This paper explores the roles that the presumption of innocence (PoI) can play beyond the criminal trial, in other dealings that citizens may have with the criminal law and its officials. It grounds the PoI in a wider notion of the civic trust that citizens owe each other, and that the state owes its citizens: by attending to the roles that citizens may find themselves playing in relation to the criminal law (such roles as suspect, defendant, convicted offender and ‘ex-offender’), we can see both how a PoI protects us, beyond the confines of the trial, against various kinds of coercion, and how that PoI is modified or qualified as we acquire certain roles. To develop and illustrate this argument, I pay particular attention to the roles of defendant (both during the trial and while awaiting trial) and of ‘ex-offender,’ and to the duties that such roles bring with them.

Antony Duff proposes a comprehensive concept of the presumption of innocence, covering the period before, during and after a criminal process, both in an official (state vs. individual) and a non-official, civic sense. By that broad usage, the concept of presumption of innocence is getting blurred and risks losing its contours. I therefore suggest to keep separate matters separate. The presumption of innocence in the narrow sense that I suggest applies only where there exists a suspicion that an individual has committed a criminal offence. The important function of the presumption of innocence in that situation is to prevent an over-extension of state power against the individual under suspicion before that suspicion has been confirmed to be true beyond a reasonable doubt. A general presumption that all people abide by the law at all times is neither warranted nor necessary. It is not warranted because experience tells us that many people break some laws sometimes. And it is not necessary because a system of civil liberties is sufficient to protect us against official or social overreach based on a suspicion that we may commit crimes.

In my response to Duff I focus mainly on the following two issues. Firstly, I examine what kind of a norm the presumption of innocence (PoI) really is and how it ontologically differs from other types of rules, principles, rationales, etc. My tentative conclusion is that a PoI does not suffice the requirement of being a dogmatic rule and, thus, has less weight than what Duff perhaps assumes.
Secondly, I examine what role the concept of innocence plays in the debate on fundamental (moral and legal) principles and the underlying rationales of a criminal law system. Although I am sympathetic to much of what Duff purports in his plea for civic trust and a parsimonious use of criminal law, I am reluctant to believe that it is really a broader version of a PoI that warrants the kind of morally decent criminal law system that he suggests normatively ought to be. In my view, most of what Duff wants to ascribe to the PoI can be derived from a principle of fairness which, in my view, is already embedded in the fundamentals of criminal law doctrine.

Duff sets out to present, not theoretical concepts, but ‘real’ principles that underlie positive law. This paper examines whether Duff’s analysis really reflects current law. To that end, this paper analyses the case law of the European Court on Human Rights. As far as his preposition that there are many presumptions of innocence is concerned, Duff seems to be right. In the case law of the European Court different presumptions can be discerned, with different rationales. However, these presumptions are a far cry from the trust principle Duff advocates. Indeed, a principle that prescribes trust cannot be found in the Court’s case law. There might be a unifying principle but if so this principle is about respect for human dignity rather than trust. This analysis serves as a basis for criticism. It is argued that the approach Duff proposes is in tension with the Court’s case law in several respects.

The presumption of innocence (PoI) is considered to be an important principle for regulating pre-trial detention. The idea is that pre-trial detention should be a last resort. However, pre-trial detention practice demonstrates that pre-trial detention does not function on the basis of a presumption of innocence but rather from a presumption of guilt and dangerousness. It must be concluded that, with regard to pre-trial detention, the PoI has a rather limited normative effect.

Factors justifying not presuming innocence are generally incorporated into the Presumption of Innocence (PoI). A confusing discourse has resulted: numerous guilt-presuming acts are deemed consistent with the PoI. I argue for an unusually broad PoI: any act that might convey to a reasonable actor that he is not presumed innocent of a punishable offence constitutes a PoI interference. Thus, academic debate need only be about the question what PoI interferences are justifiable or unjustifiable. This question must be answered using pro- and anti-PoI values. I analyse three PoI interferences in relation to Duff’s retributivist punishment theory: presumptions of guilt, vicarious liability of car owners and coercing non-suspects into proving their sobriety. Retributivists tend to castigate such procedures based on their (supposed) consequentialist rationale. I argue, however, that they might also be justified on retributivist grounds. The retributivist anti-PoI duty to punish the guilty may be the worst enemy of innocents.

In this response to my five critics, I note the strength of the arguments in favour of treating the presumption of innocence as a narrow, legal presumption that operates only within the criminal process; but I then try to make clearer my reasons for talking of different presumptions of innocence (moral, rather than legal, presumptions) outside the criminal process, in other contexts in which issues of criminal guilt or innocence arise – presumptions that guide or are expressed in the conduct of the state’s officials towards its citizens, and of citizens towards each other. Once we look at these other contexts in which criminal guilt and innocence (of past and future crimes) are at stake, we can see the importance of civic trust as a practical attitude that citizens owe to each other; and the fruitfulness of examining the various normative roles that citizens may have to play in relation to the criminal law.