‘What’s the Plan?’: On Interpretation and Meta-interpretation in Scott Shapiro’s Legality

Abstract

This chapter seeks to explain and evaluate Shapiro’s theory of legal interpretation, as it is outlined in his recent book ‘Legality’. More specifically, in the following I will try to provide (a) a reconstruction of Shapiro’s theory of legal interpretation, as it is developed in ‘Legality’; (b) an assessment of this theory of interpretation on its own terms (i.e. its internal coherence, its overall persuasiveness); and (c) an evaluation of the compatibility of this theory of legal interpretation with the general project of ‘law as plan’.

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

9.1 Introduction

Scott Shapiro’s book, Legality, is a very rich and challenging contribution to analytical legal philosophy – or, as Shapiro would say, to ‘analytical jurisprudence’.¹ The book, though quite long and rich of historical and theoretical diversions (both useful, of course, to bolster the book’s main argument), has a remarkably linear structure: it embraces one fundamental thesis about the nature of law (‘Law as Plan’), which is developed in an articulate and detailed fashion, consistently defended against possible rivals, and applied to many different facets (actually, to all facets) of the phenomenon under analysis (i.e. the law).

In sum, Shapiro’s endeavour is that of elaborating a full-fledged theory of law, in the mark of the tradition of legal positivism. Obviously enough, such a theory of law comprises also a theory of legal interpretation. Indeed, any theory of law is incomplete if it does not flesh out the consequences it is supposed to bear on matters of interpretation. A theory of law that bears no consequence at all on interpretive issues or, the other way round, a theory of interpretation that can be attached to any possible theory of law are, though not unconceivable, highly suspect.

The aim of this chapter is to explain and evaluate Shapiro’s theory of legal interpretation, as outlined in Legality – an important part of Shapiro’s theoretical enterprise that has not yet attracted, as far as I know, much interest in the already conspicuous literature on Legality.²

More specifically, in this chapter I will try to provide (a) a reconstruction of Shapiro’s theory of legal interpretation, as it is developed in Legality (Sect. 9.2); (b) an assessment of this theory of interpretation on its own terms (i.e. its internal coherence, its overall persuasiveness) and (c) an evaluation of the compatibility of this theory of legal interpretation with the general project of ‘Law as Plan’ (Sect. 9.3). By ‘compatibility’ here, I do not mean a matter of logical consistency, but rather a matter of ‘soundness’: in other words, as long as sound theory construction is concerned, a theory of interpretation should not be construed in such a way as to frustrate the general aims of the theory of law to which it relates.³

My analysis of Shapiro’s theory of legal interpretation, as it is formulated in Legality, will point to some serious problems in his account that concern the role of substantive evaluative judgements (including also moral judgements) in the process of legal interpretation and in the extraction of the ‘economy of trust’ and the unduly narrow scope of the Planning Theory of legal interpretation. I conclude that all these three problems are, in turn, related to one major problem related to the proper identification of the actors that can be considered the real authors of the legal plan.

9.2 The Planning Theory of Law: From the Concept of Law to Legal Interpretation

Roughly a good half of Legality (chapters VIII–XIII) is devoted to matters of (the theory of) legal interpretation. In other words, matters of interpretation are vital for the project of law as plan, and Shapiro is perfectly aware of this.

Let us take a brief look, then, at the theory of legal interpretation associated to law as plan.

9.2.1 A (Very) Brief Survey of the Planning Theory of Law

The first thing to note, perhaps, is that Law as Plan is not only a positivistic theory but a strong positivistic theory. Indeed, Law as Plan, or ‘the Planning Theory of Law’ (Shapiro 2011, 195) or ‘plan positivism’ (Shapiro 2011, 178), is admittedly a reformulation of exclusive, or ‘hard’, legal positivism (Shapiro 2011, 267–273), with the help of some new philosophical tools: mainly, the concept of ‘plan’ borrowed from Michael Bratman’s recent work.⁴

I will not deal here with the details of Shapiro’s elaborated theoretical framework; for my argument’s sake, suffice it to say that:

• A plan is a kind of norm (Shapiro 2011, 127–129): it is a guide for conduct (it picks out courses of action that are required, permitted, or authorized under certain circumstances) and a standard of evaluation (it is supposed to be used as a measure of correct conduct). A plan need not discipline entirely and all at once the kind of behaviour it applies to: it can be ‘partial’. Moreover, a plan is a ‘positive purposive’ entity: ‘a norm is a plan as long as it was created by a process that is supposed to create norms’ (Shapiro 2011, 128), and it disposes its subjects to follow it.⁵

• Plans reduce deliberation costs in circumstances of uncertainty: circumstances in which the relevant actors face the complexity, contentiousness, and arbitrariness of communal life, due to substantial moral disagreement between individuals, uncertainties about the more appropriate ways to achieve some valuable end, and so on.

• The law (in the sense of ‘the legal system’) is a planning organization: an organization that exercises planning activity, that is, that produces plans; the adoption of (legal) plans is called for by the ‘circumstances of legality’: ‘the circumstances of legality obtain whenever a community has numerous moral problems whose solutions are complex, contentious or arbitrary’ (Shapiro 2011, 170).

• If contrasted with other planning organizations (such as criminal organizations, for instance), the law is defined by its distinctive aim or function⁶: ‘the fundamental aim of the law is to rectify the moral deficiencies associated with the circumstances of legality’ (Shapiro 2011, 213)⁷ (‘the moral aim thesis’); this is, according to Shapiro, the fundamental function of the law.

Legal activity is a planning activity – a planning activity is the activity of making (devising, developing, and implementing) plans – and plans are norms. As a consequence, plans issued within the legal activity are legal norms, and conversely, legal norms are plans – or at least ‘plan-like’, in case they are not positively created by a planning organization but just adopted thereby (Shapiro 2011, 120).

Moreover, legal activity is a planning activity not only because it produces plans (norms) but also because it is itself structured as a plan (Shapiro 2011, 176). By this, Shapiro means that at the root of the activity of legal planning (production, modification, and enforcement of legal norms) lies a ‘master plan’ that regulates the further activities of planning to be carried on by officials, marking the relevant divisions of institutional labour between them – the formulation and adoption of new plans, their enforcement, etc. (Shapiro 2011, 176–180). In most legal systems, the master plan can be conveniently identified with the constitution (Shapiro 2011, 169, 205), even if part (indeed, even a large part or the totality) of the master plan can also be customary in character, and hence only plan-like. Such a set of fundamental rules (the master plan) exists because it is a ‘shared plan’.⁸

9.2.2 The Place of Legal Interpretation in the Planning Theory

With this few general remarks on the overall framework of the Planning Theory in mind, let us now take a closer look at the rile of legal interpretation according to the Planning Theory.⁹

Interpretation is the activity through which a legal text is given a certain meaning: any act of interpretation, then, can be subsumed under a certain ‘interpretive methodology’, that is, ‘a method for reading legal texts’ (Shapiro 2011, 304).

Every legal culture allows for a vast array of interpretive methodologies (for instance, textual interpretation, intentional interpretation, historical interpretation, purposive interpretation, and so on). Moreover, in any given case (a) more than one interpretive methodology can be legitimately available to the interpreter, and (b) the different interpretive methodologies available to the interpreter can lead to different interpretive results. The conclusion immediately follows that the interpreter, in most cases (maybe, always), is supposed to make a choice between the various interpretive methodologies available. This choice is a matter of ‘meta-interpretation’; it belongs to a meta-interpretive theory (Shapiro 2011, 304–306).

Now, the first interpretive implication of the Planning Theory is rather straight forward: if legal norms are plans, then legal interpretation is interpretation of plans (Shapiro 2011, 194). Recall that a plan is a device that is supposed to reduce deliberation costs in situations of uncertainty, complexity, controversy, etc. (the circumstances of legality), and that it does so by ‘tying the hands’ of its addressees: they will just have to follow the plan blindly,¹⁰ without making any further deliberation. Shapiro calls this the ‘Simple Logic of Plan’ (SLOP): ‘The existence and content of a plan cannot be determined by facts whose existence the plan aims to settle’ (Shapiro 2011, 275). Since legal norms are plans and plans are supposed to perform this very function, it follows that legal norms have to be interpreted (and must be amenable to be interpreted) in a way that does not resuscitate those very controversial issues that the plan was supposed to settle.¹¹

Shapiro deals with this issues under the heading of the ‘General Logic of Plans’ argument (GLOP): the interpretation of any member of a system of plans cannot be determined by a fact whose existence any member of that system aims to settle (Shapiro 2011, 311).

Shapiro praises, then, a theory of meta-interpretation that respects GLOP. This requires resorting to the concept of ‘trust’: legal systems are plans, and plans are based on certain attitudes of trust.¹² The meta-interpretation mandated by the Planning Theory, then, focuses on the actual distribution of trust within the legal system. Every plan (and every legal system) is premised on a certain system of trust management: if a plan (or, more correctly, the plan designers) shows a significant amount of trust that a given (type of) actor in a given (type of) situation will be able to overcome the circumstances of legality, then the system will authorize that actor to develop the plan (the plan will grant him interpretive discretion). If, on the contrary, the system assumes that a given (type of) actor in a given (type of) situation is not trustworthy, or not trustworthy enough – in other words, if the system assumes that the actor will not be able to overcome the circumstances of legality in a given situation – then it grants that actor a constricted role (he will just have to follow the plan blindly, without trying to develop it) (Shapiro 2011, 353).

This is what Shapiro calls the ‘economy of trust’ of a system (Shapiro 2011, 331).

The economy of trust is essential in the choice of an interpretive methodology, for the allocation of decision-making power operated by an interpretive methodology must be consistent with the economy of trust of the system. Now the rather obvious question is: ‘how are we supposed to ascertain the actual economy of trust on which the system is premised?’¹³ Shapiro stipulates two possible (families of) methods.

The first method is the ‘God’s-eye method’; it requires that each (meta-) interpreter autonomously decide about his own degree of trustworthiness: if he deems himself to be in the position of deserving a great amount of trust (either in absolute or in some specific circumstances), he will grant himself interpretive discretion; if, on the other hand, he deems himself untrustworthy, he will ‘tie his own hands’ and will defer to someone else’s judgement.

The second method is the ‘Planners method’: in such a case, ‘a meta-interpreter should not assess her own trustworthiness, but rather defer to the views of the system’s planners regarding her competence and character’¹⁴ and choose an interpretive methodology accordingly.

In other words, an interpreter who resorts to the God’s-eye method is actually resorting to his own judgement in determining the proper economy of trust of the system. On the other hand, an interpreter who resorts to the Planners method tries to refer to the economy of trust as it is designed by the plan designers (i.e. the Framers) and ‘embedded in the plans of the legal system’.¹⁵

Obviously enough, the Planning Theory requires the Planners method of meta-interpretation and firmly rejects the God’s-eye method: the former is perfectly consistent with the GLOP, whereas the latter openly violates it and reopens the ‘Pandora’s Box’ of the circumstances of legality (Shapiro 2011, 348).

But this is not yet a final verdict against the God’s-eye method and in favour of the Planners method of meta-interpretation. Indeed, the choice between the God’s-eye method and the Planners method depends on the reasons and attitudes of actual participants in the legal system: ‘In an “authority” system, the reason why the bulk of legal officials accept, or purport to accept, the rules of the system is that these rules were created by those having superior moral authority or judgement. The authoritative provenance of these rules, in other words, is deemed to be of paramount moral importance. In an “opportunistic” system, by contrast, the origins of most of these rules are deemed morally irrelevant. Officials in these regimes accept them because they recognize, or purport to recognize, that these rules are morally good and hence further the fundamental aim of law’ (Shapiro 2011, 350).

It is an empirical question whether a given legal system is the ‘authority’ or the ‘opportunistic’ type. As a matter of fact, Shapiro believes that the actual US legal system, for instance, belongs to the former kind (Shapiro 2011, 351). At any rate, the Planners method is appropriate (indeed, required) for authority systems alone.

So in an authority system, the economy of trust – the system’s planners attitudes of trust (Shapiro 2011, 351) – will be ascertained through the Planners method, which in turn results in a theory of meta-interpretation articulated in three steps:

• 1. Specification (‘What competence and character are needed to implement different sorts of interpretive procedures?’)

• 2. Extraction (‘(a) What competence and character which the planners believe actors possess led them to entrust actors with the task that they did? (b) Which systemic objectives did the designers intend various actors to further and realize?’)

• 3. Evaluation (‘Which procedure best furthers and realizes the systemic objectives that the actors were intended to further and realize, assuming that they have the extracted competence and character?’)

At the end of all this, and after taking into account also possible matters of ‘competition’,¹⁶ the meta-interpreter will be in a suitable position to individuate the proper interpretive methodology for any given kind of interpreter. This whole process, moreover, is rooted on social facts alone, such as the judgements of competence and character made by the planners and the ‘regime’s animating ideology’ (Shapiro 2011, 382; a rather similar presentation of the interpretive task is also in Shapiro 2007): legal positivism is, then, vindicated.

In sum, then, the theory of legal interpretation developed in Legality is, admittedly, doubly limited. In the first place, it is supposed to apply only to legal systems that have certain specific features – only to ‘authority systems’, as defined above. Second, the concrete bearings (the whole panoply of meta-interpretive issues) of this theory of interpretation are indeterminate, because they are radically contextual: it all depends on how a given legal system allocates amounts of trust and distrust between its officials – its economy of trust. Accordingly, if the legal system deems a certain kind of official trustworthy, or comparatively more trustworthy than some other kind of official, it will grant him a considerable amount of interpretive discretion (allowing him to use purposive styles of interpretation, for instance). If, on the other hand, the legal system distrusts a certain kind of official, or distrusts her comparatively more than some other kind of official, it will require her to keep her interpretive powers at bay, probably by adhering strictly to the wording of the rules laid down by the plan designers (or by some other more trustworthy officials).

So, in the end, Law as Plan does not require any specific interpretive methodology. Rather, it requires that each interpreter choose the interpretive methodology that appears to be the more appropriate, in relation to that interpreter’s role and place in the system’s economy of trust. And it is each interpreter’s task to ascertain his own role and place in the system’s economy of trust, under the regime’s animating ideology.

9.3 Legal Interpretation in the Planning Theory: Some Doubts and Queries

Given this general, and quite complex, theoretical framework, I will now try to assess some possible critical points in the treatment of the issues of interpretation and meta-interpretation in Shapiro’s Planning Theory.

9.3.1 Value-Free Adjudication?

According to Shapiro’s theory of law and legal interpretation, legal reasoning is necessarily amoral (Shapiro 2011, 240, 244, 266–267, 276). This means that, as far as genuine legal reasoning is concerned, the interpreter does not (and should not) resort to moral considerations: he will just have to look at social facts, in the guise of positive laws. Of course, Shapiro concedes, the resolution of a dispute may require the judge to ‘reach outside the law’, to look also at moral facts in order to adjudicate a dispute; sometimes, according to Shapiro, this may indeed prove inevitable, since laws are the product of human beings, whose ability to foresee all the relevant disputes and all the relevant features of possible disputes (indeed, whose ability to plan) is limited. Since law is a human product, and since human beings have cognitive as well as moral limitations, law is intrinsically limited.¹⁷ When the interpreter is confronted with a case whose solution (or maybe whose optimal solution) is not provided by existing law, the decision of such a case is to be attained outside the law. And in such a case, the judge would not be engaged in legal reasoning (which is necessarily amoral) but in sheer legal decision-making: he is not applying pre-existing law; he is just solving a dispute.¹⁸

Translating all this in the jargon of the Planning Theory, in some cases (hard cases) the original plan can ‘run out’, and judges are required to write a new plan on their own in order to adjudicate the case at hand.¹⁹

Now, I wonder if the portrait of the judicial task elaborated as an upshot of the Planning Theory is really illuminating. Indeed, I suspect that it obscures important parts of judicial legal reasoning.

My main contention, here, is that each and every act of interpretation necessarily involves value judgements. This is true, of course, in cases of extreme indeterminacy, gaps, conflicts of norms absent a clear criterion of composition, conflicts between legal and moral requirements, etc. (hard cases). But it is also true in cases in which there is a legal text, formulated in clear words. This derives from the following main factors:

• (a) Even clear words have to be interpreted (they do not carry their own meaning on their face), and even clear words can have ‘penumbras’ of meaning (a relative degree of vagueness and indeterminacy is always an ineliminable feature of language).²⁰ Moreover, even the interpreter’s choice of resorting to literal meaning is namely that: a choice.

• (b) A given legal text is clear only insofar as it is not ‘upset’, unsettled by a specific case which happens not to match perfectly with the formulation of the legal text (what would be the job of the interpreter in such a case? Should he resort to such controversial entities as the ‘purposes’ of the plan, even if he has been deemed untrustworthy by the plan designers – and indeed, even if he deems himself so?).

• (c) In precedent-based systems, a given case can usually be subsumed under more than one precedent; moreover, from any given precedent usually more than one ratio decidendi can be inferred.

• (d) In statute law systems, the interpreter always confronts a vast array of legal materials, not just one statute, or part of a statute. So, the interpreter has to reconstruct the ‘plan’ (the relevant legal norm) out of an array of raw legal materials, and in so doing he will be, again, faced with substantive choices.

• (e) The interpreter normally has to engage in an enquiry into the validity of the norm to be interpreted (is the relevant legal norm still in force? Does it present grounds of unconstitutionality? And so on), and this requires further interpretive activities.

What I am trying to point out with the preceding observations is that the interpreter consistently faces various substantive choices not only in hard cases but also in purportedly easy cases, and these choices are guided by substantive value judgements (ideas of reasonableness, soundness, etc.²¹). If all this is correct, then Shapiro’s distinction between legal reasoning properly understood (which is necessarily amoral) and judicial decision-making (which involves the exercise of some kind of moral judgement by judges) is misguided, because amoral legal reasoning never obtains. The only exception I can think of is when legal reasoning is aimed only at establishing that a given law is in force because some facts obtained. But this is very far from exhausting the complexity of legal reasoning even in the easiest of easy cases.

Moreover, if the argument above is correct, a significant adjustment is needed in the Planning Theory as far as legal interpretation is concerned. Indeed, since (a) value-free adjudication does not exist and (b) every interpreter is always required to add something to the plan, then (c) the difference between plan designers and plan appliers collapses. At the very least, that difference is just a matter of degree.²² But a stronger (and maybe more coherent) implication of the argument would be that judges are always not only plan appliers but also plan designers. I will return on this point later on (§ 4).

The argument of the limits of law prompts also another, related, perplexity. True, law has limits, because it is a human, social enterprise. As a consequence, it can easily happen that, in some cases, we just ‘run out of law’, at least apparently. But I find it quite unrealistic to argue that when the law runs out, the interpreter is simply entitled to reach outside the law, looking for an answer in the realm of morality. Indeed, as a matter of fact, in most contemporary legal systems I am aware of, when the law has apparently run out, interpreters do not just look for a good solution on purely moral grounds: instead, they keep on looking for guidance from the law also in hard cases.²³ In other words, positive law still exerts a kind of ‘gravitational force’ on judicial reasoning also when it does not strictly control the case at hand. This could happen by requiring that the interpreter decide the case with reference to general (or constitutional) principles, precedents, analogies, and so on. Of course, in such cases, the interpreter has to rely more heavily on evaluative, substantive considerations, and the solution may be more controversial than in other cases since, for example, different principles may support different analogies and outcomes (Hart 1994, 275). But still, the same happens also in easy cases, as we have seen. Thus, the difference between hard cases and easy cases, between legal reasoning and legal decision-making (in Shapiro’s words), is just a matter of degree.

Moreover, that judges decide hard cases by reaching outside the law is just a contingent, system-specific possibility (Chiassoni 2012): for instance, a given legal system could require judges to refrain from deciding hard cases and defer them to the legislature instead. So, the thesis that in hard cases, when the law (or plans) runs out, judges necessarily reach a decision outside the law should be understood at any rate not as a conceptual truth about the law but rather as a contingent, system-specific matter. This, in turn, would also be more consistent with the limited domain of the Planning Theory of legal interpretation. As we have seen (supra, Sect. 9.2), the Planning Theory of interpretation does not provide definite answers to interpretive problems in general but requires a careful inquiry on the economy of trust within a given legal system instead. So, a given legal system could deal with hard cases at least in one of the following ways (or even in a mix between them): it could require judges to adjudicate the case anyway, resorting to some kind of moral reasoning; or it could require judges to defer the judgement to the legislature, or to a higher, specialized court; or it could even require judges to state that the hard case at hand has no solution in the law and hence that it does not require a judicial pronouncement (non liquet).

So, it is not a conceptual truth that in hard cases judges resort to moral reasoning outside the law: they do so only if they are required by the relevant legal system to adjudicate the hard case at hand anyway, and even in these cases, in my opinion, it could be shown that positive law can still exert some control on judicial reasoning. Even in hard cases, positive law can limit the range of permissible interpretive outcomes, and legal reasoning does not become free-floating moral reasoning.

Contra Shapiro, then, my conclusion on this point is that legal reasoning is always contaminated by moral arguments, both in hard and in easy cases. We can decide to call it legal decision-making if we wish, but then we have to acknowledge that this is what judges do all the time and not just occasionally in hard cases. And moreover, is it not an important truism about the law that the job of judges is to apply the law? This truism cannot be explained by Shapiro’s idea that a ‘morally contaminated’ legal reasoning is no more legal reasoning (but sheer decision-making reaching outside the law instead), once we acknowledge that adjudication always involves value judgements.

9.3.2 Extracting the Economy of Trust from the Master Plan

According to Shapiro, the construction of the meta-interpretive theory mandated by the Planning Theory requires individuating the economy of trust embedded in the system.²⁴ Moreover, Shapiro claims that the existence and content of the master plan are a matter of descriptive fact (Shapiro 2011, 192).²⁵

As we have seen, this process involves three steps – specification, extraction, and evaluation (supra, Sect. 9.2). Here I will focus briefly on the extraction stage of meta-interpretation.

Extraction is ‘essentially an explanatory process. The meta-interpreter attempts to show that a system’s particular institutional structure is due, in part, to the fact that those who designed it had certain views about the trustworthiness of the actors in question and therefore entrusted actors with certain rights and responsibilities. The views extracted through this practice are those that best explain the construction and adoption of the texts that guide the practice.’ Moreover, ‘their [i.e. the designers’] views on the trustworthiness of actors are legally significant only insofar as they played a causal role in the actual design and adoption of the authoritative texts’ (Shapiro 2011, 361–362, italics in the original). According to Shapiro, extraction need not be a holistic (Dworkinian-style) enterprise: the meta-interpreter may be content with assessing the attitudes of trust that the system shows towards his role only. As a general criterion, low-rank officials need not embark in wide assessments of relations of trust in the system, whereas top-rank officials can often be required to do so.

The meta-interpreter will derive the designers’ attitudes of trust from ‘the structure of the legal system’²⁶ or, if needed, from a ‘more detailed historical investigation’ (Shapiro 2011, 365–366).

Shapiro is perfectly aware that this is no easy job. Legal systems usually have a very long lifespan; many generations, many different social, cultural, and political forces contribute to shaping them; and they can embed different attitudes of trust at various levels (for instance, a constitutional norm might embed a certain attitude of trust towards a certain type of interpreter, whereas a statutory norm might embed a different attitude of trust towards the same type of interpreter). ‘Because legal systems are built and rebuilt over time, usually by the hands of many individuals, it would be extremely surprising to find a coherent set of views about trust underlying the totality of the law. As a general matter, the task of the meta-interpreter is not merely to recover these disparate attitudes of trust, but also to synthesize them into one rational vision. A system’s economy of trust, thus, is constructed during meta-interpretation, not simply found’ (Shapiro 2011, 366, italics added).

Shapiro describes this process as a kind of factual inquiry on social facts (Shapiro 2011, 275, 382–383). This, in turn, grounds Shapiro’s assumption that Law as Plan is a strongly positivistic theory and also, apparently, the very idea of conceiving of law as a plan.²⁷

Now, I find it rather surprising that the extraction stage within the process of meta-interpretation can be plausibly conceived as a factual inquiry on social facts. Of course, that somebody has a certain kind of belief, ideological or ethical stance, etc., it is certainly a matter of fact. But when it comes to describing or constructing the content of such beliefs, ethical or ideological stances, etc., it is quite odd to conceive this as a factual, empirical inquiry.²⁸ All the more so if the interpreter is also required to make sense of the empirical data he collects, that is, to ‘synthesize them into one rational vision’.

For my part, I would rather argue that individuating the economy of trust of a system is not a purely (or even mainly) empirical matter: instead, it requires a substantive inquiry on the purposes of the designers, their attitudes and intentions, their ideologies, their compromises (since those ideologies can be and often are diverse and incompatible), and on how much of all that is actually written into the text of the constitution. All this requires, inevitably, the (meta-) interpreter to resort to value judgements, that is, substantive evaluative judgements of soundness such as the ones required by the ‘principle of charity’.²⁹ Once the (meta-) interpreter puts his hands into this kind of stuff, he cannot be deemed to carry on a merely empirical, factual research. And if, on top of all this, we also add that according to Shapiro the law has a fundamental moral aim (see supra, Sect. 9.2), it is hard indeed to see how moral evaluations and substantive evaluations, more generally, can be ruled out from this enterprise.

This becomes particularly clear, for instance, when the (meta-) interpreter faces internal inconsistencies in the master plan, that is, when the plan seems to rely on conflicting trust judgements. To deal with such cases, Shapiro counsels an epistemological procedure borrowed from the revision of inconsistent theories in philosophy of science³⁰ – a procedure that, to my mind, could also be described as the search for a kind of Rawlsian reflective equilibrium.³¹

Note that my objection does not involve taking sides in the alternative between internal and external point of view. I mean to say that what is in question here is not the possibility of making detached or uncommitted, instead of committed, evaluative statements: in either case (scil., internal and external evaluative statements), what is at stake is exactly this: an evaluative statement that involves the use of substantive arguments (even if, ex hypothesis, only in a detached way). And an evaluative statement is not an empirical statement.

In sum, my objection here is that conceiving of the extraction stage of meta-interpretation as a factual, empirical inquiry is a distortion, whose aim is to provide (hard) legal positivists with a self-assuring portrait of legal reasoning as based exclusively on social facts.

9.3.3 Is the Scope of the Planning Theory of Interpretation Too Narrow?

My arguments so far have dealt with problems of internal consistency of Shapiro’s theory of meta-interpretation. Now I want to advance a critical argument from a slightly different perspective.

Shapiro’s theory of meta-interpretation is based on the assumption that the legal system under consideration has certain features, namely, it has to be an ‘authority system’, as contrasted to an ‘opportunistic system’.

Recall that ‘in an “authority” system, the reason why the bulk of legal officials accept, or purport to accept, the rules of the system is that these rules were created by those having superior moral authority or judgement. The authoritative provenance of these rules, in other words, is deemed to be of paramount moral importance. In an “opportunistic” system, by contrast, the origins of most of these rules are deemed morally irrelevant. Officials in these regimes accept them because they recognize, or purport to recognize, that these rules are morally good and hence further the fundamental aim of law.’³²

Only in ‘authority systems’ that the Planners method is supposed to work. Now, a given system is an authority system or an opportunistic one depending on the attitudes of the bulk of officials. If the relevant attitude is that of according to special moral status to the planners, then we have an authority system. If, on the other hand, officials think that the plan they happen to be stuck with is morally good (regardless of any opinion on its authors), then we have an opportunistic system.

Notice that, on its face, this enterprise is different from meta-interpretation. Meta-interpretation requires, inter alia, extracting the trust attitudes of the planners. Here, instead, we are dealing with (moral?) attitudes of officials. So, we can resort to the Planners method of meta-interpretation only after we have decided, through a purportedly different procedure, that the relevant system is an authority system.

All this raises, I think, the following questions: How much agreement do we need within the bulk of officials about this feature of the system, in order to obtain the relevant kind of system? How do we ascertain it? Is it not quite possible that different officials have disparate ideas (and sometimes, even no idea at all) about that? What happens if the officials are split between the two attitudes referred above?³³ Who exactly are the plan designers, towards whom high moral respect is directed in authority systems (only the Framers, or also subsequent legislators and judges)?

I do not have precise answers to these questions, but I think the Planning Theory should. Quite to the contrary, the relevant passages in Legality are rather quick. Here we would expect an argument about the procedure to identify the relevant attitude in officials, the amount of convergence in attitudes required in order to have the relevant kind of system, as well as an argument about a description of the current situation in a given legal system (such as the US), with reference to actual attitudes of officials such as courts, legislatures, and so on. Instead of all this, Shapiro leaves us with just a few scattered impressions and a quote from a newspaper article (Shapiro 2011, 351–352), without no further evidence in favour of his (crucially important) point that the US system is in fact an authority system.

For my part, I will just point to a couple of vague intuitions. First, answering at least some of the questions listed above would require, again, an inquiry into substantive reasons, intentions, propositional attitudes, and ideologies. In short, it requires, again, a substantive evaluative inquiry, not just an empirical one.

Second, if, as I suspect, in contemporary, complex, and multiple-actor legal systems (as opposed to ‘charismatic’ ones, as Max Weber would have it) it is never or rarely the case that the bulk of legal officials accept, or purport to accept, the rules of the system simply because these rules were created by those having superior moral authority or judgement, then it follows that the whole meta-interpretive machinery deployed by Shapiro has indeed a very narrow scope of application. Arguably, most, if not all, contemporary legal systems would require a rather different meta-interpretive theory than the one envisaged by the Planning Theory.

9.4 Who Are the Planners?

The doubts I have tried to raise in the previous sections are probably compounded by a single major doubt that has already surfaced here and there in this chapter. This doubt relates to the difficulty of identifying, in Shapiro’s argument, who exactly are the planners of a legal system.

I think that Shapiro has done little to clarify this rather ambiguous point, since in many passages he makes it clear that (insofar as the US legal system is concerned) he has in mind the Framers, including also the ratifiers and the authors of the Amendments, at any rate those who originally designed the system (indeed, part of his refutation of Dworkin’s arguments takes advantage of a rich and very interesting historical digression on the trust attitudes of the Framers).³⁴ At some other junctures, Shapiro includes in the category of planners also other officials who can and do affect the original plan – more recent legislators, apparently.³⁵ This move is certainly reasonable, but the question can be raised if it is still consistent with the definition of an ‘authority system’, upon which a good deal of Shapiro’s argument on meta-interpretation is premised. Moreover, in still some other places, Shapiro says that the allocation of interpretive discretion between officials (which should be one of the main features of the master plan) is the exclusive job of legislators (Shapiro 2011, 338), who thus would turn out to be the ‘plan designers’.

To make my argument more clear, I should point to the fact that the Planning Theory is premised on (inter alia) two ideas that in the end may prove quite difficult to reconcile: (a) the idea that the logic of planning requires that the content of the plan is ascertainable without resorting to the same kind of arguments that the plan was supposed to settle,³⁶ with (b) the idea that plans can be incomplete and can be designed ‘incrementally’ (Shapiro 2011, 122–124, 199–200, 277–279).

Suppose that a judge discovers that for some kind of case, the ‘legal plan’ just happens to ‘run out’; this is rather inevitable since, according to the Planning Theory, the law has limits (in an exclusive positivism vein). And suppose that in such a case the judge decides to adjudicate the case anyway, ‘adding to’ the original plan,³⁷ which clearly involves resorting to the same kind of controversial arguments that the plan was supposed to settle. Now, how are we to describe this situation according to the conceptual framework of the Planning Theory of Law? According to idea (a) above, the plan clearly did not work,³⁸ and the judge has in some sense written a new plan³⁹; on the other hand, according to idea (b) above, the judge has just ‘added to’ the same original plan that now has been developed incrementally.

So, the Planning Theory seems to struggle with two independently sound, but apparently irreconcilable, ideas: the idea that plans perform their specific role when they eliminate the deliberation costs that affect decision-making in environments of uncertainty, moral controversy, etc. (which in turn requires that a plan can be interpreted, and must be interpreted, in a non-evaluative fashion), and the idea that a total planning would indeed be a nightmare, if at all possible: plans are capable of being developed in an incremental fashion, and it belongs to the normal structure of plans (including legal plans) that they make room for areas of indeterminacy.

The only possibility to cope with these two conflicting claims is, I think, to ‘enrol’ judges as plan designers – contrary to the structure of Shapiro’s argument (as I read it, at least) which seems to reserve the honorific label of ‘planners’ only to those high-rank officials that are in charge with structuring the plan: the Framers in the first place and maybe also legislators (‘system designers’) (Shapiro 2011, e.g. 346–349).

Now, what would be the implications of enlisting also judges in the category of planners?⁴⁰ I think that the main implication is that also the plans produced by the judiciary, the ‘doctrines’ developed by courts, shape the legal plan as a whole: the legal plan is (also) what the courts say that it is, and so all this must be taken into account in the ‘extraction’ stage of meta-interpretation (that in turn becomes perhaps even more complex and even less ‘empiric-like’ than envisaged by Shapiro).

This goes hand in hand with another similar point. At many stages of his argument, Shapiro refers to ‘system designers’ as a closed circle, a hortus conclusus. I wonder if this is a sound vision of the law, especially in contemporary, highly complex legal systems in which power relations between different actors are relentlessly negotiated – their battleground often carrying the label of ‘legal interpretation’, of course.

In contemporary legal systems, many different actors shape the law; many different actors shape and reshape their own decision-making powers – and the powers of other actors as well. In the US, after all, the power of the Supreme Court to review legislation has been established by the Supreme Court itself.⁴¹ Constitutional courts in many European countries have consistently acted so to expand their powers, for instance by stating that a number of constitutional principles would be immune from constitutional amendment (and that such amendments could then be subject to judicial review). The process of integration of European national legal systems into European Union Law has been largely dealt with by courts (both national Supreme courts and the ECJ) and so on.

So, the idea that a legal system is the product of a closed and fixed number of plan designers sounds a bit simplistic. In the picture I have drawn, all the actors I have mentioned should be described as consistently resorting to the God’s-eye approach in meta-interpretation: each of them evaluates his own degree of trustworthiness and decides accordingly. Shapiro says that the God’s-eye approach ‘frustrates the ability of the law to achieve its fundamental aim’ (Shapiro 2011, 347). Then maybe we should conclude that, according to Shapiro, a vast amount of contemporary legal systems are just ‘peripheral cases’ of legal systems.

Legal systems evolve, and their structural evolution is not only, and not always, a matter of top-down planning by legislators. The evolution of legal system is sometimes a consequence of judicial decisions; judges are not only plan appliers but sometimes are, in a very important way, plan designers as well.