In jurisprudence, a debate exists about the possibility and desirability of a rigid distinction between discovery (how a judge actually reaches a decision) and justification (how a judge publicly justifies a decision). This article shows that this debate is being muddled because of differences and ambiguities in the way that different writers use the terms ‘discovery,’ ‘justification,’ and related terms. The article argues that merely distinguishing between ‘discovery’ and ‘justification’ is not precise enough, and that we should make a distinction between different elements within each of these contexts. I propose a six-fold classification, through which we can identify reasons, acts, and processes that play a role both in the context of discovery and in the context of justification. This six-fold classification enables us to move forward from debating whether discovery and justification can be rigidly separated, towards articulating how each element (reasons, acts, and processes) has a role to play in each of the contexts (discovery and justification), and how these elements and contexts are related.

• 1 Introduction

  The objective of this article is to analyse and clarify the conceptual framework that permeates the debate about the distinction between discovery and justification in legal decision-making. While on the one hand the majority of contemporary literature either advocates for or presupposes a rigid distinction between discovery and justification, on the other hand this distinction is being challenged by those who defend that such a distinction is unachievable, undesirable, or both.
  The problem that this article identifies is that this debate is being muddled due to insufficient conceptual clarity. While both sides of this debate do not necessarily mean the same thing when they use terms such as ‘discovery’ and ‘justification’, the debate gets even more confusing when authors use different terms altogether ¹x Such as Raz’s ‘reasons for conformity’ and ‘reasons for compliance’, or Dworkin’s evaluation whether a particular decision was taken ‘responsibly.’ For more details and references, see notes 12-15 below. to refer to similar concepts and problems.
  My claim is that merely distinguishing between ‘discovery’ and ‘justification’ is not precise enough to cast clarity onto the debate. Moreover, when authors use terms such as ‘the process of discovery’ or ‘the context of justification,’ there has to be more precision as to what they mean by ‘process’ or ‘context.’ Often authors on both sides of the debate are using these different terms interchangeably while at other times they are using the same terms to refer to different things.
  Therefore, in this article I propose a more articulated distinction: one between reasons, acts, and processes in both the context of discovery and in the context of justification – making it a six-fold classification. When I use the term ‘context,’ I am referring to the group of elements that comprise either discovery or justification. In that sense, the context of discovery encompasses the reasons, acts, and processes of discovery, while the context of justification encompasses the reasons, acts, and processes of justification. Once we achieve this six-fold classification we can be more precise with our analysis of legal decision-making, for we need not be debating whether ‘discovery’ and ‘justification’ are altogether either rigidly separated or inherently intertwined. We can instead try to understand and articulate how each element (reasons, acts, and processes) interact with each other in each and between the contexts (discovery and justification).
  After this introduction, I will define discovery and justification in section 2, as well as provide a brief contextualization of the distinction between discovery and justification in contemporary jurisprudence. Section 3 will focus on the context of discovery. I will primarily rely on Anderson’s analysis of discovery in legal decision-making ²x Anderson 1996. in order to identify and define its different elements, namely acts, reasons, and processes. Section 4 will apply these same distinctions to the context of justification. Section 5 will then briefly analyse some possible distinctions and relations between the different elements of both contexts, in order to demonstrate how useful this conceptual framework is. The 6 ^th and last section will provide a case example and discuss some practical implications of the concepts and distinctions discussed throughout the article.

• 2 Discovery and justification

  In this section I will provide the definitions of discovery and justification, as well as some background to the debate regarding the rigid distinction between discovery and justification and its lack of terminological precision.
  Discovery may be understood, in an initial definition, as the process by which the judge ‘actually’ reaches a decision: ³x Anderson 1996, p. 1. a conclusion concerning a judicial problem. It includes aspects like the motivation she had to reach that decision and the reasons she had to think she is making the right choices. Justification may be understood as how the judge publicly justifies a decision. ⁴x Anderson 1996. It is the exposition of the reasons why the decision should be accepted by all society as the right decision. In broad terms, this distinction between discovery and justification is well summed up by MacCormick: ‘What prompts a judge to think of one side rather than the other is quite a different matter from the question whether there are on consideration good justifying reasons in favor of that rather than the other side.’ ⁵x MacCormick 1997, p. 16.
  As mentioned in the introduction, the problem is that there is no agreement in the literature about the use of these concepts. Different authors mention discovery and justification with slightly different meanings, while others use other terms like the distinction between the context of discovery and the context of justification, ⁶x As the distinction is commonly referred to in epistemology, e.g., Hoyningen-Huene 2006, p. 119-131. between the process of discovery and the process of justification, or between the logic of discovery and the logic of justification. ⁷x Wasserstrom 1961, p. 27. It is unclear if different authors are truly engaging with each other (and not talking past each other) when concepts such as discovery and justification are used in this broad a fashion. ⁸x For related uses of discovery/justification and similar terminology, see, e.g., Golding 1992, p. 109-125; Sartorius 1992, p. 127-146; Lyons 1992, p. 147-166; Michelon 2006, p. 114-117; Hage 2005, p. 35 et seq.; Feteris & Klossterhuis 2011, p. 253-273.
  The distinction between discovery and justification gained force in the jurisprudential debate with Richard Wasserstrom. ⁹x Wasserstrom 1961. He intended to solve the tension between Legal Formalists and American Realists by stating that the rigid distinction between discovery and justification helped to understand the disagreements between these authors: while Formalists were studying the process of justification, Realists studied the process of discovery.
  Contributing to the permanence of the lack of terminological clarity is the fact that Wasserstrom stressed the importance of justification, regarding discovery as a process that should be left aside from jurisprudence – a notion that was carried on by positivist theories of law and adjudication, such as MacCormick’s. ¹⁰x MacCormick 1997, p. 15 et seq. The focus was maintained solely on justification, and questions were usually kept constrained within two poles: decisions may be justified by reasons and decisions may be justified by processes. ¹¹x This justification by processes is not akin to my use of the term process, whether in the process of discovery or the process of justification. They are different concepts, which points to the need of this work of clarification. When authors use justification by process, they are often not referring to justification at all but to legitimacy. Think about these two examples: (1) In a given legal system, all candidates to the presidential election ought to have the same time of exposure on both television and radio programs. (2) In a given legal system, all court proceedings ought to be open to the public. In neither example the outcome (the result of the election or the judicial decision) is justified by the fact that the procedure was followed, but following the procedure is part of what legitimizes the outcome. Throughout the article it will be clear that what I mean by process of discovery and process of justification is not related to procedural legitimacy, even though they are not mutually excluding components of judicial legitimacy.
  The use of the broad term ‘justification’ to refer both to justifying reasons and to justifying processes is not the only problem, as some authors use different terms altogether for similar concepts. Dworkin, for example, in his latest book ¹²x Dworkin 2011. tried to make clear his points about how moral decisions should be evaluated when we deeply disagree. In short, he states that it is not the actual decision that matters the most, nor the string of arguments cast in its support if those arguments are taken to be independent from the decision-maker. In very tough cases we must, Dworkin states, evaluate the responsibility of the decision-maker — i.e., whether the decision was taken responsibly (although Dworkin does not really provide us with tools to carry out this task). We can argue that for Dworkin the evaluation of the decision itself – as an autonomous object – is put on a second plane. ¹³x Dworkin 2011, p. 100. This amounts to saying (if we translate into the language of this article) that we should be assessing discovery, not justification. Dworkin, however, does not make use of the discovery/justification concepts, which demonstrates how the terminology is not yet well established in the literature.
  For another example of the use of different terminologies we may refer to Raz’s justification theory. Central to his theory is his concept of exclusionary reasons. Simply put, they are exclusionary because they are (second order) reasons that exclude other (first order) reasons for action from consideration. ¹⁴x Raz 1999. When interpreting what Raz meant by this, one could be tempted to assert that he is referring to the process of discovery. This is not the case. Raz writes in the postscript to the second edition of Practical Reason and Norms that the reasons why people made decisions do not matter at all. He states that the relevance of norms is in conformity and not in compliance. In other words, it does not matter if a reason helped an individual to decide for a certain action, because all that matters is that the action was the one prescribed by the reason – even if the ‘correct’ decision to act was reached by mere chance. ¹⁵x Raz 1999, p. 179 et passim. What Raz is referring to as ‘reasons for conformity’ is in fact also a rigid distinction between discovery and justification with a focus on the latter.
  Raz’s conception of ‘reasons for compliance’ is akin, not merely to a justification by process (as mentioned above, a justification in which the steps taken in order to reach the decision act as justifying reasons) ¹⁶x MacCormick 1997, p. 15 et seq. (the steps taken in order to reach the decision) but to a consideration of the discovery process in the justification of the decision (specific reasons for discovery must have influenced the decision-making in a precise way for the decision to be justified). The former means only ‘ticking boxes,’ in the sense that certain formal conditions must have been fulfilled or certain steps must have been taken during the making of the decision, irrespective of their content. It is indeed a procedural account. The latter, or the study of the context of discovery, does take content into account. It is not a procedural account but a substantive account of the process of coming up with a decision.
  Now that the lack of terminological clarity in this debate has been pointed out, we need to start working towards more precise definitions. This will be the purpose of sections 3, 4 and 5 below. In section 3, I will use Bruce Anderson’s analysis of discovery in legal decision-making in order to identify and flesh out the differences between reasons, acts, and processes of discovery.

• 3 The context of discovery

  In trying to achieve a more precise conceptual framework for the study of discovery, Anderson’s ‘ Discovery’ in Legal Decision-Making is paramount. If one wants to comprehend the actual reasons of a judicial decision, Anderson asserts, one must investigate and comprehend the discovery process. analysing decisions taken by judges from their own point of view (and from their description of the decision-making process of those decisions), Anderson realized that the discovery process involves ‘questions and answers’ towards the solution. Searching a detailed demonstration of what goes on between the first question asked and the final decision taken, the author brings to scope Bernard Lonergan’s works: ¹⁷x Bernard Lonergan (1904-1986). Among his major works are Lonergan 1992 and Lonergan 1971.

  He [Lonergan] explicitly studies discovery or invention and testing in various fields, particularly science, in terms of puzzling, asking questions, experiencing insights and testing hypotheses. Questions arise when one is puzzled. Insights occur in response to those questions. Insights lead to the formulation of new ideas and then one tests them.

  Insights would be ‘acts of discovery’ and are distinct from sensorial experience, wonder and questions.
  According to Lonergan, insights occur in response to formulated questions in two dimensions: theoretical reasoning (knowledge for the sake of knowledge) and practical reasoning (knowledge for the sake of what to do), which he called direct insight and practical insight respectively. Both are aimed at understanding or discovering possible relations among data. Direct insights may be true or false and practical insights may be appropriate or not in a particular situation; thus they should be tested. In theoretical reasoning direct insights are tested by r eflective insights and in practical reasoning practical insights are evaluated by practical reflective insights. ¹⁸x Anderson 1996, p. 97-98. Hence, there are four ‘categories’ of insights.
  Reflective insights discover the link between prospective judgments and the sufficiency of the evidence for making judgments of fact regarding the truth or falsity of direct insights and formulations. ¹⁹x Anderson 1996, p. 98. Hence both insights involve discovery. ²⁰x Anderson 1996, p. 122.
  Practical insights are the key activity in practical reasoning. The mental activity is represented by ‘What-is-to-be-done?’ questions, which lead to discovering possible courses of action. Some of those courses of action may be impossible or unreasonable to perform, and testing those alternatives may be represented by questions that ask ‘Is-it-to-be-done?’ and such. Practical reflective insights discover the relation between the significant factors of a particular situation, the proposed course of action, and the consequences and implications of the action. Practical reflection or deliberation leads to a judgment of value that one course of action is sufficiently suitable. A decision to perform the course of action ends practical evaluation. ²¹x Anderson 1996, p. 98-99.
  Taking these elements into account, a complete expression ²²x Anderson denominates it as ‘normal expression.’ However, ‘normal’ could lead to different interpretations — one can conclude the ‘normal’ is to provide an incomplete expression. By suggesting the term complete, I am also avoiding to use the word ‘sufficient’ (used by other authors such as MacCormick), because I understand that in many cases — especially those in which parties do not disagree with the decision taken by the judge — an ‘incomplete expression’ could be considered as ‘sufficient justification’ as to justify the decision. of a judicial decision should comprise: (1) relevant evidence and relevant law; (2) interpretation of evidence and law; (3) the judgments of fact and decision regarding which evidence is relevant and which law and rulings of law are relevant; (4) possible decisions, which would be hypothetical courses of action; (5) the judgment of value, measuring consequences of the decision taken and if it is supported by the law. ²³x One important difference between what Anderson is proposing and what authors who rely on a rigid distinction between discovery and justification would accept as sufficient justification is comprised by the 4 ^th item. If there is no relevant relation between discovery and justification, all of my mistakes, dead ends, miscomprehensions and so forth, are also irrelevant for the justification of the decision. For the proponents of the rigid distinction, the ‘final product’ and the reasons that support it are completely independent from the steps one took to reach it.
  Also worth mentioning are Lonergan’s two ‘patterns’ of expression: ‘rhetoric’ and ‘axiomatic’. While the former is patterned in accord with the discovery process, the latter could be explained at its best in the axiomatic presentations of mathematics. In the legal context we could look at MacCormick’s portrait of deductive justification ²⁴x Anderson 1996, p. 152-153. as an example of axiomatic justification. Anderson demonstrates both patterns as incomplete: On the one hand, ‘the rhetoric of discovery turns out to be an authoritarian monologue. We read the case her [a judge’s] way. ’ ²⁵x Anderson 1996, p. 78. The judge must make an effort to lead the reader to agree with his opinion and conclusions and, more specifically, with the way the case was solved. On the other hand, a justification that lacks rhetorical expression will not bring the reader to adequate understanding to form an educated opinion about the justification of the final conclusion. ²⁶x Agreement or disagreement with the conclusion is irrelevant in this case, as we are assessing the justification and not the truth-value of the conclusion. Lawyers often conflate truth and justification. If we subscribe to the definition of knowledge as true, justified belief, then it follows that truth and justification are not interchangeable. It is not sufficient to demonstrate that it is true that it is raining outside in order to justify a belief I had that it is raining outside. If I never checked outside and came to the conclusion that it is raining as a result of a coin flip, my belief happens to be true but it was not justified. In jurisprudence, this mistake takes the form of believing that if a decision is in accordance with the law, this is not only a necessary but a sufficient condition of its justification. It is important that we recall here that my overall argument is an epistemological, and not a metaphysical one. For more on this distinction, see David 2001, p. 153-154. The relations between the truth-value and the justification of legal decisions are a relevant part of this debate, but an analysis of these relations has been left out of this paper due to space constraints. As both patterns by themselves are inadequate, one should aim at balance: ‘The expression must be one that finds the relevant balance between rhetorical and axiomatic presentation.’ And, Anderson continues: ‘That relevant balance is particular to each instance. There is not going to be an axiomatics of balance. ’ ²⁷x Anderson 1996, p. 157.
  Although one can explain Wasserstrom’s paradigm in terms of Lonergan’s concepts, it is not difficult to understand that both authors are not referring to decision-making processes in the same way. Is spite of that, Anderson’s work could be understood as in part trying to reconcile Lonergan’s process of problem-solving with Wasserstrom’s decision-making process as two processes: discovery and justification. The attempt is quite clear when Anderson tried calling Lonergan’s understanding phase (direct insights) ‘discovery’ and the testing phase (reflective insights) ‘justification.’ But he dropped the attempt:

  ‘To call the understanding phase “discovery” and the testing phase “justification” would continue to mask the creative role of insight in testing. To analyze reflective insight and practical reflective insight and not to stress that they are acts of “discovery” would mis-represent the creative nature of testing.’ ²⁸x Anderson 1996, p. 101.

  Anderson presents another possibility to describe the decision-making process:

  ‘(…) when a judge is actually solving a problem, theoretical and practical problem-solving are inter-related throughout the decision-making process. A judge’s attention may shift from theoretical to practical problem-solving and from practical to theoretical problem-solving more or less continuously. (…) This process would lead to a complex inter-related set of insights, judgments of fact, judgments of value, and finally a decision.’ ²⁹x Anderson 1996, p. 108-109.

  Thus, settling the differences between this complex demonstration of inter-related elements and that which rigidly distinguishes ‘discovery’ and ‘justification’ seems like an even harder if not unfeasible attempt without the more detailed terminology and definitions that I present in this article.
  There is great scepticism in the literature regarding the possibility of analysing the process of discovery, especially in legal decision-making. This scepticism was addressed by Anderson, and we can emphasize at least three of his objectives: (1) To demonstrate that the discovery process can be analysed and is vital to comprehension of judicial decisions. (2) To present discovery as a deliberate and conscious process that is not essentially arbitrary, haphazard, and irrational. ³⁰x Anderson 1996, p. 128. (3) To extend the importance of discovery, asserting it is central for legal reasoning, and even demonstrating that discovery might have an important role in legal justification.
  Discovery can be said to be rational insofar as reason acts to organize our insights as we need to order properly all hypotheses and testing, and all judgments and decisions, with the goal of finding the correct (or the most correct) solution to a problem. Unorganized insights do not make sense and cannot lead to any reasonable conclusion. Thus, one can say the process of discovery is rational and deliberate. However, it does not necessarily follow that the act of discovery (i.e., the flash of insight) is also rational and deliberate.
  Thus, the process of discovery is also distinct from the reasons of discovery (the reasons or motives which lead a judge do decide one way or the other such as the law, rules of law, a desirable consequence, or even an excessive formalism or biases and prejudices) and from acts of discovery ( insights – treated by Lonergan as ‘acts of discovery’ – which discover the reasons of discovery that are organized by the process of discovery that leads to a decision). While the discovery process is rational, the act of discovery may not be quite so. That is why being more precise in these definitions and distinctions is important: so we can analyse each of them and their features in detail.

• 4 The context of justification

  In spite of Anderson’s start in problematizing the concept of justification (e.g., by arguing that there is discovery in the process of justification) the author does not go any further. I propose the same division I made regarding discovery (including a similar nomenclature) that will lead us to process of justification, act of justification and reasons of justification.
  Anderson is talking about the process of justification when he stresses that justification, in a broader context, is the discovery of sufficient reasons to justify a decision and that testing applied to hypothetical decisions involves discovery. This process is how the judge finds reasons of justification, or in other words, the way she discovers arguments and reasons that will support her decision. Yet it is not necessarily or distinctly a part of the discovery process (as long as it does not have as a goal discovering a solution) despite having the same structure of organizing and evaluating questions and insights.
  Reasons of justification are the arguments themselves. They may be the reasons that lead the judge to take the decision (hence insights and judgments of value) or any other arguments discovered by the process of justification as good enough to justify the conclusion reached. To justify is to prove to be just, right, or reasonable. What support a decision are the reasons of justification.
  The act of justification is what Anderson and Lonergan call expression: the words (or any other language) used to expose the reasons of justification: the reasons why one thinks her decision is right, just or reasonable. The act of justification is the chance for a judge to make her decision legitimate (at least in more advanced and democratic legal systems) and it is the raw material by which the decision will be studied and evaluated.
  There is one more classification which is important to emphasize: when one takes a decision there are two levels in which there is justification: personal and interpersonal. The former regards reasons we present ourselves, as we must convince ourselves that our conclusions are correct. They are subjective and personal arguments, which may only make sense to whom is deciding. Hence personal justification is part of the discovery process and its reasons are obtained by reflective insights. As we discover more and more reasons of personal justification, we become increasingly more convinced that we are making the right decision.
  Interpersonal justification is what we expose to others. Its arguments are the reasons of justification, which are discovered by the process of justification, and it is exposed by the act of justification. Interpersonal justification may be necessary to justify the decision to other people. Rhetorical and/or axiomatic expressions are employed to demonstrate to the listener that the decision made is justified. The keyword is to convince. As Anderson correctly puts it, there is no such thing as an absolute justification: expression will be the raw material that presents the possibility for the listener or reader to understand and evaluate the decision. The reader may agree or may not agree with the reasons exposed.

• 5 Further distinctions and relations

  The most obvious distinction between discovery and justification is the one regarding acts. To reach a solution for a problem and to expose its reasons are two different – yet not opposite – actions. It is one thing how a judge reaches a decision and quite another how she justifies it. We can see that distinction clearly in two levels: logical and empirical. (1) Logically, a conclusion must be reached so that it can be, afterwards, justified. There is no such thing as the possibility to justify what is not even thought of yet. (2) Empirically, the act of reaching a solution and taking a decision is distinct from the act of exposing it. The former is subjective and introspective while the latter is objective and expositive (we use words, sounds, images or gestures so we can demonstrate why our point of view is to be accepted).
  As for the processes, three distinctions must be made: (1) The elements of the process of discovery are insights, which will be the premises for reasoning towards the solution, while the process of justification comprises both the discovery of sufficient reasons of justification and the utterances of these reasons. (2) One can only begin justifying a hypothetical decision once this hypothesis is thought or invented. So the process of justification depends on the process of discovery, and the former can only start after the latter has already been started and has shown some results. (3) There are distinctions regarding the tests of hypothetical decisions, and not paying attention to this difference may have led Anderson to present some unfair criticism of MacCormick’s work. While MacCormick asserted that the tests are made in justification (as analogous with the empirical testing in natural sciences), Anderson stressed that the testing phase was actually inserted in the process of discovery by formulating relevant questions that lead to reflective insights (direct or practical, depending on whether they are held in theoretical or practical reasoning). Instead, my conceptualization allows us to see that there are two species of testing: (a) Those mentioned by Anderson, used during the process of discovery with the objective of satisfying a personal criteria of ‘sufficiency’ or, in other words, the search for good reasons of personal justification (those which seem ‘good enough’ for the person taking the decision). (b) Tests as mentioned by MacCormick, which aim to satisfy criteria of interpersonal justification or, in other words, aim to find out good reasons of interpersonal justification to justify the decision to other people. ³¹x Reasons of interpersonal justification often operate as reasons of personal justification, i.e., I may only be satisfied with a decision once I believe I am able to justify it to other people.
  Studying the relations and distinctions between reasons of discovery and reasons of justification is surely the most problematic issue at hand. At this point, what is important to emphasize is that it is likely that the reasons of discovery are not the same reasons exposed as reasons of justification. Demonstrating this hypothesis is not difficult, as is shown in the following example. When invited to a party by his wife, John decides to stay home because he is feeling lazy. He then justifies to her that they are short in money and that it would be wise to save the expenditure on baby-sitting they would incur if they were to go to the party. This justifying reason might indeed be true, and maybe John only took the decision to stay home because he knew this reason would justify his decision, yet the decision was taken because he was lazy.
  Hence, despite the possibility of establishing a formal rigid distinction between process of discovery and process of justification, in practice we cannot separate them. First of all, as demonstrated above, there is discovery in the process of justification. Also, the process of justification needs only a hypothetical decision to get started while the process of discovery only gets to an end when the decision is reached. From the moment one discovers a hypothetical decision until the moment a decision is taken, justification may ³²x We must keep in mind that we may be referring to a decision that does not necessarily have to be justified to others. play an important role in discovery. John may only decide to stay home once he discovers a good justifying reason that he can give to his wife (i.e., the costs of baby-sitting). In this case, the process of justification is operated along with the process of discovery, as the conclusions in the former may be used in the latter, and the insights obtained in both processes will be used in the other and so on.
  For judicial decisions taken inside legal systems which do not authorize purely authoritarian solutions, strictly personal justification cannot be considered as sufficient. Interpersonal justification is not a possibility but an essential part of the decision. In this sense, to assert that the judge always first takes the decision and then searches for justifying reasons (and if she does not find those reasons she must start anew the discovery process) is to oversimplify the problem. Discovery only ends, i.e., the judge will only decide, in the moment she has discovered sufficient reasons for interpersonal justification as she knows they are necessary for the decision to be accepted at all.
  In short: (1) The process of justification can only be initiated after the process of discovery has begun. (2) The process of discovery should only finish after the process of justification has come to an end, especially when interpersonal justification is one of the decision’s intrinsic requisites. (3) All insights obtained by the process of discovery may form raw material for the process of justification, and vice-versa. (4) All reasons of justification may be used as reasons of discovery, and vice versa; thus for all those reasons it is imperative to understand both processes as intrinsically interrelated and interdependent. (5) The acts of discovery (insights) and the acts of justification (utterances) are, indeed, distinct and potentially independent instances. The possible dependence and the relations between these distinct acts, however, ought to be part of the study of legal decision-making.

• 6 Conclusion and implications for legal practice

  Anderson proposes justification as comprising both rhetorical and axiomatic patterns. The judge should expose the actual way the decision was reached (exposing even what she thinks would be her biases and prejudices) and then should expose the objective judicial reasons that authorizes the conclusion. It is a new concept, in which reasons of discovery (subjective) must be presented as reasons of justification just as much as reasons reached by deductive reasoning. With such a justification the judge would demonstrate (1) why it seems to her the right decision; (2) why it seems to her that other people should consider her decision the right decision; (3) why every citizen must accept her decision.
  What is at stake in this discussion is a practical matter. The central problem of the larger debate is: What justifies a legal decision? In other words, what kind of justification and which elements should a justification contain so that the decision ought to be considered legitimate in a given legal system? The requirement that courts justify their decisions stems from the rule of law, and the justification is thought to demonstrate that the decision is not arbitrary or in any way personal. It is part of the overall aspiration to be ruled by the law and not by men. This article is part of an effort to demonstrate that the current conceptual framework for legal justification is not adequate to the evaluation of our criteria for assessing the legitimacy and justification of legal decisions.
  Let us briefly review an example to illustrate the practical implications of the study of discovery in jurisprudence. This illustration comes from Anderson, ³³x Anderson 1996, p. 80-92. and he used it to demonstrate the feasibility of studying discovery in legal decision-making. It is an investigation into a decision-making carried out in an arbitration process, and I have chosen it for the sake of its simplicity. ³⁴x Whether an arbiter has the same requirements as a court judge to justify his decisions is irrelevant to the present purpose of demonstrating that the study of discovery, its elements, and its relations to justification can cast better light onto the criteria by which the justification of legal decisions can be assessed. In fact, Anderson’s other example (Anderson 1996, p. 61-78) to illustrate the same points is a decision by the Supreme Court of Canada. However, the case – which involves the legality of abortion in Canada – is too dense to be presented here due to space limitations. It is about a decision by an arbiter who must decide which of two insurance companies ought to pay insurance benefits to an injured woman in a car accident. The central problem of this case revolves around the concept of ‘insured person’ which is defined by the relevant statutes as ‘(a) any person who is an occupant of the described automobile and (b) the insured person if they are an occupant of any other automobile.’ ³⁵x Insurance Act R.S.N.S. 1989, c. 231 and the Claims Agreement, Insurance Bureau of Canada, Section 111, Rule 11A, apud Anderson 1996, p. 81 and 91-92. In the case at hand, the injured woman (let’s call her Z) ‘was opening the trunk of her own car when a moving car hit the adjacent parked car which, in turn, hit the woman’. ³⁶x Anderson 1996, p. 81.
  This example shows the relevance of the authoritative sources in the context of discovery, which can be transposed to the context of justifying the decision. If we took a formalist approach to justification, the decision could be that – say – ‘Company A should pay the insurance benefits because Z was not an occupant of the vehicle’. This reasoning can easily be expressed in a syllogism, which will demonstrate that the decision is in accordance with the law. However, we do not need to extend too much the interpretation of ‘occupant’ to see that, albeit perhaps with some controversy, Z could be considered an occupant. It was, after all, her own car, and she was accessing its open trunk. The problem is thus that the solution ‘Company B should pay the insurance benefits because Z was an occupant of the vehicle’ can also be expressed in a legal syllogism (albeit with particularities and some possible controversy), and could also be considered justified or in accordance with the law.
  What Anderson pointed at with this case were the ways in which the arbiter solved the case: by resorting to questions and answers, insights and formulations, judgements and a decision. What he could have pointed at as well is that the quality of the discovery process brings weight to the justification of the decision. What we see are the elements of the discovery process (including its acts and reasons) playing a role in the justification. The arbiter’s solution is better justified if the decision-making process was adequate.
  What the arbiter did in this case ³⁷x Based on Anderson’s analysis, as seen in Anderson 1996, p. 82-86. was to start asking himself if ‘was Z an occupant of the vehicle’ was indeed the right question to be asked given the circumstances and the problem posited by the insurance companies. He investigated the concrete circumstances, the applicable statutes, and searched for legal guidelines, including legislation that could provide a method to solve the case. Also relevant were the questions he realised he did not have to answer, such as the extent of the damages or the formulation of a general rule for future cases (as would be the case were his decision a binding precedent for himself or for other decision-makers).
  The arbiter moved to seek an answer in the legislation. Was there any piece of legislation that further specified the meaning of ‘occupant’? He analysed the meaning of ‘occupant’ depicted by the Insurance Act in light of the concept of ‘pedestrian’ brought by the Claims Agreement, in order to make consistent the meaning of ‘non-occupant’ with the meaning of ‘pedestrian.’ He then sought the meaning of ‘occupant’ in the ‘Words and Phrases’ Sections of both the Canadian Abridgement and the Nova Scotia Reports. Unsatisfied, he resorted to the Oxford Concise Dictionary and, having found the definition of occupant as ‘one who resides or is in a place’ he went on to ask himself if that answer was appropriate, if there were other ways still to define ‘occupant,’ or if there would be an extended meaning in the common law that would be broader than the definition given by the dictionary. So he needed to further analyse the relevant case law. Finally, he researched the legislation of other Canadian provinces and how they define ‘occupant’ and found different answers to his question, as well as how these different definitions were used in similar cases.
  Even if the conclusion the arbiter reached is the same as the one he would have reached without this whole process, his decision can now be better justified. This does not necessarily mean that his solution is now right and would otherwise be wrong. His conclusion that ‘Z was not an occupant of the vehicle at the moment the accident occurred’ can be the same, but his decision is better now. The truth-value of the decision is not to be confused with its justification. The arbiter’s decision is better justified if he can express to the reader the elements that he analysed and took into account when formulating his decision. He can do so by allowing us to grasp and assess the quality of the decision-making process. ³⁸x This is the key difference between this approach and the justification of a legal decision as proposed by, e.g., MacCormick or Raz. For them the ways in which the decision was reached are not only unnecessary but also a hindrance to proper assessment of the decision, while for Anderson these elements are precisely what allows us to better assess if the decision is justified.
  What we are seeing is that discovery does not comprise one big ‘eureka’ moment, an illuminating and arbitrary strike of creativity that shows us the solution to a problem and that is followed by a separated and independent process of justification. Hopefully the example is sufficient to demonstrate that both can be seen as intertwined processes which are part of the overall process of legal decision-making. It is also important that we are able to differentiate these processes from the acts and reasons that constitute them. In our example, acts of discovery would be the particular insights the arbiter reached. The reasons of discovery need to be scrutinized further. They can be the questions that the decision-maker asked himself (the questions that led to insights); they can also be connections of data that he identified and that led him to the insights; and they can finally be the tests and verification that were carried out to confirm the suitability of the insights. And lastly, the acts of justification are the utterances, verbal or written, that seek to justify the decision, while the reasons of justification are the reasons themselves that purportedly support the decision.
  Once we have established this conceptual framework that encompasses both discovery and justification, and their reasons, acts, and processes, we can start to ask with more precision about the elements that constitute sufficient and adequate justification for legal decisions. This will have to be left to be presented in a future opportunity. These two problems need to be addressed: (1) How should authoritative sources be addressed in the process of discovery (as opposed to solely in the process of justification) in order to better justify a decision? (2) What are the relations between individual or subjective elements of decision-making (e.g., the arbiter’s particular insights that he himself reached in that specific case) and what I call its collective elements (e.g., precedents, doctrine, legislation, a specific legal culture). My hypothesis is that the quality of these relations as well as the quality of the assessment of authoritative sources in the process of discovery (as opposed to merely their demonstration in the act of justification) have a direct impact on the weight of the justification of a particular decision. And that would be, inter alia, reason enough for the elements of discovery to figure in the process of justification.

References

• Anderson 1996
  B. Anderson, ‘Discovery’ in Legal Decision-Making, Dordrecht: Kluwer Academic Publishers 1996.

• David 2001
  M. David, ‘Truth as the Epistemic Goal,’ in: M. Steup, Knowledge, Truth, and Duty: Essays on Epistemic Justification, Responsibility, and Virtue, Oxford: Oxford University Press 2001.

• Dworkin 2011
  R. Dworkin, Justice for Hedgehogs, Cambridge, MA: Harvard University Press 2011.

• Feteris & Klossterhuis 2011
  E. Feteris & H. Klossterhuis, ‘Law and Argumentation Theory: Theoretical Approaches to Legal Justification,’ in: B. van Klink & S. Taekema, Law and Method, Tübingen: Mohr Siebeck 2011.

• Golding 1992
  M.P. Golding, ‘A Note on Discovery and Justification in Science and Law,’ in: A. Aarnio & N. MacCormick, Legal Reasoning Volume I, Aldershot: Dartmouth 1992.

• Hage 2005
  J. Hage, Studies in Legal Logic, Dordrecht: Springer 2005.

• Hoyningen-Huene 2006
  P. Hoyningen-Huene, ‘Context Of Discovery Versus Context Of Justification and Thomas Kuhn,’ in: J. Schickore & F. Steinle, Revisiting Discovery and Justification, Dordrecht: Springer 2006.

• Lonergan 1992
  B. Lonergan, I nsight: A Study of Human Understanding, Toronto: University of Toronto Press 1992.

• Lonergan 1971
  B. Lonergan, Method and Theology, Toronto: University of Toronto Press 1971.

• Lyons 1992
  D. Lyons, ‘Justification and Judicial Responsibility,’ in: A. Aarnio & N. MacCormick, Legal Reasoning Volume I, Aldershot: Dartmouth 1992.

• MacCormick 1997
  N. MacCormick, Legal Reasoning and Legal Theory, Oxford: Clarendon Press 1997.

• Michelon 2006
  C. Michelon, Being Apart from Reasons: The Role of Reasons in Public and Private Moral Decision-Making, Dordrecht: Springer 2006.

• Raz 1999
  J. Raz, Practical Reason and Norms, Oxford: Oxford University Press 1999.

• Sartorius 1992
  R. Sartorius, ‘The Justification of the Judicial Decision,’ in: A. Aarnio & N. MacCormick, Legal Reasoning Volume I, Aldershot: Dartmouth 1992.

• Wasserstrom 1961
  R. Wasserstrom, The Judicial Decision: Toward a Theory of Legal Justification, London: Oxford University Press 1961.

Noten

• 1 Such as Raz’s ‘reasons for conformity’ and ‘reasons for compliance’, or Dworkin’s evaluation whether a particular decision was taken ‘responsibly.’ For more details and references, see notes 12-15 below.

• 2 Anderson 1996.

• 3 Anderson 1996, p. 1.

• 4 Anderson 1996.

• 5 MacCormick 1997, p. 16.

• 6 As the distinction is commonly referred to in epistemology, e.g., Hoyningen-Huene 2006, p. 119-131.

• 7 Wasserstrom 1961, p. 27.

• 8 For related uses of discovery/justification and similar terminology, see, e.g., Golding 1992, p. 109-125; Sartorius 1992, p. 127-146; Lyons 1992, p. 147-166; Michelon 2006, p. 114-117; Hage 2005, p. 35 et seq.; Feteris & Klossterhuis 2011, p. 253-273.

• 9 Wasserstrom 1961.

• 10 MacCormick 1997, p. 15 et seq.

• 11 This justification by processes is not akin to my use of the term process, whether in the process of discovery or the process of justification. They are different concepts, which points to the need of this work of clarification. When authors use justification by process, they are often not referring to justification at all but to legitimacy. Think about these two examples: (1) In a given legal system, all candidates to the presidential election ought to have the same time of exposure on both television and radio programs. (2) In a given legal system, all court proceedings ought to be open to the public. In neither example the outcome (the result of the election or the judicial decision) is justified by the fact that the procedure was followed, but following the procedure is part of what legitimizes the outcome. Throughout the article it will be clear that what I mean by process of discovery and process of justification is not related to procedural legitimacy, even though they are not mutually excluding components of judicial legitimacy.

• 12 Dworkin 2011.

• 13 Dworkin 2011, p. 100.

• 14 Raz 1999.

• 15 Raz 1999, p. 179 et passim.

• 16 MacCormick 1997, p. 15 et seq.

• 17 Bernard Lonergan (1904-1986). Among his major works are Lonergan 1992 and Lonergan 1971.

• 18 Anderson 1996, p. 97-98.

• 19 Anderson 1996, p. 98.

• 20 Anderson 1996, p. 122.

• 21 Anderson 1996, p. 98-99.

• 22 Anderson denominates it as ‘normal expression.’ However, ‘normal’ could lead to different interpretations — one can conclude the ‘normal’ is to provide an incomplete expression. By suggesting the term complete, I am also avoiding to use the word ‘sufficient’ (used by other authors such as MacCormick), because I understand that in many cases — especially those in which parties do not disagree with the decision taken by the judge — an ‘incomplete expression’ could be considered as ‘sufficient justification’ as to justify the decision.

• 23 One important difference between what Anderson is proposing and what authors who rely on a rigid distinction between discovery and justification would accept as sufficient justification is comprised by the 4 ^th item. If there is no relevant relation between discovery and justification, all of my mistakes, dead ends, miscomprehensions and so forth, are also irrelevant for the justification of the decision. For the proponents of the rigid distinction, the ‘final product’ and the reasons that support it are completely independent from the steps one took to reach it.

• 24 Anderson 1996, p. 152-153.

• 25 Anderson 1996, p. 78.

• 26 Agreement or disagreement with the conclusion is irrelevant in this case, as we are assessing the justification and not the truth-value of the conclusion. Lawyers often conflate truth and justification. If we subscribe to the definition of knowledge as true, justified belief, then it follows that truth and justification are not interchangeable. It is not sufficient to demonstrate that it is true that it is raining outside in order to justify a belief I had that it is raining outside. If I never checked outside and came to the conclusion that it is raining as a result of a coin flip, my belief happens to be true but it was not justified. In jurisprudence, this mistake takes the form of believing that if a decision is in accordance with the law, this is not only a necessary but a sufficient condition of its justification. It is important that we recall here that my overall argument is an epistemological, and not a metaphysical one. For more on this distinction, see David 2001, p. 153-154. The relations between the truth-value and the justification of legal decisions are a relevant part of this debate, but an analysis of these relations has been left out of this paper due to space constraints.

• 27 Anderson 1996, p. 157.

• 28 Anderson 1996, p. 101.

• 29 Anderson 1996, p. 108-109.

• 30 Anderson 1996, p. 128.

• 31 Reasons of interpersonal justification often operate as reasons of personal justification, i.e., I may only be satisfied with a decision once I believe I am able to justify it to other people.

• 32 We must keep in mind that we may be referring to a decision that does not necessarily have to be justified to others.

• 33 Anderson 1996, p. 80-92.

• 34 Whether an arbiter has the same requirements as a court judge to justify his decisions is irrelevant to the present purpose of demonstrating that the study of discovery, its elements, and its relations to justification can cast better light onto the criteria by which the justification of legal decisions can be assessed. In fact, Anderson’s other example (Anderson 1996, p. 61-78) to illustrate the same points is a decision by the Supreme Court of Canada. However, the case – which involves the legality of abortion in Canada – is too dense to be presented here due to space limitations.

• 35 Insurance Act R.S.N.S. 1989, c. 231 and the Claims Agreement, Insurance Bureau of Canada, Section 111, Rule 11A, apud Anderson 1996, p. 81 and 91-92.

• 36 Anderson 1996, p. 81.

• 37 Based on Anderson’s analysis, as seen in Anderson 1996, p. 82-86.

• 38 This is the key difference between this approach and the justification of a legal decision as proposed by, e.g., MacCormick or Raz. For them the ways in which the decision was reached are not only unnecessary but also a hindrance to proper assessment of the decision, while for Anderson these elements are precisely what allows us to better assess if the decision is justified.