Remarks on the Theory of Appellate Decisions

and the
Ruled or Canons about how Statutes are to be Construed

1950 Vanderbilt Law Review , v. 3, p. 395, by Karl N. Llewellyn

I
One does not progress far into legal life .without learning that there is no single right and
accurate way of reading one case, or of reading a bunch of cases. For
(1) Impeccable and correct doctrine makes clear that a case holds with authority only
so much of what the opinion says as is absolutely necessary to sustain the judgment. Anything
else is unnecessary and distinguishable and non-controlling for the future. Indeed, if the
judgement rests on two, three or four rulings, any of them can be rightly and righteously knocked
out, for the future, as. being thus unnecessary. Moreover , any distinction on the facts is
rightly and righteously a reason for distinguishing and therefore disregarding the prior alleged
holding. But
(2) Doctrine equally impeccable anf correct makes clear that a case holds with
authority the rule on which the court there chose to rest the judgment; more, that that the rule
covers, with full authority, cases which are plainly distinguishable on their facts and their issue,
whenever the reason for the rule extends to cover them. Indeed, it is unnecessary for a rule or
principle to have led to the decision in the prior case, or even to have been phrased therein, in
order to be seen as controlling in the new case: (a) We there said. . . (b) That case necessarily
decided. . .
These divergent and indeed conflicting correct ways of handling or reading a single prior
case as one determines what it authoritatively holds, have their counterparts in regard to the
authority of a series or body of cases. Thus
(1) It is correct to see thatThat rule is too well settled in this jurisdiction to be
disturbed; and so apply it to a wholly novel circumstance . But,
(2) It is no less correct to see that The rule has never been extended to a case like the
present; and so to refuse to apply it: We here limit the rule. Again,
(3) It is no less correct to look over the prior applications of the rule and rework them
into a wholly new formulation of the true rule or [396] true principle which knocks out some
of the prior cases as simply mis-applications and then builds up the others.
In the work of a single opinion-day I have observed 26 different, describable ways in
which one of our best state courts handled its own prior cases, repeatedly using three to six
different ways within a single opinion.
What is important is that all 26 ways (plus a dozen others which happened not to be in
use that day) are correct. They represent not evasion, but sound use, application and
development of precedent. They represent not departure from, but sound continuation of, our
system of precedent as it has come down to us. The major defect in that system is a mistaken
idea which many lawyers have about itto wit, the idea that the cases themselves and in
themselves, plus the correct rules on how to handle cases, provide one single correct answer to a
disputed issue of law. In fact the available correct answers are two, three, or ten. The question is:
Which of the available correct answers will the court selectand why? For, since there is
always more than one available correct answer, the court always has to select.

1
 True, the selection is frequently almost automatic. The type of distinction or expansion
which is always technically available may be psychologically or sociologically unavailable. This
may be because of (a) the current tradition of the court or because of (b) the current temper of the
court or because of (c) the sense of the situation as the court sees that sense. (There are other
possible reasons a-plenty, but these three are the most frequent and commonly the most
weighty.)
The current tradition of the court is a matter of period-style in the craft of judging. In
1820-1850 our courts felt in general a freedom and duty to move in the manner typified in our
thought by Mansfield and Marshall. Precedent guided,. but principle controlled; and nothing
was good Principle which did not look like wisdom-in-result for the welfare of All-of-us. In
1880-1910, on the other hand, our courts felt in general a prime duty to order within the law and
a duty to resist any outside influence. Precedent was to control, not merely to guide;
Principle was to be tested by whether it made for order in the law, not by whether it made
wisdom-in-result. Legal Principle could not be subjected to political tests; even legislation
was resisted as disturbing. Since 1920 the earlier style (the Grand Style) has been working its
way back into general use by our courts, though the language of the opinions moves still
dominantly (though waningly) in the style (the Formal Style) of the late 19th Century. In any
particular court what needs study is how far along the process has gotten. The best material for
study is the latest volume of reports, read in sequence from page 1 through to the end: the cur
rent mine-run [run of the mill cases] of the work.
The current tetnper of the court is reflected in the same material, and represents the
courts tradition as modified by its personnel. For it is plain [397] that the two earlier period-
styles represent also two eternal types of human being. There is the man who loves creativeness,
who can without loss of sleep combine risk-taking with responsibility, who sees and feels
institutions as things built and to be built to serve functions, and who sees the functions as vital
and law as a tool to be eternally reoriented to justice and to general welfare. There is the other
man who loves order, who finds risk uncomfortable and has seen so much irresponsible or
unwise innovation that responsibility to him means caution, who sees and feels institutions as the
tested, slow-built ways which for all their faults are mans sole safeguard against relapse into
barbarism, and who regards reorientation of the law our polity as essentially committed to the
legislature. Commonly a man of such temper has also a craftsmans pride in clean craftsmans
work, and commonly he does not view with too much sympathy any ill-done legislative job of
attempted reorientation.1 Judges, like other men, range up and down the scale between the
extremes of either type of temper, and in this aspect (as in the aspect of intellectual power and
acumen or of personal force or persuasiveness) the constellation of the personnel on a particular
bench at a particular time plays its important part in urging the court toward a more literal or a
more creative selection among the available accepted and correct ways of handling precedent.
More vital, if possible, than either of the above is the sense of the situation as seen by the
court. Thus in the very heyday of the formal period our courts moved into tremendous creative

1
Intellectually. this last attitude is at odds with the idea that reorientation is for legislature. Emotiontly. it isnt. Apart
from the rather general resistance to change which normally companions orderliness of mind, there is a legitimate feeling that
within a team, team-play is called for, that it is passing the buck to thrust onto a court the labor of making a legislative job make
sense and becom e workable.

2
expansion of precedent in regard to the labor injunction and the due process clause. What they
saw as sense to be achieved, and desperately needed, there broke through all trammels of the
current period-style. Whereas the most creative-minded court working in the most creative
period-style will happily and literally apply a formula without discussion, and even with relief, if
the formula makes sense and yields justice in the situation and the case.
So strongly does the felt sense of the situation and the case affect the courts choice of
techniques for reading or interpreting and then applying the authorities that one may fairly lay
down certain generalizations:
A. In some six appealed cases out of ten the court feels this sense so clearly that lining up
the authorities comes close to being an automatic job. In the very process of reading an
authority a distinction leaps to the eye, and that is all that that case holds; or the language of
another authority (whether or not really in point) shines forth as clearly stating the true rule.
Trouble comes when the cases do not line up this clearly and semi- [398] automatically, when
they therefore call for intellectual labor, even at times for a conclusion that the law as given will
not allow the sensible result to be reached. Or trouble comes when the sense of the situation is
hot clear.
B. Technical leeways correctly available when the sense of the situation and the case call
for their use cease to be correctly available unless used furtherance of what the court sees as
such sense. There is here in our system of precedent an element of uprightness, or conscience, of
judicial responsibility; and motive becomes a factor in determining what techniques are correct
and right. Today, in contrast with 1890, it may be fairly stated that even the literal application of
a thoroughly established rule is not correct in a case or situation in which that application does
not make sense unless the court in honest conscience feels forced by its office to make the
application.
C. Collateral to B, but deserving of separate statement, is the proposition that the greater
the felt need, because of felt sense, the wider is the leeway .correctly and properly available in
reshaping an authority or the authorities; What is both proper and to be expected in an extreme
case would become abuse and judicial usurpation if made daily practice in the mine-run of cases.
All courts worthy of their office feel this in their bones, as being inherent in our system of
precedent. They show the feeling in their work. Where differences appear is where they should
appear: in divergent sizing up of what is sense, and of how great the need may be in any
situation.
One last thing remains to be said about sense. There is a sense of the type of situation
to be contrasted with the sense of a particular controversy between particular litigants. Which of
these aspects of sense a court responds to more strongly makes a tremendous difference.
Response primarily to the sense of the particular controversy is, in the first place, dangerous
because a particular controversy may not be typical, and because it is hard to disentangle general
sense from personalities and from fireside equities. Such response is dangerous in the second
place because it leads readily to finding an out for this case onlyand that leads to a c. plicating
multiplicity of refinement and distinction, as also to repeated resort to analogies unthought
through and unfortunate of extension. This is what the proverb seeks to say: Hard cases make
bad law.
If on the other hand the type of situation is in the forefront of attention, a solving rule
comes in for much more thoughtful testing and study. Rules are thrust toward reasonable
simplicity, and made with broader vision. Moreover, the idiosyncracies of the particular case

3
and its possible emotional deflections are set for judgment against a broader picture which gives
a fair chance that accidental sympathy is not mistaken for long-range justice for all. And one
runs a better chance of skirting the incidence of the other proverb: Bad law makes hard cases.
On the case-law side, I repeat, we ought all thus to be familiar with [399] the fact that the
right doctrine and going practice of our highest courts leave them a very real leeway within
which (a) to narrow or avoid what seem today to have been unfortunate prior phrasings or even
rulings; or (b), on the other hand, to pick up, develop, expand what seem today to have been
fortunate prior fillings or even phrasings.
It is silly, I repeat, to think of use of this leeway as involving twisting of precedent. The
very phrase presupposes the thing which is not and which has never been. The phrase
presupposes that there was in the precedent under consideration some one and single meaning.
The whole experience of our case-law shows that that assumption is false. It is, instead, the
business of the courts to use the precedents constantly to make the law always a little better, to
correct old mistakes, to re-correct mistaken or ill-advised at tempts at correctionbut always
within limits severely set not only by the precedents, but equally by the traditions of right
conduct in judicial office.
What we need to see now is that all of this is paralleled, in regard to statutes, because of
(1) the power of the legislature both to choose policy and to select measures; and (2) the
necessity that the legislature shall, in so doing, use languagelanguage fixed in particular
words; and ( 3) the continuing duty of the courts to make sense, under and within the law.
For just as prior courts can have been skillful or unskillful, clear or unclear, or wise or
unwise, so can legislatures. And just as prior courts have been looking at only a single piece of
our whole law at a time, so have legislatures.
But a court must strive to make sense as a whole out of our law as a whole. It must, to
use Franks figure,2 take the music of any statute as written by the legislature; it must take the
text of the play as written by the legislature. But there are many ways to play that music, to play
that play, and a courts duty is to play it well, and in harmony with the other music of the legal
system.
Hence, in the field of statutory construction also, there are correct, unchallengeable
rules of how to read which lead in happily variant directions.
This must be so until courts recognize that here, as in case-law, the real guide is Sense-
for-all-of-Us. It must be so, so long as we and the courts pretend that there has been only one
single correct answer possible. Until we give up that foolish pretense there must be a set of
mutually contradictory correct rules on How to Construe Statutes: either set available as duty
and sense may require.
Until then, also, the problem will recur in statutory construction as in the handling of
case-law: Which of the technically correct answers (a) should be given; (b) will be given and
Why?
And everything said above about the temper of the court, the temper [400] of the courts
tradition, the sense of the situation and the case, applies here as well.
Thus in the period of the Grand Style of case-law statutes were construed freely to
implement their purpose, the court commonly accepting the legislatures choice of policy and

2
Frank, Words and Music: Some Remarks on Statutory Interpretation, 47 Col. L. Rev. 1259 (1947).

4
setting to work to implement it. (Criminal statutes and, to some extent, statutes on procedure,
were exceptions.) Whereas in the Form1 Period statutes tended to be limited or even eviscerated
by wooden and literal reading, in a sort of long-drawn battle between a balky, stiff-necked,
wrong-headed court and a legislature which had only words with which to drive that court.
Today the courts have regained, in the main, a cheerful acceptance of legislative choice of
policy, but they are still hampered to some extent in carrying such policies forward by the
Formal Periods insistence on precise language.
II
One last thing is to be noted:
If a statute is to make sense, it must be read in the light of some assumed purpose. A
statute merely declaring a rule, with no purpose or objective, is nonsense.
If a statute is to be merged into a going system of law, moreover, the ourt must do the
merging, and must in so doing take account of the policy of the statuteor else substitute its own
version of such policy. Creative re. shaping of the net result is thus inevitable.
But the policy of a statute is of two wholly different kindseach kind somewhat limited
in effect by the statutes choice of measures, and by the statutes choice of fixed language. On
the one hand there are the ideas consciously before the draftsmen, the committee, the legislature:
a known evil to be cured, a known goal to be attained, a deliberate choice of one line of approach
rather than another. Here talk of intent is reasonably realistic; committee reports, legislative
debate, historical knowledge of contemporary thinking or campaigning which points up the evil
or the goal can have significance.
But on the other handand increasingly as a statute gains in ageits language is called
upon to deal with circumstances utterly uncontemplated at the time of its passage. Here the quest
is not properly for the sense originally intended by the statute, for the sense sought originally to
be put into it, but rather for the sense which can be quarried out of it in the light of the new
situation. Broad purposes can indeed reach far beyond details known or knowable at the time of
drafting. A dangerous weapon statute of 1840 can include tommy guns, tear gas or atomic
bombs. Vehicle, in a statute of 1840, can properly be read, when sense so suggests, to include
an automobile, or a hydroplane that lacks wheels. But for all that, the sound quest does not [401]
run primarily in terms of historical intent. It runs in terms of ,hat the words can be made to bear,
in making sense in the light of the unforeseen.
III
When it comes to presenting a proposed construction in court, there is an accepted
conventional vocabulary. As in argument over points of case-law, the accepted convention still,
unhappily requires discussion as if only one single correct meaning could exist. Hence there are
two opposing canons on al most every point. An arranged selection is appended. Every lawyer
must be familiar with them all : they are still needed tools of argument. At least as early as
Fortescue3 the general picture was clear, on this, to any eye which would see.
Plainly, to make any canon take hold in a particular instance, the construction contended
for must be sold, essentially, by means other than the use of the canon: The good sense of the

3
[Sir John Fortesque (13941476): English jurist. A supporter of the Lancastrian king Henry VI, he was
chief justice of the Court of Kings Bench from 1442 until 1461, when Henry was deposed by the Yorkist Edward IV.
Fortescue was attainted and fled to France with the royal family.]

5
situation and a simple construction of the available language to achieve that sense, by tenable
means, out of the statutory language.

Canons of Construction

Statutory interpretation still speaks a diplomatic tongue. Here is some of the technical
framework for maneuver. [401]

1. A statute cannot go beyond its text. First 1. To effect its purpose, a statute may be
National Bank v. DeBerriz, 105 S.E. 900 implemented beyond its text. Dooley v.
(W.Va. 1921) Penn. R.R., 250 Fed. 142 (D.Minn.
1918)
2. Statutes in derogation of the common law 2. Such acts will be liberally construed if their
will not be extended by construction. nature is remedial. Becker v. Brown, 91
Devers v. City of Scranton, 161 Atl.450 N.W. 178 (Neb. 1902)
(Penn.1932)
3. Statutes are to be read in the light of the 3. The common law gives way to a statute
common law and a statute affirming a which is consistent with it and when a
common law rule is to be construed in statute is designed as a revision of the
accordance with the common law. whole body of law applicable to a given
Bandfield v. Bandfield, 75 N.W. 287 subject it supercedes the common law.
(Mich. 1898) State v. Lewis, 55 S.E. 600 (N.C. 1906)
4. [402] Where a foreign statute which has 4. It may be rejected where there is conflict
received construction has been adopted with the obvious meaning of the statute
[by the local judiciary], previous or where the foreign decisions are
construction [by the foreign judiciary] is unsatisfactory in reasoning or where the
adopted too. Freese v. Tripp, 70 Ill. 496 foreign interpretation is not in harmony
(1873) with the spirit or policy of the laws of
the adopting state. Bowers v. Smith, 20
S.W. 101 (Mo. 1892)
5. Where various states have already adopted 5. Where interpretations of other states are
the statute, the parent state is followed. inharmonious, there is no such restraint.
Burnside v. Wand, 71 S.W. 337 (Mo. State v. Campbell, 85 Pac. 784 (Kan.
1902) 1906)
6. Statutes in para materia must be construed 6. A statute is not in pari materia if its scope
together. Milner v. Gibson, 61 S.W.2d and aim are distinct or where a
273 (Ky. 1933) legislative design to depart from the
general purpose or policy of previous
enactments may be apparent. Wheelock
v. Myers, 67 Pac. 632 (Kan. 1902)
7. A statute imposing a new penalty or 7. Remedial statutes are to be liberally
forfeiture, or a new liability or construed and if a retroactive
disability, or creating a new right of interpretation will promote the ends of
action will not be construed as having a justice, they should receive such
retroactive effect. Keeley v. Great

6
 Northern Ry., 121 N.W. 167 (Wis. construction. Falls v. Keys, 278 S.W.
1909) 893 (Tex. Civ. App. 1925)
8. Where design has been distinctly stated no 8. Courts have the power to inquire into real
place is left for construction. Federoff v. as distinct from ostensible purpose.
Birks Bros., 242 Pac. 885 (Cal. 1925) Coulter v. Pool, 201 Pac. 885 (Cal.
App. 1925)
9. [403] Definitions and rules of construction 9. Definitions and rules of construction in a
contained in an interpretation clause are statute will not be extended beyond
part of the law and binding. Smith v. their necessary import nor allowed to
State, 28 Ind. 321 (1867) defeat intention otherwise manifested.
(In re Bissell, 282 N.Y.Supp. 983 (4th
Dept. 1935)
10. A statutory provision requiring liberal 10. Where a rule of construction is provided
construction does not mean disregard of within the statute itself, the rule should
unequivocal requirements of the statute. be applied. State ex rel Triay v. Burr, 84
Los Angeles County v. Payne, 255 Pac. So. 61 (Fla. 1920)
281 (Cal. 1927)
11. Titles do not control meaning; preambles do 11. The title may be consulted as a guide when
not expand scope; section headings do there is doubt or obscurity in the body;
not change language. Westbrook v. preambles may be consulted to
McDonald, 44 S.W.2d 331 (Ark. 1931) determine rationale, and thus the true
construction of terms; section headings
may be looked upon as part of the
statute itself. Brown v. Robinson, 175
N.E. 269 (Mass. 1931); Gulley v.
Jackson, 145 So. 905 (Miss. 1933)
12. If language is plain and unambiguous, it 12. Not when literal interpretations would lead
must be given effect. Newhall v. to absurd or mischievous consequences
Sanger, 92 U.S. 761 (1875) or thwart manifest purpose. Clark v.
Murray, 41 P.2d 1042 (Kan. 1935)
13. Words and phrases which have received 13. Not if the statute clearly requires them to
judicial construction before enactment have a different meaning. Dixon v.
are to be understood according to that Robbins, 158 N.E. 63 (N.Y. 1927)
construction. Scholze v. Scholze, 2
Tenn. App. 80 (M.S. 1925)
14. [404] After enactment, judicial decision 14. Practical construction by executive officers
upon interpretation of particular terms is strong evidence of true meaning.
and phrases controls. Eau Claire State ex rel Bashford v. Frear, 120
National Bank v. Benson, 82 N.W. 604 N.W. 216 (Wis. 1909)
(Wis. 1900)
15. Words are to be taken in their ordinary 15. Popular words may bear a technical
meaning unless they are technical terms meaning and technical words may have
or words of art. Hawley Coal Co. v. a popular signification and they should
Bruce, 67 S.W.2d 703 (Ky. 1934) be so construed as to agree with evident
intention or to make the statute

7
 operative. Robinson v. Varnell, 16 Tex.
382 (1856)
16. Every word and clause must be given 16. If inadvertently inserted or if repugnant to
effect. In re Terrys Estate, 112 N.E. the rest of the statute, they may be
931 (N.Y. 1916) rejected as surplusage. U.S. v. York ,
131 Fed. 323 (C.C.S.D.N.Y. 1904)
17. The same language used repeatedly in the 17. This presumption will be disregarded where
same connection is presumed to bear it is necessary to assign different
the same meaning throughout the meanings to make the statute consistent.
statute. Spring Canyon Coal Co. v. State v. Knowles, 45 Atl. 877 (Md.
Industrial Commn, 277 Pac. 206 (Utah 1900)
1929)
18. Words are to be interpreted according to the 18. Rules of grammar will be disregarded
proper grammatical effect of their where strict adherence would defeat
arrangement within the statute. Harris purpose. Fisher v. Connard, 100 Pa. 63
v. Commonwealth, 128 S.E. 578 (Va. (1882)
1925)
19. Exceptions not made cannot be read. Lima 19. The letter is only the bark. Whatever is
v. Cemetery Assn., 42 Ohio St. 128 within the reason of the law is within
(1884) the law itself. Flynn v. Prudential Ins.
Co., 100 N.E. 794 (N.Y. 1913)
20. [405] Expression of one thing excludes 20. The language may fairly comprehend many
another. Detroit v. Redford Twp., 235 different cases where some only are
N.W. 217 (Mich. 1931) expressly mentioned by way of
example. Springer v. Philippine Island,
277 U.S. 189 (1928)
21. General terms are to receive a general 21. They may be limited by specific terms with
construction. DeWitt v. San Francisco, which they are associated or by the
2 Cal. 289 (1852) scope and purpose of the statute. People
ex rel Krause v. Harrison, 61 N.E. 99
(Ill. 1901)
22. (ejusdem generis) It is a general rule of 22. General words must operate on something.
construction that where general words Further, ejusdem generis is only an aid
follow an enumeration they are to be in getting the meaning and does not
held as applying only to persons and warrant confining the operations of a
things of the same general kind or class statute within the narrower limits than
specifically mentioned. Hull Hospital v. were intended. Texas v. U.S., 292 U.S.
Wheeler, 250 N.W. 637 (Iowa 1933) 522 (1934)
23. Qualifying or limiting words or clauses are 23. Not when evident sense and meaning
to be referred to the next preceding require a different construction. Myer v.
antecedent. Dunn v. Bryan, 299 Pac. Ada County, 293 Pac. 322 (Idaho 1930)
253 (Utah 1931) 24. Punctuation marks will not control the plain
24. Punctuation will govern when a statute is and evident meaning of language. State
open to two constructions. U.S. v. v. Baird, 288 Pac. 1 (Ariz. 1930)

8
 Marshall Field & Co., 18 C.C.P.A. 228
(1930)
25. It must be assumed that language has been 25. And and or may be read
chosen with due regard to grammatical interchangeably whenever the change is
propriety and is not interchangeable on necessary to give the statute sense and
mere conjecture. Hines v. Mills, 60 effect. Fulghum v. Bleakley, 181 S.E. 30
S.W.2d 181 (Ark. 1933) (S.C. 1935)
26. There is a distinction between words of 26. Words imparting permission may be read as
permission [may] and mandatory words mandatory and words imparting
[shall]. Koch & Dryfus v. Bridges, 45 command may be read as permissive
Miss. 247 (1871) when such construction is made
necessary by evident intention or by the
rights of the public. Jennings v. Suggs,
178 S.E. 282 (Ga. 1935)
27. A proviso qualifies the provision 27. It may clearly be intended to have a wider
immediately preceding. State ex rel. scope. Reuter v. San Mateo County, 30
Higgs v. Summers, 223 N.W. 957 (Neb. P.2d 417 (Cal. 1934)
1929)
28. When the enacting clause is general, a 28. Not when it is necessary to extend the
proviso is construed strictly. proviso to persons or cases which come
Montgomery v. Martin, 143 Atl. 505 within its equity. Forscht v. Green, 53
(Pa. 1928) Pa. 138 (1866)