• 1. Introduction

  Legal research can be pursued in different ways. In Civil Law countries, most students and researchers focus on traditional legal research, also known as doctrinal research or, in the UK, expository research.¹xIn the Common Law tradition, the focus on doctrinal research is less dominant than in the Civil Law tradition (Van Gestel, Micklitz & Rubin 2017; Kumm 2009). In this type of research, researchers systematically map the law on a certain subject, or as Smits puts it: ‘research that aims to give a systematic exposition of the principles, rules and concepts governing a particular legal field or institution and analyses the relationship between these principles, rules and concepts with a view to solving unclarities and gaps in the existing law’ (Smits 2017, 210). However, most doctrinal scholars go beyond mere construction of legal doctrine. They also evaluate the law and give recommendations for reform (Taekema 2011; Van der Burg 2019). Frequently, they also include additional disciplines in their research, so that it becomes interdisciplinary doctrinal research.
  One discipline that may be fruitfully combined with doctrinal research is philosophy. In some types of doctrinal research, legal philosophy can hardly be avoided. For research on the trias politica, the function of constitutional rights or the suggestion that a referendum should be introduced in the Netherlands, questions of a philosophical nature must be addressed, and philosophical literature is highly relevant. In fact, for almost any doctrinal subject, there is a relevant philosophical dimension so that philosophical analysis could provide more depth to the research.

  Including philosophy in a doctrinal research project may seem daunting: how can legal scholars without philosophical training give a good account of philosophical arguments or theories? Indeed, constructing original philosophical theories may be asking for too much. However, in this article we aim to show that a more modest use of philosophical methods in a doctrinal project is feasible. There are many different philosophical methods, and they are often correlated with specific philosophical traditions. A Kantian philosopher not only has different substantive views compared with those of a utilitarian, but also uses different methods. There are, however, also some general methodological approaches that can be useful and feasible for doctrinal scholars.
  The aim of this article is to give an account of three general philosophical methods that, in our view, are most relevant and feasible for doctrinal legal scholars. When teaching research methods classes to legal Ph.D. students, we could not find suitable general materials to provide a starting point for their philosophical research. Available texts usually are restricted to specific philosophical traditions or focus on substantive issues, not on methods. This text is meant as a general introduction for doctrinal scholars without a philosophical background, rather than as a contribution to philosophical debates on methods.
  This restriction to philosophy as an enrichment of doctrinal research has implications for which methods to discuss. We focus on three methods that are largely continuous with methods used in doctrinal research.²xOn methods of doctrinal research, see Smits (2017, 221-227), Kestemont (2018) and Vranken (2012). These have the advantage of being partly familiar and therefore do not require extensive philosophical training. Philosophy and law share basic features of doing research in the humanities, being interpretive and argumentative disciplines. The way these basic features are shaped in the process of doing research differs, however. In this article, we pay attention to the shared characteristics but the main focus is on the specifics of philosophical methods.
  In the following, we discuss three methods of legal philosophy: argumentation analysis and construction (Section 3), author analysis (Section 4) and reflective equilibrium (Section 5). In research projects these three methods are usually combined: a researcher may begin with studying the available literature, go on to analyse the arguments in the literature and then to submit the positions and arguments to a reflective equilibrium process. Yet the three methods can be clearly distinguished. Before we turn to the methods, we make some general remarks on legal philosophy as an auxiliary discipline to doctrinal research (Section 2).

• 2. Some Remarks on Legal Philosophy

  There are many views on what legal philosophy is. We present our own view here, in which we focus on legal philosophy as an auxiliary discipline to doctrinal research. Obviously, legal philosophy is important as an independent discipline in its own right. However, focusing on how it can enrich doctrinal research enables us to ignore various elements that are less relevant for interdisciplinary projects.³xFor an account of different methods of legal philosophy as a separate enterprise, see Coleman (2004) and Langlinais and Leiter (2016). Both focus on the debate in Anglo-American legal philosophy. A more general account of German legal philosophy (with some attention for method) is Klatt (2007). We take a broad view of the field of legal philosophy, seeing it as a part of practical philosophy that is continuous with moral and political philosophy (compare Postema 1998). All of these fields are concerned with questions of human action and practices. Legal philosophy can then be defined as that part of philosophy specifically concerned with the practice of law.
  When looking at the use of legal philosophy for doctrinal research, we suggest that it is helpful to distinguish three complementary perspectives: legal philosophy as an activity, as insights and as theories.⁴xThis distinction correlates with the threefold distinction that is common in Dutch law schools with regard to learning objectives for courses, in terms of skills (the activity of doing philosophy), insight (understanding central concepts and principles) and knowledge (of philosophical theories). These three perspectives are obviously related and rely on each other; even so, they should be distinguished. Reducing legal philosophy to one of the three does not do justice to the richness of what philosophy has to offer.
  First, legal philosophy is an activity, that of systematic reflection on law. This includes very general topics, such as the question of what law is and how it is connected to the state and to morality. It also includes reflection on fundamental principles and concepts, such as human rights, democracy and rule of law. And it includes specific questions, such as whether the law should ban cannabis, pornography or burqas. What distinguishes philosophy from doctrinal research is that the answers cannot primarily be found in legal sources. Philosophers mostly go beyond those sources, and often do not limit themselves to reflection on a particular legal order. The activity of systematic reflection on law is pursued mainly in two forms: as conceptual or analytical thinking or as normative thinking. The first form sees systematic reflection as the clarification of concepts and principles used in law. It may, for instance, clarify what legal interpretation is or how to understand the concept of rights. The second form provides normative reflection on what law should be or what principles ought to be accepted. It may, for instance, argue that law should allow euthanasia, or that freedom of religion should not be a separate human right.
  Second, legal philosophy can be seen as a collection of insights regarding central or fundamental concepts and principles, or the ideas behind the legal order. For example: what are the justifications for free speech? Here, one may use the idea put forward by John Stuart Mill, namely that free speech is necessary to maintain a public debate in which the best ideas may win out, so that the truth may prevail (Mill 1989). Usually, philosophical analysis does not lead to only one possible answer, but to a set of defensible answers – even though individual scholars may strongly believe that one of these answers is the best. In relation to free speech, other scholars have argued that it is a necessary component for human flourishing to be able to express yourself (Marmor 2018). For doctrinal scholars, insights such as these are often important because they are important building blocks for criticism as well as for constructing legal doctrine as a coherent system, in light of the fundamental principles and concepts of a legal order.
  Third, legal philosophy consists of theories. Some legal philosophers have spent much time studying a certain subject and have come up with coherent general theories as answers to its problems. For example, John Rawls developed a theory of justice, Ronald Dworkin a theory of rights and H.L.A. Hart a theory of law (Rawls 1971; Dworkin 1978; Hart 1994). These influential books are the product of thorough systematic reflection by intellectual giants. However, they have been heavily criticized by other philosophers. Philosophical theories are always controversial and subject to debate. Philosophy is not only systematic reflection by individual thinkers, but also consists of critical discussion with other philosophers. When doing philosophy, it is often good to begin by studying influential texts like these. A reader may agree with them or – partly – disagree. Understanding precisely why one disagrees may suggest directions for developing better answers. Philosophical articles often start with the exposition of a philosophical theory, followed by a critical analysis of why some of its elements are defensible and others are not, resulting in suggestions for amending the theory or sometimes completely rejecting it.
  In this article, our primary focus is on the practice of doing legal philosophy, on systematic reflection. That reflection, of course, usually builds on available philosophical theories and may result in insights about our legal order. Here, the three perspectives on legal philosophy are clearly connected. Doctrinal research is also a practice, and we are interested here in how that practice can be improved and enriched by including philosophical subprojects. When doctrinal scholars want to use philosophical insights and theories, they usually cannot simply appeal to the content of those theories: they have to be translated and transformed in order to be able to incorporate them into doctrinal analysis. What is at stake in such translation and transformation can be explained by making a distinction between internal and external perspectives on law.⁵xFor a discussion of this distinction see Tamanaha (1996).
  Doctrinal scholars typically take an internal perspective on law. They look at law as participants in the legal practice – as lawyers, but also as ordinary citizens – would do. They construct and reconstruct doctrine as the applicable law of a specific legal order. Even so, doctrinal research is moderately internal: it is open for input from outside the legal order, such as empirical and philosophical insights (Taekema 2011). Philosophical theories may take either an internal or an external perspective on law. If they take an internal perspective, their work is fairly easy to combine with the internal perspective of legal doctrine. The main difference between the doctrinal and philosophical internal perspective is that the latter makes more use of arguments that go beyond the primary sources of law, as we have already noted. This makes the difference between philosophical and doctrinal research gradual.
  However, many philosophical theories do not take an internal perspective, but an external one. They do not situate themselves as discussion partners of participants in the legal practice, but as observers or critics of that practice, or of legal orders in general. Often, their proponents aim to provide general or even universal theories about law or construct a coherent theory about a just society. They may also criticize a particular law as not meeting the standards of their theories. However, they do this not as participants, but as outsiders to legal practice. The external perspective can yield interesting insights and theories, but these are not directly applicable in the context of doctrinal research. Because the arguments are addressed to a different audience in a different context, they need to be translated and transformed to fit them within the doctrinal framework.⁶xSee the two special issues of Erasmus Law Review, 2015, 8 (2) and 8 (3) on The Incorporation Problem in Interdisciplinary Legal Research, and our Introduction to these special issues, Taekema and Van der Burg (2015). These external theories are thus sharply different from doctrinal work, in contrast to philosophies with an internal perspective, which differ gradually, and thus warrant more caution when used in a doctrinal context.⁷xThey may combine easily with other external perspectives on law, such as sociology of law. For an example of the use of Hart in such a context, see Tamanaha (1993).
  A barrier to including legal philosophy in doctrinal legal research is the perception that it is too difficult. Although this may indeed be the case, this is a problem with interdisciplinarity in general: all other disciplines have their own characteristics that may be hard to master for someone who has not studied the discipline extensively (Van Klink and Taekema 2011). We highlight two of these characteristics that are particularly visible in philosophy. First, disciplines have their own jargon and technical language, which is different from doctrinal terminology (Van Klink and Taekema 2011, 19-20). Even if a philosophical text uses the same words such as 'proof' or 'responsibility', these terms often have a meaning different from that in law. In philosophy, the problem is not merely that the discipline has its own technical terminology, but that frequently each theory has its own distinct jargon. Second, disciplines have their own methods and techniques, distinct from the methods and techniques that lawyers use (compare Taekema 2011, 47-49). Philosophical texts often require highly abstract and complex thinking.
  The characteristic of abstract thinking is perhaps the biggest obstacle to engage with philosophy because it makes direct application difficult. Very abstract and general theories – which usually take an external perspective – are not directly applicable to the more situated and concrete problems with which lawyers are concerned.⁸xCf. for a similar thesis see Vilaça (2015), arguing that most legal theory is irrelevant for conventional legal practice in European civil law countries. This is particularly true of one part of legal philosophy that is called ‘general jurisprudence’, which has a focus on understanding the concept of law.⁹xFor example, according to Julie Dickson, the core business of analytical jurisprudence is ‘to isolate and explain those features which make law into what it is’ (Dickson 2001, 17). Such legal philosophers search for general characteristics that are common to all legal orders, and as a result the analysis is extremely abstract and, therefore, less directly relevant to doctrinal research. After all, the criterion of what counts as law for the purposes of legal practice is not to be found in such a general or even universal abstract philosophy, but in the specific rules of recognition of that practice.¹⁰xFor a critique, see Vilaça (2015, 787-790) and Van der Burg 2014, 79-89). Obviously, these rules may sometimes be influenced by philosophical theories. Abstract theories are also very common in modern philosophy more generally, whether it is ethics or philosophy of science. In the fields of normative moral and political philosophy, scholars construct ideal theories, e.g. on social contracts in an imagined state of nature or principles of justice in an ideal society.¹¹xSee Robeyns (2008) for a discussion of ideal theory and non-ideal theory. Obviously, these theories are not directly applicable to concrete legal orders in our non-ideal societies, and understanding them may often require a major effort for non-philosophers.¹²xJohn Rawls, currently the most influential proponent of an ideal theory, explicitly acknowledges that it cannot be directly applied to concrete issues. The only example he discusses of how we can go from ideal theory to non-ideal theory is that of civil disobedience (Rawls 1971/1999, 244-248 and 351-355). As these abstract theories often take an external perspective, there is the need for translation and transformation when using them as arguments for law reform. This requires creative thinking on the part of the legal researcher who draws upon ideal theories. However, these theories may sometimes be used effectively in criticizing existing legal arrangements from an external perspective, and, depending on the exact legal research topic and aim, this may be useful.
  Fortunately, there is also an increasing body of philosophical work in non-ideal theory, constructing normative arguments to address real problems, which is an approach that is closer to legal practice. For example, considering the question whether vaccination against contagious diseases should be legally obligated, it is possible to draw on moral and political philosophy for supporting arguments.¹³xPierik (2018) provides arguments to support mandatory vaccination. Of course, studies like these often indirectly also build on work in ideal theory, but the advantage of relying on them is that philosophers have already done part of the difficult job of translation and transformation from ideal theory to non-ideal circumstances. Many studies in non-ideal theory are done from the internal political perspective of a particular society, or group of societies, rather than from a universal or general perspective applicable to all societies.¹⁴xThe research is often meant to apply to a particular group of societies familiar to the researcher, such as contemporary Western societies (e.g. Engster 2015, 14). Even if this perspective is not necessarily identical to the internal perspective of the legal order, the analysis is usually not radically different either. For example, interpretations of rights such as privacy and freedom of religion by philosophers reflecting on political and societal debates in particular societies are usually continuous with the legal interpretations in the legal order of that society – even if there may be differences in meaning, they will most often not be diametrically opposed. Consequently, translation and transformation from legal philosophy to doctrinal scholarship is usually fairly unproblematic when the focus is on non-ideal theory.
  With these features of legal philosophy in mind we can now turn to a discussion of three philosophical methods that we regard as feasible and relevant for doctrinal researchers. These are methods that are common in legal philosophy but that are also used widely in different fields of philosophy. For this reason, we speak of philosophical methods rather than legal philosophical methods. To avoid misunderstanding, although these are three core methods for philosophy, they are not unique to philosophy. Most disciplines, including law, study literature and analyse arguments. There is a continuity between legal scholarship and legal philosophy rather than a rigid demarcation. However, because these three methods are core methods of philosophy, philosophers have elaborated these methods more in depth than most other disciplines.

• 3. Argumentation Analysis and Construction

  The first of the three methods is argumentation analysis. Philosophy and law are both argumentative practices.¹⁵xHere, we refer both to the real-world practices of doing law and philosophy and to the academic disciplines: all of these share this characteristic. In court proceedings, the parties try to convince the judge to adopt their point of view, and for this, they provide arguments, including counterarguments against the other party’s arguments. Parliamentary procedures are shaped in a way that makes it possible to have open debates about legislative bills and policies. In legal philosophy, it is all about the force of arguments. How strong are the arguments pro and contra introducing a referendum, a change in organ donation legislation or abolishing – or increasing – the inheritance tax? Doctrinal scholars, legal sociologists and legal philosophers can complement each other here. While doctrinal scholars derive their arguments mostly from doctrinal sources and sociologists base themselves on empirical data, philosophers focus on arguments that go beyond positive law, where the conclusions are mostly of a normative nature (‘this law needs to be changed’). However, philosophers may also use arguments pro and contra a certain conception of a general notion (for example, that democracy is based on the will of the majority of the people), providing conceptual arguments rather than normative ones.
  Unlike arguments in mathematics and logic, most legal philosophical arguments do not have a strictly knockdown character. They are merely plausible, and plausibility is a matter of degree. Argumentation analysis can help determine the plausibility of each argument and of the argumentation as a whole. We can make a distinction between sound and convincing arguments.¹⁶xFor this distinction, see Van der Burg (2019, 23). Disagreement is inherent in philosophy, as it is in law and in legal research. The primary standard for determining the quality of arguments is that the argument must be sound, that is, it must be logically valid and it must be defensible.¹⁷xFeteris and Kloosterhuis (2011, 255) define a sound argument as being a logically valid one with acceptable premises. Instead of acceptable, however, we prefer the weaker notion of defensible. When editors or supervisors have to decide whether to accept an article or a dissertation, the criterion should be soundness. They need not agree with the author; they may even strongly disagree. Convincingness has a more subjective dimension, because an argument that convinces one audience need not convince some other audiences. In the academic world, there are many debates in which both parties regard each other’s positions as respectable and defensible but are not convinced by the arguments.
  Argumentation analysis may focus on different aspects of the argumentation. Here we discuss three aspects.¹⁸xThe ordering of the three is arbitrary; scholars will frequently switch between them throughout the research process.
  The first step in an argumentation analysis is making an inventory or list: to map all the arguments pro and contra a certain point of view and see if there are arguments missing in the debate. This will result in an argumentation overview, a list of arguments pro and contra. This requires collecting arguments from different sources, such as parliamentary documents, newspaper articles and philosophical publications. Often, on close analysis, important arguments or contra-arguments are found to be missing in the debate; scholars should determine whether this is the case and then add them. Therefore, the overview requires not merely collecting arguments, but also constructing new ones. This requires creative thinking.
  A second step is an analysis of the arguments separately. Is each of the arguments sound and convincing? For legal arguments, the question is whether the interpretation of a certain statute or judicial decision is correct. For empirical arguments, we should examine whether the statement is factually correct. In daily practice this can sometimes be done through direct observation, but in research, we frequently need to resort to sources such as witness statements or sociological publications to support or falsify the statement.
  In philosophical analysis of arguments, reliance on legal sources or facts is usually not enough. How do we assess whether a referendum reinforces democracy or rather weakens it? Or whether the rule of law implies that a retroactive statute violates the rule of law? There are basically two routes to construct philosophical arguments. The first is engaging one’s own creativity: by critically reflecting on existing arguments, scholars may come up with reasons for amending them or replacing them by constructing new arguments. The second is by examining normative views and general theories proposed by philosophers to see how these may support or rebut certain arguments. When doing philosophy, many researchers combine these two routes: they gather arguments about a certain question from various sources, including the views of philosophers – but not limited to these – and confront them with each other in a creative way. Argumentation analysis is therefore not merely analysing the arguments found in the literature, but also selecting, combining and creating arguments in order to construct a complete argumentation.
  The third step is the analysis of the argumentation as a whole: is the argumentation sound and convincing? Does conclusion C follow from arguments A and B? For instance, an argumentation has to be logically consistent. A familiar example of inconsistency is: ‘I did not kick my wife and, moreover, I did not kick her hard.’¹⁹xThis is just an example. There is an extensive literature on legal argumentation; for a recent overview, see Feteris (2017). More frequently, the steps in the argumentation are incomplete; for example, if someone derives a general rule from a few cases. Such inductive reasoning may sometimes provide a sound argumentation leading to highly plausible conclusions, but often it does not. From good experiences with two Dutch people, we cannot derive that all or even most Dutch people are nice; however, when we have met a hundred people from the Netherlands and all of them are nice, the generalization becomes a bit more plausible. Even if there are a decent number of cases, generalization will not be more than plausible. It is always rebuttable or defeasible. The same holds for argumentation: the conclusions are always provisional.
  Argumentation analysis may not provide knockdown arguments, but it can be highly relevant for doctrinal research. If we want to conclude that a law must be reformed – or, on the contrary, that it is not in need of reform – we must provide sound and convincing arguments for our positions. If we want to argue that the Supreme Court should revise and overrule its case law, we need sound and convincing arguments as well. Argumentation analysis can contribute to a higher quality of those arguments and help us to reject the weaker and reinforce the stronger ones. Moreover, it forces us to determine exactly what they agree or disagree with – points they may want to reject, refine or supplement.. Thus, argumentation analysis may uncover the (implicit) philosophical assumptions in legal research and practice, and it may provide philosophical counterarguments to legal statements. This makes an important component to improve the coherence and transparency of doctrinal research.

• 4. Author Analysis

  As was already apparent in Section 3, many arguments are not found in legal or political practice, but they have been developed by other scholars. Therefore, a second feasible method is the appeal to philosophers who have extensively studied certain subjects. Rather than constructing their own definition, scholars may use that of others, for example: ‘According to the philosopher Aristotle (Ethica Nicomachea, Book V) justice means that like cases are treated alike.’ Or ‘According to Warren and Brandeis privacy is “the right to be left alone"’ (Warren and Brandeis 1890).
  This method is very familiar in doctrinal scholarship. For a systematic exposition of positive law, we refer to authoritative authors, for example, to the Asser series in Dutch private law, or to Chitty on Contracts in English contract law. Sometimes doctrinal scholars, and especially legal practitioners, do not extensively study the case law, statutes and regulations themselves, but simply rely on an authoritative treatise or handbook. In our view, however, scholars should use the literature as a starting point or as a supplement to their own research, for example because it offers a good summary on a certain subject, or because it answers certain questions that case law and other sources cannot answer. When doctrinal scholars refer to an authoritative author, they cannot simply assume the author’s point of view to be the right one. They should investigate whether other authors agree or disagree with the author. Most importantly, they have to analyse what the arguments supporting the point of view entail and examine whether these arguments are sound and convincing. Doctrinal research often consists of a critical analysis of a prevailing doctrine, and this can lead to the conclusion that the arguments supporting the doctrine are not sound, plausible or convincing.
  This critical attitude is even more important in philosophy. Philosophy as a discipline is characterized by strong differences of opinions and perennial debates. Whereas in legal scholarship there is often a dominant or prevailing doctrine,²⁰xIn Dutch: de heersende leer, in German: herrschende Meinung. in philosophy this is rare. On almost every subject, there are contradictory views. John Rawls is arguably the most influential political philosopher of the 20th century, yet there is much literature criticizing his work. How can legal scholars without philosophical training deal with these controversies?
  Using ideas from a leading philosopher to support a point of view is an authority argument. It does have some strength but it is not decisive. We may increase the strength of the argument by presenting reasons for choosing this philosopher. The argument that there was no time to study other philosophers is not a strong one. Arguing that Rawls is the most quoted political philosopher of the last century and that his work on justice has had a big impact on political debates makes the argument a bit stronger. Perhaps we can add that many philosophers agree with Rawls’ views on a specific topic, for example, with regard to civil disobedience. If most critics agree with Rawls on this topic – even if they criticize other parts of his work – it makes the argument stronger. If, on the other hand, there have been many critiques on this specific topic, we should study these critiques and assess whether they are convincing.
  In the end, the reputation of a philosopher is not decisive. It is the authority of the argument, the strength of the analysis, that counts. We do not use authors because of their authority but because they have thought long and hard about certain problems and have come up with interesting and defensible ideas. We adopt their ideas if these ideas are plausible and well argued. If they are not, we can often use the criticisms on these ideas to get a better understanding of the subject. A major part of the philosophical literature consists of critical analysis of theories, a description of their shortcomings and suggestions to revise and refine the theories. In the end, scholars can never simply use a philosopher’s work. They have to study the author’s analysis and take their own stance on whether they find the theory sound and convincing.
  Researchers may also decide to study multiple authors. This means analysing the arguments and ideas of different philosophers, confronting and perhaps combining them. Authors often supplement each other, which makes it possible to combine elements from different theories. Sometimes philosophers criticize each other; then the criticisms and the differences as well as the overlap between the authors may be used to develop a new position. In the end, again, it depends on the strength of the arguments; in this case, the collection of arguments derived from the theories of the different authors, combined with the scholar’s own additional arguments. An important caveat is that it may seem that philosophers are discussing the same subject, whereas in fact they are talking past each other, because they use different (theoretical) frameworks to look at the same subject. Eclectically combining elements of different authors often leads to inconsistencies. This resembles the well-known insight from comparative law: we cannot randomly combine rules from French liability law with English rules and incorporate this combination in Dutch law. Even when comparative lawyers find certain ideas from abroad useful, they still have to examine whether the ideas are compatible with the legal system in which they want to incorporate them.²¹xCf. the discussions on legal transplants, e.g. in Watson (1974/1993). Similarly, doctrinal scholars should not eclectically and uncritically pick and combine elements from the theories of, for example, Bentham, Habermas and Foucault and then incorporate them into legal doctrine.

• 5. Reflective Equilibrium

  The basic idea of the third method is a simple one: we collect all relevant and initially plausible beliefs²²xWe follow Daniels (1996, 6) in using ‘beliefs’ as the broad term that includes all forms of information, judgments, opinions, tenets, considerations, etc. That includes beliefs about facts and empirical theories as well as about norms and values. and confront them with each other to test, partly reject and refine them until we have constructed a coherent theory. Imagine detectives who try to find out ‘who did it’ based on loose chunks of information.²³xFor an elaborate illustration of how a reflective equilibrium could be used to solve a crime, based on very weak and unreliable witness statements, see Elgin (2017, 69-73). They use forensic data to identify a number of possible suspects and check if it is possible to construct a coherent story, a scenario, for each of them based on traces, motives, testimonies, etc. This eliminates some of the suspects because they have a legitimate alibi or a lack of motive. However, even if at first the arguments against a certain scenario seem strong, it remains important to keep an open mind and ask further questions: perhaps someone lied about the alibi, or an unexpected motive emerges (such as a secret affair). Only when all elements fit together in a coherent story, when there are no inexplicable contradictions or alternative scenarios left, will they conclude that enough evidence has been collected for a convincing scenario.
  Philosophers can use a similar method. Suppose we wonder whether a poll tax (all citizens have to pay the same amount of tax per person) is just. Our starting point consists of so-called ‘considered judgments’: beliefs of which we are strongly convinced.²⁴xThe notion of ‘considered judgment’ was introduced in Rawls (1951). Rawls (1971/1999) gives various descriptions, but the core idea is that they are made by reasonable persons ‘in circumstances where the more common excuses and explanations for making a mistake do not obtain.’ (42) These judgments can be concrete, for example, that the marginal income tax rate should not frustrate incentives to work harder. Therefore, we reject a 90% rate. They can also be of a more general nature, such as the ability-to-pay principle, which means that persons who earn more should pay more, also relative to their income. Billionaire Warren Buffet considered it unjust that he paid a lower percentage of taxes on his income than his secretary (Buffet 2011). We also look at the relevant facts. How much income does the state need for the state budget; how high should a poll tax be to fund it? If we confront these judgments with the facts, we will obviously conclude that, with the current budget, a poll tax is unjust. The reason is that it will lead to billionaires paying way less tax than their secretaries and people with very low incomes paying so much tax that they will barely have enough money left to live on.
  Of course, this is a simple story, but it may suffice to illustrate the core idea of the method. We collect beliefs about which we are highly confident, analyse each of them for their plausibility and try to create a coherent account, in which we try to incorporate these beliefs. We do so by partly rejecting and partly revising some of them. The method has become especially famous for use in normative political theory and applied ethics, but it has a wider range; it can also be used – in slightly adapted versions – in science, history and aesthetics, for instance.²⁵xMore on reflective equilibrium: Rawls (1971/1999), Daniels (1996) and Van der Burg and Van Willigenburg (1998). Elgin (2017) uses it to construct coherent accounts in sciences, history and art. There are many versions of reflective equilibrium; we have constructed a version here that is most similar and continuous with doctrinal research. It is holistic.²⁶xCf. Rawls (1971/1999, 19), repeated at 507: ‘[J]ustification is a matter of the mutual support of many considerations, of everything fitting together into one coherent view.’ It does not find justification in foundations that are claimed to be absolutely certain, like a principle of utility or a Kantian categorical imperative, but in the acceptability of the coherent account at the end of the process. And this account is acceptable, because in light of all relevant beliefs, it is the best account available – even if it is always only a provisional result that can be modified in light of new information or further reflection.
  We need at least three types of beliefs to construct what may be called a narrow reflective equilibrium:

  1. Concrete considered judgments with regard to concrete cases and specific rules and assumptions.

  2. General considered judgments with regard to general principles, values and ideals.²⁷xThe distinction between general and concrete considered judgments is a gradual one. Rawls has switched in his work from concrete to general. In 1951, considered judgments are seen as judgments with regard to concrete cases; in 1971, examples include also more general principles such as equal liberty of conscience (1971/1999, 181). In the version of reflective equilibrium suggested here, we include both general and concrete judgments. In Rawls (1971/1999) with its focus on ideal theory, fundamental principles are a distinct category, but for more applied versions of reflective equilibrium like the one we develop here, it is more convenient to regard principles as a subset of the category of general considered judgments.

  3. Knowledge about relevant facts and empirical theories about human behaviour and our society.²⁸xThis can include concrete facts, but also insights from empirical disciplines. Reflective equilibrium can thus include insights from various disciplinary backgrounds and, hence, presents a frame that may enable the combination and integration of various subprojects in interdisciplinary projects.

  What are these considered judgments precisely? They are convictions about which we are highly confident, but not absolutely certain. We strongly believe that we can rely on them, for example, because they are not just covert self-interested judgments or prejudices, or reactions to accidental strong emotions. We have reflected on them carefully, analysed them critically and still strongly believe that they are correct. And we believe we have good arguments for them. However, considered judgments are not dogmas or axioms. They are provisional and always subject to review. On closer analysis we may discover that they are not defensible.
  There are various reasons to revise or reject considered judgments. An argument that seemed to support the initial judgment may be falsified. For example, we may believe that the crime figures increase in the Netherlands, but then we find convincing empirical research that shows the opposite. As a result, our view that punishments need to be more severe in order to fight the rising crime rate is no longer defensible either. We may also come to the conclusion that our initial judgment was the result of irrational responses, prejudice, bias or other factors that may question its reliability. Most frequently, the reason will be that our complete set of beliefs is internally inconsistent. Our acceptance of the ability-to-pay principle proves to be inconsistent with our acceptance of the principle that everyone should pay the same share. In such cases we have to revise or refine one of our judgments. We go back and forth between the various elements and reflect on how to resolve the inconsistencies until we have a coherent set of judgments: an equilibrium. This is why we call this method reflective equilibrium.²⁹xIn the literature, the term is used to refer both to the end result, where all our judgments are in a state of reflective equilibrium, and to the process or method leading up to that result. In this article we have chosen the latter version. Our existing considered judgments act as a starting point, but are not the end of the reflective process: they must be subject to reflection and can be revised and even rejected in this process.
  These three types of beliefs are central to philosophical analysis, but we may add other elements. For example, background theories, religious views, ideals, empirical information about moral views among the population and so on (Van der Burg and Van Willigenburg 1998). Which elements should be included in the process depends on the purpose of a research project. On the one hand, adding more elements makes the process more complex and makes it more difficult to construct a coherent theory, so we should be cautious not to include too large a set of beliefs.³⁰xElgin (2017) starts with the widest possible set of beliefs, consisting of all initially tenable commitments, where the reason to include a commitment may be extremely weak. We think it is better to start with only those beliefs of which we feel pretty certain; this makes the initial set more restricted and thus the process more feasible. Even so, there are degrees of certitude. Of some beliefs in this set, we can hardly imagine that we can be forced to abandon them; of others we more easily accept that the opposite position could be defensible as well. On the other hand, adding elements may provide more relevant information, especially critical input in the process, and thus may make the results more robust and reliable. There is no general rule governing what should be included or not. Feasibility and robustness are both important methodological ideals here, and they often point in different directions.
  For legal philosophical analysis aiming to be relevant for law, it is essential to include knowledge about the legal order(s). In that context, we should add a fourth set of elements:
  4. Knowledge about the relevant legal order(s) and about positive law.
  What knowledge about law is to be included, and how much weight it has in the balancing and refinement, depends on the purpose of research. If our purpose is to construct a philosophical theory of privacy that fits and justifies the European Court of Human Rights (ECtHR) case law, we include treaty texts and all relevant case law. Although some of the case law may be rejected as mistaken, it cannot be set aside radically. This is different if your purpose is to reform the law. If we want to assess whether a referendum would improve the Dutch political system, we include knowledge about current constitutional law in order to see whether a referendum somewhat fits into that. Even so, we may disregard certain constitutional rules if there are good arguments that they should be abolished on the basis of our considered judgments. After all, we want to investigate whether the constitutional law must be reformed. The weight of our considered judgments about democracy then becomes relatively stronger and that of positive law weaker.
  Reflective equilibrium may seem familiar to lawyers, as they use a broadly similar method.³¹xActually, John Rawls found part of his inspiration for the idea of reflective equilibrium in legal reasoning. For reflective equilibrium methods in the context of legal arguments, see Dworkin (1978, 160f). In describing and constructing legal doctrine, legal scholars collect the usual legal sources: case law, legislation and treaties. Case law can be compared to considered judgments with regard to concrete cases, and the rules of legislation and treaties to more general considered judgments. Doctrinal scholars go back and forth between the different elements until they have constructed a coherent doctrine. In this process they sometimes discover ambiguities and gaps in the law or contradictions. To resolve these, they must search for underlying arguments based on general principles and ideals, for example, the ability-to-pay principle or theories on democracy and the rule of law.³²xThis corresponds to Dworkin’s claim that judges should look at two dimensions in solving difficult cases. These are the dimensions of ‘fit’ to legal sources (especially case law and legislation) and the dimension of ‘political morality’: the underlying political and moral principles and values of the legal order. See Dworkin (1985, 143).
  Both in legal doctrine and in philosophical reflective equilibrium, coherence is the central methodological concept.³³xOn coherence, also in law, see Amaya (2015) and MacCormick (2005). Why is coherence important, and what does it mean? Coherence consists of two aspects: consistency and mutual support. It is obvious that the opinions of detectives, legal scholars or philosophers should be consistent. A judge or a scientist should not contradict herself. A judge who imposes a fine of 1000 euros for suspect A, while in the identical case of suspect B fully acquitting the suspect, acts inconsistently and undermines the law and her own reputation. However, merely avoiding inconsistencies is not enough. We also expect that judicial decisions fit into the legal system as such and that they are correct applications of the general rules and principles of their legal order; thus, the decisions should be supported by other decisions, rules and general principles. Similarly, we expect philosophical theories to be coherent, for example, that views on justified or unjustified killing with regard to abortion and euthanasia are consistent with views on capital punishment and that there is a relation of mutual support between the various views.
  Coherence is, of course, not the only criterion accepting certain theories or positions. A theory should not be the mere reconstruction of our prejudices and should not be based on invalid arguments, self-interest or questionable ideological positions. This brings us to the notion of a wide reflective equilibrium. This is a process in which various methods are used to provide for the maximum critical reflection and to allow for radical revisions of the initial set of considered judgments. It would take too much space to discuss them all, but we mention at least some methods to provide substantial critical input.³⁴xWhich methodological principles should be used varies with the purpose and context of the research project; therefore we can give here only some indications. In the literature on Rawls, this has been discussed mainly in terms of the appeal to background theories. However, this is a very broad category. The most important types of background theories focus on either methods to make the collection and processing of considered judgments more reliable, and on whether the theories of justice are feasible in light of general sociological and psychological insights.
  There are various methods to critically assess the set of considered judgments and factual beliefs. With regard to facts and empirical theories, science – and legal practice – knows many methodological requirements to establish them beyond reasonable doubt. With regard to considered judgments, similar methods may be used, e.g. checking our judgments against those of others, ascertaining whether we may be biased or self-interested, and so on.³⁵xOne famous method to avoid bias and self-interest is that of Rawls’ original position, where persons are required to decide on issues of justice behind a veil of ignorance, not knowing what their personal characteristics and position in society are. We may also widen the set by explicitly broadening our own moral horizon, e.g. by listening to people with different backgrounds and experiences, by reading literature, watching movies and documentaries, by trying to imagine what it would be to live in different circumstances (DePaul 1993). We want our conclusions to be as plausible and broadly acceptable as possible. This implies that we should actively search for means to introduce critical input, include perspectives that are different from our own so that we can really test our considered judgments and our provisional equilibria.
  A second set of methodological principles focuses on the output. The account that we come up with should be relatively simple and elegant and easy to apply in practice. It should be complete with regard to what we know of the field of study, and robust, so that it can deal with new cases. It should also be feasible in light of what we know about society and human behaviour.³⁶xRawls (1971/1999) includes general social theory, rational choice theory and psychological theories of moral development to demonstrate that his approach also is feasible in light of those background theories.
  As a method, reflective equilibrium is dynamic and open to continuing revision. It embraces and invites counterarguments as these can be brought into a new reflective equilibrium process, leading to further revision and refinement.³⁷xWhen we present the result of our reflective equilibrium process in a publication, we often simplify the actual process and ignore most of the intermediate stages and the beliefs we have rejected or refined. We present the end result in a coherent form and support it with an argumentation, including arguments pro and contra. We show thus that the coherent end result can take account of the most important beliefs; each of these beliefs can then be seen as arguments for our theory or our conclusion. Although the process is different from that of argumentation analysis, the justification of the result may resemble how we present an argumentation analysis. As we have argued previously, normative arguments are merely plausible, they are always defeasible. Reflective equilibrium does not guarantee absolute truth, but it can help to improve the degree of plausibility. The result is always provisional and open to future improvement. Because of its openness to considered judgments and knowledge of various kinds, reflective equilibrium makes it possible to combine philosophical arguments with legal arguments and factual knowledge. For this reason, and because its focus on coherence is similar to legal justification, it is a method that is useful and feasible in the context of interdisciplinary doctrinal research.

• 6. Conclusion

  Doctrinal scholars may sometimes want to include a philosophical dimension in their research. In this article, we have discussed three philosophical methods that can be used by scholars without a philosophical background, namely, argumentation analysis, author analysis and reflective equilibrium.
  In actual research projects, the three methods are not sharply separated; they are usually combined. It would be inefficient if scholars had to reinvent the wheel again and again; therefore, a philosophical subproject should often start with an author analysis of philosophical texts. Are the views of those philosophers sound and convincing? Do the views of different philosophers overlap, diverge or even conflict? This requires an argumentation analysis: what are the arguments pro and contra certain positions, and are these sound and convincing? Subsequently, these arguments and positions can be submitted to a reflective equilibrium process: are they coherent with our considered judgments and with the facts? Perhaps the arguments and positions derived from those philosophers lead to revising our considered judgments, or perhaps the considered judgments lead to criticizing and revising their arguments and views. We continue this process until we have formulated a clear point of view, a coherent theory; we may then incorporate it in our doctrinal research.
  Each method can be applied at various levels of sophistication. A doctrinal scholar could attempt to make a comprehensive study of the work of H.L.A. Hart and the body of literature building on and criticizing Hart. She could also merely read Hart’s main book and a few of the best-known articles that criticize it. Both choices can be justified. Similar degrees in levels of engagement exist with regard to the other methods. Whatever a researcher decides to do, we hope that this article has at least shown some ways of how to do it. Legal philosophy may sometimes be difficult for doctrinal scholars without a philosophical background, but it is possible and often highly rewarding.

References

• Amaya, A. (2015). The Tapestry of Reason. An Inquiry into the Nature of Coherence and Its Role in Legal Argument. Oxford: Hart Publishing.

• Coleman, J. (2004). Methodology. In: J. Coleman, K.E. Himma, & S. Shapiro (eds.). Oxford Handbook of Jurisprudence and Philosophy of Law (pp. 311-351). Oxford: Oxford University Press.

• Daniels, N. (1996). Justice and Justification. Reflective Equilibrium in Theory and Practice. Cambridge: Cambridge University Press.

• DePaul, M.R. (1993). Balance and Refinement. Beyond Coherence Methods of Moral Inquiry. London: Routledge.

• Dickson, J. (2001). Evaluation and Legal Theory. Oxford & Portland: Hart Publishing.

• Dworkin, R. (1978). Taking Rights Seriously. Cambridge, MA: Harvard University Press.

• Dworkin, R. (1985). A Matter of Principle. Cambridge, MA: Harvard University Press.

• Elgin, C.Z. (2017). True Enough. Cambridge, MA: MIT Press.

• Engster, D. (2015). Justice, Care and the Welfare State. Oxford: Oxford University Press.

• Feteris, E. (2017). Fundamentals of Legal Argumentation. A Survey of Theories on the Justification of Judicial Decisions. Dordrecht: Springer.

• Feteris, E. & Kloosterhuis, H. (2011), Law and Argumentation Theory. Theoretical Approaches to Legal Justification. In: B. Van Klink & S. Taekema (eds.). Law and Method. Interdisciplinary Research into Law (pp. 253-273). Tübingen: Mohr Siebeck.

• Hart, H.L.A. (1994). The Concept of Law. Oxford: Oxford University Press.

• Kestemont, L. (2018). Handbook on Legal Methodology. From Objective to Method. Cambridge: Intersentia.

• Klatt, M. (2007). Contemporary Legal Philosophy in Germany. Archiv für Rechts- und Sozialphilosophie, 93(4), 519-539.

• Kumm, M. (2009). On the Past and Future of European Constitutional Scholarship. International Journal of Constitutional Law, 7(3), 401-415.

• Langlinais, A. & Leiter, B. (2016). The Methodology of Legal Philosophy. In: H. Cappelen, T.S. Gendler, & J. Hawthorne (eds.). The Oxford Handbook of Philosophical Methodology. Oxford: Oxford University Press. doi: 10.1093/oxfordhb/9780199668779.013.9.

• MacCormick, N. (2005). Rhetoric and the Rule of Law: A Theory of Legal Reasoning. Oxford: Oxford University Press.

• Marmor, A. (2018). Two Rights of Free Speech. Ratio Juris, 31(2), 139-159.

• Mill, J.S., edited by Collini, S. (1989). On Liberty, with The Subjection of Woman and Chapters on Socialism. Cambridge: Cambridge University Press.

• Pierik, R. (2018). Mandatory Vaccination: An Unqualified Defence. Journal of Applied Philosophy, 35(2), 381-398.

• Postema, G. (1998). Jurisprudence as Practical Philosophy. Legal Theory, 4(3), 329-357.

• Rawls, J. (1951/1999). Outline of a Decision Procedure for Ethics. Philosophical Review, 60(2), 177-197; reprinted in: S. Freeman (ed.). John Rawls. Collected Papers (pp. 1-19). Cambridge, MA: Harvard University Press.

• Rawls, J. (1971/1999). A Theory of Justice. Cambridge, MA: The Belknap Press of Harvard University Press.

• Robeyns, I. (2008). Ideal Theory in Theory and Practice. Social Theory and Practice, 34(3), 341-362.

• Smits, J. (2017). What is Legal Doctrine: On the Aims and Methods of Legal-Dogmatic Research. In: R. van Gestel, H.-W. Micklitz, & E. Rubin (eds.). Rethinking Legal Scholarship: A Transatlantic Dialogue (pp. 207-228). Cambridge: Cambridge University Press.

• Taekema, S. (2011). Relative Autonomy. A Characterization of the Discipline of Law. In: H.S. Taekema & B.M.J. van Klink (eds.). Law and Method. Interdisciplinary Research into Law (pp. 33-52). Tübingen: Mohr Siebeck.

• Taekema, S. & van der Burg, W. (2015). Introduction: The Incorporation Problem in Interdisciplinary Legal Research Part 1: Theoretical Discussions. Erasmus Law Review, 8(2), 39-42. doi: 10.5553/ELR.000050.

• Tamanaha, B.Z. (1996). The Internal/External Distinction and the Notion of a “Practice” in Legal Theory and Sociolegal Studies. Law & Society Review, 30(1), 163-204. doi: 10.2307/3054037.

• Van der Burg, W. (2014). The Dynamics of Law and Morality. A Pluralist Account of Legal Interactionism. Farnham: Ashgate.

• Van der Burg, W. (2019). The Merits of Law: An Argumentative Framework for Evaluative Judgements and Normative Recommendations in Legal Research. Archiv für Rechts- und Sozialphilosophie, 105(1), 11-43.

• Van der Burg, W. & Van Willigenburg, T. (1998) (eds.). Reflective Equilibrium: Essays in Honour of Robert Heeger. Dordrecht: Kluwer Academic Publishers.

• Van Gestel, R., Micklitz, H.-W., & Rubin, E. (2017). Introduction. In: R. van Gestel, H.-W. Micklitz, & E. Rubin (eds.). Rethinking Legal Scholarship: A Transatlantic Dialogue (pp. 1-27). Cambridge: Cambridge University Press.

• Van Klink, B. & Taekema, S. (eds.). (2011). Law and Method. Interdisciplinary Research into Law. Tübingen: Mohr Siebeck.

• Vilaça, G.V. (2015). Why Teach Legal Theory Today? German Law Journal, 16(4), 781-820. Retrieved from https://heinonline.org/HOL/P?h=hein.journals/germlajo16&i=794.

• Vranken, J.B.M. (2012). Exciting Times for Legal Scholarship. Law and Method, 2(2), 42-62. doi: 10.5553/ReM/221225082012002002004.

• Warren, S.D. & Brandeis, L.D. (1890). The Right to Privacy. Harvard Law Review, 4(5), 193-220. doi: 10.2307/1321160.

• Watson, A. (1974/1993). Legal Transplants. An Approach to Comparative Law. Athens and London: University of Georgia Press.

Noten

• * A previous version of this text has been presented to classes at Zagreb, Queen Mary University of London and Erasmus School of Law, Rotterdam; we profited from the feedback of our students. We also want to thank Irma Bluijs, Noam Gur, Machteld Geuskens, Stuart Goosey, Dimitrios Kyritsos and Tamar de Waal for their helpful comments on previous versions of this article, and Jacqueline Brand, Alma Besic, Lieske Bottema and Haris Sabanovic for providing assistance with the research in various ways. This article is part of a series. In a second article, we hope to discuss the various purposes for which doctrinal scholars can use philosophy.

• 1 In the Common Law tradition, the focus on doctrinal research is less dominant than in the Civil Law tradition (Van Gestel, Micklitz & Rubin 2017; Kumm 2009).

• 2 On methods of doctrinal research, see Smits (2017, 221-227), Kestemont (2018) and Vranken (2012).

• 3 For an account of different methods of legal philosophy as a separate enterprise, see Coleman (2004) and Langlinais and Leiter (2016). Both focus on the debate in Anglo-American legal philosophy. A more general account of German legal philosophy (with some attention for method) is Klatt (2007).

• 4 This distinction correlates with the threefold distinction that is common in Dutch law schools with regard to learning objectives for courses, in terms of skills (the activity of doing philosophy), insight (understanding central concepts and principles) and knowledge (of philosophical theories).

• 5 For a discussion of this distinction see Tamanaha (1996).

• 6 See the two special issues of Erasmus Law Review, 2015, 8 (2) and 8 (3) on The Incorporation Problem in Interdisciplinary Legal Research, and our Introduction to these special issues, Taekema and Van der Burg (2015).

• 7 They may combine easily with other external perspectives on law, such as sociology of law. For an example of the use of Hart in such a context, see Tamanaha (1993).

• 8 Cf. for a similar thesis see Vilaça (2015), arguing that most legal theory is irrelevant for conventional legal practice in European civil law countries.

• 9 For example, according to Julie Dickson, the core business of analytical jurisprudence is ‘to isolate and explain those features which make law into what it is’ (Dickson 2001, 17).

• 10 For a critique, see Vilaça (2015, 787-790) and Van der Burg 2014, 79-89). Obviously, these rules may sometimes be influenced by philosophical theories.

• 11 See Robeyns (2008) for a discussion of ideal theory and non-ideal theory.

• 12 John Rawls, currently the most influential proponent of an ideal theory, explicitly acknowledges that it cannot be directly applied to concrete issues. The only example he discusses of how we can go from ideal theory to non-ideal theory is that of civil disobedience (Rawls 1971/1999, 244-248 and 351-355).

• 13 Pierik (2018) provides arguments to support mandatory vaccination. Of course, studies like these often indirectly also build on work in ideal theory, but the advantage of relying on them is that philosophers have already done part of the difficult job of translation and transformation from ideal theory to non-ideal circumstances.

• 14 The research is often meant to apply to a particular group of societies familiar to the researcher, such as contemporary Western societies (e.g. Engster 2015, 14).

• 15 Here, we refer both to the real-world practices of doing law and philosophy and to the academic disciplines: all of these share this characteristic.

• 16 For this distinction, see Van der Burg (2019, 23).

• 17 Feteris and Kloosterhuis (2011, 255) define a sound argument as being a logically valid one with acceptable premises. Instead of acceptable, however, we prefer the weaker notion of defensible.

• 18 The ordering of the three is arbitrary; scholars will frequently switch between them throughout the research process.

• 19 This is just an example. There is an extensive literature on legal argumentation; for a recent overview, see Feteris (2017).

• 20 In Dutch: de heersende leer, in German: herrschende Meinung.

• 21 Cf. the discussions on legal transplants, e.g. in Watson (1974/1993).

• 22 We follow Daniels (1996, 6) in using ‘beliefs’ as the broad term that includes all forms of information, judgments, opinions, tenets, considerations, etc. That includes beliefs about facts and empirical theories as well as about norms and values.

• 23 For an elaborate illustration of how a reflective equilibrium could be used to solve a crime, based on very weak and unreliable witness statements, see Elgin (2017, 69-73).

• 24 The notion of ‘considered judgment’ was introduced in Rawls (1951). Rawls (1971/1999) gives various descriptions, but the core idea is that they are made by reasonable persons ‘in circumstances where the more common excuses and explanations for making a mistake do not obtain.’ (42)

• 25 More on reflective equilibrium: Rawls (1971/1999), Daniels (1996) and Van der Burg and Van Willigenburg (1998). Elgin (2017) uses it to construct coherent accounts in sciences, history and art. There are many versions of reflective equilibrium; we have constructed a version here that is most similar and continuous with doctrinal research.

• 26 Cf. Rawls (1971/1999, 19), repeated at 507: ‘[J]ustification is a matter of the mutual support of many considerations, of everything fitting together into one coherent view.’

• 27 The distinction between general and concrete considered judgments is a gradual one. Rawls has switched in his work from concrete to general. In 1951, considered judgments are seen as judgments with regard to concrete cases; in 1971, examples include also more general principles such as equal liberty of conscience (1971/1999, 181). In the version of reflective equilibrium suggested here, we include both general and concrete judgments. In Rawls (1971/1999) with its focus on ideal theory, fundamental principles are a distinct category, but for more applied versions of reflective equilibrium like the one we develop here, it is more convenient to regard principles as a subset of the category of general considered judgments.

• 28 This can include concrete facts, but also insights from empirical disciplines. Reflective equilibrium can thus include insights from various disciplinary backgrounds and, hence, presents a frame that may enable the combination and integration of various subprojects in interdisciplinary projects.

• 29 In the literature, the term is used to refer both to the end result, where all our judgments are in a state of reflective equilibrium, and to the process or method leading up to that result. In this article we have chosen the latter version.

• 30 Elgin (2017) starts with the widest possible set of beliefs, consisting of all initially tenable commitments, where the reason to include a commitment may be extremely weak. We think it is better to start with only those beliefs of which we feel pretty certain; this makes the initial set more restricted and thus the process more feasible. Even so, there are degrees of certitude. Of some beliefs in this set, we can hardly imagine that we can be forced to abandon them; of others we more easily accept that the opposite position could be defensible as well.

• 31 Actually, John Rawls found part of his inspiration for the idea of reflective equilibrium in legal reasoning. For reflective equilibrium methods in the context of legal arguments, see Dworkin (1978, 160f).

• 32 This corresponds to Dworkin’s claim that judges should look at two dimensions in solving difficult cases. These are the dimensions of ‘fit’ to legal sources (especially case law and legislation) and the dimension of ‘political morality’: the underlying political and moral principles and values of the legal order. See Dworkin (1985, 143).

• 33 On coherence, also in law, see Amaya (2015) and MacCormick (2005).

• 34 Which methodological principles should be used varies with the purpose and context of the research project; therefore we can give here only some indications. In the literature on Rawls, this has been discussed mainly in terms of the appeal to background theories. However, this is a very broad category. The most important types of background theories focus on either methods to make the collection and processing of considered judgments more reliable, and on whether the theories of justice are feasible in light of general sociological and psychological insights.

• 35 One famous method to avoid bias and self-interest is that of Rawls’ original position, where persons are required to decide on issues of justice behind a veil of ignorance, not knowing what their personal characteristics and position in society are.

• 36 Rawls (1971/1999) includes general social theory, rational choice theory and psychological theories of moral development to demonstrate that his approach also is feasible in light of those background theories.

• 37 When we present the result of our reflective equilibrium process in a publication, we often simplify the actual process and ignore most of the intermediate stages and the beliefs we have rejected or refined. We present the end result in a coherent form and support it with an argumentation, including arguments pro and contra. We show thus that the coherent end result can take account of the most important beliefs; each of these beliefs can then be seen as arguments for our theory or our conclusion. Although the process is different from that of argumentation analysis, the justification of the result may resemble how we present an argumentation analysis.