Justifying Legal Decisions in Hard Cases
Neil MacCormick’s Universalisability Thesis
It is ‘an important aspect of the rule of law’, writes Neil MacCormick, ‘that courts and judges take seriously the established rules of the institutional normative order’. Precisely because of this the whole business of the justification of legal decisions will ‘focus on a syllogistic element, showing what rule is being applied, and how’ (MacCormick 2006: 5). According to MacCormick, Ward LJ’s final ruling in Re A demonstrates clearly that this case must be understood:
as a type-case, as a universally stated situation … [I]t is not some ineffable particular feature of this Jodie interacting with this Mary that justifies the decision but certain statable aspects of the relationship between them in the context of a particular practical dilemma.
(MacCormick 2006: 16)
Thus, while it might indeed be true that ‘particular reasons must always exist for particular decisions’, the real issue is ‘the significance of the justifying relationship between reason and decision, and whether or not this involves the universalizability of grounds of decision’ (MacCormick 2006: 3). Ultimately, ‘[t]here is … no justification without universalization … For particular facts – or particular motives – to be justifying reasons they have to be subsumable under a relevant principle of action universally stated’ (MacCormick 2006: 21).
With this updated account, MacCormick both confirms and expands the model of legal reasoning he first presented in detailed form in 1978. There, building upon H.L.A. Hart’s linguistic criteria of ‘open texture’, and his analysis of reasoning in hard cases, he suggested a process of legal decision-making made up of several stages: universalisability; consequences; coherence; consistency. Unlike Hart, MacCormick did not suggest that a judge enjoys an almost unfettered discretion in decision-making in hard cases; rather, he outlined a theory about the constraints that govern the exercise of judicial discretion when hard cases occur. Nonetheless, like Hart, MacCormick regarded open texture as an attractive feature, allowing the law an opportunity for advancement. His theory can be stated briefly. First, the principle of universalisability entails that the way a decision is made in a hard case must also hold for decisions in every such case in the future (one must treat like as like; both backward-looking and forward-looking) and involves generalisation as a first step towards identifying the relevant general category. Second, an assessment of the consequences of generalising allows one to balance universalisability, fixing the genus through a subjective judgement of value and permitting a choice between two or more possible rulings to disclose a likely rule. Next, the requirement of coherence operates to ensure that the chosen rule can be subsumed under some principle of generality already present in settled law, that it is not simply an exercise in creative interpretation but the rule is grounded in some general principle of law, the existence of which may be described as being like part of the ‘glue’ that holds law together; in other words, that what is presented is really only a making explicit of some principle already implicit in law such that the relevant rule may be seen as correctly subsumed under it (MacCormick 1978: 44). Finally, ‘consistency’ tests the non-contradictoriness of this rule in relation to other explicitly formulated legal norms within the legal system.
This fuller account of legal decision-making emerges as consistent with the account of law as ‘institutional fact’ that MacCormick developed under the idea of an Institutional Theory of Law (ITL) in association with Ota Weinberger. Within that ‘institutional’ theory of law, MacCormick (1986) claimed to provide an ontological basis for the analysis of all social action, including law. There, they adopted the theory of institutional facts as set out by G.E.M Anscombe (1958) and John Searle (1969), drawing on their observations that there are some entities that seem to exist in the world independent of our frameworks of thought, will and judgement, which they call ‘brute facts’, and others which appear not to exist in this way; for example, a goal in a football match. We cannot point to any physical thing or event and say that it, bare and simple, is a goal, and yet we do, nonetheless, talk intelligibly about a goal. Searle called these facts ‘institutional facts’, since they ‘are indeed facts; but their existence, unlike the existence of brute facts, presupposes the existence of certain human institutions’ (Searle 1969: 51).
Institutional facts, then, are explicable given these overarching institutions and exist within their systemic framework. They might be tied to specific physical acts or events but they are not identical with these physical events. Much depends on Searle’s distinction between regulative rules and constitutive rules. Whereas a constitutive rule might define what constitutes a goal, a regulative rule would specify what one does next after a goal has been scored. The objects that together make up the physical setting for the football match assume a new form of existence in them being interpreted in terms of these constitutive and regulative rules.
MacCormick develops Searle’s distinction to suggest that legal ‘institutional facts’ (such as the temporal existence of a contract between two persons) exist within the frame of reference of certain organised activities that we may term ‘institutions’ (for example, the institution of Contract that precedes any particular instantiation of it). In this regard, three features structure our use of these concepts: institutive rules, which lay down ‘the conditions which are essential to the existence of an instance of each such institution’; consequential rules, which detail the consequences that arise as a result of the establishing of an instance of an institution, and terminative rules, which outline the provisions regarding termination of instances of institutions. But there is a difference between the institution per se and instances of it:
The existence of an institution as such is relative to a given legal system, and depends upon whether or not that system contains an appropriate set of institutive, consequential and terminative rules. If it does, then the occurrence of given events or the performance of given acts has by virtue of the rules the effect of bringing into being an instance of the institution.
(MacCormick 1978: 66)
So we can envisage such institutions as being ‘structured by legal rules’ (MacCormick 1988: 76), and:
[t]his way of conceptualising the matter … makes clear the diachronic quality … of our legal arrangements, by virtue of the way it separates or ‘individuates’ institutive and terminative rules … ‘Instances of institutions’ exist in the eye of the law … from the moment of an institutive event until the occurrence of a terminative event.
And this, in turn:
makes clear the way in which … ‘momentary legal information’ connects logically with ‘diachronic legal information’. Diachronic information concerns standing arrangements … [from which] one can derive by deduction the momentary consequential duties, liberties and powers one has in respect of the given arrangement.
(MacCormick 1988: 79)
Momentary legal information is normative: ‘it tells us what ought to or must, be or be done … what can or cannot validly be achieved … [I]t is choice guiding’ on the basis of some ‘underpinning value’ without which ‘the information would lose its practical or normative quality’. But ‘[s]etting up legal arrangements will help us achieve valued states of affairs only to the extent that we have a reason to suppose that their normative consequences will be mirrored in actual behavioural outcomes’, and ‘it is not worth much if arrangements we make can largely be ignored’. Consequently, some degree of ‘[r]elative immunity from arbitrary change is in effect a necessary condition for legal arrangements and legal institutions to have the diachronic quality which … is one of their characteristic features’ (MacCormick 1988: 79–80). Inasmuch as legal reasoning is deductive, then the model towards which this tends is ‘predicate logic’, so that ‘institutional facts could almost be re-named as “normative predicates”’ (MacCormick 1988: 81).
To explain how all of this relates to the practical decision-making setting, MacCormick recounts two familiar stories, the shrewd and penetrating judgement of King Solomon in 1 Kings 3: 16–28, in respect of the competing claims of two women to motherhood of a single child, and the death of Cleopatra. He begins this analysis by juxtaposing the Solomonic judgement and the decision of the Court of Appeal in Re A: ‘the phenomenon of conjoined twins … can [easily] pose issues quite as awful as the king’s sword’ (MacCormick 2006: 15). For instance, imagine that before some contemporary tribunal we have established the rule that ‘children should be under the custody of their natural mothers’ and, together with this, we have also developed some ‘reliable evidentiary (DNA) test’. In this event, we will immediately have translated Solomon’s skilful achievement into ‘a routine practice’, albeit that ‘the real world will always be capable of throwing up surprises’. However, the point is that as soon as the application of law is problematised in this way then the issues raised must be addressed and then the most immediate issue becomes that of how to do this (MacCormick 2006: 5).
Suppose we were to regard King Solomon’s method as a model for our judgement, offers MacCormick. On this basis, we might also consider it right to posit some form of instinctive or intuitive awareness that will enable us to latch on to the particulars of the instant case to indicate the response that the rules fail to provide and, in that situation, our answer in the present case might well be thought of as providing a precedent for future cases. But a precedent can only ever be an analogy for a new decision: no two sets of events are ever exactly the same. In these circumstances, our intuition might also tell us that it would be right to have a rule and to treat the instant case as a rule-case. Nonetheless, unusual and unexpected cases will still appear and force us to ask whether our rule permits a different interpretation or if all of the facts are classified correctly. In this way, we are now beginning to ‘problematize the rule’s applicability to the case in hand’, understanding it as ‘a case of first impression’ and directing our intuitive judgement once more towards its unique particularity. Within such a scenario, quips MacCormick, every judge will require ‘to be possessed of some small share of Solomon’s wisdom’! (MacCormick 2006: 6).
In effect, every judge will face two choices: either she must regard the instant case as a rule-case or she must concede that it presents something new. But the point is that no matter how routine the case, the judge’s decision will always be a particular decision. It is not simply a question of universals but of particulars and universals; that is, of ‘particular persons … that … instantiate certain universals’ (MacCormick 2006: 9). In this sense, the reasons that a judge gives to justify her decision will always be grounded in the particular case, but what an intuitionist approach would do is help a judge to discover her intuitive capacity to determine the features of a decision that make it right. In that sense, a good decision-making procedure would be one that:
maximized opportunities [for] careful attention to all points of a problem situation, and that gave decision-making tasks to appropriate persons … endowed both with adequate attentiveness to detail and with a fairminded readiness to make no decision till in possession of all relevant reasons in any particular case.
(MacCormick 2006: 10)
But does such intuition exist? MacCormick suggests that Adam Smith’s model of the ideal, fully informed, impartial spectator may provide the best example of how to go beyond our immediate reaction to a situation towards one that represents a ‘rationalized response to the whole of a situation in all its particularity’. But he also notes that any fully developed moral agent who is ‘capable of giving allegiance to moral rules … derived from generalizing responses to recurring types of cases’ will inevitably also belong to a community, the members of which offer allegiance to such rules. In that case, any such ‘fully refined moral capacity’ could exist only as a result of ‘a more unrefined attachment to rules of a heteronomous character’ so that deciding according to rules is not inconsistent with deciding in some ‘deeper way that confronts the whole complexity of real-life situations’. Thus, it would be wrong to overemphasize the particularistic aspects of decision-making, especially when rationalized according to Smith (MacCormick 2006: 12–13).
MacCormick finds clear evidence of this procedure in the Solomonic judgement: first, Solomon infers that one of the women is ‘the mother’; then, he delivers his judgement, ‘“Give her the child … [because] she is the mother …”’ We have to understand, he argues, that in the procedure followed here the ‘“because” nexus is all-important’. Not only does the sword-drama expose the ‘true mother’ but it also reveals how being the true mother becomes the reason for awarding the child. Moreover, since this mother relation is identified as a ‘because-reason’ here, in this case, then it also becomes a ‘because-reason’ for any future cases. What this amounts to, he suggests, is merely another way of saying that reasons must be universalisable, which is precisely why Lord Justice Ward, in his closing remarks, determines that Re A, however unique, must be regarded as a type-case. It is precisely this element of universalisability that, together with the requirements of ‘consistency over time’ and ‘an overall coherence of values and principles’, provides the foundations for the whole rationality of a system of precedents (MacCormick 2006: 13–14).
Causes are always particular, says MacCormick, and, inasmuch as we are able to discern ‘sets of like cause-and-effect series, we may be able to establish inductive generalizations from them’. However, just because:
… one particular has been shown to cause another particular would be no proof that anything else has ever caused, is now causing or will ever again cause any other thing … At the level of … observation of particulars, we never observe anything. We may see the snake biting, we may see the queen dying. But we do not see this bite causing this death. And if we did … , that would be no ground at all for supposing that every such bite will be a cause of death.
(MacCormick 2006: 18)
Therefore, the relationship between particulars and generalisations is one of ‘potential falsification’ and any ‘explanatory hypothesis’ will need to ‘be capable of forming a consistent part of a coherent general theory’. In other words:
what enables us to conceptualise the death of Cleopatra is that the particular fact of the snake biting belongs as minor premise in an argument of which the major premise is a hypothesis culled from the snake-venom theory and the conclusion is the death
(MacCormick 2006: 19)
Merely to affirm that reasons for actions are particular and factual does not prove that the connection is not a ‘relevant universal’; on the contrary, to justify an act is to demonstrate that ‘upon any objective view of the matter, the act ought to have been done … given the character of the act and the circumstances of the case’. Just so, for any reason to be a justifying reason it must indicate ‘the generic nature of the act and the generic circumstances of action’ and, as soon as these are provided, ‘an implicit principle – universal in terms – is revealed’. In this way, justifying reasons are ‘conceptually distinct’ from both explanatory and motivating reasons:
There is no justification without universalization; motivation needs no universalization; but explanation requires generalization. For particular facts – or particular motives – to be justifying reasons they have to be subsumable under a relevant principle of action universally stated.
(MacCormick 2006: 21)
On one level, ‘this is irrefutable’, concedes Christodoulidis. However, ‘it is a level that concerns the delivery of explanation rather than the making of decisions’ and as soon as we recognise the importance of the distinction between these two levels this ‘throws the issue of “particularity” wide open’ once more. Of course, the decision will depend upon particulars, and it will address questions of appropriateness as well as justify the use of universal categories, but the ‘justification of the application cannot draw its reasons from “universalizability” but from the appropriateness of extending the universal … into this set of particulars’ and that ‘is a judgement that cannot be carried in the universal category but requires attentiveness to the particular’ (Christodoulidis 2006: 98).
In this sense, law can never deliver the reasons to justify a decision. It always comes too late to inform the moment of its occurrence. In this sense, in the terms stated previously, law is always the observation of a trace left behind, a multiplicity of points through which a movement has passed, rather than the experienced unity of an action itself. Universalisation may reduce indefinitely the distance that must be bridged but, like the repeated application of a halving rule, without closing the gap: some gap always remains, the distance represented by the question concerning the appropriateness of that universal continuing into these particulars. In fairness to MacCormick, however, universalism is really only seen as doing part of the work here, even if the greater part, and particularism, as in the form of the appeal to consequences, actually concludes the task. At this point, there is always the possibility that the judge will deem the circumstances of the case to present a ‘new problem’. But in the context of decision-making in a hard case, when the rules appear to have run out and the judge is faced with what appears as a new problem, how, given the prior commitment to universalisation, will she recognise the problem as a new one? What provides the cue for her recognising the inappropriateness of applying the universal rule here? How can some particulars that do not register in law as instances of a general rule register as exceptions to it? Given a prior commitment to universalisation, how is it possible that any case might be recognised as not always-already instantiating some general rule?
If the pull to universalise is grounded in a prior commitment to select from among a variety of possibilities only those features that identify a case as always already instantiating a rule then, by definition, choosing some means not choosing others. The issue then is whether those characteristics that are not chosen thereby become invisible in such a way that their exclusion also prevents their reappearance later on, their subsequent registering as significant within the system. Are we not simply brought back once more to a question over the limits of legal possibility?
Michael Detmold’s ‘Particularity Void’: The Moment of Indecision
According to Michael Detmold there are particular situations, practical questions, which universal reasoning cannot answer. For him, part of the meaning of universality, that the rule is always applied when the conditions of its application are met, presents us with a problem. It is not that we cannot use a rule when deciding a case but that rules are not self-applying. There is a gap between a rule and its application, which he calls the ‘particularity void’. What he means is that there is a difference between asking whether a rule is reasonable and whether it is reasonable to apply it. In other words, it is in particulars and not in universals that actions must be grounded, so that an assessment has to be made each time a decision is made whether the conditions of application are met. In this way, a judge cannot evade responsibility for her decisions by hiding behind the rules. She cannot meaningfully say that she sentences someone to death and at the same that she does not support the death penalty, since she must decide each time whether it is the right thing to do and also think that it is the right thing to do. In this sense, the particularity void, as he calls it, becomes the place where I must take responsibility for my decisions.
For Detmold, legal reasoning is practical insofar as it is ‘reasoning towards a decision for or against action’ and his primary concern is with a ‘judge’s practical reasoning towards the action of giving judgment’ (Detmold 1989: 436), what he calls ‘[t]he particularity of adjudication’. He gives the example of someone seeking to acquire judicial office through examination:
I am given a problem to solve consisting of facts A B and C. I work it out and … [m]y conclusion is universal: a defendant in circumstances A B and C must pay damages. But am I right? I check my reasoning and conclude I am right. I finish the exam, content. But … I go to my books after the exam to make sure. Yes … I am sure. I am now sure that I have the answer to the (universal) question …
(Detmold 1989: 455)
Nonetheless, this is still not ‘a practical answer’, argues Detmold. It will become practical only when it becomes particular. But is this merely a matter of waiting for a suitable particular to come along that accords this universal judgement?
In due course I am appointed, he continues, and my first case replicates the case of my exam. As I sit alone in my chambers contemplating judgement, why does my will not unleash itself? It is not that I doubt my conclusion: ‘I remember my reasoning very clearly’. But ‘I now have a radically different problem’, he explains, one ‘which universal (hypothetical) reasoning does not solve’. Indeed, ‘the whole problem is that no reasoning can solve it. It is particular’. It is something of which ‘nothing can be said (anything I say will be universal)’ (Detmold 1989: 456). How might we account for this?
We can take this further, Detmold suggests, by looking at the confrontation between Pierre and Davoût in Tolstoy’s War and Peace. A ‘moment of indecision’ saves Pierre from being shot as a spy on Davout’s orders. Davout, holding his rifle, looks towards Pierre; he hesitates and does not fire. At this moment, according to Tolstoy, many things pass through Davout’s mind:
Davout looked up and gazed intently at him. For some seconds they looked at one another, and that look saved Pierre. Apart from conditions of war and law [ABC] that look established human relations between the two men. At that moment an immense number of things passed dimly through both their minds, and they realised they were both children of humanity and were brothers.
(Detmold 1989: 457)
At first, this appears very like universalist reasoning and indeed Tolstoy seems to suggest as much. But for Detmold both the hesitation and the action are deeply significant:
Davout, at the moment of practicality entered the unanswering void of particularity, the realm of love, about which only mystical, poetic things can be said … ; or nothing … Judges enter this realm every day (if only they knew).
(Detmold 1989: 455–57)
In this sense, he claims, those theorists who seek to find through ‘the progressive refinement of the categories of law according to experience’ a means by which to settle these issues are mistaken; in fact, no matter how:
highly defined A B and C are … [the] problem [is] exactly the same … A judgment in respect of A B and C … cannot cross the void … [I]t can justify a judgment … a theoretical/hypothetical … right up to the void. But the final rationality of practical judgement seems in doubt …
(Detmold 1989: 458)
Detmold notes how Neil MacCormick has attempted ‘to reassert that rationality against … particularity’ through a reconsideration of the idea of justification. But, for Detmold:
th[is] act of justification is incapable of solving the problem for it immediately raises the question, why justify? and the answer, like that to the original question, will be ultimately particular, not universal; so it will have its own particularity void.
(Detmold 1989: 459)
Even MacCormick’s attempt to derive the desired universality along the lines of Adam Smith’s postulate of the ideal spectator does not successfully evade criticism, since ‘anyone’s question is anyone’s void’. Ultimately, what MacCormick and Smith both fail to demonstrate, he argues, is ‘how the impartial spectator’s judgment is not also incorrigibly particular’. In fact, we find ‘two questions of universalization’ involved in practical reasoning: on the one hand, ‘whether I am to be universalized to all moral agents judging p’ and, on the other, ‘whether p is universal or incorrigibly particular’. Moreover, ‘it is p which opens the particularity void and casts doubt on the truth of all practical judgments, subjective or objective’ (Detmold 1989: 459). We need to investigate further, he claims, this question of whether ‘the negotiation of the particularity void depends upon the particular in respect of which my action is contemplated speaking for himself’. Although reason brought Davoût to an ‘acceptance of the norm: execute all Russian spies’, nonetheless, a ‘void of reason … stood between this norm and the particular Pierre’. Davoût might just as well have responded: ‘it is reasonable to execute the enemies of France, but why should I do it?’ So, we find ‘a second particularity void: … one for subject as well as object’ and, in the end, ‘particularity holds out’. Ultimately, what this suggests is ‘a category leap: the particularity void cannot be crossed by reason’ (Detmold 1989: 464–65).
More recently, Detmold has elaborated further on the problems of particularity in adjudication identifying two problems, which he characterises as the ‘in-tray’ and the ‘out-tray’: the ‘in-tray’ is the matter of what it is that informs a decision, how it will be justified and the extent to which the informing thing is particular or universal; the ‘out-tray’ is the making of the decision and deciding to whom or to what it applies, the particular or the universal. According to Detmold, all practical judgements are of the out-tray and are radically particular: even though there are clear difficulties with the notion of deciding something about another person’s life, still the common law seeks to address itself to history, an always radically particular history, judging that history.
In other words, it is always a particular person and a particular history that form the basis of the law’s judgement (Detmold 2006: 83–94).
On this view, the gap that we recognised earlier in Re A may be seen to have at least two aspects: first, in terms of the potential asymmetry between addressor and addressee; second, in terms of the void between determination and application. In general, judges only tend to make law conservatively, says Detmold, but ‘the fullness of law as practical reason is achieved when the law that judges apply is law that has crossed the citizen subject void; when law is in a true sense the citizen’s law, when law is common law’ (Detmold 1989: 467).
American Legal Realism and the Critical Legal Studies Movement
‘American Legal Realism was the most important indigenous jurisprudential movement in the United States during the 20th twentieth century’ (Leiter 2003: 50). Indeed:
[c]urrent debates about legal reasoning are best understood as attempts to answer the central question that the realists left unresolved: How can we engage in normative legal argument without either reverting to the formalism of the past or reducing all claims to the raw demands of political interest groups?
(Singer 1988: 468)
Essentially, the legal realists argued on three fronts as to why rules alone cannot decide cases and are of limited use in predicting the way that a court will decide: first, legal rules resist any form of mechanical application because they are inherently vague and ambiguous; second, this vagueness and ambiguity means that any case can be read in a number of different ways; third, the resulting indeterminacy of abstract legal concepts leads to a manipulability of precedent through a blurring of the distinction between holding and dictum (ratio and obiter). Effectively, a judge could redefine the holdings in precedent cases to reveal alternative or multiple rules of law capable of governing the outcome in the case before her in different ways.