By David Harvey
Dworkin (1977) argues that Hart’s theory of law is insufficient in that it doesn’t explain all aspects of law. In his criticism of Hart’s account, Dworkin stipulates that Hart fails to incorporate principles into his description of what law is. However, since Dworkin’s criticisms emerged, the degree to which Hart’s theory, in fact, fails to acknowledge certain legal principles as law is unclear. In this essay I will interpret Hart’s theory in a way that stretches it to encompass principles as much as it can. For it appears that Hart can to some degree accept that principles become a part of law following their involvement in legal precedents. However, it will become apparent that Dworkin’s thesis differs from this slightly in that he believes these principles not to be grounded in their institutional use, but to be formed via a kind of accepted appropriateness over time. In this, then, Dworkin wants to argue that principles can and do legally bind judges before precedent has been set. In my argument I submit that Dworkin fails to conclusively prove his theory’s superiority. I will further claim that not only is the accuracy of the Dworkinian account descriptively dubious, but it possesses characteristics which do not make it normatively preferable to Hart’s.
To begin, I set out Hart’s theory. Hart argues that law is comprised of a combination of primary and secondary rules (1961: 79). Primary rules consist of statutes or conventions which explain both the penalties we may face for certain actions and the rights we are entitled to, either absolutely, or under certain circumstances. These rules of conduct help to guide our behaviour as we know to act in certain ways and not others to avoid undesirable repercussions. Secondary rules, on the other hand, detail the method by which laws come to be valid. The most famous of these are called the rules of recognition. Rules of recognition explain the characteristic possessed by primary rules which qualifies them as law. To take a trivial example, in the United Kingdom, bills become validated into law by passing through the elected legislature in the House of Commons and the House of Lords, before being signed by the monarch. However, this is not the only characteristic a valid rule could possess. It is important to note that Hart suggests three kinds of pedigrees for a valid rule: “their having been enacted by a specific body, or their long customary practice, or their relation to judicial decisions” (1961: 95).
Hart thereby argues that law consists of the application of valid rules. Take, for instance, the case of Daniels and Daniels v. R. White & Sons and Tarbard (1938). The plaintiffs, Mr. and Mrs. Daniels purchased a drink labelled as lemonade from Mrs. Tarbard and, later, they became ill. It was discovered that their illness was caused by their consumption of Mrs. Tarbard’s lemonade which had been contaminated with carbolic acid. Mr. and Mrs. Daniels went on to sue Tarbard for damages in compensation for their medical treatment. In court, Judge Lewis J. argued that as the case fell under the laws concerning sale by description, Mrs. Tarbard was guilty of the plaintiffs’ claim. Though Mrs. Tarbard may not have known of the contamination, unfortunately the Sales of Goods Act, 1893, required that goods sold be of merchantable quality. The judge thus followed the rule specified and hence applied the law.
However, some cases do not always seem to fall so clearly under the law. In light of this, Hart establishes the problem of the penumbra. Penumbral cases, he argues, are those which require judges to use their discretion to adjudicate appropriately in circumstances where the way in which a law should be applied to a particular case is unclear. Hart therefore holds the view that the law contains gaps. Furthermore, the fact that there are gaps in the law entails that the filling of these gaps via the exercise of judicial discretion does not have any legal nature. That is to say that, in these cases, judicial discretion involves applying values which are outside of the law to a case in order to reach a judgment. These values may invoke a moral or social duty but their application is not legally right or wrong.
This is a conclusion which Dworkin rejects. And it is on this particular distinction between Hart and Dworkin that this essay focuses. Dworkin argues contrastingly that the law is, in fact, largely complete; it has no, or very few, gaps. This is because the law is not merely comprised of laws which are either applicable or not; it is also made up of principles. Dworkin claims that principles are not accounted for under Hart’s theory because they differ from rules in two significant ways.
Firstly, unlike rules, principles do not apply to cases in an on or off fashion, but serve to detail a judge’s decision surrounding a case as she takes into account the respectable weight and significance of any relevant principles. To exemplify this Dworkin discusses a case whereby the judge cites an apparent legal principle as the reason for his decision. In Riggs v Palmer 1889, the potential heir to his grandfather’s will had in fact murdered the testator in order to obtain his inheritance. Having already been convicted of murder, the case of Riggs v Palmer contested the notion of allowing the murderer his inheritance. The court ruled in favour of the plaintiffs, arguing that to allow the grandson to profit from his crime by obtaining his inheritance was to violate the universal maxim that no person should be able to profit from her crime. This principle clearly held substantial weight in the case and significantly affected the judgment.
Secondly, Dworkin argues that principles do not enter the legal system via a kind of validating process. Instead, principles become a part of law over time through a collective sense of appropriateness; they reflect, then, our conception of what law is and should be. The latter of these features of principles leads Dworkin to conclude that Hart’s rule of recognition cannot apply to principles as there is no identifiable historical pedigree to a principle which shows its legal validity. Indeed, Dworkin does not think principles bear any kind of relation to validity, though it is not clear what Dworkin offers as an alternative for distinguishing between appropriately legal principles and other principles with a pure moral sense.
For Hart, these principles must, at least initially, have no legal sense in that they do not legally bind the decisions of judges. The Riggs v Palmer case is particularly interesting as it appeared that if merely the rules of law were followed, the decision should have gone the other way. Statutes of the time explicitly endorsed the execution of a testator’s will following their death regardless of circumstance. Additionally, not all judges were in agreement with the majority’s judgment. Dissenter, Judge Gray, argued that the court’s decision did not follow the written statutes of the law and that the court ought not to invent statutes simply to achieve a morally desirable result (Riggs v Palmer 1889). This depiction of the majority’s decision seems more in tune with the Hartian conception of law whereby judges use their discretionary power to introduce moral principles into their judgments. From Judge Gray’s point of view, there was indeed nothing of legal substance to the court’s decision.
Of course, since Riggs v Palmer, precedent became set for future similar cases. Here, Hart can account for this principle as having a kind of legal sense. For future lawyers can argue on the basis of precedent that contemporary cases be ruled in the same way. Moreover, since precedent was set, Hart can identify the origins of the principle, allowing his rule of recognition to come into play. If we return to his criteria for establishing a rule’s validity, it is clear that he identifies precedents and conventions as validating characteristics for rules as well as “[rules’] relations to judicial decisions” (Hart 1961: 95). Hence, Riggs v Palmer introduced the following idea into law: that a person ought not to profit from their crime. And in this, the principle came to relate to a judicial decision. This leads me to understand that Hart’s rule of recognition does have the capacity to identify legal principles which have been institutionalised. This is highly significant. If principles are merely a component within precedents, which Hart’s secondary rules recognise as valid law, it seems that judges are not legally required to consider the principles themselves when making legal decisions; she can simply appeal to precedents.
However, this brings me to the crux of the debate. For Dworkin’s claim is not properly satisfied by identifying and appealing to precedents for the legal incorporation of principles. Hart’s position commits him to the view that, prior to Riggs v Palmer, the principle of profiting from one’s own wrongdoing had no legal sense at all. Dworkin, on the other hand, wants to say that it did. His claim is that in Riggs v Palmer, the judges were not using their discretion, in the Hartian sense. Rather, the judges were legally bound to make the decision they did; they were legally required to follow the already legal principle that one should not profit from one’s own crime. (It is worth flagging that Dworkin does not think that principles are legally binding in the same strong way that rules are, but that they are at least required to be factored into a judge’s decisions. I shall from here on simply refer to principles as binding under the Dworkinian view, but it should be understood that this is in the weaker sense.)
It is not very obvious that the judges were legally bound to consider the principle in question. As we have already seen, not all the judges on the case were sympathetic to this Dworkinian idea. Furthermore, let us return to the case of Daniels and Daniels v R. White & Sons and Tarbard. In this case, the judge ruled that Mrs. Tabard was liable for the injuries sustained by Mr. and Mrs. Daniels, though the judge admitted that unfortunately Mrs. Tarbard was actually entirely innocent in the matter. Surely if there is one legal principle that the legal profession and public would collectively endorse, it is that one should not be held accountable for an offence in which one is an innocent party. Judge Lewis J. was certainly not moved by this principle, and admitted rather that in following the requirements of the law was sadly to rule against his own moral reasoning.
Suppose then, that there is some legal principle similar to the one I propose. What are we to make of Lewis’ judgment? Should he have ruled differently? Or, more strongly, as Dworkin would propose, was he in fact legally bound to rule differently? The resulting conclusion seems to be that if such a principle existed and legally bound the judge, the decision reached was wrong and unlawful. This at least seems doubtful. Indeed, the judge cited legal constraints as the very reason he was unable to rule in the way our supposedly legal principle suggests. Of course, Dworkin may protest that Judge Lewis simply hadn’t grasped the right conception of law.
It would require a difficult empirical enquiry to verify whether such principles are in fact legally binding due to their unwritten nature and lack of historical enactment in these cases. Dworkin argues that principles become part of the legal system over time through a sense of appropriateness by both lawyers and laymen. But, as a result, we are left uncertain about when a principle truly becomes legally binding. If the relevant legal principle did in fact exist at the time of Riggs v Palmer, certainly Judge Gray was doubtful of such. Indeed, what degree of appropriateness is needed? And who is to decide and when? These issues become prevalent as we begin to evaluate the desirability of a Dworkinian conception of law.
For having discussed the difficulties in verifying whether in fact principles have a sense of legality apart from their institutionalisation, perhaps it is worth considering a different approach to the topic. As Bayles (1991: 380) notes towards the end of his paper, it could be argued that Hart and Dworkin tackle the issue of what is law with different objectives in mind. Bayles claims that Hart’s theory of law is more descriptive whilst Dworkin’s account leans more towards an ideal of law. I am unsure as to whether Dworkin would accept that his theory is intended to be normative rather than descriptive. Yet having briefly discussed a clash between Hart and Dworkin from a descriptive perspective, it seems reasonable to see whether Dworkin’s account provides a desirable conception of law. Perhaps Dworkin can show at least that legal decisions should require significant input from a judge’s own understanding of principles.
I cannot hope to contrast every component of Dworkin’s theory with Hart’s equivalent. But allow me to consider in relation to the aspect we have discussed which conception of law we might prefer. In the case of Riggs v Palmer, might we not prefer that the judges were legally bound to deny the murderer his inheritance? That is to say that their majority decision was lawfully the right one. Indeed one supposed advantage of Dworkin’s theory is that it suggests that there is always a legally correct answer to almost every case, as the law is complete. This is Dworkin’s right answer thesis. Comparatively, for Hart, the law is incomplete, and so when a judge exercises her discretion to fill in the law, her judgment is not legally right or wrong.
It is argued that Dworkin’s complete law brings two advantages. For one, lawyers are able to submit their case with conviction that there actually is a right answer. Indeed, why else argue that one’s case be accepted by the judge in a court of law? Secondly, the fact that the law can give a correct answer to any case suggests that the law becomes more predictable, as the outcome of a case does not so heavily depend on the morally and personally fuelled discretionary powers of the judge. I will argue that neither of these supposed advantages actually occurs in practice under Dworkin’s conception of law.
We have already briefly mentioned some outstanding questions which result from Dworkin’s account in establishing the legality of principles. Let us now consider these in detail. Under Hart’s theory of law, the rules of recognition allow lawyers and laymen to know what constitutes law. But Dworkin fails to give us a way by which we might know all that law consists of. Indeed, it is central to his characterisation of principles that they do not possess a recognisable quality which can determine whether or not they are law. This is the very reason why Dworkin dubs Hart’s rule of recognition insufficient in determining all aspects of the law. But let us analyse the impact this has on the practice of law.
Under a Dworkinian conception of law, judges, in cases like Riggs v Palmer, would be supposedly legally bound by certain legal principles. Yet each judge has no way of knowing whether a particular principle has become legally binding or not. For, recall that in this case, the relevant principle cited in the judges’ decision had never before been written in any legal institutional form. The judges, then, were left only to speculate whether the principle had become appropriately legally binding, as Dworkin states that principles should become a part of law through a collective sense of appropriateness. Yet how can we expect judges to know whether or not unprecedented principles have become legally appropriate? The reality is that the judges could not have known whether their understanding was the right one. Thus Dworkin’s right answer thesis experiences difficulties in practice; even if there is a right legal answer to every case, he provides judges with no method of identifying what the answer is and whether it is reached. Remember that once a principle has been written into law through its relation to a judicial decision or other kinds of precedents, the Hartian depiction of law equally requires that judges take these into account in making judgments. Before the enactment of a principle, however, there is no written reference to the principle; not even a convention towards it. If there were, the Hartian theory would be able to capture its legal sense just as easily. The only cases which truly show the difference between Dworkin and Hart are those where nonconventional and unprecedented principles are used in law for the very first time.
A further problem arises from the Dworkinian understanding of principles. Not only are judges unable to know whether their understanding of legal principles is correct, but this same uncertainty renders advocates and laymen equally cautious of the law. It would be difficult if lawyers, who have explored every inch of written law, were to argue a case only for it to be rejected on grounds of principles which bear no precedent. Can we really argue that the judgment in Riggs v Palmer was predictable? I would argue not. It would have only taken a majority of the judges to share Gray’s concerns for the case to have concluded entirely differently. Similarly, we might argue that it should have been fairly obvious that in the case of Daniels & Daniels v Tarbard, Mrs. Tarbard should not have been held liable for her innocent actions, yet the judge ruled that she legally was. It seems, then, that the application of principles in law, particularly when precedent is lacking, remains highly unpredictable.
I raise this issue of predictability as it is largely agreed to be a virtue of law as a social decision-making system. Naturally, the predictability of law reflects the certainty of legal outcomes. Under the Hartian system of law, there is much uncertainty. Hart argues that where there are gaps in the law, or in atypical cases, those gaps must be filled at a judge’s discretion. However, though one would suppose that an understanding of law as complete would provide more certainty about the law, in practice it seems to accomplish the very opposite. For whether or not principles will be appreciated by the judge is unforeseeable. Furthermore, which principles each judge might take to hold more weight in their decision is equally unknowable. Indeed, Dworkin even argues that Judge Gray made use of an alternative legal principle in his dissent in Riggs v Palmer: that of not punishing an individual beyond the law’s statutes. Once more, if only more of the judges were of Gray’s opinion, the judgment would have been ruled entirely differently. To relate this to a Hartian conception, with an unknowable amount of potentially relevant legal principles, this inclusion of principles seems to create very few boundaries to a judge’s discretionary powers. Indeed, in cases such as Riggs v Palmer, the judgment seemed to actually contradict the statutory legal requirements. Now, Hartians might be in favour of such a ruling, and consider it as one whereby we would want to encourage adherence to moral principles in order to avoid an absurd result, but stricter legal positivists may argue this inappropriate. Either way, the Dworkinian analysis of law results in either less or the same amount of unpredictability.
And Dworkin does seem concerned with the predictability of law. Part of his theory concerns the historical chain of legal cases whereby he argues, as an aspect of his right answer thesis, that each legal decision must follow and reflect previous legal decisions. This generates both consistency and subsequent predictability. However, Hart can similarly advocate this same legal chain through his recognition of precedent as a test for legal validity. Moreover, this aspect of Dworkin’s theory cannot provide any insight in cases which Hart’s conception does not. For the very cases which include supposed legal concepts and fall outside of the scope of Hart’s rules of recognition are those where precedent does not exist. Therefore, the way in which a judge could continue the legal chain cannot be obvious or entirely foreseeable.
However, might we be happy to substitute uncertainty and predictability for more desirable outcomes in legal cases? Perhaps in Riggs v Palmer we would want to say that the judges did have a legal reason to prevent the murderer his inheritance. But, as Judge Gray argued, this could quite easily merely amount to bending the law to suit our moral intuitions. And, Hart’s depiction of law may well allow for such judgments; only they are considered to amount to purely a judge’s moral-natured discretion rather than law. Hence this isn’t a strong enough argument to suggest that Dworkin’s theory fares better than Hart’s in such cases.
To conclude, I admit that in this essay I have given Hart’s account a charitable hearing; I have somewhat distorted Hart’s theory so as to allow for it to capture Dworkinian concerns. Indeed, though it is fair to argue that Hart wants to include legal precedents in his conception of law, these certainly do not form his core thesis of what law is. I have understood Hart in his broadest sense, then, in order to identify the most fundamental difference between the scope of his theory and Dworkin’s. Therefore, whilst I conclusively deny that Dworkin’s theory is superior to Hart’s, it is untrue that Hart’s theory remains fundamentally untouched. The depiction of Hart’s theory that I have given encompasses many of Dworkin’s ideas into account such that the two depictions achieve only one significant difference.
This difference lies in whether or not a principle has a legal sense before it is recognisably institutionalised within the law. We have discussed various reasons why it seems more likely that principles do not have a legal sense before their enactment. Firstly, there does not seem to be a consensus within the legal community as to whether in cases such as Riggs v Palmer or even Daniels & Daniels v R. White & Sons and Tarbard there was a legal principle to be considered at all. Secondly, even when judges cite supposed principles as their reasons for particular judgments, it is not clear that that principle must have a legal sense. The same principles may simply have reflected the moral intuitions of the judges, as Judge Gray indicated. Unfortunately, determining the sense in which these principles are used makes for a challenging empirical investigation.
Having reached this difficulty in concluding which theory is most factually accurate, we then moved our discussion towards an evaluation of whether a system of law would be more favourable if it were working under the Dworkinian conception. In this, I have argued that despite alluring theoretical characteristics of Dworkin’s theory, such as his right answer thesis, we see no improvements in the predictability of law. Due to our inability to recognise precisely when a principle becomes legally legitimate, judges, lawyers and laymen are left unable to know what the law requires. It is therefore all very well to argue that legal decisions should require significant input from a judge’s own understanding of moral principles, but if she cannot know whether the ones she can think of are truly legal principles, her decision seems as equally arbitrary as if she were exercising discretionary powers under a Hartian conception of law.
If you want to hear more from David Harvey, you can find his Twitter @dpharvey95
Table of Cases
Daniels & Daniels v R. White & Sons and Tarbard ( 4 All E. R. 258)
Riggs v Palmer ( 115 N.Y. 506)
Bayles, M. (1991). “Hart vs. Dworkin” in Law and Philosophy, vol. 10, no. 4, pp. 349-381.
Dworkin, R. (1977). “The Model of Rules I” in Taking Rights Seriously. Cambridge, MA: Harvard University Press.
Hart, H. L. A. (1961). The Concept of Law. Oxford: The Clarendon Press.
MacCormick, N. (1978). Legal Reasoning and Legal Theory. Oxford: The Clarendon Press.
Raz, J. (2008). “The Rule of Law and Its Virtue” in Kavanagh, A. and Oberdiek, J.(eds), Arguing About Law. London: Routledge.