- 1 Ross (1953: 386 / 2013: 388; my translation). For reasons unknown, this and quite a few other key p (...)
It must fall outside a work in jurisprudence to enter into a comprehensive discussion of fundamental philosophical problems. It must be allowed to simply declare a standpoint and refer to the fact that this standpoint is shared by a significant group of modern philosophers and philosophically interested practitioners of other disciplines.1
- 2 Even if there may be some modest signs of a turning of the tide. This special issue may be one such (...)
- 3 Thus, e.g. Leiter’s claim that Scandinavian realism ‘is today more of a museum piece than a live co (...)
- 4 Another significant factor has undoubtedly been Hart’s highly influential but largely mistaken crit (...)
- 5 Already in 1967 The Encyclopedia of Philosophy held that: ‘Logical positivism, then, is dead, or as (...)
- 6 Cf. e.g. Leiter: ‘That Scandinavian Realism should have fallen out of favor is not wholly unsurpris (...)
1It is commonly agreed that Scandinavian realism has fallen on hard times since its heyday in the early and mid-20th century2 – especially in the Anglo Saxon world where many commentators have long relegated it to the pages of history.3 One significant factor in this regard has been a growing perception that the fundamental philosophical premises, on which the Scandinavians relied, are untenable.4 This is particularly true with regard to Alf Ross who, as the only member of the movement, relied primarily on logical positivism which most modern philosophers agree is largely moribund today.5 However, on the basis of the at least superficial similarities between logical positivism and the less known Uppsala school, many legal philosophers have, perhaps unfairly, extended the negative judgment also to the other Scandinavian realists.6
- 7 Cf. Leiter 2007 for a collection of papers (the earliest stem from 1997).
- 8 Cf. Holtermann 2006, Spaak 2009, Mautner 2010. The former of these works was written in Danish and (...)
2In recent years scholars sympathetic to Scandinavian realism seem in the phenomenon known as ‘naturalism’ to have found grounds for cautious optimism that a cure for this philosophical crisis might be available. Naturalism has been high on the agenda in general philosophy for the last three or four decades and following Quine’s seminal article “Epistemology Naturalized” (1969). With roughly the usual delay this general trend has in the last decade or two been taken up also in legal philosophy. In this field, the pioneering work was done from around the turn of the century by Brian Leiter with his naturalistic reconstruction of American realism,7 but in later years parallel attempts have been made to look also at Scandinavian realism through the same philosophical prism.8
3The result of these exercises depends of course on each occasion on what exactly one means by naturalism. Unfortunately, contemporary philosophy has not been particularly helpful in this regard sporting a variety of diverging and, in some cases, mutually inconsistent understandings of the term.9 Notwithstanding the full spectrum of these competing understandings, it seems reasonable to make a rough distinction between two conceptions of naturalism that have been applied in the literature on legal realism: one that is broad and inclusive, which has been applied by Spaak and Mautner in relation to Scandinavian realism, and another that is more narrow and exclusive and distinctively Quinean, which has been applied by Leiter and Holtermann in relation to American and Scandinavian realism respectively.
4Simplifying somewhat, the broad conception can be described in the following way:
[N]aturalism can intuitively be separated into an ontological and a methodological component. The ontological component is concerned with the contents of reality, asserting that reality has no place for ‘supernatural’ or other ‘spooky’ kinds of entity. By contrast, the methodological component is concerned with the ways of investigating reality, and claims some kind of general authority for the scientific method.10
- 11 Cf. e.g. Mautner: ‘Many philosophical “-isms” are naturalist, among them philosophies known as evol (...)
5Thus defined it is clear that we are not dealing with a very specific philosophical position. Depending on how the details are spelled out, we are dealing rather with a combination of a physicalist or materialist monism and an empiricist or positivist theory of knowledge that seems to fit quite a few philosophical “-isms” throughout history. In particular, this conception of naturalism quite obviously shares an immediate affinity with the general philosophical program to which the Scandinavians subscribed and in which they carefully couched their legal theory, i.e. logical positivism and the anti-metaphysical Uppsala-school.11
- 12 Thus, Spaak: ‘naturalism has again become an important topic in core areas of philosophy’. Spaak (2 (...)
- 13 Ross (1953: 386 / 2013: 388).
6It is therefore at least prima facie understandable why Spaak and Mautner have adopted this broad conception in their attempts to revive Scandinavian realism through naturalization. The problem is, however, that this approach in effect does very little to address the specific character of the Scandinavians’ current philosophical crisis – especially if we focus, as I shall do in this article, on Alf Ross whose particular logico-positivist version of Scandinavian realism has often been at the centre of debates about the movement’s current relevance. Ross simply fits in too easily on this broad understanding of naturalism, and the label is too uninformative to provide any real argument in favour of his continued philosophical viability. Naturalizing thus understood in effect requires little more than reasserting the basic tenets of his version of Scandinavian realism, thus conveying the message that his decade-long crisis has merely been the result of temporarily being out of intellectual fashion.12 But this, I submit, is a wrong reading of the challenge Ross (and, to the degree its fate is often (rightly or wrongly) tied to his, Scandinavian realism as such) is currently facing. Naturalism’s current topicality does not mean that logical positivism is back in fashion under a new name. The fact is that Ross’s fundamental philosophical standpoint is simply no longer ‘shared by a significant group of modern philosophers’,13 and this challenge has to be taken seriously if his version of Scandinavian realism is to hold continuous relevance.
- 14 Cf. e.g. Haack 2009, Leiter 2007, Feldman 2012.
7With a view to the overall meaningfulness of the naturalizing exercise there may therefore be good reasons to think that while Spaak and Mautner’s broad approach may not necessarily be intrinsically problematic it may not be particularly rewarding either. Fortunately, things look different if we apply the narrow Quinean conception of naturalism instead. This conception, which is also known as revolutionary or replacement naturalism,14 is a quite specific version of the methodological variant described above, i.e. one that focuses on epistemology and which can preliminarily be characterized as i) the explicit rejection of so-called Cartesian foundationalism, i.e. of any attempt to derive our beliefs from indubitable foundations, and ii) the view that this normative justificatory exercise should be replaced by a descriptive empirical study of how we actually form our beliefs.
- 15 In fact, his argument originally presented in Quine 1980  directly targets only the logical p (...)
8What is particularly important in the present context is that Quine explicitly considers logical positivism to be a kind of Cartesian foundationalism.15 His naturalism is, in other words, inconsistent with logical positivism because it presupposes its failure as a philosophical program. With a view to the specific philosophical challenge confronting Ross’s version of Scandinavian realism, we should therefore expect that a successful naturalization on this conception of naturalism will provide a much stronger argument for his current philosophical viability because it would incorporate the very arguments that have been launched against his current philosophical foundations in logical positivism (by Quine and others).
9For the same reason, however, we should also expect such an exercise to be substantially different and also more demanding because it requires not merely an exegetic exercise of interpretation and reassertion but rather a comprehensive philosophical reconstruction of Ross’s realist position. More specifically, it requires dismounting his legal theory from the general, logico-positivist philosophical foundations in which he deliberately and carefully placed it and reinserting it into a fundamentally different philosophical framework, i.e. replacement naturalism.
10In spite of the difficulties, I submit that this philosophical reconstruction is in fact feasible. My argument for this claim proceeds in the following way: first, I outline how Ross conceives jurisprudence in a way that fits into the program of logical positivism; second, I develop in greater details the basic tenets of Quinean replacement naturalism; third, I show how Ross’s theory of law can be adapted into this epistemology; and fourth and finally, I show how Ross’s philosophy thus transformed provides a new and critical perspective on some ongoing discussions in contemporary philosophy.
11In order to successfully dismount Ross’s legal realism from its foundations in logical positivism we must first establish how Ross himself conceived his legal theory in relation to this general philosophy of science. This in turn presupposes a very brief account of the basic tenets of logical positivism.
12In spite of belonging to empiricism in philosophy according to which the ultimate foundation for knowledge is sense data, logical positivism is at the same time closely associated with Descartes who, as a rationalist, is traditionally contrasted with empiricism. Of course the logical positivists flatly reject Descartes’ rationalist strategy of securing knowledge from a foundation of truths of reason but they clearly adopt his basic idea that knowledge should be indubitable, and that it should therefore be based on an unshakeable foundation. As good empiricists they only insist that this foundation be constituted by basic sense data. Beyond this austere foundation the logical positivists only recognize logical truths (which they consider tautological) and truths that are capable, at least in principle, of being derived logically from sense data. Any proposition that does not fit in either of these three categories is to be rejected as metaphysics.
- 16 Ussing quoted in Ross (1953: 15 / 2013: 53; Bindreiter's translation). This particular passage is a (...)
13Turning to the doctrinal study of law, it soon becomes evident that this discipline contains a long list of propositions which it is very hard to imagine could ever be derived, even in principle, from sense data. Doctrinal textbooks routinely mention normative phenomena like rights, duties, engagements, etc. In the Danish version of On Law and Justice, Ross chooses, as a random example, a sentence from a textbook on contract law: ‘The acceptor is bound to pay the bill of exchange on the due day for payment, cf. § 28 (1) Danish Bill of Exchange Act.’16
14The epistemological challenge posed by this kind of normative proposition (in Ross’s terminology a so-called directive) is that a duty cannot be reduced to basic sense data:
- 17 Ross (1953: 16 / 2013: 53; Bindreiter's translation).
But what is an ‘obligation’, and how can one empirically determine if it has arisen? The acceptance as such, accomplished by drawing some ink lines on a piece of paper, does not, among its observable consequences, seem to have one that can be called ‘obligation’.17
- 18 The following four paragraphs follows Holtermann (2015; forthcoming) closely.
- 19 Cf. Ross (1958: 65–66).
15One attempted solution to this challenge whose alleged failure Ross never tires of exposing is of course natural law.18 In Ross’s interpretation, natural law pursues a rationalist epistemological strategy: it tries to derive the validity of such normative statements of legal rules from a foundation of self-evident truths of reason; more specifically, from one fundamental, intuitively valid idea of justice which is constitutive of law, and to which all human beings, qua rational creatures, have access and will assent.19
16To Ross, however, the problem with such intuitions is that they, in contrast to sense data, are inextricably private. Intuitions can vary from person to person, and patently do so quite often. As Ross puts it in one of his most quoted passages:
Like a harlot, natural law is at the disposal of everyone. The ideology does not exist that cannot be defended by an appeal to the law of nature. And, indeed, how can it be otherwise, since the ultimate basis for every natural right lies in a private direct insight, an evident contemplation, an intuition. Cannot my intuition be just as good as yours? Evidence as a criterion of truth explains the utterly arbitrary character of the metaphysical assertions. It raises them up above any force of inter-subjective control and opens the door wide to unrestricted invention and dogmatics.20
- 21 Cf. Holtermann (2015; forthcoming) and also Ross’s criticism of Kelsen, this issue translation of R (...)
- 22 Cf. e.g. Ross (1958: 19).
17What is more often overlooked, however, is that Ross is equally dismissive (even if less hostile) of the legal positivists attempts to justify statements of legal doctrine.21 The question is, therefore, whether a legal science is ultimately possible at all on the conditions of logical positivism. Ross clearly thinks it is, and the key move in his suggested solution is a fundamental change in the perspective of jurisprudence – from being norm expressive to being norm descriptive.22 It may not be immediately obvious how this simple move should help but it is in fact crucial for the possibility of a doctrinal legal science. How is perhaps best explained through an analogy and a clarifying notion from Frege’s logic. First the analogy:
Ellen and the Christmas pixie-norm: If six year old Ellen around Christmas claims that she ought to leave some rice pudding in the addict for the Christmas pixie, we may safely assume that this is not true. We do not as a matter of fact have any duties toward imaginary creatures like Christmas pixies. If, on the other hand, Ellen’s father were to say that Ellen believes that she ought to leave rice pudding in the addict, the case is completely different. And this is so because, in contrast to the first proposition, the truth value of the latter is wholly independent of the possible existence of duties toward imaginary creatures. It depends, instead, exclusively upon whether or not Ellen actually believes in the existence of such a duty. And this is ultimately a descriptive psychological question regarding her beliefs, not a normative (i.e. norm expressive) question about the existence of duties toward Christmas pixies.
- 23 Cf. Frege (1994 : 149).
18This analogy illustrates an old insight from the philosophy of language attributed to Frege: if any given proposition P is embedded in a so-called propositional attitude report (i.e. an agent A believes that, claims that, feels that, etc. P), the truth value of that particular proposition has no bearing on the truth value of the compound proposition (the propositional attitude report).23 Whether or not things actually are the way A believes/claims/feels they are, is simply irrelevant to the truth of the whole report. In such contexts the crucial question is instead whether or not she actually believes/claims/feels that things are in the way asserted in the proposition.
19And it is precisely this shift in truth values following the introduction of a propositional attitude context which Ross aims to achieve when he insists that legal science should be norm-descriptive and not norm-expressive. Following this paraphrasing the doctrinal study of law is no longer a study of how we as legal subjects ought to behave; about which rights and duties we have. It is, roughly speaking, a study of how judges believe we ought to behave; about which rights and duties they believe we have.
20In his examples of propositions of legal science Ross consistently emphasises this paraphrasing terminologically. Thus he writes ‘that every proposition occurring in the doctrinal study of law contains as an integral part the concept 'valid (Illinois, Californian, common, etc.) law.'’24 This means that the propositions of legal science must be indexicalised to use an expression from the philosophy of language: they must be propositions about the beliefs of a particular group of people regarding a particular field; i.e., about the beliefs of Illinois, Californian, common, etc. judges regarding Illinois, Californian, common, etc. law. This addition may on occasion be tacitly implied but it can never be thought away entirely – lest the propositional attitude context and thus the actual possibility condition for an empirically respectable legal science disappear. In other words, the epistemologically problematic directives (and adhering thoughts) are placed in the mouth (and mind) of the judge.
21One might object that propositions about judges' beliefs about rights and duties (in Ross’s words: ‘the ideology of the sources of law which in fact animates the courts’25) are equally impossible to verify. After all, how do we know what people believe? This is private too. Ross is ready to acknowledge the existence of epistemological challenges also in this area. But in contrast to the difficulties facing natural law and legal positivism, he does not consider these problems to be insurmountable on empiricist conditions – for the following reason:
- 26 Ross (1958: 75; emphasis added).
If, in spite of all, prediction is possible, it must be because the mental processes by which the judge decides to base his decision on one rule rather than another is not a capricious and arbitrary matter, varying from one judge to another, but a process determined by attitudes and concepts, a common normative ideology, present and active in the minds of judges when they act in their capacity as judges. It is true that we cannot observe directly what takes place in the mind of the judge, but it is possible to construct hypotheses concerning it, and their value can be tested simply by observing whether predictions based on them have come true.26
- 27 This is also why Aarnio is wrong when he claims that ‘[t]his very element in the Rossian prediction (...)
22Correspondingly, if Ellen believes in the existence of duties towards Christmas pixies it is highly likely that she will behave accordingly around Christmas: she will remind her daddy to buy rice in the super market, to cook it when they get home, etc. And if a judge believes that the Bills of Exchange Act, section 28, subsection 1 is socially binding she will behave accordingly if a case fulfilling these conditions is brought before her court: she will order the acceptor to pay.27
23It is in view of these considerations that Ross’s theory of law becomes a predictive theory: it becomes a set of predictions of judges’ behaviour under certain specified conditions. Thus, in Ross’s final analysis an assertion A made in the doctrinal study of law that a given directive D is valid (Illinois, California, common etc.) law, becomes:
a prediction to the effect that if an action in which the conditioning facts given in the section are considered to exist is brought before the courts of this state, and if in the meantime there have been no alterations in the circumstances which form the basis of A, the directive to the judge contained in the section will form an integral part of the reasoning underlying the judgment.28
- 29 Ross (1958: 77; my emphasis).
- 30 Cf. Ross (1958: ch. 3).
24And more generally, the entire doctrinal study of law becomes a theory about ‘the aggregate of factors which exercise influence on the judge’s formulation of the rule on which he bases his decision’.29 Ultimately, these factors can be subsumed under the four sources of law: legislation, precedent, custom and the tradition of culture (“reason”).30
25This, in brief, is Ross’s standard model for the propositions of legal science. It represents his general conception of the doctrinal study of law, and illustrates how he elegantly manages to bring jurisprudence in accordance with the epistemological presuppositions of logical positivism.
- 31 In Quine's words: ‘any statement can be held true come what may, if we make drastic enough adjustme (...)
26The problem is, however, that most contemporary philosophers have come to believe that these very epistemological premises are fatally flawed. Many philosophers have contributed to the criticism of logical positivism but I will focus only on Quine's criticism for the reasons mentioned in the introduction. Put very briefly, the starting point of his argument is that it is simply impossible, even in principle, to reduce any part of our alleged knowledge to basic sense data. Neither sense data themselves, nor the propositions of logic which should make it possible to combine sense data into complexes of theory, are indubitable in the way the logical positivists seemed to presuppose.31
27This conclusion is fatal for logical positivism because it seems to imply that, contrary to the movement’s intensions, all known scientific disciplines become “metaphysics” – from hard physics to Ross’s newly reconstructed realistic jurisprudence. To put the same point differently: logical positivism has ended up in many philosopher’s nightmare: scepticism.
28This conclusion leaves but one serious problem: it is deeply counterintuitive. It simply seems too hard to believe that otherwise perfectly reliable science should ultimately be metaphysics in the pejorative sense; i.e. cognitively meaningless nonsense with no greater claim to our attention than palmistry.
- 32 Quine (1969: 79; my emphasis).
29And it is precisely this consideration that Quine takes as the starting point of his naturalized epistemology. Thus, he writes: ‘We may not be able to explain why we arrive at theories which makes successful predictions, but we do arrive at such theories.’32 We do in fact have knowledge. We are in possession of valid science. And in so far as this is true, there must be something wrong with a theory of knowledge which turns out to claim the opposite.
30The pertinent question is what it is exactly that is wrong with the logical positivists’ theory of knowledge. Considering the fact that we actually seem to be in full possession of viable science, Quine’s conclusion is that there must simply be something wrong with the movement’s Cartesian premise that deducibility from sense data is a necessary condition for science. Or, as Quine later puts it: ‘Cartesian doubt is not the way to begin.’33
31But how should we begin then if the Cartesian ideal of unshakeable foundations is mistaken? How else should we do epistemology? Answering this question necessitates a closer diagnosis of the traditional epistemological project. The Cartesian tradition is characterized by not taking anything for granted. It takes as its starting point the completely open questions: Do we have knowledge at all? Is knowledge really possible? On this conception, epistemology becomes an ideal or utopian project right from the outset. Disregarding matters in the real world, it starts out with an ideal definition of knowledge.
32But from these beginnings the possibility of ending up with a negative answer is indeed real – as the examples of Descartes and logical positivism show. From there we can indeed be forced to conclude that knowledge is unattainable. It is therefore precisely this ideal and ultimately utopian feature of Cartesian epistemology which Quine (and a growing number of modern philosophers with him) concludes is mistaken, and leads him to his anti-sceptical premise, i.e. to the earlier mentioned claim that we do in fact have valid science, even if we still haven’t figured out quite why this is so. Once we truly appreciate this fact it irreversibly changes the epistemological agenda. From this point the primary question is no longer whether or not we actually do have knowledge. Anybody can see that we do. Instead, the epistemological question becomes how we have received that knowledge as a matter of fact. In Quine’s words:
- 34 Quine (1969: 75; emphasis added).
But why all this creative reconstruction, all this make-believe? The stimulation of his sensory receptors is all the evidence anybody has had to go on, ultimately, in arriving at his picture of the world. Why not just see how this construction really proceeds?34
33And this question – as to how we, and the sciences generally, have ultimately arrived at our ‘picture of the world’ – is an empirical question. It is a question as to which factors actually influence the shaping of our beliefs about the world. This brings out the most conspicuous break with Cartesian epistemology. For as long as the starting point of epistemology is the open question as to whether or not we do in fact have knowledge, we cannot for logical reasons utilize that same knowledge to answer the question: ‘If the epistemologist’s goal is validation of the grounds of empirical science, he defeats his purpose by using psychology or other empirical science in the validation.’35
34But if we no longer ask whether or not we have knowledge, but how we arrived at the knowledge we claim we have, then we are no longer barred from using already existing knowledge:
However, such scruples of circularity have little point once we have stopped dreaming of deducing science from observations. If we are out simply to understand the link between observation and science, we are well advised to use any available information, including that provided by the very science whose link with observation we are seeking to understand.36
35And it is precisely this change in perspective which has been setting large parts of the epistemological agenda after the fall of logical positivism. The epistemological project has to a large extent become one of giving a detailed empirical description of how we have arrived at our picture of the world, and epistemology as a discipline has, correspondingly, become a descriptive rather than a normative discipline. Hence the label replacement naturalism.
36Before returning to Ross, however, we should sort out a few details pertaining to this empirical epistemological position. For while the general idea of taking the philosopher ‘out of the arm chair and into the field’37 has met with widespread sympathy, there has been considerable disagreement as to where precisely this field might be. Which empirical discipline or disciplines should take the place of armchair philosophy?
37Quine for his part was still so much influenced by the empiricist tradition that he thought it obvious that an account of ‘the construction of our picture of the world’ should be constituted predominantly by a natural scientific description of the affection of our senses. Accordingly, he thought that the key empirical epistemological discipline should be some kind of austere behavioural psychological study of the correlation between the sensory input to which we are exposed, and the ensuing output in the shape of theories of the design of this world.
38The question is, however, whether Quine still has a strong argument in favour of this continued hard scientism. When the goal was reduction to sense data this favouring seemed to make good sense. But once this reduction has been proven unfeasible and the associated ideal concept of knowledge has been abandoned it has all the appearance of old empiricist prejudice. It simply does not seem plausible that only those factors that fall within the prisms of the natural sciences should play a role in the construction of our beliefs.
39Accordingly, other modern epistemologists have rebelled against this particular part of Quine’s program. They have tried to give an empirical account of the construction of our picture of the world without, in so doing, sticking narrowly to a natural scientific recording of the ‘stimulation of our sensory receptors’.
40One early prominent example of this school of thought is Thomas Kuhn who uses history and sociology to give an unprejudiced account of how as a matter of fact physicists have arrived at their theories.38 These soft disciplines acquire this new role precisely because one of Kuhn’s conclusions is that the ‘stimulation of our sensory receptors’ plays a considerably less significant role in the construction of our scientific theories than Quine would have us believe. On the contrary, factors like education, socialization into a particular scientific tradition (a paradigm), etc. that are not captured through the narrow prism of behavioural psychology turn out to play at least as big a role.
- 39 Bloor (1991 ).
- 40 Bloor (1991: 3).
41Kuhn is supported in this conclusion by a long range of other modern epistemologists and philosophers of science who have also spoken in favour of turning epistemology into an empirical discipline. This goes, for instance, for the so-called social constructivism within sociology of knowledge and of science which unambiguously approve of an empirical assumption in the style of Quine. This can be seen, inter alia, from the key work Knowledge and Social Imagery39 from one of the leading theorists in social constructivism David Bloor: ‘All knowledge, whether it be in the empirical sciences or even in mathematics, should be treated, through and through, as material for investigation.’40 However, as amply indicated by the name, settling for a natural scientific registration of ‘stimulation of our sensory receptors’ is found to be grossly insufficient also in this school. A sociological perspective is necessary.
42Sociologists of science and of knowledge are undoubtedly right that Quine’s strictly natural scientific perspective is insufficient if the goal is to provide an exhaustive description of the creation of our science and knowledge. On the other hand, these sociologists have a marked tendency to wear the patch over the other eye. Instead of Quine’s narrow physicalist perspective we get an equally narrow sociological view. And this, no doubt, is equally unsatisfactory. The more reasonable conclusion must surely be that the new empirical epistemology becomes a multi-disciplinary endeavour. Considered as an empirical phenomenon, knowledge and science is multi-faceted. From a neutral perspective it would appear to involve human beings as biological organisms, sense-perceiving creatures, language users, psychological subjects and social creatures. Accordingly, a credible empirical replacement naturalism seems forced to include all those natural and social sciences which focus on these various aspects of the construction of our picture of the world.
43Thus far, focus has been on the methodological side of the modern empirical school and its break with Cartesian epistemology. However, with a view to seeing how it shall ultimately be possible to place Ross’s realism within this naturalistic framework, we should observe a corresponding development – for parallel reasons – on the object side of epistemology. Thus, Bloor writes with regard to the empirical epistemology that it ‘would be impartial with respect to truth and falsity, rationality or irrationality, success or failure. Both sides of these dichotomies will require explanation.’41 This attitude is in contradiction with the traditional Cartesian search only for indubitable truths. But it follows, in fact, directly from the pragmatic, non-ideal starting point of a naturalized epistemology. In the absence of ideal indubitable knowledge, it appears that we have no other option but to begin with what people actually believe they know. From some ideal point of view, these beliefs may very well be true or false, knowledge or ignorance. But this is precisely what we do not know up front. They must therefore be placed on the same footing. It is in this sense that an empirical epistemology can be said to amount to a “theory of beliefs”.
44Finally, it should be added that the suggested empirical study can be carried out on different levels of abstraction according to the particular set of beliefs which we are trying to explain at any given moment. On the one hand, we can investigate completely general, universal factors influencing the creation of the beliefs of Homo sapiens as such. Such an account would undoubtedly include a general explanation as to how the senses bring information about the surrounding physical world, how language is learned, how human beings are socialized into particular world views, etc.
45On the other hand, however, we can also carry out more specialized studies where the goal is the particular beliefs of particular groups about particular subjects. One of the things which characterize our ‘construction of our picture of the world’ is that it comes into being, particularly in the sciences, through a sharp division of intellectual labour. We should therefore expect considerable varietion with regard to the factors influencing e.g. historians’ beliefs that the Holocaust has taken place in comparison with physicists- belief in Einstein’s theory of relativity. In order to deliver satisfactory explanations of such a broad variety of beliefs we need a flexible empirical epistemology that will be able to change its lenses according to which specific construction of which specific aspect of a picture of the world we are trying to explain.
46Recapitulating the conclusions made so far, we thus get the following image of replacement naturalism as an epistemological program: i) it rejects the possibility of providing an a priori foundation for human knowledge and for the sciences; ii) it rejects the need to provide such a foundation; iii) instead, it considers knowledge to be an empirical phenomenon; iv) epistemology therefore becomes an empirical discipline engaged in providing an explanation as to how we have actually acquired the beliefs we have; and v) it becomes a multi-disciplinary activity that invokes all those empirical disciplines that may prove relevant in order to provide the desired empirical explanations of the beliefs under scrutiny.
47From this picture gradually emerges an idea of how Ross fits into the whole story. The crucial point is that the basic move made in his realist theory of law amounts precisely to an empirical turn analogous to the one made in modern naturalized epistemology: from considering, in accordance with natural law theories, a particular kind of beliefs (viz. judges’ beliefs regarding, for instance, the legal obligations of acceptors in accordance with the Bills of Exchange Act, section 28, subsection 1) as being in need of a priori justification, Ross instead considers these same beliefs to be purely empirical phenomena to be studied and explained as such.
48In this sense, Ross’s realist jurisprudence in fact becomes a highly specialised subsection of modern naturalized epistemology. It is not engaged in unfolding how human beings generally tend to form beliefs as such. But neither was, as we saw, Kuhn’s history and sociology of science. He wrote only about how physicists arrive at various theories in physics. According to Ross, legal science becomes a sociology of judges – a sociology which is engaged in finding out how precisely Illinois, Californian, common, etc. judges arrive at their particular beliefs regarding Illinois, Californian, common, etc. law. It is engaged, in other words, in exploring that niche within comprehensive empirical epistemology that can be summarised in the question: ‘Why not see how this construction [of judges’ beliefs regarding Illinois, Californian, common, etc. law] really proceeds?’42
49And, as earlier mentioned, the cornerstone of the answer to this question is that the four sources of law (legislation, precedent, custom and the tradition of culture/“reason”) jointly constitute ‘the aggregate of factors which exercise influence on the judge’s formulation of the rule on which he bases his decision’.43
- 44 Ross (1958: 261).
- 45 Quine (1980 : 43).
50Furthermore, we should notice that Ross’s motivation for this change of perspective is also analogues to that of naturalized epistemology: it has simply proven impossible to provide indubitable foundation for norm-expressive beliefs. Natural law is doomed to fail because, as Ross writes, ‘[t]he ideology does not exist that cannot be defended by an appeal to the law of nature.’44 A sentence that comes very close to Quine’s motive for parting ways with logical positivism: ‘[a]ny statement can be held true come what may’.45
51But there is an even further argument to be made that Ross’s jurisprudence can be fitted into the empirical turn. One of the most conspicuous features of naturalized in comparison with traditional epistemology is that it no longer expresses itself in the first person. Descartes’ Metaphysical Meditations is written in the first person as a personal reflection on his own doubts and despair.46 And though the Cartesian project takes a collective turn with logical positivism it continues to be pitched in the first person. It is our theories, our science that are to be derived from our sense data.
52Naturalized epistemology, on the other hand, strikes a wholly different tune. Here, characteristically, any identification between the epistemologist and the person whose belief acquisition is under scrutiny is abandoned completely. Instead, it “anthropologizes” its object by consistently portraying it in the third person – as illustrated in Quine’s words:
It [naturalized epistemology] studies a natural phenomen, viz., a physical human subject. This human subject is accorded a certain experimentally controlled input /…/ and in the fullness of time the subject delivers as output a description of the three-dimensional world and its history.47
53Basically the same anthropologizing tendency is conspicuously present when Kuhn and Bloor respectively speaks about how physicists and mathematicians arrive at their theories.
54This development is strikingly parallel to the one we have seen Ross propose in jurisprudence. In “the old days” the goal of the philosopher of law was to provide her “own” directive propositions of valid law, of legal rights and duties, with adequate epistemic justification. Epistemologically speaking, the philosopher of law and the legal scientist in general was in the same boat as the judge. But once we take the step to legal realism and introduce the propositional attitude context in order to distance the legal scientist from the problematic norm-expressive propositions, this identification is abandoned. An epistemic wedge is driven right between these actors in the legal field, and hereinafter legal science becomes, as we have seen, a theory about how judges (referred to in the third person) believe we should behave – what rights and duties they think that we have.
55Summarizing the points made so far, we thus have the following reasons to assume that Ross’s realistic theory of law can be fitted into the program of naturalized epistemology. First of all, Ross shares the fundamental empirical premise underlying the project: instead of trying to validate specific beliefs in an a priori manner they are consistently regarded as empirical phenomena and, as such, they are provided with an empirical explanation. Secondly, Ross’s motivation for this move is his scepticism regarding the general possibility of providing a satisfactory foundation for these beliefs. And finally, in virtue of this change of perspective Ross starts to express himself consistently in the third person.
56A few loose ends remain, however, before it all falls neatly into place. The most immediate challenge is perhaps that Ross clearly beliefs in the Cartesian assumption, i.e. in the need for indubitable foundations. The only problem is that the normative (i.e. norm expressive) propositions cannot meet this strong criterion. But it is precisely this need that is rejected by naturalized epistemology.
- 48 Any impression to the opposite may be due to the so-called selection effect, cf. e.g. Schauer (2009 (...)
57On closer consideration, however, it is hard to see how it could possibly become a problem for the predictive theory to abandon the Cartesian assumption in favour of a lower threshold of doubt. The study which Ross envisions is clearly feasible if the relevant criterion is now merely ability to produce predominantly successful prediction.48
58Another worry after coupling Ross and naturalized epistemology is related more to questions of overall structure in the sciences. On a verbatim account, the argument made so far seems to imply that legal science as envisioned by Ross becomes a kind of epistemology. But this should only be superficially surprising.
59As we have seen, the empirical turn involves a radical change in the way the relationship between epistemology and science is conceived. In the Cartesian tradition the core metaphor is “foundational”. Empirical science is envisioned as being placed on an a priori epistemological foundation. But after the empirical turn this picture is abandoned. Instead, we get a conception of empirical science which, as a whole, simply looks after itself. We have a long range of various disciplines that are busy studying various aspects of empirical reality. And among those, some (so-called epistemologists) have taken it upon themselves to study the creation of all sorts of beliefs within that particular species which is called Homo sapiens. And within this part of science a small subsection (so-called legal scientists or jurists) have specialised in the detailed study of how one group of people called (Illinois, Californian, common, etc.) judges arrive at their particular beliefs regarding (Illinois, Californian, common, etc.) law.
60As an illustration of this general conception of science, in which Ross thus finds his place, Quine prefers to use the metaphor of Neurath’s ship: ‘Wie Schiffer sind wir, die ihr Schiff auf offener See umbauen müssen, ohne es jemals in einem Dock zerlegen und aus besten Bestandteilen neu errichten zu können.’49
61I have so far indicated how a modified version of Ross’s legal realism can be fitted into a Quinean version of replacement naturalism. However, the influence is not only unilateral. As it turns out Ross’s extremely lucid early analysis of the predicament of legal science can also help clarify current philosophical discussions of naturalized epistemology.
62One of the things that have been heavily debated in relation to naturalized epistemology is its relationship with the anti-Sceptical assumption which critics have claimed is more ambivalent and problematic than generally indicated by Quine and other proponents of the movement. For instance, Hilary Putnam has accused the transition from a prescriptive a priori to a descriptive empirical epistemology to amount, ultimately, to an ‘attempted mental suicide’.50 As we shall see, Ross’s model for the discussion of the problems related to legal science turns out to provide an ideal model for a reformulation of this objection.
- 51 Ross (1953: 15 / 2013: 52; Bindreiter's translation).
63As we have seen, Ross uses an apparently typical normative (i.e. norm expressive) proposition of legal science (regarding the legal obligations of acceptors of bills of exchange) as a focal point for his discussion of the epistemological status of legal science. But in order to illustrate the problems allegedly particular to legal science, Ross contrasts, in the Danish version of On Law and Justice, this proposition with an equally typical proposition of natural science, viz., the so-called Boyle-Mariotte’s law, which Ross formulates thus: ‘[p]ressure and volume of a given mass of confined gas are inversely proportional.’51
64Ross’s point in introducing this proposition is that it, in contrast to the directive of legal science, appears to have epistemic value as it stands. There is no need to reformulate this proposition as a propositional attitude report in order for it to be valid science. We can, however, use Ross’s contrast of the two propositions and his consequent rewriting of the legal proposition to make an illustrative re-description of the opposition between the Cartesian tradition and modern naturalized epistemology. In the Cartesian tradition the epistemologist is ultimately interested in being able, in good philosophical conscience so to speak, to personally say or to concur with the proposition that ‘pressure and volume of a given mass of confined gas /…/ etc.’ It is this interest that is manifested in the fact that Descartes and the logical positivists consistently expressed themselves in the first person. And in this, the situation of the Cartesian epistemologist is in fact wholly parallel to the one in which the natural lawyer finds herself. She is equally interested in being able, in good philosophical conscience, to say or to concur with the proposition that, e.g., ‘the acceptor is bound to /…/ etc.’
65But this ideal possibility of being able oneself to express or to concur with particular scientific propositions is no longer the primary goal of naturalized epistemology. Here, the issue is no longer whether particular facts or states of affairs asserted in the sciences are real or not. What is at stake is the fact that a certain person or social group (in this case, physicists) believes that ‘pressure and volume of a given mass of confined gas /…/ etc.’ And it is this empirical phenomenon that we are trying to provide with a satisfying empirical explanation.
66But following the Rossian recipe, this implies that we have introduced a propositional attitude context for the proposition which was originally of prime interest to the Cartesian epistemologist. Instead of the proposition: ‘pressure and volume of a given mass of confined gas /…/ etc.’, we now have the proposition: ‘[p]hysicists believe that pressure and volume of a given mass of confined gas /…/ etc.’. And, as with Ross’s propositional attitude reports in legal science, this logically implies that the truth value of the embedded proposition (‘pressure and volume of a given mass of confined gas /…/ etc.’) is irrelevant for the truth value of the complex proposition. In other words: the truth value of the scientific proposition becomes irrelevant to the truth value of the epistemological proposition.
67Paradoxically, there is therefore an even greater logical distance between the propositions of naturalized epistemology and the propositions of the science which it is studying, than there was, originally, between the propositions of the Cartesian tradition and that same science – a fact which is strikingly at odds with the predominant metaphors applied for the two epistemological traditions in the shape of Neurath’s ship and armchair philosophy respectively.
68And this observation paves the way for a critical examination of the motivation underlying naturalized epistemology. As we have seen, naturalized epistemology appears to take as its starting point a categorical break with the strong concept of knowledge applied in the Cartesian tradition. This concept of knowledge turned out to be so strong that it led to the sceptical conclusion: knowledge was impossible. And it is this apparent insight which legitimises the very idea of a naturalized epistemology.
69As we have seen, however, Ross made the exact same move on behalf of legal science. And we know his motivation for making this turn. It did not include any abandonment of the Cartesian concept of knowledge. On the contrary, his motivation included an explicit confirmation of this concept. He introduced the propositional attitude context into legal science precisely in order for the discipline to avoid the problems that confront norm-expressive propositions vis-à-vis (the logical positivist version of) a Cartesian concept of knowledge.
70The observation that Quine and other replacement naturalists perform the exact same rephrasing as Ross does, feeds the suspicion that they may in fact be sharing his motives. This suspicion is reinforced by the fact that the naturalists make their move on similar grounds: both Ross and Quine (and others) introduce the propositional attitude context only after they have found out that they cannot provide indubitable foundations for them. This makes it plausible that the naturalized epistemologist may in fact have felt some embarrassment over expressing the scientific propositions herself.
71This suspicion of a motivational affinity is reinforced further when we take a look at the passages in which Quine reveals what exactly he expects to get out of the empirical study. Thus, in an exemplary passage he writes:
- 52 Quine(1969: 83; my emphasis). Cf. also e.g. Quine (1974: 3).
The relation between the meagre input and the torrential output is a relation we are prompted to study for somewhat the same reasons that always prompted epistemology; namely, in order to see how evidence relates to theory, and in what ways one’s theory of nature transcends any available evidence.52
72Formulations like these make it somewhat hard to convince oneself, that Quine’s enthusiasm as to the empirical study of knowledge is entirely unconnected with an expectation that it will provide us with a new argument for the sceptical conclusion. In the same way that “Two Dogmas” seems to have shown that knowledge is philosophically impossible, he appears to expect that handing over the epistemological burden to empirical science will show that knowledge remains impossible, also when described in empirical terms. In other words, Quine seems confident that the evidence/theory account will show a considerable deficit on the side of evidence also when made out in hard empirical currency.
73But this is where the comparison with Ross seems to highlight a principled difficulty of the possibility for Quine and for naturalized epistemology generally to arrive at this conclusion in a satisfactory manner. For Ross, the price to be paid in order to secure the possibility of legal science is, as we have seen, the rationality of the judge. In virtue of the introduction of the propositional attitude context any attempt to provide an epistemological foundation for the norm-expressive propositions of judges is abandoned. Copying Putnam’s flair for drama, one might say that Ross does not only attempt but in fact carries out a kind of mental “genocide” on the tribe of judges.
74Ross can do so because he is convinced that the thrust of his attack strikes solely at an ultimately dispensable province of human knowledge, viz. norm expressive beliefs. The mainland of knowledge, i.e. the territory covered by ordinary descriptive empirical science, would remain unharmed by the attack simply because it consists, semantically speaking, of assertions instead of directives, and the problems brought out concern only directives.
75But this serves to make clear what critics believe is the principled problem related to the ultimately analogues move made by replacement naturalism. For contrary to Ross, Quine and those of like mind cannot retreat to a mainland of knowledge from the scientific field which they happen to be investigating. In spite of the introduction of an intentional context this move does not imply a shift from one type of proposition to another (i.e. from directive to assertions). For this reason any dissociation from the enclosed scientific proposition (i.e. ‘pressure and volume of a given mass of confined gas /…/ etc.’) remains preliminary. Any problems that Quine and others like him expect to find in the science subjected to empirical investigation will apply equally to their own propositions. And it is in this regard Ross can be used to give new sense to the claim that a consistent naturalising of epistemology ultimately amounts to ‘attempted mental suicide’.
76Only time will tell whether this challenge ultimately constitutes a serious problem for Quinean replacement naturalism. But until then, I hope to have established a convincing argument that it is in fact possible to dismount Ross’s version of Scandinavian legal realism from its foundations in logical positivism and reinsert it into the naturalistic framework and thus for Ross and Scandinavian legal realism to continue to be in accordance with ‘a significant group of modern philosophers and philosophically interested practitioners of other disciplines.’
—Acknowledgment.— This research is funded by the Danish National Research Foundation Grant no. DNRF105 and conducted under the auspices of the Danish National Research Foundation’s Centre of Excellence for International Courts (iCourts). I am grateful to Jeppe von Platz and Martin Vinding and two anonymous reviewers for valuable comments on an earlier draft of this article.