Can one who takes Scripture to be the word of God, and who takes their independent moral judgements to be reliable, reconcile such beliefs with Scriptural injunctions that appear to permit and require evil actions? That is the Problem of Divinely Prescribed Evil. An ethicsfirst solution takes our independent moral judgements to be reliable and attempts to reconcile them with seemingly divinely prescribed evil. Amir Saemi (2024) offers a prima facie promising ethics-first solution: take Scriptural injunctions to be not moral, but legal. In this paper, I critically examine this proposal. After raising worries about Saemi’s argument for his solution, I explore his analogy with the ethics and laws of war, raise three concerns for his solution, and present a dilemma which is, ultimately, an argument against Saemi’s solution. I end with some suggestions for further inquiry into this recalcitrant problem, and analytic philosophy of religion about Islam.
Questions of evil bedevil theism. The
Scripture contains
The Qur’an appears to suggest women have a duty to be obedient to their husbands (4: 34), that men may beat their wives if they fear disobedience (4: 34), that inheritance should discriminate against women (4: 11), that women’s testimony is less credible than men’s (2: 282), that thieves and adulterers should face corporal punishment (5: 38; 24: 2), and that slavery (and sex with one’s slaves) is permissible (4: 24).
Some morally controversial passages in the Bible seem to suggest women must obey their husbands as they would God (Ephesians 5: 22–25), that cursing one’s parents (Leviticus 20: 9), adultery (Leviticus 20: 10), and homosexuality (Leviticus 20: 13) are punishable by death, that slavery is permissible (Exodus 21: 2–6; Leviticus 25: 39–55; Ephesians 6: 5; Timothy 6: 1), and that a woman should be stoned to death if her virginity prior to consummating a marriage cannot be proven (Deuteronomy 22: 20–21).
The
(Divinity of Scripture) Scripture is the words of an omnibenevolent, omniscient, and omnipotent God.
(Existence of Seemingly Prescribed Evil) There are some actions prescribed or permitted by the best interpretation of a Scriptural passage which seem immoral, according to our independent moral judgements.
(Reliability of Our Independent Moral Judgements) Our independent moral judgements reliably represent moral values, moral duties, and moral permissions. (
(Divinity of Scripture) and (Existence of Seemingly Prescribed Evil) together appear inconsistent with (Reliability of our Independent Moral Judgements). If Scripture is the word of God, and Scriptural injunctions prescribe or permit immoral acts according to our independent moral judgements, then such judgements look unreliable – since allowing and commanding us to do evil is not something a benevolent God would do.
(Existence of Seemingly Prescribed Evil) and (Reliability of our Independent Moral Judgements) also appear inconsistent with (Divinity of Scripture). If Scriptural injunctions prescribe and permit immoral acts by our independent moral judgements, and if such judgements are reliable, then Scripture looks unlike the word of a benevolent God who would not prescribe and permit immoral acts.
So, either Scripture is not divinely inspired, we misread morally controversial passages, or our moral judgements are unreliable. The problem of divinely prescribed evil is surprisingly underexplored. Or so it was until
In this paper, I critically examine the second, most developed of Saemi’s solutions to the problem. Saemi explores the problem in the context of Islam.
In the next section, I introduce Saemi’s solution. I raise two concerns about Saemi’s argument for his solution in the section “The Argument?”. In “The War (Dis)Analogy”, I explore Saemi’s analogy with the ethics and laws of war that motivates his solution, before raising three concerns for Saemi’s solution in “Three Concerns for (Legal Interpretation)”. I present, in “An Argument Against (Legal Interpretation)?”, an argument against his solution. I conclude with what this all teaches us that provides suggestions for future inquiry.
My prognosis for Saemi’s solution is unfavourable. Regardless, Saemi’s book remains invaluable. He excellently executes his “subsidiary goal” of offering overviews of moral theory and moral epistemology in Islamic thought, provides careful discussion of the problem and possible solutions, and brings contemporary normative and meta-ethics to bear upon the philosophy of religion. I offer my criticisms in the spirit of collective inquiry, advancing together towards resolving this recalcitrant problem – Saemi’s place in that inquiry is beyond reproach.
There are two general strategies for a Muslim – or any other theist – to reconcile our inconsistent triad:
(Scripture-first) Our independent moral judgements are unreliable in the face of Scriptural injunctions.
(Ethics-first) Our independent moral judgements are reliable and can be reconciled with seemingly prescribed evil.
Arguably dominant in Islamic history, (Scripture-first) views suggest that, taking Scripture at face value, our independent moral judgements are unreliable. Saemi’s discussion of these views is rich and deeply informative. My focus, however, is on his (Ethics-first) responses.
Saemi’s first (Ethics-first) solution involves suspending judgement on the reliability of our independent moral judgements and arguing it can still be “rational” for us to act upon them. I’ll focus on his second solution. This is because, firstly, it is the most developed. Secondly, Saemi’s initial response is arguably not a
Saemi’s second (Ethics-first) solution takes our independent moral judgements to be reliable. I’ll grant this. I want to ask whether Saemi’s solution plausibly resolves our initial tension
Saemi argues that the appearance of a tension is generated by assuming:
(Moral Interpretation) Actions prescribed or permitted by the best interpretation of a Scriptural passage are morally obligated or permitted.
Saemi brilliantly notices that “God or the Prophet uses
Instead, Saemi suggests that we adopt:
(Legal Interpretation) Actions prescribed or permitted by the best interpretation of a Scriptural passage are
The suggestion:
I’ll follow Saemi in assuming Scott
To illustrate this,
Revisionary just war theorists argue that the Geneva Conventions, which legally regulate the conduct of war, are thus morally mistaken: they allow killing of combatants on either side and prohibit killing non-combatants. Still, since it is often difficult to establish the just and unjust sides of a war, and to establish which noncombatants may be liable to harm, if we want to “minimise atrocities” and “avoid…possibly unresolvable” disputes, it can be justified to legislate morally suboptimal laws (
This, Saemi maintains, may have been the situation at the time of revelation. It may have been infeasible in seventh century Arabia to solve their social problems by legislating the morally correct laws. It may have even made the situation worse. The morally correct laws, we assume, would be more egalitarian. But such moral principles, if made legal laws, would deviate substantially from those socially possible for that society (for example, banning slavery). If socially impossible then, God may have been justified in legislating morally sub-optimal laws in Scripture, provided “they…presented the best available plan to solve some social problems of the newly formed [Islamic] society”. (
The moral language in Scripture can be read as
Saemi’s ingenious solution offers a
A preliminary note:
Saemi provides an argument for (Legal Interpretation), which I reconstruct here:
(P1) Scriptural injunctions were the proposed laws of Islamic society at the time of revelation.
(P2) Scriptural injunctions represented the Prophet’s planning to solve the social problems of early Islamic society. [Planning Conception of Law]
(P3) Sometimes, it is rationally necessary to legislate immoral laws to solve social problems; it can be rationally required to have immoral laws. [Feasibility Constraint]
(P4) Early Islamic society was one where it was rationally necessary to legislate immoral laws to solve social problems.
(P5) If (P4), the Prophet would have been, in principle, justified to legislate sub-optimal moral laws, through Scripture, to solve said social problems.
(C) “Therefore, Legal Interpretation is true…while actions prescribed or permitted by the best interpretation of a Scriptural passage are
I have two worries about this argument.
First, this is not a valid deductive argument for (Legal Interpretation); (P1)–(P5) do not entail
Second, (P2) begs-the-question against (Moral Interpretation). The reasoning given from (P1) to (P2) is: “at the time of revelation, Scriptural injunctions were the laws of the society. Therefore, given
Absent a valid, non-question-begging argument for (Legal Interpretation), the maximally charitable move is this: Saemi offers (Legal Interpretation) as the
Our question becomes:
Saemi draws his analogy with the ethics and laws of war to illustrate how true moral principles, and justified governing laws in non-ideal conditions, could come apart.
Even if we assumed (Legal Interpretation), however, there are difficult questions that comparing the ethics and laws of war with Scriptural injunctions and moral truths raises for Saemi. They also highlight pressure points for Saemi’s view that I will press in the follow sections.
Saemi assumes that our rational goal when considering possibly legislating morally sub-optimal laws is to “minimise the atrocities [and] avoid [possibly irresolvable] disputes”.
Herein lies a disanalogy. For in the law of war,
So, two questions: first, why assume that God or the Prophet had the aforementioned goal? We might appeal, as Saemi does, to “God’s justice”. But wouldn’t a just God have provided us the moral truths and avoided ambiguous legislation? Muslims also believe that the Prophet’s revelation was
Next, justified laws may be morally sub-optimal for different reasons. Some may be because of contingent and more easily changeable features of a society or, say, because of the nature of some activity. Whilst subject to contingent variation, war will likely always involve the “fog of war” that makes it hard to tell, say, who the just side is, or which non-combatants are liable to harm. The Geneva Conventions may be the best laws of war partly given this.
Saemi is more optimistic about the prospects of alignment of other laws with correct moral principles over time. He notes: “there might always be a distance between correct moral principles and proper legal laws, as…the Geneva Conventions shows. But as time passes, the legal laws should come closer to correct moral principles.” (
This assumes that the “obstacle” to bringing legal laws closer to morality is social feasibility and something to do with, as Saemi puts it, our “moral consciousness” expanding over time to include others within equal moral concern. But perhaps
What, then, can Saemi say to the Muslim who maintains that
The worry: the analogy with war may be
Third, distinguish between two ways in which a law can be
A law can also
A final disanalogy, then: if revisionary just war theorists are right, the Geneva Conventions only
Suppose one can be morally or rationally justified in legislating morally sub-optimal laws when their sub-optimality (and immorality) is exhausted by their
The problem: there appears to be a
But an omnibenevolent, omnipotent, and omniscient God prescribing legal injunctions that
Did the war analogy not show how this could be morally justified? But even eschewing concerns about that analogy, this objection is about the apparent tension between our initial triad
To block this revenge worry, we need an account of why God would legally require gravely immoral conduct which explains away, or otherwise addresses, the thought that God would not require gravely immoral conduct of us – morally
Lara Buchak, however, may have the goods. In her contribution in this issue, Buchak argues an essential feature of authority is to allow people to learn things they may not learn unaided. This can be done through laws that “provide
Take slavery. Buchak maintains that God could not have successfully banned slavery at the time of revelation since this would be socially impossible. But he could have, and perhaps did, command us to do “less slavery than previously” such that, over time, we come to learn through experience that it is better to end slavery. This is allegedly preferable to just telling us to stop. For society then would likely not follow that prohibition, and without “sampling”
Buchak’s paper nicely illustrates how this suggestion has several attractive features for a proponent of (Legal Interpretation). But it does not defang the revenge objection.
First, notice the
Suppose someone kills homosexuals for being homosexual and that I want to teach them the wrongness of doing so. I do this by telling them to do it less than they previously have but still allow them to stone homosexuals for this reason. Suppose I then adopted this as the method of teaching a society the wrongness of stoning homosexuals for thousands of years, in the hope that, eventually, they will realise it is wrong and stop. This seems a seriously morally wrong method of moral teaching.
Second, one might insist that Buchak’s reinforcement moral learning account was the
If one insists that a benevolent God would do this, then we need an explanation why which does not restrict God’s omnipotence or omniscience, lest we surrender God’s benevolence. But suppose an answer was forthcoming that did not threaten God’s properties. Why, then, would it have been inconsistent with such properties to have, for example, just given us individual and communal exercises of deep, immersive moral imagination instead?
The revenge worry, then, appears unscathed by Buchak’s proposal. This is a serious problem for Saemi’s solution.
My second objection concerns Saemi’s claims about what moral function remains for Scripture if (Legal Interpretation) is true. The worry is that if our independent moral judgements are reliable and help inform us that Scriptural injunctions are not moral ones, then “it is not clear whether Scripture has any moral function”. (
Saemi understands that worry as similar to an old objection to prophecy from Barahima about the relationship between reason and revelation. That relationship is a central concern of much Classical Islamic philosophy.
But Saemi misconstrues this old objection. The historical concern was identifying the epistemically
It will not suffice, then, to find
The first response, modelled on al-Fārābī and Averroës, is that “all theoretical and practical truths are discoverable by reason alone” but that Scripture can provide epistemic and motivational assistance. This, even if plausible, will not suffice. For, again, we want the unique role of Scripture here.
The second response is adapted from al-Ghāzāli and ‘Abd al-Jabbār. It maintains that “there are moral truths that don’t disagree with reason but which cannot be discovered through reason alone” and “Scripture may expand our moral consciousness by teaching us about them”. Saemi puts this as Scripture “extending” our moral obligations.
But if these obligations are, he notes, “recognisable by reason”, then why can’t I learn them without Scripture – even if with difficulty? There may be some specific things we would learn only through Scripture (prayer rules, say). But this is not much by way of
Saemi could respond that the epistemically unique role of prophecy and revelation is to be found not in moral epistemology but elsewhere. And it would suffice to save Scripture from having
If (Legal Interpretation) is true, how can we determine which claims of Scripture should be read as moral and which legal – given that both employ moral language?
Saemi cannot appeal to the distinction between the Meccan/Medinan verse distinction; those verses in the Qur’an before and after the Prophet’s migration to Medina (
Our independent moral judgements, Saemi claims, can “help inform us that Scriptural injunctions are not moral ones”. If so, that suggests a test:
(Ethics-First Test) If a Scriptural injunction, read as a moral truth, conflicts with our independent moral judgements, then that is strong epistemic reason to think that it is a legal injunction. If there is no conflict, this is strong epistemic reason to think it is a moral injunction.
Here are two worries about this test.
First, it is a
Saemi might consider that unfair, since we are granting the reliability of our independent moral judgements. But their reliability does not entail their
Regardless, a second worry. The test seems to suggest that, after “careful
But if this entails that (empirically-informed) reason can determine which injunctions are moral and which legal, then why think that Scripture provides us some
In the quote above, Saemi did say “help”, though, and not that the
This problem is serious because a resolution is necessary for
I now present a dilemma, which generates an argument against (Legal Interpretation) in the Islamic context:
(P1) If God exists, and if the Prophet is his final messenger, then, in
(P2) If (Legal Interpretation) is true, then God, in the final revelation, used ambiguous moral language to convey such morally sub-optimal legal injunctions.
(P3) So, if God exists and the Prophet is his final messenger, then (Legal Interpretation) is false. [From (P1) and (P2)]
(P4) So, if (Legal Interpretation) is true, then either God does not exist, or the Prophet is not his final messenger. [From (P3)]
(C) So, either (Legal Interpretation) is false, or it is false that God exists and the Prophet is his final messenger. [From (P3) and (P4)]
To explain, the final revelation constitutes a
(P1) claims that if God exists and the Prophet was his final messenger, then in the final revelation, God would not have used ambiguous moral language – in such a morally high-stakes situation – to convey merely legal injunctions that permit
(P1) and (P2) entail that if God exists and the Prophet is his final messenger, then (Legal Interpretation) is false – (P3). For (Legal Interpretation) entails God would have done something, assuming (P1), that is inconsistent with God’s existence and the Prophet’s revelation being final. Contraposition of (P3) gives us: if (Legal Interpretation) is true, then either God does not exist or the Prophet is not his final messenger, (P4).
A dilemma follows: either (Legal Interpretation) is false, or it is false that God exists and the Prophet is his final messenger.
This dilemma is troubling for any Muslim attracted to (Legal Interpretation). Muslims cannot reject [God exists and the Prophet is his final messenger]. So, Saemi – and any other Muslim – is forced to accept that (Legal Interpretation) is false.
(P3) and (P4) are trivial consequences of other claims, and (P2) uncontroversially follows from (Legal Interpretation). So, the action is in (P1).
Here are three reasons to accept (P1).
First, consider what would be practically rational when legislating morally suboptimal laws. Assume, following Saemi, that a rational goal in morally sub-optimal legislation is “minimise atrocities” and “avoid possibly irresolvable disputes”. If so, then it would be practically irrational for God to have used ambiguous moral language
For, firstly, Scriptural injunctions can (and have!) been read as morally permitting and requiring actions that constitute – by our judgements – moral atrocities. So, the aim of “minimise atrocities” would be unachieved. Secondly, using ambiguous moral language in the final revelation
Second, it appears inconsistent with our independent moral judgements that an allloving and all-knowing deity would, in the final revelation, through ambiguous moral language,
To see my third reason: one might reject (P1) because they take God to have equipped us to settle the question of legal or moral interpretation. But even if God did that, (P1) remains plausible. We can justify (P1) by justifying its logically equivalent contraposition:
(P1*) If God used
Here is an argument for (P1*). If God is omniscient, then God could have reliably predicted that the issue of interpreting ambiguous moral language in the final revelation would arise, risked dire moral consequences, and would likely be misread. Allowing this very morally risky interpretative issue
One
(Sceptical Theism) We do not have epistemic access to God’s actual or possible moral, prudential, or other normative reasons for or against doing any action, or for allowing any action or event to occur.
The objection is that (P1) falsely assumes that the rational goals
But (Sceptical Theism) would not reveal (P1) is false. For (Sceptical Theism) would not entail the
Appeal to (Sceptical Theism), however, generates a serious problem. For it
Perhaps as
Still, (Sceptical Theism) seems to generate an impasse. Either (Sceptical Theism) is false or it isn’t. If it is, I have provided reasons to accept (P1) and reject (Legal Interpretation) on pain of relinquishing core Islamic commitments. If (Sceptical Theism) is true, then (P1)
A second response. Perhaps (P1)
This is
Interestingly, it is difficult to assess this response
Islamic scholarship has, of course, addressed the finality of the Prophet. Such accounts are offered within complex theories of religion that take us too far afield. But, briefly, here is why appeal to these won’t easily work. These theorists seek to explain Islamic commitment to the finality of the Prophet
It is, sadly, not prosecutable here. Providing a comprehensive account of the Prophet’s finality should be a central task for philosophy of religion about Islam, as providing accounts of core Christian commitments – like (for most) the Trinity – is in philosophy of religion about Christianity. Needing an account of finality to provide this second response to my argument only strengthens that case.
Until then, I tentatively conclude that, on minimal Islamic assumptions, (Legal Interpretation) is false.
This may seem a bleak place to end. To think that would be to miss lessons learnt through engaging with Saemi’s excellent book.
First, when adopting a solution to the problem of divinely prescribed evil, one must be alive to features of the religion co-opting that solution. Saemi, of course, realises this. But these features can complicate matters in unobvious ways, even with minimal religious commitments. That is what I have suggested with respect to the finality of the Prophet’s revelation in Islam for Saemi’s solution.
Second, any plausible account of revelation in Islam must explain what the finality of the Prophet’s revelation amounts to. That would help adjudicate the argument in the previous section
Third, the consequences of (Sceptical Theism) for the problem of divinely prescribed evil require attention. I’ve argued it generates a disappointing impasse over the argument in the previous section. But if (Sceptical Theism) is true, then perhaps we are bound to only play in the realm of logical possibility. If we wanted to go further and make probabilistic judgements, on what evidence can this be done if (Sceptical Theism) is true?
Finally, some of what is needed to establish whether (Legal Interpretation) is the least costly response to our problem is now clearer. Namely, an account of the rationality and moral permissibility of legislating morally sub-optimal legal laws that require committing grave moral wrongs (that blocks the revenge worry), demonstrating how appeal to (Sceptical Theism) does not undermine any positive case for (Legal Interpretation), and an explanation of the finality of the Prophet’s revelation consistent with the moral sub-optimality of the laws provided in the final revelation.
Examples of Scripture include, at least, the Qur’an and Hadith (accepted-as-authentic accounts of the behavior and sayings of, at least, the Prophet) in Islam, the New and Old Testament in Christianity, and the Hebrew Bible in Judaism.
By ”independent moral judgements”,
For related work, see
By “Islam”, I mean the religion according to which, at least, there exists exactly one God, who is omnipotent, omnibenevolent, and omniscient, and that Muhammad ibn Abdullah (PBUH) was the final Prophet of God (the “seal of the prophets” or
Two other options: linguistic reinterpretation, or historical contextualisation, of morally controversial passages (
A point also made, as
For representatives, see
Saemi also does not state his background meta-ethical commitments. They seem broadly realist. This silence matters more than it may appear. Different realist views have varying consequences for moral epistemology, or for what moral facts could be hidden from us. And there are broadly “realist” views whose commitments radically differ from what many think “realists” must accept; see my
In the workshop for this symposium (21/02/25), Saemi confirmed this is how he is best read.
I return to this in the last major section.
A benevolent God would presumably legislate the
Notice this is a problem for
At the symposium workshop, Buchak acknowledged this cost.
I am here following
As Mohammed Saleh Zarepour reminds me, it also seems inconsistent with the famous Hadith Makarim al-Akhlaq: “
He also thinks unhelpful
Interestingly, given his solution is a “legal” one, Saemi does not engage with Islamic jurisprudence. On Saemi’s view, do jurisprudents become secular normative ethicists, social planners, and sociologists, trying to ascertain the morally best socially feasible plans? I suspect many Islamic jurisprudents would demur, if so.
I suggested in our symposium workshop that it seems theologically undesirable for any Muslim to adopt a solution that suggests God justifiably prescribes evil, renders it mysterious why God sent his final message when we were socially unprepared to follow the morally best laws, and does not rule out continual, morally-ever-improving revelation. These seem amongst the ‘costs’ of (Legal Interpretation).
When presenting the revenge objection, I argued against the best available account of why God would have, namely Lara Buchak’s.
Sceptical theism is a family of views; for discussion, see
My thanks to Mohammed Saleh Zarepour for this.
As
See, e.g.,
I am extremely grateful to my co-symposiasts, the symposium workshop audience, and especially to Lara Buchak, Paul O’Grady, Amir Saemi, and Mohammed Saleh Zarepour for excellent written comments (and to the latter for the kind invitation to contribute).