True ‘contradictions’ and conflicts in the Talmud

Abstract

In this article, we consider the presence of what appear to be true contradictions in the Talmud. We consider and reject a glut-theoretic response. We argue these apparently contradictory Talmudic passages should be understood not as ordinary propositions, but as being given under an operator. This allows us to understand these rulings as genuinely conflicting, but not genuinely contradictory. We illustrate the broad shape of such a view through consideration of Hans Kelsen's late-period philosophy of law. We also consider and reject responses to this issue given by Rashi, R. Kook, and R. Feinstein. We close by considering why a system of Jewish law which allows for jointly affirmed conflicting propositions is desirable. We claim that if, as Maimonides suggests, the fundamental project of Jewish law is the eradication of idolatry, then the law itself should remind us that God's ways are higher than our ways. One way of doing so is to resist capture by a transparent set of principles, by allowing for conflicting rulings to be affirmed. This article does not presuppose familiarity with the Talmudic corpus or Jewish tradition.

Introduction

The Talmud contains – among other things – disputes about the understanding, derivation, and application of Jewish law.¹ Such disputes are resolved by various means, including (but not limited to) appeals to hermeneutic principles, principles of inference, and principles of authority.² Standardly, disputes are resolved by some combination of these strategies. The desire to resolve such disputes stems in part from a practical concern. Much of the Talmud seeks to develop a framework of use in determining what should be done – ethically, religiously, or civically – in various kinds of situations. But it also stems from a concern with truth.

No matter how innovative and abstruse halakhic enquiry becomes, it is seen as part of, not an addition to, the original revelation received by Moses. (Moses is said to have received both the written and oral Torah).³ This principle is laid out explicitly in the Yerushalmi version of the Talmud, tractate Peah 2:4:

Rebbi Joshua ben Levi said: ‘On them, and on them; all, like all; words, the words; Bible, Mishnah, Guemara, and Aggadah. Even what a competent student will discover before his teacher was said to Moses on Sinai. What is the reason? [And even if] there is something about which one would say, Look, this is new!, his colleague will answer, It already has been forever.’

R. ben Levi's opening reference to ‘On them and on them; all, like all; words, the words’ is an allusion to Deuteronomy 9:10, and the textual clues used to prove that the Oral Law was given at the same time as, and with the same authority as, the written Torah. The implicit view is that revelation is continually growing and developing without changing; that the elaboration of the law is, in fact, a statement of the law as it always was. The ontology presupposed by such a view is intriguing, but for our purposes we need only note that the results of Talmudic disputation have, in principle, a truth value. There are rulings which are part of God's revelation of the Oral Law, and rulings which are not. Accordingly, opinions which are discounted in the Talmud are false assertions about the contents of the Oral Law; and those affirmed are true assertions about its contents. These rulings and opinions are treated as belonging to one of these kinds (false assertions, or true assertions) exclusively – in other words, false opinions are not simultaneously counted as true, and vice versa. Talmudic disputation, then, usually adheres to the Law of Non-Contradiction (henceforth ‘LNC’).

But not always. The Talmud contains apparent contradictions which are claimed to be true. Apparently-contradictory propositions are juxtaposed, explicitly recognized as in conflict, and affirmed as both true. This article will offer an account of how we should understand these apparent true contradictions in the Talmud.⁴

We will need to discuss halakhic rulings and texts, and rulings and texts from secular law. For convenience, we will use the term ‘legal’ freely across both domains. We might, for example, talk about general features of ‘legal norms’ in both halakhic and secular contexts. This is not intended to imply that all the content of the Talmud, or indeed halakhic law, can be captured by the standard concept of legality. But there is sufficient overlap when talking about the broad structural issues covered by this article.

Contradictions in the Talmud

There are contradictions in the Talmud which are accepted as unresolved rather than affirmed as true. These arise most frequently where an argument ends aporetically. Such aporias are noted and left unresolved, but there is no suggestion that this is desirable (a result of perfect disputation) or accurate (a reflection of the genuine nature of the Oral Law). Rather, the aporia is seen as caused by insufficient information; perhaps about the meaning of the terms employed, or about the intentions which stood behind the transmitted saying, and so on. Contradictions of this kind are not relevant to our discussion.

We are interested in a smaller subclass of Talmudic contradictions, where the conflicting opinions are both affirmed as true, in full knowledge of their conflicting status.⁵ A passage in Eruvin 13b has become something of a touchstone in relation to this issue –

Rabbi Abba said that Shmuel said: ‘For three years, the House of Shammai and the House of Hillel disagreed. These said: The halakhah is in accordance with our opinion, and these said: The halakhah is in accordance with our opinion. A Divine Voice emerged and proclaimed: These and those are the words of the living God. However, the halakhah is in accordance with the opinion of the House of Hillel’. Since both these and those are the words of the living God, why was the House of Hillel privileged to have the halakhah established in accordance with their opinion?

We will find this idea revisited and reinterpreted in some of the authorities we will canvas further below. The apparent existence of true contradictions raises a family of questions and problems, which attach to our understanding of the practice of Talmudic disputation and its internal norms; and to the object of Talmudic disputation and its ontology. If Talmudic disputation has both a practical and theological commitment to aim for the resolution of contradicting opinions, why are apparent contradictions like these affirmed as true? On the other hand, if accepted Talmudic rulings are features of the revealed Oral Law, what are the ontological implications for that revealed law if contradictory rulings are jointly accepted as true?

We can generate a dilemma from here. On the one hand, we might say that the Oral Law contains gluts – that is, it contains true sentences whose negations are also true. While this would help resolve anxiety about the object of Talmudic disputation, it would only deepen the problems with our account of the practice of Talmudic disputation. In particular, we would run into the problem of how we could limit the Oral Law's use of gluts in a principled way, so as to explain the apparent preference for harmonization of disputes, and application of the law of non-contradiction, elsewhere in the Talmud. (We could consider this a recherché version of the issue with controlling explosion in paraconsistent logics more broadly.) On the other hand, if we say the Oral Law is not best understood glut-theoretically, then while the object of Talmudic disputation is not worrisome, the practice of Talmudic disputation simply appears mysterious, as the affirmation of apparent contradictions seems unmotivated.

It seems we need to revise our understanding either of what Talmudic disputation is – perhaps it is not aimed at truth or consensus after all – or of what the object of Talmudic disputation is – perhaps the Oral Law includes gluts – to overcome this puzzle. But in what follows, we will argue against making either revision. We will argue that we can avoid making either of these revisions by reconsidering the nature of revelation. We will claim that the revelation of the Law takes place both through direct transmission, and generation via accepted institutions and norms. It is the products of the latter which can comfortably admit the affirmation of conflicting statements, and not the products of the former.

Dialectical norms in the Talmud

Here is an objection: ‘Your anxiety about such contradictions is anachronistic. The Talmud does not claim to be a compendium of philosophical dialogues and we should not assume that it exhibits sensitivities to philosophical concerns. We, as philosophers, are troubled by true contradictions – but the Talmud itself is not.’

To head off this objection, it would be sufficient to show that the Talmud, on its terms, recognizes contradictions as fundamentally undesirable. And it clearly does. Countless discussions begin with the locution ‘[Rabbi X] raises a contradiction’. Such discussions characteristically end either with the formulaic assurance that ‘this is not difficult’ followed by a resolution of the contradiction,⁶ or the statement that the issue ‘remains difficult’ and the contradiction stands.⁷ Furthermore, where more than one account of the contradiction being raised is available – perhaps because it is unclear exactly what problem the traditional source is raising – the account which renders the contradiction ‘not difficult’ is standardly preferred to that which renders it ‘difficult’.⁸

Indeed, a preferred tactic in rabbinic literature is to see contradictory statements as not, in fact, truly contradictory but tracking subtle differences in meaning or application. For example, in his commentary on Ketubot 57a, Rashi writes:

It is necessary to say, ‘These and those are the words of the Living God; sometimes this logic pertains and sometimes that logic pertains, because the logic changes even on the basis of a small change in the issues at hand’.⁹

The Talmud is an explicitly dialectical text which has a more or less clearly articulated set of dialectical norms to which it adheres. Chief among these is a commitment to dissolve apparent contradictions within the body of legal rulings and opinions it collates. We must now consider how to reconcile these dialectical norms with the apparent presence of true contradictions. Perhaps the most direct route would be to opt for a glut-theoretic understanding of the Oral Law. We will now consider this.

Theological motivations for a glut-theoretic approach

There is both historical and contemporary precedent for appealing to glut-theoretic accounts to smooth over philosophical difficulties raised by religious beliefs and practices. We will focus on a contemporary precedent in Christian philosophy, set by the work of Jc Beall.¹⁰ The specific philosophical difficulty which motivates Beall is the Christological notion of hypostatic union.¹¹ This is the claim – made doctrine by the council of Chalcedon – that Jesus has both a completely human and a completely divine nature. Here is Beall's thumbnail of the essential structure of the problem –

[P]1. Christ is immutable (in virtue of Christ's divine nature).

[P]2. Christ is mutable (in virtue of Christ's human nature).

[P]3. Therefore, Christ is both mutable and not mutable (via logic).

This argument relies not only on the familiar (but suppressed) entailment from x is immutable to x is not mutable, but also on the familiar ‘transference principle’ (in effect, a generalized ‘communicatio idiomatum’ doctrine) that

P0. Something that has (or exemplifies) a nature N has whatever properties are entailed by having nature N.¹²

In classical logic, the affirmation of a contradiction makes valid the derivation of any proposition/s whatsoever. This is the problem of explosion we alluded to earlier. For example, the following is a valid inference:

If (Jesus is immutable) & (Jesus is mutable), then (God does not exist).

We can plug anything we like into the final set of brackets and have it come out as valid. This is clearly a bad result on both logical and theological grounds. It is a bad result logically as it leads to triviality and incoherence- every possible sentence in our language can be validly derived from the stated contradiction. It is a bad result theologically, because heretical and orthodox statements are derivable from affirmation of hypostatic union. (A believer who accepted Beall's account of hypostatic union and employed classical logic would be obliged to see the above argument as not only valid but sound. Beall (2023, 8) explicitly acknowledges this, and cites it as motivation to move away from a classical account of logical consequence.)

Paraconsistent logics stipulate only that for some class or set of propositions, contradictions do not lead to explosion. A glut-theoretic account is the more robust position that not only are some contradictions not explosive, but some of these non-explosive contradictions are true. (It is not the acceptance of all contradictions as true. Indeed, it had better not be, on pain of collapsing into trivialism.¹³) Adoption of this approach saves Beall from having to make the theologically unpalatable choice to reject P1 or P2. He can class the doctrine of hypostatic union as one of the privileged class of gluts, non-explosive contradictions.¹⁴ A similar strategy is employed in response to contradictions stemming from identity claims proper to trinitarian theology.¹⁵ Why shouldn't we adopt a similar strategy when confronted with true contradictions in the Talmud?

Rejecting a glut-theoretic approach

There is wealth of literature attacking glut-theoretic accounts, to which we do not intend to contribute. We instead want to consider what the characteristic reasons for adopting a glut-theoretic account are, and whether they have any purchase in the Talmudic context.

The adoption of gluts is a complex and contested position which should be adopted, if at all, only under dialectical pressure. This is the characterization offered by its own proponents, who urge the adoption of gluts precisely because of their capacity to dissolve dialectical perplexities (the ‘Liar paradox’, for example). The dialectical challenges of glut-theoretic approaches – the most pressing being the need to form an acceptable rule for identifying contradictions to be treated as gluts versus those to be treated conventionally – are recognized by its proponents, and understood as part of the cost/benefit analysis that drives the glut-theorist.¹⁶

Accordingly, this article does not require a novel or sustained argument against the adoption of glut-theoretic accounts. It needs only to show that we are not subject to the kind of dialectical pressure that could render the employment of gluts appealing. We will see that the costs that the employment of gluts is designed to defray do not emerge in the Talmudic context. Differences between the Christological case and the Talmudic case will prove decisive.

True conflicts without contradiction

The religious differences between Christianity and Judaism are not directly relevant to our discussion. What is relevant are the semantic differences that follow from those religious differences. Adoption of a glut-theoretic approach about Jesus' hypostatic union, for example, is driven by a flat contradiction between ordinary propositions. We can crudely formalize the structure of the Christological problem as the need to affirm the contradictory assertion that:

A & ~A

We are not, however, obliged to understand the true contradictions which the Talmud contains as being ordinary propositions. While ordinary propositions generate contradictions when both they and their negations are asserted, this need not be true of propositions given under an operator. For example, the following assertion is not contradictory:

(Possibly)A & (Possibly)~A

If we understand the Talmud's contents as formalizable under an operator in this way, we will have the beginning of an account which can nullify challenges to our standard understanding of the practice and object of halakhah. The use of sentential operators has been explored by Beall and Henderson (2019) as a non-glut-theoretic option in responding to the Christological contradiction outlined above. In brief, the proposal is that the claim ‘Jesus is immutable’ is understood as coming under a sentential operator ‘According to the divine story’, and ‘Jesus is mutable’ is understood as coming under the sentential operator ‘According to the human story’ (Beall and Henderson 2019, 160). These ‘stories’ are explanations produced by and proper to the relevant nature. Jesus (uniquely) possesses two natures, which are non-unionizable. Thus, both opposing claims are true (according to the natures under which they are to be understood), and true of the same thing – namely, Jesus. But as the two natures – and their explanatory contents – are not unified, we cannot infer the contradiction ‘Jesus is mutable and immutable’ as there is no single nature, or covering explanatory standpoint, from which such a statement can be produced.¹⁷ Beall and Henderson are careful to block inference from ‘According to the x story, S is F’ to ‘S is F’ (Beall and Henderson 2019, 161−162). They do so by pointing out that their operator is an ‘according to x’ operator, and not an ‘in virtue of x’ operator. The latter generates such an inference; the former does not. In the language of Beall (2017, 204), the ‘according to x’ operator does not ‘release’:

We say that O releases – or exhibits release behaviour – iff for any sentence φ, there is no relevant possibility in which O(φ) is true but φ untrue.

We will also seek to handle apparent Talmudic contradictions through the use of sentential operators. But what operator would be appropriate in the Talmud's case?

Our view is that the Talmud contains claims about halakhah produced by rabbinical authorities employing accepted forms of interpretation. And so, we recommend the operator ‘(Via a process of halakhic disputation)’, or ℍ. By employing this operator, we can formalize the ‘true contradictions’ in the Talmud as follows:

ℍA & ℍ~A¹⁸

Understood in this way, the ‘true contradictions’ in the Talmud are not, in fact, contradictions. But we do not obtain this result by denying that the propositions in question really conflict (as Rashi does above), nor do we deny that they are both affirmed as true (as other authorities do, as shown below). On our view, which we will develop further below, conflicting rulings in the Talmud genuinely do endorse, recommend, or demand beliefs or courses of action which generate deep tension in being jointly held or executed. But these are conflicts rather than logical contradictions, as such rulings are not ordinary propositions. (From this point, we will refer to the passages in the Talmud where conflicting rulings are jointly affirmed as ‘true conflicts’.)

ℍ is an ‘according to x’ operator, and not an ‘in virtue of x’ operator. (ℍ can be rendered to make this clearer – ‘According to the Oral Law’, for example.) Thus, ℍ does not release; ‘ℍ x & ℍ ~x’ does not entail the bare statement ‘x&~x’. Accordingly, the Oral Law's conflicting rulings offer no motivation towards a glut-theoretic approach.

It is useful to note here that Beall (2017, 205–206) argues that the only kind of ‘According to law’ operator which does plausibly release, and would thereby potentially motivate the adoption of gluts where apparent contradictions were found, is ‘According to God's Law’. In the Jewish context, this is not quite so simple. In Judaism, divine revelation can take place directly or institutionally, via the Written Law and Oral Law, respectively. ‘God's law’ can, in a Christian context, be captured by a single operator; the Jewish context requires two.

The Oral Law, and claims about it, come under the ℍ operator, which does not plausibly release. The written law – the product of direct revelation – is different, and requires a different operator. Let ${T}$ stand for ‘According to the Written Law’. This is an operator which does plausibly release, just due to the status of scripture itself, as a record of God's will and actions. It is core to the meaning of the Torah that from –

${T}$‘God created the world’

we can immediately infer:

‘God created the world’.

This will be revisited further below.

By exploiting the semantic difference between Christology and halakhah, we will be able to account for the true conflicts in the Oral Law without opting for gluts, and without losing the claim that the Talmud's claims are truth-evaluable. Before we develop this account and consider objections to it, we will first look at how rabbinical sources, traditional and contemporary, have sought to handle the apparent existence of true contradictions in the Talmud.

Rabbinical views

Rabbinic authorities largely share an aversion to seeing true Talmudic contradictions as genuine rather than apparent. We will look at three differing approaches.¹⁹ The first we have already seen above, when we looked at Rashi's commentary on Ketubot. In that case, Rashi claimed that apparent contradictions trace subtle differences across what merely seemed to be identical cases – as such for Rashi there is no genuine contradiction or conflict in such cases. The second approach,R. Kook's, can be seen as a slightly more elaborate version of Rashi's. He writes:

The multiplicity of views which arise as a result of the differences of personalities and educational backgrounds is specifically the fact which most enriches and broadens wisdom. Ultimately, all issues will be properly understood, and it will be recognized that it would have been impossible for the Building of Peace [the Third Temple] to have been erected had it not been for all those influences which appeared to have been in conflict.²⁰

In Kook's view, the contradicting statements are only apparently contradictory. While for Rashi, the apparently contradictory propositions were, in fact, visibly distinct and non-contradictory to the careful reader, Kook's view is that these statements appear to us to be contradictory, but this appearance will be dispelled at some later, possibly Messianic, point in time.

R. Feinstein takes a third position:

And this is the matter concerning all of the disputes of our earlier and later sages, when one forbade and the other permitted: whenever the [law] was not resolved in accordance with one [of them], any [rabbi] could rule in his locality in accordance with his logic, even though the true law is only in accordance with one of them . . . Both are the words of the Living God, but the real truth, revealed before heaven, is only like one of them.²¹

For R. Feinstein, contradicting halakhic rulings are justified by virtue of having been produced by (or ‘according to . . .’) the halakhic process. This shares much in common with the account we develop in this article. However, R. Feinstein further posits that these conflicting halakhic rulings are overruled by a (presently concealed) ruling – ‘the real truth, revealed before heaven’. The ‘according to’ approach, and its tolerance of conflict, is not held to be the essence of the Oral Law, but a product of the temporary concealment of the (conflict-free) ruling by God.

All three of these approaches share the assumption that the following exclusive disjunction obtains for any part of the Oral Law; either that part of the Oral Law is non-contradictory/non-conflicting and objective or it is genuinely contradictory/conflicting and hence either merely conventional, or due to be overruled by a higher, non-contradictory ruling. We deny this exclusive disjunction. For us, the true conflicts in the Talmud are non-contradictory, conflicting, objective, and true. In developing this position, we will draw on a distinction between direct and institutional revelation.

Direct revelation and contradiction

As is well known, the traditional view is that 613 commandments were given to Moses at Sinai, and transmitted through the written Torah Moses also received. As such, the written Torah is directly revealed from God, and claims about it fall under the ${T}$ operator. If we were to find contradictions there, we would be pressed either to say that scripture is corrupt, that it has esoteric layers of meaning, or that God's word includes gluts.²² This is because ${T}$ releases, and so contradictions in this sphere would be genuine logical contradictions. But direct revelation is not the only kind of revelation recognized in the Jewish tradition.

Within the written Torah itself are passages which rabbis have taken as proof that the Written Law was not intended to be self-sufficient, but accompanied by an unwritten, Oral Law. These passages include laws which are stated incompletely and require further interpretation to distinguish between possible applications²³ and episodes in which humans are invested with the authority to answer questions about the law.²⁴ The Oral Law is taken to be a set of traditions which enabled Jewish authorities to decide between alternatives, or provide clarifications, when questions about the law were raised.

The content of the Oral Law is talked about in two ways. At points, the rabbis employ the Oral Law as, effectively, a set of facts or answers which was passed on orally. This first kind of content is of no special interest to us in this context. If all that distinguished the Oral Law from the Written Law was the fact of its being transmitted orally, this would still be a direct revelation in the relevant sense.

However, while the written law is given directly, not all of the Oral Law is; it is not merely the transmission of memorized tradition, but also a living activity in which this tradition is put to work, and operated to yield new rulings and interpretations. In this latter sense, the Oral Law is revelation via an institution – namely, the institution of rabbinical deliberation.

These are both differing ways in which the notion of Halakha l'Moshe m'Sinai – law given to Moses on Sinai – is employed in the Oral Law, one being direct transmission from an original authority and the other indirect generation by authorized individuals. It is this second aspect to the Oral Torah – as an ongoing activity, rather than a memorized set of given answers – which interests us, and offers a way to understand the apparent true contradictions in the Talmud.

This ‘institutional’ aspect to the Oral Law introduces an important difference not only in the medium used, but also in the kind of revelation in play. Simply, the Oral Law grants autonomy and power to the halakhic process, and the legitimacy of its results cannot be reduced to the direct word of God. This can be seen clearly by looking at the celebrated tale of Eliezer and the oven of Akhnai. R. Eliezer has offered a ruling on the halakhic status of an akhnai (‘snake’) oven, which the other rabbis reject. Ultimately, God himself asserts Eliezer's correctness, but is rebuked by the rabbis in the following way:

Rabbi Yehoshua stood on his feet and said: It is written: ‘It is not in heaven’ (Deuteronomy 30:12). The Gemara asks: What is the relevance of the phrase ‘It is not in heaven’ in this context? Rabbi Yirmeya says: Since the Torah was already given at Mount Sinai, we do not regard a Divine Voice, as You already wrote at Mount Sinai, in the Torah: ‘After a majority to incline’ (Exodus 23:2). Since the majority of rabbis disagreed with Rabbi Eliezer's opinion, the halakhah is not ruled in accordance with his opinion.²⁵

Eliezer had previously sought to prove his correctness by eliciting miracles from various objects, like trees and rivers, to which his antagonists replied, ‘One does not cite a proof from [a tree, or river, etc.]’. The implication is that one also does not cite a proof from God when faced with vexed questions of halakhah; rather the matter is decided by the ‘majority’. Clearly, then, the Oral Law is not merely the recitation of traditions directly received from God, but also produced by the deliberation of the majority about those traditions. In this respect, the Oral Law is revealed indirectly, via the proper functioning of an institution – namely, the institution of rabbinic disputation.

Indirect revelation and conflict – an institutional account of halakhah

It might be objected that we commonly reject institutional arrangements which produce conflicting results. For example, if tax law was generated which claimed a given accounting procedure was both illegal and not illegal, we would say that the responsible legal institution was poorly formed. It would be poorly formed because it would leave the putative taxpayer bereft of any confidence or guidance in avoiding imprisonment for tax avoidance. We can consider this a pragmatic objection, and we will turn to it in the section following this one.

We could also understand this objection in a more abstract way. The objector might have the suspicion that legal institutions which produce conflicting rulings are not merely poorly formed in practical terms, but defective in some essential way. Something about what it means to be a functioning legal institution at all has been missed. This fundamental flaw would be something like a failure properly to instantiate or follow norms which are constitutive of institutions. We can consider this a theoretical objection. We will confront the theoretical objection in this section.

There is, in fact, precedent in philosophy of law for such a view. We will give an account of this precedent below, only to show that there is space in theory of law for the kind of view we have developed; and, thus, that our proposed account cannot be immediately discounted on a priori grounds.

It is not a priori that institutions that produce conflicting rulings are fundamentally defective; in fact, there are grounds to think that the capacity to produce simultaneously valid conflicting rulings is definitive of some kinds of institutional legal validity. In his late period,²⁶ the philosopher of law Hans Kelsen divorced the structures of legal inference and validity from logical norms (and indeed, from norms generally). This was not a claim that legal norms should not be (or could not be) ordered – rather, the claim was that core logical principles, including the LNC, had no purchase on the question of which legal norms are valid. A practical consequence of this position is, for example, that the co-presence of two legal norms which conflict (in the sense of urging different behaviours or legal redress) with each other in a legal corpus is acceptable. Both conflicting norms are equally valid in principle.

Kelsen has no need of gluts to motivate this position. His account rests in part on a distinction between statements of fact and statements of norms (Kelsen understands norms as imperative statements – e.g. ‘Jail murderers’ – and hence not truth-evaluable). This move was influenced by Jörgensen's famous article ‘Imperatives and Logic’.²⁷ Kelsen takes the view that legal norms (in the language of this article, legal rulings) are justified by the fact that they exist and are enacted within (‘willed by’) the relevant institution. In other words, legal rulings are justified by what the human legal institution does, and continues to do. Hence, the LNC has no sway; if two norms recommend mutually exclusive courses of action, this does not in itself lead to a breakdown in legal practice. Indeed, norms can only be displaced by further norms specifically aimed at eliminating them (‘norms of derogation’²⁸ which expressly remove the validity of specific norms), not by the mere fact of a norm's contradicting other norms in general. Kelsen writes:

That two mutually contradictory statements should both be true, is impossible; that two mutually conflicting norms should both be valid, is possible. And this conflict cannot, like a logical contradiction, be resolved by way of knowledge, with the aid, say, of legal science. Legal science can only confirm the existence of this conflict, and must leave its resolution to the act of will of the legal authority, or to customary non-observance.²⁹

Kelsen's account explicitly entails that the response to conflicting rulings arising in an institution is a question of the institution's own decision to eradicate or tolerate that conflict. (‘Legal science . . . must leave its resolution to the will of the legal authority’.) This mirrors the Talmud's varied treatment of apparent contradictions, sometimes tolerating and often eradicating them. This suggests that Kelsen has the resources to further develop the broad approach we have sketched here. But our intention for the moment has only been to show that there is pre-existing work by central thinkers in philosophy of law which can accommodate conflicts in a similar fashion to our account of Talmudic ‘contradictions’.

The pragmatic value of conflict-prone institutions

Even if we cannot object a priori that an institution that generates conflicting rulings is poorly formed, we can certainly object on pragmatic grounds. Why should we want institutions that generate apparent contradictions – much less halakhic institutions, which are supposed to provide guidance in living a pious life?

The answer to this question must of necessity be somewhat abstract. Halakhah is not monolithic, and across differing traditions there are variations in what, exactly, a halakhic institution is understood to be doing. We will offer a general rationale by looking at Maimonides. Maimonides is a figure who (under an interpretation, and possibly with some abridgement) most traditions are eager to claim as an authority. Accordingly, hopefully this account has some general purchase.

In the Mishneh Torah, Maimonides identifies the rejection of idolatry as the fundamental purpose of Torah, such that engagement in idolatry entirely nullifies all other piety a person might show; and resistance to idolatry, properly understood, is equivalent in piety to complete knowledge of and adherence to the law.³⁰ In the Guide For The Perplexed, Maimonides glosses idolatry as the worship of an image of an intermediary between God and creation.³¹ This intermediating image need not be a literal picture – it might be a belief. Continuous with this is Maimonides's claim – within and without the Guide – that God must be conceived of negatively. Which is to say, we should acknowledge that God can in no way be captured by our concepts. Maimonides approvingly cites Isiah 55:8–9:

For My plans are not your plans, Nor are My ways your ways – declares the Lord. But as the heavens are high above the earth, So are My ways high above your ways and My plans above your plans.³²

Maimonides's enthusiasm for a negativistic view of God is that it prevents us from conceiving ideas which we might then mistake for God itself.

If Maimonides is right that the core function of the law is to eliminate idolatry, what should we expect divine law to be like? It would be structurally opposed to the worship of intermediaries between God and humanity – including worship of the law itself. It follows from this that the law could not be static, nor complete; for this would encourage us to identify our understanding of the law with the demands of relating to and understanding God itself. It follows also that the law should not be exhaustively ordered by transparent logical or ethical principles. That would encourage us to see the principles as the source of the good, and as worthy of veneration. By eluding full clarity, completion, or principle, the law can prevent us from forgetting that God's ways are higher than our ways.

All of the above, we suggest, is good reason to want a halakhic institution which fails to exclude conflicting rulings. Such an institution makes it evident that the demands of piety which derive from God, and our understanding of those demands, cannot be fully identical.

What we hope to have shown is that while the Talmud's occasional indulgence of true conflicts between legal rulings might not be ‘practical’ in the sense of ‘easily giving a quick answer’, it is perfectly practical in terms of supporting and compelling the kind of practice we want when navigating the relationship between mankind and a completely transcendent and unknowable God.

A final objection – halakhah not objective

Halakhah is not meant to be a mere practice or institution like any other. Every morning, in our prayers, we quote from Mishnah Peah and Talmud Shabbat the claim that the study of Torah (meant to include the Oral Law) is equal in merit to deeds of kindness, honouring our parents, and so on. If the Oral Law is a man-made institution, and its products merely conventional, why would this be? Put differently, when we deliberate over the correct halakhah, surely we are not deliberating over what is the best reflection of the majority opinion – rather, we are deliberating over what is required to be right (or least wrong) with God. This is what the present account is missing – the demandingness of the Law.

The answer to this objection depends on three variables. First, one's conception of the nature and function of halakhah. Second, the demandingness of one's concept of reality. Third, the extent to which one believes it is permissible to speculate about God. These three variables can be independently varied, and the required answer will differ with these arrangements. No one article could exhaustively account for the possible variations, and the change in the answer required caused by each of them. For example, the Lithuanian Brisker tradition canvassed by Sinai and Golding (2016) holds that halakhic deliberation occasions the production of new facts by God which underwrite the reality and truth of the results of that deliberation. This combines a commitment to a demanding conception of reality (second variable) with the view that speculation about God in positive terms is permissible (third variable). Conversely, if one took the view that halakhah was designed to elicit moral instruction, and that positive speculation about God was impermissible, then there would be less motivation to take a demanding conception of reality to be relevant. We can find a version of such a combination of views in Maimonides's Guide.³³

Given the diversity of permissible views in Judaism, within and outside orthodoxy, it would not be possible to furnish a unifying solution to the objection that our account renders halakhah ‘not real’. But we can rebut the claim that on our account halakhah is merely conventional. Our position is not that halakhah is a man-made institution. On the contrary – we follow the rabbinic dictum that the Oral Law is divinely revealed. Our claim is that it is revealed both directly, and institutionally. Halakhic institutions are objectively constitutive of the latter part of the Oral Law. These institutions are, simply, the rabbinate and its ongoing development of its own norms, interpretations, and ascriptions of authority.

The objector might further ask – but what is it that these institutions are about? Are their rulings made true by virtue of matching the eternal will of God, or just by virtue of being produced by them? And how do we tell what is a genuine part of the institution and what is not? If a small group of rabbis rules that lighting fires on the sabbath is permissible, can we say with confidence that they are not part of the institution? And what about large-scale differences in halakhah – between reform, conservative, and orthodox, say? Who is to say which is the authoritative branch – if any?

These are genuine problems – but problems for everyone thinking about halakhah. And these problems present differently depending on the three variables outlined above, among others. As such, they are beyond the scope of this article.

Conclusion

In this article, we have offered an account of the ‘true contradictions’ in the Talmud, claiming that they are, in fact, true conflicts. We argued that they are not to be understood as instrumental (as for R. Kook and Rashi), nor are they resting on mere conventionalism (as for R. Feinstein), and nor are they to be understood as resting on a glut-theoretic view of truth. Rather, such conflicting rulings should be understood as products of the institutions which produced them, and by which they are justified and made true. These institutions are directly given by God, but their results are not. Accordingly, these institutions, which are ‘not in heaven’, can generate true and justified rulings which conflict with each other without needing to appeal to exotic theories of truth, or needing to deny that Talmudic disputation is generally aimed at the avoidance of conflicts where possible.