Ashworth on Principles
Abstract and Keywords
This chapter asks the logically prior question of what we mean by principles. Section 1 first clears up a couple of preliminary issues about the scope of the discussion. Sections 2 to 5 explore, by turns, four properties which Ashworth seems to ascribe to principles: generality, special force in argument, non-instrumentality, and categoricality. Section 6 reflects on links that may be thought to hold between these four properties, and arrives at a somewhat sceptical conclusion.
Does Andrew Ashworth's peerless body of work1 on criminal law and the criminal justice system have a unifying theme? The most striking is his emphasis, at almost every turn, on the importance of principles. Ashworth sees principles in the law and advocates fidelity to them. He also stands up for principles that, in his view, ought to guide and constrain the law's development and official use, even if they are not to be found in the law. In what he acknowledges to be a triumph of hope over experience, Ashworth yearns for a principled criminal law as well as a principled approach, among politicians and officials, to decisions about whether and how to resort to it. Here are two typical passages—two among many—expressing that yearning:
Legislatures frequently create or re-enact offenses without proper consideration of the extent of their conformity with general principles or of the justification for departures. The argument here is for a more structured and more principled approach.…In an ideal system, the legislature would make a principled inquiry into the need for a criminal sanction, the form of the offense, the relative seriousness of the crime, and the case (if any) for derogation from first principles, before enacting any offense.2
(p.4) [My] main purpose has been to develop two lines of argument. The first is that the criminal law is indeed a lost cause, from the point of view of principle…The second line of argument is more constructive, in seeking to identify a principled core of criminal law. The core consists, it is submitted, of four interlinked principles…It is not claimed that they should be regarded as absolute rules, and indeed at various points above some possible qualifications to them have been discussed. Derogations from them should be argued as derogations, and should be principled in themselves.3
These passages, and many others like them, give us various clues as to what Ashworth thinks a principle is. He typically provides conceptual orientation by connecting principles (more or less contrastively) with rules,4 values,5 policies,6 doctrines,7 interests,8 and various other things. But so far as I am aware he never spells out the criteria for something to qualify as a principle, or for someone to qualify as principled. Nor is it incumbent upon him to do so. He is not writing primarily for philosophers. Yet there is sufficient philosophical depth in Ashworth's writings to make the question ‘What exactly does he mean by a principle?’ a live one, and one to which we might reasonably expect to find an answer by working back from the many scattered remarks he makes in which principles are mentioned or endorsed.
That answer is important to anyone who wants to think critically about Ashworth's views on the criminal law and the criminal justice system. Until we know what a principle is, it is hard to know whether Ashworth's warmth towards principles and their use is warranted. Not all moral and political philosophers share the warmth. Some, indeed, are antagonistic towards principled thinking in practical matters. But could that simply be because they have a different view from Ashworth of what a principle is, or of what it means for something or someone to be principled? Quite possibly. In what follows I will attempt to work out Ashworth's views on these matters, adding some critical commentary, and some comparisons with other possible views, as I go along. Section 1 clears up a couple of preliminary issues about the scope of our discussion. Sections 2 to 5 explore, by turns, four properties which Ashworth seems to ascribe to principles: generality, special force in argument, non-instrumentality, and categoricality. Section 6 reflects on links that may be thought to hold between these four properties, and arrives at a somewhat sceptical conclusion.
1. Some Preliminaries
On any plausible view, including Ashworth's,9 principles, policies, and rules have this much in common. All are taken to provide reasons (or in other words to (p.5) militate in favour of or against something) by any person who endorses them. But do they provide the reasons they are taken to provide? Only—one might assume—if they are good principles, policies, and rules. If one endorses them, necessarily one takes them to be good. But one may be mistaken. In which case one takes a principle, policy, or rule to provide a reason that it doesn’t actually provide. In truth it does nothing to support the thing that it is supposed by its endorsers to support. At any rate, so one might assume.
Yet there is a rival view, very much alive in British politics and public culture. According to this view, the mere fact of endorsement makes a principle, policy, or rule reason-giving for the endorser, even if the principle, policy, or rule is (otherwise) worthless and does not deserve the endorsement. For example, politicians are criticized by some for departing from their own policies, however dire. They are also admired by some for sticking to their own principles, however bizarre or fanatical. Indeed, one common way of understanding what makes a person ‘principled’—where this is treated as an honourable thing to be—is that he or she sticks to his or her principles doggedly, never mind which principles they are.
Occasionally Ashworth speaks as if this is his understanding. He would apparently prefer politicians that have and stick to ‘their own principles’.10 He also seems to set some store by the fact that certain principles are already endorsed by the law, as if the law should be assessed according to its conformity with the principles that it already endorses, because it endorses them.11 However, most of the time Ashworth's interest in principles (and for that matter in policies and rules) appears to be more critical. He is interested in which principles and policies and rules they are, and he wants them to be the right ones. We all know that he would not enthuse about a system of criminal justice that stuck to its principle of presuming guilt, or its policy of remanding suspects in custody indefinitely, or its rule that no crimes can ever be excused. Would he nevertheless regard it as a slight redeeming feature of such a system that at least it has its principles and sticks to them? Or would he think, as I do, that having and following its dreadful principles might make the system even worse than it would be if it did the very same dreadful things without any principles at all?12 Ashworth's position in this debate is less clear. Even if he does not think that principles mitigate evildoing, he may think, with Lon Fuller,13 that they militate against evildoing.14 The question need not, however, detain us for long. For the purpose of understanding what exactly a principle might be, we will concern ourselves (and assume that Ashworth concerns himself) with sound principles only, and likewise, for parity, with sound policies and rules.
We will also avoid landing Ashworth with any commitment to the desirability of having principles, or of being principled, outside the roles of law-maker and (p.6) law-applier (and some allied roles). Many who harbour doubts about principles and principled reasoning are thinking of the ways in which such things might figure in non-institutionalized aspects of everyday life.15 They are thinking about ordinary people who do what they do (refusing to pay their bank charges, boycotting an airline, declining to eat meat, advocating capital punishment, buying fair-trade chocolate, wearing eyeliner to school, not talking to their neighbours, speaking French, demanding compensation, etc) ‘on principle’ or ‘as a matter of principle’. George Bernard Shaw lampooned such people as quintessentially English moral obfuscators, concealing their reasons for action behind a veil of righteousness with a view to forestalling any call (even from themselves) for a proper justification.16 André Gide expressed the same charge less affectionately: ‘One shouldn’t expect any kind of sincerity from [people of principle].’17 I have sympathy for these negative reactions. But I do not think that they—or the various more measured and focused philosophical critiques that echo their themes18—typically have the same force in relation to the use of principles by law-makers and law-appliers as they have in relation to the use of principles by non-institutional agents. In fact one possible worry about the prestige that some people attach to principled thinking outside the law is that it may betray too legalistic a view of life, placing too much faith in certainty, clarity, prospectivity, generality, consistency, finality, and other desiderata of legality, and sacrificing thereby some attention to the particularity of each situation.19 The law needs to live up to the desiderata of legality—one might well think—precisely because most of what life throws at us, including its everyday moral demands, does not live up to them at all. By this backdoor route one could arrive at a critique of the place of principles in practical thought, and in the art of life, which makes a unique exception for, and even gives principles pride of place in, the special business of legal thought and legal practice.
This might be Ashworth's own position. He says nothing to suggest that his hallmark enthusiasm for principles extends beyond the making and applying of law (and allied activities). And he often juxtaposes his advocacy of a principled criminal (p.7) law with remarks on the importance, especially in the criminal law, of the desiderata of legality. His case for a ‘more principled approach’ to criminal legislation relies on principles that are ‘strongly related’ to those of ‘certainty, consistency, comprehensibility, and accessibility’.20 Elsewhere he writes:
The…disadvantages [of vaguely defined criminal defences] are that the broader issues of principle raised by certain arguments for exculpation are unlikely to be considered thoroughly, and that the law cannot function properly as a source of guidance for conduct if the boundaries of permissible conduct are undefined…[C]oherence is an important value in the criminal justice system—coherence not merely as the absence of contradiction, but more positively as a network of mutually supporting rules and principles.21
Is there an implicit ‘therefore’ after the conjunction in the first sentence? Does Ashworth think that the law's failing to confront ‘broader issues of principle’ is bad at least partly because it detracts from the ability of the law to guide people (ie because it detracts from the law's compliance with the rule of law)? That is a credible view, at least in the common law systems, where there is often nothing to rely upon for guidance except a line of cases.22 If it is Ashworth's view, as I will assume it is, then it already puts some distance between him and those who are enthusiasts for principled thinking across the whole arc of human life. I think the existence of that distance is much to Ashworth's advantage, but I will not be arguing the point any further here.
2. Principles as General
In some cases the courts, according to Ashworth,
…make high statements of principle, which may raise hopes that a consistent framework is to be established…In the past, any hopes of a consistent judicial approach have usually been dashed, as the supposed principle is progressively whittled away or, more damningly, simply ignored.23
Here we have another passage connecting a principled criminal law with one of the desiderata of legality—this time, consistency. But what is meant by ‘consistency’? Does it mean absence of conflict, which is what it is usually taken to mean in discussions of the desiderata of legality? Apparently not. It is true that, when a principle is ‘simply ignored’, that makes for a new conflict in the law. When a principle is ‘whittled away’, however, the reverse is true. Whittling away leaves less law than we had before, and so tends to reduce legal conflict.24 What is sacrificed by (p.8) the whittling away is not consistency in the sense of conflict-avoidance, but consistency in the sense of generality, quite a different desideratum of legality. As the principle is whittled away it applies to fewer cases. This, for Ashworth, is already a loss to the law. More importantly for our purposes, it turns a principle, according to Ashworth, into a ‘supposed’ principle.
This tells us one thing that Ashworth believes about principles. He believes that principles provide not just reasons but general reasons, reasons of general application. Or does he? Quite often, as in the first passage quoted at the start of this essay, he speaks of the importance of ‘general principles’.25 That may be taken to imply that there can also be non-general principles. But there are other possible readings. Perhaps all principles are general but some principles are more general than others, so that ‘general principles’ means ‘more general principles’. Or perhaps—more likely—‘general principles’ is a pleonasm used by Ashworth to remind us of the built-in generality of all principles.
That principles provide reasons of general application, or have a built-in generality, seems to me undeniable. But as it stands the point does not help to differentiate principles from anything else in the neighbourhood. All reasons are of general application. If the fact that it is raining is a reason to put up an umbrella, then the fact that it is raining on Monday is a reason to put up an umbrella on Monday, the fact that it is raining on Tuesday is a reason to put up an umbrella on Tuesday, the fact that it is raining on Wednesday is a reason to put up an umbrella on Wednesday, and so on. A complication arises, to be sure, if Wednesday's rain, unlike Monday's or Tuesday's, is combined with high wind. Not only is the fact of the high wind a distinct reason not to put up an umbrella (the umbrella will be wrecked and then will need to be replaced in time for Thursday's windless downpour). It also has the consequence that the fact of the rain on Wednesday is not after all a reason for putting the umbrella up (the umbrella, wrecked as soon as I put it up, won’t keep me dry for even a moment). Yet that does not go to show that the reason we are considering, the advertised reason to put up an umbrella, is of less than perfectly general application. It only goes to show that we have not yet stated the reason in full.
The advertised reason to put up an umbrella, more fully stated, is the fact that it is raining coupled with the fact that putting up the umbrella will help to keep me dry in the rain. Since ex hypothesi the second fact does not hold on windy Wednesday, the reason to put up the umbrella does not extend to that day. Fine tuning of the reason, revealing that it is incompletely stated even after this emendation, can go on for many further steps. There is the fact that I will be cycling to work on Friday, the fact that on Saturday I will be in training for a damp Scottish fell-run, the fact that I deserve a good soaking after my grumpy remarks last Sunday about people with umbrellas, and so on. In fact it is possible to distinguish different rainy days and different potential umbrella users along a potentially infinite number of lines in respect of the umbrella-favouringness of (p.9) the rain. At each step, however, what we may call the generality thesis holds: nothing counts as a rationally salient difference between any two situations unless it is a difference between two generic types of situations, unless there are facts that hold in one of the situations that do not hold in the other, facts that add up to explain why the situations are different in respect of the reasons for action that each affords to whom.26 If one is to preserve rational intelligibility, one cannot say ‘that reason won’t apply next Sunday’ without identifying the imaginably repeatable feature or features of next Sunday in virtue of which the reason supposedly won’t apply then.
Is there some stronger sense of ‘general’, going beyond this requirement of rational intelligibility, in which principles are supposed by Ashworth to be general? Probably there is. But it is not at all clear what it is. His remark about ‘whittling’ principles away such that they turn into merely ‘supposed’ principles suggests that he might want to deny that reasons are principled as soon as any proviso is entered—as soon as the high wind breaks the constant umbrella-favouringness of the rain, or (to generalize) as soon as the constant rational salience and valence of any given fact is interrupted or complicated by the introduction of another fact. Among moral philosophers, the word ‘principles’ is sometimes reserved for reasons of such pristine constancy.27 It is armed with this understanding of what principles are; indeed, that some have come to doubt whether morality (or practical life as a whole) has much or any space for principles.28
But Ashworth does not help himself to this ‘purist’29 account of principles. Recall what he says about derogations:
Derogations from [principles] should be argued as derogations, and should be principled in themselves.30
Doesn’t ‘derogation’, when done a bit at a time, come to the same thing as whittling? If so, why is the process presented here as consistent with the whittled-away principle remaining a principle, when elsewhere it demotes it to the status of a ‘supposed’ principle? Is it perhaps a matter of degree, such that an apparently principled reason becomes less principled with each successive whittling, even when the whittling itself conforms to the generality thesis? That is a possible interpretation of what Ashworth says. But notice that, in arriving at or defending this interpretation, it does not help much to know that for a principle to remain a principle, any whittlings must be ‘principled in themselves’. To know what that means we would already need to know how ‘principled’ derogations differ from (p.10) those that simply meet the requirements of the generality thesis, ie those that are rationally intelligible. And that is the very thing that we are trying to find out.
3. Principles as Powerful
Maybe the answer lies in Ashworth's remarks about the force of principles, by which I mean their ability to defeat countervailing considerations in cases of conflict. He believes that ‘[a]rguments of principle cannot be easily overridden’.31 Maybe he thinks that whatever makes principles resistant to override also makes them resistant to derogation. In fact, he does not always cleanly distinguish override from derogation, and he may not attach much importance to the distinction. For example, he writes:
In certain spheres there may be other values and interests that are regarded as so strong as to displace the general principle of equal treatment.32
What does ‘displace’ mean? Does it mean ‘render inapplicable’? Or does it mean ‘defeat’, which, on the contrary, presupposes continued applicability? In a similar vein Ashworth writes:
These are put forward as core principles. It is not claimed that they should be regarded as absolute rules, and indeed at various points above some possible qualifications to them have been discussed.33
The word ‘qualification’ brings to mind what above I called a ‘proviso’, a limit built into the scope of the principle. But in that case, why is a ‘qualified’ principle being contrasted with an ‘absolute’ rule, which is presumably a rule of such force that, whatever the limits of its application, it defeats all opposition wherever it applies? Maybe what Ashworth sometimes presents as the generality of principles is not so much a distinctive scope as a distinctive force that allows them to prevail over other reasons so that they are more often (hence: more generally) decisive?34
The suggested contrast with ‘absolute rules’ is intriguing in a second way. It suggests that for Ashworth all principles, or at least some of them, are rules, even if not absolute ones. That sounds right to me. It was Ronald Dworkin who encouraged us to think otherwise. He famously proposed that no principles are rules, because rules ‘are applicable in an all or nothing fashion’ whereas ‘[p]rinciples…have a dimension of weight or importance’.35 This contrast is doubly misleading. First, it confuses the scope question with the force question. The weight of a principle is part of its force, and tells us nothing about when it applies. The supposed ‘all or nothing’ application of a rule, conversely, tells us nothing about its force. Secondly, and more importantly, rules necessarily do have a dimension of (p.11) ‘weight or importance’. They can conflict with other rules and then it has to be decided which rule is the more important. It is true that when two legal rules conflict, and one is found to be more important, it is tempting for lawyers to tidy up by declaring the more important rule to be (henceforth) the only rule, or by attempting to accommodate such force as the less important rule has by way of exception or proviso to the more important. But this points to nothing special about the rational force of rules. It only draws attention to the pressure on the law to conform to the desiderata of legality, including the consistency desideratum, in its dealings with rules (whatever their rational force may be before and after those dealings).36
Yet there is certainly something special about the force of rules. Rules are what Dworkin calls ‘trumps’.37 They defeat at least some countervailing considerations by kind and not (only) by weight. It follows that on some occasions they do not need to rely on their weight (or their weight alone) in order to prevail. What Ashworth calls an ‘absolute’ rule is presumably one that never needs to rely on its weight in competition with other considerations, except for other considerations of the same kind, because it defeats them all by kind alone. They are all trumped, or (as it is sometimes more technically put) ‘excluded from consideration’ by the rule.38 That Ashworth regards at least some principles as having at least some of the same exclusionary force is readily apparent. True, this is not entailed by his claim that ‘[a]rguments of principle cannot be easily overridden’, which is consistent with the view that principles are just relatively weighty reasons. But consider:
[T]he principle of mens rea ought to operate as the primary restraint upon the pursuit of other aims such as social defense. This, indeed, is the significance of the doctrine of fair opportunity…True, public safety is one of the reasons for having criminal laws, but it does not follow that, where the element of social danger arising from certain conduct is high, this supplies a strong justification for dispensing with the doctrine of fair opportunity.39
So the principle of mens rea is not defeated by mere increases in the weight of the considerations of ‘social defence’ that conflict with it. Why not? Why does victory not turn simply on how weighty the conflicting considerations of ‘social defence’ become, and how slight is the departure from the principle of mens rea that they call for? For Ashworth things are plainly not so simple. The countervailing considerations of ‘social defence’ are defeated by kind rather than by weight. Or at any rate they are downgraded by kind in such a way that they count for less than their (p.12) weight in competition with the principle of mens rea.40 And the point may apparently be generalized to a range of other principles, if not to all. Once the law is ‘objectionable in principle’ there is often, in Ashworth's work, a relatively quick move to the law's being ‘absolutely indefensible’.41 The move is quick, I take it, because principled objections have a built-in advantage in their conflict with other considerations. They do not need to rely on their weight alone to prevail.42
4. Principles as Non-instrumental
The passage just quoted introduced us, in passing, to another recurrent theme in Ashworth's treatment of principles. Here it comes into the foreground, and with it comes a puzzle:
In considering the interaction between the principles [of criminal legislation] and other arguments, usually based on social defense, we have noticed various criteria for derogation from the principles.43
The puzzle: why should the contrast class of ‘other arguments’ be differentiated, even ‘usually’, by a value that figures in them (‘social defence’), unless principles are also differentiated by some value that figures in them? Another iteration of the puzzle:
Greater use of fixed penalties, of plea bargaining, and summary trials in England and Wales might all be seen as examples of managerialist techniques that sacrifice principle…for effectiveness.44
Why is ‘principle’ contrasted with ‘effectiveness’ here? That does not seem to be a contrast of like with like. A principle (we have already gleaned) is a rule or something like it. So it is something in which values could be embodied, reflected, or encapsulated, whereas ‘effectiveness’ is surely just one of the values that might be so embodied, reflected, or encapsulated. One might think that there has been some kind of category mistake when a device for engaging with value, a principle, is contrasted with one or more of the values that it might be used to engage with.
Or maybe not. One might instead be reminded of another contrast once drawn by Ronald Dworkin, namely the contrast between principles and policies. A policy, for Dworkin, ‘sets out a goal to be reached’ (or a ‘present feature to be protected from adverse change’), whereas a principle is ‘a requirement of justice or fairness or some other dimension of morality’, where morality is understood not to require the (p.13) reaching or maintaining of some goal.45 Ashworth too draws a principle–policy contrast, and he may well mean to draw it along Dworkinian lines:46
[Principles] amount to strong arguments based on moral or political foundations rather than absolute precepts…Reference is also made below to certain policies founded on arguments about what is expedient rather than what is right in principle.47
English criminal law both is shaped and ought to be shaped by a number of principles, policies and other standards and doctrines. One of the purposes of this chapter is to draw together and to discuss critically some of the foremost principles that ought to exert an influence on the substance of English criminal law…The justifications for upholding each principle are discussed, and each principle is followed by a policy or other instrumental goal that may often run counter to the principle in practice.48
So a policy is a kind of ‘instrumental goal’. What does that mean? The expression may be elliptical. To judge by Ashworth's main examples of policies, it is not the goal that is supposed to be instrumental, but rather the policy that is supposed to be instrumental towards the goal. The goal is the thing that is served instrumentally; it is a state of affairs to the realization of which actions conforming to the policy contribute, to the extent that they do, in a causal as opposed to a constitutive way. That is why—to return to the puzzling contrast between ‘principle’ and ‘effectiveness’—the question of effectiveness arises in connection with a policy. But it does not arise, we may glean, in connection with a principle. For a principle is not there to serve a goal, or perhaps even to contribute to a state of affairs, unless you count the state of affairs in which the principle has been conformed to, a state of affairs to the realization of which the conforming action made a constitutive, as opposed to a causal, contribution.49
This way of distinguishing principles from policies has odd implications. It entails that Bentham's ‘principle of utility’50 is not an unsound principle, for it is not a principle at all; likewise the ‘precautionary principle’ favoured by some policy-makers,51 the ‘harm principle’ set out by John Stuart Mill,52 and the Roman Catholic ‘principle of subsidiarity’.53 None are principles. Nor, for that (p.14) matter, are the desiderata of legality the principles that Ashworth himself proclaims them to be. As he says:
[A] legal system…should adhere to rule-of-law principles in its criminal law, by ensuring fair warning, maximum certainty of definition, subjective requirements for criminal liability, and so on.…By these means, as H.L.A. Hart put it: ‘First, we maximize the individual's power at any time to predict the likelihood that the sanctions of the criminal law will be applied to him; Secondly, we introduce the individual's choice as one of the operative factors determining whether or not these sanctions shall be applied to him.’54
Maximal or optimal freedom under the law is a goal, which is served only instrumentally by the desiderata of legality.55 There is a live question of their effectiveness in securing it.56 So if those desiderata must be classified as either principles or policies, then according to the criterion under scrutiny here, which is supposedly Ashworth's own criterion, they should surely be classified as policies, not—pace Ashworth—as principles. This pushes us towards the conclusion that the criterion under scrutiny here is not, after all, Ashworth's criterion, or at any rate not the criterion he is looking for.
5. Principles as Categorical
Something went astray in Ashworth's characterization of a policy as an ‘instrumental goal’. My suggested explanation is this. Talk of a ‘goal’ sends out two different signals. On the one hand it suggests something that can be pursued instrumentally. On the other hand it suggests something to which one need not, but might, become rationally committed. In the latter sense, I have my goals and you have yours. Some of my reasons for action are given to me by my goals. Given my goal of running the Glasgow half-marathon, for example, I have an extra reason to go running this morning that many people lack. Notice that this is an example of a goal in the second sense which is not a goal in the first. At the end of my training, I can run the Glasgow half-marathon only by running it, and then what I do will contribute constitutively, not causally, towards my fulfilment of the goal.57 Governments and public authorities can have goals in this sense too, qualifying as goals not by virtue of the instrumentality of the actions taken in their pursuit, but rather by virtue of the rational optionality, as we might call it, of the pursuit itself. The institution has some rational latitude in respect of which goals it adopts, and when it adopts a goal as its own it acquires new reasons that it would not otherwise have had.
(p.15) In the sense that Ashworth often has in mind, it seems to me, the reasons given by policies belong to this class of goal-dependent reasons. Possibly policies are themselves goals in the relevant sense (so Ashworth was right to speak of a ‘policy or other…goal’) and that is why the reasons to which policies give rise are goal-dependent.58 Principles are different. One has reason to conform to sound principles irrespective of one's goals. They supply what are sometimes called ‘categorical’ reasons, reasons which one cannot lose merely by changing one's goals.
This is the contrast between principle and policy that is suggested by the link that Ashworth sometimes forges between policy questions on the one hand and ‘collective goals’ or ‘social goals’ on the other, where such goals, and the priority among them, ‘should be a matter for democratic (participatory) decision-making’.59 Here ‘social goals’ are not goals for a society (as Dworkin's use of the same expression sometimes suggests)60 but goals of a society. Their adoption gives the government, and the authorities more generally, reasons different from those they would have had, had different goals (or different priorities among goals) been adopted. Policy depends on such adoptions, thinks Ashworth. But principle prevails irrespective of the adoption of any goals by anybody, whether democratically or otherwise.
You may think that this way of distinguishing principle from policy (or this aspect of the distinction) flies in the face of the fact, already noted, that people often have their own principles. Clearly they may adopt principles in the course of adopting or shaping their goals; they may even end up with highly principled goals (to save the planet, to uphold justice, etc). Yet once they have their principles, people hold those principles to give them categorical reasons for action. They do not regard their principles as binding on them only for as long as they continue to have the same goals. It is often thought to be part of being a ‘principled’ person; indeed, that one resists a change of personal goals that would allow one to escape from the tyranny of what one now regards as principles, by allowing one to stop regarding them as principles and to start regarding them as something more like policies. Such resistance requires a certain capacity for obstinate self-deception that may provide the beginnings of an explanation for the ribald and scathing views that Shaw and Gide respectively took of principled people. But that is not yet an indictment of principles themselves. It is an indictment of the natural tendency to elevate what are really goal-dependent reasons to the status of categorical ones. Having had some categorical reason to ϕ, and having responded to it by making (p.16) ϕing one's goal, it is a short step to self-righteously thinking of all one's reasons to ϕ as categorical ones, forgetting that some of them come of the fact that one adopted ϕing as one's goal. The designation ‘principled’ is sometimes reserved for the tiresome people who do this.
There is nothing in this that need worry Ashworth, to whom (you will recall) we are attributing no views about the role of principles in personal life. Yet there is much in it to support what I take to be part of his view about what principles are, viz. that the reasons they give us, when they are sound enough to give us reasons, are categorical ones. If that were not so then the foregoing remarks about the psychology (or the pathology?) of ‘principled’ people would not hang together as they do.
6. Combining the Criteria
There is a view according to which morality is a body of principles, where all principles are understood to give reasons for action that are (a) distinctively general, not only satisfying the generality thesis to which all reasons are subject, but also being (as I put it above) pristine in their constancy; (b) of absolute force, defeating all competing reasons by kind; (c) non-instrumental, in the sense that the value in conforming to them is constituted rather than caused by that conformity; and (d) categorical, in the sense that their rational hold over us does not depend on what personal goals we have at the time. In Kant's work all of these features of morality were said to follow from morality's unconditionality or universality, a vague idea that was interpreted in numerous different and often inconsistent ways by Kant. Each of the proposed criteria (a) to (d) reflects a different interpretation of that vague idea. As has often been said, Kant's urge progressively to strip away all kinds of conditionality from morality left less and less rational space for morality to inhabit.61 Probably, if moral considerations must meet all of the criteria that Kant set for them (which (a) to (d) do not exhaust) then there are no valid moral considerations. Even criteria (a) to (d), without more, are capable of eliminating the whole of moral life, depending on their further interpretation. And if (a) to (d) are understood as criteria for something to qualify as a principle, moral or otherwise, then quite possibly there are no sound principles of anything anywhere in the world.
There is no reason to think that Ashworth is wedded to the Kantian illusion of an unconditional morality. And as we have seen, he is certainly not wedded to conditions (a) to (d) as conditions for something to count as a principle. Yet, as we have also seen, he echoes all of them in his work. He embraces diluted versions of (a) and (b). He expects constancy from his principles, but he is no ‘purist’ about it (section 2 above). He regards principles as having a special force, an ability to prevail by kind and not only by weight, but he does not expect it to be absolute (p.17) (section 3). In the spirit of (c), he contrasts arguments of principle with instrumental arguments, but does not uphold the contrast in all his examples (section 4). Only (d), towards which he only gestures, seems to square with everything he says (section 5).
Nevertheless, in the role he gives to principles throughout his work, Ashworth does seem to court, or give succour to, the loosely Kantian idea that criteria (a) to (d) fit naturally together. He gives the repeated impression, including in many of the passages quoted above, that considerations with the force of principles should be expected to apply with the generality of principles, that considerations with the force of principles should also be expected to be categorical, that categorical considerations should be expected to be non-instrumental considerations, that non-instrumental considerations should be expected to be general and categorical, and so forth. In this essay, I have tried to show how these false expectations arise.
By and large, I have suggested, they come of small and understandable conceptual slips. As we saw in sections 2 and 3, it is all too easy to run together the question of scope raised by (a) with the question of force raised by (b), thanks to the equivocality of words like ‘derogate’ and ‘displace’. As we saw in sections 4 and 5, it is all too easy to confuse the criterion of non-instrumentality in (c) with the criterion of categoricality in (d), thanks to the different signals sent out by the word ‘goal’. We did not stop to add, but it would not take too long to show, that the status of being categorical is easily and commonly taken to entail having a special force in conflict, thereby linking (d) to (b), and compounding the other small slips with another.
These small slips add up to give a misleading picture of practical rationality as a whole, even as it bears on the law. Although at some times and in some places there may be a case, a rule-of-law case, to build up a body of law that works at a high level of generality, that is treated as forceful in argument, and that is presented as non-instrumental and not contingent on political goals, there is nothing otherwise—nothing apart from this body of law—to suggest that categorical considerations tend to have more rational force than goal-dependent ones, that non-instrumental considerations tend to have more rational force than instrumental ones, that more general considerations tend to have more rational force than less general ones, that less general considerations are more likely to be instrumental and goal-dependent than non-instrumental and categorical, that goal-dependent considerations are more likely to be instrumental than non-instrumental, and so on. In the ordinary logic of reasons, all these properties vary independently of each other. In particular, there is no shortcut to showing that a certain consideration has the force necessary to defeat another consideration in a particular conflict; all one can do is show that that is the force it has in the context of that particular conflict. Its generality, its non-instrumentality, and its goal-independence tell us nothing about how it is going to fare in competition with other considerations.
All of this is consistent with holding, with Ashworth, that ‘[a]rguments of principle cannot be easily overridden’ and that principles provide ‘strong arguments’, even for the purposes of debate outside the courts (eg in forming or legislating criminal justice policy). But the points I have just made tend to suggest (p.18) a reversal of Ashworth's inference. Ashworth gives the impression, in various remarks we have quoted, that an argument of principle cannot easily be overridden because it is an argument of principle, ie because it already satisfies his other criteria for being a principle, such as (d) and his diluted version of (a). I tend to think, by contrast, that an argument counts as an argument of principle because it cannot easily be overridden. In other words, we use the language of principle, as I think Ashworth himself does in practice, to assert that something is important (that it carries a lot of weight, or wholly or partly excludes some other considerations). It may be that, within the relevant class of important considerations, we restrict the title ‘principle’ to those that also happen to satisfy some other criteria, such as (a) or (d) or both. But there is no reason to think that satisfaction of those other criteria is what gives a consideration its importance. Its importance always calls for some independent defence.
(*) Professor of Jurisprudence, University of Oxford.
(1) I will be drawing on the following sample: A Ashworth, ‘Towards a Theory of Criminal Legislation’ (1989) 1 Crim LR 41 (hereafter ‘Towards a Theory’); A Ashworth, Principles of Criminal Law (3rd edn, Oxford: Oxford University Press, 1999) (hereafter Principles 3e); A Ashworth, ‘Testing Fidelity to Legal Values: Official Involvement and Criminal Justice’ (2000) 63 MLR 633 (hereafter ‘Testing Fidelity’); A Ashworth, ‘Is the Criminal Law a Lost Cause?’ (2000) 116 LQR 225 (hereafter ‘Lost Cause’); A Ashworth, ‘Is Restorative Justice the Way Forward for Criminal Justice’ (2001) 54 Current Legal Problems 347 (hereafter ‘Restorative Justice’); A Ashworth, ‘Criminal Justice Reform: Principles, Human Rights and Public Protection’  Crim LR 516 (hereafter ‘Criminal Justice Reform’); A Ashworth, ‘A Change of Normative Position: Determining the Contours of Culpability in Criminal Law’ (2008) 11 New Criminal Law Review 232 (hereafter ‘Normative Position’); A Ashworth, ‘Defending the Criminal Law: Reflections on the Changing Character of Crime, Procedure, and Sanctions’ (2008) 2 Criminal Law and Philosophy 21 (with Lucia Zedner) (hereafter ‘Defending the Criminal Law’); A Ashworth, Principles of Criminal Law (6th edn, Oxford: Oxford University Press, 2009) (hereafter Principles 6e). By referring to two editions of Principles of Criminal Law published ten years apart I am enlarging the stock of material on which we may draw to establish and confirm Ashworth's views. The book has been substantially rewritten over that period. Nevertheless, where I quote passages from Principles 3e that do not appear in Principles 6e, it is, I believe, only the formulations that have been abandoned and not the views that they represent.
(2) ‘Towards a Theory’, 41 and 54.
(3) ‘Lost Cause’, 253 and 255.
(4) eg ‘Testing Fidelity’, 635; ‘Lost Cause’, 245; Principles 6e, 45.
(5) eg ‘Towards a Theory’, 41; ‘Defending the Criminal Law’, 39.
(6) eg ‘Lost Cause’, 225; Principles 6e, 45.
(7) eg ‘Testing Fidelity’, 642; ‘Normative Position’, 236.
(8) eg Principles 3e, 49.
(9) Principles 3e, 59.
(10) ‘Lost Cause’, 256.
(11) Principles 3e, 59; ‘Criminal Justice Reform’, 528.
(12) A recurrent theme of Hannah Arendt's work, especially The Origins of Totalitarianism (New York: Harcourt Brace Jovanovich, 1973).
(13) L Fuller, The Morality of Law (revised edn, New Haven, CN: Yale University Press 1969), 154.
(14) See ‘Testing Fidelity’, 634–5, for remarks of Ashworth's that might be read as supporting either the ‘mitigate’ or the ‘militate against’ view.
(16) ‘There is nothing so bad or so good that you will not find Englishmen doing it; but you will never find an Englishman in the wrong’, says Napoleon in Shaw's The Man of Destiny. ‘He does everything on principle. He fights you on patriotic principles; he robs you on business principles; he enslaves you on imperial principles; he bullies you on manly principles; he supports his king on loyal principles, and cuts off his king's head on republican principles.’ GB Shaw, Plays Pleasant (Harmondsworth: Penguin, 1946), 205–6.
(17) ‘[Les gens à principes] sont, reprit Ménalque en riant, ce qu’il y a de plus détestable en ce monde. On ne saurait attendre d’eux aucune espèce de sincérité; car ils ne font jamais que ce que leurs principes on décrété qu’ils devaient faire, ou, sinon, ils regardent ce qu’ils font comme mal fait.’ A Gide, L’Immoraliste (Paris: Mercure de France, 1972), 116.
(18) Shaw and Gide do not make clear whether they are casting aspersions on all uses of principles, or whether there are some honest uses that survive the critique. A philosophical critic could echo the theme by explaining what it is about principles that makes them so open to dishonest use, even if they are also capable of having honest uses. I take some first steps in section 5 below.
(19) See Bernard Williams’ critique of principled moral thinking in ‘Persons, Character and Morality’ in A Rorty (ed), The Identities of Persons (Berkeley: University of California Press, 1976), and his critique of the ‘peculiar institution’ view of morality in his Ethics and the Limits of Philosophy (London: Fontana, 1985), Ch 10.
(20) ‘Towards a Theory’, 41.
(21) ‘Testing Fidelity’, 634.
(22) For a measured defence, avoiding common-law romanticism, see J Raz, ‘The Politics of the Rule of Law’ (1990) 3 Ratio Juris 331.
(23) Principles 6e, 166.
(24) Unless what is whittled away is a closure rule for resolving a legal conflict, eg a rule giving priority to statute over common law.
(25) ‘Towards a Theory’, 41; ‘Lost Cause’, 255; Principles 6e, 154; ‘Testing Fidelity’, 644.
(26) For more on this thesis see J Raz, Engaging Reason (Oxford: Oxford University Press, 1999), 219–25.
(27) See eg M Singer, ‘Moral Rules and Principles’ in A I Melden (ed), Essays in Moral Philosophy (Seattle: University of Washington Press, 1958), 169: ‘Moral principles hold in all circumstances and allow of no exceptions; they are invariant with respect to every moral judgment and every moral situation.’
(28) See notably J Dancy, Ethics Without Principles (Oxford: Oxford University Press, 2004), and many of the essays, largely inspired by Dancy's work, in B Hooker and M Little (eds), Moral Particularism (Oxford: Clarendon Press, 2000).
(29) ‘Towards a Theory’, 45.
(30) ‘Lost Cause’, 255.
(31) ‘Restorative Justice’, 369.
(32) ‘Lost Cause’, 245.
(33) ‘Lost Cause’, 255.
(34) cf Principles 3e, 59: principles are ‘strong arguments…rather than absolute precepts’.
(35) R Dworkin, Taking Rights Seriously (London: Duckworth, 1977), 24 and 26.
(36) In the early work that we are drawing on here, Dworkin seems to be committed to another thesis which would make it impossible for him to grant this. He seems to be committed to the thesis that, when a case is rightly decided, it does not change the law. Hence, in the (rightly decided) cases in which I would say that the court is resolving the conflict between two rules, Dworkin would say that there was no conflict to resolve. See R Dworkin, ‘No Right Answer?’ in PMS Hacker and J Raz (eds), Law, Morality, and Society (Oxford: Clarendon Press, 1977). I assume that Ashworth has no truck with this thesis.
(38) The language of exclusion, and the associated apparatus, is owed to J Raz. See his Practical Reason and Norms (London: Hutchinson, 1975), 35 ff.
(39) ‘Towards a Theory’, 53.
(40) On the ability to downgrade the weight of competing reasons as a type of exclusionary force, see S Perry, ‘Judicial Obligation, Precedent and the Common Law’ (1987) 7 OJLS 215, 222–3.
(41) ‘Defending the Criminal Law’, 46–7. See similarly ‘Lost Cause’, 11–12.
(42) Which leaves open the possibility that they can still ‘on occasion’ be outweighed even by ordinary reasons: Principles 3e, 59.
(43) ‘Towards a Theory’, 57.
(44) ‘Defending the Criminal Law’, 39–40.
(46) Ashworth sometimes refers to Dworkin's work on nearby questions (eg ‘Towards a Theory’, 44 and 45; Principles 6e, 24 and 35), but I have not found an instance where Ashworth names Dworkin as the source of the principle–policy contrast that he has in mind. Hence my note of hesitation here.
(47) Principles 3e, 58–9.
(48) Principles 6e, 45.
(49) cf Dworkin, n 35 above, 22–3, where it is noted that any action can be presented as contributing to the state of affairs in which that action has been performed. Dworkin's suggestion that this is a way of disguising a principle as a policy suggests that the distinction between the two drawn in the text above is the same one that he is advancing.
(51) See eg J Cameron and J Abouchar, ‘The Precautionary Principle: A Fundamental Principle of Law and Policy for the Protection of the Global Environment’, (1991) 14 Boston College International and Comparative Law Review 1.
(52) JS Mill, ‘On Liberty’ in JS Mill, Utilitarianism; On Liberty; Essay on Bentham (Warnock: Glasgow, 1962) 126, 136.
(53) Pope John Paul II, Centesimus Annus (Rome, 1991),  and .
(54) ‘Normative Position’, 237.
(55) See J Gardner, ‘Introduction’ in HLA Hart, Punishment and Responsibility (Oxford: Oxford University Press, 2008).
(57) Maybe even my training contributes constitutively to my fulfillment of the goal, depending on how you interpret the goal.
(58) I tend to think that A's policies are A's goals concerning how (constitutively or causally) A is to ϕ, which give A goal-dependent reasons to ϕ in that way, where the reasons to ϕ which militated in favour of adopting such policies may or may not themselves have been goal-dependent. If I am right about this, then one may have a policy for conforming to a principle. That seems to me to be an advantage of the account. I am not so sure, however, that Ashworth would find it an advantage. Sometimes he seems to expect policy to be founded on principle: eg ‘Lost Cause’, 242 (policy-making must be on principled grounds) and 255 (only principled derogation from principle is acceptable). On other occasions, however, it seems that he envisages a more exclusive contrast according to which policies are ‘founded on arguments about what is expedient rather than what is right in principle’: Principles 3e, 60.
(59) Principles 6e, 26.
(61) This is the main lesson of T Nagel, ‘Moral Luck’ (1976) 50 Proceedings of the Aristotelian Society Supplementary Volume 137.