The GULS Law Review

Getting you through the GU law degree!

header photo

A Dworkinian Approach to Law and Mercy


This article comes from Daniel Peixoto Murata, a visiting student to the Glasgow School of Law and a student at the University of São Paulo. Daniel would like to thank Professor Emilios Christodoulidis for his comments and reading suggestions on the subject of this esssay. 



 We live in unfair societies, yet law is said to be general and is applied – in most cases – equally to everyone. This, of course, is problematic and thus the temptation to flex law trough mercy emerges. In this essay, I will analyse the possibility of law to accommodate mercy, however I will argue for the incompatibility of those ideas. This is because once we realise that law is an argumentative practice concerning rights[1], mercy cannot figure on it, for it would mean mercy’s own denaturation or law’s. Law – I take this as one premise for my argument[2] – is a practice concerned about disagreement regarding rights. Legal disputes are about which rights people might have. Mercy, even if rational, is by definition impossible to be translated into rights language.


 I will start with some considerations about mercy, and on this I will rely on Scott Veitch’s insights. He argues that mercy operates by the criteria given by a ‘construction of a commonality[3]. This suggests that mercy emerges when an individual sees in the other (the subject of mercy) an equal, a “brother”. Veitch also points out that context regarding mercy is relevant in order for it to have criteria. To quote Veitch on the idea of mercy’s criteria: “the justifiable use of mercy depends less on the particular decision, but on the criteria that are used. There are different perspectives from which to see this. () That is, there may be reasons why mercy should be granted, and the fact that we (from a descriptive perspective) cannot predict or table them in a formulaic sense does not mean that they exist as non-criteria[4]. What follows from this is that when we look at mercy through the lens of the one seeking it, we can entertain if the pledge is founded on an emotion or an argument, for example. Similarly, looking through the lens of the one that grants mercy brings about other things that might figure as criteria: empathy or a moral principle can do the job here. Mercy, thus, presents itself in a variety of conditions in which it can be appropriate[5].


The question about the criteria for mercy is relevant for us, as it is clear in the following examples: is appropriate for a male judge to grant mercy to a rapist, instead of the female victim? Can we say that is appropriate to leave the mercy question in the hands of the private capital owner, regarding the theft of a minor material good in his industry by the hands of a poor person that sold it to get money in order to buy food? Such examples show how the possibility of law’s mercy is even more problematic than it seems at first sight, because in principle we can’t even point who is in position to grant it.


 Rights, by the other hand, are political claims that work as trumps towards other political considerations[6]. To put it in other words, rights have priority over consequentialist or utilitarian measures. Legal rights are a special kind of rights, for they correspond to political rights that were dignified with legality, thus becoming enforceable through courts or other kinds of institution when needed[7]. The content of most relevant rights, we should note, is not given easily, for it is not only possible but routine to have people disagreeing about what means the right of freedom of speech or the right of property, to give two powerful examples.


 I want to argue – with Dworkin – that most rights are interpretive concepts. An interpretive concept is one open to constructive interpretation. In turn, constructive interpretation is made through the imposing of a purpose – a value regarding it – in order to present it in ‘its best light’. Such purpose explains the concept when it is able to describe the properties it has and to present a moral justification for it. One should notice that an interpreter cannot give any purpose he desires to the concept. This is  because concepts have institutional history, they have contexts in which they are given, constraints to the available interpretations[8]. Constructive interpretation “is a matter of interaction between purpose and object”, in Dworkin’s words[9]. This equals to say that rights are open to theoretical disagreements about their content[10]. Relevant legal disputes are likely to be about what is the best interpretation of a given right.


There are two qualifications that I must make: firstly, not every right is an interpretive one, some legal fields such as tax law usually give little leeway for interpretation. Secondly, arguably Dworkin’s great insight for legal theory is the realisation that law worthy the name has a justificatory character, which means that law is an argumentative practice concerned with the reason giving of judicial decisions, one can use this insight without the need to embrace Dworkin’s defence of liberalism or his ‘right answer thesis[11], the idea that ideally Judge Hercules, in possession of overwhelming knowledge, time and resources can find the right answer in the law for every single case through interpretive exercise[12]. This amounts to say that we can recognise that judges (should) strive in order to find the best possible answer to the case at hand, but this does not mean that such answer actually exists[13].


 I will try to show how theoretical disagreements within legal practice are not compatible with mercy. Rights are defended or denied through claims that present themselves as legal reasons (and this holds if we are positivists or not), and such reasons intend to have binding force towards the judge. To put in another words, if someone is asking the Judiciary to grant some right, and let’s say the plaintiff actually has a case, we would deem legally wrong a decision that denied such right, as a rights-violating decision. This, in most legal systems, would provide a legal reason towards the judicial appeal. The question for our mercy inquiry then becomes: can mercy be translated into a right?


The answer – as anticipated – is negative. The reason for that is quite straightforward: if someone can make a rights claim, then we are not talking about mercy anymore, since it makes no sense to say that someone has a legal right to be pardoned. To put this in another words: even if law and morality are necessarily connected, or even if law is a branch of morality[14], this does not mean that law and mercy belong to the same branch of morality. It is simply not appropriate to talk about mercy within legal argumentative practice. Law claims – arguably – formal justice, which means the formulation of legal claims in a general rule-like fashion that treats similar cases in similar terms. Mercy’s criteria cannot be translated into such legal claims, for it would mean mercy’s own disappearance, or law’s implosion[15]. As Christodoulidis puts it: “To claim that this [mercy] can be accommodated within law is to stretch the plasticity of law to a point beyond recognition; to a point where normativity would be compromised; to a point where law would be too poorly selective to perform even elementary functions in society - notably its function of providing some certainty of expectations. The emergence of the 'known commonality' as criterion for judgement can thus, to argue against Veitch, be only at the expense of law. To law the particularity of the affective encounter is invisible, the particularity could not have pierced the legal terms of its exclusion.”[16]


 Debates and legal decisions might not use the language of rights, but they are the necessary context to make sense of such controversies. The idea of mercy and its ‘flourishing commonality’ plays no role here, by the contrary: if a judge were to “legally forgive” a torturer, based only on his ethical perceptions (for example, let’s say the judge was touched by the fact that the torturer needs to provide sustenance to his family), without any legal reason for it, he would certainly be harshly criticised on legal grounds. If he were to give such legal reasons, then we would be constructing a right, and mercy would be already redundant.


 My thesis should not be read in a conservative or repressive tone. Law can still claim legitimacy in an unjust society through its efforts in the search for the best possible answer[17]. This search, translated in a justificatory effort in legal practice, I believe, is what grants law authority, for this cannot be founded on arbitrary will alone. By the other hand, once we recognise that law cannot be merciful, we’re short of – one might say – distracting resources that alleviate us from the burden of debating the very political issue of the nature and purpose of law and legal institutions. We have no option but debate the political decisions judges make every day. I conclude with Dworkin’s words, since they nicely summarise the view I endorse: “We construct a conception of law an account of the grounds needed to support a claim of right enforceable on demand in that way by finding a justification of those practices in a larger integrated network of political value. We construct a theory of law, that is, in the same way that we construct a theory of other political values of equality, liberty, and democracy. Any theory of law, understood in that interpretive way, will inevitably be controversial, just as those latter theories are.”[18]


[1] DWORKIN, Ronald. Laws Empire (Cambridge: Harvard University Press, 1986), chapters 1 to 3, and DWORKIN, Ronald. Justice in Robes (Cambridge: Harvard University Press, 2006), chapters 1 and 6.

[2] I hold this premise in dealing with contemporary law. This is far from true when we take the early modern legal systems, or those of antiquity. What matters here is, as Loughlin says, that “rights discourse has today become one of the major forces shaping the development of legal orders throughout the world and an especially powerful form of political expression”. LOUGHLIN, Martin. Sword & Scales (Oxford: Hart Publishing, 2000), pp.198-199.

[3] VEITCH, Scott. ‘Doing Justice to Particulars’ in Emilios Christodoulidis (ed.). Communitarianism and Citizenship (Aldershot: Ashgate publishing, 1998), pp. 226-233.

[4] Idem, pp. 232.

[5] Ibid., pp. 232.

[6] DWORKIN, Ronald. Justice for Hedgehogs (Harvard: Harvard University Press, 2011), pp. 329. This claim is repeated in many dworkinian texts, such as Taking Rights Seriously and A Matter of Principle.

[7] DWORKIN, Ronald. Justice for Hedgehogs (Harvard: Harvard University Press, 2011), pp. 331.

[8] DWORKIN, Ronald. Laws Empire (Cambridge: Harvard University Press, 1986), chapter 2. This idea is repeated in many dworkinian texts, such as Justice in Robes and Justice for Hedgehogs.

[9]DWORKIN, Ronald. Laws Empire (Cambridge: Harvard University Press), 1986, pp. 52.

[10] Idem, chapters 1 and 2.

[11] DYZENHAUS, David. ‘Recrafting the Rule of Law’ in David Dyzenhaus (ed.). Recrafting the Rule of Law (Oxford: Hart Publishing, 2000), pp. 9.

[12] DWORKIN, Ronald. Taking Rights Seriously (London: Duckworth, 1978), chapter 4.

[13] This, of course, is not our concern in this article. It suffices to say that we can (but do not necessarily need to) endorse Dworkin’s right answer thesis, but we can also embrace some reasonableness or adequacy thesis.

[14] Dworkin apparently moves from the necessary connection thesis, spoused in Taking Rights Seriously, to the law as a branch of morality thesis, stated at Justice for Hedgehogs.

[15] CHRISTODOULIDIS, Emilios. ‘‘Truth and Reconciliation’ as Risks’ in Social and Legal Studies (Jun 1, 2000), pp. 188.

[16] CHRISTODOULIDIS, Emilios. ‘The Irrationality of Merciful Legal Judgement’ in Law and Philosophy (Vol. 18, No. 3, May, 1999), pp. 224.

[17] I stress that by best possible answer, one does not need here to postulate a “right answer thesis”. This of course is a possibility.

[18] DWORKIN, Ronald. Justice for Hedgehog (Harvard: Harvard University Press, 2011), pp. 404-405.



Go Back