Summer School on Law and Logic
European University Institute & Harvard Law
School
14-25 July, 2014
Florence, Italy
Scott Brewer
Session 8.1.2
(post 07-24-14)
Legal-Theoretic approaches to the logic of legal
interpretationJurisprudential issue: structural vs.
practical enthymemicity
(i) The theorist of informal (including legal) argument must
reconstruct enthymematic arguments in order to explain,
from a theoretical point of view, what logical form they
have in general -- deductive, inductive, analogical, or
abductive. That is, what is not perspicuous in the manner of
presentation of an informal argument, and what therefore
calls for theoretical explication, is its logical type (inductive,
deductive, etc.). We may call this type of nonperspicuity
'structural enthymemicity'.
(ii) The practical legal interpreter (the judge, the lawyer,
etc.) must also engage in interpretive reconstruction of legal
arguments, but from a point of view different from that of
the theorist, and with different aims and criteria of success.
The practical legal interpreter reconstructs judicial
enthymemes in order to be guided in various ways (for
example, to conform conduct to rules that may emerge from
those arguments, or to apply rules to instant cases) by
enthymematic legal arguments in particular circumstances.
From the point of view of this practical interpreter, what is
nonperspicuous in an argument (say, a precedential judicial
opinion) is the guidance it provides in the case sub judice.
Which type of enthymemicity seems to be addressed by
Holmes in The Path of the Law, and by Dewey in Logical
Method and the Law?
2

Structural is the central theme.

Dewey, as we shall see, seems to believe that no
legal argument, no "practical argument," is ever
deductive.
Defeasibility (factual defeasibility – contrast legal rule
defeasibility)
 Exercise. Consider this argument:
(P1) Jones confessed to shooting Smith
(P2) Each of five witnesses testified
that he or she saw Jones shoot Smith
(P3) Jones' fingerprints were found on
the gun recovered at the scene of
Smith's shooting.
Therefore, (C) Jones shot Smith
(1) If P1, P2, and P3 were all true, how strong would they be as
"evidence" for the Conclusion C
(2) Now suppose that the whole set of P1 through P7 are true
(including, that is, P1, P2, and P3). In that case, with the larger set
of propositions as a background, how strong would P1, P2, and P3
be as "evidence" for C, under the conception of "evidence"
reflected in the FRE?
P4
Jones was beaten by the police and ordered to
confess.
P5
Each of the five witnesses was bribed by the
prosecutor to testify that he she saw Jones shoot Smith.
P6
Fingerprint evidence is reliable only 40% of the
time.
P7
The technicians in laboratory to which the gun was
sent for fingerprint analysis were both incompetent and
corrupt.
3

Defeasible argument:
Less formal definition: An argument from premises to a
conclusion is defeasible if and only if the argument is one in
which it is possible that the addition of some premises to the
argument's original premises can undermine the degree of
evidential warrant that the original premises provide for the
conclusion.
More formal definition: An argument from premises P1-Pn to
conclusion C is defeasible if and only if the argument is one in
which it is possible that the addition of some premise(s), Pn+1,
to P1-Pn, can undermine the degree of evidential warrant
premises that P1-Pn provide for C.
Queries:
(i) Is it possible that a deductively valid inference is defeasible -i.e., is it possible to add any premise to a deductively valid
inference that undermines the force of the evidential warrant the
original premises provide for the conclusion?
(ii) In the example of Jones and Smith offered above, is the
argument from P1, P2, and P3 as premises to the conclusion h
defeasible or indefeasible?
One question has been central to both general jurisprudence and
the jurisprudence of logical form: Whether and to what extent
legal arguments, especially the legal arguments offered by judges,
can be adequately represented as valid deductive inferences.
Can some such arguments be thus represented? Many? A large
percentage? None? Legal Realists like Holmes and John Dewey,
as well as other major jurisprudential thinkers like H.L.A. Hart,
Ronald Dworkin, and judge-theorists like Antonin Scalia, believe,
and some of them explicitly argue, that something important is at
stake in providing a philosophical explanation of legal argument
that correctly answers this question. Why might they think this
question important, and why might we?
For the philosophers in this group (though this may motivate
others less), one answer is just that, as a matter of increasing the
4
stock of philosophical knowledge, it is valuable to be able to
understand and explain the structure of a type of reasoning, legal
and more specifically, judicial, that seems (at least, seems "pretheoretically") to be a central component of legal decision-making.
In this way answering the question derives some of its value from
the value human communities have placed on the practices and
institutions of law for thousands of years. That value, the
millennial value of understanding legal practices and institutions,
is reflected in the robustness of the debate about how best to
explicate the concept of "law."
Another answer, closely related to the first, links two values that
are and have been widely held, indeed are and have been almost
pervasive in many human cultures for thousands of years.
One value is the value to a society of institutions and practices of
law, the other is the value to a society of institutions and practices
of reason and the faculty of reasoning.
These values become linked synergistically: we want to know
whether and to what extent the institutions and practices of law
that seem so important to society are capable of being guided by
reason and, for given legal systems, whether and to what extent
they are in fact guided by reason.
Now, the concept of "reason" itself is obviously highly abstract
and has been the subject of never ending attention and speculation
probably since humans acquired self-reflective cognitive powers.
It certainly has been a core philosophical interest for several
thousand years.
And while it's long been clear there are many types of reason and
many types of reasoning, one type has been regarded by many
philosophers and others in the broader culture as prima inter
pares: deductive reason and deductive reasoning. Perhaps the first
time that philosophers and others saw the power of deductive
reason literally graphically illustrated was with the publication of
Euclid's Elements in ancient Greece.
Geometric reasoning, and its axiomatic deductive method, came to
be a dominant metonym for "reason" itself, or anyway reason at its
5
best. Since that time a great many disciplines, including empirical
science, social science, philosophy, and law,1 have sought to
harness the power of deductive systems to advance their projects
of explanation and social coordination. So while it may seem to be
a relatively technical and narrow concern, what a great many
theorists -- including Legal Realists, Legal Positivists, and Natural
Lawyers -- all regard the question (1) as of vital theoretical and
practical importance. For reasons (!) just noted, question (1) has
in turn been associated with another
What did the legal realists deny?: Conceptualist vs.
Legal Realist claims about deduction in legal argument
John Maxcy Zane was a "conceptualist" legal theorist, as that term
is defined by Hart.2 Zane seemed to believe that legal arguments
could be accurately represented as an a priori deductive system.
Zane's conceptualism is revealed in his succinct (if somewhat
unclearly phrased) declaration that
[I]t must be perfectly apparent to anyone who is willing to
admit the rules governing rational mental action that unless
the rule of the major premise exists as antecedent to the
ascertainment of the fact or facts put into the minor
premise, there is no judicial act in stating the judgment.
(Zane, German Legal Philosophy, 16 Mich. L. Rev. 287
(1918) (emphasis added))
1
See for example the very illuminating discussion in Hoeflich, Law and
Geometry: Legal Science from Leibniz to Langdell 30 Am. J. Legal Hist. 95
(1986).
2
Hart's definition in The Concept of Law, Chapter VII, is:
Different legal systems, or the same system at different times, may
either ignore or acknowledge more or less explicitly . . .[the] need for
further exercise of choice in the application of general rules to
particular cases. The vice known to legal theory as formalism or
conceptualism consists in an attitude to verbally formulated rules
which both seeks to disguise and to minimize the need for such choice,
once the general rule has been laid down." (COL, 2nd ed. @ 129)
6
It seems that the likeliest interpretation of Zane's phrase 'no
judicial act' is that a judge who offers a legal argument to resolve a
case that is not a deductive argument hasn't even succeeded in
making a (judicial) legal argument at all.
The "Zany"
conceptualist view, as we might call it, seems to be that:
(i) Deductive form is a necessary condition of a "judicial
act."
Note that proposition (i) entails:
(ii) No judicial-legal arguments are defeasible.
By sharp contrast to a Zane-like view, many legal theorists,
including many or most or even all Legal Realists, hold the view
that many, perhaps most or even all, judicial legal arguments are
defeasible. We can articulate two possible versions of the claim
that judicial legal arguments are defeasible (the "defeasibility
thesis"), which I'll call the "strong version" and the "moderate
version."
Defeasibility thesis (two versions)
Strong version: (iii) No judge's legal argument can be
accurately represented as a valid deductive inference
note that (iii) entails
(iv) All judicial-legal arguments are defeasible.
Moderate version: (v) In every set of a judge's legal
arguments that are offered to resolve a case, there is at
least one very important argument that cannot be
adequately represented as a valid deductive inference; note
that (v) entails
(vi) In every set of a judge's legal arguments that
are offered to resolve a case, there is at least one
very important argument that is defeasible.
7
Query: to which version of the defeasibility thesis is Holmes, or Dewey,
committed?
To which version does they seem committed when they both
assert that "general propositions do not decide concrete cases"?
What does it even mean to say that a general proposition does not
"decide" a concrete case? (What would it mean to say that a
general proposition does "decide" a concrete case?)
Perhaps it means that general propositions, applied as major
premises in a syllogism (as the general proposition 'All men are
mortal' appears in the syllogism "All men are mortal; Socrates is a
man; therefore, Socrates is mortal") are never decisive in a
"concrete" cases?3
But even on that reading questions remain. Does Holmes mean
that general propositions, applied as major premises in a
syllogism, play no role at all in deciding "concrete" cases (this
would be what I called the strong version of the defeasibility
thesis), or play no significant role in deciding concrete cases, even
though they may play an incidental role (this would be the
moderate version of the defeasibility thesis)?
Dewey who offers the most focused discussion of the type of logic
that is needed to represent adequately judges' legal arguments.
Neil MacCormick correctly notes that Dewey's article Logical
Method and Law "is one of the great foundational texts of
American legal realism." [Rhetoric and the Rule of Law 33.]
In the discussion that follows, my principal concern is Dewey's
view of legal argument and the role of valid deductive inference
therein.
3
Holmes seems to suggest something like this in this famous passage in The
Path of the Law: "You can give any conclusion a logical form, " but "[b]ehind
the logical form lies a judgment as to the relative worth and importance of
competing legislative grounds, often an inarticulate and unconscious judgment,
it is true, and yet the very root and nerve of the whole proceeding."
8
But like other Legal Realists, Dewey is unclear in which version
of the defeasibility thesis he endorses.4 Nevertheless, for reasons
I'll offer below, I believe he may fairly and accurately interpreted
to endorse the strong version of that thesis, labeled (iiia) above.
So now, from a logical point of view, the issue is joined between
the conceptualist and the (Deweyan) Legal Realist about as starkly
as possible. While the Deweyan Legal Realist endorses:
Holmes-Dewey Legal Realist (???): (iv) All judicial-legal
arguments are defeasible
"Zany" conceptualist: (ii) No judicial-legal arguments are
defeasible.
We could not have a (logically) stronger opposition between a
Zaney conceptualist view and a strong Legal Realist view like the
view Dewey appears to hold. Encountering such a stark clash of
claims about the nature of judicial legal arguments engenders for
us this question: What kind of philosophical argument can we
make to "adjudicate" such extremely opposed views?
Both Hart and the "deductive punctuated equilibrium" model of
legal argument provide such arguments, as does Dewey in Logical
Method and Law. I now turn to consider Dewey's argument in
detail. After doing so, and after having argued that there are very
significant problems with his explanation of legal argument, I'll
present two closely related competing views of legal argument
that attempt to avoid the kind of mistake Dewey makes.
4
Neil MacCormkick notes this unclarity in Dewey's thesis and tentatively
attributes to Dewey the weak thesis (iva), but to Legal Realism, for which
Dewey was such a powerful influence, the strong thesis (ii):
I follow [Dewey] in thinking the certainty we can have in law is, at
best, qualified and defeasible certainty. Perhaps this was the very
thing [Dewey] had in mind. Surely it was at leas part of what he had
in mind. Anyway, it has become a dominant theme in American
jurisprudence right through the twentieth century that logic and
formalism had no place in law. [MacCormick, Rhetoric and the Rule
of Law 33 (emphasis added)]
9
Holmesian anti-logic and the "fallacy of logical form"
theses
It is now a commonplace that Holmes’s declaration, "The life of
the law has not been logic; it has been experience" is what Tom
Grey has aptly called the "central slogan of legal modernism."5
Holmes first offered it in a review of Langdell's book on
contracts,6 and repeated it prominently at the opening of The
Common Law, where it serves as part of an extended admonition
about the limitations of "logic" in the best explanation of common
law doctrines:
"The object of this book is to present a general view of the
Common Law. To accomplish that task, other tools are
needed besides logic. It is something to show that the
consistency of a system requires a particular result, but it is
not all. The life of the law has not been logic: it has been
experience. The felt necessities of the time, the prevalent
moral and political theories, intuitions of public policy,
avowed or unconscious, even the prejudices which judges
share with their fellow men, have had a good deal more to
do than the syllogism in determining the rules by which
men should be governed. The law embodies the story of a
nation's development through many centuries, and it
cannot be dealt with as if it contained only the axioms and
corollaries of a book of mathematics."7
I shall refer to this basic thesis as Holmes’s "anti-logic."
The anti-logic thesis is not a passing fancy on Holmes part.
Rather, he maintained and repeated it (though not always in the
5
Thomas Grey, "Langdell's Orthodoxy," University of Pittsburgh Law Review
45 (1983): 1-3.
6
Oliver W. Holmes, Book Review, American Law Review 14 (1880): 233-234
(reviewing C.C. Langdell, Summary of the Law of Contracts (1880)).
7
Oliver W. Holmes, The Common Law, ed. Mark DeWolfe Howe (Boston:
Little, Brown and Company, 1963), 5. Holmes iterates the assertion, with an
application to contract law, about logic and experience: "The distinctions of the
law are founded on experience, not on logic." Ibid., 244. Again, in slightly
different terms, he repeats it, observing (optimistically) that "the law is
administered by able and experienced men . . . who know too much to sacrifice
good sense to a syllogism." Ibid., 32.
10
same words) for at least twenty-five years, from an 1880 review of
Langdell's contracts book to his 1905 dissent in Lochner, in which
he declared that "[g]eneral propositions do not decide concrete
cases.
The decision will depend on a judgment or intuition more subtle
than any articulate major premise."8
The anti-logical thesis is of course, also a centerpiece of The Path
of the Law (hereinafter Path).9
Holmes’s Anti-Logic Within the Larger Structure of Path of
the Law: Path's Four Jurisprudential Theses
Path's Four Main Jurisprudential Theses
To understand the content of Holmes’s anti-logic thesis in Path, it
will be helpful to locate it among the other major theses that make
up the essay.
Holmes was not a careful systematic jurisprudential thinker,
although he certainly had many flashes of brilliant jurisprudential
insight, both with regard to relatively narrow doctrinal issues and
with regard to more abstract and traditional issues of the sort
addressed by legal positivists and natural lawyers.
Despite Holmes’s lack of systemic attentiveness, Path can
profitably be read and examined for its insights into systemic
jurisprudence.
Path advances at least four major theses, all of which concern one
overarching topic: the best explanation of legal institutions,
doctrines, and reasoning. The four theses are:
8
Lochner v. New York, 198 U.S. 45, 76 (1905). That this is the same theme as
"law is not logic but experience" is revealed by comparing it to Holmes’s
second explicit reference to logic and experience in the chapter on void
contracts in The Common Law. There, he asserts that the "distinctions of law
are founded on experience, not on logic." Holmes, The Common Law, 244.
9
Oliver W. Holmes, "The Path of the Law," in Collected Legal Papers (New
York: Harcourt, Brace and Howe, 1920), 167.
11
(1) Prediction Thesis. Law (at least, the law from the
point of view of a lawyer) is a prediction of the decisions
courts will make, backed by the use public force.
(2) Separation Thesis. Legal norms (rules, doctrines,
principles, etc.) are not identical with moral norms, and
their conflation has caused much confusion in the study of
actual legal systems.
(3) Anti-Logic Thesis. The law of a living legal system,
such as that in the U.S., cannot be adequately explained as
an axiomatic deductive system, in large part because there
is inevitably a significant role played by the "inarticulate"
in a judge's discernment and application of the law.
(Note that Holmes offers the (in)famous "bad man" as a
heuristic device to illustrate and argue for theses 1, 2, and
3.)
(4) Rational Reform Thesis. It is (a) possible and (b)
normatively desirable to effect rational reform in the law –
reform that is both properly attentive to and properly
critical of history and tradition.
"Anti-logic" Thesis and the "Fallacy of Logical Form"
After advancing and illustrating the Separation Thesis, Holmes
turns to the second of his two "principles for the proper
understanding of law," a principle to which he refers as "the
fallacy of logical form."10
According to Holmes, the "fallacy" is "the notion that the only
force at work in the development of the law is logic."
This argument is of central concern and I shall consider it in detail
below.
Five Senses of 'Logic' in Path
There is much wisdom and insight in the four theses Holmes
proffers, but along the Path there are also some crucial missteps.
10
Ibid., 184.
12
Among them, in my view, is Holmes’s analysis of the role of
"logic"11 in legal reasoning, doctrine, and institutions.
That analysis comes in his treatment of the second of what he
refers to as two "first principles for the study of this body of
dogma or systematized prediction which we call the law,"12 which
he also calls "two pitfalls" that "lie perilously close to the narrow
path of legal doctrine"13 and two "fallacies."
The first "first principle" is the Separation Thesis (the associated
"fallacy," apparently, is failure to recognize its truth).
The second is the "fallacy of logical form" -- "the notion that the
only force at work in the development of the law is logic."14
To assess the cogency of Holmes arguments about this "fallacy,"
we must first discern what exactly Holmes meant in speaking of
"logic" and "logical form."
This is no small task, since Holmes used the term 'logic' in the
essay in several rather different senses without being defining and
explaining which sense of the term he had in mind at different
points.15 This lack of systemic care is important. The anti-logical
11
I follow the standard philosophical convention of using single quotation
marks to mention words, and double quotation marks to quote a speaker's (or
group of speakers') use of a term. (Thus, logic is the systematic study of
rational inference patterns; 'logic' is a five letter word; a good deal of talk about
"logic" in American legal academia is misguided.)
12
Holmes, "The Path of the Law," 169.
13
Ibid., 178.
14
Ibid., 180. The analysis of the "fallacy of logical form" takes up about five
pages, about 14% of the total pages of the essay. Ibid., 180-84. This is roughly
equal to the space Holmes devotes to the prediction thesis (pages 168-73) and
to the separation thesis (pages 173-79). Of course discussions of these theses
overlap substantially, so it is hard (and unnecessary) to fashion an accurate
measurement of space Holmes devoted to each. I offer the rough calculation
only because it is worth noting that Holmes devotes nearly 50% of the essay to
arguing the possibility and desirability of rational reform and of sketching a
program therefor.
15
Holmes was, in many ways, a master of language, as witness the terse
and powerful aperus that have earned him that reputation. But on a
broader scale his language is full of pitfalls. The meaning of the most
central concepts in his writings, such as "philosophy," "principles,"
"logic," and "experience" would have to be clearly defined from within
Holmes’s own argument before an attempt to explicate his ideas in a
more coherent and consistent way could possibly succeed.
13
thesis has such a misleading but powerful impact on the thinking
of generations of law students, lawyers, judges, and scholars about
the ways in which it is both possible and normatively desirable to
recognize and promote the life of articulate reason in legal
decision making.16 Obviously Holmes did not explain his
understanding of logic in legal decision making only in Path;
rather, the discussion in Path reflects a view that Holmes
repeatedly articulated, and it will be helpful occasionally to repair
to passages in other works to help discern his various meanings of
'logic'.
In Path, Holmes used the term 'logic' in at least five significantly
different senses. What he said about "logic" is true of only some
of the varied referents for that term. He thus slides quite close to a
logical "fallacy" of his own (namely, equivocation). The five uses
are these:
(i)
'Logic' as one of a set of roughly synonymous terms,
including 'sensible,' 'reasonable,' 'warranted,' 'advisable.’
For example, “this really was giving up the requirement of
a trespass, and it would have been more logical, as well as
truer to the present object of the law, to abandon the
requirement altogether";17 "there are some cases in which a
logical justification can be found for speaking of civil
liabilities as imposing duties in an intelligible sense."18
(ii)
"Logic" as syllogistic inference (or some other type of
deductive inference). For example, "there is a concealed,
half conscious battle on the question of legislative policy,
and if any one thinks that it can be settled only
deductively, or once and for all, I only can say that I think
he is theoretically wrong."19
Mathias Reimann, "The Common Law and German Legal Science," in The
Legacy of Oliver Wendell Holmes, ed. Robert Gordon, (Stanford: Stanford
University Press, 1992), 146.
16
See the discussion below in Section Error! Reference source not
found..Error! Reference source not found..
17
Holmes, "The Path of the Law," 188.
18
Ibid., 175.
19
Ibid., 182-83. Although the interpretive evidence for it is slightly indirect,
Holmes does use 'logic' in Path in sense (ii). At the start of page 180, Holmes
introduces the concept of a "fallacy" in the "notion that the only force at work in
the development of the law is logic." (Later, at page 184, he refers to this
fallacy by the phrase "the fallacy of logical form.") Further down the page, in
14
(iii)
"Logic" as a formal deductive system, with axioms, rules
of inference, and theorems, as in geometry. For example,
"the danger of which I speak is . . . the notion that a given
[legal] system, ours for instance, can be worked out like
mathematics from some general axioms of conduct."20
(iv)
"Logic" as a rationally discernible pattern of cause and
effect. For example, "The condition of our thinking about
the universe is that it is capable of being thought about
rationally, or, in other words, that every part of it is effect
and cause in the same sense in which those parts are with
which we are most familiar. So in the broadest sense it is
true that the law is a logical development, like everything
else."21
(v)
"Logic" as a set of argument types, individually invariant
but distinct from one another. For example, "The training
of lawyers is a training in logic. The processes of analogy,
discrimination, and deduction are those in which they are
most at home. The language of judicial decision is the
language of logic. And the logical method and form flatter
that longing for certainty and for repose which is in every
human mind."22
Query: which Sense of 'Logic' is an appropriate target for a
claim of "fallacy" (the "Fallacy of Logical Form")?
what is clearly an explication of the "fallacious" view of "logic," he refers to the
mistaken view that
a given [legal] system . . . can be worked out like mathematics from
some general axioms of conduct. . . . So judicial dissent often is
blamed, as if it meant simply that one side or the other were not doing
their sums right, and, if they would take more trouble, agreement
inevitably would come.
Ibid., 180 (emphasis added). The salient feature in Holmes simile – legal
reasoning is mistakenly thought to be "like mathematics" – is that mathematics
(and "doing sums") is a deductive process. Thus, I take Holmes’s reference to
deduction at the end of page 180 to be a exegesis of his use of 'logic' at the start
of that page, and in that way he uses 'logic' in sense (i). Ibid.
20
Ibid.
21
Ibid.
22
Ibid., 181.
15
Senses (i) and (ii) – Unproblematic
Use (i) is a common, non-technical use of 'logic' and plays no
troublesome role in Holmes’s anti-logic. Nor does use (ii) present
any real problem. Deduction is of course one type of "logical"
inference (only one among several, as I shall discuss in greater
detail below), and it would surely be a serious jurisprudential
mistake to believe that "the only force at work in the development
of law" is deductive logic. (Pace Holmes, it is difficult to find
theorists who endorse this belief, and Langdell is pretty clearly not
among them – again, more on this later.) What Holmes intends to
label the "fallacy of logical form" is what he takes to be a
particular jurisprudential view about deduction – namely, the view
that an actual legal system can be formalized in a way that allows
deductive inference of results in particular cases.23 Thus,
Holmes’s real target is not deduction per se but some view –
exactly what view we shall have to consider – about the role that
deduction either does actually play in legal argument (a
descriptive claim), or can possibly play in legal argument (a
conceptual claim), or should play in legal argument (a conceptual
and normative claim). The problems come with senses (iii) and
(iv).
Senses (iii) and (iv) – Very Problematic
It might seem that the target of Holmes’s anti-logic is
sense (iii) of 'logic' – more precisely, the view that actual legal
systems are deductively axiomatizable. But here the assertions
that comprise Holmes’s anti-logic become problematically
unclear. Holmes concedes that the proposition "the only force at
work in the development of the law is logic" is true "in the
broadest sense" (broad along what metric, one wonders) for sense
(iv) of 'logic'.24 This proposition is true, Holmes seems to believe,
by virtue of the rather Kantian view that "the postulate on which
we think about the universe is that there is a fixed quantitative
relation between every phenomenon and its antecedents and
consequents."25 As Holmes also seems to recognize, this
23
See text at note ***.
See Holmes, "The Path of the Law," 180.
25
Ibid. We know that Holmes admired Kant, at least in part, for at the end of
Path Kant is a central part of his epideictic tribute to the power of the intellect:
"To an imagination of any scope the most far-reaching form of power is not
money, it is the command of ideas. . . . Read the works of the great German
24
16
concession is a significant threat to the coherence of his antilogical and prediction theses. To see why, note that the prediction
thesis relies, at least implicitly, on the idea that judicial behavior,
like other motions and behaviors of the universe (whether
products of the intentional mind or not – Holmes does not appear
to distinguish the intentional from the purely physical), has a
rationally discernible causal structure.26 The whole idea of
"prophesying" the law seems to rely on the assumption that, in
discerning the causal structure of judicial behavior, the lawyer or
judge must examine examples of judicial behavior encountered in
experience and recorded in case reports, generalize inductively,
and predict, "prophesy," on the basis of deduction ('logic' in sense
(ii)). Thus if 'logic' is used in sense (iv), Holmes’s own prediction
thesis would be an instance of the "fallacy of logical form" unless
he can distinguish this use from a different use of 'logic' by other
theorists who, in Holmes’s view, really did commit the "fallacy."
Does he distinguish his thesis successfully?
I think not. He does tell us that the "danger of which I
speak is not the admission that principles governing other
phenomena also govern the law, but the notion that a given [legal]
system, ours for instance, can be worked out like mathematics
from some general axioms of conduct."27 That is, those who
commit the fallacy of logical form, unlike Holmesian predictors,
think that the law is or could be deductively axiomatized – sense
(iii) of 'logic'. But this brings up a tricky issue for the anti-logic
thesis. As many scholars have observed, Langdell was a chief
target for Holmes’s anti-logic.28 It is also well remarked that
jurists, and see how much more the world is governed to-day by Kant than by
Bonaparte." Ibid., 201-02. Reimann argues that Kant, in the very influence he
exercised over the "great German jurists," was at least the superficial target of
Holmes anti-logical thesis in The Common Law. His real target, suggests
Reimann, was Langdell, but Holmes the mere HLS lecturer could not, for
political reasons, directly attack the dean of Harvard Law School, where
Holmes might like a permanent job. Reimann also points out that Langdell's
view of the role of logic in law was quite different from that of many of the
"great German jurists," and that the views of at least one of them, von Savigny,
were very consonant with Holmes’s own views, though Holmes never conceded
the point. See Reimann, "German Legal Science," 146.
26
See text at note ***.
27
Holmes, "The Path of the Law," 180.
28
See Grey, "Holmes and Legal Pragmatism," 818. In correspondence with
Pollock, Holmes said of Langdell's book on contracts:
17
Langdell, in the brief passages in which he discusses the matter
(probably too brief to get a clear picture of his view), seemed to
think of the system of legal concepts as one that was generated by
means of inductive generalization from decided cases, rather than
from some a priori axiomatic structure, and only later applied
deductively.29 Langdell, like Holmes, saw a crucial role for logic
in sense (iv) in the "legal scientist's" discernment of legal rules
and principles. Langdell also saw a crucial role for the subsequent
use of deductive inference (logic in sense (ii)), when the
inductively discovered30 rules and principles were later applied to
individual cases; surely Holmes’s prediction thesis sees an
important role for deductive inference working on the rules and
principles discovered from experience. In this way, Langdell's
conception of the role of "logic" was much closer to Holmes’s
than Holmes acknowledged.
Despite these similarities, real differences of opinion about
the role of logic in legal argument seemed to remain between
Holmes and Langdell. Langdell was far more sanguine than
Holmes about the possibility of organizing the inductivelygenerated rules and principles into a coherent conceptual order
that could later be applied to individual cases apodictically. In the
helpful terms Tom Grey brought to the analysis of "Langdell's
orthodoxy," Langdell may have believed, along with other "legal
scientists" of his day, that empirically (and inductively) generated
A more misspent piece of marvelous ingenuity I never read, yet it is
most suggestive and instructive. I have referred to Langdell several
times in dealing with contracts because to my mind he represents the
powers of darkness. He is all for logic and hates any reference to
anything outside of it, and his explanations and reconciliations of the
cases would have astonished the judges who decided them. But he is a
noble old swell whose knowledge, ability and idealist devotion to his
work I revere and love.
Oliver W. Holmes to Frederick Pollock, Boston, 10 April 1881, HolmesPollock Letters: The Correspondence of Mr. Justice Holmes and Sir Frederick
Pollock, 1874-1932, ed. Mark DeWolfe Howe (Cambridge: Belknap Press,
1961), 16-17.
29
See M. H. Hoeflich, "Law and Geometry: Legal Science from Leibniz to
Langdell," American Journal of Legal History 30 (1986): 95. See Reimann,
"German Legal Science," 108-110; Grey, "Langdell's Orthodoxy," 29-30;
Anthony J. Sebok, "Misunderstanding Positivism," Michigan Law Review 93
(1995): 2054.
30
Actually, the inductions that both Holmes and Langdell contemplated relied
on an initial abductive inference as well, as do all inductive inferences.
18
legal rules and principles could in fact be organized into a system
that was "complete" (i.e., such as to provide one right answer to
every case), "conceptually ordered" (i.e., consisting of lower level
rules that could be derived from a smaller set of higher order
principles that were themselves coherent), and "formal" (i.e., such
as to provide apodictic certainty for individual legal decisions).31
Although Holmes himself aspired in much of his work to render
areas of law into conceptually ordered systems, he was also quite
skeptical of the ability of any "legal scientist," himself included,
actually to organize legal rules and principles so as to allow for
one right and certain resolution – such as deduction could in
theory provide – of every case.32 For Holmes, blind, "inarticulate"
and irrational forces were too powerfully present in legal
decisionmaking for Langdell's conceptualistic goals to be
realizable:
[T]he logical method and form flatter that longing for
certainty and for repose which is in every human mind.
But certainty generally is an illusion, and repose is not the
destiny of man. Behind the logical form lies a judgment as
to the relative worth and importance of competing
legislative grounds, often an inarticulate and unconscious
judgment, it is true, and yet the very root and nerve of the
whole proceeding.33
This is a real disagreement, but if it is this view that Holmes was
targeting as the "fallacy of logical form," then the fit between the
Langdellian ontology and epistemology of law (inductive
discovery, conceptual ordering, deductive application) and the
fallacy of logical form Holmes is tenuous at best. Again, in
Holmes’s terms, the "fallacy" is "the notion that the only force at
work in the development of law is logic."34 If 'logic' in this
proposition is being used in either sense (ii) (deductive inference)
or sense (iii) (a deductive system), then that proposition seems not
to describe Langdell's view at all; as noted above, Langdell
accorded a vital role to both inductive inference (close kin to
'logic' in Holmes’s sense (iv) – rationally discernible causal) and
31
32
33
34
See Grey, "Langdell's Orthodoxy," 6-11.
See text at note ***.
Holmes, "The Path of the Law," 181.
Ibid., 180 (emphasis added).
19
to deductive inference ('logic' in sense (ii)), operating in a system
conceptually ordered by the "legal scientist" ('logic' in sense (iii)).
But Holmes conceded that in sense (iv) of 'logic', which Langdell
thought a vital part of legal analysis, logic was a vital force in the
development of law (indeed, he even suggested, "the only force,"
albeit only "in the broadest sense"35). In sum, though there was
genuine disagreement between Holmes and Langdell about the
role of deductive logic in legal reasoning, the disagreement was
not nearly as great as Holmes made it out, in part because Holmes
mischaracterized the complexity of Langdell's own views about
the role of different modes of logical inference in legal argument.
(I discuss those different modes in some detail in the next section.)
**Sense (v) – Most Promising
This brings me to a final point. Holmes seems unclear about what
he himself understood to be within the scope of the kind of "logic"
involved in the "fallacy of logical form." The "fallacy of logical
form" seems to be the view that law can be organized into an
axiomatic system in such a way as to allow for apodictic
resolution of individual cases. But at a crucial point in the antilogic section of the paper, Holmes speaks as if it is neither solely
deduction ('logic' in sense (ii)) nor solely a deductive system (logic
in sense (iii)) that he has targeted, but rather something much
more inclusive, namely, 'logic' in sense (v):
[J]udicial dissent is often blamed, as if it meant simply that
one side or the other were not doing their sums right, and,
if they would take more trouble, agreement inevitably
would come.
This mode of thinking is entirely natural. The training of
lawyers is a training in logic. The processes of analogy,
discrimination, and deduction are those in which they are
most at home. The language of judicial decision is mainly
the language of logic. And the logical method and form
flatter that longing for certainty and repose which is in
every human mind.36
35
36
Ibid.
Ibid., 181.
20
Now, it seems, the fallacious view about the possible role of
"logic" in legal argument that Holmes is challenging embraces not
only "deduction" but also "the processes of analogy and
discrimination" – i.e., disanalogical argument! A critique of the
role of "logic" in legal reasoning that is this inclusive is surely a
far cry from the narrower and much more plausible view that law
cannot be organized into an axiomatic system that is deductively
applicable in every case. One has to be not just skeptical, but
skeptical to an implausible extreme, to deny that logic in the broad
sense of a patterned form of inference (including deduction and
analogy) plays a vital role "in the development of law." And if the
response on behalf of Holmes is that he is not critiquing the view
that "logic" plays a vital role in the development of law, but is
instead critiquing the literal belief that "the only force at work in
the development of law is logic," well, we must ask who ever
believed that logic, in any sense, was the only force at work in the
development of law? If the proposition Holmes uses to describe
the "fallacy of logical form"37 is taken literally, it seems no one,
including the most "formalist" and deductivist of the legal
scientists – German, British, or American – could have believed or
endorsed that. The ball is in Holmes’s court: if he really means it
literally, he must show us that he is not attacking a straw theory.
Ironically, perhaps, it is this last of the five conceptions of logic
that is the most promising for a deep and cogent explanation of the
role of different modes of logical inference in legal argument. As
so often, even when Holmes is misguided and somewhat
confused, his suggestions are fertile. In Part II of this paper I take
up that "suggestion" and explore the role, not just of "logic" in the
narrow sense of deduction, but in the broad sense of patterned
inference. Marking the different patterns of logical inference that
can be used in legal argument, that are used in legal argument,
and that should be used in legal argument comprises the work of
an ongoing intellectual enterprise. I call that enterprise the
jurisprudence of logical form.
Dewey's defense of Holmes' general anti-deductivism
The most challenging claim, and the most suggestive about his
view of the role of defeasibility and deduction in judicial legal
37
I have quoted that proposition several times. See, for example, the text at
note 14.
21
argument, comes in this following passage in the Dewey article,
Logical Method and Law:
Take the case of Socrates being tried before the Athenian
citizens, and the thinking which had to be done to reach a
decision. Certainly the issue was not whether Socrates
was mortal; the point was whether this mortality would or
should occur at a specified date and in a specified way.
Now that is just what does not and cannot follow from
a general principle or a major premise. Again to quote
Justice Holmes, "General propositions do not decide
concrete cases." No concrete proposition, that is to say
one with material dated in time and placed in space,
follows from any general statements or from any
connection between them. [Logical Method and Law, p.
22.5 (emphasis added)]
Does this assertion have direct implications for structural enthymemicity,
practical enthymemicity, or both? (See definitions above.)
The Scope of Dewey's claim
Dewey's assertion raises at least two questions of interpretation.
First, what is the scope of Dewey's claim? Does he intend to limit
the scope of his claim to the domain of legal reasoning (including
legal reasoning in 5th century Athens and in early 20th century
America – or maybe including legal reasoning everywhere and
everywhen?) but not extend the claim to reasoning in other
domains, such as the domain of moral reasoning, instrumental
("prudential") reasoning, or reasoning in the empirical sciences?
Second, what exactly is the meaning of the claim, or, more
precisely, what does Dewey intend to be the referent of 'them' in
his assertion that "No concrete proposition, that is to say one with
material dated in time and placed in space, follows from any
general statements or from any connection between them"? Does
'them' refer to any connection between general propositions, or
does it refer to any connection between general and concrete
propositions?
Did Dewey intend his claim to be limited to legal reasoning? To legal and
moral reasoning? To all types of practical reasoning?
22
It seems likely that Dewey does not intend to limit his claim to
legal reasoning, nor even to the more general category of practical
reasoning (practical reasoning is reasoning about what one ought
to do -- this includes as "sub domains," as species within its genus,
moral reasoning, legal reasoning, and instrumental-prudential
reasoning
See for example Dewey's characterization of "'practical'
reasonings" as "reasonings leading up to decision as to
what is to be done," Logical Method and Law, 10 Cornell
L. Q. at 18.9
Also compare Joseph Raz' related distinction of "practical"
and "theoretical" authorities, Raz, Authority, Law, and
Morality, at 195.9). But the views Dewey advances in this
article are striking in that he seems committed to absorbing
all theoretical reasoning (reasoning about what one ought
to believe) into the domain of practical reasoning.
Some evidence for this comes in his assertion, quoted above,
quoting Holmes and referring to Socrates. Dewey's reference to
Socrates' execution surely is intended to evoke the syllogism so
familiar (in Dewey's day too, one speculates) to students of
traditional syllogistic logic:
All men are mortal
Socrates is a man
Therefore, Socrates is mortal
And his deliberate invocation of that non-legal deductive
syllogism when explaining his claim about general and concrete
propositions suggests that he believes that his thesis is not limited
to legal argument.
Additional evidence is found later in the article, where he
explicitly considers the scope of his inquiry into “logical method,”
and defines ‘logical theory’ as
an account of the procedures followed in reaching
decisions . . . in which subsequent experience shows that
they were the best [procedures] which could have been
23
used under the conditions. [Logical Method and Law, 10
Cornell L. Q. at 17-18.]
In like vein, he asserts:
It will be said that [the foregoing] definition [of ‘logical
theory’] . . . , in confining logical procedure to practical
matters, fails to take even a glance at those cases in which
true logical method is best exemplified: namely, scientific,
especially mathematical, subjects.” [Logical Method and
Law, 10 Cornell L. Q. at 18 (emphases added)]
Note the dialectic here: Dewey He considers one possible
challenge to this conception of "logic":
Challenge: The scope of his analysis of "logic" is "logical
method in legal reasoning and legal decision," and in those
cases the legal reasoner is engaged in practical reasoning,
the type of reasoning that confronts "the necessity of
settling upon a course of action to be pursued." [Logical
Method and Law, 10 Cornell L. Q. at 18.3]
So perhaps not all "logic" is assimilable to practical
reasoning, and Dewey is claiming to offer a philosophical
explanation only of the species of practical logical
reasoning and not of the genus of all logical reasoning,
including theoretical reasoning?
Dialectic: Dewey's reply: No.
Having offered that suggestion, Dewey chooses not to rest
his analysis of "logic" on that answer, which he
characterizes as "partial" and "ad hoc." [Logical Method
and Law, 10 Cornell L. Q. at 18.3.]
Having acknowledged that others might regard his
definition of logical theory as ill-suited to explain
reasoning in "scientific and mathematical subjects," he still
suggests that even the mathematician’s “logical
procedures” can be assimilated to practical reasoning,
since “every thinker, as an investigator, mathematician or
24
physicist as well as ‘practical man,’ think in order to
determine his decisions and conduct—his conduct as a
specialized agent working in a carefully delimited field.”
[Logical Method and Law, 10 Cornell L. Q. at 18-19]
After offering these preliminary remarks, Dewey considers
what he clearly understood was then – and is still today –
the far more dominant explication of the concept of logic ,
namely, "an affair of propositions which constitute . . .
with a view to the utmost generality and consistency of
propositions.” [Logical Method and Law, 10 Cornell L. Q.
at 19].
But again, while acknowledging some utility for this
conception of “logic,” he seems to suggest that logic so
conceived is always only a means, not an end in itself, a
“means of improving, facilitating, clarifying the inquiry
that leads up to concrete decisions.” [Logical Method and
Law, 10 Cornell L. Q. at 19].
Dewey also invokes the case of legal reasoning to
advance an argument whose conclusion seems to be
that logic is always “ultimately an empirical and
concrete discipline” (Logical Method and Law, 10
Cornell L. Q. at 19), and that “logic must be . . . a logic
relative to consequences rather than antecedents, a
logic of prediction of possibilities rather than a
deduction of certainties.” (Logical Method and Law, 10
Cornell L. Q. at 26).
Later in his argument, he asserts the more general proposition that
and “logic is really a theory about empirical phenomena, subject
to growth and improvement like any other empirical discipline.”
[Logical Method and Law, 10 Cornell L. Q. at 27 (emphasis
added)].
In this striking claim Dewey seems aligned with the extreme
empiricist view advanced, for example, by John Stuart Mill:
[T]he foundation of all sciences, even deductive or
demonstrative sciences, is Induction; . . . every step in the
25
ratiocinations even of geometry is an act of induction; . . .a
train of reasoning is but bring many inductions to bear
upon the same subject of inquiry, and drawing a case
within one induction by means of another. [J. S. Mill, A
System of Logic 147 (8th ed. 1959).38]
Is this view of deductive logic -- assimilating it either to practical
reasoning (as Dewey suggests at several points) or to empirical
reasoning (as he suggests at other points) convincing?
Here are some reasons to doubt his claim. Dewey seems to trade
on the vagueness of the idea of the kind of "concrete decision" that
is a mathematician or logician must make when attempting to
offer a formal mathematical or logical proof.
If Dewey means to include within the scope of “concrete
decisions" Andrew Wiles’ “decision” to offer the proof he offered
of Fermat’s last theorem, then maybe Dewey is right in some
sense. But something seems off here, for that kind of “decision”
does not seem to support Dewey's contention that “logic is
ultimately an empirical and concrete discipline.”
38
Compare also the striking and intriguiging assertion by Chalres Sanders
Peirce, one of America's greatest formal logicians, also, like Holmes and Felix
Cohen, a theorist in the "pragmatist" tradition:
It may seem strange that I should put forward three sentiments,
namely, interest in an indefinite community, recognition of the
possibility of this interest being made supreme, and hope in the
unlimited continuance of intellectual activity, as indispensable
requirements of logic. Yet, when we consider that logic depends on a
mere struggle to escape doubt, which, as it terminates in action, must
begin in emotion, and that, furthermore, the only cause of our planting
ourselves on reason is that other methods of escaping doubt fail on
account of the social impulse, why should we wonder to find social
sentiment presupposed in reasoning? As for the other two sentiments
which I find necessary, they are so only as supports and accessories of
that. It interests me to notice that these three sentiments seem to be
pretty much the same as that famous trio of Charity, Faith, and Hope,
which in the estimation of St. Paul, are the finest and greatest of
spiritual gifts. Neither Old nor New Testament is a textbook of the
logic of science, but the latter is certainly the highest existing authority
in regard to the dispositions of heart which a man ought to have.
[Charles S. Peirce, Three Logical Sentiments [CP 2.655]
26
What Dewey’s (and Mill's) hyper-empiricism fails to account for
is the distinction between an a priori and a posteriori discipline.
Empirical support is need for empirical disciples but empirical
support does not seem needed for a priori disciplines. Indeed it's
hard to imagine what kind of empirical support one could get for
such logical inference rules such as modus ponens and universal
instantiation.
However convincingly (or not) Dewey does seem to endorse the
thesis that all "logical" reasoning, even reasoning in mathematics,
is a species of practical reasoning and a species of empirical
reasoning.
This leads us to our next interpretive challenge in understanding
Dewey's explanation of legal argument: when he says that "[n]o
concrete proposition, that is to say one with material dated in time
and placed in space, follows from any general statements or from
any connection between them," what is the referent of 'them'?
Does he intend "them" to refer to general and concrete
propositions, or instead only to concrete propositions?
Is Dewey's claim coherent? Immediate and fairly straightforward
counter-example?
Premise:
Everything is identical to itself at all times
and in all places.
Conclusion: Socrates was identical to himself in Athens
in 399 B.C.
This argument has the logical form of what logicians refer to as
the inference rule of “universal instantiation”
(x) Fx
Fa
If so, is Dewey committed to denying that the conclusion of this
argument follows deductively from its premise?
It's hard to imagine a more general proposition than the
proposition asserted in the premise of this argument. And the
27
conclusion also seems to be a paradigm instance of a "concrete
proposition" in Dewey's sense.
Rule skepticism and induction: Holmes, Dewey, Cohen
The basic patterns of inductive inference:
generalization and inductive specification
inductive
Recall that in an inductive argument, the truth of the premises
cannot guarantee the truth of the conclusion, but when they are
well chosen, their truth can warrant belief in the truth of the
conclusion to some degree of probability. There are two main
varieties of inductive inference: inductive generalization and
inductive specification.
Inductive generalization
Inductive generalization involves generalizing from particular
instances. The premises of this type of argument report features of
the particulars, and its conclusion states a probabilistic
generalization that is inferred from those particulars. In the notes
below we'll use two examples to illustrate the form of inductive
generalization. One is the Knapp judge's analysis of logical
relevance in the case he was deciding, the other is a simplified
example from empirical science (induction is one of the
foundations of all empirical scientific reasoning).
Where
'1 . . . n'
instances
stands for a set of individual
' '
stands for one property that
the individuals 1 . . . n
have been noted to possess
''
stands for another property
the individuals 1 . . . n
have been noted to possess,
the pattern of inductive
generalization is:
28
(1 )
1 is both  and  (i.e., has both characteristics, 
and )
[e.g., Person A made a factual assertion and
Person A spoke truly.]
[e.g., Bird A was a raven and Bird A was
black.]
(2 )
2 is both  and 
[e.g., Person B made a factual assertion and
Person B spoke truly.]
[e.g., Bird B was a raven and Bird B was
black.]
(3 )
3 is both  and 
[e.g., Person C made a factual assertion and
Person C spoke truly.]
[e.g., Bird C was a raven and Bird C was
black.]
.
.
.
(n)
n is both  and 
[e.g., Person N made a factual assertion and
Person N spoke truly.]
[e.g., Bird N was a raven and Bird N was
black.]
(n+1) There were [few or no] observed instances of an 
that was  and also was not-
[e.g., There were few persons who made a
factual assertion and did not speak truly -Knapp: "even in the greatest liars, . . .
where they lie once they speak truth 100
times."]
29
[e.g. No ravens were observed to be nonblack]

h: [Probably] [All or Most]  's are 
[e.g., Knapp: Probably, most persons who
make factual assertions are persons who
speak truly.]
[e.g. Probably, all ravens are black.]
Note that inductive arguments are arguments about evidence and
the hypotheses the evidence is said to support. Thus, the premises
of an inductive argument are evidentiary propositions (the " " in
our  - h schema) and the conclusion is a hypothesis that the
evidence is offered to support the "h" in our  - h schema.
Inductive specification
The other type of inductive inference is inductive specification.
Instead of reaching a conclusion about a class of individuals, an
inductive specification offers a conclusion about one individual,
based on a generalization about the classes to which that
individual belongs. Again, we illustrate the form of this argument
by reference to the two examples offered above:
In the Knapp example, the inductive specification is the
argument that a great many persons (Knapp endorses the
claim that the ratio is 100 to 1!) who made factual
assertions spoke truly, therefore, some individual person D
who made a factual assertion (or perhaps the next
individual person who will makes a factual assertion -- see
the note in the following paragraph) is also likely to have
spoken truly (or likely will speak truly).
In the raven example, the inductive specification is the
argument that a great many (actually, in this example, all)
ravens were black, therefore, some individual raven was
black (or perhaps the next observed individual raven will
be black -- again, see the note in the following paragraph).
30
Note that inductive analogies are a basic form of
argument for making predictions based on
empirical evidence -- predictions, for example,
about the next person we encounter who will make
a factual assertion, or the next raven we will see.
The abstract form of the argument is this:
(1 through  n ) 1 through n have all been both  and 
(i.e., has both characteristics,  and ).
[e.g., Person A through Person N all made a
factual assertion and spoke truly.]
[e.g., Bird A through Bird N all were ravens
and black.]
(n+1) There were few observed instances of an  that was
 and also was not-
[e.g., There were few Persons who made a
factual assertion and did not speak truly.]
[e.g. No ravens were observed to be nonblack.]
Therefore, h: Some individual n+1 probably has both 
and .
Some Person (perhaps some person we
encounter in the future) who makes a
factual assertion probably spoke (or
probably will speak) truly.
Some Bird (perhaps some bird we
encounter in the future) who is a raven
probably is black.
31
Characteristic common to
specification
inductive generalization and
inductive
In order for to assess how convincing an inductive inference is,
one must assess the premises or conclusion according to several
criteria (cf. Steven Barker, Elements of Logic (5th ed. p. 187)):
 the number of observed instances
 the degree of shared characteristics among the
identified characteristics
 the degree of unshared characteristics among the
identified characteristics
 the logical strength of the conclusion ("all," "some,"
"probably," "very likely" etc.)
 the explanatory relations among the identified
characteristics
Note that probabilistic judgments ("probably," "almost certainly,"
"more likely than not," etc.) are always relative to some evidence.
See Barker, op. cit. p. 184: "[P]robability when understood as
rational credibility is a relative matter . . . . The very same
conjecture takes on different degrees of probability relative to
different amounts of evidence."
Note also that there is (or should be) a close relation between our
judgment regarding how probable we think a conclusion is relative
to the evidence offered in the inductive argument and the logical
strength of the conclusion (e.g., the conclusion "All individuals
that have property  have property " is logically stronger than
the conclusion "Some individuals that have property  have
property "; in the examples above, the conclusion in raven
induction was logically stronger than the conclusion in the Knapp
induction).
Although it may be counter-intuitive, note that the greater the
logical strength of the conclusion of an inductive argument, the
lower is the probability of that conclusion relative to the evidence
stated in the argument's premises. (Logician Stephen Barker
makes this point as follows: "The more sweeping the
generalization that we seek to establish, the less is its probability
relative to our evidence." (see Barker op. cit. p. 187)). Do you see
why this is true?
32
Some uses of inductive inference in legal argument
Using induction to find the law (the lawyer) -- Holmes' bad
man
"Take the fundamental question, What constitutes the law?
You will find some text writers telling you that it is
something different from what is decided by the courts of
Massachusetts or England, that it is a system of reason,
that it is a deduction from principles of ethics or admitted
axioms or what not, which may or may not coincide with
the decisions. But if we take the view of our friend the bad
man we shall find that he does not care two straws for the
axioms or deductions, but that he does want to know what
the Massachusetts or English courts are likely to do in fact.
I am much of his mind. The prophecies of what the courts
will do in fact, and nothing more pretentious, are what I
mean by the law." [The Path of the Law]
Using induction to make the law (the lawyer, the judge, the
legislator, the administrative agent, the constitutional
adopter): Dewey on the practical, experiential, inductive
nature of "logical" method -- in law and elsewhere
"If we recur then to our introductory conception that logic
is really a theory about empirical phenomena, subject
to growth and improvement like any other empirical
discipline, we recur to it with, an added conviction:
namely, that the issue is not a purely speculative one, but
implies consequences vastly significant for practice. I
should indeed not hesitate to assert that the
sanctification of ready-made antecedent universal
principles as methods of thinking is the chief obstacle
to the kind of thinking which is the indispensable
prerequisite of steady, secure and intelligent social
reforms in general and social advance by means of law
in particular. If this be so infiltration into law of a
more experimental and flexible logic is a social as well
as an intellectual need." [Logical Method and the Law 26]
"If we trust to an experimental logic, we find that
general principles emerge as statements of generic
33
ways in which it has been found helpful to treat
concrete cases. The real force of the proposition that all
men are mortal is found in the expectancy tables of
insurance companies, which with their accompanying rates
show how it is prudent and socially useful to deal with
human mortality. The 'universal' stated in the major
premise is not outside of and antecedent to particular
cases; neither is it a selection of something found in a
variety of cases. It is an indication of a single way of
treating cases for certain purposes or consequences in
spite of their diversity." [Logical Method and the Law 22]
Question for Dewey: Is that true of 'F = MA' as well?
Using induction to find "legislative facts" – recall "Dewey's
Dream": Brown v. Board of Education, 347 U.S. 483, 493-95
(1954)
"We come then to the question presented: Does
segregation of children in public schools solely on the
basis of race, even though the physical facilities and
other 'tangible' factors may be equal, deprive the
children of the minority group of equal educational
opportunities? We believe that it does. . . . To separate
[Negro schoolchildren] from others of similar age and
qualifications solely because of their race generates a
feeling of inferiority as to their status in the community
that may affect their hearts and minds in a way unlikely
ever to be undone. The effect of this separation on their
educational opportunities was well stated by a finding in
the Kansas case by a court which nevertheless felt
compelled to rule against the Negro plaintiffs: 'Segregation
of white and colored children in public schools has a
detrimental effect upon the colored children. The impact is
greater when it has the sanction of the law; for the policy
of separating the races is usually interpreted as denoting
the inferiority of the negro group. A sense of inferiority
affects the motivation of a child to learn. Segregation with
the sanction of law, therefore, has a tendency to (retard)
the educational and mental development of Negro children
and to deprive them of some of the benefits they would
receive in a racial(ly) integrated school system.' . . . .
34
Whatever may have been the extent of psychological
knowledge at the time of Plessy v. Ferguson, this
finding is amply supported by modern authority.
[FN11] Any language in Plessy v. Ferguson contrary to
this finding is rejected.
FN11. K. B. Clark, Effect of Prejudice and
Discrimination on Personality Development
(Midcentury White House Conference on Children
and Youth,
1950); Witmer and Kotinsky,
Personality in the Making (1952), c. VI; Deutscher
and Chein, The Psychological Effects of Enforced
Segregation: ASurvey of Social Science Opinion,
26 J.Psychol. 259 (1948); Chein, What are the
Psychological Effects of Segregation Under
Conditions of Equal Facilities?, 3 Int. J. Opinion
and Attitude Res. 229 (1949); Brameld,
Educational Costs, in Discrimination and National
Welfare (MacIver, ed.,1949), 44--48; Frazier, The
Negro in the United States (1949), 674--681. And
see generally Myrdal, An American Dilemma
(1944).
We conclude that in the field of public education the
doctrine of 'separate but equal' has no place. Separate
educational facilities are inherently unequal. Therefore, we
hold that the plaintiffs and others similarly situated for
whom the actions have been brought are, by reason of the
segregation complained of, deprived of the equal
protection of the laws guaranteed by the Fourteenth
Amendment. This disposition makes unnecessary any
discussion whether such segregation also violates the Due
Process Clause of the Fourteenth Amendment. . . . .
Is there no deductive reasoning in legal reasoning?
Cohen's challenge in The Ethical Basis of Legal
Criticism pages 216-219 (1931)
But elementary logic teaches us that every legal decision
and every finite set of decisions can be subsumed under an
35
infinite number of different general rules, just as an infinite
number of different curves may be traced through any point or
finite collection of points. Every decision is a choice between
different rules which logically fit all past decisions but
logically dictate conflicting results in the instant case. Logic
provides the springboard but it does not guarantee the success
of any particular dive.
If the doctrine of stare decisis means anything, and one can
hardly maintain the contrary despite the infelicitous formula tions which have been given to the doctrine, the consistency
which it demands cannot be a logical consistency. The consistency in question is more akin to that quality of dough which is
necessary for the fixing of a durable shape. Decisions are
fluid
particular decision. See WAMBAUGH, STUDY OF CASES (2d ed. 1894) c. 2; Salmond,
T h e o r y o f J u d i c i a l P r e c e d e n t s (1900) 16 L. Q. REV. 376; GRAY, op. cit. supra
note 11, at § 555; BLACK, JUDICIAL PRECEDENTS (1912) 40; MORGAN, op. cit. supra
note 29, at 109-10; Goodhart, Determining the Ratio Decidendi of a Case (1930) 40
YALE L. J. 161. Logical objections to this conception are dismissed by Professor Morgan as
"hypercritical" and 'too refined for practical purposes." But Professor Oliphant, who
refuses to be deterred by such warnings (see his reply in M u t u a l i t y o f Ob l i g a tion in
Bilateral Contracts at Law (1928) 28 COL. L. REV. 997 n. 2 to Professor Williston's
charges of scholasticism, The Effect of One Void -Promise in a Bilateral Agreement
(1925) 25 COL. L. REV. 857, 869) has suggested an alternative conception that is logically
sound and practically far more useful. Rules of increasing generality, each of them linking
the given result to the given facts, spread pyramid-wise from a decision. The possibility of
alternative modes of anaylsis makes a decision the apex not of one but many such
pyramids. No one of these rules has any logical priority; courts and lawyers choose among
competing propositions on extra-logical grounds. Oliphant, A Return to Stare Decisis
(1928) 6 AM. LAW SCHOOL REV. 215, 217-18; and cf. LLEWELLYN, BRAMBLE BUSH
(1930) 61-66; Bingham, What is the Law? (1912) 11 MICH. L. REV. 1, 109, 111 n. 31.
The picture clearly suggests that the decision bears to the rules the same relation that
Professor Whitehead has traced between a point and the surfaces that would ordinarily be
said to include the point. See THE PRINCIPLES OF NATURAL KNOWLEDGE (1919) c. 8;
THE CONCEPT OF NATURE (1919) c. 4.
40 Loc. cit. supra, note 39.
36
1931]
LEGAL CRITICISM
217
until they are given "morals." It is often important to conserve with new
obeisance the morals which lawyers and laymen have read into past decisions
and in reliance upon which they have acted. We do not deny that importance
when we recognize that with equal logical justification lawyers and laymen
might have attached other morals to the old cases had their habits of legal
classification or their general social premises been different. But we do shift the
focus of our vision from a stage where social and professional prejudices wear
the terrible armor of Pure Reason to an arena where human hopes and
expectations wrestle naked for supremacy.
No doubt the doctrine of stare decisis and the argument for consistency
have a significance which is not exhausted by the social usefulness of
predictable law. Even in fields where past court decisions play a negligible
role in molding expectations, courts may be justified in looking to former
rulings for guidance. The time of judges is more limited than the boundaries of
injustice. At some risk the results of past deliberation in a case similar to the
case at bar must be accepted. But again we invite fatal confusion if we think of
this similarity as a logical rather than an ethical relation. To the cold eyes of
logic the difference between the names of the parties in the two decisions bulks
as large as the difference between care and negligence. The question before
the judge is, "Granted that there are differences between the cited precedent
and the case at bar, and assuming that the decision in the earlier case was a
desirable one, is it desirable to attach legal weight to any of the factual differences
between the instant case and the earlier case?" Obviously this is an ethical
question. Should a rich woman accused of larceny receive the same treatment as
a poor woman? Should a rich man who has accidentally injured another come
under the same obligations as a poor man? Should a group of persons, e. g., an
unincorporated labor union, be privileged to make all statements that an individual
may lawfully make? Neither the ringing hexameters of Barbara Celarent nor
the logic machine of Jevons nor the true-false patterns of Wittgenstein will
produce answers to these questions.
What then shall we think of attempts to frame practical legal issues as conflicts
between morality, common sense, history or sociology, and logic (logic playing
regularly the Satanic role) ? One hesitates to convict the foremost jurists on
the American bench of elementary logical error. It is more likely that they
have simply used the word "logic" in peculiar ways, as to which they may find
many precedents in the current logic textbooks.41
See M. R. Cohen, The Subject Matter of Formal Logic (1918) 15 JOUR.
OF PHIL. 673.
4'
37
218
YALE LAW JOURNAL
[Vol. 41
Bertrand Russell has warned us :
"When it is said, for example, that the French are 'logical',
what is meant is that, when they accept a premise, they also accept everything that a person totally devoid of logical subtlety
would erroneously suppose to follow from that premise. . . .
Logic was, formerly, the art of drawing inferences ; it has now
become the art of abstaining from inferences, since it has appeared that the inferences we feel naturally inclined to make are
hardly ever valid." 42
If we construe the word "logic" in the light of this warning,
we may readily agree with Mr. Justice Holmes when he
asserts that "the whole outline of the law is the resultant of a
conflict at every point between logic [viz. hasty generalization]
and good sense",43 and find some meaning in the statement of
Judge. Cardozo that "the logic of one principle" prevails over the
logic of another 44 or in his pride that "We in the United States have
been readier to subordinate logic to utility."
13
42
RUSSELL, SKEPTICAL ESSAYS (1928) c. 7 (Behaviorism and Values)
HOLMES, COLLECTED LEGAL PAPERS (1920) (Agency) 49, 50.
CARDOZO, NATURE OF THE JUD ICIAL PROCESS, (1921) C. 1 (Introduction.
99.
The
Method of Philosophy) 41. Judge Cardozo illustrates (op. cit. 38-39) the method of logic
or philosophy, which is distinguished from the methods of history or evolution, of custom
or tradition, and of sociology, with the rule that one who contracts to purchase real
property must pay for it even though, before the sale is actually completed, the property
is substantially destroyed. This, he maintains, is the projection to its logical outcome of
the principle that "equity treats that as done which ought to be done," a principle which
does not apply to the sale of chattels which did not come under the jurisdiction of
Chancery. But what sort of princip le is this? It is certainly not a logical principle, i.e.,
a proposition certifiable on logical grounds alone, since it is obviously false. If it were
true no plaintiff in equity could ever obtain a judgment since he could never in the face
of such a rule show that the defendant had not done what he ought to have done. Would
it not be quite as logical for a court to say "equity does not treat that as done which
has not been done"? If a rule is undesirable we do not make it less undesirable by
deducing it from another rule too vague to be liked or disliked and then concentrating our attention on the process of inference rather than the premise. What is in question in the case
proposed is not a logical probleim or a choice of judicial methods but a conflict of social
interests, and there is much that may be said in favor of throwing upon the party who
contemplates future enjoyment of a definite piece of real property the risk of its destruction
and the necessity of insurance. But what may thus be said bears no peculiar imprimcitur of
logic. See also CARDOZO, THE GROWTH OF THE LAW (1924) 79-80.
44
C AR D O Z O , T H E G R O W T H O F T H E LAW ( 1 9 2 4 ) 7 7 . This is said with regard to the
tendency in recent decisions (of which Judge Cardozo's opinion in MacPherson v. Buick
Mfg. Co., 217 N. Y. 382, 111 N. E. 1050 (1916) is a noteworthy landmark) to extend
the scope of a manufacturer's obligations to the ultimate consumer with regard to the
quality of the product.
4
'
38
19311
LEGAL CRITICISM
219
We may have to interpret the word "logical" as synonymous with
"aesthetically satisfying" in order to understand the statement of Mr. Justice
Brandeis and Mr. Warren that a distinction between cases where "substantial mental
suffering would be the natural and probable result" of an act and cases "where no
mental suffering would ordinarily result" is not logical though very practical.46 Such
an identification of the rules of logic with those of intellectual aesthetics seems to
be assumed at times by Judge Cardozo as well.47
No verbal definition is intrinsically objectionable. But it seems fair to
suggest that the use of the word "logic" in the senses exemplified in these
typical passages seriously lowers the probability of clear thinking on the
relation between law and ethics. Most of us think of logic as the most general
and formal of the sciences.48 Upon that basis we may say, paraphrasing a remark
of Mr. Justice Holmes, that conformity with logic is only a necessity and not a
duty. The bad judge is no more able to violate the laws of logic than he is to violate
the laws of gravitation. He may, of course, ignore both. It is not our purpose
to deny that there would be less judicial stumbling were courts more constantly
aware of the logical relations between particular and universal, between premise
and conclusion, between form and content.
IV
The theory which denies ethical justiciability to law, in whole
or in part, cannot be maintained. Its superficial plausibility
Again the rejected "privity" analysis of the situation seems to be peculiarly "logical" because
it permits the deduction of an undesirable rule from another undesirable rule which is too
vague to arouse the resentment which the deduced rule arouses.
See also ibid. 83, where "adherence to logical and advancement of utility" are balanced in
terms of "the social interest which each is capable of promoting."
' , Warren and Brandeis , The Right of Privacy (1890) 4 HARV. L. REV. 193,
reprinted in SELECTED ESSAYS IN THE LAW OF TORTS (1924) 122, 126.
4
"If I am seeking logical consistency, the symmetry of the legal struc ture, how far shall I
seek it?" CARDOZO, NATURE OF THE JUDICIAL PROCESS 10, and cf. ibid. 33-34.
"If it was so, it might be; and if it were so, it would be, but as it isn't, it ain't. That's
logic." CARROLL, THROUGH THE LOOKING GLASS C. 4. And see WITTGENSTEIN,
TRACTATUS LOGICO-PHILOSOPHICUS (1922)1 §§ 6.1, 6.1262; M. R. Cohen, op. cit. supra note
41; Hoernle, Review of SCIENCE OF LEGAL METHOD (1918) 31 HARV. L. REV. 807; Russell,
PRINCIPLES OF MATHEMATICS, (1 9 0 3 ) c . 1 ; Ad ler, La w a n d t h e M o d e r n M i n d : A
S y mp o s i u m (1 9 3 1 ) 31 CoL. L. REV. 99-101; Keyser, On the Study of Legal Science
(1929) 38 YALE L. J. 413.
47
48
220
YALE LAW JOURNAL
[Vol. 41
arises from the narrow connotation given to the terms ethics and morality when they are extruded from the field
of legal criticism. The falsity of the theory arises from the fact that, along with the promptings of
"conscience", the principal values of life are banished from the juristic consciousness and an inadequate
"practical" ethics substituted. The invalidity of the inference by which this theory is established arises
from the fallacy (guaternio terminorum) by which the extrusion from legal criticism of "ethics" in its
broadest sense is inferred from a denial of its legal importance in its narrower connotation. Finally, the
confusion of the theory lies in the indeterminate character of the system of values substituted by our
jurists for what they call "ethics" and "morality."
Law is just as much a part of the domain of morality as any other phase of human custom and conduct. It
has no special purpose, end, or function, no restriction of moral scope, other than that variable restriction
which its positive and practical nature may impose in the way of limitations of efficacy and applicability.
We may, if we like, call the good which law can achieve "justice." But if "justice" means anything
less than that total, it is not a valid basis of legal criticism. To say that something or other is beyond the
"proper scope" of law is either to say that law on that subject will bring about more harm than good or it is to
indulge in meaningless verbiage. The evaluation of law must be made in terms of the good life, and to
demonstrate the nature of this standard is the task of ethics, and more particularly, of morality. Difficult as that
task is and uncertain as its conclusions have been, it is a vicious illusion to suppose that the task of
statesman or judge is less difficult, or that his conclusions can be more certain.