Against the Input View of Legal Gaps

1. Introduction

Are there gaps in Polish law⁴? Of course – at the very least the legal text presupposes their existence⁵. What role do they play in the judicial practice? There are two major lines of response to this question. First, and fairly uncontroversially, legal gaps function within the context of de lege ferenda proposals. To say that there is a gap in the law is to draw the legislators’ attention to a deficiency in the legal text and call for its rectification by an appropriate legislative measure. Since in this picture the legal gaps are merely results of legal interpretation, I will call it the output view of legal gaps. Second, and more controversially, legal gaps are sometimes taken to play the role of premises in judicial reasoning. In particular, the identification of a gap in the law is thought to justify the deployment of an otherwise impermissible method of interpretation. For instance, Article 300 of Polish Labour Code⁶ instructs the interpreter to draw an analogy with the Civil Code⁷ only if there is a gap in the labour law⁸. This and similar

provisions, which I will present shortly, seem to be motivated by the following considerations: on the one hand, reasoning by analogy leaves significant room for judicial creativity, thus – arguably – it should not be a generally applicable method of statutory interpretation⁹. On the other hand, the discretion inherent to the analogical reasoning is still less harmful than the complete unforeseeability generated by gaps in the law. This is easily generalized to what I shall call the input view of legal gaps, on which legal gaps trigger the application of otherwise inapplicable methods of statutory interpretation. It is my goal in this paper to show that the input view is self-defeating. My argument exploits the fact that there is no reliable way of identifying the existence of a legal gap available to the proponent of the input view. Consequently, there are cases in which the judge has discretion to decide whether they are dealing with a gap in the law or not. I assert that in some of these cases the input view actually increases the scope of judicial discretion, instead of curbing it. In the second section I introduce the input view against the background of Jerzy Wróblewski’s conception of ideologies of judicial application of law. In particular, I argue that the input view is rooted in a formalistic approach to legal interpretation. In the third section I discuss the relationship between legal gaps and judicial discretion and argue that the proponents of the input view should understand legal gaps as situations in which the judge enjoys a significant amount of discretion. In the fourth section I introduce my main argument that the input view systematically increases the scope of discretion in a certain class of cases. In the fifth section I discuss three alternative accounts of legal gaps and show that none of them help proponents of the input view evade my objection. Finally, in the sixth section I briefly discuss the implications of my argument.

2. The input view

My goal in this section is to describe the features of the input view. On a first approximation, it can be described as stating that assertions of existence of legal gaps are premises in legal reasoning in pretty much the same way as “Josef K. killed a man” is a premise in the reasoning of a judge who applies the norm “whoever kills a man is punishable by imprisonment” (therefore, Josef K. is punishable by imprisonment). Even though – as far as I know – no one has explicitly defended the input view in writing, I believe that there are good reasons to think that it is a part of “folk legal theory” in many liberal democracies whose legal culture of which is dominated by the legacy of logical positivism.

2.1. The ideology of judicial application of law

Before a more precise definition of the input view is provided, it will be useful to situate it within a broader legal-political context. In doing so, I will rely primarily on

mention legal gaps. Nonetheless, “cases unregulated by the labour law” are commonly referred to in the labour law literature as “legal gaps”. See, e.g. L. Florek, Prawo Pracy [Eng. Labour Law], Warszawa 2015, A. Wypych- -Żywicka, Refleksje nad art. 300 Kodeksu pracy [Eng. Reflections about Article 300 of the Labour Code], “Studia Iuridica Lublinensia” 2015/3, p. 102, or the decision of the Supreme Court of 1 March 2018 (III PK 18/17).

Wróblewski’s seminal work on the ideology of judicial application of law¹⁰. According to Wróblewski, “the ideology of judicial application of law makes clear and ultimately justifies how the court ought to apply the law” and “defines its general direction and the attitude of the court to the law in force”¹¹. Unsurprisingly, Wróblewski denies that any of the existing ideologies has been exhaustively spelled out. Nonetheless, he believes that we may coarsely distinguish three basic ideologies: the ideology of bound judicial decision-making, the ideology of free judicial decision-making, and the ideology of legal and rational judicial decision-making. The first two can be pictured as the opposite extremes of a spectrum of ideologies. For the sake of simplicity I will ignore the third one, which is an attempt at reconciling the most valuable insights from first two¹². It is important to emphasize that Wróblewski’s ideologies are not meant as faithful reconstructions of the views of any particular thinker or group of thinkers, but rather as somewhat exaggerated abstract models, whose primary purpose is to illuminate the main tendencies in thinking about the role of judges. For this reason, I will not try to ascribe the input view to any particular theorist, despite citing some of them while explaining how I understand it. The ideology of bound judicial decision-making has its roots in the liberal ideology of the Enlightenment. Its primary goal is to protect individual liberty from the arbitrariness of judicial decisions. Thus, the proponents of the ideology of bound judicial decisionmaking readily embrace legal positivism. In particular, they take the system of positive law to be complete and closed. Accordingly, they picture judicial decision-making as a highly “mechanical” process, where the judge’s reasoning is entirely descriptive (as opposed to evaluative). In doing so, they endorse Montesquieu’s ideal of a judge as a mouthpiece of the law¹³. The ideology of free judicial decision-making arose from the criticism of various tenets of the ideology of bound decision-making. As such, it is not affiliated with any particular political agenda, even though it had been adopted by various political movements across the twentieth century. Champions of free judicial decision-making point out that the idea of a complete and closed legal system is a phantasy, for gaps in the law – as matter of empirical fact – exist. Thus extra-legal considerations inevitably enter into almost any judicial decision. However, they do not see this as a negative phenomenon, for a value-loaded judgment of a wise judge is much more likely to be just than one stemming from unreflective adherence to the letter of the law¹⁴. In general, Wróblewski is more sympathetic towards the ideology of free judicial decision-making¹⁵. Nonetheless, he points out that, as opposed to its rival, the ideology lacks a systematic and positive account of how judges should apply the law in practice. It would thus seem reasonable to take the systematic theory offered by the ideology of bound judicial decision-making as one’s starting point and improve it by accommodating the insights offered by the proponents of the free judicial making. In fact, I hypothesize that a hybrid ideology along these lines, which combines a strong positivistic foundation with the acknowledgement of the existence of gaps in the law, is a popular

approach in contemporary liberal democracies. Of course, such a hybrid ideology can take many different forms. In this paper I would like to focus on one of them, the hallmark of which is the input view of legal gaps¹⁶. Without going into unnecessary speculation, let me roughly characterize it as the ideology of bound judicial decision making which makes the minimal possible concession to the insights of the ideology of the free judicial decision-making, just in order to make theoretical room for the existence of legal gaps. In the fourth section I will argue that this approach is self-effacing. First, however, let me identify some symptoms of this ideology. As we have pointed out, ideologies of judicial application of law are in general independent of any particular thinkers. One may thus reasonably wonder, how one should go about identifying an ideology. In meeting this epistemological challenge, I will follow Wróblewski’s lead. In his view, an ideology can be reconstructed by recovering its elements from, inter alia, the following sources: 1) legal rules formulating particular directives of judicial application of law; 2) justifications of judicial decisions; 3) general theoretical reflections concerning the application of law¹⁷.

In subsections 2.2–2.4, I examine each of those sources in turn. This will enable me to finally formulate the input view in subsection 2.4.

2.2. Legal rules formulating particular directives of the judicial application of law

Consider the following rules of adjudication¹⁸: 1) Article 1 of the Swiss Civil Code (Code civil Suisse/Schweizerisches Zivilgesetzbuch) requires that “in the absence of a provision the court shall decide according to customary law, and in the absence of customary law, in accordance with a rule that it would make as a legislator”¹⁹. 2) According to Article 4 of the Louisiana Civil Code, “when no rule for a particular situation can be derived from legislation or custom, the court is bound to proceed according to equity. To decide equitably, resort is made to justice, reason, and prevailing usages”²⁰. 3) According to Article 12 of the Italian General provisions concerning law (dispo- zioni sulla legge in generale): “If a controversy cannot be decided by means of a determined provision, one has to take into account provisions which regulate similar cases or analogous matters; if the case still remains dubious, one decides according to the general principles of the state legal order”²¹.

What they all have in common is that their antecedents make an explicit reference to a gap in the positive law. Their structure suggests that they are not just general guidelines for interpretation. Rather, they find application only in exceptional cases in

which there exists a particularly high risk of issuing an arbitrary judgment (otherwise the reference to legal gaps would be redundant). Once this is conceded, it should not be controversial to classify them as duty imposing rather than as power conferring in Hart’s sense²². Arguably, the duty they impose is twofold. First, they oblige the officials not to apply “non-standard” methods of interpretation (custom, equity, analogy, respectively) when there is no gap in the law. Second, they oblige the officials to apply “non-standard” methods of interpretation, should they identify a gap in the legal provisions relevant for deciding the case at hand.

2.3. Justifications of judicial decisions

A quick search through the database of the Polish Supreme Court’s judgments reveals that something like the distinction between standard and non-standard methods of interpretation is often present in how Polish courts think about legal gaps. For instance, in Supreme Court’s judgment of 27 February 2019 (II PK 285/17), we read “The district court explained that we are not dealing with a gap in the law because basic methods of interpretation suffice to settle the issue in dispute”²³. Furthermore, courts often explicitly state that reasoning by analogy, which is commonly considered to be a non-standard method of interpretation, can only be applied if there is a gap in the positive law. Thus, for instance in the decision of 13 June 2018 (III PZP 2/18), the Supreme Court states: “This lack of regulation cannot be classed as a genuine legal gap, which would require filling by means of an analogy with the provisions of the Labour Code”. Similarly, in the decision of 24 April 24 2018 (V CZ 22/18), the Court says: “In such a situation there are no grounds for the application by analogy of Article 5191(1) of the Civil Procedure Code²⁴, because there is no gap in the law”.

2.4. General theoretical reflections concerning the application of law

Inasmuch as the input view presupposes a  hierarchy between standard and non- -standard methods of interpretation, it meshes well with formalistic approaches to legal interpretation, the hallmark of which is the reductionism of interpretative premises²⁵. The underlying assumption of formalistic accounts is that certain interpretative methods are inherently more objective than the others and that the officials can only resort to the less objective methods if the more objective ones did not yield a complete answer to the legal question at hand. Even though one may disagree over which methods of interpretation count as formalistic and which do not, the formalists seem to be univocal in assigning priority to the linguistic methods of interpretation. For the sake of argument, I will henceforth assume that determining the lexical meaning is the sole formalistic method of interpretation. As far as I can see, nothing in my argument depends on that. In order to appreciate the need for the input view, the reader should notice that a major motivation for embracing judicial formalism is to reduce the scope of judicial discretion.

However, even the most hard-nosed formalists concede that legal sources fail to determine a unique answer to every legal question. The inevitability of such indeterminacies entails the inevitability of judicial discretion. In some cases the discretion is harmless, for instance when a defendant’s guilt has been established beyond reasonable doubt and the judge is free to determine a prison sentence anywhere in between 36 and 40 months. A standard story about these situations has it that the legislators intentionally conferred discretion on the judges in order to enable them to consider the peculiarities of the case at hand. In other cases, however, judicial discretion is believed to be a bad thing. In particular, it has been claimed that if judges have discretion in deciding about rights and obligations, then they are de facto creating retroactive law²⁶. It is precisely these cases that are invoked to justify the need to supplement the formalistic picture with the input view. In such cases, the advantage of curbing the scope of judicial discretion outweighs the price of admitting non-formalistic methods, such as analogia legis, analogia iuris, or a reference to travaux préparatoires. However, it has to be emphasized that these devices can only be used once the existence of a normative gap has been established by reference to the canon of formalistic methods of interpretation. Otherwise, it would be possible for judges to exploit the gapfilling devices in order to decide contrary to the unambiguous statutory provisions when the outcome of literal interpretation would clash with their axiological prefe rences. And this is precisely what the formalists want to avoid. Now, in order to allow for gap-filling devices while preventing judges from using them in “normal” cases, the proponent of the input view is committed to a sharp distinction between the following two stages in the process of gap-filling. First, the official faced with a legal question uses the generally accepted methods of legal interpretation to discern a normative gap within the positive law. Second, by way of an exception, the official deploys a special interpretative method in order to fill in the gap. Having said that, I am finally in position to formulate the two theses that constitute the input view of legal gaps:

Identification thesis: Legal gaps are identified exclusively by using formalistic methods of interpretation. Filling thesis: Legal gaps are filled by using a non-formalistic method of interpretation.

Two clarifications are in place. First, the sets of formalistic and non-formalistic methods are understood to be mutually exclusive. Second, even though my formulation of the filling thesis refers to an indefinite non-formalistic method, it is crucial that any instance of the input view choses a particular one, e.g. analogy. This is so because the filling thesis is considered to be an alternative to a situation in which the judges not only exercise discretion, but are also free to justify their discretionary judgement by reference to any non-formalistic sources they see fit.

3. Judicial discretion and legal gaps

In the previous section I have argued that the central motivation for adopting the input view of legal gaps is to limit judicial discretion. It would therefore not be unreasonable to expect that one’s understanding of legal gaps links the notion to the existence of judicial discretion. And, indeed, such a link is explicit in some of the most influential

treatments of legal gaps in the contemporary analytic philosophy of law. Thus, for instance, Joseph Raz defines a legal question as a question all the possible answers to which are legal statements. Accordingly, he says that a legal gap occurs if and only if the law does not give a complete answer to a legal question²⁷. Thus, in Raz’s account, every case in which the judge enjoys discretion involves a legal gap²⁸. However, this does not mesh well with the input view. For in many cases discretion is valuable and, arguably, intended. In fact, it is often pointed out that there are good rational reasons for the legislators not to settle conclusively all of the legal questions in advance. Ultimately, it is impossible for them to foresee every possible state of affairs to which a given norm may be relevant. It is thus argued that in some cases the value of flexibility outweighs that of certainty²⁹. Once we concede the practical indispensability of judicial discretion, it becomes clear that its existence by itself does not offend the rule of law. This becomes especially clear with respect to evaluative standards, the legislative precisification of which would be clearly counterproductive. Timothy Endicott aptly illustrates this point with the provision according to which the plaintiff should be granted damages sufficient to make them as well off as if the tort had not been committed:

Suppose that a successful plaintiff has suffered a moderately serious back injury. One dollar in compensation would not make him as well off as if the injury had not been caused (it would be an insult). A billion dollars would be excessive: it would exceed what is required to make the plaintiff as well off as if the injury had not happened. So how much does the legal standard require? It is quite true that disputes of damages will be formulated as competing conceptions of welfare. But the problem for understanding the law is not only that the appropriate principles of compensation in such cases are open to controversy, but also that there is no precise sum that the vague standard (on any conception) demands³⁰.

Even though such examples did not receive too much attention in the literature devoted to judicial discretion and legal gaps in particular, it seems arbitrary to deny that in determining the quantum of damages the judge necessarily exercises discretion. Indeed, even though the standard of comparative welfare seems to rule out certain answers to the legal question of what the relevant amount of damages is, it seems that there is an entire spectrum of answers that clearly meet it. If we suppose that ten thousand dollars is a correct answer to the question at hand, then it would be difficult to deny that ten thousand and one dollar is a correct answer as well. However, it is doubtful that even the fiercest formalist would argue that the situation discussed by Endicott involves a legal gap, at least not in the sense that it needs to be filled. This leads us to a conclusion that not every case involving judicial discretion constitutes a legal gap. Let me therefore suggest an alternative understanding of gaps in the law, one that appeals to the idea that discretion comes in degrees³¹. Consider the case of an unlaw-

ful destruction of a movable item. If one statute empowers the judge to order the perpetrator to pay the damages to the owner of the destroyed item, whereas according to another statute the judge can order the perpetrator either to pay damages or to replace the destroyed item, then, other things being equal, it is intuitive to say that the latter statute confers more discretion on the judge than the former. Of course, there is no straightforward correlation between the number of lawful solutions from which the judge can choose and the amount of discretion the statute confers on her. It is worthwhile to notice that a judge who decides whether using a skateboard is a violation of the “no vehicles in the park” rule only decides between two solutions, namely yes and no, whereas a judge determining the quantum of damages for a tortious injury typically has many more options. Nonetheless we are intuitively more inclined to classify the former situation, rather than the latter, as involving a gap in the law. Indeed, even though legal theorists have identified a plethora of factors that can contribute to an increase of discretion, such as a conflict of norms, the vagueness of statutory language, or the unforeseeability of certain states of affairs, it seems impossible to portray the amount of discretion generated by a given legal question as a simple function of such factors. This is mostly due to the fact that these factors belong to different explanatory levels and, therefore, it seems impossible to order them with respect to the exact amount of discretion each of them generates. The absence of a clear-cut procedure for determining the amount of judicial discretion involved in a particular case is one of the reasons why judges sometimes unconsciously exercise their discretion. This, however, does not automatically invalidate their decisions³². Nonetheless, in order to exercise their discretion responsibly, they should be aware of its scope³³. This point is of utmost importance for the proponents of the input view, whose goal is to limit the scope of judicial discretion. In fact, I believe that reflection on the amount of judicial discretion is the only plausible way of identifying legal gaps available to formalists broadly construed. The reason is that formalists would like to limit the usage of non-formalistic arguments to the cases when it is absolutely necessary. It would fly in the face of their central commitment if they allowed judges to invoke non-formalistic arguments in order to justify the use of non-formalistic arguments. Now, I believe that the most charitable reading of the input view has it that legal gaps are identified by reference to the amount of judicial discretion that is left after the formalistic methods of adjudication have been applied. Even though it is doubtful whether it is possible to come up with a simple procedure of determining the exact amount of discretion, we have already seen that there are cases in which the judge can confidently ascertain (a) that they have some discretion; and (b) that they are definitely not dealing with a legal gap. Mutatis mutandis it seems quite plausible that there are cases in which the judge can confidently ascertain (a) that they have lots of discretion; and (b) that they are definitely dealing with a legal gap (and therefore should apply a relevant gap-filling method). Therefore, for the purposes of the input view for a legal question to involve a legal gap just is for judge to have a significant degree of discretion. Unfortunately for the proponents of the input view, it would be at best wishful thinking to claim that there is a sharp cut-off point between the cases involving lots of discretion and those involving some-but-not-lots-of discretion. Since on the input view

the law-applying official can only invoke evaluative arguments once they discerned the existence of a gap, in those borderline cases the judges have discretion in deciding whether a given legal question involves a gap or not; and, consequently, whether they are permitted to subsidiarily apply a non-formalistic method of interpretation. In the following section, I argue, contra the input view, that the provisions designed to curb judicial discretion in such borderline cases, actually preserve it. My argument exploits the fact that on the current understanding of legal gaps judges in some cases have to exercise discretion to decide whether they are dealing with a legal gap or not. Knowing that, one may already object that, even if I am right, the proponents of the input view can surely help themselves to some other account of legal gaps. I do not think this is the case. However, I postpone the discussion of this objection until the fifth section, after my central argument is in place.

4. The objection

In this section I am going to argue that there is a class of legal questions in which the input view is bound to fail by its own lights – instead of reducing the scope of discretion, it actually preserves it. The cases that I have in mind are characterized by two features: 1) They involve an amount of discretion that places them at the borderline between legal gaps and mere judicial discretion. In such cases the law-applying officials have discretion to decide whether or not they are dealing with a legal gap and, consequently, whether the relevant gap-filling provision is applicable. 2) If the judge applies the gap-filling device, then it provides them with an unequivocal answer to the legal question at hand³⁴.

Having the two conditions in place, let me offer an example to demonstrate what exactly is wrong with the input view. For the sake of simplicity, let us assume that the only formalistic method in the judge’s arsenal is linguistic interpretation, whereas it is part of their theory of interpretation that they could (and should) apply analogous reasoning only once they identified a legal gap. Suppose that a judge is called to decide whether a teenager has violated a “no vehicles in the park” rule by riding into a park on a skateboard. After consulting their personal experience and a number of authoritative dictionaries, the judge reaches the conclusion that the letter of law does not provide a determinate answer to the legal question at hand. Consequently, the judge becomes aware that the law confers on them some discretion in deciding the case. However, they are uncertain whether the amount of discretion they have in this particular context is sufficient to speak of a gap in the law. Indeed, they are dealing with a borderline case of a legal gap. Suppose now that the terms of use of every other park in that town each contain a “no vehicles in the park” rule and a definition of a vehicle according to which a skateboard is a vehicle. We may

fairly safely assume that the analogia legis with other parks’ terms of use unequivocally points to an interpretation whereby the teenager violated the park’s terms of use³⁵. Now, in the case at hand we can distinguish between the following three decisions that the judge can reach: 1) The judge ascertains the existence of a gap. Therefore, they are obliged to consider the analogia legis when making (and justifying) her decision. They declare the teenager guilty. 2) The judge denies the existence of a gap. Therefore, they cannot rely on analogia legis and have discretion either to declare the teenager guilty or to declare them innocent. 2a) The judge declares the teenager guilty without relying on analogia legis. 2b) The judge declares the teenager innocent without relying on analogia legis.

Even though each of the answers 1), 2a), and 2b) is lawful on the input view, it seems natural to say that both 2a) and 2b) are simply wrong. The reason is that by denying the existence of a gap, the judge has deprived themselves of additional interpretative (and justificatory) resources. The mere fact that by ascertaining the existence of a gap they would gain access to additional ways of reaching and justifying their decision should lead us to denying that they had discretion in ascertaining the existence of the gap in the first place. However, this contradicts the stipulation that they had discretion to say that there was no legal gap. One natural reaction to the argument just presented would be to claim that what it really shows is the untenability of the stipulation that in some cases the existence of a gap is indeterminate. Even though it may be difficult to formulate a general criterion for identifying legal gaps, in each particular case the judge has sufficient resources to determine whether there is a gap or not. Even if there is some merit to this response, it certainly does not support the input view. Notice that our preference for scenario 1) over scenarios 2a), and 2b) stems from the holistic comparison of the reasonings instantiated by each of them. The reason to claim that the judge should not deny the existence of a gap is that by doing so they would deprive themselves of additional interpretative resources. However, it is impossible to justify the contribution of these resources without actually entertaining the conclusions to which they lead. Therefore, the use of analogical reasoning actually helped us identify the gap, rather than fill it. However, this clearly contradicts the central commitment of the input view, i.e. that there is a sharp distinction between the stage of identification of a gap and the stage of gap-filling. The upshot of the argument is that there exists a class of cases in which adopting the input view actually broadens the scope of judicial discretion, instead of limiting it.

5. How not to identify legal gaps

In this section I discuss three alternative ideas about how to identify legal gaps and argue that neither of them is capable of providing a precise cut-off between legal gaps

and other instances of legal indeterminacy. The theories I scrutinize appeal to legislative intent, legally relevant properties, and disagreement of competent agents, respectively. Let us first consider the idea that whether a given legal question involves a legal gap depends on whether it was the legislator’s intent to leave it unregulated³⁶. It is not difficult to see the axiological appeal of this view – ultimately, one of the reasons why the notion of legal gaps is so controversial is that it touches upon the delicate relation between the legislative branch and the judicial branch. Therefore, it would be clearly desirable from the perspective of the separation of powers if the judges only stepped in in cases of obvious legislative slips. Unfortunately, this is an unattainable ideal. Richard Posner aptly points out that “cases where legislatures explicitly delegate policy-making tasks to courts are rare, so the difference between the accidental and the deliberate gap has little practical significance”³⁷. The force of this observation can only be strengthened if we accept more general worries about legislative intent as a guide to legal interpretation³⁸. In fact, in his seminal paper on legal gaps, Legal Reasons, Sources and Gaps, Raz listed the indeterminacy of intention among the causes of legal gaps³⁹. Another notion that has been employed to determine the existence of legal gaps is that of a legally relevant property⁴⁰. In order to understand this idea we have to introduce the notion of a generalized legal case (e.g. murder, transfer of property to a third party, etc.). It is assumed that to each generalized legal case corresponds a set of mutually independent relevant properties, whose combinations constitute the universe of particularized cases falling under the generalized case. Now, one speaks of a legal gap if the law remains silent with respect to at least one of the particularized cases⁴¹. This idea is commonly illustrated with the example of restitution of legal estate (generalized case):

Assume that a statute stipulates that (1) the restitution of legal estate is obligatory, if the transferee is in good faith, the transfer is made with consideration and the transferor is in bad faith; and (2) the restitution of legal estate is obligatory if the transfer is made without consideration. Assume now that the transferor is in good faith and the transfer is made with consideration but the transferee is in bad faith. Is the restitution of legal estate obligatory? The norm does not answer the question. A gap occurs⁴².

Let us now unpack this example a little bit. The properties relevant to the restitution of legal estate are then: the transferee’s good faith (A), the transferor’s good faith (B), and consideration (C). Let us assume for the sake of simplicity that bad faith is just lack of good faith, and accordingly symbolize the transferee’s and transferor’s bad faith as ~A and ~B, respectively. Now, according to rule (1), restitution is obligatory if A&~B&C. According to rule (2), the restitution is obligatory in any case in which ~C. However, the law does not explicitly say if restitution is obligatory, when the transferee is in bad faith, the transferor in good faith and the transfer is made with consideration (~A&B&C). Aleksander Peczenik claims that the law contains a normative gap – it does not give a determinate answer to the legal question posed by such a situation.

At first glance, this strategy seems like a good candidate for precisely delineating the set of legal gaps. Unfortunately for the proponents of the input view, the relevant property strategy is overly inclusive. As Fernando Atria aptly notices, the mere fact that the positive law does not regulate a particularized case does not even entail that the judge exercises discretion in deciding it⁴³. For instance, by Peczenik’s lights, most western legal systems contain a legal gap as to the permissibility of sexual intercourse between consenting adults⁴⁴. Yet, this is clearly not the sort of legal gap the proponents of the input view are interested in. The upshot is that the notion of legally relevant properties does not get them any closer to precisely determining the class of legal gaps that fulfill the antecedent of provisions such as Article 1 of the Swiss Civil Code⁴⁵. The final criterion I am going to discuss is the disagreement between competent actors. As far as I know, no author has proposed to identify legal gaps by reference to disagreement. However, the notion is quite commonly believed to be a symptom of two phenomena closely related to legal gaps, namely vagueness and hard cases. The idea would be to say that there is a legal gap if the competent parties sincerely disagree on whether a given particularized case is regulated or not. The intuitive appeal of this idea as a criterion for discerning legal gaps lies in the fact that the existence of sincere disagreement implies that at least two people are inclined to decide a particularized case in different ways and that some arguments can be presented to support each of the conflicting solutions. Nonetheless, disagreement itself seems to be a vague notion. How much divergence of opinion constitutes disagreement? The only non-arbitrary sharp answer seems to be: any divergence. According to this reply, it suffices that at least one person breaks the consensus to classify the situation as a disagreement and, thereby, as a gap. However, this is strongly unappealing, for “there is no requirement to align your usage with that of the first maverick who first breaks ranks”⁴⁶. One may be tempted to reject Mark Sainsbury’s proposal, as long as it is a competent maverick. However, this reply only transforms the problem into one of drawing a clear line between who does and who does not count as a competent actor for the purposes of identifying legal gaps⁴⁷. The upshot of this section is that none of the criteria for identifying legal gaps proposed in literature provide a way of drawing a clear line between legal gaps and other cases involving judicial discretion. Of course, I do not claim to have proven that it is logically impossible to draw such a line. However, the burden of providing such a criterion is on the proponents of the input view. In order for their view to be functional, it does not suffice to believe that there is such a criterion. They should provide judges with a principled way of discerning legal gaps that could be used as a justification of applying such provisions as Article 1 of the Swiss Civil Code.

I began this paper by identifying what I call the input view of legal gaps and pointing to its formalistic provenance. Central to the view is the positing of a clear-cut distinction between the identification of a gap in the law and filling it. I have argued that in this picture none of the criteria of identifying legal gaps allow the formalist to conclude that there is a clear-cut distinction between legal gaps and other cases of judicial discretion. Moreover, I exploited this conclusion to show that the input view, contrary to its discretion-curbing aspirations, systematically leads to an undesirable increase in judicial discretion. I have presented the input view as a central feature of an ideology of judicial application of law, which can be tentatively called “minimally enlightened bound judicial decision-making”. Its underlying idea is to make theoretical space for the existence of legal gaps, while making as few concessions to the ideology of free judicial decisionmaking as possible. The upshot of my paper is that such an approach is not feasible. In particular, if one admits non-formalistic methods of interpretation into one’s theory of interpretation, one should not hope that it is possible to circumscribe their application only to the exceptional cases in which formalistic methods do not suffice. This observation, in turn, blurs the very distinction between formalistic and non-formalistic methods of interpretation. By casting the input view in terms of Wróblewski’s ideologies, I decided not to ascribe the view to any particular individual. Doing so makes me prone to the charge that no one holds this view. Even if the objection is valid, I think it does not defeat the purpose of my paper. For at the very least it shows us how not to think about the notoriously vexed issue of legal gaps. Let me here rehearse the main theses I associated with the input view: 1) Legal gaps are premises in legal reasoning. 2) Legal gaps are identified exclusively by using formalistic methods of interpretation. 3) Legal gaps are filled by using non-formalistic methods of interpretation. 4) It is possible to draw a sharp distinction between formalistic and non-formalistic methods of interpretation. 5) A legal gap occurs when a judge has a significant amount of discretion in deciding a given legal question.

Even if no actual theorist or practitioner believes – consciously or unconsciously – the conjunction of those 5 theses, I submit that each of them has some prima facie appeal. The lesson to be drawn from my paper is that we should not give in to this appeal too quickly.