LAWS, PLANS AND INTELLIGIBILITY: DEFENDING LEGAL POSITIVISM

The debate between legal positivists and antipositivists has progressed to new points of contention. In recent years, a new positivistic theory of law has been put forth by Scott J. Shapiro, called ‘The Planning Theory of Law’. This paper aims to demonstrate how the Planning Theory is able to withstand a powerful antipositivistic objection by Mark Greenberg that social facts, by themselves, are incapable of grounding legal facts in an intelligible manner. Building on David Plunkett’s reply to Mark Greenberg in ‘A Positivist Route for Explaining How Facts Make Law’ (2012), this paper demonstrates how conceptual facts provided to us by the Planning Theory are able to account for the intelligible and reason-based manner in which social facts ground legal facts, thereby creating law without appealing to value facts or morality. A. INTRODUCTION A central debate in the philosophy of law for decades has been whether the existence of law is determined simply by social (or descriptive) facts, or necessarily by value (moral) facts1 also. Philosophers of law have accordingly divided into legal positivists and legal antipositivists respectively. Names like H.L.A Hart and Ronald Dworkin famously dominated the debate, but in the last decade the arguments have progressed and the positions sophisticated. It is into this new phase of debate that this paper enters. Two recent developments are relevant for our purposes. The first is a philosophically novel objection to legal positivism by Mark Greenberg2 arguing that social facts alone cannot account for the intelligibility of law, that is, the reason-based relationship between legal facts and their determinant facts, since social facts cannot by themselves explain their own relevance in determining law. The second relevant development is a fresh account of the nature of law by Professor Scott J. Shapiro, called the ‘Planning Theory of Law’ (henceforth ‘the Planning Theory’).3 Although firmly positivistic, the account departs in many ways from Hart’s ‘laws as rules’ theory of law.4 * Advocate and Solicitor (Singapore). I am grateful to Professor Sylvie Delacroix for supervising an earlier version of this paper. I am also grateful to Michael Workman and an anonymous reviewer for their helpful comments. All errors, omissions and infelicities are entirely my own. 1 Henceforth I will use the term ‘value facts’. 2 Mark Greenberg, ‘How Facts Make Law’ (2004) 10 Legal Theory 157. 3 Scott J. Shapiro, Legality (HUP 2011). 4 H.L.A Hart, The Concept of Law (OUP 1961). UCL Journal of Law and Jurisprudence 26 In this paper, my thesis is that the Planning Theory, properly framed, can answer Greenberg’s challenge and account for the intelligibility of law. Building on David Plunkett’s Reply to Greenberg,5 I will argue that conceptual facts about law can explain the relevance of social facts to determining legal facts and in a way that is equally able to account for our intuitions about legal reasoning that Greenberg appeals to. My strategy is thus to concede Greenberg’s observations about the reason-based relationship between legal facts and its determinant facts (what he terms the ‘rationalrelation doctrine’), but to deny that it is a reason to reject legal positivism, since positivists have an equally plausible way of accounting for the intelligibility of law without appealing to value facts. Further, in demonstrating how an appeal to conceptual facts allows the positivist to meet Greenberg’s challenge, I hope to draw attention to an interesting connection between the epistemological question of how we know legal facts and the metaphysical question of how legal facts are determined. B. THE PLANNING THEORY AND THE ARGUMENT FROM INTELLIGIBILITY In this Part, I set out the context of the debate within which my argument in this paper is advanced: Shapiro’s constitutive account of law in the form of the Planning Theory, and Greenberg’s argument from intelligibility – that there is a reason-based relationship between legal facts and their determinant facts which is problematic for legal positivists. 1. The Planning Theory The central claim of the Planning Theory is that legal activity is a form of social planning.6 Legal institutions and, by extension, legal systems, are instances of organisations of social planning,7 and legal norms are just the set of plans or planlike norms produced by legal institutions.8 What are plans? Plans are positive, purposive entities that aim to settle deliberative questions about what is to be done and so guide conduct over time to achieve goals set by the planners.9 They are positive entities in that they are created and sustained by rational agents, unlike the norms of morality which are (on some accounts) 5 David Plunkett, ‘A Positivist Route for Explaining How Facts Make Law’ (2012) 18 Legal Theory 139. 6 Shapiro (n 3) 155. 7 ibid 154–192. 8 ibid 225. 9 ibid 118–153. Laws, Plans and Intelligibility: Defending Legal Positivism 27 naturally-occurring. They are also purposive in that they are not just created norms, but created to be norms. Plans also settle deliberative questions about what to do. Human beings do not have the cognitive resources to think about every single decision there is to make at every moment of every day. The deliberative costs would be too high. Therefore, to reduce those costs we create and apply plans. Plans also guide conduct over time by placing normative pressure on a person who accepts the plan to follow it, or at least not disregard it without weighty reasons. This pressure comes from the principle of instrumental rationality: once I have intended the ends, I must rationally also intend the means. An important feature of plans is that they can be shared, in that they can be created to be applied to a group of people who may not include the planners. However, the subjects of the plan need not have the same mental state as the planner for it to be accepted by all. Where plans involve many people, widespread alienation can occur without negating the existence of a plan. For example, in a large company, the Chief Executive Officer comes up with a master plan for the whole corporation, with subplans and tasks for each department and individual. He does so because he aims to improve the long-term prospects of the company. However, the majority of his employees may in fact have no interest in the success of the company and thus be alienated from their jobs. Nevertheless, they may accept their roles in the plan and do what they are instructed. Shapiro’s key move is to see legal institutions as instances of social planning: According to what I will call the “Planning Theory of Law,” legal systems are institutions of social planning and their fundamental aim is to compensate for the deficiencies of alternative forms of planning in the circumstances of legality. Legal institutions are supposed to enable communities to overcome the complexity, contentiousness, and arbitrariness of communal life by resolving those social problems that cannot be solved, or solved as well, by nonlegal means alone.10 Therefore, if legal institutions are social planning organisations, then legal norms are just the set of plans or planlike norms created by the activity of such organisations.11 10 ibid 171 (emphasis in italics in original). 11 ibid 225. UCL Journal of Law and Jurisprudence 28 Crucially, Shapiro’s account of plans as positive entities guarantees the positivity of his account of legal norms.12 Under the Planning Theory, the existence of a legal norm, just like the existence of a plan, is never determined by reference to value facts, but social facts alone.13 For Shapiro, plan positivism leads to legal positivism, which makes it potentially vulnerable to Greenberg’s argument from intelligibility, to which I now turn. 2. The Intelligibility of Law Greenberg begins with premises that are uncontroversial as between positivists and antipositivists. First, in any legal system there is a substantial body of determinate legal content, meaning to say that there are many true legal facts in that system; a legal fact being any true legal norm.14 Second, these legal facts are not ontologically primitive; they do not exist as basic facts about the universe. It is not a natural law of the universe that it is illegal to drive faster than 50km/h on city roads. Rather, legal facts are themselves determined by more basic facts.15 Third, among these basic facts are social facts, that is, descriptive facts about what was said and done by individuals. These social facts include anything that was said and done by legal officials, such as legislation, judgments, executive orders, etc.16 Greenberg then begins his challenge by asking, ‘what is the relationship between these social facts and legal facts?’17 It is uncontroversial that the relationship is at least metaphysical, in that legal facts, to an extent, consist in social facts. What we say and do in Parliament, as judges or executives, to some extent makes up the legal facts of the system. However, is the relationship between legal facts and social facts brute metaphysical, such that the social facts entail the legal facts without anything further? An example of a brute metaphysical relationship is the relationship between the chemical composition of water and the physical phenomenon of water itself. Whether something is water is simply a question of whether its chemical composition is H2O. 12 ibid 231–233. 13 ibid 178. 14 Greenberg (n 2) 162. 15 ibid 164. 16 ibid 162. 17 ibid 163. Laws, Plans and Intelligibility: Defending Legal Positivism 29 Once an object’s chemical composition is H2O, it just is water, without further reasons why H2O should be water. A brute metaphysical relationship runs counter to our intuitions about the process of legal reasoning.18 The only access that we have to legal facts is through our social facts. For example, I cannot simply discover by meditation the principles of contract law; rather, I h

In this paper, my thesis is that the Planning Theory, properly framed, can answer Greenberg's challenge and account for the intelligibility of law. Building on David Plunkett's Reply to Greenberg, 5 I will argue that conceptual facts about law can explain the relevance of social facts to determining legal facts and in a way that is equally able to account for our intuitions about legal reasoning that Greenberg appeals to. My strategy is thus to concede Greenberg's observations about the reason-based relationship between legal facts and its determinant facts (what he terms the 'rationalrelation doctrine'), but to deny that it is a reason to reject legal positivism, since positivists have an equally plausible way of accounting for the intelligibility of law without appealing to value facts. Further, in demonstrating how an appeal to conceptual facts allows the positivist to meet Greenberg's challenge, I hope to draw attention to an interesting connection between the epistemological question of how we know legal facts and the metaphysical question of how legal facts are determined.

B. THE PLANNING THEORY AND THE ARGUMENT FROM INTELLIGIBILITY
In this Part, I set out the context of the debate within which my argument in this paper is advanced: Shapiro's constitutive account of law in the form of the Planning Theory, and Greenberg's argument from intelligibility -that there is a reason-based relationship between legal facts and their determinant facts which is problematic for legal positivists.

The Planning Theory
The central claim of the Planning Theory is that legal activity is a form of social planning. 6 Legal institutions and, by extension, legal systems, are instances of organisations of social planning, 7 and legal norms are just the set of plans or planlike norms produced by legal institutions. 8 What are plans? Plans are positive, purposive entities that aim to settle deliberative questions about what is to be done and so guide conduct over time to achieve goals set by the planners. 9 They are positive entities in that they are created and sustained by rational agents, unlike the norms of morality which are (on some accounts) naturally-occurring. They are also purposive in that they are not just created norms, but created to be norms. Plans also settle deliberative questions about what to do. Human beings do not have the cognitive resources to think about every single decision there is to make at every moment of every day. The deliberative costs would be too high.
Therefore, to reduce those costs we create and apply plans. Plans also guide conduct over time by placing normative pressure on a person who accepts the plan to follow it, or at least not disregard it without weighty reasons. This pressure comes from the principle of instrumental rationality: once I have intended the ends, I must rationally also intend the means.
An important feature of plans is that they can be shared, in that they can be created to be applied to a group of people who may not include the planners. However, the subjects of the plan need not have the same mental state as the planner for it to be accepted by all. Where plans involve many people, widespread alienation can occur without negating the existence of a plan. For example, in a large company, the Chief Executive Officer comes up with a master plan for the whole corporation, with subplans and tasks for each department and individual. He does so because he aims to improve the long-term prospects of the company. However, the majority of his employees may in fact have no interest in the success of the company and thus be alienated from their jobs. Nevertheless, they may accept their roles in the plan and do what they are instructed.
Shapiro's key move is to see legal institutions as instances of social planning: According to what I will call the "Planning Theory of Law," legal systems are institutions of social planning and their fundamental aim is to compensate for the deficiencies of alternative forms of planning in the circumstances of legality.
Legal institutions are supposed to enable communities to overcome the complexity, contentiousness, and arbitrariness of communal life by resolving those social problems that cannot be solved, or solved as well, by nonlegal means alone. 10 Therefore, if legal institutions are social planning organisations, then legal norms are just the set of plans or planlike norms created by the activity of such organisations. 11 10 ibid 171 (emphasis in italics in original). 11 ibid 225.
Crucially, Shapiro's account of plans as positive entities guarantees the positivity of his account of legal norms. 12 Under the Planning Theory, the existence of a legal norm, just like the existence of a plan, is never determined by reference to value facts, but social facts alone. 13 For Shapiro, plan positivism leads to legal positivism, which makes it potentially vulnerable to Greenberg's argument from intelligibility, to which I now turn.

The Intelligibility of Law
Greenberg begins with premises that are uncontroversial as between positivists and antipositivists.
First, in any legal system there is a substantial body of determinate legal content, meaning to say that there are many true legal facts in that system; a legal fact being any true legal norm. 14 Second, these legal facts are not ontologically primitive; they do not exist as basic facts about the universe. It is not a natural law of the universe that it is illegal to drive faster than 50km/h on city roads. Rather, legal facts are themselves determined by more basic facts. 15 Third, among these basic facts are social facts, that is, descriptive facts about what was said and done by individuals. These social facts include anything that was said and done by legal officials, such as legislation, judgments, executive orders, etc. 16 Greenberg then begins his challenge by asking, 'what is the relationship between these social facts and legal facts?' 17 It is uncontroversial that the relationship is at least metaphysical, in that legal facts, to an extent, consist in social facts. What we say and do in Parliament, as judges or executives, to some extent makes up the legal facts of the system. However, is the relationship between legal facts and social facts brute metaphysical, such that the social facts entail the legal facts without anything further?
An example of a brute metaphysical relationship is the relationship between the chemical composition of water and the physical phenomenon of water itself. Whether something is water is simply a question of whether its chemical composition is H2O. 12 ibid 231-233. 13 ibid 178. 14 Greenberg (n 2) 162. 15  Once an object's chemical composition is H2O, it just is water, without further reasons why H2O should be water.
A brute metaphysical relationship runs counter to our intuitions about the process of legal reasoning. 18 The only access that we have to legal facts is through our social facts. For example, I cannot simply discover by meditation the principles of contract law; rather, I have to look at what was said and done by Parliament in legislation and judges in case law to know what the law is. However, if the relationship between social and legal facts were brute metaphysical and not reason-based, even after looking at legislation and case law, I still would not be able to tell what the law is, because I would not know how to extrapolate from the mass of social facts before me to arrive at determinate legal facts about contract law.
Reason-based extrapolation from social facts is evident in daily legal practice.
When judges interpret statutes, they write lengthy judgments justifying their interpretations. When lawyers argue in court, they do not simply submit a list of cases to the judge; they seek to justify on the basis of reasons why previous cases support certain legal propositions and not others. It is implicit in our legal practice and discourse that social facts and legal facts relate in a reason-based way. A certain set of social facts support a certain set of legal facts over others on the basis of reasons. Without reasons, the law would simply be indeterminate, because the process of arriving at legal facts from social facts would be opaque and unintelligible.
Further, we would be unable to understand the effect that a change in social facts would have on the law of the system. Suppose that we have a certain set of social facts A and we know that social facts A results in legal facts B. Now suppose that there was a change in social facts A, and we are asked, 'how do legal facts B change?' If the relationship between social and legal facts were brute metaphysical and not based on reasons, then we could not answer the question. For example, we would have no idea whether the deletion of a single word 'the' in a constitution would have no legal effect, or invalidate a single legal doctrine, or invalidate the entirety of constitutional law in the system. Therefore, the relationship between legal facts and social facts cannot be brute metaphysical; it must be reason-based. This is what is meant by 'intelligibility', which Greenberg also calls the 'rational-relation doctrine'. 19 Rational agents must be able to 18 ibid 164. 19 ibid 163. understand why a certain set of determinants of legal facts (such as social facts) results in one set of legal facts and not another. 20 There must be reasons why deleting a single word 'the' from a constitution has the effect that it has on the law. We may not agree what the effect is, but we agree that there is a correct answer, and that the answer is correct because of reasons. To be clear, Greenberg is not making the claim that the content of the law must be rationally intelligible, but rather the relationship between determinant social facts and legal facts must be rationally intelligible. 21

The Problem for Positivists
Having set out this requirement of intelligibility, Greenberg then argues that social facts alone cannot explain law, because they cannot provide the reasons for their own contribution to legal facts. 22 This is essentially a version of Hume's law: in the same way that descriptive facts about the universe cannot by themselves provide reasons for normative conclusions, social facts by themselves cannot provide reasons for claiming that one set of legal facts is correct instead of another. 23 Certainly, there are some social facts that purport to determine how other social facts determine legal facts. For example, it might be written in a constitution that 'only the speeches of a majority in the Supreme Court shall determine the law'. However, in order to decide what, according to that constitution, is the contribution of Supreme Court judgments to the law, we must first decide what that constitution's own contribution to the law is. Essentially, we cannot appeal to social facts B to determine the legal effect of social facts A without first asking what the legal effect of social facts B is.
Further, even if a constitution purported to determine its own contribution to the law, perhaps with the statement 'this Constitution shall be the supreme law of the land', we must still decide how that particular statement contributes to the law. And if one were to point to another set of social facts to explain that contribution, then we have to first decide what this other set of social facts' legal contribution is, ad infinitum. This the preliminary question of what counts as a legal practice that is relevant to determining legal content, as opposed to mere social practices that are irrelevant.
Greenberg's point is this: social facts cannot, without begging the question, determine their own role and relevance to determining legal facts. They are normatively inert; they cannot be a basic source of reasons. Thus, if social facts relate to legal facts in a reason-based way, and social facts cannot provide those reasons, then social facts cannot be the only necessary determinants of legal facts. Something else besides social facts must be the source of reasons that explain the intelligibility of law.
Greenberg suggests that there are two possible candidates that could fulfil the role of intelligibility-enabling facts -conceptual facts and value facts. 24 (Value facts being, as defined in my introduction, genuine moral facts, and conceptual facts being facts about concepts which are constituent components of our thoughts -I will expand on this below). However, Greenberg believes that value facts are a much stronger candidate than conceptual facts for successfully satisfying the rational-relation doctrine. 25 Therefore, according to Greenberg, this is a reason to reject legal positivism.

C. THE INTELLIGIBILITY OF PLANS
In this Part, I advance my substantive argument in this paper. First, I set out a basic understanding of concepts and conceptual facts. Second, I show that the constitutive account of law in the form of the Planning Theory is a result of conceptual analysis that therefore gives us access to certain conceptual facts about law. Finally, I demonstrate how exactly those conceptual facts enable the intelligible, reason-based relationship between social facts and legal facts that allows the positivist to satisfy the rationalrelation doctrine.
As mentioned above, my broad strategy will be to grant Greenberg's observations about the reason-based relationship between legal facts and social facts, but to demonstrate, building on Plunkett's work, that conceptual facts, as opposed to value facts, are just as capable of satisfying the rational-relation doctrine, and in fact do so in a way that equally accounts for our intuitions about law and legal reasoning that Greenberg appealed to. 24 Greenberg (n 2) 187. 25 ibid 187-190.

Concepts and Conceptual Facts
It is helpful to understand concepts in their everyday context. 26 Concepts are the constituent component of our thoughts that our words aim to express. When we use words like 'water' or 'cat' to refer to things in the world, we are applying our concepts [WATER] 27 and [CAT]. Given this relationship between concepts and language, the way we find out more about our concepts is by examining our word usage. If you and I use the words 'water' and 'cat' in the same instances, then to that extent we have shared concepts of [WATER] and [CAT].
As we examine the instances in which we use the word 'cat', we discover common properties in the objects that must obtain if they are to fall under our concept of [CAT] and thus be correctly labelled 'cat'. Such properties might include being a mammal, having four legs, having a tail etc. This process of examining our instances of word-usage in order to determine those properties is called conceptual analysis. 28 Note that we do not actually have to present physical objects to rational agents to examine their word usage. Conceptual analysis can also happen by way of hypotheticals. Chalmers and Jackson note this in their summary of conceptual analysis: Analysis of a concept proceeds at least in part through consideration of a concept's extension within hypothetical scenarios, and noting regularities that emerge. This sort of analysis can reveal that certain features of the world are highly relevant to determining the extension of a concept, and that other features are irrelevant. What emerges as a result of this process may or may not be an explicit definition, but it will at least give useful information about the features in virtue of which a concept applies to the world. 29 Conceptual analysis helps uncover the essential properties of an object X that determine whether it falls under a given concept. These essential properties are conceptual facts about the concept. Suppose we discover through conceptual analysis that in order for some animal to fall under our concept of [WHALE] it must be a mammal, since it is that which distinguishes it from [FISH]. If that is true, then it is a conceptual fact that whales are mammals. This means that if I possess the concept 26 I broadly adopt Chalmers' and Jacksons' theory of concepts in David Chalmers and Frank Jackson, 'Conceptual Analysis and Reductive Explanation' (2001) 110 Philosophical Review 315. My aim is to assume a basic and uncontroversial theory of concepts. 27 I adopt this type to indicate when I am referring to a concept as opposed to the use of words. 28 Frank Jackson, From Metaphysics to Ethics: A Defence of Conceptual Analysis (OUP 2000) ch 2. 29 Chalmers and Jackson (n 26) 322.
[WHALE], then I can reason that, should I encounter a hypothetical animal X, if X falls under my concept of [WHALE] then X is also a mammal. This also illustrates another feature of conceptual facts, for it is not just that the hypothetical animal X is a mammal, but that it also falls under my concept of [MAMMAL]. To summarise, a rational subject who possesses a given concept understands the properties that must obtain in order for something in the world to fall under that concept.
These essential properties are uncovered through conceptual analysis, involving an 30 Plunkett (n 5) 184. 31 Heterochromia is a difference in coloration, usually of the iris, but also, for example, hair or skin. 32 Plunkett (n 5) 184. examination of our word usage. These essential properties constitute conceptual facts that are able to relate claims in different vocabularies with each other and illuminate something about that relationship.

Conceptual Facts from the Planning Theory
Given this basic account of concepts, can we view Shapiro's project in Legality as conceptual analysis giving us access to conceptual facts about law? The starting point must be that Shapiro considers himself to be undertaking conceptual analysis.
Throughout the book he addresses two big questions about law: the Identity Question and the Implication Question.
The Identity Question 'is to ask what it is about X that makes it X and not Y or Z or any other such thing. … A correct answer to the Identity Question must supply the set of properties that make (possible or actual) instances of X the things that they are.
The identity of water, to take another example, is H2O because water is just H2O. Being H2O is what makes water water.' 33 In this sense, Shapiro aims to give an account of the properties of law that make it law in the same way that H2O makes water water.
In asking the Implication Question, 'we are not so much interested in what makes the object the thing that it is but rather in what necessarily follows from the fact that it is what it is and not something else. … to take a trivial finding, mathematicians have discovered that 3 is a prime number. While being a prime number is not part of number 3's identity (being the successor of 2 is), we might still say that it is part of the nature of 3 because being 3 necessarily entails being prime.' 34 In this sense, Shapiro aims to uncover law's necessary properties, the properties that it 'could not fail to have'. 35 In asking these two questions about law, therefore, Shapiro sees himself as undertaking a quintessentially conceptual analysis of law: 'Analytical philosophers have traditionally approached both identity and implication questions by means of a distinctive methodology. … I will refer to it here as conceptual analysis.' 36 This is also borne out by his methodology: 'The key to conceptual analysis, then, is the gathering of truisms about a given entity'. 37 The data that Shapiro collects for his analysis consists of truisms, defined as 'those truths that those who have a good understanding of how legal institutions operate (lawyers, judges, legislators, legal 33 Shapiro (n 3) 8. 34 ibid 9. 35 ibid. 36 ibid 13. 37 ibid. scholars, and so on) take to be self-evident, or at least would take to be so on due reflection.' 38 Many of these truisms would be based on intuitions about law, which Shapiro sees as having a provisional role to play in conceptual analysis. 39 Essentially, Shapiro proceeds by examining our intuitions about when and whether we identify something as law in a hypothetical scenario. If our intuitions lead us to identify a judge's written speech as law, but not the Prime Minister's speech, then those intuitions hint at a potential property of law. Indeed, chapters 5 and 6 of Legality aim to draw on our intuitions about when something counts as law. There, Shapiro builds a very complex planning system from a very simple case of a shared plan to cook dinner together, hoping that by the end we would intuitively identify that planning organisation as a legal system, or at least, see intuitive similarities between complex planning organisations and legal systems. Of course, intuitions can be flawed, and it is the point of a theory to correct them, which is why they only play a provisional role. Conceptual analysis aims to identify the sources of conflict between our intuitions and rationally resolve them, which necessarily requires correcting them. 40 It is therefore reasonable to claim that Shapiro is undertaking the kind of conceptual analysis described in the previous section. Shapiro examines our intuitions about when and whether we would call something 'law' in order to uncover the properties that constitute our concept of law. To the extent that this methodology characterises his project in Legality, it delivers conceptual facts about law.

The Intelligibility-Enabling Role of Conceptual Facts
We must now examine whether these conceptual facts, delivered to us by the Planning Theory, can illuminate the reason-based relationship between legal facts and social facts that the rational-relation doctrine requires. In order to do so, we must first be clear on what these conceptual facts are. We can derive the following conceptual facts from the theses of the Planning Theory set out in Part B above: (1) The legal facts (norms) of a system are just the set of plans created and applied by legal institutions.
(2) Plans are positive, purposive entities that aim to settle deliberative questions about what is to be done and so guide conduct over time to achieve goals set by the planners. 38 ibid 15. 39 ibid 17. 40 ibid 17.
(3) Combining (1) and (2), legal facts are positive, purposive entities that aim to settle deliberative questions about what is to be done and so guide conduct over time to achieve goals set by the system's planners.
How do these conceptual facts account for law's intelligibility? In the previous section, we saw that conceptual facts relate claims in different domains.
[HETEROCHROMIA], for example, relates facts in biochemistry with phenomenal observations. Similarly, once a subject possesses the concept [LEGAL INSTITUTION], conceptual fact (1) allows her to understand that social facts about what was said and done in legal institutions (legislation, judgments etc.) determine legal facts in a system; that there is a relationship between social facts and legal facts.
However, this does not go far enough. Greenberg's challenge is not just that there is a relationship between social facts and legal facts, but that the relationship is reasonbased and thus intelligible to rational agents. At this point, we turn to the Planning Theory of Meta-Interpretation for answers.
It is here that I hope to build on Plunkett's reply to Greenberg. In his reply, Plunkett persuasively demonstrated that conceptual facts (in particular those conceptual facts described by the Planning Theory) are in principle capable of fulfilling the intelligibility-enabling, reason-providing role that Greenberg sees as crucial among the determinants of law. However, as he acknowledges himself, 41 his argument takes on an abstract focus as a proposed positivist route of how an appeal to conceptual facts might meet Greenberg's challenge given the sorts of facts he thinks conceptual facts are (which I have gratefully adopted). 42 For example, Plunkett's argument goes so far as to show that conceptual facts are capable of making intelligible the facts that are needed to meet the rational-relation requirement, such as that 'creatures like us, using the legal concepts that we have, can understand the fact that certain social facts ground legal content', 43 but he does not go further to identify which facts are made intelligible by conceptual facts, and how those facts provide reasons for one set of legal facts over another. In other words, what Plunkett does is to lay crucial philosophical groundwork about the nature of conceptual facts and how they are the kind of intelligibility-enabling facts that a positivist might appeal to to meet the rational-relation doctrine, but he does not go further to demonstrate the process of extrapolating from a specific set of social facts to legal facts by appealing to conceptual facts.
In this paper, however, I hope to push further down his proposed route and demonstrate, at the ground level, how exactly the conceptual facts delivered to us by the Planning Theory enable an individual legal reasoner (Dorothy, whom we will meet later) to derive a determinate set of legal facts from social facts in an intelligible, reasonbased way without appeal to value facts. The reason I think this demonstration is helpful is that part of the persuasiveness of Greenberg's account of how facts make law is its intuitive appeal -that it accords with and explains our everyday experiences of legal practice and reasoning. In fact, I believe this is a good reason to accept the rationalrelation doctrine. However, if it can be demonstrated that an appeal to conceptual facts, rather than value facts, in satisfying the rational-relation doctrine can equally explain our everyday experiences of legal practice and reasoning, then the rational-relation doctrine is no longer a reason to prefer antipositivism to positivism.
My thesis is that the Planning Theory of Meta-Interpretation allows us to explain those everyday experiences, once it is understood not just as a framework for adjudicating between different interpretive methodologies, but as a framework for adjudicating between different ways of determining legal facts from social facts. For Greenberg, the answer to the question of how to choose an interpretive methodology follows from the answer to the question of how to choose between candidate ways of determining legal facts from social facts (hereafter I will adopt Greenberg's term 'models' to refer to these candidate ways). To put it another way, how one account of law answers the question 'which interpretive methodology is correct?' is the same way it answers 'how do social facts rationally determine legal facts?'. If legal positivists are right, then to choose an interpretive methodology one must undertake empirical inquiry about what people said and did in certain instances to derive support for one interpretive methodology over another. If natural lawyers are right, then the only way to choose an interpretive methodology is to engage in moral and political philosophy to show that considerations of fairness or democracy, for example, support one interpretive methodology over another. 47 Therefore, given the connection between models and interpretive methodologies, examining how the Planning Theory determines questions of which interpretive methodology to adopt will provide the clearest answer as to how the Planning Theory satisfies (or attempts to satisfy) the rational-relation doctrine.

The Planning Theory of Meta-Interpretation
As mentioned, one implication of seeing laws as plans is that laws are tools designed to achieve certain ends or goals: 'Legal activity also seeks to accomplish the same basic goals that ordinary, garden-variety planning does, namely, to guide, organize, and monitor the behaviour of individuals and groups.' 48 This applies at the level of an entire legal system. A legal system is a planning system designed to achieve numerous and complex moral or political ends. 49 Further, these ends are too complex for a single individual to achieve, so the plans of a legal system aim to guide and organise behaviour across time and between persons, often many different persons. In doing so, the legal system assigns roles to various actors; some are plan-creators such as legislators and some are plan-appliers such as law enforcement agencies and judges (although common law judges would be both plan-creators and plan-appliers). These roles are often expressed through legislation, with instructions on how to carry out those roles and for what end, for example: 'A Constitutional Court shall be created consisting of six Supreme Court Justices who are charged with upholding the Constitution and in such matters their judgment shall be supreme.' Such instructions make up the master plan of a system.
How the master plan is formulated will indicate how much trust is to be assigned to different actors in furthering their objectives. In general, the more detailed the instructions given, and therefore the less discretion, the less trust is seen to be given to that actor. Correspondingly, the more discretion afforded to an actor, the more trust is seen to be placed in him. Such judgements of trust and distrust make up what Shapiro calls the 'economy of trust'. 50 Since the ends of a legal system are numerous and complex and can only be achieved by many individuals playing different roles, one of the key functions of plans is to manage this economy of trust. 51 This is important in the context of meta-interpretation because Shapiro argues that questions of meta-interpretation cannot be resolved a priori. Questions of appropriate interpretive methodology (textualism vs purposivism, etc.) can only be resolved by reference to the economy of trust presupposed by the actual legal system in question. 52  appropriate for a legal system that affords its actors a great amount of trust, but inappropriate in a legal system that is miserly in its judgements of trust.
We can illustrate this interplay between judgements of trust and interpretive methodologies with a simple, non-legal example. Andrea is a chief engineer in a nuclear plant. Once a month she calls one of her senior engineers, Brie, and assigns her the task of checking the centrifuges in the plant. Now because she is a senior engineer with many years of experience, Andrea's instructions can be as general as 'go make sure the centrifuges are okay.' This is because Andrea places a high level of trust in Brie. Suppose Brie is replaced by a younger employee, Camille, fresh out of university. When it is Camille's turn to check the centrifuges, because she is much less experienced, Andrea gives her much more detailed instructions on how to check the centrifuges. Andrea writes down exactly what ranges of levels are acceptable, exactly what temperature each centrifuge must be at, and the exact steps that Camille must take to check all the centrifuges. This is because Andrea places a lower level of trust in Camille. Now suppose Camille approaches you asking how she should interpret the instructions given to her by Andrea. Should she adopt a literal, textualist interpretation, being faithful to every digit of the instructions? Or should she take a purposive approach? Suppose you have not looked at the instructions yet. Does it make sense to tell Camille that because the point of the plan is to achieve a certain end, that is, that the centrifuges are checked, she should always interpret the plan in light of that end?
Of course not, because that would miss the point of the question. Camille does not know how to check the centrifuges. Further, once you have looked at the instructions, you realise that the economy of trust underlying the plan is one where Camille is afforded little trust, and it is the plan that is intended to settle all questions on how to achieve its purpose. Thus, the appropriate methodology is not purposive, but literalist, because to do otherwise would frustrate the economy of trust presupposed by the plan.
The principle illustrated is this: the question of which interpretive methodology to adopt is to be answered according to the objectives of the legal system set by its planners and in light of their judgements of trust and distrust on various actors in furthering those objectives. 53 To determine what the proper interpretive methodology is, we first ascertain the economy of trust presupposed by the system in light of its 53 ibid 370.
objectives. This involves looking at the plans of the system to ascertain its judgements on different actors and thus the measure of trust afforded to them. Shapiro calls this 'extraction'. 54 Then, we rank interpretive methodologies according to how well those extracted objectives are served if the interpretive methodology in question is followed by the relevant actors. This means that interpretive methodologies that afford a high amount of discretion to those actors to whom the system attributes a low amount of trust are unlikely to serve those objectives very well. Correspondingly, interpretive methodologies that afford a low amount of discretion to those actors are likely to serve those objectives much better. The interpretive methodology that best furthers those objectives is the correct one. If there is a tie between two, then both interpretive methodologies are correct, and the law is indeterminate at those points. 55 Note that this does not mean that one interpretive methodology will apply across an entire legal system. In a single legal system there may be some areas where the system attributes a greater amount of trust to certain actors (judges, civil servants etc.) and others where the system attributes a lower amount of trust to those same actors.
The meta-interpreter must decide the level of generality at which she is to conduct her meta-interpretive exercise. 56 Of course, this is a very crude version of the Planning Theory of Meta-Interpretation. The principle I have stated thus far only discriminates between interpretive methodologies of varying degrees of discretion, but not types, and therefore does not yield unique answers given the rich variety of interpretive methodologies available. Shapiro develops a more sophisticated version in Legality, but this crude version suffices for our purposes. Our aim is to see how the Planning Theory's approach to Meta-Interpretation gives insight into how it could answer Greenberg's challenge.

Satisfying the Rational-Relation Doctrine
Both Shapiro and Greenberg see a connection between choosing from candidate ways that social facts determine legal facts and choosing between different interpretive methodologies to adopt when interpreting sources of law. Both exercises involve discriminating between different ways of determining legal facts from a given set of social facts. 54  What was the Planning Theory's approach to Meta-Interpretation? The interpretive methodology that best serves the objectives of the system in light of its presupposed economy of trust is the correct one. What sorts of facts does this involve? Social facts. According to Shapiro: That some set of goals and values represents the purposes of a certain legal system is a fact about certain social groups that is ascertainable by empirical, rather than moral, reasoning … This account of legal interpretation is positivistic in the most important sense, namely, it roots interpretive methodology in social facts. That a legal system has a certain ideology is a fact about the behaviour and attitudes of social groups. 57 Therefore, the Planning Theory discriminates between candidate ways of determining legal facts from social facts by appealing to further social facts, in particular, facts about the attitudes of the system's planners. This immediately raises a problem. Have we not already granted the inability of social facts to explain the reasonbased relationship between social and legal facts? How can we appeal to more social facts?
The problem disappears when we become clear on which facts are actually doing the intelligibility-enabling in the process of extrapolating legal facts. We can see this clearly when we consider how the legal reasoning process works from the perspective of a hypothetical legal reasoner, Dorothy.
Suppose Dorothy wants to know what the correct answer is to a question of contract law in her legal system. For argument's sake, let us say the question is whether the law permits the use of extrinsic material to determine the meaning of contractual clauses. She has on her table a whole mass of social facts such as (as a representative sample) what was said and done by parliamentarians and judges, the annual mean historical rainfall levels in her country for the past 100 years and the slogans chanted by protestors that marched past her home last month. However, she has no idea (yet) whether and how any of them are relevant to determining the question of contract law.
Nevertheless, she knows that she is here to determine what the law is, and so she must appeal to her concepts about law, in this case, the conceptual facts about legal institutions and legal facts accessible to her through the Planning Theory. She appeals to conceptual fact (3): that legal facts are positive, purposive entities that aim to settle 57 ibid.
deliberative questions about what is to be done and so guide conduct over time to achieve goals set by the system's planners.
This enables her to do two things. First, she correctly identifies that what was said and done by parliamentarians and judges is relevant to determining the law, but not her country's historical weather data or last month's street protestors. Second, she understands that within the relevant social facts there are two sub-groups of social facts.
The first sub-group of social facts, Social Facts X, are facts about the attitudes of the system's planners regarding the objectives of the system and its economy of trust.
The second sub-group of social facts, Social Facts Y, are facts about what was said and done in the planning process. Conceptual fact (3) tells her that these two groups of social facts stand in a certain relationship with each other in determining legal facts, which is that Social Facts Y must determine legal facts (in this case a question of contract law) by reference to Social Facts X.
At this point we should pause to note what has transpired. Appealing to conceptual fact (3) has enabled Dorothy to identify which social facts are relevant and which are not to determining legal facts. Conceptual fact (3) has also revealed to Dorothy that the relationship between social facts and legal facts is reason-based, i.e that the relationship between Social Facts Y and the legal facts to be determined is reason-based, and that those reasons must be provided by Social Facts X. However, she does not yet know what those reasons are. In other words, there is a higher-order intelligibility enabled here, which points out the relevance of certain social facts and not others to determining legal facts, but not (yet) the lower-order intelligibility of identifying the reasons why those social facts result in one set of legal facts and not another.
Greenberg draws attention to this distinction when arguing that value facts enable intelligibility, but he elides these two levels of intelligibility: By contrast, value facts are well suited to determining the relevance of law practices, for value facts include facts about the relevance of descriptive facts. or in what way. The significance for the law of the fact that a certain model is all-things-considered better than others is simply the fact that that model is better than others. 58 In other words, Greenberg believes that in the case of value facts, the higherorder intelligibility (the relevance of value facts themselves) is self-evident. We need not question his claim at this point, we simply need to note the two levels of intelligibility involved in his account, because our appeal to conceptual facts differs in this respect. The higher-order intelligibility enabled by conceptual fact (3) answers Greenberg's objection to the extent that he claims positivists cannot determine the relevance of some social facts as opposed to others in determining legal facts. What remains to be shown is the lower-order intelligibility, that is, the reasons provided by those relevant social facts to prefer one candidate way of determining legal facts over others.
Returning to Dorothy, conceptual fact (3) has helped her to identify the relevant social facts, and that determining legal facts from those relevant social facts must be done by reference to two sub-groups of social facts, Social Facts X and Social Facts Y, such that out of all the possible methods of determining legal facts from Social Facts Y, the method that is best justified according to Social Facts X is the legally correct method.
How, then, do Social Facts X provide reasons for favouring one model of extrapolating legal facts over another? In order for these facts to provide reasons, they must be the sort of facts that are capable of providing reasons the way that value facts do in Greenberg's account; that is, Social Facts X must not only provide the relevant considerations being served (eg democracy), but the relevance of certain practices in relation to serving those considerations (eg an intentionalist model of statutes). It is only then that Social Facts X can provide reasons for favouring a certain model over others, in the same way that 'democracy supports an intentionalist model of statutes' does.
Recall that Social Facts X are facts about the attitudes of the system planners regarding the objectives of the system and its intended methods of achieving those objectives (ie the economy of trust). How do these facts provide reasons for preferring one set of legal facts over others? 58 Greenberg (n 2) 187 (emphasis added).

Again, by consulting conceptual fact (3), Dorothy understands that Social Facts
X are facts about the objectives of a legal system and its intended methods of achieving those objectives (ie the economy of trust) from the perspective of the system's planners.
For example, it could be that, from the perspective of the system's planners, the aim of the law is to foster and facilitate economic activity, and therefore rigid contractual rules that promote certainty serve that purpose. Social Facts X thus provide not only the relevant considerations (the maximisation of economic activity), but also the means of serving those considerations (rigid contractual rules that promote certainty). They therefore serve the same function as value facts in providing reasons for favouring one model over another.
Dorothy now has everything she needs to extrapolate the correct set of legal facts from Social Facts Y. The correct set of legal facts to extrapolate from Social Facts Y is just the set of legal facts that are best justified according to Social Facts X -the objectives of the legal system and its intended methods of achieving those objectives.
To put it in terms of her example, the answer to the question of whether the law permits the use of extrinsic materials in interpreting contracts (the legal fact) is determined according to the model of extrapolation from what was said and done by parliamentarians and judges (Social Facts Y) that best accords with the system's planners' view that the maximisation of economic activity is best served by rigid contractual rules that promote certainty (Social Facts X). What the legally correct answer is will depend on the specifics of Social Facts Y which I have left unspecified in the example, but the point here is that Dorothy now has the resources she needs to determine the legally correct model. Both higher-order and lower-order intelligibility are thus enabled by reference to conceptual fact (3) and the rational-relation doctrine is satisfied.

D. TWO OBJECTIONS
In this section I address two possible objections to my response to Greenberg's challenge. These are not the only possible objections, but brevity precludes a wider survey.

Conceptual Facts Are Grounded in Social Facts
One objection arises from the view of conceptual analysis that I have stated above. That is, if conceptual analysis involves examining our word usage, it seems likely that conceptual facts are made up of nothing more than social facts about our word usage or our dispositions to use certain words under certain conditions. Therefore, we have simply returned to the same problem: social facts are once again being used to determine their own relevance. Plunkett addresses this objection in his Reply: Recall Greenberg's argument that social practices cannot determine their own relevance. Because of this argument, one might want to argue that views of concepts as complex facts about dispositions will not work for the positivist in responding to the argument in How Facts Make Law. In short, the argument here would be that since social facts (such as facts about dispositions) cannot "determine their own relevance," it follows that conceptual facts that are equivalent to facts about dispositions cannot help explain the relevance of legal practices to legal facts. 59 As a preliminary point, it may not necessarily be the case that concepts are metaphysically reducible to social facts about our word usage just because we discover our concepts by studying our word-usage. In other words, just because epistemic access to conceptual facts is only available through an examination of social facts, it does not necessarily follow that conceptual facts are thereby metaphysically constituted by social facts.
Nevertheless, given the close connection between studies of word usage and conceptual analysis in my argument, it is at least reasonable to hold such a view of conceptual facts, and so for the sake of argument I will grant the point that conceptual facts are indeed metaphysically constituted by social facts.
The initial response to this objection is that it does not stick insofar as it makes the more specific claim that certain social facts cannot determine their own relevance to legal facts. Greenberg's argument is sometimes framed this way: that law practices themselves cannot adjudicate between ways in which those practices contribute to the law. 60 If that is the objection, then the Planning Theory's model of meeting the rationalrelation doctrine above has parried it, for there are in fact two sub-groups of social facts: Social Facts X, being the dispositions of the system's planners towards the objectives of the system and the means of achieving them, and Social Facts Y, being social facts about what was said and done in the planning process ie law practices. Social Facts Y do not determine their own relevance to legal facts. Instead, Social Facts X provide 59 Plunkett (n 5) 188. 60 Greenberg (n 2) 178. independent standards which determine the relevance of Social Facts Y. Plunkett makes this point in his Reply: More broadly, we can distinguish between the following two claims: (1) there are certain social facts that both (a) ground legal facts, and (b) determine the relevance of those social facts; versus (2) there is a set of social facts that (a) ground the legal facts, and another set of social facts that (b) make it that those token social facts are relevant to constituting the legal facts. It is not clear to me why proponents of conceptual facts who see them as grounded in social facts need be committed to the first schema. Put together, this means that Greenberg's argument that legal practices cannot determine their own relevance is not sufficient by itself to establish that appealing to conceptual facts (where, in turn, one understands concepts in terms of dispositions) undermines one's ability to develop an effective response to the argument in How Facts Make Law. 61 However, a counter-response arises -we may have avoided the specific framing of Greenberg's objection, but not the wider-framed objection that social facts alone cannot determine the relevance of any or any other social facts to legal facts. That is to say that if the relevance of Social Facts X and Y to legal facts is determined by conceptual facts, and conceptual facts are in turn determined by further social facts about word usage ('Social Facts Z') then we have simply pushed the problem further back. What determines the relevance of Social Facts Z?
Recall that the rational-relation doctrine does not preclude the reliance on social facts in all relationships of determination, but rather in relationships of rational determination. 62 That social facts are normatively inert is a problem where the relationship of determination is reason-based. On the other hand, if the relationship of determination is not reason-based but brute metaphysical, then it does not matter whether social facts are normatively inert. For example, it is not a problem for a chemist seeking to give a constitutive account of water that the determinant fact of water, a chemical composition of H2O, is normatively inert. He does not need to provide reasons why H2O is water. It simply is without further reasons. On the other hand, it is a problem for a positivist seeking to give a constitutive account of law that its determinant facts, social facts, are normatively inert, because the relationship of determination is reasonbased, and social facts, being normatively inert, cannot provide those reasons.
With this in mind, it is not clear that the relationship between conceptual facts and its determinant Social Facts Z is reason-based, as opposed to brute metaphysical.
If concepts are the constituent component of our thoughts that words express, and the content of a concept is determined by the set of conditions under which we use certain words, then it is reasonable to think that the relationship between Social Facts Z, social facts about our word usage, and conceptual facts is simply brute metaphysical. That is to say conceptual facts obtain as long as enough people use a word enough times in specific conditions. If this is true, then the fact that Social Facts Z are normatively inert is not a problem for the legal positivist who appeals to conceptual facts. Social Facts Z simply determine conceptual facts without further reasons.

Extraction requires value facts
Another objection may be raised that there are some value facts secretly doing the work in the process of extraction. Recall that the Planning Theory of Meta-Interpretation requires one to determine the objectives and the economy of trust of a legal system by interpreting social facts, including the institutional history of the system. The objection might be levelled that there is some value fact doing the work here -why are social facts about the dispositions of the system's planners relevant to determining what the law is now? This objection is understandable given that the Planning Theory of Meta-Interpretation bears resemblance to antipositivist interpretivist theories. For interpretivists, the institutional history of a system is relevant to determining present legal facts because of the moral concern arising from the fact of state coercion. 63 Therefore, the objection is that unless the positivist is able to provide an account of why the institutional history of a system is relevant without appealing to value facts (which would seem to be a natural fit) the Planning Theorist is probably secretly relying on some value facts to do the work.
The solution is to once again appeal to conceptual fact (3). If laws are plans or plan-like norms whose function is to guide and organise the conduct of members of a community over time and across persons, then it is a matter of conceptual necessity that the question of what the plan is must be determined in accordance with the attitudes and dispositions of the system's planners in the first place. In other words, it is inherent in the concept of a plan that one looks backwards to the institutional history of the planning process to determine how the plan was intended to guide and organise conduct 63 Ronald Dworkin, Law's Empire (HUP 1986)  in any given situation. Therefore, to the extent that one appeals to value facts to determine what the law is, one commits a conceptual error. The Planning Theory therefore has a plausible account of the relevance of institutional history to the determination of legal facts without appeal to value facts and so the objection is met, at least to the extent that our response to the first objection about the nature of conceptual facts holds.
A follow-up criticism might then be raised that, even if institutional history is relevant by virtue of conceptual facts, extraction nevertheless seems to require reasoning about value facts -specifically, the dispositions of the system's planners towards the values of the system. To respond to this criticism, it's helpful to explain how Shapiro understands this process of determining the objectives and economy of trust underlying a legal system: 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. 64 Therefore, in extraction, one is determining what kind of values the system's planners must have in order to explain their influence on the institutional structure of the system. Undoubtedly, this involves inquiry about value, but it does not follow that such explanations are only possible by reference to genuine value facts. To return to my earlier example, Dorothy need not have any particular attitude towards wealthmaximisation for her to determine that the system's designers aimed at wealthmaximisation. Of course, she needs to already know what wealth-maximisation is in order to identify it as an objective of the system, but it does not follow that she is thereby appealing to a genuine value fact. Wealth-maximisation, to her, may simply be a concept, rather than a value fact. Similarly, one need not treat Aryan superiority as a value fact in order to determine that the Third Reich aimed at Aryan superiority; one need only consult the concept of Aryan superiority in order to arrive at that explanation.
Value facts are unnecessary in the inquiry.

E. CONCLUSION
I have demonstrated above through Shapiro's Planning Theory, and specifically the Planning Theory of Meta-Interpretation, that conceptual facts are capable of satisfying the rational-relation doctrine, and that therefore positivism is able to meet Greenberg's challenge in How Facts Make Law.
It is interesting to note the broad structure of Greenberg's challenge -it begins from an account of legal reasoning and ends with conclusions about the nature of law that are inconsistent with legal positivism. What this reveals is that the metaphysical question of the nature of law is deeply connected to the epistemological question of how we as rational beings know what the law is. Therefore, it seems that any constitutive account of the nature of law must also provide a plausible account of legal reasoning. 65 Though it remains to be seen whether an appeal to conceptual facts withstands further scrutiny, I would tentatively conclude that conceptual facts can play a crucial role in a positivist account of legal reasoning -one that can equally accommodate our intuitions about legal practice and what legal reasoning entails.
The rational-relation doctrine also throws light onto an interesting relationship between the intelligibility of law and the normativity of law, in that the way one accounts for the former has implications for the way one accounts for the latter. If, as Greenberg suggests, the correct model of determining legal facts is the one that best accords with independently true value facts (eg democracy favours an intentionalist model of statutes) then the fact that a proposition has been determined to be legally correct itself provides reasons for action for the individual legal reasoner. 66 On the other hand, if the correct model of determining legal facts is the one that best accords with the system's planners' views on the objectives of the system and the intended means of achieving those objectives, then determining the correct set of legal facts does not, without more, provide reasons for action. In other words, just because Dorothy has settled the question of what the law is does not settle the question of what she should do. Instead, the law only provides reasons for action from a certain 65 On this note, one reason to favour the Planning Theory over Hartian positivism is its ability to plausibly account for Dworkin's theoretical disagreements. Under the Planning Theory, theoretical disagreements are understandable because determining the criteria of legality in a legal system requires reflection on the values underpinning it, ie extraction, which legal officials can legitimately disagree on. However, the Hartian positivist cannot explain theoretical disagreements because the criteria of legality, the rule of recognition, is necessarily the subject of continuing consensus. 66 This proposition is one that he develops more fully in Mark Greenberg, 'The Moral Impact Theory of Law' (2014) 123 Yale Law Journal 1288. perspective, that is, the perspective of the system's planners. This is what Shapiro calls 'the legal point of view': According to the Planning Theory of law, to say that X is legally obligated to do A (where "legal" is used perspectivally) is to say that from the legal point of view, X is obligated to do A. The legal point of view purports to represent the moral point of view. It states that the norms of the legal system are morally legitimate and binding. Since norms of the legal system are plans, the legal point of view claims that the plans of the system are morally legitimate and binding. 67 This would situate the debate about the intelligibility of law within the wider inquiry into the normativity of law. If it is true that one must appeal to value facts to even determine what the law is, then it is difficult for the positivist to maintain that the existence of a legal fact is independent of the merits of its content, and therefore that legal obligations do not necessarily track genuine moral obligations. Greenberg's challenge therefore reveals a bridge not only across epistemology and metaphysics, but also metanormative theory, one that might prove fruitful for attempts to integrate various questions in analytical jurisprudence.