1 Introduction

Like many other debates in analytic jurisprudence, the discussion about whether law exists and how we can account for its existence is very much alive—although, admittedly, not in very good shape. As has been the case in almost every other area of philosophy, the naturalistic ideology has long been the dominant view in jurisprudence. For example, some prominent legal theorists, e.g., Ross (1959), Moore (2002), and Leiter (2007), have proposed that answering ontological questions about law requires engaging in ‘serious’ metaphysical work to establish how it fits within the natural world. We may call this an inflationary view. Yet, while this line of thought has attracted the attention of more contemporary legal scholars (including those, e.g., Gizbert-Studnicki (2016) and Marmor (2019), who are now interested in providing a ‘metaphysical explanation’ of how law is grounded in more fundamental or basic entities), we should be aware that it comes with some major issues (e.g., sceptical challenges related to the ‘metaphysical principle’ that we need to use in order to establish the ‘naturalisation’ of our legal reality). As a way to resist this orthodoxy and avoid falling prey to these challenges, I consider here a ‘deflationary’ alternative.

Based on Amie Thomasson’s meta-ontology, I suggest applying the familiar methods of conceptual analysis and empirical investigation to answer whether legal entities (e.g., legal systems, legal rules, legal officials, legal activities, etc.) exist. I do this by following her ‘easy’ methodology, viz., well-formed ontological questions (e.g., are there forks? or are there chairs?) are easily answered by clarifying—via conceptual analysis—the application conditions of the corresponding sortal terms (e.g., ‘fork’ and ‘chair’) and determining—via empirical investigation—whether they are satisfied (see Thomasson, 2009a).

Though Thomasson’s easy ontology has been widely discussed in other areas of philosophy, including social ontology (see, e.g., Garcia-Godinez, 2023), little attention has been paid to its applicability to legal philosophy. The only exception are those legal scholars interested in elucidating the artifactual nature of law; in particular, Burazin (2018) and Ehrenberg (2018). Yet, my suggested approach differs from their projects both in providing a comprehensive analysis of legal entities (not only of legal systems) and in connecting it with an expressivist account of legal discourse (the purpose of which is to demystify the normative character of certain legal statements).

To present this deflationary approach to legal ontology, I structure the paper as follows. I start by introducing the topic in terms of the ontological question about legal entities (in § 2). Then, I focus on one inflationary proposal, highlight some of its extant problems, and motivate the search for an alternative (in § 3). Next, I introduce (in § 4) the deflationary approach to legal ontology, making clear the advantages of taking Thomasson’s easy route. I continue (in § 5) by building on her recent work on modal normativism to elaborate on the way in which we can elucidate legal normativity based on an expressivist account of legal discourse. Finally, I conclude (in § 6) with a brief summary.

2 Legal (meta)ontology

What is legal ontology? Briefly, legal ontology is the study of the legal world. Like social ontology, which “is the study of the nature and properties of the social world” (Epstein, 2018), legal ontology is concerned with analysing those entities that constitute legal reality. So, as Moore puts it: the point of legal ontology is “to enquire into the sort of things legal that we may plausibly suppose to exist” (2002: 620); for example, legal objects (e.g., laws and contracts), legal events (e.g., prosecutions and jury trials), and legal relations (e.g., rights and obligations). However, as Moore (idem) also asks, if such things exist, how can we ‘picture’ or account for their existence?

Following Eklund (2006), we can say that when we ask this kind of question, we get into a meta-ontological level of enquiry; that is, we do not ask (the first-order question) what is there? but rather (the second-order question) what is the nature of those questions about what there is? Thus, for instance, the meta-ontological question about artifacts is not (e.g.) are there forks? but instead what are we asking when we ask ‘are there forks’? And similarly, the meta-ontological question about law is not (e.g.) are there legal systems? but what are we asking when we ask ‘are there legal systems’? To put it succinctly, then, while legal ontology is about legal entities (i.e., what there legally is), legal meta-ontology is about the methodology, epistemology, and semantics of legal ontology (i.e., how we can come to decide, know, and talk about what there legally is).

Whereas most proposals about these kinds of second-order questions in legal philosophy are not explicit about engaging in meta-ontological issues, we can understand their views precisely as offering one approach or another to legal meta-ontology. However, by subscribing to some form of inflationary view (viz., that answering ontological disputes requires employing a certain ‘robust’ metaphysical principle), they introduce further challenges to the prospects of common-sense jurisprudence (viz., the need to engage first in serious metaphysical work in order to then be able to account for the existence of legal entities). So, with the intention of releasing legal philosophers from this pressure, I submit here a ‘deflationary’ alternative (along the lines of Thomasson’s easy ontology) which can help us answer ontological questions about legal reality without appealing to any robust metaphysical principle, and so without falling prey to any of such sceptical challenges.

3 An inflationary approach to legal ontology

Following important developments in metaphysics, some legal theorists have proposed interesting accounts of legal meta-ontology (e.g., Ross, 1959, Moore, 2002, and Leiter, 2007). Yet, while differing in various, significant respects, they all have in common a background assumption, viz., answering ontological questions about law requires adopting an inflationary approach to legal ontology; in particular, employing the ‘robust’ metaphysical principle of ‘scientific method’ to decide what there legally is. To understand how this assumption introduces further challenges to common-sense jurisprudence, I briefly present some of their core claims.

To begin with, let me clarify what I mean by an ‘inflationary’ approach to ontology. As Thomasson (2015) explains, the difference between inflationary and deflationary approaches to ontology lies in the principles that they use to answer the ontological question are there Ks? (where ‘Ks’ can be artifacts, propositions, fictional characters, numbers, subatomic particles, etc.). Thus, if deciding whether there are Ks requires employing a metaphysical principle, e.g., Ks exist if and only if Ks are posited by our best scientific theory, or are mind-independent, or have a real nature, or have irreducible causal powers, etc., then the proposed view is committed to a meta-ontology that takes existence questions to be a matter of ‘metaphysical discovery’. For example, if one has strong naturalistic (or even physicalistic) preferences, to answer whether forks or sport teams exist, one needs to discover first if such things are crucial for a scientific theory to provide the best explanation of the world (or if their compound nature, rather than their individual parts, are causally potent).

But this kind of methodology has important shortcomings. For instance, it cannot help us account for the existence of a great number of common-sense entities (e.g., artifacts and social groups) as they do not satisfy those principles. So, as Thomasson (2015: 115–121) argues, since the proposed principles do not work as general or ‘across-the-board’ criteria of existence, this gives us a reason not to approach ontology from an inflationary perspective. But how is it then that some legal scholars insist that we should commit ourselves to a ‘robust’ metaphysical principle in order to answer ontological questions of law? Why, that is, should we believe that legal ontology is a matter of metaphysical discovery?

To illustrate some of the difficulties we run into when adopting an inflationary perspective to legal ontology, let me consider one prominent case, viz., legal realism. Without going into so much detail about this view, I want to show that its commitment to a ‘robust’ metaphysical principle (viz., scientific method) introduces unattainable challenges for the legal theorist, thus making it more appealing for them to go deflationary. Let us see.

When defending what he calls ‘jurisprudential realism’ against ‘jurisprudential idealism’, Ross says that the former’s underlying thought is “the desire to understand the cognition of law in conformity with the ideas of the nature, problems and method of science as worked out by modern, empiristic philosophy” (1959: 67). Moreover, since he thinks that all science is concerned with natural facts, he takes “all scientific statements about reality [to be] subject to experimental test” (idem).Footnote 1 But why should we think that jurisprudence is constrained by such a naturalistic conception of science? For Ross, it is because the ‘principle of verification’ characteristic of empirical domains also applies to legal science, which “means that the propositions about valid law must be interpreted as referring not to an unobservable validity or “binding force” derived from a priori principles or postulates but to social facts”; in particular, “the actions of the courts” (1959: 40).

One problem with this view, though, is the fundamental presupposition that statements of law are representational, and so should be tested against an ‘observable’ background. But representationalism, as Price (2013) has extensively argued, is a serious handicap for naturalism. For example, when introducing ‘the placement problem’, he says: “If all reality is ultimately natural reality, how are we to ‘place’ moral facts, mathematical facts, meaning facts, and so on? How are we to locate topics of these kinds within a naturalistic framework, thus conceived?” (2013: 6). Essentially, there are two options.

  1. (A)

    If one accepts naturalism with representationalism (like standard legal realists, such as Ross, Leiter, and Moore, do), then there is no other choice but trying to squeeze those facts into the category of natural facts (including ‘psychological’ or ‘behaviouralist’ interpretations thereof). Yet, sticking to representationalism forces the naturalist to respond first to some persistent and serious challenges; for example, the naturalistic fallacy, the problem of specifying (without over- or under-inclusiveness) the notion of ‘natural facts’, the apparent commitment to some form of scientific idealism, the consequent irrelevance of metaphysics if everything is meant to be settled by science, etc. So, unless legal philosophers are prepared to undertake this Herculean task, they better consider the second option.Footnote 2

  2. (B)

    If one accepts naturalism without representationalism (as Price does), then one can reject (e.g.) that moral or mathematical statements are about or correspond to natural facts. However, by giving up the representational presupposition, legal realists will also give up the thesis that ‘the principle of verification’ applies to jurisprudence, and so that the scientific method is decisive for determining what there legally is. In other words, the non-representationalist legal realist will not be committed to the ‘robust’ metaphysical principle characteristic of ontological naturalism, and so will not vindicate the use of metaphysical tools to ‘discover’ legal reality.

So, what should the legal realist do? To put it simply, if the legal realist wants to account for legal reality without having to engage first in gruelling metaphysical debates, so that they can focus instead on more practical and pressing issues, they should opt for (B) and consider a deflationary alternative to ontological naturalism. However, since the point generalises to any other inflationary approach, the overall conclusion here is that any view committed to a ‘robust’ metaphysical principle (be it scientific method, mind-independence, causal efficacy, etc.) imposes on itself significant metaphysical challenges that prevent it from answering straightforwardly whether Ks exist. Thus, as Thomasson (2009a) suggests, our best way to get out of this predicament is to take the easy route.

Though before going down this path, let me clarify one point. Given the recent literature on ‘metaphysical explanations’ (rather than ‘metaphysical principles’) that is now thriving in metaphysical debates about legal reality, someone might ask whether I should consider this trending discussion before moving to the deflationary approach to legal ontology. Yet, while this might sound like a reasonable demand, I will resist it for two reasons. First, that these projects are different; and second, that ‘metaphysical explanation’ is not necessarily incompatible with a deflationary meta-ontology. Let me elaborate.

As to the first: whereas I am interested here in answering how can we account for the existence of legal entities?, the proponents of metaphysical explanations about law are rather concerned with answering what grounds legal facts? (see, e.g., Gizbert-Studnicki, 2016 and Marmor, 2019). But crucially, as Schaffer (2009: 347) tells us, the grounding question (what grounds what) presupposes existence (what there is). So, in other words, the grounding project starts by presupposing an ontology and then goes on to explain its fundamental structure. Thus, when it comes to legal reality, the grounding approach (which assumes that there are legal entities and seeks then to determine their ‘grounds’) comes apart from those other projects in legal meta-ontology (including this one) which aim precisely to help us account for the existence of legal entities.

As for the second: given the contested nature of ‘grounding’ in metaphysics, we may be relieved to know that it is also possible to adopt a deflationary perspective to ‘metaphysical explanation’ (along the lines of, e.g., Locke’s, 2020 normativist approach), and then reject the assumption that metaphysical (non-causal) explanatory claims, e.g., grounding claims, are about worldly metaphysical facts. However, while I think that this is a promising project, I leave its examination for another time. In what follows, I only consider a deflationary approach to legal ontology (not to metaphysical explanations of legal ontology).

4 A deflationary approach to legal ontology

As mentioned at the outset of this paper, I want to present here a deflationary approach to legal ontology. The motivation behind this project is to offer an alternative to inflationary views. After noting above some problems with endorsing a legal meta-ontology (viz., legal realism) committed to a robust metaphysical principle (viz., scientific method), I will now consider Thomasson’s easy ontology.

In order to show how it can help us answer ontological questions about legal entities without forcing us to accept a ‘robust’ criterion of existence, I will start by introducing (however briefly) some of its main elements. Through many works (but especially in her 2007, 2008, 2009a, and 2015), Thomasson has developed a deflationary approach to existence questions about abstract and concrete objects (e.g., numbers, propositions, fictional characters, properties, social groups, and artifacts).Footnote 3 Without going into all the details, let me just outline the core theses of her approach.

4.1 Thomasson’s easy ontology

Unlike those other views which suggest that there is a robust or substantive criterion of existence (e.g., being posited by our best scientific theories, being mind-independent, having irreducible causal powers, etc.) based on which we can decide whether there are entities of a certain kind (e.g., numbers, propositions, artifacts, etc.), Thomasson’s approach tells us that we can resolve well-formed ontological queries simply by means of conceptual analysis and (in some occasions) empirical investigation (2015: 20).Footnote 4 To put it simply, the purpose of the easy approach is to provide a non-inflationary methodology, epistemology, and semantics of the first-order question are there Ks? Now, to understand the potential of her approach, let me introduce both her existence (E) and reference (R) principles.

According to Thomasson’s view, there is a semantic relationship between ‘exist’ and ‘refer’, such that we can move up and down from reference claims (at the meta-language level) to existence claims (at the object-language level) without changing their truth value. For example, if it is true that our term ‘K’ refers, then (via semantic descent) it is also true that Ks exist. Thomasson (2008: 65) makes this explicit by introducing the following principle:

(E) Ks exist if and only if ‘K’ refers.

So, if there is some x to which we can refer by ‘K’, then we have that Kx exists, which can be read (in quantificational terms) as ∃x (Kx) (2008: 67). However, while (E) says what it is for Ks to exist, it does not say what it is for ‘K’ to refer. Taking ‘K’ as a sortal term (by means of which we can refer to entities of a certain kind), then it must be associated with certain semantic rules of use, a subset of which establish application conditions (i.e., conditions for the correct application of the term on a particular occasion). Thomasson (2008: 67) expresses this requirement in terms of a second principle:

(R) ‘K’ refers if and only if the application conditions for ‘K’ are fulfilled.

Importantly, to determine whether those conditions for the application of ‘K’ are fulfilled, we only need to carry out both conceptual analysis (to clarify what those application conditions are) and empirical investigation (to find out if there is something that satisfies such conditions).Footnote 5 Thus, given the analytic entailment between these principles, we can see how the easy approach is indeed very straightforward, viz., by (E) and (R) we have that existence claims amount simply to true conceptual and empirical claims (i.e., claims about the satisfaction of application conditions) (Thomasson, 2008: 74).

Based on these semantic principles, Thomasson then goes on to introduce what she calls ‘transformation rules’ (2009a: 5). A transformation rule, in general, says that if certain conditions are met, then there is an entity of kind K (e.g., if there are particles arranged tablewise, then there is a table). So, by following this rule, we can analytically get the existence of Ks (e.g., tables) from the reference of the sortal term ‘K’ (e.g., the satisfaction of the application conditions of the sortal term ‘table’) (2009a: 14). In other words, the transformed sentence ‘there are Ks’ or ‘Ks exist’ can be derived analytically from the sentence ‘the application conditions of ‘K’ are met’ (2007: 163).

As made out from the very beginning of its development, Thomasson’s account follows from Carnap’s deflationary view about existence: “Existence claims must be made using a language, and […] must involve using the meaningful terms of that language with their extant application conditions” (2009a: 7). Now, to appreciate how this methodology can help us answer ontological questions about legal entities, I presently submit an easy legal ontology. Then, in the next section, I will elaborate further on how this account can also help us deal with more practical and substantial issues in jurisprudence, viz., the normative character of certain legal statements. For this, I will introduce (in § 5) a functionalist approach to legal normativity.

4.2 Easy legal ontology

The purpose of this subsection is to argue for the following thesis:

Th1: Legal entities, such as legal objects (e.g., legal norms), legal events (e.g., jury trials), and legal relations (e.g., rights and obligations) exist, and we can easily account for their existence by means of conceptual analysis and empirical investigation.

Before discussing it any further, I should highlight here that Th1 is not an ontological, but a meta-ontological thesis. It tells us what we need in order to answer existence questions about legal entities, not what legal entities are. So, regardless of their nature, the easy approach is meant to help us account for their existence.

To understand Th1, allow me first to clarify the easy approach with an example. Let us suppose that the semantic rules associated to our sortal term ‘fork’ include “an object with two or more prongs used for lifting food to the mouth or holding it when cutting”.Footnote 6 As competent speakers, we can use the term ‘fork’ to refer to an object that satisfies the application conditions mentioned in those semantic rules (viz., an object with two or more prongs suitable for lifting food). So, if there is such an object, then (by R) we can correctly refer to it and thus (by E) we can truly say that a fork exists.

In the legal context, I argue, something similar is the case. Legal terms (e.g., ‘judge’, ‘contract’, ‘trust’, ‘trial’, etc.) are sortal terms associated with application conditions. So, according to the semantic principles from the easy ontology, if those application conditions are satisfied, then the legal term refers, and so (via semantic descent) there exists a legal entity (e.g., a judge, a contract, a law, a trial, etc.). For example, let us imagine that, within a certain legal framework, for someone x to be a judge, x must satisfy conditions Ci… Cn. So, if anyone, say, John, satisfies those conditions, then the term ‘judge’ will correctly apply to John, thus making it analytically true that there exists a judge (within the relevant legal framework).Footnote 7

The crucial presupposition here, of course, is that legal terms are sortal terms that purport to refer (within a highly institutionalised context) to objects, events, relations, etc. But how plausible is this presupposition? Like many other (natural or formal) languages, legal language works as a general framework for talk and thought, and so it helps us organise the way in which we come to express and think about certain aspects of our reality. As such, it is by way of establishing conditions of use (at the meta-language level) that legal terms can refer to such aspects (at the object-language level). Though, admittedly, this is not always an easy task. In fact, in legal philosophy we are already acquainted with two hard problems concerning this linguistic role of legal terms.

On the one hand, we have problems of interpretation; and, on the other, problems of evidence. The first category concerns questions about the semantic rules associated to a particular legal term. For instance, despite our best efforts in articulating the exact meaning of the term ‘trust’, it may not be clear whether ‘being established by deed’ is part of its application conditions. So, in order to use the term appropriately, we will have to clarify its meaning though a certain form of conceptual analysis. Before elaborating on how to do this, though, let me notice that legal terms, as MacCormick (1998) maintains, are institutional terms, the application conditions of which are formally articulated by ‘institutive’ rules.Footnote 8 As such, then, the conceptual analysis involved here may be constrained by formal processes and mechanisms that are not common to more ordinary conceptual inquiries.

To explain the role of conceptual analysis in elucidating the semantic rules (particularly, the application conditions) of the legal terms, let me introduce the distinction between descriptive and normative conceptual work. As Thomasson (2012) argues, conceptual analysis is necessary for determining conceptual truths (e.g., the semantic content of a sortal term). Yet, this kind of analysis is not meant to be limited only to stating or describing what the actual rules of use associated to a certain term are, for it can also be about negotiating and deciding which rules they should be. In offering a comprehensive account of both, Thomasson has developed a descriptive but also a normative approach to conceptual work. In the latter case, she has advanced a ‘pragmatic’ view.

Simply put, Thomasson argues that when it comes to deciding whether or not to introduce a term into our linguistic framework, we should expect the answer to be linked to pragmatic considerations; particularly, regarding its function within the language (2017b: 374). Thus, for example, to answer the normative conceptual question about whether and how to use the term ‘fork’, we should look at its linguistic function, which may in turn involve addressing some further issues:

Why is it useful to have the relevant term in our vocabulary (or concept in our repertoire)? What role does this concept play (perhaps along with allied terms and concepts) in our overall conceptual system? What we would be missing if we lacked such a term or concept? What did having that concept do for societies that enabled them to carry on and reproduce their conceptual system, including use of the concept at issue? (2017b: 374, n. 8)

Clearly, although responding to these questions is not always easy, there is still no reason to take those issues to be a matter of any ‘metaphysical’ discovery. Indeed, we only need to recognise, as Thomasson does (2021), that sortal terms (along with their associated semantic rules of use) are human creations, and so are susceptible of change and further specification. However, while these are already common beliefs within conceptual engineering scholarship, other areas in philosophy have yet to catch up. For instance, in the case of legal philosophy, some people (e.g., Dickson, 2001, Raz, 2009, and Shapiro, 2011) believe that the concept of ‘law’ (and perhaps other legal terms) tracks essential or necessary properties, without realising how problematic this is.Footnote 9

But assuming that legal terms do not have their referents fixed once and for all, does not mean that just anything can be their referents. We need to agree first on what they are meant to refer to before using them to evaluate whether there is anything that satisfies their corresponding application conditions. But how can we come to this agreement? Recent literature in conceptual engineering supports the idea that conceptual disputes (e.g., about the concept of ‘race’, ‘gender’, and ‘disability’) can be solved by meta-linguistic negotiations (e.g., Plunkett & Sundell, 2013; Plunkett, 2015; Belleri, 2017; Thomasson, 2017a). Without delving so deep into these waters, let me simply mention the general point. A meta-linguistic negotiation is “a metalinguistic dispute that concerns a normative issue about what a word should mean, or, similarly, about how it should be used, rather than the descriptive issue about what it does mean” (Plunkett, 2015: 828). In jurisprudence, a seminal version of this kind of methodology has already been suggested by Bix (1995: 469): “Some of the disagreements present within conceptual “debates” might be better understood as disagreements regarding the best answer to a particular question or problem”. So, as he says, while there may not be a ‘right’ or ‘wrong’ conceptual definition, there may be one that is ‘more or less useful’ or ‘more or less convenient for certain purposes’ (1995: 470).

Now, to recap, while important, the problems of interpretation that accompany legal terms can be dealt with by carrying out conceptual analysis, where this may involve both descriptive and normative conceptual work. So, for example, to answer the conceptual question what are the application conditions of ‘trust’ or ‘judge’ or ‘trial’? the interpreter can survey the actual linguistic practice (i.e., how competent speakers use the terms in the relevant circumstances) but can also engage in a meta-linguistic negotiation and discuss the various reasons (e.g., political, moral, scientific, etc.) in favour of one or another alternative meaning. Though, in either case, the expected outcome will be the determination (however temporary and refutable) of the conditions for the correct application of the term on a particular occasion.Footnote 10

But even if we manage to establish the semantic rules that constitute the application conditions of a legal term, we may still have another challenge to meet, viz., solving specific problems of evidence. As mentioned above, this is the second category of problems that emerges from the linguistic role of legal terms, viz., to refer to legal entities. I briefly comment on it here.

In his now classical discussion about legal reasoning in adjudication, MacCormick (1994) elaborates on the conceptual limits of deductive justification and mentions that, when it comes to the application of rules, ‘findings of facts’ (e.g., whether the killing of someone was intentional, or whether the signing of a document was done under no coercion, or whether an election or appointment complied with the formal procedure, etc.) are usually contested matters. Yet, to solve these issues, which have to do, as he says, with “the question ‘Is r, s, t an instance of p for the purposes of applying if p then q?” (1994: 95), we only need to set a standard or process of proof that can be used within the legal context for deciding if, and to what extent, certain empirical propositions are true. In his words: “The process of ‘proof’ is a process of establishing that certain propositions are for legal purposes to be considered true, specifically for the purpose of the litigation in question” (1994: 27).

Though mostly intended to be a criterion of truth in the courtroom, the idea of ‘standard’ or ‘process’ of proof can be understood more generally as a criterion for determining whether the application conditions of a legal term (e.g., ‘homicide’, ‘contract’, ‘judge’, ‘trust’, etc) are satisfied. Of course, any respectable standard or process of proof will introduce further complexities to the easy legal ontology as it will likely include, amongst others, logical, mathematical, and even scientific principles (e.g., consistency, probability, and testability).Footnote 11 However, the crucial point of the easy approach to legal ontology is not threatened by this; what makes the easy approach deflationary is not denying that empirical methods are indispensable for determining what there is, but rather denying that such methods correspond to ‘metaphysical’ principles in a way that we need to discover first some metaphysical truths before addressing any other, more mundane issue (e.g., whether John was truly appointed as a judge).

What matters for the easy legal ontology instead is simply that there is a way for us to decide whether something (be it an object, an event, a relation, etc.) satisfies the application conditions of a legal term. That is, to the extent that we have in the institutional, legal context a mechanism to discern ‘true’ from ‘false’ empirical claims, we can run without difficulty this deflationary methodology. Moreover, as Thomasson suggests, it is precisely in this kind of context that we can easily get to know what we need: “if we know that the legislature has ever voted to approve a bill, we know there is law; if we know that individuals have ever fulfilled the relevant requirements, we know that there is a marriage; if we know that someone has ever filled the relevant paperwork, we know that there is corporation” (2015: 102). Nothing more mysterious than confirming whether an empirical claim is true is required to make solid inferential claims (via appropriate transformation rules) about what legally exists.

Before concluding this section, let me summarise the main contribution of this discussion. After considering (in § 3) some problems with an inflationary approach to legal ontology; in particular, with the view that to answer ontological questions about legal entities we need to employ a naturalistic ‘metaphysical’ principle (e.g., scientific method), I proposed here an alternative route. Following Thomasson’s easy ontology, I suggested that to answer whether a legal entity exists, we need only conceptual analysis (involving both descriptive and normative conceptual work) and empirical investigation (as required by the relevant standard or process of proof). Yet, though they come with certain difficulties, these two elements are significantly less demanding than the inflationary methodology (e.g., legal realism). For one, they do not require the legal theorist to engage first in any ‘serious’ metaphysical discussion in order to decide if there is anything legal. While problems of interpretation and evidence will always arise, the methods for solving them are more familiar and manageable than the pressing challenges that follow from subscribing to one or another metaphysical principle.

On top of this, as I presently show, the deflationary approach to legal ontology can also help us understand the practical relevance of existent legal entities (from judges and their decisions to laws and entire legal systems). By taking on the easy legal ontology, that is, we can demystify not only the ontology but also the normative character of our legal reality.

5 A functionalist account of legal normativity

Since according to the easy legal ontology, answering existence questions about legal entities does not require any hard metaphysical work, it spares the legal theorist the time and energy to focus on more substantial questions. One of such questions concerns the normativity of law; in particular, what does it mean to say that such-and-such legal entity is normative? As an attempt to provide a unified account of legal reality, including an explanation of its normative character, I introduce here an expressivist analysis of legal discourse. Yet, to be precise, this is not a first-order analysis of legal normativity (e.g., concerning the reasons that can make a legal decision or a legal rule justified); instead, it is a second-order (or meta-ethical) analysis of legal statements.

The motivation behind this analysis is to show that we can elucidate legal normativity in a way that coheres with the deflationary approach to legal ontology submitted above. To put it concisely, my suggestion is that by taking legal entities (e.g., judges, laws, contracts, etc.) to be institutional artifacts created to fulfil a certain function, we can understand their normativity in terms of the commitment addressed to using those artifacts in the way determined by the semantic rules associated to their corresponding legal terms. So, more specifically, I argue here for the following thesis:

Th2: Normative legal statements correspond to object-language expressions of commitment towards using legal entities according to the function determined by the semantic rules associated to legal terms; and descriptive legal statements correspond to meta-language expressions of those semantic rules.

To understand Th2 and the way in which it can help us elucidate legal normativity, let me start by clarifying how it follows from the deflationary approach to legal ontology. For this, I will mention first the relationship between Thomasson’s easy ontology and her normativist account of metaphysical modality. This will provide the background to appreciate the relevant connections between meta-ontology and meta-ethics.

5.1 Thomasson’s modal normativism

Besides developing a deflationary approach to ontology, Thomasson has advanced what she calls a ‘normativist’ account of metaphysical modality (see, e.g., her 2013 and 2020b). Briefly, she argues for an anti-representationalist view about modal discourse, according to which “talk about what is ‘metaphysically necessary’ does not aim to describe modal features of the world, but rather, provides a particularly useful way of expressing constitutive semantic and conceptual rules in the object language” (2013: 143).

Against those views that take modal discourse to be about ‘real modal properties’ which are meant to be ‘discovered’ through serious metaphysical work, Thomasson argues that “we should not think of modal claims as attempting to describe modal features […] that explain what makes them true, and we should not think of modal knowledge as coming through anything like contact with modal features. Instead, we can begin by seeing modal discourse as performing a more basic function: conveying norms” (2013: 145). For example, the modal claim ‘Necessarily, all bachelors are men’ is not about any modal properties of bachelors or men, but an expression of the semantic rules associated to ‘bachelors’ and ‘men’, and the analytic entailment between them, viz., that, in any context, the satisfaction of the application conditions of the former guarantees the satisfaction of the application condition of the latter. So, in using those terms (at the object-language level), we can express modal claims, the truth of which is determined by the semantic rules that we can mention and then examine (at the meta-language level) via conceptual analysis. The normativist view, in other words, holds that metaphysical modality is not about discerning what the modal features of reality are but about expressing semantic rules (and their logical consequences) under conditions of semantic descent.

Indeed, this is a crucial element of Thomasson’s normativism as it is what explains how we can get to know modal truths without running into the common difficulties associated to descriptivist theories of modality (e.g., the problem of determining how exactly we can acquire modal knowledge if modal truths are not empirical but metaphysical). In short, she takes modal truths to be ‘discoverable’ via conceptual analysis, and so modal knowledge to be ‘acquired’ via conceptual competence, sometimes as combined with empirical knowledge (2013: 152).

In her most recent work, though, Thomasson (ms) has extended her view from modal to moral discourse. In particular, she has argued that when talking about moral facts and properties, we are not talking about ‘worldly’ facts and properties (susceptible of ‘metaphysical discovery’), but about a distinctive function of moral discourse, viz., to influence or regulate behaviour without having to appeal to a prior or formal authority. Thus, for example, we can talk about the moral fact killing innocent people is wrong, or the moral property killing innocent people has the property of wrongness, although neither of them is meant to track worldly facts or properties, nor needs ‘explanatory truth-makers’. Instead, they only express what is implicit in our moral framework, viz., that we must not kill innocent people. But there is nothing mysterious about this; we can understand this deontic modality in a way that is continuous with the way in which Thomasson understands metaphysical modality, i.e., as a form of making explicit semantic rules of use (e.g., the application conditions of the action-type ‘killing innocent people’ imply that there are no circumstances under which anyone may perform an action-token of it).Footnote 12 As I will now explain, this deflationary meta-ethics can help us account for the distinctive function of legal discourse, viz., to regulate behaviour through the normative constraints imposed by legal entities.

5.2 The normativity of legal entities

Like any other artifact framework, a legal framework has both ontological and normative import, viz., it provides competent speakers with semantic rules, not only for the use of legal terms (e.g., ‘judge’, ‘contract’, ‘trial’, etc.), but also for the use of the entities to which we refer by those terms (e.g., what are the responsibilities of a judge, what obligations follow from signing a contract, what are the stages of a trial, etc.). Though not expressed in this way, this point has already been made by those legal theorists who support the so-called ‘artifact theory of law’ (e.g., Burazin 2018, Ehrenberg 2018, and Roversi 2018). For example, when elaborating on the normativity of artifacts, Ehrenberg says:

If artifacts communicate their usage, they thereby carry a very basic normativity of recognition. That is, the very fact that something is an artifact means that it is generally created to serve a purpose and to communicate that purpose. When someone creates a token artifact of a certain kind, she thereby communicates an intention for the relevant audience to recognize it as an instance of that kind of artifact. The artifact therefore comes bundled with a norm of recognition or identification (2018: 184–185).

As I understand it, Ehrenberg’s idea here is that when people create an artifact of a certain kind (e.g., a table, a chair, a fork, etc.), they create something that is intended to fulfil a certain function, as conceptualised or specified in the relevant artifact kind (viz., ‘table’, ‘chair’, ‘fork’, etc.). As he acknowledges, this idea is based on Thomasson’s theory of public artifacts: “Amie Thomasson tells us that public artefacts […] have ‘receptive’ features that signal how the object is to be used or treated” (2016: 154). Indeed, Thomasson (2007, 2014) explains the construction of an artifact token as requiring both the existence of a type-description and the existence of an object that satisfies such a description; and then she argues that, when referring to the object by the artifact kind, the author and the target audience have (though perhaps only implicitly) the intention not only of categorising the object as an artifact token but also (and more importantly) of participating in the practice of recognising the object as having a certain function (determined by the type-description associated to the artifact kind).

Then, by utilising this theory, Ehrenberg takes himself to reveal ‘where normativity creeps into the descriptive premises’ of artifact creation; in particular, the creation of law as an institutional artifact (2016: 148). Regardless of the merits of his project, though, I agree with him that we introduce normativity into our legal reality by establishing functions for our legal entities. And so, I believe that we can elucidate legal normativity not by looking at ‘normative properties’ or ‘normative facts’, but by clarifying the intentions of those involved in the appropriate practice of recognition.

Against this background, then, I propose that we can understand the various things that we can do with our legal discourse, e.g., making normative and descriptive legal statements. In the case of normative (or internal) legal statements, they are object-language expressions of commitment towards using legal entities according to the functions established by the semantic rules associated to the relevant legal term (e.g., “judges are responsible to ensure that the accused gets a fair trial”). In the case of descriptive (or external) legal statements, on the other hand, they are meta-language expressions mentioning the function of legal entities as determined by the semantic rules associated to a legal term (e.g., “according to the concept of ‘judge’, those who hold judicial office are responsible to ensure that the accused gets a fair trial”).Footnote 13

Though I cannot say much here about the difference between ‘normative’ (or internal) and ‘descriptive’ (or external) legal statements, the importance of this distinction, which can be traced back to Hart (1994: 102–103), is a familiar topic within legal philosophy. To put it in a few words: A ‘normative legal statement’ can be a normative premise (e.g., a legal rule) that plays some justificatory role within legal reasoning, while a ‘descriptive legal statement’ can be a description of what is taken to be the case within the legal context (e.g., that a rule satisfies the criteria of legal validity). So, in this sense, whereas internal legal statements are usually uttered to trigger normative consequences, external legal statements are normally uttered to state an actual (or hypothetical) state of affairs.

To illustrate this, let me consider the following example. Let us suppose that a competent speaker says “the speed limit in Ireland is 80 km/h on regional roads”. If uttered as a normative statement, the speaker intends the audience (though perhaps only herself) to use the legal entity (viz., the speed limit rule) as a normative premise (or practical reason) in her legal reasoning. As such, the purpose of the utterance is to express a commitment towards using the entity according to the function determined by the semantic rules associated to the legal term (viz., ‘speed limit’). On the contrary, if uttered as a descriptive statement, the speaker intends the audience (though, again, perhaps only herself) to mention the legal term (with its associated semantic rules). So, in this case, the purpose of the utterance is not to express a commitment towards using the rule as a justificatory premise in legal reasoning, but to state its semantic content.Footnote 14

Yet, while it seems that only normative legal statements have practical implications, the truth is that descriptive legal statements are also practically relevant. For one, descriptive legal statements that mention (rather than use) legal terms can help us state and then negotiate their conceptual content (i.e., the semantic rules). As such, then, as we have seen in relation to the application conditions of legal terms, we can take on conceptual analysis to clarify and decide what the function of legal entities is (e.g., the function of judges, contracts, trials, speed limits, etc.). And here again it becomes crucial to appreciate the contrast (and the relationship) between descriptive and normative conceptual work. While the former can help competent speakers to clarify what the content of their legal terms is (e.g., by elucidating their semantic rules or defining their inferential profile in light of the actual linguistic use or practice), the latter can help them to decide what the content of their legal terms should be (e.g. by negotiating their semantic rules or changing their inferential profile in light of normative, such as moral or political, considerations). So, although downgraded now by most contemporary, metaphysically-minded legal scholars, conceptual analysis is (or should be) at the centre of analytic jurisprudence. In this sense, I very much sympathise with Spaak, when he says: “The analysis of fundamental legal concepts, such as the concept of law, the concept of a legal norm, the concept of a legal right, or the concept of legal validity, is a central jurisprudential task, so the question of the proper method for analysing concepts should be of considerable interest to jurisprudents” (2016: 81).

To round-up, let me just stress the general outcome of this and the previous sections. While some legal scholars prefer to engage in ‘hard’ metaphysical work to ‘discover’ not only what there legally is, but also what makes our legal reality normative, we can now see that, to an important extent, these are rather ‘easy’ tasks that require only conceptual analysis and empirical investigation. So, instead of embarking on a difficult journey through dense metaphysical jungles, the legal theorist can easily respond to ontological and normative questions about legal reality and legal discourse by undertaking both conceptual and empirical work.

6 Conclusion

Against ‘inflationary’ views about legal meta-ontology, I have argued here for a ‘deflationary’ alternative. In particular, I proposed, following Thomasson’s easy ontology, that there is no need to appeal to a robust metaphysical principle (e.g., scientific method) in order to answer whether legal entities (e.g., legal objects, legal events, and legal relations) exist. By employing the semantic principles of ‘existence’ and ‘reference’ from Thomasson’s approach, I elucidated how we can say that there is a legal entity (e.g., a judge or a contract) based solely on the satisfaction of the application conditions of a legal term (e.g., ‘judge’ or ‘contract’). Then, I considered two familiar issues related to this methodology, viz., the problem of interpretation and the problem of evidence. For the first, I utilised Thomasson’s theory of conceptual analysis (as involving both descriptive and normative conceptual work) to clarify the way in which competent speakers can acquire the relevant knowledge about the semantic content of legal terms, but also how they can solve their conceptual disagreements. And for the second, I mentioned how a standard or process of proof can help disputants to decide on the truth of an empirical claim.

In the second part of the paper, I offered a functionalist account of legal normativity. More specifically, in order to explain the normative character of our legal reality, I elaborated on Thomasson’s modal (and moral) normativism and then introduced an expressivist account of legal discourse. Here, I suggested that competent speakers can do various things by uttering legal statements. For instance, through normative (or internal) legal statements, they can express their commitment towards using legal entities according to the function specified in the semantic rules associated to the corresponding legal terms; whereas through descriptive (or external) legal statements, they can mention those rules and then engage in their conceptual analysis, including meta-linguistic negotiations. So, while the purpose of this was not to define the function or functions of legal entities (since this can only be determined through a first-order investigation), I claimed that it takes nothing other than appreciating their functional role within a general legal framework to understand how, as Ehrenberg puts it, normativity creeps into our legal reality.

In conclusion, then, instead of assuming that we must do ‘serious’ metaphysical work before being able to answer ontological and normative questions about law, we should step back (in the way that Thomasson urges) and consider the ‘easy’ alternative. With this, we get what common-sense jurisprudence needs, viz., a simple (as opposed to a ‘metaphysically explanatory’) realism about law and its normativity.