus, these two respected scholars
concluded that Holmes as a judge simply abandoned his enlightened theory
of judicial decisionmaking (Grey 1989; Tushnet 1977).

9. Holmes’s Scientific Theory of Judicial Decision-making

A closer look at Holmes’s scientific theory of judicial decision-making elab-
orated in The Common Law and “The Path of the Law” may help us decide
whether his performance as a judge was consistent with that theory. For
Holmes, the theological and metaphysical entities assumed by normative
justifications for legal rules are unreal, or at least unknowable. The only
scientific justification for a common law decision, therefore, is social policy—
the socially beneficial consequences of that decision. Every decision pro-
motes some social policy, Holmes thought, whether or not that policy was
the conscious reason for the decision (Holmes 1963, 31–2). Paradoxically,
then, judges who share the theological or metaphysical confusions of their
society may decide wisely for that society precisely because they share its
errors. Take, for example, judges in a primitive society who share its belief
that inanimate objects have life and will. They may punish an offending tree
that fell on the plaintiff only because they believe the tree willed the harm
(Holmes 1963, 10–6), but in doing so they unconsciously promote beneficial
social policy (Holmes 1963, 34–6). By satisfying the plaintiff’s passion for
vengeance, the judges forestall private acts of vengeance that might lead to
feuds or violence. By deciding the way the people think they ought to, the
judges preserve the effectiveness of the law.

Holmes recognized that people were unlikely to follow a rule contrary to
the society’s fundamental beliefs. Thus, social policy itself imposes a limit 
on the progressive development of the law. Judges cannot get too far ahead
of their society or the law will lose its effectiveness (Holmes 1963, 36–42). In
most cases, this is not a problem, because judges will share the community’s
prevailing beliefs and therefore unconsciously adopt rules that preserve the
effectiveness of the law. Positivist judges, who have escaped from their society’s
metaphysical or theological modes of thought, however, must recognize 
the social policy of preserving the effectiveness of the law as a limit on their
ability to implement other social policies.

Holmes thought that the common law may nevertheless improve in a
number of ways. First, it may become more effective in achieving the socially

44 Patrick J. Kelley

© Blackwell Publishers Ltd 2002.



desirable consequences that justify it, whatever those consequences may be.
The law must be knowable to be effective, because it operates by threatening
actors with legal consequences if they take certain action (Holmes 1963,
42–3). Judges can make the law more effective, then, by making it more clear,
definite, and certain. Those subject to the law will know better what legal
consequences follow certain actions, and they are then in a better position to
bring their conduct into line with the law.

Second, the law may improve as the community discards outmoded theo-
logical and metaphysical beliefs. When the community no longer believes
that trees have life and will, for instance, its judges need no longer punish
harm-causing trees. The old rule will either pass away or be fitted out with
new policies more congenial to the community’s current beliefs (Holmes
1963, 8–33).

Third, judges who realize that all common law rules are based on social
policy may consciously improve the law by adopting rules that more effect-
ively implement current policies or that implement a new policy preferable
to the one underlying the old rules (Holmes 1897, 466–8). These decisions
can be made scientifically, however, only after scientific studies showing the
consequences of particular rules and comparing the social advantages of
different consequences (Holmes 1897, 468–9).

10. Holmes’s Reflections on Judging

Holmes’s progressive historicism not only informed his theory of judging, it
also defined his predicament as a judge: What should a positivist judge 
do when his society is still mired in theological and metaphysical modes of
thought, and the scientific studies needed to make scientific choices among
legal rules are not available? Holmes’s response to this predicament can be
seen in his personal reflections on judging, embodied in an article (Holmes
1899) and a speech (Holmes 1962) toward the end of his twenty years on the
Massachusetts Supreme Judicial Court. We can determine his personal
approach to judging from these reflections, read against the background of
his scientific theories embodied in The Common Law and “The Path of Law.”

First of all, Holmes thought, the positivist judge ought to adhere strenu-
ously to the doctrine of stare decisis,4 as that makes the law more reliable,

Holmes, Langdell and Formalism 45

© Blackwell Publishers Ltd 2002.

4 Holmes 1962, 156: “It has seemed to me that certainty is an illusion, that we have few scientific
data on which to affirm that one rule rather than another has the sanction of the universe, that
we rarely could be sure that one tends more distinctly than its opposite to the survival and
welfare of the society where it is practiced, and that the wisest are but blind guides. But we have
a great body of law which has at least this sanction that it exists. If one does not affirm that it is
intrinsically better than a different body of principles which one could imagine, one can see an
advantage which, if not the greatest, at least, is very great—that we know what it is. For this
reason I am slow to assent to overruling a decision. Precisely my skepticism, my doubt as to the
absolute worth of a large part of the system we administer, or of any other system, makes me
very unwilling to increase the doubt as to what the court will do.”



certain, and knowable, and hence more effective in achieving its socially bene-
ficial consequences, whatever they may be. In interpreting and applying
precedent, however, the positivist judge should isolate the social policy
behind the prior decision and reason from that to determine its scope; the
judge should not reason from the metaphysical or theological justifications
given by prior courts.5 Finally, judges can and should make technical decisions
establishing a clear line between competing lines of precedent. Line-drawing
decisions seem peculiarly apt for making the law more fixed, definite, and
certain because a line-drawing decision can determine just how far each of
two competing strands of cases extends. Since it clarifies the precise limit of
both strands, it seems doubly beneficial (Holmes 1899, 232–3). A judge may
sometimes have the opportunity to make a pure policy choice: for example,
when an old precedent based on outmoded policy needs to be given a new
policy base. But there will rarely be scientific studies that settle the policy
question. Holmes (1962, 156) said: “We have few scientific data on which to
affirm that one rule rather than another has the sanction of the universe, …
we rarely could be sure that one tends more distinctly than its opposite to the
survival and welfare of the society where it is practiced ….” The judge ought
not express his personal wish, then, but recognize that different portions of
society want different things, and adopt “the resultant, as nearly as [he can]
guess, of the pressure of the past and the conflicting wills of the present”
(Holmes 1962, 156). Holmes recognized that there was no scientific way to
choose among different beneficial consequences once and for all, so Holmes
never articulated substantive criteria for preferring one set of consequences
over another. For Holmes, the “worth” of competing social ends depended
only on the intensity of the competing desires for those ends, which varies over
time (Holmes 1899, 231). The only scientific way to choose between ends
would be to measure the intensity of those competing desires scientifically.
Holmes (1899, 242) held out hope that someday this could be done, while
recognizing that it was still a distant, hoped-for ideal. The scientific studies
of social consequences Holmes recommended in “The Path of the Law” were
more immediately feasible, but they would not provide criteria for choosing
between two competing social policies.

46 Patrick J. Kelley

© Blackwell Publishers Ltd 2002.

5 See Holmes’s criticism of a particular piece of judicial reasoning: “Now here the reasoning
starts from the vague generalization Right, and one asks himself at once whether it is definite
enough to stand the strain. If the scope of the right is already determined as absolute and
irrespective of motive, cadit quaestio, there is nothing to argue about. So if all rights have that
scope. But if different rights are of different extent, if they stand on different grounds of policy
and have different histories, it does not follow that because one right is absolute, another is—
and if you simply say all rights shall be so, that is only a pontifical or imperial way of forbidding
discussion” (Holmes 1899, 241).



11. Holmes’s Performance as a Judge

Holmes’s performance as a judge seems perfectly consistent with Holmes’s
theories. Holmes’s obsession with line drawing and his adherence to pre-
cedent are consistent with the positivist injunction to make the law more fixed,
definite, and certain, and hence more effective in achieving its social policy.
Every legal precedent promotes some social policy, according to Holmes, so
even technical legal decisions without reference to specific social policies are
perfectly consistent with Holmes’s theory that social policy underlies every
common law rule. Holmes’s apparent reluctance to make explicit policy
choices and his deference to the policy choices of the legislature and to public
opinion also reflect both his positivist commitments and his conclusion that
there is no scientific way to establish that one policy is better than another.

Holmes as a judge was innovative, however, in three recurring kinds of
cases. First, whenever possible, Holmes gave policy-based reasons for accepted
rules or doctrines, without changing the rules or doctrines themselves.6 In
this, he was evidently attempting to substitute the real, policy based reasons
for the theological or metaphysical reasons previously given by judges.

Second, even without scientific studies of consequences, Holmes could
adopt new common law rules when they were clearly called for by his positivist
analysis alone. His analysis in The Common Law, for example, called for 
the adoption of wholly external standards of liability in tort and criminal law
(Holmes 1963, 42–3, 61–2, 86–90, 115–29). Holmes derived this conclusion
from the purpose of the law—“to induce external conformity to rule”
(Holmes 1963, 42)—and the observation that external or objective standards,
in which the personal motives or intentions of the actor are irrelevant, are
more effective than subjective standards in inducing external conformity
(Holmes 1963, 86–9). Holmes as a theorist needed no scientific studies to
establish that conclusion, so Holmes as a judge never hesitated to adopt
external or objective liability standards in tort and criminal law.7 Careful study
bears this out. My review of all Holmes’s tort and criminal law opinions 
in the Massachusetts Supreme Judicial Court concluded that Holmes did
everything he could to move tort and criminal law in Massachusetts towards
completely external liability standards (Kelley 1992, 283–352).

Finally, Justice Holmes could apply theorist Holmes’s theory of spe-
cification, too, without any additional scientific policy studies.8 That theory
was simple. As experience with the danger of certain conduct under certain
circumstances accumulates, judges should substitute specific rules of lia-
bility for the general negligence standard of danger foreseeable by the

Holmes, Langdell and Formalism 47

© Blackwell Publishers Ltd 2002.

6 See, e.g., Rideout v. Knox, 148 Mass. 368 (1889); White v. Duggan, 140 Mass. 18 (1885);
Commonwealth v. Peaslee, 177 Mass. 267, 59 N.E. 55 (1901).
7 See, e.g., Commonwealth v. Pierce, 138 Mass. 165 (1884); Davis v. Rich, 180 Mass. 235 (1902);
Cotter v. Boston Railroad Co., 180 Mass. 145 (1901).
8 See, e.g., B. & O. Railroad v. Goodman, 275 U.S. 66 (1927); McCarvell v. Sawyer, 173 Mass. 540
(1899); Pinney v. Hall, 156 Mass. 225 (1892); Lorenzo v. Worth, 170 Mass. 596 (1898).



ordinary reasonable man (Holmes 1963, 89–98). This process of specification
makes the law more definite, fixed, and certain and therefore more effective
as a deterrent to dangerous conduct. When the judge is not sure what experi-
ence teaches about the danger of this conduct under these circumstances, he
should leave the question to the jury, which is likely to have a clearer view
of what experience teaches. But after a series of similar jury cases, the judge
should embody the consistent jury decisions in a specific rule laid down 
by the court to resolve similar cases in the future. Even without a series of
jury cases, however, the court can announce a specific rule if it understands
fully what experience teaches about the danger of certain conduct under
certain circumstances. “[W]hen standards of conduct are left fo the jury,”
Holmes said (1963, 100–1), “it is a temporary surrender of a judicial function
which may be resumed at any moment in any case when the court feels
competent to do so.”

Even in these last two kinds of innovation, where Holmes as a judge
seems to adopt a new rule, Holmes did not believe he was implementing his
own view of public policy. Instead, in each kind of case, Holmes no doubt
believed he was just taking the next step in the necessary evolution of the
law from the metaphysical to the positivist stage. The change in each kind of
case was simply in the form of the law, moving to a form that made the law
more fixed, definite, and certain and hence more effective at achieving the
consequences—the social policy—that already justified the prior law.

12. The Common Core of Legalism in Holmes and Langdell

We can see, then, that in their prescriptions for judicial decisionmaking, the
theories of Holmes and Langdell converge. The point of convergence is at
what we might call extreme legalism: the position that judges ought to decide
a current case based on the rules established by past precedent, without ap-
peals to special claims of justice in an individual case, and without appeals
to any higher-order normative principles beyond or in back of the prior
precedent.

Holmes arrived at this extreme legalism by the following route. All laws
are scientifically reducible to laws of antecedence and consequence: If you
do X and Y happens, judges will order that Z be done. Moreover, laws are
only justified by their consequences. Law has only practical ends: its social
consequences for the community or policy. But there is no scientific basis for
choosing among policies, even with scientific studies of the social conse-
quences of alternative laws. The judge’s role is not, therefore, to choose among
different policies but to root out misguided metaphysical or theological
reasoning, relate each law to the social consequences that alone justify it, and
make the law more effective in achieving that policy by making the law more
clear, definite, and certain, and therefore more effective in influencing
human behavior.

48 Patrick J. Kelley

© Blackwell Publishers Ltd 2002.



Langdell arrived at this extreme legalism by a different route. If law is
separate and distinct from morality, any scientific understanding of law
must clearly separate the law from normative justifications for the law. The
best way to do that is to isolate the true or correct legal meaning of legal
doctrines. A scientific judge, then, should decide cases objectively, by logical
application of these purified legal doctrines to the facts of the case. This would
keep the judge’s own personal moral judgments, which may differ from judge
to judge, out of the law. The law will then be more even-handed, predictable,
and knowable.

For both Holmes and Langdell, then, appeals to justice in the individual
case and to higher-order normative principles related to justice were not
persuasive. For Holmes, this was because justice was a metaphysical fiction,
and an individual case is simply an occasion to set out a rule that will influ-
ence human behavior in a certain way in the future. For Langdell, this was
because there are always equally plausible arguments about what justice
requires in an individual case, there is no scientific way to determine the
uniquely just result, and objective, predictable results are achievable if 
we focus simply on the logical application of purified legal doctrine.

For Langdell, then, and to a large extent for Holmes, judicial decision-
making in a common-law system is appropriately legalistic. Courts should
simply apply pre-existing rules and doctrines—purified by Langdell or reduced
by Holmes—to make the legal system predictable, consistent, and knowable.

When Holmes lifted his critique of Langdell’s methodology out of his
review of Langdell’s second edition and dropped it, without reference to
Langdell, into the first lecture of The Common Law, he changed its context 
so radically that he changed its apparent meaning. In the context of his first
lecture, Holmes’s opening lines are an attack on all syllogistic judicial reason-
ing from general normative principles and a harbinger of his conclusion 
in that lecture that what always justifies judicial decisions are the social
consequences of those decisions. Those who took this meaning and read 
it back into Holmes’s critique of Langdell subsequently saw Langdell as a
representative of the mythical “formalist” school of legal reasoning they
thought was accepted by most late nineteenth-century judges. They did
not see Langdell as he was: A philosophically naive, trendy, scientism
wannabe9 whose legalistic, “purified doctrine” methodology would have
been scorned by all the great judges of the nineteenth century precisely
because it failed to account for the ordinary method of judicial reasoning
from the ultimate normative principles that shape legal doctrines and limit
their application.

There is an irony here. The historical success of Holmes’s broadly under-
stood transplanted critique of Langdell has swamped Holmes’s apparent
narrow critique of Langdell’s methodology. Holmes’s narrow critique seemed

Holmes, Langdell and Formalism 49

© Blackwell Publishers Ltd 2002.

9 American colloquialism for one who imitates and thus seems to “want to be” another.



to be this: A methodology of purified doctrinal terms collapses into a pur-
poseless legalism that ignores judges’ traditional concern for justice in the
individual case, based on broader normative postulates. But Holmes’s pro-
posed judicial methodology, in which judges would look beyond the individual
case only to the social consequences of their decision, was also antithetical to
ordinary judicial reasoning from general normative principles. Furthermore,
since Holmes’s positivism left him an agnostic as to the relative worth of
different consequences, his practical advice to judges, as well as his practice
as a judge himself, was to decide cases so as to make the law more fixed, def-
inite, and certain, and thus more effective at achieving whatever consequences
in fact justified it. So Holmes’s philosophically sophisticated scientism 
led him, like Langdell, to clarify and purify legal doctrine so that it could be
applied objectively, consistently, and predictably. Holmes and Langdell were
legal soul mates, differing only in the sophistication with which they
understood the trendy scientism of their age.

Southern Illinois University School of Law
Carbondale, Il 62901-6804

USA
email: pkelley@siu.edu

References

Baker, John Hamilton. 1979. An Introduction to English Legal History. London:
Butterworths.

Gilmore, Grant. 1974. The Death of Contract. Columbus: Ohio State University Press.
——— . 1977. The Ages of American Law. New Haven: Yale University Press.
Grey, Thomas. 1983. Langdell’s Orthodoxy. University of Pittsburgh Law Review 45:

1–53.
——— . 1989. Holmes and Legal Pragmatism. Stanford Law Review 41: 787–870.
Holmes, Oliver Wendell Jr. 1870. Codes, and the Arrangement of the Law. The

American Law Review 5: 1–13.
——— . 1871. Book Notice. The American Law Review 5: 539–40.
——— . 1872. Book Notice. The American Law Review 6: 353–54.
——— . 1877. Primitive Notions in Modern Law: No. II. The American Law Review 11:

641–60.
——— . 1880. Book Notice. The American Law Review 14: 233–5.
——— . 1897. The Path of the Law. Harvard Law Review 10: 457–78.
——— . 1899. Law in Science and Science in Law. Harvard Law Review 12: 443–63.
——— . 1946. Letter to Frederick Pollock, April 10, 1881. In Holmes-Pollock Letters,

16–8. Cambridge: Harvard University Press.
——— . 1953. Letter to Harold Laski, September 15, 1916. In Holmes-Laski Letters, 

vol. 1. Ed. Mark DeWolfe Howe, 20–2. Cambridge: Harvard University Press.
——— . 1962. Twenty Years in Retrospect (December 3, 1902). In The Occasional

Speeches of Justice Oliver Wendell Holmes. Ed. Mark DeWolfe Howe, 154–7.
Cambridge: Harvard University Press.

——— . 1963. The Common Law. Ed. Mark DeWolfe Howe. Cambridge: Harvard
University Press. (1st ed. 1881.)

50 Patrick J. Kelley

© Blackwell Publishers Ltd 2002.



Howe, Mark DeWolfe. 1957. Justice Oliver Wendell Holmes: The Shaping Years 1841–
1870. Cambridge: Harvard University Press.

Kelley, Patrick J. 1990. Who Decides? Community Safety Conventions at the Heart of
Tort Liability. Cleveland State Law Review 38: 315–90.

——— . 1991. Proximate Cause in Negligence Law: History, Theory, and the Present
Darkness. Washington University Law Quarterly 69: 49–105.

——— . 1992. Holmes on the Supreme Judicial Court: The Theorist as Judge. In 
The History of the Law in Massachusetts: The Supreme Judicial Court 1692–1992. 
Ed. Russell K. Osgood, 275–352. Boston: Supreme Judicial Court Historical
Society.

Langdell, Christopher C. 1870. A Selection of Cases on the Law of Contracts, Part 1.
Boston: Little, Brown & Company.

——— . 1871. A Selection of Cases on the Law of Contracts. Boston: Little, Brown &
Company.

——— . 1879. A Selection of Cases on the Law of Contracts, with a Summary of the Topics
Covered by the Cases. Boston: Little, Brown & Company.

——— . 1880. A Summary of the Law of Contracts. Boston: Little, Brown and Company.
(1st ed. in Langdell 1879.)

——— . 1887. Harvard Celebration Speech. Law Quarterly Review 3: 123–25.
LaPiana, William. 1994. Logic and Experience: The Origin of Modern American Legal

Education. New York: Oxford University Press.
Little, Eleanor, N. 1954. The Early Readings of Justice Oliver Wendell Holmes.

Cambridge: Harvard Library Bulletin.
Llewellyn, Karl N. 1960. The Common Law Tradition: Deciding Appeals. Boston: Little,

Brown & Co.
Maine, Henry Sumner. 1970. Ancient Law: Its Connection with the Early History of

Society and Its Relation to Modern Ideas. Gloucester: Smith. (1st ed. 1861.)
Mill, John Stuart. 1969. Auguste Comte and Positivism. In Collected Works of John

Stuart Mill, vol. 10: 261ff. Toronto: University of Toronto Press. (1st ed. 1865.)
Milsom, S. F. C. 1981. Historical Foundations of the Common Law. 2nd ed. Toronto:

Butterworths.
Phelps, Edward J. 1892. Methods of Legal Education I. Yale Law Journal 1: 139–43.
Pollock, Frederick. 1876. Principles of Contract at Law and in Equity. London: Stevens

and Sons.
Reimann, Matthias. 1992. The Common Law and German Legal Science. In The

Legacy of Oliver Wendell Holmes. Jr. Ed. Robert Gordon, 72ff. Stanford: Stanford
University Press.

Savigny, Friederich Carl von. 1848. System des heutigen römischen Rechts. Vol. 1. Berlin:
Veit & Comp.

Simon, Walter Michael. 1963. European Positivism in the 19th Century: An Essay in
Intellectual History. Ithaca: Cornell University Press.

Simpson, A. W. B. 1981. The Rise and Fall of the Legal Treatise. University of Chicago
Law Review 48: 632–79.

Sorell, Tom. 1991. Scientism: Philosophy and the Infatuation with Science. London:
Routledge.

Speziale, Marcia. 1980. Langdell’s Concept of Law as Science: The Beginning of Anti-
Formalism in American Legal Theory. Vermont Law Review 5: 1–37.

Tiedeman, Christopher G. 1892. Methods of Legal Education III. Yale Law Journal
1: 150–8.

Tushnet, Mark. 1977. The Logic of Experience: Oliver Wendell Holmes on the
Supreme Judicial Court. Virginia Law Review 63: 975–1052.

Holmes, Langdell and Formalism 51

© Blackwell Publishers Ltd 2002.