Reasonable Doubt in Islamic Law

Yale Journal of International Law
Volume 40
Article 3
Issue 1 Yale Journal of International Law
2015
"Reasonable Doubt" in Islamic Law

Intisar A. Rabb
Follow this and additional works at: http://digitalcommons.law.yale.edu/yjil

Part of the Law Commons
Recommended Citation
Intisar A. Rabb, "Reasonable Doubt" in Islamic Law, 40 Yale J. Int'l L. (2015).
Available at: http://digitalcommons.law.yale.edu/yjil/vol40/iss1/3
This Article is brought to you for free and open access by Yale Law School Legal Scholarship Repository. It has been accepted for inclusion in Yale
Journal of International Law by an authorized administrator of Yale Law School Legal Scholarship Repository. For more information, please contact
[email protected].
Article
"Reasonable Doubt" in Islamic Law

Intisar A. Rabbl
IN TRO D U CTION ....................................................................................................................................... 42
1. THE RECURRING CASE OF M'IZ AND QUESTIONS OF ISLAMIC LEGITIMACY ........................... 51

A. Early Debates on Approaches to Islamic Criminal Law .............................................. 52
1. The Recurring Case of Md 'iz ........................................................................... 52
2. The Interpretive Debates about M r'iz ................................................................. 54
a. The Case for Enforcing the Punishment: Strict Textualist Readings
of the Case of Md 'iz........................................................................... 55
b. The Case for Avoiding the Punishment: Pragmatic Textualist
Readings of the Case of M iz ............................................................ 57
3. The Significance of M d'iz ............................................................................... 59
B. The Role of Muslim Jurists in Defining Structures of Islamic Law and Governance ..... 61
1. Islamic Law as Religious Morality .................................................................. 62
2. Islamic Law as Institutional Competence ....................................................... 64
3. Islamic Law as Institutional Constraint ........................................................... 67
II. THE CONVERSION OF "REASONABLE DOUBT" FROM JUDICIAL PRACTICE TO LEGAL TEXT ......... 69
A. The Canonization of Doubt Through Early Judicial Practice .................................... 70
1. Case of the Mysterious Pregnancy ................................................................. 71
2. Case of the Falsely Accused Butcher .............................................................. 72
3. Case of the Absentee Husband ...................................................................... 74
B. The Textualization of Doubt Through Juristic Interpretation .................................... 76
III. THE DEPLOYMENT OF "REASONABLE DOUBT" TO LIMIT EXECUTIVE POWER ......................... 78

A. The Systematization of Islamic Law in a Political Culture of Violence .................... 79
B. The Generalization of Doubt in Response to Excessive Punishment ......................... 82
1. Moral Anxieties Revealed: Death Is Different ................................................ 83
2. Moral Anxieties Compared: Death Is Dramatically Different ......................... 86
3. Moral Anxieties Expanded: A General Jurisprudence of Doubt ..................... 90
C ONCLU SIO N........................................................................................................................................... 93
t Professor of Law, Harvard Law School. Thanks for many helpful conversations and
comments to William Alford, Mary Bilder, Tomiko Brown-Nagin, Ash BAli, Michael Cook, 1. Glenn
Cohen, Christine Desan, Charles Donahue, William Eskridge, Mohammad Fadel, Noah Feldman, Lani
Guinier, Wael Hallaq, Daniel Hulsebosch, Marion Katz, Adriaan Lanni, John Manning, Daniel
Markovits, Hossein Modarressi, Roy Mottahedch, William Nelson, Vlad Perju, Kim Scheppele, Jed
Shugerman, Carol Steiker, Laura Weinrib, James Whitman, and Adnan Zulfiqar; to participants of the
Legal History Roundtable at Boston College Law School, the "junior" and regular faculty workshops at
Harvard Law School, and the faculty workshop at NYU Law School; to my co-panelists at the American
Society of Legal History session on Late Antique and Medieval Islamic Criminal Legal History; and to
the tremendous efforts of Allison Day, Leslie Esbrook, Vivaan Nehru, Andrew Walchuk, and the other
editors of the Yale Journal of InternationalLaw. Unless otherwise noted, all translations from Arabic
and Persian are my own.
THE YALE JOURNAL OF INTERNATIONAL LAW [Vol. 40: 41
Legal interpretationtakes place in afield ofpain and death.
- Robert M. Cover'
INTRODUCTION
Doubt pervades most laws, 2 becomes alarming when it arises in criminal

law, 3 and seems especially discomfiting in the context of Islamic criminal law.
In the high-stakes area of American criminal prosecutions, dubious facts or
ambiguous laws can result in unjustified deprivations of life, liberty, or
property.4 Criminal prosecutions in early Islamic contexts bore similar risks,
which were compounded by the idea of a divine lawgiver who had outlined a
set of fixed criminal laws and harsh punishments. 5 In this system, there was no
legislature to update the law, no high court to authorize departures from it, and
no prosecutor charged with proving facts beyond a reasonable doubt. 6 These
features made-and continue to make--doubt in Islamic criminal law not only
concerning but also ubiquitous.
In the face of harsh punishment in various other legal traditions, those
charged with enforcing criminal laws acknowledged doubt and sought to
allocate its burdens fairly in matters of both fact and law. For example, Late
Antique Roman jurists developed the legal canon in dubio pro reo, which
1. Robert M. Cover, Violence and the Word, 95 YALE L.J. 1601, 1601 (1986) (footnote
omitted).
2. On the "open-textured" and thus indeterminate nature of general rules, standards, and
principles when applied to ordinary cases, see H.L.A. HART, THE CONCEPT OF LAW 125-28 (2d ed.
1994). On the discomfort with doubt that textualism seeks to alleviate, see Peter Linzer, The Comfort of
Certainty: Plain Meaning and the ParolEvidence Rule, 71 FORDHAM L. REv. 799, 802 (2002), which
discusses psychological comforts derived from perceptions of textualist "plain meaning" as providing
"predictive stability," and notes that "[i]t is reassuring to believe that the words on a page provide order,
and order is unquestionably comforting."
3. For an overview of American criminal law, see generally JOSHUA DRESSLER,
UNDERSTANDING CRIMINAL LAW (6th ed. 2012). Recognition of the special nature of criminal law led
the United States Supreme Court to make federal criminal law exclusively legislative and to require
criminal legislation to be clear. See McBoyle v. United States, 283 U.S. 25, 27 (1931) ("Although it is
not likely that a criminal will carefully consider the text of the law before he murders or steals, it is
reasonable that a fair warning should be given . . . of what the law intends to do if a certain line is
passed. To make the warning fair, so far as possible the line should be clear."); United States v. Hudson
& Goodwin, 11 U.S. (7 Cranch) 32, 34 (1812) (barring federal common law crimes).
4. See, e.g., United States v. Bass, 404 U.S. 336, 348 (1971) (noting an "instinctive distaste
against men languishing in prison unless the lawmaker has clearly said they should" (quoting Henry J.
Friendly, Mr. Justice Frankfurter and the Reading of Statutes, reprinted in HENRY J. FRIENDLY,
BENCHMARKS 196,209 (1967))).
5. For an accessible overview of Islamic criminal law, see generally RUDOLPH PETERS,
CRIME AND PUNISHMENT IN ISLAMIC LAW (2005), which describes Islam's three categories of offenses:
(1) fixed, nondiscretionary crimes and penalties, including four agreed-upon offenses: illicit sexual
relations, false accusations of illicit sexual relations, theft, and intoxication; and three of disputed status:
apostasy, blasphemy, and highway robbery; (2) the laws of murder and personal injury; and (3)
discretionary penalties.
6. See BERNARD WEISS, THE SPIRIT OF ISLAMIC LAW 38, 101-09 (1998) (identifying a
"textualist-intentionalist bent" among Muslim jurists to describe their historical reliance on text to
determine divine legislative intent, and illustrating the jurist-centric textualist-intentionalism in the
interpretation of the law of theft based on the Qur'dn and related sources).
2015] "ReasonableDoubt" in Islamic Law
requires judges to acquit in cases of factual doubt. 7 Medieval English and

Continental European judges and jurors regularly avoided convictions when
they had doubts about the fairness of punishment! Contemporary civil lawyers
in France and Germany also frequently appeal to the presumption d'innocence
(presumption of innocence) and to Unschuldsvermutung (innocence-9
presumption), respectively, as requirements for acquittal in cases of doubt.
Likewise, the United States Supreme Court has directed judges and juries to
convict only if presented with factual proof of criminal culpability, proved
beyond a reasonable doubt.10 The Court also continues to apply the ancient rule
of lenity, albeit haltingly, as a principle instructing judges to construe
ambiguous penal statutes narrowly. 1' In all these ways, judges in disparate
legal systems have regularly devised strategies to avoid doubt's harmful
consequences in criminal law.
By contrast, a now popular notion of Islamic law favors a decidedly strict
textualist account of Islamic legal theory that leaves little room for discretion or
7. See Konrad Moser, "IN DUBIO PRO REO": DIE GESCHICHTLICHE ENTWICKLUNG DIESES
SATZES UND SEINE BEDEUTUNG IM HEUTIGEN DEUTSCHEN STRAFRECHT 77 (1933), cited in George P.
Fletcher, Two Kinds of Legal Rules: A Comparative Study of Burden-of-Persuasion Practices in
Criminal Cases, 77 YALE L.J. 880, 881 n.5 (1968).
8. See JAMES Q. WHITMAN, THE ORIGINS OF REASONABLE DOUBT 334-36 (2008); see also
JOHN H. LANGBEN, THE ORIGINS OF ADVERSARY CRIMINAL TRIAL 334-35 (2003) (describing common
truth-defeating techniques that prosecutors, jurors, and judges used to mitigate the application of capital
punishment for petty crimes in eighteenth- and nineteenth-century England).
9. Fletcher, supra note 7, at 880-81 nn.2-3 (describing these presumptions as modem
Continental restatements of the ancient Roman law maxims in dubio pro reo and nulla poena sine
culpa).
10. In re Winship, 397 U.S. 358, 362-63 (1970) ("[P]roof of a criminal charge beyond a
reasonable doubt is constitutionally required .... The standard provides concrete substance for the
presumption of innocence-that bedrock 'axiomatic and elementary' principle whose 'enforcement lies
at the foundation of the administration of our criminal law."') (citations omitted).
11. For the Court's earliest citation of the lenity rule, see UnitedStates v. Wiltberger, 18 U.S.
76, 95 (1820), stating that "[tihe rule that penal laws are to be construed strictly is perhaps not much less
old than construction itself." For recent applications, see, for example, Burrage v. United States, 134 S.
Ct. 881, 891 (2014), which applied the rule of lenity to reverse a sentencing enhancement against a
defendant for distributing cocaine that "result[ed]" in death, where the prosecution could not prove that
the defendant was the but-for cause of death; and United States v. Santos, 553 U.S. 507, 512 (2008),
which applied the rule of lenity to a money-laundering statute that left in doubt whether the statute's
prohibition on using "proceeds" from unlawful activity applied to "profits" or "receipts" from illegal
gambling activities. For recent rejections, see, for example, Abramski v. United States, 134 S. Ct. 2259,
2272 n.10 (2014), a 5-4 decision in which the majority concluded that the rule of lenity did not apply to
defeat the Court's defendant-disfavoring construction of a gun law provision because, "[a]lthough the
text creates some ambiguity, the context, structure, history, and purpose resolve it"; and United States v.
Hayes, 555 U.S. 415, 429 (2009), a 7-2 decision rejecting the rule of lenity on the grounds that the text,
purpose, and the drafting history clarify that "Congress defined 'misdemeanor crime of domestic
violence' to include an offense 'committed by' a person who had a specified domestic relationship with
the victim." Last Term, Justice Scalia registered his frustration with the Court's halting and inconsistent
application of the rule in strong terms: "If lenity has no role to play in a clear case such as this one, we
ought to stop pretending it is a genuine part of our jurisprudence. Contrary to the majority's miserly
approach, the rule of lenity applies whenever, after all legitimate tools of interpretation have been
exhausted, 'a reasonable doubt persists' [about the criminality of an act] .... " Abramski, 134 S. Ct. at
2281 (Scalia, J., dissenting) (quoting Moskal v. United States, 498 U.S. 103, 108 (1990)). For further
analysis, see Note, The New Rule of Lenity, 119 HARv.L. REV. 2420 (2006) (identifying a narrowed rule
of lenity in recent years through a review of the Rehnquist Court); and Sarah Newland, The Mercy of
Scalia: Statutory Construction and the Rule of Lenity, 29 HARV. C.R.-C.L. L. REV. 197 (1994)
(discussing Justice Scalia's support for lenity as a textualist tool for statutory interpretation).
doubt. To be sure, the textbook version of Islamic legal theory posits an

absolute form of legislative supremacy--divine legislative supremacy-
whereby God is the sole lawgiver. 12 In this scheme, God "legislated" a
comprehensive scheme of laws in the seventh century through a combination of
scripture (Qur'dn) and prophetic practice (Sunna). The Qur'dn and Sunna-the
latter as contained in collections of prophetic reports called .adfth-comprise
Islam's foundational legal texts. 13 These texts specify a set of fixed criminal
laws, violations of which trigger severe penalties ranging from corporal to
capital punishment. 14 In the most strictly textualist version of this theory,5
divine legislative supremacy meant applying Islamic criminal laws broadly.1
On this account, a combination of strict textualism and divine legislative
supremacy left no obvious mechanism or need for Muslim judges to
acknowledge, much less accommodate, doubt.
This popular account does not match the historical record. A close look at
the sources for Islamic legal history suggest that, when adjudicating criminal
cases, medieval Muslim judges and jurists faced the same factual and legal
doubts that afflicted judges in other legal traditions. With different institutional
structures and greater interpretive authority, these medieval Muslim judges and
jurists carved out a pronounced role for doubt, and in the process, further
defined and updated Islamic criminal law. They packaged their doctrine in the
form of a statement calling on judges to "avoid criminal punishments in cases
of doubt."' 16 I call this statement Islamic law's doubt canon: a doctrine that
came to be oft-repeated as a broadly recognized legal maxim. 7 The historical
record shows that the doubt canon was applied more frequently than the early
foundational texts, read alone, seem to require. The strict textualist approach to
doubt in Islamic law stood in contrast to a more prevalent tendency to construe
criminal laws narrowly-lenity style-whenever ambiguities arose from
application of old texts to new facts, or to avoid conviction andpunishment-
12. See, e.g., MUjAMMAD BAQIR AL-$ADR, LESSONS IN ISLAMIC JURISPRUDENCE 73-86 (Roy
P. Mottahedeh trans., 2003); MUHAMMAD B. IDRIS AL-SHAFII, ISLAMIC JURISPRUDENCE 102-05 (Majid
Khadduri trans., 1961).
13. Muslims regard the Qur'dn as the first source of Islamic law and understand Muhammad
to be the last prophet to have received revelations directly from God, which he recorded in the Qur'an.
Muslims take Mubammad's life example (Sunna) to be a second source of Islamic law. The Sunna was
recorded in thousands of disparate reports of the Prophet's words and actions, called hadth (and for
mainstream Shi' Muslims, the Surma also includes reports from the series of twelve Im~ms who
succeeded the Prophet). For further discussion of the sources of Islamic law, see MOHAMMAD HASHIM
KAMALI, PRINCIPLES OF ISLAMIC JURISPRUDENCE 16-55, 58-60 (3d ed. 2003); and HOSSEIN
MODARRESSI, AN INTRODUCTION TO SHII LAW: A BIBLIOGRAPHICAL STUDY (1984).
14. See PETERS, supranote 5.
15. See infra Subsection I.A.2.
16. See infra Section ll.B.
17. For a comparison between legal canons in Islamic law and American law, see Intisar A.
Rabb, Islamic Legal Maxims as Substantive Canons of Construction: Hudfid-Avoidance in Cases of
Doubt, 17 ISLAMIC L. & SOC'Y 63 (2010) [hereinafter Rabb, Islamic Legal Maxims]. Elsewhere, I have
referred to the doubt canon as "the Islamic rule of lenity," when emphasizing the aspects of legal
ambiguity inherent in the concept of shubha. See Intisar A. Rabb, The Islamic Rule ofLenity, 44 VAND.
J. TRANSNAT'L L. 1299, 1327-51 (2011) [hereinafter Rabb, Islamic Rule of Lenity]. By contrast, this
discussion is more comprehensive. I seek here to show the expansive definition and historical legal-
institutional implications ofshubha as factual, legal, and moral doubt.
2015] "Reasonable Doubt" in Islamic Law
reasonable doubt style-whenever evidence of the facts was questionable to

proving the claim. 18 In sum, as we will see, doubt became the framework for
key concepts and institutional functions undergirding the entire system of
Islamic criminal law.
Now it should be noted at the outset that the Islamic doctrine of doubt
came to have a meaning that was considerably different and far more expansive
than the Anglo-American doctrine of reasonable doubt. The Arabic word for
the concept, shubha, was a term of art that referred to doubt and ambiguity of
all types. That is, the Islamic doctrine of doubt referred to legality principles
and mitigating factors as diverse as the presumption of innocence, the rule of
lenity, doctrines of mistake of fact and mistake of law, and mercy.19 Rather
than the Anglo-American idea of a principally fact-based standard of proof, the
Islamic doctrine covered factual doubt, legal doubt, and even moral doubt
about the propriety of punishment.
Moreover, it is worth mentioning the extent to which the doubt canon
pervaded both legal circles and popular legal consciousness. The canon appears
in the earliest works of Islamic law with roots in the seventh and eighth
centuries. 20 One jurist living in the ninth century claimed that the doubt canon
was one of only a handful of principles on which all Muslim jurists had come
to a consensus. 21 In the tenth century, judges and jurists began to invoke the
canon regularly in legal treaties, judicial manuals, and reports of court cases
and other proceedings. 22 By the eleventh century, the canon was so well
entrenched that it appeared in literature as popular as the Arabian Nights.23 A
century later, it appeared in a noteworthy historical chronicle famously set to
verse by a twelfth-century Muslim historian writing from modem-day Portugal:
"The rule of 'avoiding punishment in cases of doubt' / Is a hadTth reported by
all scholars of reputable clout.' 24 Thereafter, the canon is featured in treatises
collecting legal maxims that proliferated in the thirteenth through sixteenth
centuries in virtually every Islamic legal school from every major center of the
18. See infra Subsection I.A.2.

19. For a more expanded discussion of the various types of "doubt" that the doubt canon
entailed, see infra Subsection 11.B.3.
20. See, e.g., ABC YCSUF, KITAB AL-KHARAi 303 (Dar al-Isldh 1981) (citing seventh-century
jurist Ibrahum al-NakhaTs citation of the canon on anonymous authority).
21. See IBN AL-MUNDHIR, AL-DMA' 113 (Dir al-Jindn 1986).
22. See, e.g., 2 ABC BAKR AL-RAzT JA$$AS, AI-.KAm AL-QUR'AN 108 (Muhammad al-$diq
QamhawT ed., Dar lhyd' al-Tur~th al-'Arabi 1986) (citing the doubt canon as a prophetic report in
conjunction with a discussion of Qur'Rnic legal rules); 1 AL-QAkp AL-TANOKHT, NIsHwAR AL-
MUIjAPARA WA-AKHBAR AL-MUDHAKARA 135 ('Abbd al-SMl1ji ed., Dar Sdir 1971) (describing a
famous episode of a reported incident through which Abdi YtIsuf is said to have become a judge for
using the doubt canon-cited as a prophetic report-to benefit the 'Abbdsid caliph Hadrn al-Rashid);
see also infra note 50; infra Section ll.B.
23. 2 ALF LAYLA WA-LAYLA night 298, at 527 (Ddr Sdir 1999). For an English account of
this story involving an Iraqi governor's invocation of the canon, and related stories from other sources
of medieval popular Arabic literature, see Maribel Fierro, Idra'i al-hudfld bi-l-shubuhat: When Lawful
Violence Meets Doubt, 5 HAWWA: J. WOMEN MIDDLE E. & ISLAMIC WORLD 208, 208-09 (2007).
24. 7 ABC AL-IASAN B. BASSAM AL-S-ANTARTNI, AL-DHAKHTRA FIMAItASIN AHL AL-JAZIRA
357 (Ibsin 'Abb~s ed., Dar al-Thaq~fa 1979) ("Inna dar a l-hudfidi bi l-shubahit /la-hadthun rawhu
kullu l-thiqdt.").
THE YALE JOURNAL OF INTERNATIONAL LAW [Vol. 40:41
Muslim world.25
Surprisingly-given its ubiquity-the doubt canon was not always well-
known to Islamic law. In fact, it turns out that the canon was not at all known
to the earliest Muslim jurists as a prophetic statement. The emergence and
growth of reasonable doubt during that period suggests that Muslim jurists'
interpretive moves were doing more work than their legislative supremacy
claims would suggest. The earliest written sources for Islamic law contain no
clear textual directives for "reasonable doubt," and certainly not in the form of
the doubt canon endowed with a prophetic pedigree. It is only the later sources,
some four centuries after the Prophet's death (after the tenth-century "closing
of the doors of interpretation"), that identify a prophetic-textual basis for doubt,
as a legal canon so solid that it came to be regarded as a prophetic report.
Despite indications that early Muslim jurists did not regard the doubt canon as
a hadath in the seventh through ninth centuries, by the tenth and eleventh
centuries, Muslim jurists spoke about the doubt canon as if it had been a
foundational text from the beginning. Yet, some contemporary sources assume
reasonable doubt doctrines were always present in Islamic law, and others
completely deny that it exists. Which is it? My examination of the principal
legal and historical sources from the long founding period of Islamic law
(seventh to eleventh centuries) explores what happened in the interim.
My central claim is that Muslim jurists generated a textual doctrine of
doubt in response to the changing socio-political context of the violent eleventh
century at a time when they were systematizing the law. In doing so, they
sought to define the boundaries of Islamic legitimacy for enforcing criminal
punishments in an attempt to decrease dubious caliph-ordered executions. Such
executions had always been problematic but were cast into relief with the
breakup of the Muslim empire and the accompanying need for jurists to further
systematize Islamic law. 26 Remarkably, these jurists used claims of doubt to
depart from the outcomes that Islam's foundational legal texts seemed to
authorize. Even more remarkably, they did so while claiming fealty to the text,
as faithful agents of a divine lawgiver who established the criminal laws in the
first place. That is, in view of Islam's strongly textualist ideal, these jurists hid
their role in constructing these doctrines under cover of textualism and divine
25. See, e.g., 2 AL-'Izz B. 'ABD AL-SALAM, AL-QAWA'ID AL-KUBRA 279-80 (Nazlh Kamdl
.Hammdd & 'Uthm5n Jumu'a Iumariyya eds., Dar al-Qalam 2d ed. 2007) (citing and listing various
applications of this standard formulation of the canon); IBN NUJAYM, AL-AS-BkH WA'L-NA/7,'IR 142
(Muhammad Muti' al-Hafiz ed., Dir al-Fikr 1983) (same); 4 SHMAB AL-DIN AL-QARAFI, ANWAR AL-
BUROQ F! ANWA' AL-FUROQ 1307 (D~r al-Salam 2001) (same); JALAL AL-DIN AL-SUYOTT, AL-ASHBAH
WA'L-NAZA'IR FTQAWA'ID WA-FURO' AL-SHAFI'IYYA 236-38 (Muhammad al-Mu'ta~im bi-llgh al-
Baghddi ed., Dar al-Kitdb al-'Arabi 1998) (same); see also INTIsAR A. RABB, DOUBT INISLAMIC LAW:
A HISTORY OF LEGAL MAxIMS, INTERPRETATION, AND ISLAMIC CRIMINAL LAW (2014) [hereinafter
DOUBT INISLAMIC LAW] (reviewing in Chapters 6 and 8 the history of Islamic substantive, procedural
and interpretive doubt after the eleventh century, in Sunni and Shl'! contexts, respectively).
26. Notoriously difficult to define, Islamic legitimacy as I use it most closely aligns to
Richard Fallon's helpfiul categories from American law of the legal, sociological, and moral criteria by
which jurists successfully assert their authority to define law. See Richard H. Fallon, Jr., Legitimacy and
the Constitution, 118 HARV. L. REv. 1787 (2005). For further analysis in the Islamic legal context, see
infra Section I.B.
legislative supremacy. How could Muslim jurists both claim to be textualists

and depart from the texts? Why did they? In the course of answering these
questions, my goal is to systematically uncover the means and motives by
which Muslim jurists managed to convert a practice-based doctrine of doubt
into a textual rule.
First, the means. I claim that Muslim jurists canonized, textualized, and
generalized the Islamic doctrine of doubt based on judicial practice. That is,
they identified a doctrine of doubt in early cases, which-with repetition as a
matter of judicial practice-became a legal canon in its oft-repeated, standard
form. This process represented the "canonization" of doubt. Muslim jurists then
converted the judicial practice of avoiding punishments in doubtful cases into a
legal text requiring it. To accomplish this "textualization" of doubt, they
claimed that Mulammad himself had uttered the statement calling on judges to
"avoid criminal punishments in cases of doubt." By giving it a prophetic
pedigree, these jurists rendered the canon into a normative legal text. Having
given the canon a textual basis of authority, these jurists then engaged in
sustained interpretive activity around punishment-mitigating doctrines and
heightened procedures for both capital and ordinary crimes, which together
represented the "generalization" of doubt. With a vastly expanded scope, the
doubt canon became a unifying "text" that jurists used to define crime and to
determine when it was proper for judges to enforce or avoid Islam's harsh
punishments.
Next, the motives. What explains this radical transformation of the
Islamic doctrine of doubt from judicial practice into binding legal text? The
answer, I will argue, has to do with the systematization of Islamic
jurisprudence accompanying the socio-political context surrounding the
breakup of the Islamic empire in the eleventh century. At the time jurists were
systematizing Islamic criminal law, they also witnessed rampant executions at
the hands of new local rulers seeking to solidify control of the central Muslim
lands (present-day Iran, Iraq, and surrounding lands). The risk of unjustified
political executions exacerbated existing moral anxieties on the part of the
jurists responsible for outlining the boundaries of legitimate punishment. As
the basis of Islamic law shifted from past practice to a system of texts,
doctrines of doubt would be helpful to Muslim jurists seeking to reinforce their
interpretive authority over law, ameliorate their anxieties over excessive
punishment, and perhaps to lessen the rates of punishment. Moreover, a textual
basis for doubt could provide a stronger argument against punishment
whenever judges could claim to be beset by factual or legal doubt. That is, a
textual rule allowed Muslim jurists to argue that their push against punishment
came not from personal anxieties surrounding executions but from the divine
law itself. In this way, even if they could not stop excessive punishments, at the
very least they could decline to authorize executions. They could accomplish
this by using the doubt doctrine to draw the contours of legitimate punishment
narrowly.
Part I examines the Case of Md 'iz, a landmark case from Islamic law's
founding period that vividly illustrates the debates surrounding interpretation of
Islam's foundational legal texts related to criminal law. Part II presents the
means by which Muslim jurists canonized, textualized, and generalized the
doubt canon. Part III explores the motives for which these jurists converted the
judicial canon into a foundational text. In the end, I conclude that Muslim
jurists operating in the eleventh century were systematizing the law and needed
a textual doctrine of doubt to solidify their authority, ameliorate their moral
concerns at authorizing punishment on dubious grounds, and check the political
ruler's arbitrary and excessive use of punishment. Even if the jurists could not
prevent executions, the newly textual status of the doubt canon rendered it a
powerful tool for them to argue against the legitimacy of punishment as a tool
for state violence and social control.
Before discussing the medieval cases and controversies in Islamic legal

interpretation that led to its doctrines of doubt, it may be worth mentioning a
note about their modem relevance. The surprising turn of events that led to the
construction and prevalence of doubt in Islamic criminal law is like a modem-
day U.S. constitutional amendment that passes without going through any
legislation or ratification process and that history forgets was not a part of the
original document. 2 The forgotten history of doubt in Islamic law is crucial to
understanding the most pressing questions surrounding the reemergence of
Islamic law as state law in modem constitutional and criminal contexts. To be
sure, the Islamic state is on the rise, again.28 Since the 1970s, leaders of some
thirty-eight Muslim-majority countries have incorporated Islamic law into their
constitutions as "a source" or "the source" of state law-including at least one
in the aftermath of the Arab uprisings, which began in 2010.29 During that time,
the legislatures of over a dozen countries enacted Islamic criminal law codes.3 °
27. For an insightful analogy in the American context comparing the statutory approach of an
early U.S. Supreme Court case to constitutional amendment, see infra note 70.
28. See generally NOAH FELDMAN, THE FALL AND RISE OF THE ISLAMIC STATE (2008)
(analyzing the increasing popularity of calls for sharTa as a basis for state law in Middle Eastern
contexts); RAN HIRSCHL, CONSTITUTIONAL THEOCRACY (2010) (surveying the increasing role of
Islamic and other religious laws as modem state laws).
29. See Dawood I. Ahmed & Tom Ginsburg, ConstitutionalIslamization and Human Rights:
The Surprising Origin and Spread of Islamic Supremacy in Constitutions, 54 VA. J. INT'L L.
(forthcoming 2015) (providing an empirical review of all Islamic law constitutional clauses); Intisar A.
Rabb, We the Jurists:Islamic Constitutionalism in Iraq, 10 U. PA. J. CONST. L. 527 (2008) (providing a
qualitative review of one such clause in the context of three common types of Islamic
constitutionalization). For developments after the Arab uprisings, see S.A. Arjomand, Middle Eastern
Constitutional and Ideological Revolutions and the Rise of Juristocracy, 19 CONSTELLATIONS 204
(2012); Karim Mezran, Constitutionalism and Islam in Libya, in CONSTITUTIONALISM IN ISLAMIC
COUNTRIES: BETWEEN UPHEAVAL AND CONTINUITY 513 (Rainer Grote & Tilmann J. R6der eds., 2012);
Intisar A. Rabb, The Least Religious Branch? Judicial Review and the New Islamic Constitutionalism,
17 UCLA J. INT'L L. & FOREIGN AFF. 72 (2013).
30. Penal codes based on Islamic Law include Afghanistan, PENAL CODE of 1976; Brunei,
SYARIAH PENAL CODE of 2014; Iran, QANON-I MUJAZAT-I ISLAMI-YI IRAN [ISLAMIC PENAL CODE] of
1991, as amended in 1996 and 2013; Kuwait, QANON AL-JAZA' [PENAL CODE], Law No. 16 of 1960, as
amended by Law No. 31 of 1970; Libya, QANON AL-'UQOBAT [PENAL CODE] of 1953, as amended by
Law No. 70 of 1973 and Law No. 4 of 2002; Maldives, PENAL CODE ACT NO. 1/81 of 1961, as amended
in 2014; Oman, QANON AL-JAZA' [PENAL CODE] of 1974; Pakistan, PENAL CODE of 1860, amended by
Hudood Ordinance, Law Nos. 7 and 9 of 1979; Qatar, QANON AL-'UQOBAT [PENAL CODE], Law No. 14
of 1971, as replaced by Law No. 11 of 2004; Sudan, PENAL CODE of 2003; United Arab Emirates,
Just as Islamic law arises regularly in constitutional politics, Islamic criminal

31
law is increasingly on the agenda in parliamentary and provincial politics.
A major feature of Islamic constitutional and criminal law is the
significant degree to which its partisans seek to deploy history for present use.
The new Islamic constitutions and codes require many judges in the Muslim
world to apply Islamic law in their decisions.32 These judges tend to appeal to
conceptions of Islamic law drawn from its foundational texts and
understandings from the ever-authoritative founding period. 33 At least one
American firm operating in the Muslim world has appealed to classical rules of
Islamic criminal law to address liability in a wrongful death action.34 Local
QANON AL-'UQOB.T AL-ITI.ADI [UNIFORM PENAL CODE], Law No. 3 of 1987; Yemen, [QANYN] AL-
JARKIM WA'L-'UQOBAT [CRIME AND PUNISHMENT CODE], Law No. 12 of 1994; as well as provinces in
Malaysia (Kelantan), SYARIAH CRIMINAL CODE (II) ACT OF 1993 [STATE OF KELANTAN]; Nigeria (ten
northern Nigerian states between 2000 and 2001: Bauchi, Gombe, Jigawa, Kaduna, Kano, Katsina,
Kebbi, Sokoto, Yobe, Zarnfara), BAUCHI STATE PENAL CODE of 2001; GOMBE STATE PENAL LAW of
2001; JIGAWA STATE PENAL CODE, Law No. 12 of 2000; KADUNA STATE PENAL CODE of 2002; KANO
STATE PENAL CODE of 2000; KATSINA STATE PENAL CODE of 2001; KEBBI STATE PENAL CODE, Law
No. 21 of 2000; SOKOTO STATE PENAL CODE of 2000; YOBE STATE PENAL CODE, Law No. 8 of 2001;
ZAMFARA STATE PENAL CODE, Law No. 10 of 2000; see also 4 PHILIP OSTIEN, SHARIA
IMPLEMENTATION IN NORTHERN NIGERIA, 1999-2006: A SOURCEBOOK (2007) (citing and reproducing
these Nigerian state penal codes where available); HAUWA IBRAHIM, PRACTICING SHARIAH LAW, app. D
(2012) (providing the text of the Sokoto and Zamfara State Shariah Penal Laws of 2000); and Indonesia
(Aceh), QANUN JINAYAT [PENAL CODE] OF 2009, as amended in 2014. Non-codified practices have
been reported informally in Algeria (in 1993), parts of Iraq controlled by the so-called Islamic State of
Iraq and the Levant (since 2014), Northern Mali (since 2012), the Swat Valley Region of Pakistan (since
2004), Somalia (since the 1990s), and Syria (since 2013).
31. For analysis, see generally Rudolph Peters, The Islamization of Criminal Law: A
ComparativeAnalysis, 34 DIE WELT DES ISLAMS 246 (1994). For country-specific studies, see ASMA
JAHANGIR & HINA JILANI, THE HUDOOD ORDINANCES: A DIVINE SANCTION? (1988); Mohammad
Hashim Kamali, Punishment in Islamic Law: A Critique of the HudFidBill of Kelantan, Malaysia, 13
ARAB L.Q. 203 (1998); Ziba Mir-Hosseini, Criminalising Sexuality: Zina Laws as Violence Against
Women in Muslim Contexts, 8 SUR - INT'L J. HUM. RTS. 7 (2011); Asifa Quraishi, Her Honor: An
Islamic Critique of the Rape Laws of Pakistanfrom a Woman-Sensitive Perspective, 18 MICH. J. INT'L
L. 287 (1996); Abdel Salam Sidahmed, Problems in Contemporary Applications of Islamic Criminal
Sanctions: The Penaltyfor Adultery in Relation to Women, 28 BRIT. J. MIDDLE E. STUDS. 187 (2001).
32. See, e.g., CLARK B. LOMBARDI, STATE LAW AS ISLAMIC LAW IN MODERN EGYPT: THE
INCORPORATION OF THE SHARI'A INTO EGYPTIAN CONSTITUTIONAL LAW (2006); Clark B. Lombardi,
Egypt's Supreme Constitutional Court: Managing Constitutional Conflict in an Authoritarian,
Aspirationally "Islamic" State, 3 J. COMP. L. 234-5 3 (2008).
33. E.g., Case No. 45 of Judicial Year 28, 12 SCC 1359 (Sup. Const. Ct. 2009) (Egypt)
(citing legal canons and Islamic principles in upholding personal status law reforms requiring that
marriages be registered to be recognized as valid); Case No. 23 of Judicial Year 20, 12 SCC 307 (Sup.
Const. Ct. 2007) (Egypt) (citing Qur'dnic verses and historical juristic interpretations of family law
provisions to uphold an Art. 2 "sharl'aclause" challenge to alimony requirements of personal status law
reforms); Case No. 62, Judicial Year 19, 12 SCC 92 (Sup. Const. Ct. 2006) (Egypt) (citing and applying
Qur'5nic verses and legal canons on contractual performance to uphold a challenged property law
statute). I have used SCC as the English initials for the name of the Supreme Constitutional Court and
its reporter, al-Mahkama al-Dustariyya al-'Ulya. For a translation and analysis of one SCC case, see
Nathan J. Brown & Clark B. Lombardi, Translation: The Supreme ConstitutionalCourt of Egypt on
Islamic Law, Veiling and Civil Rights: An Annotated Translation of Supreme ConstitutionalCourt of
Egypt Case No. 8 ofJudicialYear 17 (May 18, 1996), 21 AM. U. INT'L L. REv. 437 (2006).
34. See Rob Quinn, U.S. Firm Calls for Islamic Law in Lawsuit: Blackwater Boss Seeks
Sharia Law to Dodge Afghan Crash Suit, NEWSER (June 19, 2008, 7:51 AM), http://www.newser.com
/story/30373/us-firm-calls-for-islamic-law-in-lawsuit.html (describing Blackwater's request to apply
Islamic law in a federal suit for the deaths of three U.S. soldiers in Afghanistan, as the firm would
escape liability under classical Islamic "criminal" law rules of personal injury). The suit ultimately
settled.
politicians have incorporated classical Islamic criminal law rules into state and
regional codes, usually in attempts to assert power as well as an Islamic
identity based on the popular notion of Islamic law as a text-based system of
harsh criminal rules. 35 Criminal defense lawyers operating in these regimes
have appealed to classical Islamic criminal law precepts-including the doubt
canon and the Case of Mj 'iz-to defend their clients, usually
36
in attempts to
bring more pragmatic understandings of Islamic law to bear.
In all of these modem manifestations, Islamic law reveals itself to be both
a textualist and an originalist legal tradition. That is, any invocation of Islamic
law in a modem constitution or code generally contemplates an appeal back to
Islam's foundational texts and its founding period.37 Yet, the scholarly and
judicial reviews of the history of doubt in Islamic law have been
impressionistic at best and overlooked at worst. 38 This Article explores the
35. The overwhelming majority of modern Islamic criminal law codes exclude the doubt
doctrine based on what seems to be an ahistorical textualist view of Islamic law adopted by the
parliamentarians who enacted them for political gains. See Vanja Hamzi6, Nigeria, in CONTROL AND
SEXUALITY: THE REVIVAL OF ZINA LAWS IN MUSLIM CONTEXTS 119 (Vanja Harnzi6 & Ziba Mir-
Hosseini eds., 2010) (arguing that "the introduction of hudad laws came as a consequence of political
crisis and opportunism characterized by patriarchal oppression and the misuse of religion as a means of
societal control") (citations omitted). The very texts of the codes suggest that, in a departure from
historical Islamic legal practice, the modem parliamentarians drafting them likely consulted
foundational and other early texts-which typically do not include the doubt canon. Neither trained in
Islamic law nor in Islamic history, these politicians attempted to construct new laws based on the Qur'dn
and hadth as well as basic abridgements of legal works, failing to account for the judicial practices that
historically incorporated doubt or to the developed works of substantive Islamic criminal law and legal
maxims, which included the doubt canon as central and counted the doubt canon as a foundational text.
For a list of codes, see supra note 30. One exception to that rule is a recent reform to the Iranian Penal
Code, which incorporated the doubt canon after much wrangling between textualist and pragmatic
jurists, the latter of which-trained as specialists in classical Islamic law and often in history-in
conjunction with secularly trained comparative lawyers and criminal law experts, demanded it. The
juristic, parliamentary, and popular debates giving rise to the recent Iranian criminal law reforms are the
subject of another study. For the relevant provisions, see QANiJN-I MUJAZAT-I ISLAMI-I IRAN [IRANIAN
ISLAMIC PENAL CODE] 1392 [2013], arts. 120-21.
36. A prominent example is the case of Amina Lawal, a woman convicted in 2002 of adultery
and sentenced to death under Nigeria's penal code. Lawal's lawyers successfully appealed to overturn
the conviction based on citation to the Case of Md iz, the doubt canon, and other procedural lessons that
were regular features of historical Islamic legal practice. See IBRAHIM, supra note 30, at 148-83
(describing the role of the defense counsel at trial and on appeal); OSTMEN, supra note 30, at 52-82
(providing case transcripts of the Amina Lawal proceedings); Philip Ostien & Albert Dekker, Sharia
and National Law in Nigeria, in SHARIA INCORPORATED: A COMPARATIVE OVERVIEW OF THE LEGAL
SYSTEMS OF TWELVE MUSLIM COUNTRIES IN PAST AND PRESENT 589-93 (Jan Michiel Otto ed., 2010)
(describing the adoption and application of Islamic criminal law in Northern Nigeria).
37. Cf FELDMAN, supra note 28, at 4-5 (comparing Muslims' "active and continuing
engagement with the constitutional past" to the way in which "Madison, Jefferson, and Hamilton
continued to shape the American constitutional tradition from beyond the grave [such that] it is
impossible to understand arguments about the American Constitution today without taking these
founding fathers into account").
38. I am aware of five articles and four monographs on the Islamic doubt doctrine (besides
my own), all of which are recent, and almost all of which (with the exception of the article by Maribel
Fierro) are Arabic or Persian catalogs of the legal doctrine that do not analyze its history or social
context. The articles are Muhammad Bahrdmi, BarrasT va tahlTl-ifiqh! va huqfq!-i qd 'idah-i 'tudra' al-
huduid bi'l-shubahit', 5-6 DIDGAHHA-YI IJUQOQI 19 (1997-8); Mubammad Mdsawi Bujnflrdi, LW
qd'idah-ifiqhi" Qa'idah-i 'tudra' al-.udfid bi'l-shubahat,' 8 FASLNAMAH-I DIDGAHHA-Y IJUQOQI 11
(1986-87); Fierro, supra note 23; Muhammad Muhammadi GilnI, Ashind 7 bd q& idah-i 'al-Hudfid
tudra' bi'l-shubahat,' 5 MAJALLAH-I QAI)A'T VA-IUQ3QI-I DADGUSTARI-I JUMHORI-I ISLAMI-I RAN 15-
16 (1996-97); RiO Usthdl, Qd 'idah-idar', 34 FIQH- AHL-I BAYT 46 (2005) (Part 1); and 37 id.at 71-94
history in question.
I. TRE RECURRING CASE OF MA 'IZ AND QUESTIONS OF ISLAMIC LEGITIMACY
This section examines the debates about the famous Case of M 'iz, and
why these debates-and the jurists who engaged them-mattered so much
historically to definitions of Islamic law and governance. I argue that medieval
Muslim jurists' debates about Md'iz were their attempts to define the proper
boundaries of Islamic legal interpretation, institutions, and legitimacy.39 On the
interpretive level, jurists' internal debates about M5'iz reveal disputes about
how best to resolve uncertainty and doubt at the level of adjudication. On the
institutional level, the debates point to the scope of juristic competence to
resolve doubt. And on the level of legitimacy, the debates suggest that Muslim
jurists strategically used doubt and claims of their institutional competence to
resolve it as tools for asserting their own authority and thereby defining the
classical Islamic constitutional structure for legitimate applications of criminal
04
law.
Throughout this discussion, it is important to bear in mind that, in
historical Islamic contexts, there were no constitutionally defined institutions of
law and governance. 4' But to say that the institutions were not defined top-
down by a formal constitution is not to say that these institutions did not exist.
An informal constitutional arrangement unfolded whereby a class of scholar-
jurists assumed the power to interpret sharT'a as a system of law, and
monarchical rulers called caliphs assumed the power to enforce it.42 Between
them were judges, whom the caliph appointed and who looked to the jurists for
formulations of law. 43 From this arrangement, scholars commenting on Islamic
(Part II). The monographs are 'AQILA HUSAYN, AL-SHUBAHAT AL-MUSQITA LI'L-uUDOD (Dir Ibn Hazm
2003); SA'ID MANS13RI, QAVA'ID-I FIQH: BAKHSH-I JAZA'I-FIQH-1 TATBTQI-YI QAIDAH-I DAR' (Tadbir
1997); MUUAMMAD HASAN AL-RABBANT, QAIDAT AL-DAR' (Mu'assasat al-Nashr al-Islni al-Tabi'a li-
Jami'at al-Mudarrisin 2007); and SA'ID B. MISFIR AL-DAGHGHAR AL-WADI'!, ATHAR AL-SHUBAHT Fl
DAk' AL-HUDOD (Maktabat al-Tawba 1998).
39. Cf Fallon, supra note 26, at 1849-51 (outlining an "overall" legitimacy that combines
legal, sociological, and moral criteria for judicial authority).
40. Other scholars have recognized the link between jurists and legitimacy. E.g., FELDMAN,
supra note 28, at 2; WAEL B. HALLAQ, AUTHORITY, CONTINUITY, AND CHANGE IN ISLAMIC LAW 159-
83 (2001). But none have examined closely how jurists used interpretation to define institutional roles,
the elaboration of that process in significant part through criminal law, or the reasons that explain or
justify both moves. This section seeks to take up that task in the course of arguing that Muslim jurists
used specific interpretive maneuvers to expand their power over criminal law and limit that of the
executive through defining the contours of legitimate punishment.
41. In fact, there were no constitutions until Muslim-majority countries adopted American-
and European-style constitutions at the turn of the twentieth century (in the Ottoman Empire in 1876 and
in Iran in 1906), and then again after the post-World War II colonialism in large parts of the Muslim
World. Even where there were constitutions, there was a weak tradition of constitutionalism and few
inroads to democracy until the 2010 uprisings in the Arab world. See NATHAN BROWN, CONSTITUTIONS
IN A NONCONSTITUTIONAL WORLD (2002).
42. FELDMAN, supra note 28, at 2.
43. See WAEL B. HALLAQ, SHARI'A: THEORY, PRACTICE, TRANSFORMATIONS 176-81 (2009).
To be sure, several other officials exercised criminal law jurisdiction outside of the scheme. However,
this tripartite scheme provided the main pivots for law and order, and-as elaborated below-even in
those contexts, the caliphal officials were in principle subject to juridical articulations of law that set the
upper limits of punishment.
governance structures have discerned distinct institutions of law and

governance, and a separation of powers of sorts between them.44 What remains
to be explored is precisely how Muslim jurists defined the boundaries of those
powers organically and-in some respects-from the bottom up. Here, I aim to
demonstrate, using the illustrative example of Md'iz, how these definitions of
institutional power and institutional relations took shape through Muslim
jurists' interpretive debates over criminal law. In short, through these
interpretive debates, Muslim jurists used reasonable doubt to define their own
power vis-A-vis other institutions.
A. Early Debates on Approaches to Islamic Criminal Law
Why so much ado about Ma'iz? For starters, the Case of M 'iz is a rather
complicated basis for an early precedent, as it is not entirely clear what
happened. As a text, it contained a "foundational uncertainty" that sparked
debates over its meaning and precedential value for criminal law.45 Here is the
full account:
1. The Recurring Case of Md'iz

Md'iz was a zealous new convert, who apparently had repeatedly
committed adultery (zind). He came to the Prophet to confess, begging to be
punished. The Prophet initially sent Md'iz away, declining to hear the case.
Md'iz came back a second time and a third, each time renewing his confession
and requesting punishment, each time with the same result. On the fourth time,
the Prophet finally spoke. He asked some of Md'iz's neighbors about the
defendant's state of mind. The Prophet then suggested that Md'iz had not
committed adultery within the full meaning of the term, but perhaps had merely
"kissed or winked or looked at" another woman. Md'iz insisted that he was46of
sound mind and had committed adultery within the full meaning of the term.
In this case, the Prophet eventually pronounced a verdict of guilt, but did
not punish him. In fact, he said nothing about punishment. Not waiting for
specific instructions from the Prophet, the townspeople took the matter into
their own hands and enforced the harsher of the two punishments: death by
stoning. When the townspeople informed the Prophet that they had carried out
44. FELDMAN, supra note 28, at 2, 4, 68 (noting an "institutional separation of scholars and
state").
45. For use of this term to refer to analogous problems that arise in constitutional and
international law, which also lack legislatures to specify and update legal norms, see Darryl Levinson &
Jack Goldsmith, Law for States: InternationalLaw, ConstitutionalLaw, Public Law, 122 HARV. L. REV.
1791, 1794 (2009).
46. See 9 IBN ABI SHAYBA, MUSANNAF nos. 21254, 29240-41 (Hamad b. 'Abd Allah al-
Jumu'a & Muhammad b. Ibr'him al-Lal.idan eds., 2007). In addition to this pre-canonical collection of
hadith, which records the earliest known version of the case, the Case of Md 'iz also appeared in the two
main canonical collections of prophetic reports. See 4 MUHAMMAD B. ISMAIL AL-BUKHARI, SAHIj no.
6824 (Mutafa al-Dhahabi ed., Ddr al-Hadlith 2000) [hereinafter BUKHARI]; MUSLIM B. HAJJAJ AL-
QUSHAYRI, SAIT, nos. 4198-4206 (AbOi Suhayb al-Karami ed., Bayt al-Afkdr al-Dawliyya 1998)
[hereinafter MUSLIM].
capital punishment, he exclaimed in dismay that they "should have let M!'iz
go!" If Md'iz had repented, the Prophet added, "God would surely have
accepted his repentance," which would have absolved him of punishment.47
Adultery in Islamic law was a serious moral infraction, with high criminal
stakes. The Qur'dn prohibits fornication and adultery (both called zin5) and
directs judges to "flog those guilty of zind, whether male or female, with 100
lashes each.",48 In addition, though stoning was not a Qur'dnic punishment,
early Muslims understood the Sunna as also authorizing death by stoning for
adultery convictions. 49 If his confession were to be believed, Md'iz had
violated the Islamic law against adultery and would be liable for punishment.
From his concluding statement, it appears that the Prophet had been
reluctant to punish Md'iz, and on one reading, he had never intended to enforce
the punishment at all. Even though he had convicted Md'iz, he had "doubt" (as
most later Muslim jurists would term it), perhaps about whether the law
covered the act of the defendant, whether the procedures deployed to assess the
defendant's act and his intent were sufficient to prove that he had committed
zind, or whether punishment was warranted. It was quite possible that Md'iz
would repent from the crime, or that he already had done so through confessing
and then attempting to flee, in which case the Prophet may have preferred to
avoid punishment. On these grounds, the Prophet-who of all people should
have been beholden to follow God's commands-was willing to set aside a
strict reading of the text.
This case sparked tremendous debate amongst Muslim jurists. 50 These
47. See IBN ABI SHAYBA, supranote 46. The other ninth-century sources collecting prophetic
reports of early "cases" like these are replete with similar stories of individual confessions and hesitance
to punish. See, e.g., BUKHART, supra note 46, no. 6823 (reporting that a man came to the Prophet
confessing to having committed a serious crime multiple times, until the Prophet responded that his
joining of congregational prayers indicated repentance, so God had forgiven him).
48. See THE QUR'AN 17:32 (prohibiting zint--defined as fornication or adultery); id. at 24:2
(specifying a punishment of flogging or home imprisonment).
49. On the introduction of reports about stoning as a Qur'dnic punishment, see Hossein
Modarressi, Early Debates on the Integrity of the Qur'dn: A Brief Survey, 77 STUDIA ISLAMICA 5-10
(1993).
50. The case featured in the criminal law section of all the extant major collections of
prophetic reports (which provided the raw materials for the later tenth- and eleventh-century
systematization of law), from the eighth century through the eleventh century. See BUKHAR, supra note
46; see also 1 ABO AL-HASAN AL-MAwARDI, KITAB AL-IHUDOD MIN AL-.tkwT AL-KABIR 206 n.1
(Ibrahim b. 'Ali Sanduqji ed., 1995) (listing other major hadith sources, including 'Abd al-Razziq,
Ahmad b. Hanbal, Drimi, AbQ Ddwid, Tirmidhi, Ibn Majah, Ibn al-Mundhir, TahwI, Tabarani, al-
.Hakim al-Nayssbiri, and Bayhaqi). The case also featured in the criminal law chapters of all major legal
treatises (where the debates played out) from the eighth century through the present day. For a sampling
of those discussions in legal treatises, in chronological order by author, see ABO YOSUF, IKHTILAf ABT
H ANTFA WA-IBN ABT LAYLA 156 (Abi 'l-Wafa' al-Afghsni ed., Matba'at al-Wafa- 1938); ABO YOsUF,
KITAB AL-ATHAR 157 (AbY 'l-Wafia' ed., Lajnat lhya' al-Ma'drif a-Nu'mniyya 1936); 7 MLTUAMMAD
B. IDRIS AL-SHAFI'T, UMM 498 (Ahmad Badr al-Din Hassin ed., D~r Qutayba 1996); 2 JA$SA$, supra
note 22, at 108; 3 id. at 263-64; 11 ABO AL-HUSAYN AIMAD B. MUHAMMAD AL-QUDORT, TAJRID: AL-
MAWSO'A AL-FIQHtYYA AL-MUQARANA 5891, 5949-50 (Muhammad Abmad al-Sirj & 'AlI Jumu'a
Mulhammad eds., DAr al-Salgm 2004); MAWARDI, supra; 12 IBN IJAZM, AL-MUttALLA Bt'L-ATHAR 18-
22 ('Abd al-Ghaff-r Sulaymtn al-Bind.ri ed., Dar al-Kutub al-'lmiyya 1988); 2 BURItAIJ AL-DIN AL-
MARGHINANI, AL-HIDAYA SHARI BIDAYAT AL-MUBTADI' 735-36 (Muhammad Mulhammad T8rmir &
debates emerged in large part because the jurists could not agree on the precise
factual account or the normative significance of the case. They generally
agreed on the eventual outcome: Md'iz was punished despite the Prophet's
objections.51 They also agreed that the case was important: its provenance in
the founding period with the Prophet presiding as a judge made it a source of
normative guidance for future generations. But they disagreed about which
outcome (punishment
52 or release), and therefore which interpretive approach,
was correct.
2. The InterpretiveDebates about Md'iz
Muslim jurists devised two interpretive approaches to the Case of M 'iz,

which produced opposite normative outcomes. In essence, the precedential rule
arising out of the Case ofMd 'iz differed radically depending on whether jurists
evaluated the case according to textual as opposed to contextual or pragmatic
criteria. The first option, enforcing punishment, seemed to accord best with the
text. The second option, avoiding punishment, seemed to depart from the text
and be responsive to extratextual context. Yet most jurists--claiming to be
textualists-understood the case to require the second, extratextual option:
avoiding punishment. What was the basis for their claim and the tenor of the
debates by which the strictly textualist jurists begged to differ?
One interpretive approach, which I call "strict textualism," advocated
looking only at the Qur'dnic text to determine matters of guilt and punishment.
According to this approach, the townspeople were correct to punish Md'iz. He
had violated the textual prohibition against adultery, as the Prophet had
concluded in his pronouncement of guilt. The townspeople were merely
enforcing the sanction they believed (however incorrectly) to be required by
the Qur'dnic text. Even though an American or civil lawyer might expect the
judge to order punishment, on their account, all that strict textualists needed to
authorize punishment in the early Islamic context was the conviction itself. The
punishment was provided for in the divine text, which superseded the authority
of the judge (including the Prophet acting as a judge). In other words, the idea
H~fiz 'Ashiir Hfiz eds., Dar al-Salm 2000); and IBN ABI AL-QASIM, AL-WAD11I FT SHAR.U
MUKHTA$AR AL-KHIRAQT 413 ('Abd al-Malik b. 'Abd Allah b. Duhaysh ed., Dar Khidr 2000). For a
similar discussion in a well-known treatise of Islamic political theory, see IBN QAYYIM AL-JAWZ1YYA,
AL-TURUQ AL-1:IUKMIYYA FT 'L-SIYASA AL-SHAR'IYYA 87 (Muhbammad Jaml Ghdzi ed., Matba'at al-
Madani 1978).
51. To be sure, there are some questions about what happened: whether the Prophet
authorized punishment and then regretted it, whether Md'iz had in fact repented verbally or through
trying to escape once the punishment had commenced, or whether the Prophet's reasons for rebuking the
townspeople was out of mercy. Gaps in the "record" of this kind were typical for early cases, which
were reported in the hadTth literature with extremely sparse details about the litigants' conversations, the
outcome of the case, and the reasoning. These gaps helped provide the jurists with room for interpretive
maneuvering, that is, filling in the gaps about what happened with normative rules about what should
have happened based on the doctrines that they later devised-as this Article aims to illustrate through
tracing the evolution of the doubt doctrine from and as applied to the Case ofM 'iz. For further notes on
the historiography of cases in early Islamic legal sources, see infra note 148.
52. For example, compare QUDORI, supra note 50, at 5949-50 (demonstrating the pragmatic
textualist approach of a Hanafi jurist) with 12 IBN HAZM, supra note 50, at 18-22 (exemplifying the
strict textualist approach of a ZhirI contemporary of Qudfii).
of divine legislative supremacy was represented in the text rather than in

judicial discretion-even that of Prophetic authority. This explained why the
people had carried out the punishment to the objection of the Prophet. This
approach gave rise to the strict textualist rule that judges should enforce
punishments whenever caliphal officials could provide judges with some
evidence that a crime had been committed.
The second approach, which I call "pragmatic textualism," suggested that
even Qur'dnic textual rules were qualified by extratextual concerns that
affected whether both convictions and punishments were warranted in
particular cases.53 Jurists adopting this view read the Qur'dnic text in the
context of the Prophet's objection to punishment to conclude that the objection
should have carried the day. For them, the Prophet had identified
countervailing considerations-having to do with repentance and rehabilitation,
mercy, and antipathy toward the death penalty on dubious grounds-which
should have prevented punishment. This approach gave rise to the contextualist
rule that judges should avoid punishments whenever such countervailing
considerations threw the propriety of punishment into "doubt."
a. The Case for Enforcing the Punishment: Strict Textualist

Readings of the Case of Md'iz
For the jurists who maintained that the townspeople had properly
enforced the punishment against Md'iz, it was God himself who had required
the punishment. God was the sole legislator. The dictates of His texts therefore
bound both the Prophet and the townspeople. That is, not even the Prophet
could depart from the rule requiring punishment. The strict textualists looked to
the Case of Md 'iz alongside other early cases both to justify their interpretive
approach to that case and to distill specific textualist legal precepts about when
to enforce or avoid punishment.
Their argument was colorfully advanced by the Muslim jurist known as
53. I use this term to reflect the concept of pragmatic, practical, or dynamic statutory
interpretation elaborated by William Eskridge. See WILLIAM N. ESKRIDGE, JR., DYNAMIC STATUTORY
INTERPRETATION 275-306 (1994) (noting the ongoing use of legal canons as a feature of dynamic
interpretation); William N. Eskridge, Jr., Norms, Empiricism, and Canons in Statutory Interpretation,66
U. CHI. L. REV. 671, 671 (1999) (describing the eclectic approaches to statutory interpretation deployed
by judges and scholars as pragmatic); see also William N. Eskridge, Jr. & Philip P. Frickey, Statutory
Interpretationas PracticalReasoning, 42 STAN. L. REV. 42 321, 322 n.3 (1990) ("By 'practical reason,'
we mean an approach that eschews objectivist theories in favor of a mixture of inductive and deductive
reasoning (similar to the practice of the common law), seeking contextual justification for the best legal
answer among the potential alternatives."). To be sure, John Manning has insisted that modem
textualism is different from old formalist textualism, in that the modem version accommodates
"contextual readings of statutory text" by recognizing that "language has meaning only in its social and
linguistic context." John Manning, The Absurdity Doctrine, 116 HARV. L. REV. 2387, 2455 (2003); see
also John Manning, Textualism and the Equity of the Statute, 101 COLUM. L. REV. 1, 108-15 (2001)
("[M]odem textualists acknowledge that language has meaning only in context."). Acknowledging
Manning's inclusion of semantic and social context as his version of what Eskridge labeled the "new
textualism," Eskridge called Manning's approach "contextualist textualism." See William N. Eskridge,
Jr., No Frills Textualism, 119 HARv. L. REV. 2041, 2051 (2006) (reviewing ADRIAN VERMEULE,
JUDGING UNDER UNCERTAINTY: AN INSTITUTIONAL THEORY OF LEGAL INTERPRETATION (2006));
William N. Eskridge, Jr., The New Textualism, 37 UCLA L. REV. 621-91 (1990).
Dawtid al-Zahir-that is, "D wdd the Textualist"-who established the strict
textualist school of Islamic law.54 To begin with, he argued, God required
judges to appeal only to the Qur'dn and authentic, clear reports of the Sunna
without permitting human interpretation." His approach to criminal law was to
apply each rule according to the plain meaning of the text. Whenever a crime
was established, judges were to enforce the associated punishment.5 6 Further, a
judge was not to express personal preferences or exercise any discretion in such
cases. 57 Considerations of mercy, moral concerns, or contextual facts were
irrelevant.
For proof, textualist judges looked to the Qur'an itself, which
commanded in the context of adultery prohibition and punishment: "do not let
mercy dissuade you from imposing the punishment." 58 Textualists also looked
to reports of other early cases that more clearly showed instances where the
Prophet acknowledged that even he was required to apply the punishments
specified in the Qur'dnic texts, without exercising discretion of his own. For
example, in the Case of the MakhzigmT Thief, Muhammad had convicted a
woman for theft. In that case, the elite members of society petitioned the
Prophet to avoid punishment and let the convict go. The petitioners were
members of Muhammad's family, as was the woman convicted of stealing. The
Prophet responded that even his hands were tied from pardoning the woman,
where the evidence proved her guilt. "Would you intervene on a matter
involving God's laws?" he asked. "I swear by God that even if Fatima [my own
daughter] had stolen, I would cut off her hand!", 59 Here, the Prophet was
54. For an outline of his textualist legal theory, as presented in his son's law manual, see
Devin Stewart, Muhammad b. Ddwtvd al-Z5hir's Manual of Jurisprudence,al-Wu~til i15 Ma'rifat al-
USal, in STUDIES IN ISLAMIC LEGAL THEORY 99, 139-54 (Bernard G. Weiss ed., 2002), which notes that
Dawtd made inferences from textual indicators, but rejected the use of analogy, equity, and other forms
of interpretation used by other Muslim jurists of his time. For a fuller account of his jurisprudence, see
'ARIF KHALIL MUI.AMMAD ABO 'ID, IMAM DAWOD AL-ZAHIRI WA-ATHARUH F 'L-FIQH AL-ISLAMI (Dir
al-Arqam 1984). Note that this brand of Islamic textualism differs from "first-" and "second-generation"
American textualism-though the former provides an analytical hook by which to grasp analogous
Islamic legal debates in the Zdhiri school. See John F. Manning, Second-Generation Textualism, 98
CALIF. L. REV. 1287 (2010) (outlining the tenets of old and new, or second-generation, textualism);
John F. Manning, Textualism and Legislative Intent, 91 VA. L. REV. 419, 420-21 (2005) (same).
55. Stewart, supra note 54, at 157-58 (arguing that allowing "interpretation" would permit
each judge to rule according to "whim, [giving a ruling] opposite that determined by the fancy of [the
next judgel").
56. MUIIAMMAD AL-SHATTT, RISALA FT 'L-MADHHAB AL-IMAM DAWOD AL-ZAHIRI 25
(Matba'at Rawdtat al-Shfm 1912) (collecting legal opinions attributed to Dawid b. 'AR from various
works, including the strict enforcement of criminal punishment even in cases where a criminal
defendant who confessed to a crime subsequently retracted the confession).
57. See Stewart, supra note 54, at 156-57 (quoting Dwfid al-Zfhiri making a case against
interpretation by quoting Qur'5n 4:105, "We have revealed the Book unto you with the truth so that you
may judge between mankind by that which God shows you. Do not be a pleader for the treacherous,"
and commenting that God did not say, "by that which you think for yourself' or "by that to which your
preferences and perceptions lead you").
58. THE QUR'AN 24:2. For analysis in a medieval legal-exegetical treatise on the import of
Qur'Anic verses about issues of law, see 3 ABO BAKR IBN AL-'ARABI, AHKAM AL-QUR'AN 334-35 (Ddr
al-Fikr 1978).
59. The term for "God's laws" here is hudfdAlldh. The early, Qur'dnic meaning of the term
referred simply to "Islamic law." The term later came to mean "Islamic criminal law." For reports of
this case, see 8 ABO BAKR AL-BAYHAQI, SuNAN no. 17004 (Muhammad 'Abd al-Qfdir 'Ata' ed., 1994);
referring to the Qur'dnic punishment for theft. 60 Together, the Case of M 'iz
and the Case of the Makhzfimr Thief were both foundational cases that textualist
judges understood to mean that even the Prophet's authority was restricted to
text. Like him, all judges were bound to a version of divine legislative
supremacy that required them to follow the strict meaning of the texts.61
Two centuries later, as other jurists were systematizing doctrines in their
respective schools of law, Dawild's famous disciple Ibn Hazm further
elaborated the Islamic legal school of strict textualism. Ibn Hazm began with
the common point that the Qur'dn and Sunna were the only valid sources of
law. For him, these sources specified that judges were to enforce criminal
punishments whenever the evidence established that a crime had been
committed. Not even the Prophet could diverge from the strict dictates of the
text. 62 Thus, the Case of Md'iz shed no light on whether and how the judges
should handle doubt. Precisely because the meaning behind the Prophet's
rebuke of the townspeople for punishing Md'iz was unclear, his statement was
not normative. According to Ibn Hazm, to interpret the Prophet's statement as
doubt about the propriety of punishment, as most jurists did, was mere
speculation. For him, the case itself did not suggest any legal role for doubt at
all. The only relevant question for criminal liability was whether there was
proof in court that a defendant had violated a textual prohibition. 63 Neither
doubt nor any other contextual consideration was relevant for the strict
textualists, because none of the authentic prophetic texts had mentioned those
factors.64 In short, Ibn Hazm and his strict textualist followers saw themselves
as arguing for text over spirit.
b. The Case for Avoiding the Punishment: Pragmatic
Textualist Readings of the Case of M5'iz
Strikingly, most Muslim jurists used the Case of Md iz to conclude that

punishment should have been avoided. For them, several considerations should
have overridden the apparent textual commands supporting punishment, and
Md'iz should have thus been released. Accordingly, they read this case to
support a prospective rule of avoiding punishments in similar cases of doubt.
Moreover, they invoked the Case of Mj'iz to support the doubt canon that
called on judges to "avoid criminal punishments in cases of doubt." 65 In effect,
BUKHART, supra note 46, no. 6788; MUSLIM, supra note 46, no. 1688; 4 A..MAD B. SHU'AYB AL-
NASA'T, AL-SUNAN AL-KUBRA 330 (Iasan 'Abd aI-Mun'im al-Shalabi ed., 2001); and 4 ABO DAWOD
AL- SIJISTANI, SUNAN nos. 4373-74 (Muhammad 'Abd al-'Aziz al-Khalidi ed., 1996).
60. THE QUR'AN 5:38. For problems that Muslim jurists faced in interpreting the verse and
determining the associated punishment, see WEISS, supranote 6, at 104-09.
61. See 8 IBN HAZM, supra note 50, at 252.
62. 12 IBN HAZM, supra note 50, at 18, 22.
63. Id.
64. For Ibn H azm's rejection of the prophetic origin of the doubt canon, see id. at 47-63. For
discussion, see Rabb, Islamic Legal Maxims, supra note 17, at 111-13.
65. E.g., MARGHINANT, supra note 50; MAWARDT, supra note 50, at 206; QUDORI, supra note
50, at 5949-50. For further discussion, see Rabb, Islamic Rule of Lenity, supra note 17, at 1327-49.
they argued for spirit over text. 66

Against strict textualist accusations to the contrary, the jurists who read
the Case of M 'iz to support a rule for avoiding punishment asserted that they
were more faithful to the divine legislative intent. That is, the intent behind the
adultery rules was not strict obedience designed to resolutely elicit punishment,
but to elicit moral consciousness-which would be served by repentance
(signifying rehabilitation) as much as it would be by punishment. One example
is the prominent Hanafi jurist, Qudfri, a contemporary of Ibn Hazm. QudirT
argued that the Case of Md 'iz stood for the proposition that crimes could only
be established by confessions if the defendant freely confessed four separate
times that he or she had committed the crime. 67 Moreover, for QudfirT, the case
indicated that if a defendant retracted his or her confession even once before
the punishment was completed, the permission to enforce the punishment
would drop. The same would be true if one of four witnesses-a Qur'dnically
required burden to prove an adultery charge-perjured him or herself before
the punishment was carried out. Applying these rules to Md'iz, Qudiri argued
that Md'iz's attempted escape was tantamount to a retracted confession, and
perhaps to repentance as well.68
This reading, against the accusations of "spirit over text," explained why
the Prophet blamed the townspeople for not letting Md'iz go. This reading also
clarified that the proper judicial response to a criminal case where there was a
conviction accompanied by possibilities for repentance was to avoid
punishment. In defending this view, Qudiri and the majority of medieval
Muslim jurists in effect claimed, like Dawaid and Ibn Hazm, to be textualists
devoted to the principle of legislative supremacy. Further, most jurists asserted
that the strict textualists, not they, had impermissibly used the text to justify an
outcome that the divine lawgiver had not intended. The divine intent, they held,
was not punishment for punishment's sake, but punishment only when there
was no "doubt," on a contextual assessment, that punishment was in fact due.69
These jurists' arguments about how to determine the divine legislative
intent-if not through the strict textualist means-are drawn out more fully in
Parts II and III. For now, suffice it to say that the Case ofMj 'iz was about both
sets of jurists' antipathy toward doubt-that of the strict textualists and the
pragmatic textualists alike. Though their strategies for resolving doubt differed,
both framed their arguments about the fate of Md'iz as a search for legislative
intent.
66. This formulation recalls language from Church of Holy Trinity v. United States, 143 U.S.
457 (1892), which presented a conflict between a strictly textual reading and a purposive reading that
went beyond the text of an immigration and labor statute. See Holy Trinity, 143 U.S. at 459 ("[W]c
cannot think Congress intended to denounce with penalties a transaction like that in the present case. It
is a familiar rule, that a thing may be within the letter of the statute and yet not within the statute,
because not within its spirit, nor within the intention of its makers."). For the significance of the spirit-
versus-text controversy and the analogous importance of Holy Trinity, see infra note 70.
67. QUDORI, supra note 50, at 5949-50.
68. Id.
69. The looming question would be how to define doubt, which was expansive, as described
further below in Subsection III.B.3.
3. The Significanceof Md'iz
The Case of Md 'iz became iconic in Islamic criminal law. 70 It is not that
the case by itself created the doctrine of doubt that lies at the heart of the story
about interpretation in Islamic law. To be sure, it was a part of the series of
early landmark cases on which Muslim jurists drew to outline a doctrine of
doubt-common law fashion-over the course of several centuries, until they
textualized the doctrine in the tenth and eleventh centuries. But the Case of
M 'iz itself was illustrative rather than generative of that end.
So why did debates about Md'iz matter so much? They were significant
because they went to issues of constitutional structure and legitimacy. I contend
that the Case of Ma 'iz demonstrates how jurists used interpretation to define
criminal law. They also used it to define institutions of law and governance,
organically and bottom-up.
Here is how. Recall that, in historical Islamic contexts, an informal
constitutional arrangement unfolded featuring a type of separation of powers:
the class of jurists assumed the power to interpret Islamic law, the caliphs
asserted the executive power to enforce it, and judges shuttled between the
two. 71 Neither judges nor jurists possessed the prerogative to make law, which
Islamic legal doctrine yields exclusively to a divine Legislator (who legislated
once in the seventh century). Thus, Islamic textualism debates are framed in
terms of legislative supremacy-a super-strong version of divine legislative
supremacy that requires a limited role of judges in order not to derogate from
divine directives. Islamic textualism also comes with a heavy intentionalist
imperative, prompting jurists to focus on legal texts to discover divine
legislative intent. These observations-read alongside the Case of M(Viz-
suggest that, whenever Muslim jurists interpreted texts alongside cases like that
of Md'iz, they defined their own role against that of other political actors.
70. For a sense of the importance of this case to Islamic interpretation debates, consider the
discussion surrounding Holy Trinity, and the conflict between text and extratextual context, in American
statutory interpretation debates. See, e.g., Frederick Schauer, ConstitutionalInvocations, 65 FORDHAM
L. REV. 1295, 1307 (1997) ("Church of the Holy Trinity v. United States is not only a case, but is the
marker for an entire legal tradition, a tradition ... emphasizing ... that there is far more to law than the
plain meaning of authoritative legal texts .... ") (citations omitted); cf ESKRIDGE, supra note 53, at 209
(labeling the case a "sensation" that inaugurated a trend of judicial use of legislative history); ADRIAN
VERMEULE, JUDGING UNDER UNCERTAINTY 88 (2006) (observing that most judicial and academic
commentaries on legislative history and statutory interpretation theory begin with Holy Trinity as "the
leading case in the legislative history debate," known for "endorsing countertextual interpretive
techniques"); Antonin Scalia, Common Law Courts in a Civil-Law System: The Role of United States
Federal Courts in Interpreting the Constitution and Laws, in ANTONIN SCALIA, A MATrER OF
INTERPRETATION 3, 18-23 (Amy Gutman ed., 1997) (describing Holy Trinity as the "prototypical case
involving the triumph of supposed 'legislative intent' . . . over the text of the law"); Adrian Vermeule,
Legislative History and the Limits of JudicialCompetence: The Untold Story of Holy Trinity Church, 50
STAN. L. REv. 1833, 1835 (1998) (observing that Holy Trinity "elevated legislative history to new
prominence by overturning the traditional rule that barred judicial recourse to internal legislative
history"). For discussion of the interpretive process in Holy Trinity as a constitutional amendment akin
to the textualization process I trace in describing the history of doubt in Islamic law, see John Manning,
The Eleventh Amendment and the Reading of Precise ConstitutionalTexts, 113 YALE L.J. 1663, 1686
(2004).
71. FELDMAN, supranote 28, at 2; see also supra note 42 and accompanying text.
Unlike the modem American context, where three branches are in dialogue
with each other in contests over institutional functions, medieval Islamic legal
contexts lacked an active legislature or other defined branches with which
jurists would communicate institutionally. 72 Instead, through textual inter-
pretation, jurists defined and adopted institutional roles separated not by
constitutional structure but by function. For example, in the exercise of their
juridical-cum-judicialfunction,they assumed an institutionalist stance that they
claimed was wholly subordinate to a divine legislator.73 Yet, their claims to
textualism cloaked other legislative functions that jurists also assumed, as
outlined in the next Part.
Moreover, Muslim jurists debated the fate of Md'iz against the backdrop
of divine legislative supremacy and with texts that were far from static or clear.
Questions of authenticity and meaning that arose from the dynamic nature of
the sources-that is, the texts in which the Case of Md 'iz and related early
cases were recorded-left ambiguities that pragmatic jurists took to demand
interpretation. Accordingly, debates between jurists about Md'iz played out in
the legal sources, including collections of prophetic reports about criminal law,
legal treatises, manuals for judging, advisory opinions about how to resolve
hard cases in criminal law, and collections of legal maxims about doubt, as
well as in historical sources, including judicial biographies and historical
chronicles that offer anecdotal evidence of criminal trials and punishment. 74 In
short, debates about Md'iz persisted in many of the extant sources for legal
theory and legal practice (which were general in nature and anecdotal at best)
from the ninth century through the present. For the jurists writing during the
early founding period, in the lead-up to the eleventh-century systematization of
Islamic law, there was only one legitimate way to resolve the debate about
Md'iz. They had to answer the broader question: what did God require? In this
way, despite (or perhaps because of) the fact that they could not simply ask the
lawgiver to update or clarify the law, what emerges from their treatment of the
case is that the jurists debating the fate of MR'iz framed their debates as a quest
for divine legislative intent.
All in all, debates about Md'iz and other interpretive controversies were
especially significant in Islamic law because Muslim jurists enjoyed an
outsized role in making authoritative pronouncements of law. Their
pronouncements distinguished the Islamically legitimate from the illegitimate,
and in the process, defined a delicate "constitutional" balance between key
institutions of Islamic law and governance.
72. On the dialogic nature of the communication between courts, Congress, and executive
agencies, see, for example, sources cited supra note 70.
73. For an elaboration of a similar idea in American statutory interpretation, see Jane S.
Schacter, Metademocracy: The Changing Structure of Legitimacy in Statutory Interpretation, 108
HARv. L. REV. 593, 594 (1995), noting that, when judges interpret text, they are implicitly defining their
own role "by defining the goal and methodology of the interpretive enterprise, and by taking an
institutional stance in relation to the legislature."
74. See generally RABB, supra note 25.
20151 "'ReasonableDoubt" in Islamic Law
B. The Role ofMuslim Jurists in Defining Structures of Islamic Law and

Governance
How did scholars-who had no official religious designation and no
military might to back them-gain control over definitions of Islamic law?
Moreover, how did they come to define the very structures of legitimate
Islamic law and governance in medieval contexts? How did their definitions
govern or extend from the Case of Md 'iz and other early judicial practices over
which these jurists commented and presided?
The process of jurists assuming interpretive authority over Islamic
criminal law-in tandem with the rest of Islamic law-was gradual. As other
historians have recognized, the expansion of the jurists' role came in response
to the new sociopolitical realities of a quickly expanding empire. 75 Less well
studied is how those processes played out. In this Article, I propose that the
expansion of their power unfolded 76
to track expanding institutional functions of
jurists, and did so in three stages.
The assertion of scholarly control over Islamic law and questions of
legitimacy began with the death of the Prophet, who had embodied both
religious and political leadership. Under him, religious law and governance
were questions of religious morality.77 When he died, the religious and political
functions split. The jurists took over questions of religion and law, and caliphs
took over matters of governance. 78 The general populace saw Muslim jurists as
the legitimate interpreters of Islamic law and understood the caliphal authority
to enforce the juristic interpretations as a power derivative of Islamic law.
Definitions of Islamic law and governance thus became matters of institutional
competence, each institution tasked with carrying out functions best suited to
the expertise or effective power of its members. 80 Criminal law, as one of the
few areas of public law, was an area of shared jurisdiction between jurists and
caliphs. In this area, Muslim jurists attempted to use their popularly recognized
authority to interpret Islamic law to outline the legitimate types of crime that
the community would recognize and the legitimate types of punishment that the
executive could enforce. 81 In this sense, although the caliphs often skirted
norms of legitimacy by punishing expansively in their jurisdiction of political
authority, the jurists continually attempted to use Islamic law as an institutional
75. As detailed below, this picture emerges from the amalgamation of a wide array of
prosopographical sources, historical chronicles, and reproductions of documentary evidence, such as
judicial appointment letters, caliphal edicts, and reports of prophetic and post-prophetic events. See infra
Subsections I.B.1-3.
76. See, e.g., infra notes 82-96 (religious morality), 97-114 (institutional competence), 115-
123 (institutional constraints).
77. See infra Subsection I.B. 1.
78. For an accessible treatment of this well-known scheme, see, for example, HALLAQ, supra
note 43, at 197-220, describing the relationship between political authorities, jurists, and judges in early
Islamic political theory (siydsa shar'iyya).
79. Id.
80. See infra Subsection I.B.2.
81. See infra Subsection I.B.3.
constraintagainst executive overreach. As I argue below, each of these stages

is crucial to understanding the jurists' role and significance in interpretive
debates over episodes like the Case of M6 'iz and related questions of Islamic
legitimacy.
1. Islamic Law as Religious Morality
The earliest Muslims conceived of Islamic law as a unified system of

religious morality over issues of both law and governance embodied in the
person of the Prophet. As noted above, they took Islamic law to consist of
direct commands from God that Muhammad had delivered in the text of the
Qur'dn and through the example of his divinely inspired life. 82 Accordingly,
the earliest Muslims looked to Muhammad for religious guidance on how to
conduct their affairs in all spheres of life, private and public. His guidance took
the form of moral directives, political leadership, and dispute resolution, as in
the Case of M 'iz. 83 As leader of the young community of Muslims,
Muhammad combined power over all spheres of law and governance during his
lifetime.84
With Muhammad's death in 632, the functions of legislation, governance,
and adjudication ceased to be combined in a single person. His death caused a
crisis of succession and leadership, which would radically change the political
organization of the young Muslim community. 85 A series of four caliphs took
over political leadership.86 However, these first four caliphs did not assert the
divine connection to God that the Prophet had claimed, and therefore could not
assert the exclusive prerogative to define Islamic law. 8 Rather, these caliphs
exercised political authority by virtue of most Muslims having agreed to
"elect" a single leader after the Prophet and therefore follow caliphal leadership
88
in military and economic affairs.
The extent of these caliphs' religious authority depended on their
82. See HUGH KENNEDY, THE PROPHET AND THE AGE OF THE CALIPHATES: THE ISLAMIC
NEAR EAST FROM THE SIXTH TO THE ELEVENTH CENTURY 45 (2d ed. 1986).
83. When Muhammad migrated from Mecca to Medina in 622, he became the head of a
fledgling Islamic state. The people of Medina had invited him to be their head and agreed to implement
his new religious law. This migration (hyra) was an all-important event in the history of Islamic law and
governance. Tellingly, the Islamic calendar does not begin when the Prophet first received revelation as
a member of a persecuted fringe group in Mecca. It begins After Hijra (AH)-that is, after his migration
to Medina some 300 miles north-precisely because he had become the head of state. In Mecca,
Muhammad had gained only charismatic religious authority over followers with whom his message of
morality and ethics resonated. But in Medina, he gained political and legal authority as well. For a
general overview of law and governance during Muhammad's lifetime, see HALLAQ, supra note 78, at
27-71; and KENNEDY, supranote 82, at 50-81.
84. For overviews of the legal structure during the Prophet's lifetime and immediately after,
see SUBHI MAHIMA$ANI, THE PHILOSOPHY OF JURISPRUDENCE INISLAM 15-16 (Farhat J. Ziadeh trans.,
1961) (1946); and Roy P. Mottahedeh, Introductionto SADR, supra note 12, at 4-14.
85. The events following the Prophet's death have been labeled the first "crisis of Muslim
history." MAHMOUD M. AYOUB, THE CRISIS OF MUSLIM HISTORY: RELIGION AND POLITICS IN EARLY
ISLAM 7 (2003).
86. See KENNEDY, supranote 82, at 50-81.
87. See WAEL B. HALLAQ, THE ORIGINS AND EVOLUTION OF ISLAMIC LAW 38-39 (2005).
88. See KENNEDY, supra note 82, at 50-81.
scholarly competence in articulating norms drawn from their association with

the Prophet himself. 89 To be sure, the caliphs issued religious and legal
directives in addition to appointing judges and leading the military. Yet, their
command over issues of religious law was constrained. 90 For example, one of
Muhammad's former companions successfully challenged the second caliph,
'Umar b. al-Khattab, in an arbitration proceeding over the destruction of
borrowed property. Even though the decision was against him, the caliph was
so impressed with the arbiter's reasoning and scholarly acumen that he
appointed him a judge. 91 This and other anecdotes suggest that early Muslims
understood the first four caliphs to be leading members of the learned circles of
the Prophet's former companions and family 92
members, without definitive
authority over matters of religion and law.
The assassination of the fourth caliph 'AlTi just 40 years after the
Prophet's death inaugurated a dramatic shift in leadership from religious to
monarchical rule. A man by the name of Mu'dwiya was the governor of
Syria-then, a garrison outpost-and distant cousin of the slain third caliph.
Seeking to avenge his cousin's death, Mu'dwiya rallied troops to march against
'Ali, whom he accused of taking insufficient measures to bring the assassin to
justice. Ultimately, Mu'dwiya succeeded in wresting control from 'Ali, who
was killed in battle by a defector from his troops. Mu'dwiya established the
first of a series of monarchical dynasties called caliphates that presided over the
Muslim world in some form or another until the end of the Ottoman Empire in
1924. 9"
The new caliphate was forced to acknowledge that it had little basis to
assert authority over definitions of law. Like the first four successors to the
Prophet, the heads of the new dynasties assumed the title of caliph, which
literally meant "deputy" or "representative." 94 The term had connoted a sense
of prophetic religious morality under the first four caliphs, who were
themselves learned scholars of the prophetic life. But as military-backed rulers,
89. See HALLAQ, supra note 87, at 34-39 (describing shifting bases of Islamic legal and
political authority following the Prophet's death in year 632).
90. See, e.g., 10 ABO AL-'ABBAS AL-QALQASHANDI, SUBH AL-A'SHA FT$NA'AT AL-INSHA'
21, 79, 359 (al-Mu'assasa al-Mi~riyya al-'Amma 1981) (providing evidence that the caliphs appointed
judges and instructed them with general policies on judging, such as the famous letter from the second
caliph, 'Umar b. Khatpb, to AbQ Mlsd al-Ash'ari); AL-SHARIF AL-RAI)T, NAHJ AL-BALAGHA letter 53,
at 426-45 ($ubhi al-$Rlih ed., 1967) (explicating judicial and governing policies sent from 'All to Malik
al-Ashtar upon his deployment to Egypt to take over as governor). 'Umar also instituted lasting changes
over his ten-year rule (634-644), significantly to public law. See MAHMAsANI, supra note 84, at 110-14
(listing his directives concerning alms-tax, divorce, slave law, theft crimes, sex crimes, and discretionary
punishments).
91. WAKI', AKHBAR AL-QUI)LT 357 (Sa'id Muhammad al-Lalb.h m ed., 'Alam al-Kutub 2001)
(recounting the judicial appointment story of Shurayh).
92. See HALLAQ, supra note 87, at 29-56. For a contrary view, see PATRICIA CRONE &
MARTIN HINDS, GOD'S CALIPH: RELIGIOUS AUTHORITY IN THE FIRST CENTURIES OF ISLAM 2-3 (1986),
arguing that caliphs after Ab Bakr saw themselves as not just political authorities but also religious
authorities.
93. For a survey of Muslim dynasties, including a list of Umayyad and 'Abbasid rulers, see
CLIFFORD E. BOSWORTH, THE NEW ISLAMIC DYNASTIES (2d ed. 2004).
94. 2 IBN MANZOR, LISAN AL-'ARAB 299-303 (Dar $5dir 1997).
the new caliphs did not claim to be scholars of the Prophet's life, nor therefore
could they assume authority over definitions of Islamic law. That is, caliphs did
not-and, no doubt, could not-assume sweeping religious authority over even
criminal law. That much was evident from their episodic but always failed
attempts to control the jurists. 95 Stripped of any scholarly competence under
Mu'dwiya and his successors, the term "caliph" largely lost its religious
meaning. The historical sources portray the early caliphate as having
transformed from an office of moral leadership concerned with establishing' 96a
just social order to one of tribal loyalties based on "might makes right."
Political leadership had shifted to the new caliphs, who no longer shared any
part of the religious morality of the Prophet, aside from the symbolic title of
"successor." Instead, as the next section will elaborate, religious leadership-
and the moral ability to define Islamic law-had shifted almost entirely to the
jurists.
2. Islamic Law as InstitutionalCompetence
Within half a century of Islam's advent, leadership had gone from a

charismatic prophet to a militaristic caliph, with stark consequences for
questions of legitimacy in Islamic law and governance. Under the new
caliphate, the scholars insisted that their continuous study of the Prophet's life
provided them with a special institutional competence to discern the meaning
of Islamic law. That is, these jurists increasingly asserted the power to "say
what the law is." 97 They were willing to cede political leadership to the caliphs,
but for religious matters, they conceived of themselves as the "heirs to the
prophet." 98 In the eyes of the populace, the scholars were correct. 99 The
scholars' stance resonated with the political and spiritual sensibilities of enough
factions that the jurists were ultimately successful in their claim to the
95. See MUHIAMMAD QASIM ZAMAN, RELIGION AND POLITICS UNDER THE EARLY
'ABBAsIDs: THE EMERGENCE OF THE PROTO-SUNNI ELITE 82-85 (1997) (detailing the failed proposal of
the vizier Ibn al-Muqaffa' to place law under the authority of the caliph by attempting to capture the
jurists).
96. AYOUB, supra note 85, at 54-57 (arguing that this transformation to tribal loyalty began
with 'Uthman); cf LOUISE MARLOW, HIERARCHY AND EGALITARIANISM INISLAMIC THOUGHT 14-16
(1997) (detailing a social egalitarian bent during the early Islamic period that became explicitly
hierarchal during 'Uthm~n's time and that Mu'awiya's assumption of leadership based on bloodlines
ended "the opportunity for social equalising" on the basis of piety criteria).
97. See CHIBLI MALLAT, THE RENEWAL OF ISLAMIC LAW: MUHAMMAD BAQER AS-SADR,
NAJAF AND THE SHI'T INTERNATIONAL 79 (1993) ("[Tlhe quintessential constitutional question is about
who ultimately holds the power to 'say what the law is.' In view of the centrality of the shartfa in the
definition of an Islamic state, this issue represents the essential problem of contemporary Islamic law.").
Here, Mallat was referencing Chief Justice John Marshall's famous exposition of the judicial power
under the U.S. Constitution in Marbury v. Madison, 5 U.S. (1 Cranch) 137, 177 (1803) ("It is
emphatically the province and duty of the judicial department to say what the law is."). For a long-view
survey of the major themes that have arisen in American constitutional interpretation based on Justice
Marshall's famous formulation, see CHARLES FRIED, SAYING WHAT THE LAW IS: THE CONSTITUTION IN
THE SUPREME COURT (2004).
98. For discussions of this hadlith in the context of Sunni law, see FELDMAN, supra note 28, at
26. For ShVT law, see LIYAKAT N. TAKIM, THE HEIRS OF THE PROPHET: CHARISMA AND RELIGIOUS
AUTHORITY IN Sm'ITE ISLAM (2006).
99. For the early development of this shift, see generally ZAMAN, supra note 95.
institutional competence to define Islamic law.°00 In this way, the religious

morality once articulated by the Prophet had turned to the 101institutional
competence to determine that morality in legal terms by the jurists
The caliphs and jurists depended on one another. The jurists, armed only
with popular perceptions of religious legitimacy but no military weapons,
relied on caliphal enforcements of their proclamations of law. Likewise, the
only way for caliphs to recover enough religious legitimacy to ensure political
stability was to establish a working relationship with the jurists. 10 2 The two
camps settled on an arrangement whereby the caliphs ceded most interpretive
authority to the scholars, while retaining control over judicial appointments and
law enforcement. 103 In turn, these scholars developed the law largely
independently of state involvement and control. 104 Early judges (who were not 05
themselves scholarly experts) were to consult the learned circles of jurists.
This arrangement made Islamic law both "jurists' law" and state law. 1' 6 The
judiciary provided "a mechanism for enforcing [Islamic law] by the state....
Judicial authority
10 7
came from the caliph, but the law to be applied came from
the scholars."'
The "founding period" (seventh to eleventh centuries) saw important
developments in Islamic law, as the jurists defined Islamic law in progressively
more corporate and systematic terms. In corporate terms, the informal scholarly
circles of the earliest period coalesced around key figures from the eighth and
ninth centuries to form distinct schools of legal interpretation. 108 There were
100. This led to an uneasy balance with the government over law as well as more intense
discussions in legal-religious circles about what the law was. See Mottahedeh, supranote 84, at 6-8.
101. On the moral and "epistemological" basis for juristic authority and legitimacy, see
HALLAQ, supra note 40; cf ARON ZYSOW, THE ECONOMY OF CERTAINTY: AN INTRODUCTION TO THE
TYPOLOGY OF ISLAMIC LEGAL THEORY 1 (2013) ("From a very early period ...Muslims came to treat
the question of legitimacy along explicitly epistemological lines.").
102. See HALLAQ, supranote 40, at 1-24.
103. HALLAQ, supra note 43, at 197-221 (describing this tripartite scheme as a "Circle of
Justice" in which jurists, caliphs, and judges formed co-equal arcs necessary to complete the circle).
104. For accounts of the early schools, beginning in the Umayyad reign and lasting until the
formalization of legal doctrine under the early 'Abbdsids, see generally HARALD MOTZKI, THE ORIGINS
OF ISLAMIC JURISPRUDENCE: MECCAN FIQH BEFORE THE CLASSICAL SCHOOLS (Marion IH. Katz trans.,
2002); and NURrr TSAFRIR, THE HISTORY OF AN ISLAMIC SCHOOL OF LAW: THE EARLY SPREAD OF
HANAFISM (2004).
105. HALLAQ, supra note 87, at 52 (noting that "the sources are frequently unclear as to
whether or not these specialists were always physically present in the court, but [that] we know from the
beginning of the second century (c. 720 AD) judges were encouraged to seek the counsel of these
learned men and that, by the 120s/740s, they often did").
106. FELDMAN, supra note 28, at 27 (citing JOSEPH SCHACHT, INTRODUCTION TO ISLAMIC
LAW 5 (1964) ("[I]t was created and developed by private specialists; legal science and not the state
plays the part of a legislator, and scholarly handbooks have the force of law.")) (on the application of the
term "jurists' law" to Islamic law).
107. On parallel developments in the Shi' context, see, for example, HOSSEIN MODARRESSI,
CRISIS AND CONSOLIDATION IN THE FORMATIVE PERIOD OF SH'rrE ISLAM 29 (1993), noting that the
early mainstream Imami Shi'i community believed the authority of the Imams to have been founded on
them being learned and pious scholars; WILFERD MADELUNG, RELIGIOUS SCHOOLS AND SECTS IN
MEDIEVAL ISLAM (1985); WILFERD MADELUNG, THE SUCCESSION TO MUIAMMAD: A STUDY OF THE
EARLY CALIPHATE (1997); and compare Etan Kohlberg, al-UsuIA1-Arba 'umi 'a, 10 JERUSALEM STUDS.
IN ARABIC AND ISLAM 128-66 (1987).
108. For a comparison of these schools to corporations, see SHERMAN JACKSON, ISLAMIC LAW
hundreds of these schools in the first half of the founding period.10 9 By its end,
the multitude of schools had boiled down to four main Sunni schools, a
minority Sunni strict textualist school and a principle Shi'l school, also with a
textualist strand." 0 The head SunnY scholars, now regarded as the "founders" of
the surviving schools, all lived during the first half of the founding period:"'
Abu Hanifa of the Hanaff school, Mdlik of the Mdliki school, Shrfi'T of the
Shafi'T school, and Ibn Hanbal of the HanbalT school. The Zahiri strict
textualist school-though it eventually died out-was started by Dfwfld the
Textualist. The main ShI'T school traced back to this period as well: Ja'far al-
.Sdiq,the Sixth Imdm-with whom most of the proto-Sunni scholars happened
to have studied-articulated
12
the core doctrines and principles of interpretation
for Shi'i law."
Alongside the founders of the legal schools, another group of scholars
devoted themselves to collecting reports of the prophet's words and actions,
which formed the basis for the Sunna. These scholars sifted through tens of
thousands of prophetic reports in an attempt to distinguish false from authentic,
and therefore normative, statements of law." 3 Together, books outlining each
school's doctrines and the prophetic reports containing the Sunna would
provide the raw materials for more systematic treatises on Islamic law.
By the eleventh century, jurists had begun organizing the ad hoc rules and
multiple texts from the Prophet and early scholars into sophisticated treatises of
law. These jurists drew on this amorphous body of school doctrines and
prophetic reports to produce Islam's definitive works of legal doctrine (fiqh),
l 4
legal theory (us.il al-fiqh), and political theory (siydsa shar'iyya)." Through
this process, each legal school distinguished its doctrines and methodological
approaches through debates about the meaning of a common set of "texts": the
Qur'an and the Sunna. It was these treatises that laid down the definitive
doctrinal and methodological rules of Islamic law, including criminal law and
doubt.
AND THE STATE: THE CONSTITUTIONAL JURISPRUDENCE OF SHIHAB AL-DIN AL-QARAFT (1996).
109. See GEORGE MAKDISI, THE RISE OF COLLEGES: INSTITUTIONS OF LEARNING IN ISLAM
AND THE WEST 2 (1981) (estimating up to five hundred such proto-Sunni schools, most of which died
out by the end of the ninth century).
110. See M.AiMAsANT, supranote 84, at 63-66; MODARRESSI, supra note 107, at 52-55.
111. See, e.g., MAIMASANT, supranote 84, at 17-19 (calling this period the Golden Age, which
accompanied the major translation movement from Greek into Arabic, the flowering of literature, and
the patronage of arts and sciences).
112. See Mottahedeh, supra note 84, at 8-9.
113. For discussions of the collection and authentication process of prophetic reports (hadTth)
in the Sunni context, see JONATHAN BROWN, THE CANONIZATION OF AL-BUKHRT AND MUSLIM: THE
FORMATION AND FUNCTION OF THE SUNNI HADITH CANON 9 (2007), listing the six canonical Sunni
compilations of .adth-reports of the Prophet's words and practices. In the SI'I context, see
MODARRESSI, supra note 13, at 4-5, listing the four canonical ShV'1 compilations of reports of the
Prophet and Imdns' words and practices.
114. For a history of the legal theoretical works in the Sunni context, see WAEL B. HALLAQ, A
HISTORY OF ISLAMIC LEGAL THEORIES: AN INTRODUCTION TO SUNNI U$OL AL-FIQH 33 (1997). For the
Shi'! context, see MODARRESSI, supra note 13, at 7.
3. Islamic Law as InstitutionalConstraint
The division between political and religious leadership shaped a

constitutional structure for the premodern Muslim world necessary for stability
and legitimacy, but was threatened by political developments in the eleventh
century. From the rise of Islam until the rise of independent Muslim-majority
nation states after World War I, this common constitutional structure provided
that a "Muslim ruler governed according to God's law, expressed through
principles and rules of the sharT'a that were expounded by scholars.' 15 In the
eleventh century, as local rulers wrested power from the caliph, they threatened
to disturb the delicate balance between political and religious rule. These
developments made it important for jurists to reiterate their vision of the proper
institutional roles of jurists, caliphs, and judges to ensure legitimate and stable
rule. In short, these developments prompted leading Muslim jurists of that time
to use the Islamic law of doubt to more concretely articulate rules of
constitutional constraint. That is, in significant part through claims of doubt, I
contend that these jurists sought to place constraints on executive power.
One of the leading scholars of that time was the prominent figure Abi al-
Hasan al-Mdwardl, who undertook the task of restating the proper relationship
between spheres of politics and law. He was both a distinguished jurist of the
Shafi'i school and a high-ranking government official working principally in
Baghdad, which was the seat of the Muslim empire. He served as diplomat for
two 'Abb~sid caliphs just as the dynasty began to break up into smaller
principalities in the eleventh century. 116 In fact, one of his tasks was to help the
caliphs negotiate new power relationships with regional leaders to whom the
11 7
caliph was forced to yield significant power.
Mawardi wore two hats: that of the jurist articulating the domain of law,
and that of the government official delineating the domain of politics. The
written result was his famous work of Islamic political theory, al-Ahkim al-
sultaniyya (The Ordinances of Government). 118 In it, he drew on the early
history of contested political and legal authority in an attempt to memorialize
the constitutional scheme of legitimate power distribution so important to
helping preserve the integrity of the empire. Through it, Mawardi aimed to
articulate an Islamic legal theory that gave a wide scope of interpretive power
to jurists and placed constitutional constraints on the caliph.
M~wardT's starting point was the derivative authority of the caliph to
enforce Islamic law according to its juristic-scholarly articulation. The jurists
had won the argument of who possessed the religious morality and institutional
115. FELDMAN, supra note 28, at 2.

116. For a review of his life and works, see H.A.R. GIBB, Some Considerationson the SunnT
Theory of the Caliphate, in STUDIES ON THE CIVILIZATION OF ISLAM 141-65 (Stanford J. Shaw &
William R. Polk eds., 1962). For a comprehensive study of his work, see Henri Laoust, La Pens(e et
t'actionpolitiquesd'al-Mdwardl, 36 REVUE DES ETUDES ISLAMIQUES 11 (1968) (Fr.).
117. See GIBB, supra note 116.
118. ABC AL-HASAN AL-MAWARDI, AL-AIIKAM AL-SULTANIYYA WA' L-WILAYAT AL-DINIYYA
250 (Muhammad Fahmil al-Sirjni ed., 1978).
competence to interpret Islamic law. Accordingly, the earlier constitutional

scheme continued to hold sway: the caliph's rule was legitimate so long as he
agreed to uphold shar a-which included the Qur'dnic directive to "command
right and forbid wrong," as defined by the jurists. 1 9 By Mawardi's time, it was
a recognized point of consensus that only jurists trained in the interpretive
methods of 20 one of the recognized schools of law were qualified to define
Islamic law.1
Within that framework, Mawardi was at pains to more fully detail the
spheres of political and religious-legal authority over public law, including
criminal law. He acknowledged that criminal law was a part of public law,
necessary to ensure public safety. Public law ordinarily fell in the caliph's
jurisdiction. But Mawardi defined the meaning of public law narrowly,
portraying it as a short list of duties required of the caliph to justify his rule.
Instead of yielding all authority over public law to the caliph, MawardT insisted
that this was an area of shared jurisdiction. Jurists still maintained the
prerogative to12articulate the criminal law rules, and the caliphs were still to
enforce them. 1
MawardT's point to the power-seeking local rulers called the Seljtqs was
this: a separation of powers had obtained between the executive authorities and
the jurists. Maintaining that separation was crucial for the survival of the
premodem Muslim state itself.123
What I have elucidated so far is a general theory of Islam's constitutional

structure of separated powers, defined from the bottom up, in ways quite
different from the constitutionally defined branches of American law and
government. The Case of Ma'iz reflects the phenomenon of bottom-up
definitions of institutional relationships through interpretations of criminal law.
Included in the collections of foundational texts and often cited in the juristic
assertions of interpretive authority, this case illustrates the process and
significance of legal interpretation at the intersection of law and politics, which
historically were often linked through criminal law.
As the foregoing discussion should make clear, the debate over the Case
of Md 'iz encompassed an area of law that had a public face of constitutional
119. See generally MICHAEL A. COOK, COMMANDING RIGHT AND FORBIDDING WRONG IN
ISLAMIC THOUGHT (2001) (detailing the understandings and applications of this directive in Islamic
history).
120. See Intisar A. Rabb, Islamic Legal Minimalism, in STUDIES IN ISLAMIC LAW AND
TRADITION 256-74 (Michael Cook et al. eds., 2012) (noting that mainstream Sunni communities
required trained jurists to interpret Islam's foundational legal texts).
121. See MAWARDI, supra note 118, at 250 (including enforcement of hudrd laws in a short
list often public duties that the executive authority of the state must fulfill).
122. See FELDMAN, supra note 28, at 44-48 (describing the jurists' relationship with the
executive as the classical Islamic separation of powers scheme).
123. For an analogous meditation by a Seljdk scholar and vizier who, like Mawardl, wrote as
an officer of the state as well as a prominent Shfi'l jurist, see generally IMAM AL-HARAMAYN AL-
JUWAYNT, GHIYATH AL-UMAM FTILTIYATH AL-ZULAM (Khalil al-Man tir ed., Dfr al-Kutub al-'Itmiyya
1997). For an analysis, see Rabb, supranote 120.
proportions. Criminal law questions were "public" in that they required input
from jurists to articulate the law, political rulers to enforce the law, and judges
to mediate between the two. 124 They were "constitutional" in that jurists used
criminal law, in part, to constitute their own power and to negotiate relations
between these major institutions of Islamic law and governance. In other
words, during Islam's founding period, Muslim jurists had quickly gained vast
power to define Islamic law by successfully asserting a superior institutional
competence to articulate religious morality. They used this power to impose
constitutional constraints on the legitimate exercise of caliphal-executive
power.
As other historians and scholars of Islamic law have acknowledged, these
stages demonstrate the centrality of Muslim jurists and their interpretive
debates to questions of Islamic legitimacy in institutions of law and
governance. 5 Some historians have further clarified that criminal law was the
locus of considerable contestation among jurists and political authorities. 126
What other scholars in the field have not yet grasped is how jurists asserted that
power to define institutional relationships specifically through criminal law,
and why. The following discussion seeks to address both how and why by
examining the means and motives by which Muslim jurists used specific
interpretive maneuvers to expand their power over criminal law and limit that
of the executive.
II. THE CONVERSION OF "REASONABLE DOUBT" FROM JUDICIAL PRACTICE

TO LEGAL TEXT
In arguing that Muslim jurists used criminal law to expand their own
power and limit that of the caliph, I return to the central idea that the jurists'
power was itself constrained to certain interpretive ideals. From their internal
debates, it is clear that Muslim jurists believed their definitions of law to be
legitimate only if they could plausibly claim adherence to foundational texts
and to the divine legislative supremacy ideal. 127 The illustrative example is the
Case of Md'iz. All Muslim jurists believed that the option of avoiding
punishment was legitimate, in the Case of M 'iz and in subsequent cases, only
if that outcome represented divine legislative intent. But identifying avoidance
of punishment as the divine intent was not easy to accomplish through debating
the case alone. Islamic legal theory maintained that the divine legislative intent
was best represented through foundational texts, but the case itself contained a
foundational ambiguity as to what had happened. Jurists vehemently debated
whether the Prophet's pronouncement of guilt but his preference to avoid
124. Criminal law was unlike Islamic ritual law or family law, which could often be settled
privately. For overviews of ritual law and family law, see HALLAQ, supra note 78, at 271-95.
125. E.g., HALLAQ, supra note 40, at 166-83 (recognizing in general terms that jurists used the
power of their perceived legitimacy to help define institutional roles).
126. For an excellent study of the intellectual currents and enforcement mechanisms of
criminal law among Seljfq jurists and rulers, see CHRISTIAN LANGE, JUSTICE, PUNISHMENT AND THE
MEDIEVAL MUSLIM IMAGINATION 39 (2008).
127. See supra note 49 and accompanying text.
punishment set a precedent for punishment avoidance, especially given other

foundational texts that seemed to require enforcing punishments. 128 If
textualism and divine legislative supremacy were so central to the jurists'
claims of legitimacy, how did' they introduce doctrines of doubt that generally
required judges to avoid punishment in cases like that of Md'iz?
In a move astonishing for both its boldness and its near invisibility, the
jurists transformed the doubt doctrine from reports of early practices into a
legal text. This move was astonishing because it would seem wholly opposed
to the values of Islamic textualism and divine legislative supremacy. It was also
astonishing because the transformation of the canon from a judicial practice to
a legal text was so effective that few later jurists were even 29
aware of the
dubious nature of the prophetic pedigree for the doubt canon.1
Instead, most Muslim jurists largely adopted the doubt canon as a core
prophetic principle of Islamic law, which they used to resolve debates about
Md'iz and other criminal defendants. This resolution unfolded in three stages.
First, these jurists read the Case of Md'iz to introduce doctrines of doubt as a
matter of early judicial practice, e.g., "avoid criminal sanctions in cases of
doubt": the canonization of doubt. 130 Second, these Muslim jurists actually
converted this canon into a foundational text, which, as a prophetic report,
carried the full force of divine legislation: the textualization of doubt. Third,
they then managed to establish the doubt canon as the central principle of
interpretation to govern questions of doubt that arose in Islamic criminal law:
the generalization of doubt. I address the first two processes below, and the
third in Part III.
A. The Canonizationof Doubt Through Early JudicialPractice
The process of articulating doctrines of doubt first came through juristic

readings of early judicial practices. The authors of the eleventh-century legal
treatises systematically emphasized early precedents in favor of avoiding
punishment. 31 For them, the point was to demonstrate that the facts were not
always what they seemed, which was another way of saying that criminal cases
frequently produced factual doubts. These doubts were enough to require
heightened procedures before warranting convictions and punishment. From
reviewing dozens of early cases, jurists concluded that extensive doubts
permeated criminal cases. Additional procedures could alleviate those doubts
and, when applying the rule of "no punishment" in cases of doubt, these added
procedures served as tools for judges to avoid punishments in many cases. In
128. Id.
129. Only the traditionalist, strict-textualist jurists-the .AhirTs and some Hanbalis-were
consistently attuned to the non-prophetic pedigree of the doubt canon. This realization caused many of
them to reject the doubt canon both as a hadith (in prophetic attribution) and as a substantive canon (in
legal application). See IBN HAZM, supra note 50. For further discussion, see Rabb, Islamic Legal
Maxims, supra note 17, at 106-13 (detailing Ibn Hazm's objections to the doubt canon, which turned out
to be largely ineffective).
130. See Rabb, Islamic Rule ofLenity, supra note 17, at 1327.
131. See infra Subsection l11.B.3.
the jurists' estimation, the trend of judicial practices from the founding period
illustrated how often the founding figures avoided punishment. This frequent
avoidance helped them determine how Md'iz should have fared retrospectively
and how criminal defendants
32
should be treated prospectively. Consider the
following three cases.'
1. Case of the Mysterious Pregnancy
One report hails from eleventh-century Yemen, a generation after M'iz.

The local judge Abfi Mdisa al-Ash'ari was asked to preside over a criminal case
where a woman, who was either married or previously married, was accused of
committing adultery. 133 Recall that zinj int 34
this context was a capital crime, as
early Muslims had come to understand it.
The woman's situation looked suspicious. She was pregnant and seems to
have been unmarried-likely having been previously married-but the
pregnancy was not from her (former) husband. The woman swore an oath that
she had not committed a sex crime. Instead, she claimed that she woke up one
night to find a man on top of her, whom she did not36know, and who fled soon
after.1 35 Essentially, she claimed to have been raped.'
Judge Abii Miisd was perplexed. If the woman had admitted to adultery,
the textual rule might have guided him to convict. But she had not, leaving him
as nonplussed as she presented herself to be. He needed to consult a more
informed and authoritative jurist. Accordingly, the judge wrote to the caliph-
cum-jurist 'Umar in Medina, asking how he should proceed. In response,
'Umar summoned Abii Mflsa along with the accused woman and a few
witnesses from her town. The woman recounted the facts, and repeated her
denial. As the Prophet had with Md'iz, 'Umar asked the witnesses about her
character and mental state. They vouched for her and praised her generously.
The facts did not decisively resolve the case.
The situation was as follows: The woman had neither admitted to being
guilty nor was she known to be of bad moral character. For Abl Mis and
'Umar both, it was clear that some sex crime had taken place. But it was not
clear that the woman was culpable. To the contrary, the woman may not have
been a perpetrator of adultery, but rather a victim of rape. 137 'Umar remarked
132. For a collection of twenty-five early criminal cases demonstrating trends of alternate
avoidance and enforcement of Islamic criminal punishment, see RABB, supranote 25, app. A.
133. See IBN ABI SHAYBA, supra note 46, no. 28,970; WAKI', supra note 91, at 71-72.
134. See supra note 49.
135. It is unclear from the facts recounted in the report whether she was married and who
brought the accusation, the sources mention that she was a thayyib (non-virgin), which placed her in a
category of female offender for which jurists of the eleventh century understood the case to be a sex
crime that would require stoning. Her husband (if indeed she had been married), or another family
member, would have made the accusation. See IBN ABI SHAYBA, supra note 46, no. 28970.
136. A common strategy for Muslim women of this period to contest adultery accusations in
other contexts appears to have been to claim that they were raped. See, e.g., Delfina Serrano, Twelve
Court Cases on the Application of Penal Law Under the Almoravids, in DISPENSING JUSTICE IN ISLAM
473, 475,491 (citing 11 IBN HAZM, supra note 50, at 291-93).
137. In another version of the story, 'Umar says to the woman, "perhaps you have been raped."
that he feared God's wrath "if [he] sentenced this woman to death." Instead of
punishing her, he fed her, clothed her, and instructed her people to treat her
well. 138
Later Muslim jurists took this case to be one of legal doubt, for which
heightened procedures were required, which in turn helped determine whether
the law against fornication and adultery covered the acts of someone who had
not formed the criminal intent to violate the law. 'Umar's questioning about her
mental state and character paralleled the Prophet's questioning of Md'iz. In this
case, was the pregnancy, without a confession or witnesses, sufficient to
convict of adultery? The answer was no. The doubt about the law's application
to the woman who claimed to have been raped, jurists concluded, explained
why 'Umar avoided punishment, using procedure to lead to that result.
2. Case of the Falsely Accused Butcher
Similar evidentiary doubts emerged in reports about an early murder in

Medina. The Case of the Falsely Accused Butcher presented the question of
whether the Qur'dnic rule of retaliation, which permitted the death penalty for
anyone convicted of murder, 39
applied to a man who claimed after conviction
that the evidence was false.1
During the reign of the Muslim community's fourth caliph 'All, a type of
early police force in the small Arabian town of Medina was out patrolling.
Members of this patrol came across a man in the town ruins holding a blood-
stained knife and standing over the corpse of a man who had just been stabbed
to death. The patrol arrested the man with the knife. Upon arrest, he
immediately confessed: "I killed him."
The suspect was brought before 'All, the beloved cousin and son-in-law
of the Prophet Muhammad (who had died less than three decades before in
632). 'AlI was the fourth caliph according to the Sunnl account of successors to
the Prophet and the first Imam in the competing Sh'l account, and a revered
founding figure in both communities. He presided over criminal trials in his
capacity as leader of the young Muslim community from 656 to 661, as had the
Prophet before him. Upon hearing the defendant's story, 'All reportedly
sentenced him to death, in accordance with the Islamic law of retaliation for
homicide and personal injury: a life for a life.
Before the sentence was carried out, another man rushed forward, telling
the executioners not to be so hasty. "Do not kill him. I did it," he announced.
'Ali turned to the condemned man, incredulously. "What made you confess to a
See 12 SHWtA"B AL-DIN AL-QARA-F, AL-DHAKHIRA FTFURO' AL-MALIKIYYA 60 (Dir al-Gharb al-Islami
1994). For a similar version in an earlier source, see 7 'ABD AL-SALAM B. SA'ID SAItNfON, AL-
MUDAWWANA AL-KUBRA 2444 (D.r $dir, n.d.) (equating sex with a sleeping woman to rape).
138. IBN ABT SHAYBA, supra note 46, no. 28,970. For discussion of the jurists' antipathy
toward capital punishment read into scenarios like this one, see infra Section 1I.B.
139. For the Qur'5nic rle of retaliation, see THE QuR'AN 2:178-79; id. at 4:92; and id at 5:45.
For a discussion of the Islamic "reforms" to the ancient Near Eastern practice, see PETERS, supra note 5,
at 40.
murder that you did not commit?!" he asked. The man explained that he
thought 'All would never take his word over that of the patrolmen who had
witnessed a crime scene wherein all signs had pointed to him as the perpetrator.
In reality, the man explained, he was a butcher who had just finished
slaughtering a cow. Immediately after the slaughter, he needed to relieve
himself, so he entered the area of the ruins, bloody knife still in hand. Upon
return, he came across the dead man and stood over him in concern. It was then
that the patrol encountered him. Figuring that he could not plausibly deny
having committed the crime, he confessed to the "obvious" and decided to
leave the matter in God's hands.
The second man offered a corroborating story. He explained that he was
the one who had murdered a man for his money and then had fled upon hearing
sounds of the patrol approaching. On his way out, he passed the butcher
entering the area and then watched the events unfold as the butcher had
described them. Once the butcher was condemned to death, however, the
second man felt compelled
1 40
to step forward. He did not want the blood of two
men on his hands.
This case presented yet a new set of doubts. Whereas the previous case
addressed legal doubt arising from questions of criminal intent and thus
criminal culpability, this case highlighted problems of factual doubt. That
someone was criminally culpable was clear. There was a dead man in the town
ruins who was brutally stabbed to death. But the conflicts between
circumstantial evidence and competing confessions created doubt about who
was culpable. There was a problem in the evidence, which seemed impossible
to resolve. The lack of certain proof against either defendant may then have
caused 'Alli to avoid punishment.
Later jurists read this case to support the rule that judges should avoid
enforcing the death penalty even when the ordinarily acceptable forms of
evidence seemed dispositive. 141 Typically, murder convictions required either a
confession or two eyewitnesses, preferably corroborated by knowledge of the
witnesses' credibility or by circumstances pointing to the reliability of the
testimony. Here, there was the perfect trifecta of evidence: a confession,
testimony of multiple eyewitnesses, and corroboration by strong circumstantial
evidence. Yet, even then, there was doubt. There was a plausible alternative
140. IBR,AM B. HASHIM AL-QUMMI, QAI),kYA AMIR AL-MU'MININ 'ALT B. ABI TALIB 88-89,
238 (Fris Hassin Karim ed., Mu'assasat Amir al-Mu'minln 2003). Both Sunni and Shil' scholars have
cited this episode as an example of clever doubt jurisprudence. For Sunni treatments, see, for example,
IBN QAYYIM AL-JAWZIYYA, supra note 50, at 82-84 (quoting Qaddya 'AlTand Aj,'ib (referring to the
ninth-century works by Ibrdhim b. Hashim al-Qummil and others called Qadtyd Amir al-Mu 'minin or
'Aj'ib ahkdm AmTr al-Mu minTu, now published as MUHSIN AL-AMIN AL-', MILI, 'AJA'IB AtI K'AM
AMIR AL-MU'MININ 'ALT B. ABT TALiB (Markaz al-Ghadir li'l-Dirasft al-Islmiyya 2000))). For Shit
treatments, see, for example, 2 AL-HURR AL-'AMILI, WASAIL AL-SHI'A 172 (Mu' assasat Al al-Bayt li-
Iby5' al-Turdth 1956); 7 MUIAMMAD B. YA'QOB AL-KULAYNI, KAF 289 ('Ali Akbar al-GhaffarT ed.,
1957) (quoting the story from lbrdhm b. Hshim al-Qumml as received through that author's son, 'Ali
b. Ibrahim al-Qummi).
141. See infra Section III.B.
142. See PETERS, supranote 5, at 12-17.
story told by a second man who confessed to the crime, and no eyewitnesses or
reliable confessions. Because of pervasive problems in certain types of proof,
jurists read this case as support for avoiding capital punishment altogether.
3. Case of the Absentee Husband
Another set of doubts appeared in reports of a third case, which combined

the judgments of the second and fourth caliphs-cum-scholars, 'All and 'Umar,
respectively. The reports of this case noted that a man had gone traveling for
two years and returned to find his wife pregnant. 143 The husband accused his
wife of adultery. When he brought the case to 'Umar to adjudicate, 'Umar
determined that the woman was guilty of adultery and sentenced her to death.
When 'All heard about the decision, he rushed to 'Umar, saying, "[I]f this is
your decision as to the woman, then what will happen to her (unborn) son?"
'All managed to convince 'Umar to delay the sentence until the woman gave
birth. Once the son was born, he convinced 'Umar to postpone the sentence
until the baby was no longer reliant on nursing for sustenance. During this
time, perhaps as 'All had intended, the husband claimed44paternity of the child,
effectively rescinding the earlier accusation of adultery. 1
Umar observed this turn of affairs with surprise and regret. He had
initially sentenced the woman to death. If he had carried out the punishment, he
would have been at fault, because the husband claimed the child and dropped
the basis for criminal liability and punishment. 14'Umar 5
expressed his
appreciation to 'Ali and retracted the punishment order.
This case depicts 'AllI using various procedural strategies in advising
'Umar in order to avoid enforcing punishment even after the adultery
conviction. The husband's eventual "admission" of paternity-however
dubious-eventually served to cancel the punishment order altogether. Later
jurists read 'AI to have special insight or "judicial acumen" that events might
take this turn. 146 Indeed, he perhaps engineered matters to achieve this
outcome. These later jurists read this case as yet another instance of the
imperative for judges to identify contextual factors as bases for what might be
regarded as moral doubt about the propriety of punishment,
47 which demand the
use of procedural strategies to avoid punishment.1
These cases are important not for their historical accuracy, but for the
uses to which later jurists put them. 48 Pragmatic jurists writing at the end of
143. QUMMI, supra note 140, at 119 (citing variants in later Shi'T sources, including works by
prominent eleventh- to fourteenth-century ShV'T jurists Mufid and al-'Allma al-Hilli, as well as works
by prominent Sunni scholars Bayhaqi, Bqillani, and Ibn Hajar).
144. Id.
145. Id.
146. See IBN QAYYIM AL-JAWZIYYA, AL-TURUQ AL-I.IUKMIYYA FT'L-SIYA A AL-SHAR'IYYA 3,
43-80 (Dar al-Kutub al-'Ilmiyya 1995).
147. Id.
148. Reports of these cases from the seventh century are available in written literature only
from the eighth to ninth centuries at the earliest. Rather than taking them as verbatim reports of what
the founding period in the eleventh century and afterward read these early cases
to express doctrines of doubt. 49 They then used these cases to support their
view in the Case of Mj'iz that the Prophet's comment that the townspeople
should have avoided punishing M 'iz was a normative rule based on the
doctrine of doubt. In that case, the Prophet had tried to deter Md'iz from
confessing, had come up with alternative explanations of the defendant's
actions that did not fall within the ambit of the law in order to excuse Md'iz,
and then had censured the people for carrying out the criminal punishment
against him. In short, one could characterize their treatment of the Case of
Md'iz as reflecting an understanding that the Prophet had expressed factual,
legal, and moral doubt.
Drawing on that precedent, Muslim jurists interpreted the Case of the
Mysterious Pregnancy to be an instance of legal doubt, with 'Umar displaying
an immediate willingness to ascribe no fault to the pregnant woman who
signaled she had no criminal intent and thus was not covered by the scope of
the law against fornication and adultery. They further interpreted the Case of
the Falsely Accused Butcher as one of factual doubt, wherein 'AlI declined to
punish the man who had confessed to committing murder, when he was no
longer certain of the criminal culpability of either defendant where there were
two competing confessions. And they interpreted the Case of the Absentee
Husband as one of moral doubt, where 'AlI and 'Umar managed to avoid
punishing a woman for adultery through procedural delay tactics where the
context of her husband's acceptance of the child and the child's reliance on the
mother militated against the propriety of capital punishment. In sum, the judge
in each case found (or created) some type of factual, legal, or moral doubt-
sometimes in combination-that called criminal culpability into question.
These various types of doubt provided the basis for avoiding criminal
punishment in each case.
Read together, these early cases represented judicial practices that
resonated in juristic circles as normative, even if they were not as airtight as
arguments based on text. For the later jurists considering these early cases, the
outcomes pointed to an overall tendency of early Muslim judges and founding
figures to avoid punishment through heightened procedural and other
requirements, even where there were solid indications that some crime had
taken place. Having distilled this understanding from early judicial practices,
the jurists expressed it frequently in 5the
0
form of the doubt canon: avoid
criminal punishments in cases of doubt.'
actually happened, I argue that the case reports indicate ideas that the later generations of scholars came
to associate with Islam's earliest period. In a related context, Roy Mottahedeh has used sources from the
Biiyid dynasty in the late 'Abbgsid era to uncover the symbolic importance and practice of taking oaths
in processes through which local rulers gained political legitimacy. See ROY P. MOTTAHEDEH,
LOYALTY AND LEADERSHIP IN AN EARLY ISLAMIC SOCIETY passim (1980).
149. E.g., QUDORI, supranote 50, at 5949-50.
150. See, e.g., MUAMMAD AMIN AL-ASTARABADT, AL-FAWA'ID AL-MADANIYYA 326
(Mu'assasat al-Nashr al-Islrni 2003); 9 SuLAYMAN B. KHALAF AL-BAJI, MUNTAQA 168-71, 175-76,
232 (Maktabat al-Thaqdfa al-Diniyya 2004); 6 ABO HAMID AL-GHAzALT, AL-WAsIT 443-46 (Dr al-
Salem 1997); 2 IBN AL-'ARAB!, supra note 58, at 604; 2 IBN 'ABD AL-BARR, KAFl 1069-73 (Maktabat
Yet as a statement of judicial practice, the doubt canon would have

limited reach. For staunchly textualist jurists, judicial practices would have
little normative weight, as they were neither Qur'dnic nor prophetic statements.
For maximal authority and legitimacy, any canon had to be rooted in a
foundational text. 151 That move came next for the doubt canon.
B. The Textualization ofDoubt Through JuristicInterpretation
Once the jurists had expressed the doctrine of doubt in the form of an oft-
repeated canon located in early judicial practices, the jurists' next move was to
turn the canon into a text. My claim is that the jurists literally asserted that the
divine lawgiver had laid down the rule through the Prophet himself. They did
not do so based on any one text or at any one identifiable moment. Rather, they
did so based on their interpretations of the early .hadTth literature alongside
judicial practices that seemed to consistently pursue a policy of punishment
avoidance in cases of doubt, including the Case of M 'iz and other early cases.
By the end of the tenth and eleventh centuries, interpreting these cases in the
aggregate, virtually all Muslim jurists came to conceive of the doubt canon as a
foundational text issued by the Prophet himself.152 This move represented a
radical change from the early conceptions of the doubt doctrine as judicial
practice;this later conception of doubt had been transformed into a rule of text.
A study of the legal literature from the first three centuries of Islam's
advent reveals that no jurist understood direct statements about doubt to have
been prophetic in origin. To be sure, they knew of the doubt canon expressed
through its standard formulation: "avoid criminal punishments in cases of
doubt." Yet, they regarded it as statement of judicial53 practice, not as a prophetic
haaith. It simply was not a foundational legal text.
In this vein, the founders of Islam's multiple legal schools tended to cite
the doubt doctrine in their treatments of criminal law without claiming it to be a
prophetic directive. Abu Hanifa's circle in Iraq expounded the doubt doctrine
early on in the form that has been popularized amongst most subsequent jurists:
al-Riydl al-Hadithah 1980); 2 IBN RUSHD AL-HAFID, BIDAYAT AL-MUJTAHID 632 (Dar al-Kutub al-
'Ilmiyya, 1996); 16 IBN RUSHID AL-JADD, BAYAN 324 (Dar al-Gharb al-islamI 1988); 9 ABC BAKR B.
MAS'OD AL-KAsANI, BAD,' AL-SANAI' 4150 (Zakariyy Yfisuf 1968); MAWARDI, supra note 50,
'Alli
at 207-15; 9 SHAMS AL-A'IMMA AL-SARAKHsT, MABSOT 66 (AbOi 'Abd Allah Muhammad Hasan
Ism5't1 al-Shdfi'i ed., Dar al-Kutub al-'Ilmiyya 2001); ABC ZAYD AL-DABOSI, KITAB AL-ASRAR, in
Salim Ozer, Debbusi'nin "El-Esrarfi'l-UsulveT-Furu "' Adli Eserinin Tahkik ve Tahlili 151, 1229
(1997) (unpublished Ph.D. thesis, Erciyes Oniversitesi) (on file with author).
151. For approaches to early Islamic history and historiography in other contexts, see generally
FRED M. DONNER, NARRATIVES OF ISLAMIC ORIGINS: THE BEGINNINGS OF ISLAMIC HISTORICAL
WRITING (1998), which argues that early Islamic sources present a "kernel" of historical truth; TAYEB
EL-HIBRI, REINTERPRETING ISLAMIC HISTORIOGRAPHY: HARON AL-RASHID AND THE NARRATIVE OF
THE 'ABBASID CALIPHATE (1999), which identifies early Arab historical writing as moralizing rather
than presenting a register of facts; R. STEPHEN HUMPHREYS, ISLAMIC HISTORY: A FRAMEWORK FOR
INQUIRY (1991), which surveys major contemporary approaches to the field; and CHASE F. ROBINSON,
ISLAMIC HISTORIOGRAPHY 118-23 (2003), which notes that political patronage colored historical
narratives and, where independence was lacking, often presented facts best suited to legitimate the
sponsoring regime.
152. Rabb, Islamic Legal Maxims, supra note 17, at 100-06.
153. For a more in-depth survey, see id.at 5.
"avoid criminal punishments in cases of doubt."'1 54 His prominent disciple, Abu

Yiisuf-who, in the eighth century, became the first chief judge of the Islamic
empire-included the canon in his treatise on public law. a 5
The leaders of schools elsewhere in Iraq, Central Arabia, Egypt, Spain,
and the other major centers of the growing Islamic empire recognized practices
driven by the doubt doctrine as well. 156 In Medina, Malik invoked it, to avoid
capital punishment in a case of alleged adultery. 157 In Andalusia, Malik's
student Ibn al-Qdsim used claims of doubt to avoid punishments also in cases
of alleged adultery. 5 In Egypt, Sh~fi'l applied the doubt doctrine as well.
Shdfi'T's application of the doubt doctrine is particularly striking, because his
work of legal theory emphasizes the need to rely only on textual sources of
law. 159 Yet he too recognized and tried to accommodate the persistence of
factual doubts in the evidence. He also advocated applying the doubt 160
canon
where witness testimony conflicted as to the specific facts of a crime.
The same doctrine of doubt also featured in the proto-Shil' legal circles in
Kufa, Iraq. The doubt canon appears, without prophetic authority, in a work of
'All's judgments collected in the ninth century. In one case, a man gave his
wife a slavewoman and then had sex with her. When the woman complained to
'Al, accusing her husband of illicit sexual relations, 'All resolved the matter by
avoiding capital punishment. 161Moreover, when he was still caliph, 'AlI
advised his faithful companion Malik 62al-Ashtar to apply the doubt canon when
sending him to be governor of Egypt.'
In sum, all of the founding jurists invoked doctrines of doubt to avoid
154. One of Abo Hanlfa's two most prominent students, Shaybani, attributed instances of the
doubt canon's application to his teacher, who in turn drew on opinions of the seventh-century jurist
Ibrihtm al-Nakha'. See MUJAMMAD B. HASAN AL-SHAYBANI, KITAB AL-A/THtAR 136 (Khadlja
Mubammad Kamil ed., Idarat al-Qur'fln wa'l-'Ulum al-Islamiyya 1998).
155. AbO Yasuf included discussions of the doubt canon in his treatise on taxation, which
contains sections on criminal and other areas of public law. See ABID YOsur, KrrAB AL-KHARAr 303-05
(Mubammad Ibr-ahim al-BannV' ed., Ddr al-Ildh 1981).
156. E.g., 7 'ABD AL-RAZZAQ, AL-MUSANNAF F! 'L-IIADITH 210 (Iab-h al-A'Zam! ed., al-
Maktab al-IslmT 1972) (reporting on avoidance practices in Medina of the Umayyad caliph 'Umar b.
'Abd al-'Aziz (r. 717-720) for alleged sex crimes); WArd', supra note 91, at 617 (reporting cases in
which the Syrian Umayyad judge Fadldla b. 'Ubayd al-Anffri avoided punishment).
157. See 2 MALIK B. ANAS, AL-MUWATTA' 393 (Bashshar 'Awwad Ma'rf ed., Dk al-Gharb
al-Isldin 1996) (holding, for example, that judges were not to punish in a case where a man permitted
his slavewoman to have sex with another man, even though giving the permission and the act were
illegal).
158. See 16 SAI3NON, supranote 137, at 236 (noting Mdlik's position that judges should avoid
punishment for adultery in a case of alleged infidelity where there were disputes about whether the
marriage was consummated, and thus completed legally).
159. See generally SHAFt'l, supra note 12 (elaborating a jurisprudence based on his call for
authentic foundational texts as the most authoritative bases for law).
160. SHArFI', supra note 50, at 52-53 (invoking the doubt canon to avoid corporal punishment
in cases of theft, though requiring the thief to pay afine in such cases).
161. See QuMM!, supranote 140, at 253-54.
162. See IBN SHU'BA, TULTIAF AL-'UQOL 126, 128 ('Ali Akbar al-Ghaffar! ed., Maktabat al-
Sadfiq 1376). This source is dubious, and the doubt canon does not appear in the version of the letter
recorded in RAVI, supra note 90. Yet this version indicates a certain regard for the doubt canon in the
ShVI communities that Ibn Shu'ba frequented in the tenth or eleventh century, during a period when the
canon was growing in importance.
punishment, but none understood the doubt canon to be a prophetic statement.

Moreover, the scholars who were devoted to authenticating the many reports of
prophetic statements never conferred on the canon in its standard form any
stamp of authenticity and therefore never included the reports in their
collections. 163 However, with the breakup of the empire and the need for
increased juristic authority to ameliorate excessive political policies, all this
would change.
Remarkably, after the tenth century, virtually every Muslim jurist

invoked the canon not as a requirement of judicial practice, but as a prophetic
statement.164 That is, during this period, the majority of jurists-the Hanafi,
Maliki, and Shdfi'i, as well as Shi'T schools, all of whose founders had
deployed the doubt canon without prophetic attributions-now came to regard
the canon as prophetic.1 65 The need to avoid capital punishment, I argue, had
grown so urgent that jurists needed a doubt doctrine of unassailable authority to
justify avoiding punishment. They gained one by attributing the canon to the
Prophet directly.
The textualization of doubt was a stroke of organic genius. If the doubt
canon was a prophetic directive, it would become a foundational text that
elevated its authority to a position above jurists, caliph, and judges. With a firm
textual basis, the canon would stamp Muslim jurists' doubt jurisprudence with
an incontrovertible mark of authority. Specifically, a textual rule could more
readily justify the normative outcome as one of avoiding punishment in the
Case ofMd 'iz. More generally, a textual rule would offer a direct expression of
divine legislative intent, expositions of which enjoyed unquestioned legitimacy.
All of this occurred just when jurists needed to bolster their authority-which,
ironically, was accomplished through claims of doubt-in the violent eleventh
century.
I1. THE DEPLOYMENT OF "REASONABLE DOUBT" TO LIMIT EXECUTIVE

POWER
It was not inevitable that the doubt canon would come to dominate
criminal law. Jurists and caliphs alike could have just as well used the divine
legislative supremacy ideal to emphasize their subordination to the law and its
dictates of criminal law enforcement-as had the Prophet in the Case of the
MakhzamT Thief 166 Likewise, the Case of Md'iz could have been read to
emphasize the validity of capital punishment and the imperative of law
enforcement, despite the harsh consequences and what might have been the
163. For a more detailed survey, see Rabb, Islamic Legal Maxims, supra note 17, at 69-77,
which suggests that these developments reflected a shift in early understandings of legal validity in
Islamic law, from reliance on practice to authoritativeness of text.
164. The earliest reliable juristic attribution of the doubt canon to the Prophet is that of the
.Hanafi jurist Jas5$, 3 supra note 22, at 330. Such attributions increased in the eleventh century.
165. See supra note 150; see also Rabb, Islamic Legal Maxims, supra note 17, at 100-06.
166. See supra note 59 and accompanying text.
Prophet's personal inclinations toward mercy. Ddw-dd al-ZdhirT and other

textualist jurists had made these very arguments.
Yet the vast weight of the Islamic legal tradition was on the opposite end
of the scale. Mawardi and other mainstream jurists of his time had extracted a
principle of avoiding punishment from the early precedents and used them to
make the Islamic doubt canon the organizing principle of criminal law. In light
of the legislative supremacy and textualist ideal, I have detailed how criminal
law went in this surprising direction. I will now advance an argument as to
why. In brief, I argue that the jurists' early canonization of doubt and their later
textualization and generalization of the doctrine in the eleventh century came in
response to the political culture of violence that accompanied the breakup of
the Muslim empire at a time when Muslim jurists were keen on systematizing
Islamic law.
A. The Systematization of Islamic Law in a Political Culture of Violence
The early cases, by requiring more procedure, had put on display some of
the moral concems of the early juristic community as they were systematizing
Islamic law. That is, the jurists responsible for articulating the law used doubt
to express concerns about draconian punishments in the political sphere that
percolated in the background of, and were subject to, the developing norms of
Islamic law. The local rulers who seized control in the central Muslim lands of
the eleventh century (and in the later 'Abbasid period) asserted power primarily
through military force and punishment as a means of social control. 167 As I
argue here, their excesses prompted the jurists to constrain the caliphal reach
over Islamic criminal law. That is, as jurists began to systematize the law, they
did so against this early legacy of political violence and arbitrariness, which
continued-and indeed increased-after the 'Abbasid grip on absolute power
began to loosen.
To be sure, previous regimes were known for excess. Mu'dwiya's use of
military force to depose 'AI and establish the Umayyad dynasty became
standard practice for dynastic regime change in the premodem Muslim world.
The dynasty following his, that of the 'Abbasids, rose to power on the heels of
an army that marched from the East to take control from the Umayyads in
750.168 The Umayyad and 'Abbasid dynasties both indulged in disparate and
excessive applications of punishment. 169 Their excesses prompted the jurists
167. See generally LANGE, supra note 126 (detailing the particularly violent Seljdiq use of
punishment for social control).
168. See generally MOSHE SHARON, BLACK BANNERS FROM THE EAST: THE ESTABLISHMENT
OF THE 'ABBASID STATE-INCUBATION OF A REVOLT (1983) (examining the revolt that brought the fall
of the Umayyads and rise of the 'Abbfsids); MOSHE SHARON, REVOLT: THE SOCIAL AND MILITARY
ASPECTS OF THE 'ABBASID REVOLUTION: BLACK BANNERS FROM THE EAST 1 (1990) (same).
169. See JONATHAN P. BERKEY, THE FORMATION OF ISLAM: RELIGION AND SOCIETY IN THE
NEAR EAST, 600-1800, at 84-85 (2003) (noting that while some of the opposition came from a stance
championing Arab ascendancy, most of the opposition was religious, as diverse pockets of scholars and
other members of the elite became increasingly distressed at reported caliphal divergences from
Qur'anic and prophetic principles ofjustice).
not only to distinguish themselves from the caliph, but to form a "pious
opposition" against him, seeking 70to restrict as much as possible the caliphal
reach over Islamic law writ large. 1
Major developments in the eleventh-century political center of the
Muslim empire led to significant changes in the administration of criminal
justice and juristic responses to it. Namely, break-off local dynasties wrested
ever more control from the central 'Abbdsid caliph. Notable amongst them
were the Seljulqs, a group of Central Asian warriors who assumed power over
Iraq and greater Persia in 1055.171
Under the Seljiiqs, violence became more excessive and public.
Executions were especially rampant, accompanied by floggings, public
shaming parades, and imprisonment.172 The term that Mdwardi had devised for
"political governance" (siyrsa)to demarcate the jurisdiction of the caliph from
that of the jurists had transformed. By the Seljiiq period, this term for "political
governance" had come to mean "punishment" or "execution., 173 A prominent
jurist of the time remarked that the "sultans of today rely on punishment
[siydsat] and awe,' 174 that is, the type of awe75 surrounding kings, induced by the
threat or application of severe punishment.'
The sultan-appointed and jurist-advised judges played a significant role in
the administration of criminal justice. Some scholars have suggested that the
political officials were able to skirt the jurist-advised ordinary courts and
transfer criminal justice administration to officials who fell solely under the
authority of the caliph. 176 Moreover, they have suggested that the caliph had
free jurisdiction over crimes that carried discretionary penalties. In particular,
these scholars have pointed to the "extraordinary" courts of equity over which
the caliphal officials presided, as well as to law enforcement officials who
sometimes presided over discretionary penalties. 177 But more likely, as
170. See Mottahedeh, supranote 84, at 6-8.

171. See BOSWORTH, supra note 93, at 185-88; see also WILFERD MADELUNG, RELIGIOUS
TRENDS IN EARLY ISLAMIC IRAN 54-93 (1988).
172. LANGE, supranote 126, at 39 (discussing the chronicles ofIbn al-Athir and Ibn al-Jawz!j.
173. See, e.g., NIZAM AL-MULK, SIYASATNAMAH 172 (Hubert Drake ed., 1962) (equating
siyasat to punishment); cf LANGE, supra note 126, at 42 (citing Bernard Lewis, Siydsa, in INQUEST OF
AN ISLAMIC HUMANISM 3, 4-7 (A.H. Green ed., 1984)).
174. ABO .AMm AL-GHAZALT, BOOK OF COUNSEL FOR KINGS 148 (F.R.C. Bagley trans.,
1964); see also ANN 1CS. LAMBTON, STATE AND GOVERNMENT IN ISLAM: AN INTRODUCTION TO THE
STUDY OF ISLAMIC POLITICAL THEORY 124 (1981).
175. See MOTrAHEDEH, supranote 148, at 185.
176. See l IBN KHALDOJN, MUQADDIMA 455-58 (Franz Rosenthal trans., 1958). Following lbn
Khaldan, see EMILE TYAN, HISTOIRE DE L'ORGANISATION JUDICIARE EN PAYS D'ISLAM 603 (1960).
177. E.g,, KNUT VIKOR, BETWEEN GOD AND THE SULTAN: A HISTORY OF ISLAMIC LAW 191
(2005) ("The sultan makes his verdict freely; he is not bound by the SharT'a rules in any way .... ).
While an evaluation of the actual operation of maz5lir courts and other types of executive law-
enforcement institutions like the shurta police force is beyond the scope of this paper, it is perhaps
sufficient to note that the Islamic legal theory through which Muslim jurists constructed rules for
legitimacy required that caliphal officials justify their rules with respect to the maximum standards that
jurists set in the fixed criminal law. See Mathieu Tillier, Qdcfis and the Political Use of the Mazalim
Jurisdiction,in PUBLIC VIOLENCE IN ISLAMIC SOCIETIES: POWER, DISCIPLINE, AND THE CONSTRUCTION
OF THE PUBLIC SPHERE, 7TH-18TH CENTURIES CE 42 (Christian Lange & Maribel Fierro eds., 2009)
(detailing the use of ma~5lim and shurlajurisdictions); Emile Tyan, Judicial Organization,in LAW IN
2015) "ReasonableDoubt" in Islamic Law
Mdwardi had intimated, there were spheres of shared jurisdiction between the
ordinary courts on the one hand and the extraordinary, political institutions on
the other, a scheme wherein even caliphal officials' exercises 178
of discretion
were supposed to be subject to juristic articulations of law.
One example of this type of shared jurisdiction was the police role in
criminal prosecution and law enforcement. 179 The police were an arm of the
caliphal power whose main job was to secure public safety, prosecute crime,
and enforce criminal punishments upon conviction in ordinary courts. 180 To be
sure, the police did not always bring their charges through the courts, and they
often imposed outsized punishments. For example, when a man of one
neighborhood in Baghdad killed 181
his sister and her lover, the police rode out and
destroyed the entire quarter.
But as exemplified in the articulation of the doubt canon during this time,
episodes like these fueled, rather than quashed, juristic opposition to excessive
and illegitimate punishment. Such incidents provided a powerful public symbol
of excessive punishment, which jurists could highlight as illegitimate executive
overreach in a way likely to garner the sympathies of the populace. Such
instances also pushed the jurists to make clearer statements of criminal law and
procedure to which the caliph and police force would have done well to adhere,
at least in part, in order not to compromise their legitimacy.
In this vein, the other official who played a role in criminal law
enforcement-the market-inspector (muhtasib)-"declined in people's
estimation when rulers neglected [the office] and conferred it on men of no
repute, whose goal was to make profit and get bribes. ' 82 As the name would
suggest, the inspector monitored the quality of goods and guarded against
price-fixing in the marketplace. Alongside the police, the inspector also had
punitive authority, usually over non-capital crimes. The inspector sought to
curb "undesirable" social practices like open alcohol consumption and singing
through punishments such as detention, flogging, and public humiliation (by
parading offenders around town).1 83 Like the judges-who sometimes rotated
to the position of muhtasib-the inspectors were formally limited by the
THE MIDDLE EAST: ORIGIN AND DEVELOPMENT OF ISLAMIC LAW 259 (Majid Khadduri & Herbert J.
Liebesny ed., Lawbook Exchange 2008) (1955) (same).
178. MAwARDT, supranote 118, at 250.
179. See 1 MUjAMMAD 'ABDALLAH AL-BATALYAWSI, AL-IQTI)AB FTSHARII AL-KUTrAB 159
(Mu tafa al-Saqqd & Hamid 'Abd al-Majid eds., al-Hay'a al-Misriyya al-'Ham li'l-Kit~b 1961); 12 IBN
AL-JAWZT, AL- MUNTAZAM FTTA'RIKH AL-UMAM WA-L-MULOK 124 (Muhammad 'Abd aI-Qfdir 'At'
& Mu~tafai 'Abd al-Qadir 'AtA' eds., 1992); 13 id.at 316; HUGH KENNEDY, THE ARMIES OF THE
CALIPHS: MILITARY AND SOCIETY IN THE EARLY ISLAMIC STATE 13-14 (2001).
180. See Intisar A. Rabb, Police, in THE PRINCETON ENCYCLOPEDIA OF ISLAMIC POLITICAL
THOUGHT 427, 427-28 (Gerhard B6wering et al. eds., 2012); see also LANGE, supra note 126, at 48-51
(detailing the role of the police under Seljiiq rule).
181. 17 IBN AL-JAWZI, supranote 179, at 99; see also LANGE, supra note 126, at 52.
182. MAWART, supra note 118, at 258; cf Henry F. Amedroz, The Hisba Jurisdiction in the
Ahkam Sultaniyya ofMtwarda, J. ROYAL ASIATIC SOC'Y 77 (1916).
183. See LANGE, supra note 126, at 56 (describing bisba activity under the Seljoqs); cf.
KRISTEN STILT, ISLAMIC LAW IN ACTION: AUTHORrrY, DISCRETION, AND EVERYDAY EXPERIENCES IN
MAMLUK EGYPT (2011) (describing hisba activity in Mamlak Egypt).
jurists' articulation of Islamic law. 184 That is, for maximum legitimacy, the
jurist-defined Islamic criminal laws provided the upper limits of the
discretionary
1 85
punishments that the inspector possessed the authority to
enforce.
Punishment on the part of the political rulers implicated questions about
the legitimacy of the caliphs' rule. Even if the jurists were not in agreement
with the severity of punishment-and they were not-"[t]he overwhelming
persuasive power of public punishment" left no question about the rulers'
authority.186 The difficulty lay in the rulers' claim to legitimacy, attacks against
which were the only real arrow in the jurist's quiver. At the same time, Muslim
jurists knew that they relied on the new rulers' enforcement power for the rule
of law. They therefore did not seek to completely delegitimize Seljfiq rule, but
to constrain it. 87 This explains why they needed to convert the doubt doctrine
into a textual rule.
B. The GeneralizationofDoubt in Response to Excessive Punishment
The jurists' appeal to the newly textual authority of the doubt canon
reflected moral anxieties stemming from the ongoing political culture of
violence combined with the factual uncertainty of criminal culpability in many
cases, both of which were exacerbated by the breakup of the empire in the
eleventh century. That is, the clear and present danger of death in the quick-to-
punish political context planted serious anxieties among many of the jurists,
who worried about frequent executions of dubious justification at precisely the
moment when they were trying to better articulate and systematize Islamic law.
Still, why should the jurists have understood doubt to include heightened
criminal procedures when the burdens of proof in Islamic criminal law were
already so high? The answer has to do with the nature of the punishments that
were of most concern. If jurists could deploy doubt doctrines to require
procedures that tended to avoid capital punishment rather than authorize it, they
could ameliorate any anxieties over authorizing execution on dubious grounds.
Conceivably, they could also thereby restrict the rates of capital punishment. In
184. E.g., MU'AYYAD AL-DAWLA, 'ATABAT AL-KATABA 52 (Muhammad Qazvlni & 'Abbds
Iqb5l eds., 2005) (reproducing an appointment diploma of hisbajurisdiction for ajudge).
185. See, e.g., LANGE, supranote 126, at 58 (citing 'ABD AL-RAIJMAN B. NA$R AL-SHAYZARI,
NIHAYAT AL-RUTBA FTTALAB AL-IjISBA 9 (Matba'at Lajnat al-Ta'lif wa'l-Tarjama wa'l-Nashr, 1946));
'ABD AL-RAHMAN B. NA$R AL-SHAYZARI, THE BOOK OF THE ISLAMIC MARKET INSPECTOR (Ronald P.
Buckley trans., 1999). Of course, adhering to this upper-limit was not always followed in practice. See,
e.g., STILT, supra note 183, at 89-106; TYAN, supra note 176, at 649 (relating a number of examples
from historical sources from other periods of muhtasibs exceeding the extent of punishments prescribed
by the hadd penalties).
186. LANGE, supra note 126, at 43 (arguing further that such coercive power left "no doubt"
about the ruler's legitimacy). While the author is correct to note that the authority was left virtually
unquestionable, my reading of the sources suggests that the legitimacy, by contrast, was questionable-
for which the jurists responded with a doctrine of doubt.
187. As Christian Lange has observed, Muslims much preferred the draconian criminal justice
system of the Seljitk dynasty to the utter chaos that reigned when those authorities lost power in
Khuras~n for some time beginning in 1153 until the Mongol invasion a century later. See LANGE, supra
note 126, at 244-46.
other words, where the stakes were life or death, using the doubt canon to
assure more procedure was the juristic attempt at alleviating their own
anxieties and ameliorating widespread practices of capital punishment.
188
1. MoralAnxieties Revealed: Death Is Different
Muslim jurists historically had been skittish about any involvement with
political authorities, which they deemed unseemly and corrupting for people of
moral conscience. The Muslim jurists' general skittishness toward judging, I
suggest, was exacerbated in capital contexts, which principally explained the
rise of the doubt doctrine in Islamic law.
Though some jurists certainly served as state-appointed judges and
assumed other government posts, they were a distinct minority. Many who
were nominated for official judgeships or other positions stubbornly refused to
serve, preferring to remain independent of the political apparatus.'89 The
sources are replete with instances of their recoil at being nominated to judge.
Famously, Abu Han1fa was said to have endured torture for his refusals."tg The
main chronicle for the early history of judges opens with prophetic cautions
against judging. 191 This inclination against judging represented just some of the
moral anxieties surrounding judging in accordance with God's law, and the
consequences of not getting it right. When it came to the death penalty, Muslim
jurists feared answering in the afterlife for wrongful deaths occasioned by their
allowance of punishment that was not in fact warranted. 192 Moreover, against
188. The American experience with the death penalty-from which I draw the title of this
section-well articulates the "difference" of death. When placing a moratorium on the death penalty in
1972, the Supreme Court cited the need for procedure, in light of the fundamentally different nature of
the death penalty, and the need to avoid executions that were so arbitrary as to be "wanton and freakish"
in nature. Furman v. Georgia, 408 U.S. 238, 295 (1972) (Brennan, J., concurring) (imposing what was to
become a four-year moratorium on the death penalty). In his concurrence, Justice Brennan explained
that this outcome had to do with the fact that "[d]eath is a unique punishment, . . . in a class by itself."
Id. at 286-89 (Brennan, J., concurring); see also id.at 306 (Stewart, J., concurring) ("The penalty of
death differs from all other forms of criminal punishment, not in degree but in kind."). When capital
punishment was reinstated in the 1970s, the Court subjected it to heightened procedures above that of
ordinary crimes. See Gregg v. Georgia, 428 U.S. 153, 188 (1976) (joint opinion of Stewart, Powell, and
Stevens, JJ.) ("[The] penalty of death is different in kind from any other punishment" and "unique.");
Woodson v. North Carolina, 428 U.S. 280, 305 (1976) (joint opinion of Stewart, Powell, and Stevens,
JJ.) ("[The] penalty of death is qualitatively different from a sentence of imprisonment, however long.").
For an analysis of the consequences of America's constitutional and procedural regulation rather than
abolition of the death penalty, see Carol S. Steiker & Jordan M. Steiker, Entrenchment andor
Destabilization? Reflections on (Another) Two Decades of Constitutional Regulation of Capital
Punishment,30 LAW & INEQ. 211 (2012).
189. See, e.g., ABO 'ABD ALLH AMMAD B. AL-HARITH AL-KHUSHANT, QUIAT QURTUBA
kMUH.
79-93 (Yhsir Salama Ab5 Ta'ma ed., Ddr al-SamT'i 2008) (listing instances of ten judges who refused
the judgeship); WAKI', supra note 91, at 24-25.
190. For a discussion of this and his extreme antipathy toward the ruling authorities and
censure ofjudges like his rival Ibn Abi LaylI who "colluded" with them, see MUi .AMMAD ABO ZAHRA,
ABO HANIFA: HAYATUH WA-'A$RUH, ARA'UH WA-FIQHUH 37-48 (Dhr al-Fikr al-'Arabi, 2d ed. 1965).
191. See WAKI', supra note 91, at 19 (introducing biographical reports on judges with the
"[section] mentioning [hadith and other] reports announcing the gravity of assuming a judicial post over
people and that whoever assumes [such a post] has been [metaphorically] slaughtered without a knife").
192. See Baber Johansen, Vom Wort- zum Indizienbeweis: Die Anerkennung der richterlichen
Folter in islamischen Rechtsdaktrinen des 13. und 14. Jahrhunderts[From Proof by Word to Proof by
Indicia: The Recognition of Judicial Torture in Islamic Legal Doctrines of the Thirteenth and Fourteenth
the backdrop of excessive executions in the Seljliq culture of political violence,

Muslim jurists were signaling through increased rules of procedure that they
had come to see the death penalty as different from punishments for ordinary
crimes.
Notably, the landmark cases to which the jurists appealed in order to
insert heightened procedures from the early judicial practices were capital
crimes. That fact, I suggest, sparked the doubt doctrine as a rule of more
procedures than Islamic criminal law ordinarily required. Procedures for non-
capital crimes-including theft, intoxication, and sexual slander-ordinarily
required two witnesses or a voluntary confession. 193 A conviction for ordinary
crimes warranted punishment that ranged from public flogging to hand
amputation: harsh, but not deadly. By contrast, the Qur'dn required four
eyewitnesses to sex crimes, where Muslims understood the penalty to be
death. 194 In this way, convictions for capital crimes-sex crimes and murder-
typically required more procedures.
By itself, this increased witness requirement for sex crimes did not lead to
the doubt doctrine, which was nowhere contained in the early canonical texts.
Only when combined with early judicial practices did this thinnest Qur'dnic
rule of increased procedure lead to a textual rule of generalized and procedural
doubt. In the Case of Md 'iz, the Prophet had exclaimed "if only you would
have let him go." This statement served to enlarge the space for punishment-
avoiding procedures where anxieties and considerations of mercy (such as
possibilities of repentance) were enough to avert punishment even where there
was a determination of guilt. Moreover, in the eleventh century, the Muslim
jurists analyzing the three early cases read alongside the Case of Ma'iz
determined that more procedure-far beyond an increased witness 195
requirement-was necessary for capital crimes of adultery, rape, and murder.
That is, jurists added a vast matrix of procedures to the four witness
requirement by repeatedly citing the early judicial practices and the textual rule
of doubt to avoid punishment in capital contexts. 196 Multiple confessions,
judicial investigations into mitigating circumstances, avoidance of textual
ambiguity, and other procedures became folded into the Islamic doctrine of
Centuries], 28 IUS COMMUNE 2, 11-12 (2001) (noting medieval Muslim jurists' fears of the legal and
physical consequences of punishment for false convictions, namely, the doctrine that the judge
responsible would suffer the same punishment in the afterlife that he had falsely inflicted in this life); cf.
LANGE, supra note 126, at 101-15 (noting the connection between this-worldly sins and fear of other-
worldly punishment).
193. 2 ABO AL-HASAN AL-MAWARD!, ADAB AL-QAQT 14-16 (Ri'asat Diwn al-Awqdf 1972).
194. For the four-witness requirement for sex crimes, see THE QUR'AN 4:15, and for the
related four-oath procedure of mutual imprecation (plus a fifth invoking God's wrath if found to be
lying) for spousal accusations of zinil, called li &n, see id. 24:7, 9. Another procedure contained in the
.hadth corpus, but not in the Qur'dn, was the elaborate process of swearing multiple oaths to assign
liability for homicide where the evidence proving a crime was inconclusive. On the fifty-oath procedure
called qasama (after which, capital punishment could not apply), see Rudolph Peters, Murder in
Khaybar: Some Thoughts on the Origins of the Qasdma Procedure in Islamic Law, 9 J. ISLAMIC L. &
Soc'Y 132 (2002).
195. See supra Section II.A.
196. See supra notes 131-163 and accompanying text.
doubt. In this way, the jurists outlined rules of heightened procedures as

necessary for the smooth functioning of criminal law, at least initially, because
of thefinal nature of capital punishment.
Demonstrating the jurists' preference to avoid rather than enforce
punishment, the canon assumed formulations that sounded like the old
Blackstonian quote that it is better to let ten men go free than to punish one
unjustly.1 97 Thus one version stated: "Avoid h.udfid punishments against
Muslims to the extent possible; if there is any way out, then release [the
convict], as it is better that the Imam make a mistake in pardoning than in
punishing." 198 The second caliph, 'Umar, reportedly said when acting in his
capacity as a judge: "If you come to me [with a criminal dispute], question to
the utmost extent possible, for I would rather make a mistake in pardoning than
a mistake in punishing."' 199 Muslims understood these early figures and these
varied formulations to be expressing a policy that more procedure before
punishment was warranted in capital cases. Explicitly, they elaborated and
justified this policy though the textual basis of the doubt canon. But implicitly,
the historical and legal records suggest that the policy was motivated by the
extent to which the jurists' own moral anxieties indicated that politically
favored sentences of death were jurisprudentially different.
In Seljiiq times, judges thus used this sort of doubt to support a textually
grounded doubt canon. Together, these judicial practices and the textual canon
could allow a morally anxious judge, to claim-in the face of pressures from
political officials to issue execution orders on doubtful facts-that his hands
were constrained by the Islamic law of doubt.2 °0
The juristic expression* of their anxieties came through decreased
legitimation of punishments until and unless there was factual certainty of a
crime. Such certainty was exceedingly rare. Only increased procedures could
yield the level of factual certainty that could alleviate their moral anxieties, and
otherwise the procedures would permit-and indeed require-avoidance of
punishment altogether. In this way, more procedure allowed jurists to
acknowledge the persistence of doubt and to increase the scope of their
authority to regulate punishment. Their hope likely was that the additional
197. See 4 WILLIAM BLACKSTONE, COMMENTARIES *358 ("[1]t is better that ten guilty persons
escape than that one innocent suffer.").
198. 10 'ABD AL-RAZZAQ, supra note 156, at 66; BAYHAQI, supra note 59, at 413; 4
D.RAQUTNT, SuNAN 62-63 (Mu'assasat al-Risala 2004); IBN ABI SHAYBA, supra note 46, at 360; 5
TIRMIDH, SuNAN 112-13 (al-Maktaba al-Salafiya 1965).
199. BAYHAQI, supranote 59, at 414; cf 1aN ABT SHAYBA, supranote 46, at 359.
200. Notably, although medieval Muslim jurists used the doctrine of doubt to express their
discomfort with the arbitrariness of capital punishment, they were careful to note that they were not
attempting to vitiate criminal punishment altogether. Instead, these jurists took criminal law prohibitions
to be justified-Bentham style-as rules aiming to deter crime. See, e.g., MAwARDI, supra note 50, at
99 (".Hudud are punishments by which God deters (zajara bihd) people from committing prohibited
[acts] and encourages them to follow [His] commands."). They also acknowledged a retributive
justification for criminal sanctions for defendants like M5'iz, who requested criminal punishment
believing it to provide this-worldly divine retribution that would expiate his sins and result in
forgiveness. Id. at 100. For an analogous meditation on contemporary criminal law responses to mass
atrocities, see MARTHA MINoW, BETWEEN VENGEANCE AND FORGIVENESS 9-24 (1998).
procedures would lower the rate of execution, or at least perceptions of its

legitimacy. Put differently, even if they could not stop punishment, jurists
could at least remove the imprimatur of religious-legal legitimacy for
punishment that only they could confer.
2. MoralAnxieties Compared: Death Is DramaticallyDifferent2 0 1
These types of moral concerns about judging and doubt were not unique
to medieval Muslim jurists and judges. Their Christian counterparts in England
and continental Europe were themselves morally anxious when it came to
dealing with doubt, and likewise created procedures designed to allay those
concerns. Shared across these disparate contexts, morality-based responses to
doubt were clearly regular features of criminal law in medieval religious
communities, even when their precise motives and means markedly differed.
As comparative legal historian James Whitman has argued, the
reasonable doubt doctrine that eventually emerged out of medieval Christian
contexts was "designed [not] to make it more difficult for jurors to convict, [but
to make it] easier, by assuring jurors [as well as witnesses and judges] that
their souls were safe if they voted to condemn the accused., 20 2 In the modem
scheme, finding facts particular to criminal elements became the equivalent of
establishing guilt "beyond a reasonable doubt," without special regard for truth
or moral concern about the soul of the one responsible for convicting a
defendant.203 This observation is significant inasmuch as it radically differs
from modem understandings of reasonable doubt as a safeguard for defendants'
rights. It is also significant in the extent to which it sheds light on the differing
trajectories of doubt rooted in the medieval Islamic and Christian worlds.
Despite the historical, institutional, and contextual differences between
the two legal systems, medieval European and English approaches to doubt
display significant parallels to the Islamic jurisprudence of doubt. For example,
medieval European and English judges and jurors initially declined to punish in
the face of doubt out of concern for their own souls. They feared-in addition
to the legal and physical consequences that sometimes obtained from
demonstrably false verdicts-the spiritual consequences of unjust verdicts,
201. I use this title to express the continuing death-is-different theme articulated in modem
American law, to refer to the extra procedures derived by medieval Christian and Islamic legal actors in
cases of capital punishment. See Schriro v. Summerlin, 542 U.S. 348, 363 (2004) (Breyer, J., dissenting)
(emphasizing that capital punishment requires extra procedural cautions not just because death is
different, but because of the "dramatically different nature of death").
202. WHITMAN, supra note 8, at 4, 11; id. at 194 ("The underlying concern was not with
protecting the defendant at all. It was with protecting the jurors."); cf Charles Donahue Jr., Proofby
Witnesses in the Church Courts of Medieval England: An Imperfect Reception of the LearnedLaw, in
ON THE LAWS AND CUSTOMS OF ENGLAND: ESSAYS tN HONOR OF SAMUEL E. THORNE 127, 132-34
(Morris S. Amold et al. eds., 1981).
203. For more on the somewhat contrasting notions that modem American liberalism relies on
constitutional due process values to align questions of reasonable doubt with defendants' rights, but that
the reasonable doubt doctrine emphasizes fact-based inquiries that sometimes obscure truth within the
context of procedural norms such as the exclusionary rule that may otherwise prejudice defendants, see
WHITMAN, supranote 8, at 203, 207, 334-36.
particularly when criminal convictions triggered punishments of death. 204

Heightened procedures for criminal trials thus grew out of a world of moral
anxieties facing judges about the very act '2of
06
judging.20 5 The idea was that God
Himself was "the great avenger of justice.
Promoting that view, the Church defined a complex rubric of procedures
that formed part of its moral "theology of doubt," which episodically shifted to
allow legal actors to alternately avoid and enforce punishment. Initially, the
Church introduced procedures designed to absolve the misgivings of legal
actors of the Middle Ages concerned about the stakes of getting convictions
wrong. An example was the use of the Ordeals, by which God became the
divine arbiter of right and wrong. When the Church abolished the Ordeals at
the Fourth Lateran Council in 1215, it aimed to transform judges from
"ministers of bloodshed" to "ministers of the law." 20 7 On the Continent, these
developments occasioned an almost immediate shift from judges to witnesses
as the bearers of moral responsibility for punishing serious crime. Following
rigorous canonical procedure provided cover for the thirteenth-century
European judge, who could assert whenever he condemned a defendant to
death: "It is the law that kills him, and not 1.,,208 In England, the Church
developments occasioned a shift from judges to jurors as the bearers of this
same moral responsibility, accompanying the rise of the jury in the same
century. The English judge could make a similar claim about the law,
deflecting both responsibility and doubt. The shift of doubt and moral
responsibility from judges to witnesses and jurors in many criminal cases
meant that "royal jurisdiction [had] finally displaced divine jurisdiction.2 0 9
This shift had been workable for a few centuries, when generalized
procedures shielded the newly responsible legal actors from personal
responsibility for the harshest consequences of their decisions. But when the
state instituted procedures that again placed moral responsibility for specific
punishment decisions on English jurors in the sixteenth and seventeenth
centuries, the latter found their consciences taxed once again and devised 210
new
strategies to avoid corporal and capital punishment in the face of doubt.
Over the course of these several centuries, multiple punishment
avoidance strategies manifested whenever judges, witnesses, or jurors
perceived their own moral responsibility for conviction in the face of doubt.
204. See I JAMES FITZJAMES STEPHEN, A HISTORY OF THE CRIMINAL LAW OF ENGLAND 573
(Routledge 1996) (1883) ("[Medieval judges often dreaded] the responsibility-which to many men
would appear intolerably heavy and painful-of deciding . . . upon the guilt or innocence of a
prisoner.").
205. See WHITMAN, supra note 8, at 11 (noting the work of "anthropologists and historians of
religion [demonstrating that] . . . anyone in the premodern world involved in the killing of another
person subjected himself to the risk of bad luck, bad karma, bad fate, or some kind of vengeful divine
retribution") (citations omitted).
206. Id. at 11, 192-93, 198-99.
207. Id. at 90.
208. Id. at 94, 105.
209. Id. at 138.
210. Id. at 161-63.
For example, witnesses often refused to testify, and judges found strategies to
avoid findings of guilt in ordinary crimes that they did not feel warranted
capital punishment. 211 Juries also expanded the scope of the old "benefit of
clergy" doctrine, allowing first-time offenders in fourteenth-century
ecclesiastical courts to receive lighter sentences in formulations available to
literate members of the Church.212 Moreover, prosecutors frequently
"downcharged" crimes or "downgraded" the value of stolen goods to arrive at
more lenient sentences than the statute would ordinarily require.213 And judges
regularly barred criminal convictions or construed criminal laws narrowly
when they had doubt about the relationship between the evidence2 14
and criminal
culpability or fairness of punishment for an alleged crime.
Taken together, the procedural manifestations of doubt took on multiple
forms that persist in modem American criminal law: the criminal law void-for-
vagueness doctrine and its "junior version," the rule of lenity; 215 the
presumption of innocence; 216 the principle of legality; 2 17 the rule against federal
2 19
common law crimes; 218 doctrines of mistake to absolve criminal liability;
and of course heightened standards and burdens of proof-including
reasonable doubt. Possessed of only a general awareness of the roots of
reasonable doubt and lenity, Justice John Marshall famously stated in his
nineteenth-century invocation of the American rule of lenity: "The rule that
penal laws are to be construed strictly is perhaps not much less old than
220
construction itself.,
211. JEROME HALL, THEFT, LAW AND SOCIETY 92-96 (1935) (noting that judges used to
"invent technicalities in order to avoid infliction of the capital penalty").
212. See LEONA CHRISTINE GABEL, BENEFIT OF CLERGY IN ENGLAND IN THE LATER MIDDLE
AGES (Octagon Books 1969) (1929); see also JOHN HAMILTON BAKER, AN INTRODUCTION TO ENGLISH
LEGAL HISTORY 513-15 (4th ed. 2002) (1971) (overview); J.M. BEATTIE, CRIME AND THE COURTS IN
ENGLAND 1660-1800, at 141-45 (1986) (examples); John Langbein, Shaping the Eighteenth-Century
CriminalTrial: A View from the Ryder Sources, 50 U. CHI. L. REV. 1, 37-41 (1983) (further examples).
213. LANGBEIN, supra note 8, at 334-35 (citing, inter alia, BLACKSTONE, supra note 197, at
239); see also LEON RADZINOWICZ, A HISTORY OF ENGLISH LAW AND ITS ADMINISTRATION FROM
1750, at 83-106, 138-64 (1948).
214. For the development of these reasonable doubt practices in English common law and in
continental European law, see WHITMAN, supra note 8, at 334-36. As for the development of lenity
practices, there is debate about when lenity became prevalent, but agreement that it was in full force by
the eighteenth century. See LANGBEIN, supra note 8, at 335 (noting that the rule predated the eighteenth
century but did not become prevalent until then); WHITMAN, supra note 8, at 185-200; Livingston Hall,
Strict or Liberal Construction of Penal Statutes, 48 HARv. L. REV. 748, 750 nn.12-13 (1935) (tracing
the consistent use of the rule of lenity back to mid-seventeenth-century England).
215. See HERBERT PACKER, THE LIMITS OF CRIMINAL SANCTION 95 (1977). For recent
applications, see supra note 11.
216. See Taylor v. Kentucky, 436 U.S. 478 (1978) (labeling the term the "assumption of
innocence"); Coffin v. United States, 157 U.S. 432 (1895) (tracing the concept back to Roman law,
canon law, and early common law); Cummings v. Missouri, 71 U.S. 277 (1866).
217. See McBoyle v. United States, 283 U.S. 25, 27 (1931) (stipulating that the criminal law
texts and punishments must be clear before a criminal defendant may be punished for violating the law).
218. United States v. Hudson & Goodwin, 11 U.S. (7 Cranch) 32, 34 (1812). For further
analysis, see Dan M. Kahan, Lenity andFederal Common Law Crimes, SuP. CT. REV. 345 (1994).
219. See DRESSLER, supra note 3, at 153-65 (mistakes of fact); id. at 167-80 (mistakes of law).
220. United States v. Wiltberger, 18 U.S. (5 Wheat.) 76, 95 (1820). On the reasonable doubt
doctrine, see In re Winship, 397 U.S. 358, 362-63 (1970) ("[P]roof of a criminal charge beyond a
reasonable doubt is constitutionally required .... The standard provides concrete substance for the
While medieval Muslims and Christians shared moral concerns and

devised punishment-avoidance procedures to alleviate doubt, it is clear that the
shape of, impetus behind, and goals targeted by their respective doubt-based
procedures were quite different. The differences between medieval Islamic and
Christian treatments of doubt have to do with differences in theology and
institutional structure. In contrast to the prominent role medieval Muslim jurists
played in constructing the law, their English and European counterparts were
not autonomous or even first actors in that task. The latter group had more
scholastic and less singular power to legitimately define law. Moreover, it was
two competing and highly centralized institutions that defined law for English
and European judges: the Church and the Crown. The Crown eventually won
the jurisdictional battle over criminal procedure, but by then, the Church had
developed a moral theology that absolved most legal actors of the
consequences of doubt-which together led to procedures that would allow
doubt to facilitate rather than bar convictions.22 1
In contrast, medieval Muslim judges and jurists-lacking a centralized
authority in the form of a unified church or state to define the law-continued
to see doubt as a matter of moral concern. This perspective, in turn, informed
the scope of what they saw as their divinely delegated power to define Islamic
law, theology, and morality in criminal law and other arenas. Wracked by
doubt and quite aware of their institutional position as exponents of Islamic law
in "pious opposition" to the state, particularly in matters of criminal
punishment, I argue that Muslim jurists pursued at least three goals when
defining a jurisprudence of doubt. They used doubt to alleviate a deep sense of
moral concern with false convictions, to rein in rather than work with the
political rulers in punishment, and to construct a version of Islamic law in line
with their ideals of divine legislative supremacy. With no moral theology to
absolve doubt and no state to assert absolute control over punishment or the
authorization of punishment, medieval Muslim jurists had both moral and
institutional reasons to deploy a procedure-laden, robust doctrine of doubt.222
As in the Christian historical context, Islamic doctrines of doubt meant
much more than the modem notion of uncertainty about facts in evidence.
Further, similar to the English and European doctrines of doubt that reflected a
cautionary stance toward criminal law generally and the death penalty
particularly, Islamic doctrines of doubt came to be expansive. As the next
section details further, Islamic doubt came to encompass everything from
presumption of innocence-that bedrock 'axiomatic and elementary' principle whose 'enforcement lies
at the foundation of the administration of our criminal law."' (quoting Coffin v. United States, 156 U.S.
432, 454 (1895))).
221. WHITMAN, supra note 8, at 131 ("[T]he kings of England were involved in political
conflicts both with the church and with a recalcitrant, and sometimes rebellious, feudal nobility. In the
early days of the development of English law, those conflicts expressed themselves largely as conflicts
over jurisdiction.").
222. The modem situation is quite different. It was only much later that the breakdown of
traditional moral and institutional structures would lead to the wholesale dismissal of doubt in most
modem Islamic criminal law contexts-a fact underscored by this examination of its otherwise forgotten
history. For further discussion see RABB, supra note 25, at 320-21.
questions of fact and law to issues of the propriety of punishment and mercy.
3. Moral Anxieties Expanded: A General JurisprudenceofDoubt
In the Islamic context, the execution-averse anxieties behind the jurists'

doctrines of doubt quickly spread beyond the capital context. Jurists advised
judges to apply the doubt canon to ordinary crimes as well. Further, they
interpreted doubt expansively to include rules for more procedure as well as to
adjust the substantive definitions of crimes. All this created what I call a
generalized "jurisprudence of doubt" that would come to dominate Islamic
criminal law.
The unifying "text" for diverse jurists to rally around was the doubt canon
and expansive definitions of "doubt" itself--over which jurists had long
asserted interpretive authority. In addition to calling for more procedure, most
jurists saw in the doubt canon numerous manifestations of doubt. That is,
Islamic doctrines of doubt meant procedural and evidentiary requirements of
proof beyond a reasonable doubt. They also meant an Islamic rule of lenity to
cover interpretive doubt or ambiguity. Muslim jurists further used the doubt
canon to avoid punishments for any acts whose criminality was doubtful in
light of the "interpretive differences" across schools.223 They also read the
doubt canon to require judges to avoid punishment where defendants claimed
mistake-of-law or mistake-of-fact. 224 Additionally, they read it to require
operation of the principle of legality and the presumption of innocence. 225 In
short, Muslim jurists relied on the doubt canon and used it to advocate
avoidance of punishment broadly on both procedural and substantive grounds.
Much of the jurists' generalization of doubt jurisprudence occurred in the
context of Islam's unique system of legal pluralism. Although Muslim jurists
had assumed the authority to interpret the law, they never agreed on how
conflicting texts should be interpreted. Unlike most modem contexts, these
jurists historically operated in a system where no high court issued a single
decision in the face of competing interpretations. Instead they opted to
accommodate multiple interpretations. This doctrine constituted Islam's system
of legal pluralism: "the [institutional] ability to countenance a plurality of
equally authoritative legal interpretations. 2 26 Under this system, any rule that
223. QARAFT, supranote 25 (labeling this type of difference "doubt in interpretive approach").
224. They systematically articulated these doctrines in new works dedicated to legal maxims
under the Mamlik dynasty in Egypt and Syria, which rose with the complete collapse of the 'Abbasid
caliphate. For example, the Sh~fi'i jurist Suyaiti identifies three additional types of doubt: mistake-of-
law, mistake-of-fact, and interpretive difference. SUYOTI, supra note 25, at 237. Compare the M5liki
jurist, Shih~b al-Din al-Qarffi, who identified those three categories of doubt. QARAFI, supra note 25, at
1307-09. Notably, the concepts had existed before in thefiqh treatises of the late 'Abbasid era. See, e.g.,
GHAZALT, supra note 150, at 443-44 (identifying mistake of fact, mistake of law, and "doubt in the
interpretive approach to permissibility"); 7 YAI.YA B. SHARAF AL-NAWAWI, RAWIDA 306-12 (al-Maktab
al-IslAmT 1991) (identifying interpretive doubt as "doubt in the juristic perspective or interpretive
approach"); see also 11 'ABD AL-KARTM AL-RAFI'!, AL-'AzTz 145-47 (Dar al-Kutub al-'llmiyya 1997)
(same).
225. See, e.g., RABB,supra note 25 (describing these notions as elaborated in Shil' contexts).
226. See JACKSON, supra note 108, at 142 (quoting the definition of legal pluralism advanced
was valid in one Sunni school was to be recognized as valid in all.

In criminal law, this legal pluralism presented a challenge. How were
jurists to handle interpretive diversity? Islamic criminal law doctrine held that
the Lawgiver prohibited crime through clear statements, and the doubt canon
required factual certainty before implementing punishment. Yet differences in
interpretation could create uncertainties in substantive definitions of crime as
well. That is, jurists from different schools of interpretation could regard the
same act as a crime in one school but not a crime in another school.
Without a path to uniformity, some jurists constructed a doctrine of
interpretive doubt as a means to accommodate, if not resolve, conflicting
interpretations of law. This type of doubt differed from the dueling interpretive
approaches to the Case of Md'iz between strict textualists and pragmatic
textualists. In that duel, early Muslim jurists disagreed on whether they were to
strictly adhere to the dictates of the text, or were instead to look beyond the text
for contextual clues on how to pragmatically read the text. They were arguing
about whether the text alone mattered or whether the text should be interpreted
within a particular context to give it proper effect. By contrast, interpretive
doubt arose among one segment of the dueling approaches, the pragmatic
jurists, who had decided that context mattered but admitted to the problem of
textual ambiguity that arose whenever they considered context. Over time, they
increasingly concluded that the foundational texts did not clearly address new
factual scenarios that judges were likely 2 27
to encounter. Nor did those texts
clearly prohibit all undesirable conduct.
Consider, for instance, the interpretive difference concerning valid forms
of marriage. The early Islamic historical sources suggest that the Prophet
initially allowed "temporary marriages," wherein a couple specifies a date upon228
which their union would dissolve automatically without divorce proceedings.
But later Sunni jurists accepted that the second caliph 'Umar had subsequently
outlawed the practice. 229 The disputed status of this form of marriage, if
deemed invalid, could have stiff consequences in criminal law. Sex in an
by distinguished Egyptian jurist Shihab al-Din al-Qaraft during his tenure as a judge in the multiple
legal school-jurisdictions of thirteenth-century Mamliik Egypt).
227. For medieval Muslim jurists' definitions of textual ambiguity, see, for example, 4
MUSTAFA MUHAQQIQ DAMAD, QAVA'ID-I FIQH 54 (Sazimdni Mut~li'-i va Tadvin-i Kutub-i 'Ulcim-i
InsdnT-i Ddnishgihhd 2001), which defines textual ambiguity as uncertainty as to whether a legal text
covers a new set of facts because of silence concerning the new facts, generality in the import of a
statement, or conflicts between the legal text at issue and another equally authoritative legal text; and 2
IBN 'ABD AL-SALAM, AL-QAWA'ID AL-KUBRA 279-80 (Nazlh Kam~l Hammdd &'Uthmgn Jumu'a
Dumayriyya eds., 2007), which describes textual ambiguity as texts that suggest both prohibition and
permissibility.
228. There is evidence that temporary marriages continued well into the first century of Islam
in the Meccan school of Ibn 'Abbas and in the proto-ShIFl circles in Medina and Kufa. This rule
continues as a valid form of marriage in Shi'! law today, perhaps drawing from the Meccan school. See
Wilferd Madelung, 'Abd Allah b. 'Abbcs and Shrite Law, in LAW, CHRISTIANITY, AND MODERNISM IN
ISLAMIC SOCIETY: PROCEEDINGS OF THE EIGHTEENTH CONGRESS OF THE UNION EUROPEtENNE DES
ARABISANTS ET ISLAMISANTS 13, 15-16 (Urbain Vermeulen & J.M.F. van Reeth eds., 1998).
229. For a discussion of first-century temporary marriage rulings and reports of 'Umar's
prohibition, see MOTZKI, supra note 104, at 142-46. For the settled Sunni position of prohibition, see,
for example, II MAwARDI, supranote 50, at 449.
invalid marriage would open a couple to accusations of extramarital sex-

where penalties ranged from flogging to death. In view of these stakes, jurists
interpreted the doubt canon to support a presumption that allowed for laws
deemed invalid in one school to absolve criminal liability of anyone accused of
violating that law if it were deemed valid in another school. For them, though
illegal, temporary marriage could not expose parties to criminal charges of
extramarital sex. To do otherwise would unfairly impose criminal liability
where the perpetrator did not have clear notice of the230law in an area where even
the jurists could not agree on the content of the law.
Interpretive doubts were plentiful and only resolved through consistent
interpretation of authentic texts. These sorts of doubts arose even more
abundantly in the realm of prophetic reports than they did with respect to the
Qur'dn. Disagreements about the authenticity and varied wording of the
prophetic reports produced enormously diverse interpretations. Without a
single authoritative corpus of these prophetic reports, each school of law
resolved questions about the definitions of law through varied interpretive
approaches particular to their own schools. As one scholar of Islamic legal
studies has argued, it was arguably "consistency, or the ability to demonstrate
an undifferentiated commitment to a stable and unchanging theory of
interpretation, [that provided] the criterion" to judge the validity and legitimacy
of diverse legal opinions.231
In this process, the prophetic pedigree of the doubt canon did significant
work in framing the debate and turning the doubt doctrine into a widely
accepted legal maxim. Further, it allowed the majority of jurists to easily render
the doubt canon into a core principle of criminal law to govern not only capital
punishment, but ordinary punishment as well. Through the canon, jurists
sought to insert softening principles into the legal corpus. With a textual origin
for the rule, they could assert that the doubt canon had always been part and
parcel of a scheme intentionally designed by the divine Lawgiver, and that it
bound judges who aimed to adhere to the ideal of divine legislative supremacy.
In the ensuing generations, writing in the thirteenth- to fourteenth-century
Mamlilk era, the jurists authored multiple works on legal interpretation
organized around legal canons that set forth policies for particular areas of
law. 232 These collections of legal canons identified the doubt canon as the
guiding principle of Islamic criminal law, for both capital and non-capital
punishments. Importantly, the doubt canon was all-encompassing, covering
substance as well as procedure. Unlike American criminal law, where rules of
reasonable doubt and lenity, for example, form two of many different
principles of criminal law, Islamic doctrines were swept under a single header
of "doubt." The sweep coincided with-and, I argue, served as a mechanism to
230. See, e.g., QARAFI, supra note 25, at 1309 (noting that "only astute jurists can discern
[many debatable Islamic legal] rules" and that even they often disagree).
231. See Sherman A. Jackson, Fiction and Formalism: Toward a FunctionalAnalysis of Usul
Al-Fiqh, in STUDIES IN ISLAMIC LEGAL THEORY, supranote 54, at 178.
232. See supranote 17. These discussions are elaborated in RABB, supranote 25.
bolster-the jurists' institutional claims of the authority to define Islamic law.

This narrative of the canonization, textualization, and generalization of
the Islamic doubt canon rule provides a powerful illustration of the way jurists
defined their own power and placed it in opposition to the policies of a caliphal
executive. The more the politics of power drove political officers to disregard
the jurist-defined moral imperatives, the more concerned the scholarly classes
became with insisting on those imperatives and clarifying the law. 233
Punishment by government officials during this period was "fundamentally
arbitrary, informal, and unpredictable., 234 It is this backdrop of excessive
punishment generally and rampant executions particularly that drove the
attempts of jurists of that time to rein in the excesses through claims of doubt.
The historical record does not allow us to conclude with any certainty whether
jurists were effective in their attempts. Yet, I argue that their interpretive
maneuvers reveal the jurists' interest in defining the precedents and principles
from the authoritative founding period that would clarify the legal requirements
for and against punishment. They did so with a clear sense of their own
authority as the class popularly possessed of the institutional competence to
define issues of law and morality. In the big picture of the premodern Islamic
constitutional structure, the caliphs needed the support of these jurists in order
for the populace to perceive them as legitimate rulers. Thus, even if the jurists
could not directly prevent excessive punishments, they could prevent rulers
from trying to legitimate the punishments under the rubric of Islamic criminal
law.
CONCLUSION
The history of Islamic notions of doubt reveals how central social-

political and institutional contexts were to Muslim jurists' constructions of
Islamic law in this purportedly textualist legal tradition. I have argued that
Muslim jurists developed a textual doctrine of doubt in response to the
eleventh-century political context of excessive state violence that accompanied
the breakup of the empire and weakening of institutional structures for which
they sought to systematize the law. These jurists managed to convert a judicial
practice expressing a preference for recognizing doubt in order to avoid
dubious punishments into afoundationaltext requiring it.
This process puts on display the mechanisms by which Muslim jurists, all
self-proclaimed textualists, assumed multiple institutional roles, hidden under
cover of textualism and deference to divine legislative supremacy. Early in
Islamic history, Muslim jurists had adopted a judicial power to define Islamic
law. They gradually assumed a legislative power to update the law by
introducing new texts in the form of a legal maxim-the doubt canon-that
they claimed had been issued by the Prophet Muhammad. They then deployed
this canon liberally in their efforts to limit executive power over criminal law
233. Mottahedeh, supranote 84, at 6-8.

234. LANGE, supranote 126, at 42.
enforcement or at least define the contours of legitimate punishment. All of

these interpretive strategies responded to the changing institutional contexts in
which the jurists operated and the institutional relations that they helped define.
In short, Muslim jurists used interpretation instrumentally to specify legal
outcomes and to define their own institutional roles.
Further, this process helps demonstrate the significance of the
controversy surrounding the Case of Md'iz. That case represents a radical
tension in Islamic law between Muslims jurists' formal appeal to textualism
and their functional appeal to varied tools of interpretation to define
institutional roles. Muslim jurists operated in a world where they only
implicitly acknowledged that institutional context mattered to definitions of
law. Explicitly, their only legitimate basis of authority-as religious exponents
of law-was to appeal to Islam's foundational legal texts and to the ideal of
divine legislative supremacy. A look at the changing societal and institutional
contexts enables us to uncover the significance of the debates about the Case of
Mj'iz and the subsequent emergence of the doubt doctrine. When factual and
political contexts created morally complex realities, these jurists drew on early
judicial practices to canonize, textualize, and generalize robust doctrines of
doubt.
The growth of the Islamic doctrine of doubt, and its transformation from
a practice-based canon into a textual rule of expansive scope, reveals the close
link between textual interpretation, social context, and institutional legitimacy
in medieval Islamic law and society. In the end, Islamic criminal law was less
about what the "text" said, and more about how Muslim jurists assumed
institutional power to actively construct Islamic law when confronting myriad
contexts about which the text did not say much.

Reasonable Doubt in Islamic Law

Uploaded by

Copyright:

Available Formats

Reasonable Doubt in Islamic Law

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Reasonable Doubt in Islamic Law

Uploaded by

Copyright:

Available Formats

Yale Journal of International Law

"Reasonable Doubt" in Islamic Law

Follow this and additional works at: http://digitalcommons.law.yale.edu/yjil

"Reasonable Doubt" in Islamic Law

IN TRO D U CTION ....................................................................................................................................... 42

1. THE RECURRING CASE OF M'IZ AND QUESTIONS OF ISLAMIC LEGITIMACY ........................... 51

III. THE DEPLOYMENT OF "REASONABLE DOUBT" TO LIMIT EXECUTIVE POWER ......................... 78

C ONCLU SIO N........................................................................................................................................... 93

Legal interpretationtakes place in afield ofpain and death.

Doubt pervades most laws, 2 becomes alarming when it arises in criminal

requires judges to acquit in cases of factual doubt. 7 Medieval English and

doubt. To be sure, the textbook version of Islamic legal theory posits an

reasonable doubt style-whenever evidence of the facts was questionable to

18. See infra Subsection I.A.2.

legislative supremacy. How could Muslim jurists both claim to be textualists

Before discussing the medieval cases and controversies in Islamic legal

Just as Islamic law arises regularly in constitutional politics, Islamic criminal

I. TRE RECURRING CASE OF MA 'IZ AND QUESTIONS OF ISLAMIC LEGITIMACY

governance structures have discerned distinct institutions of law and

A. Early Debates on Approaches to Islamic Criminal Law

1. The Recurring Case of Md'iz

This case sparked tremendous debate amongst Muslim jurists. 50 These

2. The InterpretiveDebates about Md'iz

Muslim jurists devised two interpretive approaches to the Case of M 'iz,

of divine legislative supremacy was represented in the text rather than in

a. The Case for Enforcing the Punishment: Strict Textualist

Strikingly, most Muslim jurists used the Case of Md iz to conclude that

they argued for spirit over text. 66

3. The Significanceof Md'iz

B. The Role ofMuslim Jurists in Defining Structures of Islamic Law and

constraintagainst executive overreach. As I argue below, each of these stages

The earliest Muslims conceived of Islamic law as a unified system of

scholarly competence in articulating norms drawn from their association with

Within half a century of Islam's advent, leadership had gone from a

institutional competence to define Islamic law.°00 In this way, the religious

3. Islamic Law as InstitutionalConstraint

The division between political and religious leadership shaped a

115. FELDMAN, supra note 28, at 2.

competence to interpret Islamic law. Accordingly, the earlier constitutional

What I have elucidated so far is a general theory of Islam's constitutional

II. THE CONVERSION OF "REASONABLE DOUBT" FROM JUDICIAL PRACTICE

punishment set a precedent for punishment avoidance, especially given other

The process of articulating doctrines of doubt first came through juristic

1. Case of the Mysterious Pregnancy

One report hails from eleventh-century Yemen, a generation after M'iz.

2. Case of the Falsely Accused Butcher

Similar evidentiary doubts emerged in reports about an early murder in

Another set of doubts appeared in reports of a third case, which combined

Yet as a statement of judicial practice, the doubt canon would have

"avoid criminal punishments in cases of doubt."'1 54 His prominent disciple, Abu

punishment, but none understood the doubt canon to be a prophetic statement.

Remarkably, after the tenth century, virtually every Muslim jurist

I1. THE DEPLOYMENT OF "REASONABLE DOUBT" TO LIMIT EXECUTIVE

Prophet's personal inclinations toward mercy. Ddw-dd al-ZdhirT and other

A. The Systematization of Islamic Law in a Political Culture of Violence

170. See Mottahedeh, supranote 84, at 6-8.

B. The GeneralizationofDoubt in Response to Excessive Punishment

the backdrop of excessive executions in the Seljliq culture of political violence,

doubt. In this way, the jurists outlined rules of heightened procedures as