7/28/2019 m Kramer
1/29
M AT T HE W H. KR AM E R T h ro win g L i g h t o n th e Ro l e o f Mo ral P rin ci p l es in th e Law
THROWING LIGHTON THEROLE
OFMORALPRINCIPLESIN THE
LAW:
FurtherReflections
Matthew H. Kramer
Churchill College, Cambridge
Inclusive Legal Positivism, as und erstood throughout this ar ticle, con sists in
the fo llowing thesis: It can be the case, though it need n ot be th e case, that
a norms consistency with some or all of the requirements of morality is a
precondition fo r the nor ms status as a law in this or th at jurisdiction. While
such a precondition for legal validity is no t inherent in th e concept of law,
it can be imposed as a threshold test und er the Rule of Recogn ition in an y
particular legal regime. That test, which can be applied by the officials in
such a regime to all of the legal no rms therein or to only some subset of
tho se no rms, is on e of the criteria th at th e officials use for ascertaining th e
law. Insofar as a criterion of that sort does prevail in any particular legal
system, then, some degree o f moral worth iness is a n ecessar y condition for
the legally authoritative force of each norm that is validated thereunder.
Inclusive Legal Positivism, which readily accepts the possibility of such a
state of affa irs, is inclusive because it a llows that moral precepts can figure
amo ng the criteria th at guide of ficials a scertainment of th e law. Inclusivisttheo rists reject th e view that every criterion of law-ascertainment in every
possible legal system is focused on n on no rmative matters of provenan ce. At
the same time, the Inclusivists are positivists because they also reject the
view that ever y possible lega l system includes moral tests amon g its law-as-
certaining criteria. An Inclusive Legal Positivist insists that such tests are
contingent features, rather than essential features, of the systems of law
wherein they are applied.
Incorpora tionism, as understood throughout this article, con sists in the
following thesis: It can be the case, though it need not be the case, that anorms correctness as a moral principle is a sufficient cond ition for its status
as a legal nor m in th is or th at jurisdiction. Albeit the role of mo ral correct-
ness as a sufficient condition for legal validity is no t inherent in the con cept
of law, it can ob tain a s such un der the Rule of Recognition in an y particular
legal regime. An Incorporationist theorist maintains that moral principles
115
Legal Theory, 8 (2002), 115143. Prin ted in t he U nited States of America
Copyright Cambridge University Press 13523252/02 $9.50
7/28/2019 m Kramer
2/29
recognized as legally determinative in this manner are indeed preexistent
legal nor ms, no twithstand ing th at they have perhaps never been laid d own
in any explicit sources such as legislative enactments or judicial rulings.
When officials do engage in a practice of treating th e mora l sound ness ofno rms as a sufficient cond ition for the norms legal authoritativeness, they
have thereby incorporated moral principles into their legal systemeven
before some or all of the applicable principles have received any explicit
recogn ition . An Incorpora tionist th eorist, who read ily accepts the possibil-
ity of such a state o f af fairs, is no netheless a legal positivist because he insists
that the incorporation of moral principles into a legal systems array of
norms is contingent rather than inevitable.
In a recent issue of this jour na l, I defen ded Inclusive Legal Positivism and
a version of Incorporation ism against some powerful objections leveled atthem by Scott Shapiro.1 Shapiro has now replied to that earlier essay of
mine, as has Jules Coleman from th e camp o f Incorporationism.2 In what
follows, I shall endeavor to parry their ripostes in order to clarify my
wholehearted defense of Inclusive Legal Positivism and my wary defense of
Incorporationism. Because Shapiro has emerged as the fo remost critic of
Incorporation ism and Inclusive Legal Positivism, an d because Coleman has
for years been the lead ing proponent of the Incorpo ration ist doctrine, their
arguments deserve very serious attention indeed.
I. HARTAND INCLUSIVELEGALPOSITIVISM
The first and fo remo st aim o f my earlier ar ticle was to rebut Shapiros claim
that the norms validated as laws under an Inclusive Rule of Recognition
cannot con stitute peremptor y reasons for decisions by officials and citizens.
Whereas any genuinely authoritative legal directives exclude their address-
ees from acting legitimately on considerations that run athwart th e direc-
tives terms, a man date validated by an Inclusive Rule of Recognition
requires its addressees to d eliberate abo ut the merits of giving ef fect to th e
mandate. So Shapiro has contended: The problem is that [any norm
validated by an Inclusive Rule of Recognition] cannot make a practical
difference in theway thatrules are supposed to make practical d ifferences:
if the agent is required to determine wheth er the rule ought to be fo llowed
on the merits, then it can count neither as an epistemic tool for authorita-
tive d esignation nor as a peremptor y reason for action (Shapiro, Way,
502, emphasis in original). In response to this line of argument, I main-
1. Matthew Kramer, How Moral Principles Can Enter into theLaw, 6 LEGAL THEORY 83 (2000).I was respon ding there largely to Scott Shap iro, On HartsWayOut, 4 LEGAL THEORY 469 (1998)(h ereinafter cited as Shapiro, Way). I wish to than k Wil Waluchow, Emran Mian, and JulesCo leman for some en lighten ing excha nges of views on the issues discussed herein.
2. Scott Shapiro, Law, Moral ity, and the Guidance of Conduct, 6 LEGAL THEORY 127 (2000)(hereinafter cited as Shapiro, Law); Jules Coleman, Constrain ts on the Criteri a of Legali ty, 6LEGAL THEORY 171(2000) (h ereinafter cited a s Coleman , C on straints).
116 MATTHEWH.KRAMER
7/28/2019 m Kramer
3/29
tained that a legal norm validated by an Inclusive Rule of Recognition can
perfectly well partake of peremptoriness. What cannot be true, of course, is
tha t the scope o f the peremptor y sway of such a n orm is all-encom passing.
H owever, restrictions on the scope o f a norms peremptoriness are familiarand straightforward and are not in any way at odds with the nature and
existence of tha t perempto riness.
Thus, for example, if a legal norm Nis valida ted und er an Inclusive Rule
of Recognition that withholds validation from any norm that is grossly
unfa ir, the scope of Ns peremptor y sway do es not encompass gross unfa ir-
ness but d oes encompass many oth er con siderations such a s ordinar y un-
fairness and unwisdom. No factors covered by the peremptor y reach of N
are permissible bases for decisions or actions that are contrary to Ns
requirements. Some oth er factors such a s gross unfairness fall outside thatreach and can thus warrant d ecisions or action s contrar y to Ns demands.
However, the upshot of such a state of affairs is not the inexistence of Ns
peremptoriness, but merely its limitedness in scope. Hence, despite
Shapiros ingenious arguments, a nor m validated as a law under an Inclu-
sive Rule of Recogn ition can make a practical difference in exactly the way
tha t an y authoritative rule is supposed to m ake a practical difference.
A. Peremptoriness and Applicability
Shapiro in his rejoinder does not contest the general correctness of my
defense of Inclusive Legal Positivism, but he declares that H .L.A. Hart could
not ha ve availed himself of my defense. When seeking to vind icate Inclusive
Lega l Positivism, I took for gran ted (arguendo) a conception o f authoritative
no rms as exclusiona r y reasons-for-actiona conception associated origi-
na lly with G eoffrey Warnock and Joseph Raz, and subsequently drawn upon
by Har t. I ha d presumed that Sha piro was relying on just such a conception
and that he was attributing such a conception to Hart. However, he nowsubmits that Harts account of normativity dif fers from the War-
no ckian/Razian accoun t in ways tha t leave H art un shielded b y my defense
of Inclusive Lega l P ositivism.
Und er the Warnockian/Razian analysis of autho ritative normativity, there
is ample room for restrictions on the scope of a norms peremptoriness.
Shapiro argues tha t H arts analysis, by contra st, ha s ruled out such restric-
tions. He contends that Hart viewed authoritative norms as directives that
put an end to all questions about whether conformity is ad visable
(Sha piro, Law, 165, emphasis add ed) . Cleaving to tha t view of n ormssoShapiro avouchesH art could n ot fa ll back upon a n otion of scope-restric-
tions in o rder to maintain th at th e man da tes validated by an Inclusive Rule
of Recognition are genuinely peremptory. H ence, Shapiro con cludes, my
vindication o f the Inclusivist stance does no t redound to Ha rts benefit at all.
Shapiro is alert to the principal difficulty that stands in the way of his
Throwing Light on the Role of Moral Principles in the Law 117
7/28/2019 m Kramer
4/29
renewed critique o f H art: th e wholesale implausibility of a conception of
legal mandates as directives whose peremptoriness is always unlimited in
scope. Surely, it would be argued , no rule has absolute force, tha t is, must
be followed no matter what the reason (Shapiro, Law, 165). One shouldcertainly shrink from attributing to Hart the outlandish view that legal
no rms are invariably end owed with absolute force. Shapiro af firms tha t the
appropriate tack here is to distinguish carefully between limitations on a
norms peremptoriness and limitations o n a norms applicability. H e sug-
gests tha t, although H art believed in th e unrestricted scope of an y authori-
tative norms peremptoriness, such a belief was accompanied by an
awareness that the cond itions for the applicability of a norm are typically
subject to a number of restrictions. Shapiro expresses this point as follows,
by indicating how H art would an alyze a situation in which a comman der hasord ered his troops to complete a ten-mile training h ike: The defeasibility
of a n orders requirements can be explained in terms of the ord ers appli-
cability conditions. The comman ders ord er car ries with it a set of implicit
exceptions, e.g., the troops are required to complete the 10 mile hike unless
they, or their command er, are in serious da nger. . . . There is, thus, no need
to explain the defeasibility of [n orms] by making reference to the techn ical
concept of a partially exclusionary scope. Such cases can easily be ex-
plained in terms of the familiar notion of applicability conditions
(Shapiro , Law, 16566).3
H aving d rawn this distinction between limits onperemptor iness and limits on applicability, Shapiro concludes that H art d id
not posit any limits of the former sort. Consequently, Shapiros original
critique does tell against Hart, its specific target: [S]ince peremptory
reasons [in H art s ana lysis as described by Shapiro] have un iversal scopes,
no one can deliberate about th e merits of following a rule and still treat the
rule as a peremptor y reason for action (Shapiro, Law, 166).
Now, although the chief purpose of my earlier article was to uphold
Inclusive Legal Positivism rather than to defend Hart, a few queries about
Shapiros argument ought to be raised here. First, the only passage thatShapiro cites in ord er to substantiate h is claim about Harts reliance on the
no tion of applicability-conditions is from The Concept of Law. That is, the
passage comes from H arts early work rath er than from his later work. What
makes this citation problematic is that Shapiro, both in his original article
and in his more recent piece, has emphasized that the conception of
norma tivity in the earlier work of H art is significantly divergent from the
con ception that is to be found in H ar ts later work (Shapiro, Way, 492-94;
Shapiro, Law, 162). Given the marked shift in Harts position on the
matt er over time, an extract from his early writings is no t a reliable ind icatorof his views at the much later juncture when his conception of norms as
peremptor y reason s-for-action gained expression. To know that H art was
3. Slightly od d, in my view, is the suggestion th at th e no tion o f ap plicability cond ition s isany less clearly a tech nical con cept tha n is the n otion of a pa rtially exclusionar y scope.
118 MATTHEWH.KRAMER
7/28/2019 m Kramer
5/29
still atta ched to the no tion of limited applicability-conditions in distinction
from limited peremptoriness, we would have to be presented with a ger-
man e passage from his later writingsa passage in which he gave voice to
tha t abiding atta chmen t. Shapiro offers no such passage, and (so far as I amaware) there is none.
Second, even if we leave aside the point just mad e, the examples on which
Shapiro trades are n ot really apposite. H e d iscusses Harts view of circum-
stan ces in which a comman der s ord er should not be followed or in which a
persons promise should not be kept. In oth er words, he discusses situat ions
tha t clearly involve the inapplicabilityof certa in peremptor y reason s-for-ac-
tion rather than the inexistencethereof. Und er an Inclusive Rule of Recogni-
tion, however, the very existence of legal nor ms as legal norms isexactly what
is at stake. Amanda te that does no t satisfy the moral test(s) laid down by sucha Rule of Recogn ition in some particular legal system is no nexistent therein
as a legally valid norm. Thus, if we are to flesh o ut H art s und erstan ding of
auth oritative norms in a way tha t is relevan t to the deba te over the tenability
of Inclusive Legal Positivism, we have to focus on circumstances where the
very existence of such no rms is in question . (We might, for instance, ponder
a situation in which someon e promises to commit a murder. What is in seri-
ous doubt is whether the promise has any morally binding force at all, rather
tha n merely whether it will fail to be applicable an d d etermina tive in some
settings.4
) When we con centra te specifically on the matter of the existence oflegal norms as such, we find tha t H art time and again observed that an Inclu-
sive Rule of Recogn ition d eals with precisely that matter; he d id not d iscuss
any such Rule of Recognition by reference to applicability-conditions at all.
In his early work, for example, he wrote tha t [i]n some systems, as in th e
United States, the ultimate criteria of legal validity explicitly incorporate
principles of justice or substantive moral values, and that any enactment
passed by American legislatures in violation of constitution al mora l tests is
liable to be treated as ul tra vi res, and declared lega lly invalid by the cour ts to
the extent th at it conflicts with the constitution al provisions.5
In his laterwriting, he chose virtually the same word s to reiterate his view: [ I]n some sys-
4. Hart h imself, in th e essay where he a nalyzed legal norms as content-indepen dent per-emptor y reasons for action, referred to th e kind o f situation th at I am envisaging. H is remarksquite plainly tally with my interpretation of his mature con ception of authoritative norms:[T]he giving of a promise is intended to be a reason not merely for the promisor doing theaction when the time comes but for excluding n ormal free deliberation abo ut the m erits ofdoing it. . . . This is true even though the range of po ssible actions which o ne may validlypromise to do is not unlimited and d oes not include grossly immoral actions or those intendedto be harmful to the promisee. H.L.A. Hart, Commands and AuthoritativeReasons, inESSAYS ON
BENTHAM 243, 255 (O xford : Clar end on Press, 1982).5. H.L.A. Ha rt, THE CONCEPT OF LAW 199, 70 (Oxford: Clarendon Press, 1961) (hereinafter
cited as Concept) . SeealsoH.L.A. Hart , Positivi sm and theSeparation of Law and Morals, inESSAYS INJURISPRU DENCE AND PHILOSOPHY49, 5455 (O xford : Clarend on Press, 1983). ( The essay just citedwas origina lly published in 71 HARV. L. REV. 593 [1958].) I am of course aware that laws in theUnited States are sometimes found to be unconstitutional as applied, rather than unconstitu-tional tout cour t. While highlighting the latter type of uncon stitutionalitythe type usuallyun derconsideration in the deba tes over Inclusive Legal PositivismI am in no wayd enying the form er.
Throwing Light on the Role of Moral Principles in the Law 119
7/28/2019 m Kramer
6/29
6. H.L.A. Hart, Postscript, in2 THE CONCEPTOF LAW238, 247 (O xford: Clarendon Press, 1994).
tems of law, as in the U nited States, the ultima te criteria of legal validity . . .
explicitly incorporate besides pedigree, principles of justice or substantive
moral values, and these may form the content of legal constitutional re-
straints.6
Whenever H art adverted to the centra l issue ad dressed by InclusiveLegal Positivism, he treated it not as an issue pertaining to limitations on
no rms applicability, but as an issue perta ining to limita tion s on the sheer ex-
istence of norm s and their peremptoriness. Shapiros exposition of H art s ac-
count o f authoritative normat ivity, with examples involving an unfo llowed
comm and and an unkept promise, has tended to obscure this point.
Third , and perhaps most importan t, even if we grant that H art in his later
work distinguished sha rply between limits on peremptoriness an d limits on
applicability, an d even if we gran t that h e took the peremptoriness of legal
no rms to be all-embracing, we sha ll have no grounds for accepting Shapirosconclusionhis conclusion, q uoted abo ve, that since peremptor y reason s
[in H art s analysis] h ave universal scopes, no one can deliberate about th e
merits of following a rule and still treat the rule as a perempto ry reason fo r
action (Shapiro, Law, 166). The very distinction Shapiro high lights, the
distinction between limits on applicability and limits on peremptoriness, is
what blocks his inference. Within a model of legal mandates as norms
end owed with all-enco mpa ssing perempto riness an d restricted applicability,
any deliberation about the merits of complying with a mandate will be
cha racterized as deliberation abo ut the applicability of the man da te in someor all sets of circumstances. Mora l tests laid d own by an Inclusive Rule o f
Recognition will be cha racterized as giving rise to deliberation o f exactly this
sort. To be sure, such a characterization is far from cogen t and illumina ting,
because the model of autho ritative normativity attributed to Hart by Shapiro
is itself far from cogent and illuminating. Nevertheless, even within the
unsatisfactory confines of tha t model, there is abund an t room for avoiding
an y eq ua tion between deliberat ing-on -th e-mer its-of-fo llowing-a-mandate-M
an d failing-to-trea t-M-as-a-peremptory-reason -for-action. Accord ing to the
aforementioned model, Mwill be trea ted as a peremptor y reason -for-actionwhenever it has been deemed to be applicable. According to tha t model, the
question to be an swered by deliberation abo ut the merits of fo llowing Mis
whether the cond itions for its applicability are ind eed present. Such delib-
eration, then, d oes not con stitute any incursion into Ms peremptoriness.
Ergo, even if we concede (arguendo) that Shapiros recountal of Ha rts ma-
ture con ception of autho ritative normativity is accurate, we should grasp
tha t H art could quite coherently leave open the possibility of Inclusivist
constraints on the bindingness of legal nor ms.
B. A Coda to the Critique
H aving submitted that H art in his later work perceived th e peremptoriness
of laws as unrestricted in scope, Shapiro proceeds to consider wheth er Hart
120 MATTHEWH.KRAMER
7/28/2019 m Kramer
7/29
could fruitfully have modified his position. He contends that such a move
would have been problematic:
According to Hart, the identity of a legal system is fixed by its rule ofrecogn ition. Because the ru le of recognition is a du ty-imposing r ule, it follows
tha t a rule is a member of a certain legal system on ly if a judge is und er a duty
to apply such a rule in a case that comes before her. H owever, if legal rules
were permitted to have partially exclusiona r y scopes, a judge might be under
a duty to apply a rule in some cases, but have discretion to apply the same
rule in other cases. The latter would obtain whenever the rule, even though
applicable, is defea ted by reasons tha t fell outside its (partially) exclusionar y
scope. The d uty of a judge to apply applicable ru les would thus be indetermi-
nateit would obtain in some cases and not in others. Accordingly, the
identity of any legal system would be indeterminate. (Shapiro, Law, 166.)
Though Shapiro goes on to avow that th e problem broached in this para-
graph is not insurmountable, we ought to query whether it is in fact a
problem at a ll. Let us no te two dubious aspects of h is argument.
First, when a puta tive legal norm d oes not pa ss the moral test(s) impo sed
by an Inclusive Rule of Recognition, it is not tr ue tha t a judge has discretion
to choose between applying and not applying the norm. Instead, he is
under an obligation to invalidate the norm or at least to abstain from
applying it. Accord ingly, the req uisite result in such circumstances is no t inany way indeterminateexcept, of course, when there is no determinate
answer to the question whether the ostensible legal mand ate satisfies the
relevan t mora l test(s). To be sure, when the latter q uestion does no t lend
itself to a uniquely correct answer, a judge in the Inclusivist legal system will
have genuine discretion to opt between applying and not applying the
ostensible mand ate. H owever, tha t discretion does no t stem from anything
peculiarly problematic abo ut limits on the peremptoriness of legal norms;
rath er, it stems from the occurrence of a h ard case in which a key point of
con tention is no t un ivocally resolvable. Any oth er d ifficult case tha t h ingeson an issue tha t is no t univocally resolvable (even th rough the in trod uction
of n on -source-based standard s) will similarly engend er a need fo r the exer-
cise of genuine discretion, regardless of whether the refractory issue is
connected with Inclusivist constraints. Contrariwise, when the question of
the fulfillment or no nfulfillment of an Inclusivist mora l test d oes lend itself
to a uniquely correct a nswer, a judge in the regime th at imposes the test
do es no t enjoy any discretion o f the sort described by Shapiro.
Second, also doub tful is Shapiros claim that th e ident ity of a legal system
would b e indeterminate as a consequence o f the indeterminacy that mightexist in some hard cases. On the one hand, it is generally true that a legal
system will witness some ha rd cases for which n o uniq uely correct resolu-
tions a re available even in principle. As a result, the law to be a scertained
in any particular legal system (In clusivist or non-Inclusivist) is not fully
determinate. On the other han d, th e indeterminacy tha t obtains is present
in respect of only a small propor tion of the interaction s amo ng people tha t
Throwing Light on the Role of Moral Principles in the Law 121
7/28/2019 m Kramer
8/29
might engender legal proceedings. Most such interactions never lead to
lawsuits at all, because the legal implications of the various modes of con-
duct are so clear-cut. Even amo ng the disputes tha t d o eventua te in legal
proceedin gs, most are straigh tfor wardly resolvab le. The d ifficult cases stud-ied by law studen ts and legal scho lars constitute only a small proportion of
the lawsuits that actually materialize and a still smaller proportion of the
lawsuits tha t could have ma terialized if people had been foolishly obstinate
enough to engage in them. Even some difficult cases lend themselves to
uniquely correct resolutions, and thus the instances of genuine indetermi-
nacy and strong discretion within a legal system are marginal. Although the
ha rd cases involving full-blown indeterminacy will of course be manifold in
an enormous legal system (as will wrongly decided cases), and although
some o f tho se ha rd cases may well be prominent and important, the largemajority of issues add ressed by a legal systems workings are rout ine rath er
than vexing. Numerous though the truly knotty cases may be in an immense
legal system, they are vastly outnumbered by cases that can be handled
readily. Outright indeterminacy is peripheral rather than central.
Therefore, when con fronted with Shapiros claim ab out th e indetermi-
nacy of the identity of an y legal system, we should wonder why the ind eter-
minacy of the law in a small proportion of circumstances would un do the
determinacy of the overarching legal system within which those circum-
stan ces arise. Plain ly, much depends here on the purpose of the an alysis orenquiry in connection with which the identity of some legal system has
become a matter of interest. If our investigations require us to specify a legal
systems doctrinal details with considerable precision when we are elaborat-
ing the nature of the overall system, we might have groun ds for affirming
tha t the absence o f uniq uely correct answers in some h ard cases is sufficient
to warrant our pronoun cing the identity of the system to be ind etermina te.
Much more often, however, an enq uiry into the bearings of an overall legal
regime will no t necessitate such detailed precision . Margina l ind eterminacy
can be ignored o r a t least can be perceived a s falling well shor t o f un der-mining the determinacy of the regime as a whole. For most theoretical and
practical purposes, we are justified in concluding that the identity of a legal
system is not imperiled by the open-end edness of some of its fron tiers.
Let us ponder a n ana logy. Suppose tha t somebody needs to carr y out a
highly accurate census of the American population. If the guidelines that
inform the census have not unequivocally answered certain q uestions about
membership in the American population, and if there are n o supplemen-
tar y bases for uneq uivoca l answers to th ose question s, then some degree o f
indeterminacy will be involved in the conducting of th e census. For exam -ple, suppose that neither the guidelines nor any other relevant decisional
bases have ordained an answer to the question whether American citizens
living perman ently abroad (in Cambridge, England, for example) are to be
classified as members of the American population. Even though the people
occupying th is indeterminate status typically for m on ly a ver y small propor-
122 MATTHEWH.KRAMER
7/28/2019 m Kramer
9/29
tion of the population, somebody assigned to produce an extremely accu-
rate census might have to conclude that the existence of such people
renders the American population indeterminate. For most purposes, how-
ever, the existence of a few million American citizens residing permanen tlyabroad whose status is no t settled by the census guidelines, ex hypothesiis
hardly sufficient to warran t an ascription of indeterminacy to the American
population as a whole. For most purposes, an ascription of determinacy to
the overall population can withstand some peripheral unsettledness.
Shapiro might reply that, in the long passage quoted above, he does not
have in mind difficult cases; instead , he is talking ab out indeterminacy of a
much more wide-ran ging kind. Such a rejoind er would deflect the objec-
tion presented in my last few paragraphs, but it would trigger again my
previous objection. That is, it would seem to posit indeterminacy where infact th ere is none. As has been indicated , when th e moral tests imposed by
an Inclusivist Rule of Recognition are either determinately satisfied or
determinately unsatisfied by some apparent legal nor m N, o fficials in the
Inclusivist regime will not enjoy any strong discretion in regard to the
validation and application of N. They are under an obligation to recognize
Nas a legal no rm when the mora l tests (a long with tests of provenance) a re
determinately met, and they are under an obligation to decline to recogn ize
Nas a legal norm when those tests are determinately unmet. Un less Shapiro
is talking a bout d ifficult cases in which neither of these uneq uivoca l obliga-tions is present, his references to indeterminacy are q uite puzzling .
II. INCORPORATIONISMAND ILLUMINATION
Although the paramount purpose of my earlier article was to vindicate
Inclusive Legal Positivism in response to Shapiros formidable critique, a
secondary aim was to commend a limited version of In corporation ism while
disapproving of an extreme version. The basic idea underlying extreme
Incorporation ism was first articulated pro minently by Philip Soper but has
undergone considerable development and refinement at the hands of
Coleman.7 While my earlier article queried one main aspect of Colemans
theor y, I h arbored no d oubts about the significance an d sophistication o f
his contribution s. Moreover, on most issues surro und ing Inco rpora tionism,
his work and mine are in full agreement. Indeed , as will be obser ved, bo th
7. Amon g the relevant pub lications by Co leman are the following: Incorporationism, Conven-
tional ity, and the Practical Di fference Thesis, 4 LEGAL THEORY 381 (1998) (hereinafter cited asColeman, Incorporationism); Second Thoughts and Other First Impressions, i nANALYZING LAW257, 25878 (Brian Bix ed., Oxford: Oxford University Press, 1998); Authori ty and Reason, inTHE AUTONOMY OF LAW 287 (Robert George ed ., Oxford : Clarend on P ress, 1996) (hereinaftercited as Coleman, Authority); Negative and Posit ive Posit ivism, inMARKETS, MORALS, AND TH ELAW 3, 1227 (Cambridge: Cambridge University Press, 1988). For Sopers pioneering contri-bution, seehis Legal Theory and the Obli gation of a Judge: TheHart / Dworkin Dispute, 75 M IC H . L.REV. 473, 50915 (1977).
Throwing Light on the Role of Moral Principles in the Law 123
7/28/2019 m Kramer
10/29
Shapiro a nd Coleman h ave overestimated th e extent to which I have sough t
to d istance myself from Co lemans robust Incorpora tionist stance.
A. Unchallenged Truth
The modest version of Incorporationism that I espouse is marked by a
curtailment of the range of circumstances to which the general Incorpo-
rationist thesis pertains. Like anyone who upholds some variant of that
thesis, I accept that a norms correctness as a moral principle can be a
sufficient condition for its status as a legal norm in this or that system
of law. H oweverunlike Colemans thoroughgoing Incorporationist doc-
trinemy moderate Incorporationist theory submits that, even in legalsystems where moral correctness does amount to a sufficient condition
for legal validity, it cannot amount to such a condition in most cases.
Only in hard cases do any Incorporationist criteria in a Rule of Recog-
nition become activated. On ly in such cases are mo ral precepts incorpo-
rated into a legal regime as some o f its nor ms. If a scheme of governan ce
were instead to treat mo ral cor rectness as a sufficient cond ition for legal
validity in all circumstances (rather than only in the circumstances that
breed hard cases), it would lack the regularity and coordination that are
essential features of anything rightly classifiable as a regime of law. In asociety where every legal judgment by judicial and executive officials is
a moral judgment through and through, the consentaneity among the
multifarious determinations of the officials will be too low to enable the
emergence or continuation of a genuine system of legal norms. If the
operations of a regime are characterized by a welter of discordant deci-
sions and clashing rationales not only at its margins but throughout its
core, then it lacks the regularity that is ind ispensable for its very existence
as a legal regime.
As was repeated ly ackno wledged in my earlier article, my rejection of anytho rough going In corporation ist position is based not purely on conceptual
analysis but also on an empirical assumption . Though my empirical premise
is general and uncontroversial, it is indeed empirical rather than strictly
philosoph ical. What I assume is tha t, except in static an d highly ho mogene-
ous societies not much larger than a handful of families, there will be
widespread disagreement among people over the answers to most moral
questions; in any large and complex society, extensive d ivergences among
peoples moral outlooks are bound to be present. Now, this empirical
assumption is scarcely novel or co ntroversial. It is accepted by both Shapiroand Coleman , and has often figured saliently in legal and political philoso-
phy. Hart, for example, relied on such a premise when discussing the
disadvan tages of a regime o f primar y rules without seconda ry rules (Con-
cept, 89-90). Some strand s of Rona ld D workins critiques of legal positivism
in TakingRightsSeriouslydepend on ba sically the same empirical assumption
124 MATTHEWH.KRAMER
7/28/2019 m Kramer
11/29
about the ineliminability of mo ral con troversy.8 H ence, although my disap-
proval of any robust Incorpora tionist stance is partly empirical ( at a very
abstract level) an d is no t purely a matter o f con ceptual analysis, the genera l
drift o f the belief that un derpins my reservations is certainly no t unfamiliarin the domain of legal philosophy. In my earlier article, I therefore pre-
sumed tha t the o rientation and limits of my misgivings would be plain.
Unfortunately, both Shapiro and Coleman have taken me to be insisting
on the falsity (or to be denying the truth) of the robust Incorporationist
thesis. Each o f them retort s to my earlier article in part by claiming that my
cha llenge to th e aforemen tioned th esis has tend ed to reveal that that thesis
is trueas if I had end eavored to suggest other wise. Let us first look briefly
at this claim by Shapiro and Coleman , and then a t a bolder affirmation o f
the truth of th e robust Incorporation ist thesis. As will be seen, I can and d o
8. See especial lyRonald Dworkin, TAKING RIGHTS SERIOUSLY 65, 6667 (Cambridge, MA:H ar vard U niversity Press, 1978). See also id. at 3944, 5455, 5968. I am here looking askanceat Coleman s repeated conten tion th at Dworkins critique of a proto-Incorpora tionist versionof legal po sitivism is purely conceptual (Coleman, Constraints, 17980, 181). Q uite plainlyempirical as well as conceptual is the D workinian claim th at any reference to m orality amon gthe criteria o f legality would enta il controversy sufficient to preclude the rule of recognitionfrom gener ating the con vergen ce necessar y to establish its existence as a regula tive social rule(Coleman, Con straints, 179). Dworkin is trading on the same basic empirical assumption o nwhich I rely; no purely conceptual analysis can tell us whether the degree o f con vergence
stemming from the use of moral criteria will or will not be insufficient to sustain the existenceof a Rule of Recogn ition that encompasses such criteria. Bo th Dworkins argum ent and my ownrejection of thorou ghgoing Inco rpora tionism are partly empiricalthough I am more explicit
than Dworkin in drawing attention to that characteristic.I should no te a further a nd partly related point o n which I disagree with Pro fessor Co leman
about certain aspects of Dworkins early work. Along the following lines, Coleman takesexception to Dworkins claim that the availability of moral principles as touchstones foradjudication will eliminate or virtually eliminate judicial discretion: Of course, even as discre-tion owing to the availability of legally binding resources decreases (as the set of availableresources increases), the possibility of d iscretion owing to vagueness increases (as th e set o fcont roversial mora l predicates legally bind ing on officials increases). The same pred icates tha treduce discretion on one front increase it on another: a consequence of Dworkins own viewtha t mora l pred icates are essentially controversial. To defeat the positivist view of discretion,Dworkin has to do more than show that the set of available legal resources on which a judge
can d raw goes beyon d rules (Coleman , Constraints, 172 n.5; Coleman, In corporationism,406 n.42). Coleman s retort ten ds to b lur the d istinction b etween weak and strong discretion,i.e., between discretion in the application of a dispositive standard and discretion in theabsence o f a ny d ispositive stand ard . D workin ascribes to legal positivism the view that judgesenjoy strong discretion in har d cases. H e seeks to rebut th at view by pointing out tha t judgesare obligated and empowered to resort to moral principlesas determina tive stan dard s for theirdecisions in such cases. He can readily acknowledge that the application of those principleswill a lways involve weakdiscretion because the cor rectness or in correctn ess of any applicationis not strictly demonstrable. He can therefore readily acknowledge that a greater reliance on
mora l principles will entail an increase in th e exercise of weak discretion. H owever, such anacknowledgment does not at all impair or belie his thesis that the availability of moralprinciples as adjudicative touchstones will significantly reduce the need for judges to exercise
strongdiscretion. In shor t, Dworkin could en dorse the first two sentences in the latest q uota tion
from Coleman, so long as we recognize that the augmented discretion and the diminisheddiscretion are of different types. Once we do recognize that point, however, we can see that thethird sentence in the quotation is unsustainableat least if the phrase positivist view of
discretion refers to the positivist view of discretion as characterized by Dworkin.
Throwing Light on the Role of Moral Principles in the Law 125
7/28/2019 m Kramer
12/29
endorse that bolder affirmation without retracting any of my previous
animadversions on the thoroughgoing Incorporationist thesis itself. My
reason s for eschewing that th esis will become more eviden t here than in my
earlier article.Shapiro and Coleman both contend th at the thoroughgoing Incorpora-
tionist thesis is a cond itional proposition . As Shapiro writes: [T]he th esis
of robust incorporationism is conditional in nature. The robust incorpora-
tionist claims tha t i fa judicial convention arises tha t treats mora l appropri-
aten ess as the sole condition of legality, thensuch a convention al rule coun ts
as a rule of recognition (Shapiro, Law, 156, empha ses in original).
Coleman similarly submits tha t h is version of Incorporation ism is properly
und erstoo d as a conditional claim: If there are mo ral criteria of legality tha t
are a ccepted by officials from an internal point of view and practiced by asufficient n umber o f them, and if the bulk of the population complies with
the rules valid under those criteria, then there is a legal system in which
mora lity is a con dition of legality (Co leman , Constraints, 182). H aving
formulated the robust Incorporationist position as a conditional proposi-
tion, Shapiro and Coleman can then easily demonstrate that my severe
doubts about the sustainability of a robustly Incorporationist Rule of Rec-
ognition tend to underscore the truth of the position that I am assailing.
After all, the antecedent in the conditional proposition is almost always
falsewhich means that the overall proposition is almost always true re-gard less of the truth -value o f its consequen t. Moreover, because th ere a re
no groun ds for thinking th at the consequent will be false when the an tece-
den t is true, we can safely conclude that the cond itional claim of thoro ugh-
going In corporation ism is invariably true.
Now, there is absolutely no reason for me to o bject to the n otion tha t the
centra l thesis of rob ust Incorpora tionism presented a s a conditional propo-
sition is invariably true. On the contrary, I shall presently argue that a
slightly bolder rendering of that thesis is likewise true. Before turning to
tha t alternative rendering, however, we should glan ce at two aspects of th eShapiro/Coleman riposte that may elicit some dissatisfaction in the read er.
First, the riposte has delineated the thoroughgo ing version of Incorpora-
tionism in a mann er that seems somewhat ad h oc. H eretofore, the central
thesis of Incorporationism has sometimes been expressed in other than a
conditional form. It has quite frequently been expressed as an indicative
proposition, which is precisely the form in which Shapiro articulates it at
the outset of his latest article: There is some possible legal system where
the legality of a norm does depend on some of its moral properties
(Sha piro, Law, 129).9
Second, when the robust Incorporationist position
9. For further indicative renderings of th e central thesis of Incorpo rationism, see, e.g.,Coleman, Autho rity, 30506; Co leman, Incorpora tionism, 382. For some cond itiona l for-mulations, seeBrian Leiter, Realism, Hard Positivi sm, and Conceptual Analysis, 4 LEGAL THEORY533, 535 (1998); Sco tt Shap iro, TheDif ference that Rules Make, inANALYZING LAW 33, 57 (BrianBix ed., Oxford: Clarend on Press, 1998). Slightly more oblique in its cond itiona lity is Coleman ,
126 MATTHEWH.KRAMER
7/28/2019 m Kramer
13/29
is formulated as a conditional proposition, it becomes gratuitously unad ven-
turous. The truth of such a proposition is consistent with the necessary
falsehoo d o f both its ant ecedent and its conseq uent.
Let us, then, concentrate on a version of thorough going Incorporation-ism th at is presented as an ind icative claimthe claim tha t there is at least
on e possible legal system wherein the cor rectness of no rms as mora l prin-
ciples is in all circumstances the suff icient con dition fo r their status as legal
norms. If a n indicative proposition along these lines is true, then neither
the antecedent nor the consequent of the conditional variant of robust
Incorporationism is necessar ily false. Tha t is, the ind icative var iant of robust
Incorporationism is bolder an d thus more interesting than its cond itiona l
counterpart. Furthermore, it is true. Contrary to what Shapiro and Cole-
man have supposed, I have always believed th at th e robust Incorporation istthesis in its indicative for m (as well as in its conditional fo rm, of course) is
true.
Why, therefore, did my earlier article reject the doctrine o f thoroughgo-
ing In corporation ism, and why do I con tinue to reject it? The answer to th is
question lies no t in a mistaken perception o f tha t do ctrine as false but in a
perception of it as high ly misleading and un illuminating. For a clarification
of this point, we should recall another argument by Hart in which he
resorted to some elementar y empirical claims. In his famous discussion of
the minimum conten t of natural law (Concept, 189-95), he pointed to somebasic characteristics of h uman beings and the world wherein they live, and
he concluded therefrom that the legal system in any viable society larger
than a ha nd ful of families must include fund amental interdictory manda tes
tha t coincide in con tent with cardinal precepts of morality.10 Now, suppose
tha t an oppon ent were to point out tha t H arts empirical premises do not
strictly necessitate his conclusion. Such an opponent would accurately
maintain that, even if we grant Harts premises (as any sensible person
should), we ought not to regard as logically incoherent the idea that a
large-scale society can endure without basic legal prohibitions on murder,unprovoked assault, extensive vandalism, and other serious misdeeds.
Though the sustainability of a society without those prohibitions is stagger-
ingly unlikely, it is not logically impossible. This imaginar y opponen t would
thus coun ter H art s position with the following thesis: There is at least one
possible large-scale society tha t can end ure in the absence of any fund amen-
tal legal proscriptions. Ought we to think that such a rejoinder would
expose some significant shortcomings in H art s discussion of th e minimum
content of natura l law? Hart would doubtless have been unimpressed. H e
Autho rity, 28788, a nd considerably more o blique is Jules Coleman and Brian Leiter, LegalPositi vism, inA COMPANION TO PHILOSOPHY OF LAW AND LEGAL THEORY 241, 251 (D ennis Pa tter-son ed., Oxford: Blackwell, 1996).
10. I have argued elsewhere that H arts discussion of th e minimum conten t of natura l lawis fully consistent with legal positivisms insistence on the separability of law and morality. Seemy IN DEFENSE OF LEGAL POSITIVISM ch. 9 (O xford: Oxford U niversity Press, 1999).
Throwing Light on the Role of Moral Principles in the Law 127
7/28/2019 m Kramer
14/29
would have recommended that we dismiss the opponents positionnot
because it is false, but because (though true) it is trivial and profoundly
misleading. It aridly describes a bare logical possibility tha t n ever h as been
actualized and never will be actualized. Tho ugh the opponents thesis iscorrect in submitting that there is not any logical or conceptual incoher-
ence in the n otion of a dura ble large-scale society where no legal restraints
whatsoever are placed on serious misdeeds, the likelihood of the emer-
gence of such a society is so formidably remote tha t it should be disregard ed
by anyone con structing a theor y of law or government. While the oppo-
nen ts thesis is no t strictly wron g, it is wholly misguided in wha t it empha -
sizes. It highlights a fan ciful state o f af fairs that n ever ha s arisen a nd never
will arise, and it consequently tends to obscure the fact that that state of
af fairs will no t obta in un der any minimally plausible set o f circumstances.Now, admittedly, the robust version of Incorpora tionism is no t as outland-
ishly misleading as the imagina r y opponen ts riposte to H art. Although th e
sustainability of an unalloyedly Incorporationist regime of law in an y society
larger th an a han dful of families is prodigiously unlikelynot least because
the task of law-ascertainmen t is carried out by a governments myriad
executive o fficials as well as by judicial of ficials11it is no t as close to being
unimaginab le as is the sustainability of a large-scale society without any legal
curbs on serious wron gs. H ence, a do ctrine that asserts the possibility of an
unalloyedly Incorporation ist legal regime within a large-scale society is no tquite as misdirected in its emphasis as is the thesis described in my last
paragraph . All the same, th e robust version o f Incorporation ism h ighlights
the prospect of a state of affairs that never has obtained and never will
obt ain in a ny large-scale society. It conseq uently tends to obscure the im-
portan t fact that such a state o f a ffairs will no t emerge (in a large-scale
society) under any credible set of circumstances.
Thus, while the extreme Incorporationist thesis is trueas I have always
readily accepted12it is unhelpfully distracting and unilluminating and
11. At one juncture, Shapiro points out that a robustly Incorpora tionist Rule of Recogn itioncould o btain in a legal system with on ly two judges, b oth of whom are twin sisters, atten d th esame chur ch[,] a nd are like-minded on eth ical issues (Sha piro, Law, 156). As isa ttested by myearlier articles several references to th e possibility of a rob ustly Incorpora tionist Rule of Recog-nition in the legal system of an extremelysmall and h omo geneous society, I have no wish whatso-ever to dispute the truth of Shapiros observation. All the same, his observation is irrelevant tothe sorts of societies on which I am concen trating. With r eference to m edium-scale an d large-scale societies, where multitudes of executive/adm inistrative o fficials as well as judicial officialsare obliga ted and empowered to engage in the activity of law-ascertainmen t, a scenario of a two-judge legal regime (with no executive/administrative officials, apparent ly) is perilously close to
the Austinian depiction of the sovereign as a single person. Ha rt aptlylabeled the latter depictionas a disastrouslycon fusing piece of mythology (Ha rt, Concept, 110).
12. Ind eed, as can be gath ered from Matthew H . Kramer, LEGAL THEORY, P OLITICAL THEORY,AND DECONSTRUCTION 12830 (Bloomington: Indiana University Press, 1991), I am more in-clined th an H art to classify a robu stly Incorpora tionist primitive regime of law as a regime of law.
In my earlier article on In clusivism and In corpor ationism, the on ly sentence tha t might seemto impugn the truth of the robust Incorpo rationist thesis is in the final paragraph, where Icontend that robust Incorporationism as a jurisprudential position is unsound (Kramer,
128 MATTHEWH.KRAMER
7/28/2019 m Kramer
15/29
needlessly problematic. Its virtues are fully shared and its shortcomings
avoided by a modera te version of Incorpora tionism. To be sure, a ph iloso-
pher as nimble as Coleman can man age to d evelop a number of insights by
defending the robust Incorporationist thesis against various conceptualcha llenges. Neverth eless, those same cha llenges must be parried by moder-
ate Incorporationism, and indeed the latter doctrine can handle them
more smooth ly (as will be further argued presently).
Moreover, a mod est Incorporation ist thesis is peculiarly suitable fo r the
accomplishment of the purpose which Incorporationism was devised to
fulfill: viz. , the purpose of fending off some of Dworkins early attacks
against legal positivism, which were focused on the role of moral principles
as adjudicative touchstones in hard cases. Coleman expresses bemusement
over my distinction between robust and moderate Incorporationism, alleg-ing tha t it generate[s] more smoke than light. H e contend s that distin-
guishing among kinds or categories of cases in which [Incorporationist]
criteria migh t apply . . . is . . . no t a par ticularly helpful dimension on which
various rules of recogn ition can be ar rayed, and he asserts tha t my position
would still fail to answer the objection that motivates [Incorporationism]
in th e first place (Coleman , Con stra ints, 178 n.17, 183). Yet, because the
motivating objection to which Coleman adverts is Dworkins observation
tha t mora l principles are invoked by Anglo-American judges to resolve
some ha rd casesa fact of Anglo-American ad judication tha t D workinbelieves (or believed) to be incompatible with the positivist model of
lawmy singling out of modest Incorporationism in contradistinction to
thoroughgoing Incorporationism is entirely apposite.
To grasp fully the aptn ess of the modera te Incorporation ist thesis in th is
context, we should reflect briefly on wha t it claims. Tha t thesis does not
den y anything af firmed by the robust Incorpora tionist doctrine, but simply
asserts less. It mainta ins tha t th ere is at least one possible large-scale lega l
system in which the moral correctness of norms is sufficient for their status
as legal no rms in some or all hard cases. Because this mod erate thesis doesno t in an y way deny the bolder claim of thoroughgo ing Inco rporationism,
it does no t make the mistake of suggesting tha t that bo lder claim is false. At
the same time, it draws atten tion to the precise point of conten tion between
Dworkin and legal positivists; it asserts no more and no less than what is
necessary to rebut Dworkins criticism. In other words, it asserts what is
necessar y to respond effectively to th e crux that gave rise to In corporation -
ist theories. A thesis focused on hard cases is singularly appropriate for
overcoming an objection that is focused on hard cases. A more sweeping
thesis is superfluous. (Furthermore, with regard to m y highlighting o f the
suprano te 1, at 108). In that sentence, a s should ha ve been clear fro m th e rest of my article, Iwas not using th e term un soun d in its technical philosophical sense. I was not asserting th atthe robust Incorporationist position is derived from false premises or is invalidly derived, andtha t it therefore m ay be false. Instead, I was using u nsound mo re loosely to d enote un illumi-nat ingness and misleadingn ess.
Throwing Light on the Role of Moral Principles in the Law 129
7/28/2019 m Kramer
16/29
distinction between easy cases and hard casesthough not with regard to
certain oth er featuresmy espousal of modest Incorporation ism manifestly
runs parallel to Harts discussion of linguistic and normative determinacy
in the seventh chapter of The Concept of Law. H art too k the view that theexistence of a legal system is consistent with pervasive disagreement in the
systems penumbra but is inconsistent with such disagreement in the sys-
tems core. I correspondingly take the view tha t th e existence of an y large-
scale regime of law is con sistent with of ficials reliance on mora l cor rectness
as the sufficient cond ition fo r legal validity through out the regimes penum-
bra , but tha t it is incon sistent with any compa rable reliance throughout the
regimes core. My view follows from a combination of th e H artian view and
the general empirical assumption that underlies my championing of mod -
est Incorporationism.)In short, because the Dworkinian attack that impelled the development
of In corporation ism was concerned with the role of moral precepts in hard
cases, it can most pertinently be repelled by a thesis oriented toward that
role in exactly such cases. O f course, to ma inta in as much is not per se to
establish that modest Incorporationism can surmount all the conceptual
hurdles that confront it. Shapiros trenchant challenges to Incorporation-
ism must be met by a modest version of the doctrine as well as by a
thoroughgoing version. The rest of this essay will seek to reinforce my
earlier articles efforts to show that a moderate Incorporationist positioncan indeed meet those challengesnot by refuting them (for they are
irrefutable), but by defusing them.
Before we move on to re-encounter Shapiros critiques, a quick caveat
should be entered. Nothing in this subsection is meant in any way to
depreciate the importan ce and value of purely conceptual argumenta tion.
When such argumenta tion is illuminating and fruitfulas it very often is,
no t least in Colemans and Shapiros workit is clearly preferable to argu-
mentation that relies in part on empirical assumptions. Indeed, my earlier
articles defense of Inclusive Legal Positivism was strictly conceptual. Mycomplaint herein is on ly about conceptual theses tha t do n ot shed light on
any past, present, or future states of af fairs (in societies tha t are not tiny and
high ly homogeneous). Such th eses are mo re misleading than instructive.
B. What Goes On in Hard Cases
In my earlier article I unreservedly acknowledged the correctness of
Shapiros astute observation that the moral principles validated under anIncorporationist Rule of Recognition do not constitute a ny reasons-for-
action beyon d those con stituted by the In corporation ist criteria themselves.
His observation plainly applies to a moderately Incorporationist Rule of
Recogn ition as much as to a ro bust version . My tack in my earlier article was
no t to gainsay the sound ness of Sha piros argument, but to remove its sting
130 MATTHEWH.KRAMER
7/28/2019 m Kramer
17/29
when it is invoked against modest Incorporationism. Specifically I con-
tended that, a lthough incorporated moral precepts cannot furn ish judges
or citizens with any independen t reasons-for-action (i.e., reasons beyond
those furnished by an Incorporationist Rule of Recognition itself), thatinability is untro ubling in ha rd cases because those precepts are called upon
to play an altogether different role therein. Because the hard cases lie
outside th e d eterminate ambits of the source-based legal n orms, and be-
cause a mildly Incorporationist Rule of Recognition predetermines the
correct ways of resolving any disputes that do lie outside those ambits, hard
cases will no t be decided by reference to legal n orms that serve a s partly
independent reason s-for-action . Mora l precepts invoked to hand le tho se
cases are obviously no t such reasons, then, b ut are mean s of specifying the
requirements ord ained by the Incorporation ist stran d in the Rule of Recog-nition. Though the precepts do not add to what the Rule of Recognition
deman ds, they enable its dema nd s to be spelled out.
In my earlier article, I described the function of moral precepts in hard
cases as that of terminating disputes and achieving closure and re-
solv[ing] the points at issue.13 The last-men tioned ph rase in particular can
be faulted fo r a lack of clar ity, because it has led Sha piro to miscon strue my
argument. H e writes: When a rule is used to resolve a dispute, the rule is
treated a s a reason to decide the dispute in a ccordance with the rule. To
resolve disputes, therefore, is to guide evaluative behavior. If my argumentis correct, then moral rules that lack pedigrees cannot resolve disputes
because these rules cannot motivate a judge to resolve a dispute in a
manner differently than he might have, had he not appealed to the rule
(Sha piro, Law, 159). Tho ugh Shapiro s misread ing of my remarks is quite
und erstan dable, it is indeed a misread ing that impels his retort. My accoun t
of the invocation of moral principles in hard cases does not attribute to
them the role of independent bases for reach ing certain decisions in tho se
cases and for eschewing contrar y decisions. Rather, the principles as formu-
lated by judges are simply the vehicles through which th e dispute-termina t-ing effects of the Incorporationist strand in the Rule of Recognition
become discernible and gain expression.
When moral precepts are add uced in h ard cases within a modestly Incor-
poration ist legal system, th ey operate in two chief ways. First, they ena ble or
facilitate the end eavors of judges to figure out what the Rule of Recognition
requires. As Shapiro states, a judge might appeal to moral principles
themselves to a id him in applying the rule of recogn ition (Shapiro, Law,
159 n.65). Of course, th is heuristic function do es not consist in providing
some new reason s for any particular decisions; all decision-determiningreasons in ha rd cases are provided by the mildly Incorporation ist criterion
in the Rule of Recognition. The moral principles heuristic function con-
sists instead in clarifying an d distilling the d eman ds of the aforemention ed
13. Kramer, supranote 1, at 10102.
Throwing Light on the Role of Moral Principles in the Law 131
7/28/2019 m Kramer
18/29
criterion. While cr ystallizing or concretizing those d emands, the moral
principles do no t add thereto (o r subtract therefrom). They encapsulate at
an intermediate level the implications of th e Incorpora tionist instruction in
the Rule of Recognition, and they thereby supply judges with a cognitivehan dle on that abstract instruction.
In add ition, the moral precepts cited by the modera tely Incorporation ist
judges are the means by which they express to citizens and their fellow
of ficials the req uirements of their Rule of Recognition. Although th at Rule
of Recognition has already settled the proper resolution of each hard case
to which th ere is a un iquely correct mora l answer, the invoca tion of a mo ral
principle enables the judges in such a case to indica te persuasively what the
already-settled resolution is. By add ucing a moral principle to specify quite
concretely the deman ds of morality in the con text of a knotty dispute thatis not resolvable through recourse to source-based laws, th e judges signal
the moral sound ness of their decision and its conseq uent con formity with
the prevailing standa rds of law-ascertain ment. They reaf firm to citizens the
moral appropriateness and requisiteness of th eir verdict. They reaffirm to
their fellow of ficials their compliance with the modera tely Incorporation ist
Rule of Recognition which makes moral requisiteness the touchstone for
dea ling with h ard cases.
Thus, when moral precepts are a pplied to the fa cts of h ard cases within
a mildly Incorporationist regime, they perform both a heuristic functionand a confirmatory function. Moreover, they are fully genuine norms in
tha t th ey require certain outcomes and disallow con trar y outcomes. Wha t
they cannot d o, of courseif they have been chosen correctlyis to req uire
anything other than wha t is required by the mildly Incorpora tionist Rule of
Recogn ition itself. For Shapiro , tha t inability is suff icient to disqualify those
sourceless mora l precepts from being rightly classified as legal norms. He
adheres to a ten et designa ted a s the Pra ctical Difference Thesis, which h e
states as follows: Legal rules must in principle be capable of securing
compliance by making a difference to an agents practical reasoning(Shapiro, Law, 129). Because properly incorporated moral principles
never lead to conclusions other th an those dictated by the Rule of Recog-
nition under which the principles have been incorporated, they do not
satisfy the Practical Difference Thesis. They are therefore on ly ersatz legal
norms, in Shapiros view.
Let us note a key ambiguity in the Practical Difference Thesis. Is its
implicit quantifier universal or plurative? That is, does the thesis assert
something about alllegal nor ms or only about mostlegal norms? Shapiro
obviously intend s the thesis to be taken a s a conceptua l claim con taining auniversal quan tifier, whereas I subscribe to it a s a partly conceptual and
partly empirical claim containing a plurative quantifier. Let us probe his
argument in favor of the th esis as he con strues it.
Shapiro conn ects the Practical Difference Thesis to a broad er conception
of law, which he labels as functiona list (Shapiro, Law, 167-69). H e
132 MATTHEWH.KRAMER
7/28/2019 m Kramer
19/29
correctly remarks tha t Ha rtian legal positivism attributes to law the function
of guiding cond uct by means of the nor ms which a legal system sets forth
as authoritative standards of behavior. Shapiro presumes that laws fulfill-
ment of this function entails the capacity of every legal norm to guideconduct: [A]ll legal rules have, as their function , the guidance of cond uct.
And because a no rms guidance of conduct involves its making a difference
to peoples practical reason ingby requiring pa tterns of behavior tha t are
not necessarily already required by the Rule of Recognition in the legal
system where the norm is locatedShapiro can conclude that no r ule tha t
is in principle incapab le of making a practical difference can be a legal rule
(Shapiro, Law, 168).
The suspect link in this chain of reason ing is the premise th at the
accomplishment of laws function entails the capacity of everylegal n ormto guide conduct. We n eed to be to ld why a legal systems performan ce
of that function excludes the possibility that mostof the systems norms,
as opposed to allof its norms, are endowed with a guiding capacity. Re-
sponding to some criticisms largely along these lines from Wilfrid
Waluchow and Kenneth Himma, Shapiro offers the following argument
in support of his premise: [I]n the case of legal rules and legal institu-
tions[,] such an inference [an inference that a part has the function F
just because the whole has the function F] is sound. For legal rules are
themeansby whichlegal systems guide conduct. We can say tha t the functionof legal rules is to guide conduct because they have been produced by
legal institutions in order to guide conduct (Shapiro, Law, 169, empha-
sis in origina l). The argument in th is quota tion d oes no t clinch the ma tter,
for the quotations last two sentences are just as ambiguous as the initial
formulation of the Practical Difference Thesis. Each of those sentences
can stand up per fectly well even if th e phra se legal rules in each of th em
is taken to denote mostlegal norms rather than allsuch norms. A legal
system can admirably carry out its function of laying down authoritative
directives to steer human conduct into prescribed channels, even if someof its norms do not contribute to that function but instead play only
heuristic and confirmato ry roles. Shapiro canno t settle the choice between
a un iversal quan tifier and a plurative quan tifier for h is Pra ctical Difference
Thesis by adverting to the role which H art attributed to law. An ascription
of that role to law is consistent with the Practical Difference Thesis under
either interpretation.
Indeed, the overall guiding function of law can hardly avail Exclusive
Legal Positivists such as Shapiro, because the difficult cases in which moral
principles get invoked as decisive stand ard s are ca ses in which the source-based legal norms do not provide determinate guidan ce. Gaps or con flicts
or ambiguities in the ar ray of source-based n orms give rise to lacunae in
the guiding sway of law th at are filled (in some regimes) by officials
resorting to moral principles. For the Exclusive Legal Positivist, the over-
coming o f those lacunae does no t in an y way mean tha t the invoked moral
Throwing Light on the Role of Moral Principles in the Law 133
7/28/2019 m Kramer
20/29
principles are legal nor ms. As substitutes for legal norms, the principles fill
in where the law has run out. H ence, an Exclusive Legal P ositivist such as
Shapiro, who accepts Harts characterization of laws function, must also
accept that any particular legal system can and does perform that functionnotwithstanding the occurrence of lacunae in the systems matrix of
source-based norms. Most disputes that arise in the jurisdiction of each
system are covered by such a matrix, but some of the disputes fall within
the lacunae. In other words, when Exclusive Legal P ositivists and mod erate
Incorporationists endorse H arts function alist conception of law, th ey are
at one in thinking that a legal regimes guiding role is performed across
most circumstances but not across all circumstances. Neither the Exclusive
Legal Positivist nor the mod est In corporation ist believes tha t the guidance
offered by a legal system through its source-based nor ms is seamless o rcompreh ensive. Each believes that at a ny given juncture th e source-based
norms omit from their coverage a number of possible sets of circum-
stan ces. All the same, each believes tha t the guiding fun ction of law can be
fulfilled an d is fulfilled. H ence, the Exclusivist an d the mod erate In corpo-
rationist differ only over the question how the moral precepts invoked in
ha rd cases sho uld be classified. Exclusivists will perceive th ose precepts as
nonlegal nor ms to which judges appeal when the guidance of legal norms
(i.e., source-based no rms) has run out. Modest Incorpora tionists will per-
ceive tho se precepts as non guiding legal no rms tha t are add uced by judgesfor heuristic and confirmatory purposes when the guidance of source-
based legal norms has run out. That is, the sole issue in controversy is
whether the moral principles are nonlegal norms or nonguiding legal
no rms. Such an issue cannot be resolved by reference to th e role of law as
an institution that establishes source-based stand ards for the purpose of
steering conduct. After all, the Exclusive Legal Positivist and the moderate
Incorporationist can agree not only in attributing such a role to law, but
also on the following points: The source-based standard s do n ot pro vide
determinate guidance in hard cases; the role of law can be achieved and isach ieved despite the ina bility of th e source-based stand ard s to yield d eter-
minate guidance in certain circumstances; the moral principles relied
upon by mildly Incorporationist judges in hard cases perform no more
than heuristic and confirmatory roles. The only area of disagreement,
then, is the question whether the nonguiding principles invoked in hard
cases are legal or non legal. Any dispute over that question cann ot be put
to an end by recourse to the role of law mentioned just above, because
neither side to the dispute contests that role.
If we are to break out of the deadlock just described, we shall have to
marshal arguments that go beyond a focus on laws basic function. My
earlier article sough t to present just such an a rgumen t by main tain ing tha t
the considerations that lead us to acknowledge the status of some custom-
ar y norms as legal mand ates should lead us also to acknowledge th e status
134 MATTHEWH.KRAMER
7/28/2019 m Kramer
21/29
of incorpora ted mo ral principles as legal norms.14 Tha t complicated line o f
reasoning will not be recapitulated h ere, because Shapiro does not ad dress
it at all. Nonetheless, what should be noted in the present con text is that
tha t line of reasoning has been put for ward in an effort to keep the debatebetween Exclusive Legal Positivists and modest Incorpora tionists from
amo unting to no more than an exchan ge of rival stipulation s. My argument
tries to of fer a route for moving beyond the frustrating standof f that would
ensue from such an exchange.
Let us close this subsection by pond ering the nature of Shapiros Practical
Difference Thesis, as perceived by a propon ent o f mod erate Incorporation -
ism. Construed as a purely conceptual claim, that thesis should be re-
jectedregardless of whether we deem its implicit quantifier to be universal
or plurative. Coleman is quite correct to maintain as much. If the schemeof governance in a very small an d stable commun ity achieves considerab le
regularity in its operations on th e ba sis of a robustly Incorpora tionist Rule
of Recognition, then the conditions for the existence of a legal system are
satisfied even though the Practical Difference Thesis is not.
By contra st, when the Practical Difference Thesis is construed as a partly
empirical claimwhich is how a modest Incorporationist will generally
cho ose to construe itit is fully acceptable as long as its implicit qua nt ifier
is plurative rather than universal. So understood, that thesis submits that
most legal norms constitute partly independent reasons for citizens andof ficials to beha ve in certa in specified ways. Left open is the possibility that
some legal norms in this or that system are incapable of constituting such
reasons (tho ugh they are capable of performing h euristic and confirmatory
functions); moral precepts are just such norms when invoked under a
mildly Incorporation ist Rule of Recogn ition in h ard cases where the source-
based legal standards do not yield determinate guidance. Now, we can
knowto some degree as an empirical matterthat most of the norms in
an y large-scale system of law will be pa rtly independen t reasons-for-action
that are ascertainable as such. Where most of the norms applied by alarge-scale scheme of govern an ce do not form such reasons, the scheme will
lack the regularity and coordination th at are ind ispensable for the existence
of a legal regime. Ergo , in a large-scale society where a legal system pre-
vailsi.e., in just about any such society that is tenable at allmost of the
auth oritative norms applied by of ficials for the assessmen t o f con duct will
amo unt to reason s-for-action tha t a re not wholly derivative of the reason s
furnished by the societys Rule of Recognition itself. Such is the message of
a plurative and partly empirical version of the Pra ctical Difference Thesis,
which effectively encapsulates the conceptual an d empirical claims of mod-est Incorporation ism. Tha t thesis reserves a place for no n-source-based
legal norms, but the place is correctly characterized as the domain of
exceptionality rath er than th e domain of typicality.
14. Id. at 103-07.
Throwing Light on the Role of Moral Principles in the Law 135
7/28/2019 m Kramer
22/29
C. Regularity in Irregular Circumstances
Coleman attributes to me the notion that the officials who run a mildly
Incorporationist legal system will arrive at decisions in hard cases in acoordinated fashion as if they were applying straightfor ward source-based
legal nor ms. H e writes: I am unpersuaded by Kramers view tha t mod est
Incorporationism can sustain a legal practice in hard cases, but not more
genera lly. Surely if there are lots of ha rd cases, then on Kramers view the
controversiality of morality ought to underm ine the effo rts of o fficials to
resolve disputes in a coordinated fashion. Controversy is controversy re-
gardless of the scope of application. Even a narrow scope of applica-
tion hard casescan be a big problem, since the narrowness of scope says
nothing at all about how often controversy can arise (Coleman, Con-straints, 183 n.24). Had my earlier article put forward the view Coleman
ascribes to it, it would ind eed have been vulnerab le to his strictures. In fact,
however, I sought to emphasize therein that the likelihood of intractable
disagreements amo ng the officials about the appropriate outcomes in hard
cases is over whelmingly high . Their shared adheren ce to a mildly Incorpo -
ration ist Rule of Recogn ition does no t prevent th em from diverging ma rk-
edly in their selection and application of moral principles for dispatching
difficult disputes.15 Shapiro acknowledges that I took account of the high
probability of judicial d isunity, but h e maintains tha t my effort to d eal withit was unsuccessful. He queries whether o fficials can abide by the In corpo-
rationist criterion in a moderately Incorporationist Rule of Recognition,
given that the officials decisions in hard cases will be markedly at variance
with one an oth er, an d given th at un iquely cor rect answers will be available
in some hard cases under the Incorporationist criterion. In a hard case
where a uniq uely correct answer is available an d where on ly some officials
ar rive a t tha t answer, there is no relevan t convergence-of-behavior to the
extent that is necessary for the existence of a conventional norm. Because
this pattern o f nonconvergence will obtain in a large majority of hard cases,the Incorporationist criterion in a moderately Incorporationist Rule of
Recognition cann ot exist as such. After all, if that criterion d oes materialize,
it materializes as a conventional norm that underlies and presupposes the
convergent behavior of officials in hard cases. Because that convergent
behavior is missing, however, the Incorporationist criterion must itself be
nonexistent. Shapiro presses this point forcefully:
Rules of recognition are con ventiona l rules, an d conventional r ules must be
practiced. It is not enough that people try to conform to such ruleswhatmatters is tha t they do as they are req uired to do. If most people who en tered
churches neglected to take off their hats, but instead mistakenly removed
their shoes, then th ere would be n o convention in th at community requiring
the removal of ha ts in church . Hen ce, if Kramer lacks confidence that judges
15. Id. at 102.
136 MATTHEWH.KRAMER
7/28/2019 m Kramer
23/29
will no rmally conform to th e rule of recogn ition in hard cases, then it cann ot
be said that a rule of recognition exists that regulates these hard cases.
(Shapiro, Law, 160.)
Though Shapiros argument is powerful, it is not irrebuttable. One con-
sideration tha t part ly defuses his argument is tha t in some hard cases there
will not be any uniquely correct answers. In each hard case for which no
determination is uniquely correct, all the officials who operate a mildly
Incorporationist regime will be acting in accordance with the Incorpora-
tionist criterion in their Rule o f Recognition so long as their favored
decisions and justifications are within a range of acceptable decisions and
justifications. Within that range there will be ample room for clashing
verdicts. Hence, in hard cases that lack uniquely correct solutions, the
intractability of disagreement among officials is fully consistent with an
adherence by every official to an Incorporationist criterion for handling
such cases. Manifold divergences amon g o fficials concerning the extension
of such a criterion are compatible with a general consensus on the crite-
rions intension.
More significant in the present context is a way of coping with the crux
posed when a uniquely correct resolution of a difficult case is available
under a moderately Incorporationist Rule of Recognition. In any such set
of circumstances, only those officials who arrive at the uniquely correct
verdict are acting in strict confo rmity with the Incorpora tionist criterion in
their Rule of Recogn ition . Should we accord ingly conclude tha t there is no
convergence on that criterion by the o ther officials? And should we thus
conclude ( with Shapiro) that there is no sufficiently convergent behavior
to ground the existence of the Incorporationist criterion as a conventional
norm?
In reply to th ese q uestions, we should have recourse to a d istinction tha t
has been highlighted by Coleman in virtually all of his writings on legal
positivism: the distinction between disagreements over content and dis-
agreements over application .16 In a hard case where one outcome is
uniq uely correct und er a moderately Incorporation ist Rule of Recognition,
the o fficials who plump in goo d fa ith for oth er outcomes are engaged in a
controversy over the application(rather than th e basic nature) of the Incor-
porationist criterion in their Rule of Recognition. Even though they are
wrong and their fellow officials are correct about the application of that
criterion , they are ind eed differing with their fellow officials over matters of
application rather than over the matter o f the criterions basic tenor. Their
misguided implementa tion of that criterion does no t amoun t to a rejection
of it an d a n embrace of some other law-ascertaining benchmark. They are
16. Colemans work in this area has been valuable, but it do es not squarely address theprecise problem that I shall be pondering: the question whether, in a hard case where oneoutcome is uniquely correct, officials in a mildly Incorporationist regime who opt for otheroutcomes are conforming to the Incorporationist criterion in their Rule of Recognition.
Throwing Light on the Role of Moral Principles in the Law 137
7/28/2019 m Kramer
24/29
at one with their fellow of ficials in upho lding an Incorpora tionist criterion
in th eir Rule of Recognition, th ough they are less astute than their fellow
of ficials in understand ing what tha t criterion requires in part icular settings.
Two potential rejoinders to the argument in the last paragraph must bedeflected. First, in line with some o f the remarks in Dworkins early assaults
on legal positivism,17 a critic might o bject tha t my argument presupposes a
consensus among mildly Incorporationist officials on some formulation of
the Incorpora tionist criterion in their Rule of Recognition . Yet, the critic
would retort, a construa l of the officials interaction a long th ese lines would
implausibly portra y their differences over the Incorporation ist criterion as
disputes over borderline applications of terms. Second , a critic might simply
point out that some of the officials in the envisaged legal system are
misapplying the Incorporationist criterion, perhaps even in every hard casewhere a uniq uely correct answer is available. H ow, then, can we rightly say
tha t those officials are ad hering to that criterion?
These two o bjections can be parried by way of the same genera l observa-
tion. On the one hand, the objections should hardly be dismissed out of
hand. As the first of them rightly suggests, a depiction of the officials
dishar mony as a set o f quarrels over borderline uses of words is far from
compelling. Although a widely accepted formulation of the Incorporation-
ist criterion in a Rule of Recognition may well crystallize within the h ighly
reflective and deliberative processes of jural decision-making, one cann otoften credibly maintain that the clashing verdicts of officials in hard cases
are products of divergent d efinitions of the words in any such formulation .
Likewise, the secon d of the posited ripostes is warran ted in submitting tha t,
if strict compliance in the form of correct implementation is an essential
earmark of adherence to a conventional norm, then the officials who
misapply an Incorpora tionist criterion in some hard cases are no t adhering
to that criterion. If there were no alternative basis for deeming them to be
in compliance with that criterionthat is, if the only sort of compliance
were strict compliance involving correct a pplication sthen the second ofthe po sited rejoind ers abo ve would be q uite telling in regard to hard cases
with un iquely correct outcomes.
On the other hand, fortunately, both of the aforementioned rejoinders
can be countered by a focus on the patterns of officials arguments and
justifications. Contrary to what the first objection maintains, the existence
of the Incorporationist criterion in a modestly Incorporationist Rule of
Recogn ition d oes not d epend on the emergence of a formulation to which
all officials subscribe. Instead, it depends on a practice of justification in
which the Incorpora tionist criterion is immanent as a shared presupposi-tion ( and also as an effect). When the o fficials in a moderately Incorpora-
tionist regime present th eir decisions in ha rd cases, they seek to explain and
justify those decisions by reference to considerations that are identifiably
17. Rona ld Dworkin, supranote 8, at 56, 6263.
138 MATTHEWH.KRAMER
7/28/2019 m Kramer
25/29
moral. Un derlying this practice of moral justification is the Incorporation ist
criterion in their Rule of Recognition. Though the officials might give
expression to that criterion in a formulation on which they concur, they
might cleave to it simply as a tacit preconception that informs their d eci-sions and their articulated rationales. In either event, we can usually best
und erstan d the disaccord ance in hard cases (within a mildly Inco rpora tion-
ist legal system) as d isagreements over the implications a nd rankings of
moral values rath er than over the meanings of words.
Con trar y to what is con tend ed by the second of the potential objections,
strict compliance with the Incorporationist criterion in the Rule of Recog-
nitionviz., compliance that involves correct applications of the concepts
in th at criterion is no t th e on ly species of compliance. In ha rd cases with
uniq uely correct outcom es, of ficials who ar rive at th e wrong decisions cannonetheless be upholding their Incorporationist criterion for dealing with
such cases, so long as they endeavor plausibly to explain tho se decisions by
reference to pertinent reasons. In this con text, the phrase pertinent rea-
sons deno tes any consideration s that are identifiable as ethically signifi-
cant factors that should be taken into account by someon e who is passing
moral judgment on th e problems which th e officials have to ad dress. Inso-
far as the off icials credibly tie their verdicts to such factors, they are enga g-
ing in the process of mora l assessmen t an d moral justification which their
Incorporationist criterion requires of them. Although they are advocatingerroneous outcomes, they are thereby misapplying their Incorpora tionist
criterion rather than shifting to some other touchstone. Their upholding
of that criterion is evidenced by the fact that they appeal to germane
reasons for their decisions. In other words, they acknowledge the general
tenor of th e Incorpo rationist strand in their Rule of Recognition by com-
mending their determinations on the sorts of grounds which that strand
makes relevant. The fact that t
Top Related