Life After Prison
July 21, 2025
Limits Of The Criminal Law
July 22, 2025
Crime
Functions of Criminal Law
Few deny that one function of criminal law is to deliver justified punishment. Some go further and claim that this is the sole function of criminal law (Moore 1997, 28–29). Call this the punitive view.
Rules of criminal procedure and evidence, on this view, help facilitate the imposition of justified punishment, while keeping the risk of unjustified punishment within acceptable bounds. Rules of substantive criminal law help give potential offenders fair warning that they may be punished.
Both sets of rules combat objections we might otherwise make to laws that authorize the intentional imposition of harm. To combat objections, of course, is not itself to make a positive case for such laws.
That case, on the punitive view, is made by the justified punishments that criminal courts impose. This is not to say anything about what the justification of punishment is. It is merely to say that criminal law is to be justified in punitive terms.
Some object that this focus on punishment is misplaced. The central function criminal law fulfills in responding to crime, some say, is that of calling suspected offenders to account in criminal courts (Gardner 2007, 80; Duff 2010c, 16).
consequences.
As far as the law is concerned, criminal conduct is to be avoided. This is so whether or not we are willing to take the consequences.
It is possible to imagine a world in which the law gets its way—in which people uniformly refrain from criminal conduct. Obviously enough this is not the world in which we live. Imagine D is about to ϕ. If ϕing is a crime, reasonable force may permissibly be used to prevent Doing.
Police officers and private persons alike have powers to arrest D, and reasonable force may permissibly be used to make arrests effective.
This view puts the criminal trial at the centre, not just of criminal proceedings, but of criminal law (Duff 2013a, 196). Trials invite defendants to account for themselves either by denying the accusation that they offended, or by pleading a defence.
The prospect of conviction and punishment puts defendants under pressure to offer an adequate account. Call this the curial view. It differs from the punitive view in two ways. First, part of the positive case for criminal law is independent of the imposition of punishment.
Second, part of the positive case for imposing criminal punishment is dependent on the punishment being part of a process of calling to account. The following two paragraphs expand on both these claims.
part3
One might also claim that criminal law alone calls defendants to account. But punishments are imposed in civil proceedings—exemplary damages are the obvious case.
And it is arguable that civil proceedings also call defendants to account—that they too invite defendants to offer a denial or plead a defence; that they too use the prospect of legal liability to put defendants under pressure to account adequately (Duff 2014a).
In response, one might try to refine the function that is distinctive of criminal law. Perhaps criminal law’s function is to respond to public wrongs (whether by calling to account or punishing such wrongdoers), whereas the function of civil law is to respond to private wrongs.
What we should make of this proposal depends on what a public wrong is (Lamond 2007; Lee 2015; Edwards and Simester 2017). To make progress, we can distinguish between primary duties—like duties not to rape or rob—and secondary duties—like duties to answer, or suffer punishment, for rape or robbery. We incur duties of the latter kind by breaching duties of the former.
If the public/private distinction is cashed out in terms of primary duties, then responding to public wrongs cannot be distinctive of criminal law. Many wrongs are both crimes and torts. So the two bodies of law often respond to breaches of the same primary duty.
A more promising proposal looks to secondary duties. Perhaps criminal law’s distinctive function is to respond to wrongs on behalf of us all—to discharge secondary duties owed to the community (Duff 2011, 140). Perhaps the function of civil law is to respond to wrongs on behalf of some of us—to discharge secondary duties owed to individuals.
This might be thought to explain why criminal proceedings, unlike civil proceedings, are controlled by state officials: why officials can initiate proceedings that individual victims oppose, and discontinue proceedings that victims initiate.
The view described in the previous paragraph conceives of criminal law as an instrument of the community—a way of ensuring that the community gets what it is owed from wrongdoers. Call it the communitarian view.
If we combine this with the curial view, the distinctive function of criminal law is to seek answers owed to the community. One might doubt that the functions of criminal and civil law can be so neatly distinguished. It is arguable that civil law sometimes responds to wrongs on behalf of all of us—civil proceedings can be brought against
on the basis that her conduct is a nuisance to the public at large, or on the basis that D is a public official whose conduct is an abuse of power.
Lorem ipsum dolor sit amet, consectetur adipiscing elit. Ut elit tellus, luctus nec ullamcorper mattis, pulvinar dapibus leo.
Part4
Had the creation of those offences been an unqualified success, there would have been nothing for which to punish anyone.
One might hold that criminal law’s sole function is to prevent criminal wrongdoing. Call this the preventive view. Defenders of this view need not say that we should enact whatever laws will achieve the most prevention. That X
is the function of Y does not entail that we are justified in doing whatever will achieve most of X with Y
. That cutting is the function of knives does not entail that knife-holders are justified in cutting whatever they see. Holders of the preventive view can, in other words, accept the existence of constraints on prevention, that are not themselves justified in preventive terms (Hart 1968, 35–50). What they cannot accept is a positive case for criminal law that is not preventive.
Some hold a mixed view that combines elements of those considered above (Alexander and Ferzan 2009, 3–19; Simester and von Hirsch 2011, 3–18; Tadros 2016, 159–172). One way to construct such a view is by distinguishing between primary and secondary functions.
Primary functions are those that, when all else is equal, we have most reason to want the law to fulfil. Secondary functions are those we have reason to want the law to fulfil if it fails to fulfil its primary functions.
Criminal law’s primary functions, it is plausible to think, are preventive. Ceteris paribus, we have most reason to want criminal law to bring about a world in which wrongs like theft or murder do not occur.
Failing that, we have reason to want criminal law to call thieves and murderers to account, and to punish those who have no adequate account to offer.
There is some scepticism about mixed views. For some, the worries are conceptual. Moore claims that justified punishment must be imposed for reasons of desert, and that for this reason the punitive and preventive functions cannot be combined.
We are unable to ‘kill two birds with the proverbial one stone, for by aiming at one of the birds we will necessarily miss the other’ (Moore 1997, 28). Several replies are available. First, even if this is a problem for a mixed view of punishment, it need not be for a mixed view of criminal law.
Grant that punishment must be imposed for reasons of desert. It does not follow that criminal offences cannot be created for reasons of prevention. Criminalization and punishment are different acts and can be performed for different reasons (Edwards and Simester 2014). Second, to claim that X
is part of the positive case for criminal law—that it is one of criminal law’s functions—is not to claim that X
should be part of the mission of every criminal justice official (Gardner 2007, 202). Reasons that help make a positive case for our actions are often reasons for which we should not act. That one will be financially secure is a reason to get married, but one should not get married to be financially secure.
Similarly, to say that prevention helps make a positive case for criminal law—and for punishment—is not to say that judges should punish for that reason.
Other worries about mixed views are pragmatic (Duff 2010a). As criminal wrongdoing will persist whatever we do, the preventive function sets criminal law an insatiable goal. There is a standing risk that lawmakers who pursue that goal will deprive us of a criminal law that fulfills its other functions.
Consider again the curial view. Plausibly, we have reason to account for wrongs like theft and fraud in criminal court, but no reason to account for every interaction with property or all misleading statements from which we stand to gain. If defendants are to be called to account for the wrongs, it is these that must be criminalized.
To criminalize trivialities—in pursuit of preventive ends—is to drain criminal proceedings of their intrinsic value (Duff 2010b). No doubt these are important worries. But they do nothing to suggest that we should reject a mixed view. At most, they show that lawmakers also should not take prevention to be part of their mission.
As we already saw, this conclusion does not show that prevention is not part of the positive case for criminal law. And it may anyway be too strong. Lawmakers who exclude prevention from their mission may refuse to create crimes that would prevent a great deal of harm.
The cost of refusing to create these crimes might be greater than the cost of calling people to account for trivialities, and this might be so even when alternative means of prevention are factored in. If it would, criminal law’s preventive function should be part of the law-making mission: it is a function lawmakers should indeed aim to fulfil.
As to the first, we often have reason to account for our actions to others. We can leave open for now the precise conditions under which this is so. But it is plausible to think that if Alisha steals from Bintu, she has reason for account for the theft, and that if Chika intentionally kills Dawn she has reason to account for the killing.
Defenders of the curial view argue that criminal proceedings are of intrinsic value when defendants (are called to) offer accounts of themselves that they have reason to offer in criminal courts (Gardner 2007, 190–191; Duff 2010c, 15–17). Imagine Alisha stole from Bintu because she was under duress. Imagine Chika intentionally killed Dawn to defend herself or others.
Neither of these defendants, we can assume, is justifiably punished. On the punitive view, criminal law’s function does not stand to be fulfilled. On the curial view, things are different. Alisha and Chika both have reason to account for their behaviour—to explain what they did and why they did it.
Criminal proceedings invite each to provide that account and put each under pressure to do so. Assuming Alisha and Chika have reason to account in a criminal court, proceedings in which they (are called to) do so are of intrinsic value. One of criminal law’s functions is fulfilled even if no-one is, or should be, punished.
To endorse the curial view is not, of course, to say that we should do away with criminal punishment. But it is to say that the connection between trial and punishment is not merely instrumental.
Some think that the facts that make punishment fitting—say, culpable wrongdoing—obtain independently of criminal proceedings themselves (Moore 1997, 33). We use those proceedings to ensure that said facts are highly likely to have obtained—that D
is highly likely to have culpably committed a wrong. On the curial view, the fact that D has been tried and found guilty (or has entered a guilty plea) is itself part of what makes it fitting that D is punished.
The fitting way to respond to criminal wrongdoing, on this view, is to call the wrongdoer to account for her wrong. To call D to account is to attempt to both (a) get D to answer for wrongdoing (as occurs in court), and (b) get D to confront wrongdoing for which she has no satisfactory answer (as occurs when D is punished).
So it is only because D has first been tried and found guilty (or has entered a guilty plea) that punishment counts as a fitting response to D’s wrong (Gardner 2007, 80; Duff 2013a, 205). We can see the implications of this view by imagining a world in which trials are abolished, because some new-fangled machine allows us to identify culpable wrongdoers with perfect accuracy. Having no doubt that D is guilty, we simply impose punishment on D
. On the curial view, the punishments we impose are inherently defective: they are not imposed as part of a process of calling to account. Though our new-fangled machine might justify doing away with trials—once we factor in how expensive they can be—we would lose something of value in doing away with them.
If criminal law does have a particular function, we can ask whether that function is distinctive of criminal law. We can ask, in other words, whether it helps distinguish
criminal law from the rest of the legal system. It has been claimed that criminal law is distinctive in imposing punishment (Moore 1997, 18–30; Husak 2008, 72).
Part3
More importantly, one might claim that in the case of paradigmatic crimes—like robbery, rape, or battery—criminal law responds to wrongs on behalf of individuals—on behalf of those who have been robbed, raped, or battered.
On this view, a positive case for criminalization need not await the finding that D owes something to the whole community. It is at least sometimes enough that D owes something to those D has wronged, which D would fail to provide in the absence of criminal proceedings.
Those who reject the communitarian view might be thought to face the following difficulty: they might be thought to lack an explanation of official control over how far criminal proceedings go.
If criminal law seeks what is owed by wrongdoers to the wronged, doesn’t official control amount to theft of a conflict properly controlled by the two parties (Christie 1977)? Not necessarily.
First, we should not always require the wronged to have to pursue those who have wronged them. Second, we should not always support those who think themselves wronged in pursuing alleged wrongdoers.
As to the first point, some are trapped in abusive relationships with those who wrong them. Others are susceptible to manipulation that serves to silence their complaints. Some wrongdoers can use wealth and social status to stop accusers in their tracks.
As to the second point, the temptation to retaliate in the face of wrongdoing is often great. It is all too easy for the pursuit of justice to become the pursuit of revenge, and for the perceived urgency of the pursuit to generate false accusations.
Official control can help vulnerable individuals—like those described above—to get what they are owed. And it can mitigate the damage done by those trying to exact vengeance and settle scores (Gardner 2007, 214–216).
It can ensure that those in positions of power cannot wrong others with impunity, and reduce the likelihood that vindictiveness begets retaliation, which begets violent conflict from which all lose out (Wellman 2005, 8–10).
We can add that criminal proceedings may help protect others against being wronged in future. Those wronged may have a duty to give up control of proceedings to provide this protection (Tadros 2011c, 297–299).
These remarks suggest an alternative to the communitarian view. According to the alternative, the secondary duties of concern in civil and criminal proceedings are typically one and the same.
The positive case for criminal law’s involvement is not that it discharges duties of interest to the criminal law alone, but that it enables duties of general interest to be discharged less imperfectly than they otherwise would be—than they would be if the criminal law took no interest in them.
Call this the imperfectionist view. What is distinctive of criminal law, on this view, is not its function but its mode of functioning: the way it fulfills functions shared with other bodies of law.
Some writers seek criminal law’s distinctiveness in a different place. What is distinctive about criminal law, they claim, is that it publicly censures or condemns. This expressive function is sometimes associated with criminal punishment (Husak 2008, 92–95).
Because other bodies of law sometimes punish, and because punishment typically—perhaps necessarily—expresses censure (Feinberg 1970), the expressive function is at least partly shared.
But the message sent by criminal law is not sent only at the sentencing stage. It is sent the moment a guilty verdict is reached by a criminal court—by the declaration that D
has been criminally convicted (Simester 2005, 33–36). The social significance of conviction is very different to that of (say) the verdict that D is a tortfeasor: the former verdict conveys, in and of itself, that D’s conduct reflects badly on D
. Though additional detail may generate the same conclusion in the case of a civil verdict, such detail is not required in the case of criminal conviction. If this is right, the distinctiveness of criminal law turns out not to consist in the fact that it provides for punishment.
It turns out to consist (at least in part) in the provision of a technique for condemning wrongdoers which does not require that we punish to condemn.
So far, we have focused on the functions criminal law fulfills in response to the commission of crime. It is plausible to think, however, that criminal law’s functions include preventing crime from occurring.
We can see this by asking what success would look like for the criminal law. Would criminal law have succeeded if all thieves and murderers were tried and punished? Or would it have succeeded if there was no theft or murder, because criminalization resulted in would-be thieves and murderers refraining from such wrongs?
Notice that to pose these two questions as alternatives is not to deny that punishment might be justified in preventive terms. It is rather to suggest that resorting to punishment to achieve prevention is already a partial failure for the criminal law. It is a failure to deter those who, ex hypothesis, have already committed criminal offences.
Part5
Criminal Responsibility
Imagine that D takes V’s property without V’s consent. Is D criminally responsible for the taking? Not necessarily. In English law, D commits the offence of theft only if D acts dishonestly, and intends for V
to be deprived of her property permanently. Theft is one of many offences commission of which depends on one’s state of mind. Elements of offences that require particular mental states are known as men’s rea elements. Other elements are known as actus reus elements.
Some claim that if D
satisfies all elements of an offence—if D commits the actus reus with men’s rea—this suffices to make D criminally responsible, but not to make D criminally liable. Responsibility is understood here as answerability (Duff 2007, 19–36). While we are answerable to the courts for committing offences, we may avoid liability by offering satisfactory answers in the form of defences.
This account of criminal responsibility—call it the answerability account—relies on a distinction between offence and defence to which we will return. One argument for the answerability account invokes rules of criminal procedure and evidence.
To obtain a conviction, the prosecution must prove that D committed the offence beyond a reasonable doubt. D can safely remain silent in the absence of such proof. If the prosecution makes its case, D has strong prudential reasons to prove a defence: without one, D will be convicted and punished.
The best explanation of these rules, so the argument goes, is that offending acts generate a duty to answer that is otherwise absent. This duty explains why, when we have strong reason to believe D to be an offender, we put D
under extra rational pressure to explain herself in court.
Some think that, on closer inspection, our rules of procedure and evidence fail to support the answerability account and help to undermine it. Those rules tell prosecutors to consider evidence both that D
committed an offence and that D lacks a defence. If there is strong evidence that D killed in self-defence, D should not be prosecuted. This matters here for the following reason. When prosecutors decide whether to prosecute, they are deciding whether D should be called to answer for what D did.
The fact that prosecutors should not prosecute if D clearly killed in self-defence, suggests that those who have defences are not answerable in court. It suggests that we owe the criminal courts answers not for acts that are offences but for acts that are crimes—for offending acts which do not satisfy an available defence.
Obviously enough, it is for crimes that we are criminally liable. If responsibility is answerability, and we are answerable for crimes, the conditions of criminal responsibility and the conditions of criminal liability are one and the same. The answerability account, as described above, then fails (Duarte d’Almeida 2015, 239–267).
