Criminalisation as a Speech-Act: Saying Through Criminalising

Abstract

The act of criminalising conduct has been understood by many theorists as a form of communication. This paper proposes a model, based on speech-act theory, for understanding how that act of communication works. In particular, it focuses on analysing how and where wrongfulness can appear in this speech-act, if one were to argue, as many theorists do, that part of what is being communicated through criminalisation is the wrongfulness of the target conduct. I argue that the act of criminalisation is best understood as an indirect speech-act, which both asserts and declares normative facts, the utterance of which makes it the case that a conduct is now criminalised. Within this speech-act, wrongfulness can appear as an implicit assertion of the wrongness of the conduct, which has to be inferred by hearers from the context of utterance. The paper then briefly discusses the upshots of this model, mainly that it allows a clearer picture of how criminalisation conveys meanings, as well as leaving open the question as to whether it makes sense to think that the wrongfulness being conveyed is specifically of a moral kind.

1 Introduction

It seems to be a consensus among criminal law theory scholars that when a conduct is formally criminalised,¹ some form of communication is happening in doing so. Simester and Von Hirsch propose that criminalisation “speaks directly to subject-citizens”²; Tadros believes that it “expresses to citizens that they are under obligations” and that the criminal law “communicates moral duties to citizens”³; Duff states that the criminal law “should be seen as a communicative enterprise”.⁴ It is not surprising that theorists propose this kind of analysis of the criminal law and particularly for criminalisation, since creating a criminal offence—and thereby bringing the criminalised conduct into the purview of the criminal law—seems to carry with it many potential meanings, which makes sense to see as being presented through communication or expression—call these “communicative theories” of criminalisation.⁵ For some theorists, what is being communicated is best explained as a conveying of the moral wrongness of the conduct being criminalised, for which those who commit those crimes must answer for, such as certain strands of legal moralism in criminalisation theory.⁶ Though arguments have been presented as to why one might think moral wrongness is part of that communication,⁷ an important question that has not been fully answered by criminalisation theorists is how this communication actually happens.

This paper seeks to provide a theoretical model to explain how this communication works, and how the meanings that are carried by criminalising a conduct manifest. In doing so, it is intended to provide analytic tools for both criminalisation theorists that work on understanding or explaining the act of criminalisation, as well as those who seek to provide normative theories that justify (or not) criminalisation as a practice. To do this, I will propose to view the act of criminalisation as the performance of a speech-act which I will call a criminalisation claim. The use of speech-act theory has long been present in the philosophy of law,⁸ but it has not been specifically discussed in the context of criminalisation theory. Though it may seem intuitively correct that something is being said when a conduct is criminalised, how that gets to happen has not been clearly defined by theorists up to this point. One might think that this is because the alluded intuition can be taken for granted, since theorists could simply unpack that intuition without needing to allude to any understanding of how that communication might happen.⁹ Aside from the analytic incompleteness that this would entail, there are two important benefits to having a clear theoretical model for how communication happens by criminalisation, especially for those theorists who want to argue that wrongness has something to do with that communication. First, because it allows a theorist to specifically show where, if at all, wrongness appears in the act of criminalising. In having this speech-act model, they can do so not by having to point to the conduct being criminalised itself and trying to argue that its presence is what signals wrongfulness in the act of criminalising it, but instead by showing how wrongness is part of what the act of criminalising entails as an act of speech—that we are being told that the conduct is wrongful.

Second, by not showing how and where wrongness (of any kind) appears in the act of criminalising, normative theorists who propose that wrongness plays an important role in justifying the act of criminalisation leave open many questions that are relevant to our metaethical understanding of what is entailed in claiming that wrongness has a role to play, and there are good reasons to not leave these questions open.¹⁰ In particular, if one is to believe that the act of criminalisation is an instance of moral talk (since the conduct criminalised is a moral wrong, for example), then it is important for theorists to be clear as to what are the features of that moral talk, and part of doing that requires clarity as to how that kind of talk manifests in the act of criminalising. From there, theorists can then evaluate whether moral talk is the kind of talk that criminalisation is apt for, or if in fact it is more accurate to describe criminalisation as something different. As will become clearer below, this is especially important if instances of moral talk in criminalisation are done implicitly, requiring an inference from context of both the presence of wrongness and its specific nature as a moral wrong.

The paper begins by stating the basic thesis that I wish to defend: that criminalisation as a speech-act is best seen as an indirect speech-act, that includes both asserted and declared normative facts. I then provide some of the basic terminology and speech-act theory to understand what the basic thesis entails—readers who are already familiar with speech-act theory and John Searle’s work might move directly to Sect. 3. I then make an important distinction between wrongfulness in a broad and a narrow sense, and discuss why it makes sense to think that wrongfulness is involved in criminalisation in the first place. With that groundwork set, in Sect. 4 I focus on what normative facts are being declared—i.e. brought into existence through the speech-act—in criminalisation claims. I identify four declared facts, which I call the labelling fact, the liability fact, the obligation fact, and the responsibility fact. I discuss each in detail and show how these facts fit into some existing criminalisation theories.

In Sect. 5, the paper shifts its attention to wrongfulness more specifically, and I argue that wrongfulness in a broad sense appears in criminalisation claims as an implicit assertion—i.e. the report of a pre-existing fact—and, as such, that it has to be inferred by hearers from the context of utterance. The section ends with some discussion of some of the challenges that this implicit nature of the assertion entails. The paper concludes by showing some of the upshots of using this theoretical model for criminalisation, making it easier for criminalisation theorists to identify the targets for their arguments.

2 The Basic Thesis and Speech-Act Theory Terminology

So, what are the contents of a criminalisation claim? To answer this, I will need to explain some basic concepts of speech-act theory, but here is the basic thesis that I will be defending: a criminalisation claim is an indirect speech-act which both asserts and declares normative facts¹¹ and, through the utterance of the speech-act, a target conduct becomes a criminal offence. I will explain the terminology used within that basic thesis and then look at the actual contents of criminalisation claims.

First, a quick explanation of what ‘illocution’ means is necessary, since understanding the rest of the terms in the basic thesis requires knowing this. The term comes originally from Austin,¹² and refers to what is being done in saying something (i.e. the speech-act itself) rather than just the literal or linguistic meaning of the words being used (in Austin’s terms, locutions). Some classic examples of illocutions are common verbs like requesting, questioning, bargaining, promising, commanding, and so on—if I say the phrase “where is the library?” the illocution is a question, which is being performed by uttering the phrase “where is the library?”.

Next, indirect speech-acts are, as Searle puts it, when “one kind of illocutionary act can be uttered to perform, in addition, another type of illocutionary act.”¹³ For example, saying ‘could you pass the salt, please?’ is a good example of this kind of speech-act (and is based on a similar example by Searle).¹⁴ One meaning of the speech-act is a mere question: the speaker is asking whether the hearer is indeed capable of passing the salt. But it seems quite natural to understand the meaning of this utterance as not just a question, but a request: the speaker is asking the hearer to pass the salt to them. The utterance of ‘could you pass the salt, please?’ also carries with it a kind of ‘marker’ of the nature of the request: by asking ‘please’, the speaker lets the hearer know that they are trying to follow the norms of etiquette by asking politely. As you can see, all these meanings are happening at the same time as the utterance ‘could you pass the salt, please?’. You will also notice that these additional meanings are not literally uttered in the speech-act—they are obtained from what Searle calls “mutually shared background information”¹⁵ and from the capacity of the hearer to make inferences based on that background information. I will use an example based on Searle’s work¹⁶ to illustrate this point.

Suppose that we have two people, Amy and Beatrice, that have the following exchange:

• (1) Amy: Let’s go to the cinema tonight.

• (2) Beatrice: I have to study for an exam.

(2) is an indirect speech-act performed by Beatrice because, using Searle’s terminology, the primary illocutionary act performed by Beatrice is a rejection of an invitation, and the secondary illocutionary act is an assertion of the fact that Beatrice has an exam that she has to study for.¹⁷ Notice that the secondary illocution Beatrice performs is done in the act of saying what she literally says—she says that she has to study for an exam. The primary illocution, however, is not literal in the same way, because nothing in the words that Beatrice uses directly states that she is rejecting Amy’s invitation. In fact, she could use the same words in a sentence that cancels out the primary illocution of rejecting an invitation if she added more words, such as: “I have to study for an exam, but I’ll do it when we get back from the cinema”. And yet, it seems quite plausible that Amy (and us) can infer, if Beatrice says nothing more than what she says at (2) as a response to (1), that she is rejecting the invitation to the cinema. How does this happen? Searle tells us that a hearer of (2) needs to have an inferential strategy in order to realise that an indirect speech-act is happening, and he describes it thusly: “the inferential strategy is to establish, first, that the primary illocutionary point departs from the literal, and, second, what the primary illocutionary point is.”¹⁸ In addition to this inferential strategy, a hearer will also need “a device for finding out what the ulterior illocutionary point is”,¹⁹ which is “derived from the theory of speech acts together with background information.”²⁰ Thus, for the inference to work, a hearer needs to be able to identify a discrepancy between the literal meaning of the utterance and the potential illocution being performed in the speech-act. Once this is done, they can move on to determine what that illocutionary point is, based on their understanding of the background information that applies to the communication taking place.

Lastly, with regards to speech-act theory, Searle proposes a taxonomy of categories for possible illocutionary acts,²¹ where he proposes five kinds of illocutions: assertives, directives, commissives, expressives and declaratives (sometimes called ‘exercitives’ but I will use Searle’s original terminology for simplicity). Of these five, I wish to focus on two: assertives and declaratives. Assertives are illocutions that commit the speaker to something being the case (i.e. the truth value of what is being said), whereas declaratives are illocutions the successful performance of which guarantees that the propositional content of what is being said corresponds to the world. In other words, assertions are reports of states of affairs which the speaker believes to be true, whereas declarations are illocutions which create new states of affairs through their performance. For example, when a meteorologist utters the phrase “it is raining outside”, it is an assertion—it is a report of how things actually are and, if the speaker is being sincere (i.e. they are not knowingly saying something false), they actually believe the report to be true (which we can corroborate by looking outside, so the assertion is also falsifiable). In contrast, when the city official utters the phrase “I hereby pronounce you husband and wife”, they are making a declaration—by uttering that phrase, they are inserting a new fact into the world, and if they are successful (i.e. they have the required authority to do so) then that fact comes to be true in the act of uttering it. There is a potential further sub-division of declaratives, proposed by Bach and Harnish,²² of ‘effectives’ and ‘verdictives’, where the former are utterances that when issued by the right person under the right circumstances produce institutional states of affairs, whereas the latter are reports of something being the case and, in doing so, make that the official line.²³ However, I will prefer Searle’s terminology since I wish to remain agnostic as to which specific type of declarative illocutions are happening in a criminalisation claim—choosing between effectives and verdictives may be based on metaethical assumptions of the kind of normative facts that are being declared in criminalisation claims, so I will not impose these assumptions on the framework I am proposing here.

3 Two Senses of 'Wrongness' and Wrongness Within Criminalisation

If one believes that wrongness is part of what is being said in the act of criminalisation, seeing said act as an indirect speech-act makes sense when we actually look at how criminal offences are worded and then compare that to how many theorists understand what is happening when a conduct (Φ) is criminalised. For example, if a theorist wants to suggest that criminal offences convey a claim about the wrongfulness of Φ, then the only option for this to be the case is to understand criminal offences as indirect speech-acts, since there is no offence that literally says anything like “it is morally wrong to Φ”, “citizens shall not Φ” or “it is prohibited to Φ”. For example, s. 47(1) of the Criminal Law (Consolidation) (Scotland) Act 1995 states that “Any person who has with him in any public place any offensive weapon shall be guilty of an offence”—nowhere in the statute does it say explicitly that the relevant conduct is wrongful. It only states that performing the conduct makes one ‘guilty of an offence’ and then sets a liability for doing so.²⁴ One might think this is the case because carrying an offensive weapon is a mala prohibita offence, but offences traditionally seen as mala in se are also expressed in a similar fashion. The offence of rape is set out in the Sexual Offences (Scotland) Act 2009, s.1, which states:

“(1) If a person (“A”), with A’s penis-

• (a) without another person (“B”) consenting, and

• (b) without any reasonable belief that B consents,

penetrates to any extent, either intending to do so or reckless as to whether there is penetration, the vagina, anus or mouth of B then A commits an offence, to be known as the offence of rape.”

Once again, there is no explicit signalling of the wrongfulness of this conduct in the wording of the offence, nor of the fact that the conduct is prohibited—it describes the relevant conduct and assigns to it the label of an offence, specifically the label of rape. And yet, it does not seem implausible to believe that the wrongfulness of the conduct at play is being conveyed through criminalising it—there is something in the fact that it is being criminalised that lets us know we are being told, through the law, that the criminalised conduct is wrongful. I argue that the best way of understanding this is that it is an additional illocution that is being performed at the same time as the utterance of the act which criminalises Φing.

Additionally, there are two potential senses in which one can read ‘wrongness’ or ‘wrongfulness’ in the context of this debate, and more broadly in any kind of normative debate. The first sense is a narrow understanding of ‘wrong’ as referring to that which is morally wrong, i.e. that which is wrong according to morality. The second sense is a broader understanding of ‘wrong’ as a more general normative concept—something which is wrong with regards to some kind of normative standard, as in the wrong move in a game, the wrong choice for obtaining a particular goal, the wrong gesture or action according to etiquette, and so on. When I speak of ‘wrongs’ or ‘wrongness’ simpliciter, I will be referring to this latter sense of wrongs, and I will explicitly say ‘moral wrongs’ or ‘moral wrongness’ when I mean the narrower sense of wrongfulness.

I will be arguing that wrongfulness in a broad sense plays an important role in a criminalisation claim, but whether wrongfulness in a narrow sense does so is an open question. That being said, one could disagree with this entire approach to criminalisation in the first place—why think that wrongfulness (broad or narrow) has any role to play in criminalisation? In other words, why must our theorising about criminal law, including criminalisation, deal with wrongs at all? Answering this question is precisely where the model I will be proposing is useful.

I am willing to grant the point of the question, however, with regards to wrongfulness in a narrow sense. In fact, it is a question that underlies criminalisation theory more broadly—one needs to explain why morality is the appropriate or relevant normative domain for criminalisation, and to do so two prior questions need to be addressed: what does ‘morality’ mean in this context and why is it relevant to the criminal law. One possible answer is to point to the fact that condemnation and punishment are part of the central functions of criminal law, and that the only way to justify these practices is by using moral wrongfulness. Note, however, that this is a slightly different question to the one I posed earlier—it is not on why we might think the criminal law deals with wrongs, but specifically it is about the relevance of morality as the right normative domain from which to obtain the wrongs that criminalisation purportedly deals with.

With regards to wrongfulness more broadly, however, the question becomes more problematic. That is, believing that any kind of wrongfulness has no role to play within criminalisation entails that, for example, criminal offences do not include a wrong in any sense. This means that the conduct or action included within a criminal offence would not be able to be considered a kind of shortcoming with regards to some kind of normative standard—it would not be an action that can elicit appropriately a response of ‘that’s not the right thing to do’ in any sense. In other words, it would require us to accept that including a conduct into a criminal offence is not, in any important sense, normative in the first place—creating a criminal offence would not involve stating a norm about the conduct in question.²⁵ This is because, as Gerritsen puts it, norms are “standards, rules, principles or conventions that involve correctness conditions, rather than statistical normalities. Norms involve correctness conditions because they categorise our behaviour and attitudes as correct or incorrect, depending on whether they are in accordance with the norm in question.”²⁶ Accepting that criminal offences do not include the statement of some kind of norm about the conduct being criminalised seems counter-intuitive, as it would lead to accepting this kind of modus tollens argument:

• (1) If criminal offences include conduct that are wrongs, then criminal offences include a statement of norms (i.e. standards including correctness conditions for behaviour) about the conduct.

• (2) It is not the case that criminal offences include a statement of norms about the conduct.

  Therefore, (from (1) and (2)),

• (3) It is not the case that criminal offences include conduct that are wrongs.

Premise (1) is a reflection of what I have explained above: the presence of a concept of wrongness necessitates there being a norm involved for there to be something wrong in the first place. Premise (2) does not, however, seem plausibly true. This is because norms establish favouring or disfavouring relations—based on correctness conditions—between an action (in this case, the criminalised conduct) and a fact that counts in favour of or against the action. If this relation seems like a good explanation for what is going on with a criminal offence—that is, making a conduct count as an offence is best explained as establishing a normative relation between the conduct and a fact that favours or disfavours its performance—then criminal offences must be (or include) some kind of norm, and therefore at least some form of wrongness needs to be present in the fact that it is being criminalised (albeit, not necessarily of a moral kind). So, at the very least, if a criminal offence can be seen as including some kind of norm about the conduct in question, then at least some form of wrongness will necessarily be involved—wrongness in the sense of incorrectness with regards to being in accordance with the norm in question. And, on the contrary, if we were to argue that no form of wrongfulness is involved, then criminal offences could not be seen as including norms of some kind about the criminalised conduct, moral or otherwise. Thus, it does make sense to think of wrongfulness in a broad sense as an appropriate conceptual tool to understand how criminalisation manifests, and particularly as part of what is being said through a criminalisation claim.

Now that I have laid this groundwork, we can now better understand both the basic thesis I proposed earlier, and can now move on to look at the contents of a criminalisation claim more closely.

4 Declared Normative Facts

As I said above, a criminalisation claim is an indirect speech-act that both asserts and declares normative facts, and makes it the case that a conduct becomes a criminal offence. Since I am arguing that it is an indirect speech-act, it has both a primary and a secondary illocutionary act happening at the same time, and the primary illocution is not going to be explicitly stated in the speech-act. What is stated, however, are the effects of a conduct becoming a criminal offence.²⁷ These are, I argue, the normative facts that are being declared through the speech-act, in Searle’s terms. Importantly, it is from these facts that are being declared that a hearer would need to be able to tell, from the speech-act and the background information or context of utterance, that an additional, primary illocution is happening at the same time.

As for what is being declared by a criminalisation claim—what new facts are being created in criminalising Φ—I propose that there is one descriptive fact and three legally relevant normative facts. The first and more obvious descriptive one I call the labelling fact, which is the assigning to a particular legal category (i.e. a particular criminal offence) of Φing. We declare that the factual elements that compose Φ (i.e. the elements of the offence) are hereby labelled as the particular criminal offence that is being established by the criminalisation claim. So, for example, if Φ were an attack on another person with some relevant intention (depending on the jurisdiction in question), by criminalising Φ we are declaring the assigning of the label “assault” to the elements which constitute Φing. Sometimes, however, this labelling fact is not necessarily present in the act of criminalisation, in the sense that the offence being created is not given a particular name such as ‘assault’ or ‘culpable homicide’, but is rather ex post named simply by its location in legislation—‘the offence under s. 47(1) of the Criminal Law (Consolidation) (Scotland) Act’, to repeat a previous example.

Additionally to the labelling fact, there are three other new normative facts being declared. I will call them the obligation not to Φ, the liability to punishment for Φing, and the responsibility for Φing. The first normative fact I will explain is the new liability when someone Φs, specifically linked to a breach in the previously declared obligation not to Φ. By criminalising Φ, we are letting people know that it is hereby declared that whoever Φs has breached a legal obligation not to Φ and, therefore, is liable to face a particular consequence. Normally, this consequence would be some form of criminal punishment, and though this may not necessarily need to be something like imprisonment, it will still be condemnatory in nature.²⁸ Again, this liability did not exist, as a matter of fact, until it was declared by the act of criminalising Φ—there may have been other kinds of consequences associated to Φing, but not the ones that result from the criminal law. This new liability is particularly important for those who attribute the distinctiveness of criminalisation to the kind of consequence people might face for performing the criminalised conduct, specifically when it comes to criminal punishment—in spite of any kind of other sanctions that may be associated to Φing (pre-legal or not), the act of criminalising creates a new, specific liability for the kind of sanction that is only available to criminal offences. Husak, for example, states that “the most basic questions to be answered by a theory of criminalization is: for what conduct may the state subject persons to punishment?”²⁹ Moore also considers that a theory of punishment goes hand in hand with a theory of criminalisation,³⁰ so he would also be interested in this liability.

Any search for justification for criminalisation that is based on the kind of sanction that is applied to offenders will be interested in what kind of liability is being declared and what kind of consequence people will be liable for. As stated earlier, this liability also includes the consequence of being convicted for the offence in question, which can have its own distinct consequences that are different from mere punishment—having a criminal record, going on particular special registries (like some sexual offenders), as well as all the potential stigmas that are associated with being convicted of a crime independent of something like a prison sentence.³¹ So, not only is there a liability for the punishment itself, but also for all that comes with the conviction that may lead to that punishment.

The second normative fact is the fact that criminalising Φ creates a new legal obligation with regards to Φing. Specifically, it creates an obligation not to perform Φ in view of the fact that it is now a criminal offence. This normative fact does not need to be stated explicitly in the secondary illocution of a criminalisation claim, but it is possible to view it as a part of the speech-act because of creating a liability for punishment for Φing. In this sense, both of these normative facts are connected to each other, and happen simultaneously—creating a liability to punishment entails creating an obligation not to perform the act for which one is liable to be punished. This is not an evaluative claim, in the sense of saying that the act is being punished because it is wrong not to follow the obligation being created, but rather it is a claim of normative entailment for the liability to punishment to make sense. That is, since the speech-act is creating a legal liability to a consequence (like punishment), there needs to be a correlative obligation to which the liability is responding. Hence, in declaring a liability to punishment, the speech-act is simultaneously declaring an obligation.

Notice that this declaration does not mean that this new, created obligation is the only obligation that exists normatively speaking with regards to Φing. However, it does mean that criminalising Φ actually creates a new obligation, a legally relevant one that is officially integrated into the criminal law of a jurisdiction and declares that, from now on, citizens are under a legal obligation not to Φ. In essence, this new fact by which a legal obligation is created establishes a new reason not to Φ, which can be added into a decision-making process regarding Φing along with any other pre-existing reasons that one may identify with regards to Φing, but there now is an additional reason being created simply by the fact that Φing has been criminalised. The obligation is specifically in the form of a prohibition—this is precisely what Simester and Von Hirsch are referring to with the idea of an ‘instruction’ given through criminalising.³² The normative force of the ‘should-not’ that said instruction or prohibition carries with it is in a way reinforced by creating new normative facts about the performance of Φ, thereby making it part of the normative domain of the law. An obligation to not-Φ is created—it is declared that from now on, there is hereby an obligation, relevant to the criminal law, not to Φ.

Tadros has a related view with regards to what is being done when criminalising, particularly for what he calls the “core cases of criminalisation” in which “the criminal law expresses to citizens that they are under obligations. […] it demands that citizens must not perform the act prohibited by the criminal law because those acts are seriously wrongful. The criminal law thus communicates moral duties to citizens.”³³ Tadros’ view, however, is not clear on whether these obligations are, in the terms used here, asserted or declared. On the one hand, he accepts that there might be prior moral demands on the conduct being criminalised, so the ‘communication’ of moral duties might be a kind of mere assertive reinforcement of that putative previous duty. On the other hand, it seems that if the expression happening when criminalising a conduct is a demand on citizens not to perform Φ, that expression is adding something new to the normative status of Φing, rather than just reminding hearers of a prior moral duty. Note that saying that there is an obligation not to Φ because it is morally wrongful does not necessarily imply that the obligation is a pre-existing moral one—moral wrongfulness can be acting as a justifying reason for the creation of a new kind of obligation (like a legal one), which seems more in line with what Tadros is proposing.

This distinction is important because it entails two different views of what the law does when using moral wrongfulness in the context of criminalisation. One option is that the law is mirroring morality and simply reporting back what is purported to be the moral matter of fact with regards to both the moral properties of the conduct and the obligations that are generated from that fact. The other option is that the law is only using the fact that moral wrongfulness is a property of the conduct in question as a basis for creating new normative facts, which can function independently of the moral obligations one might purport to associate with the moral wrongfulness of the conduct.

Duff sees criminalisation more like the first option mentioned above, because to him a criminalisation claim is ‘declaring’ (or asserting, in the language I have been using thus far) rather than ‘prohibiting’ the conduct in question, in view of the fact that “its role is not to make wrong what was not already wrong, but to declare that these pre-legal wrongs are public wrongs”.³⁴ In this sense, criminalisation claims are not prohibitions because the authority by which we refrain from Φing is derived from the pre-legal moral wrongness (and thereby the moral duties which that would entail) rather than from some kind of respect for what the law says as an authoritative restraint on conduct.³⁵

Perhaps a useful way to put it is to see the assertion/declaration distinction as the difference between what is being said and what is being done through a criminalisation claim.³⁶ For what is being said, Duff’s position seems correct if we assume that the wrongfulness in question is pre-legal: the wrongfulness of the conduct being criminalised is not derived from the fact that it is prohibited, but instead we are asserting (‘declaring’ for Duff) its wrongfulness through criminalising the conduct—recognising the fact that it is wrong and, therefore, worthy of being included as a criminal offence. For what is being done, however, Duff’s position needs to be adjusted: prior to the criminalisation of the conduct, there may have been moral (pre-legal) duties and liabilities towards Φing, which might lead us to morally judge people for Φing as wrongdoers and as worthy of moral blaming, and may have given moral reasons to refrain from the conduct in question. But there were no legal duties nor liabilities with regards to Φing qua criminal offence up to that point, so there necessarily must be some kind of new normative fact—creating its own new sets of reasons, duties and obligations—being established through its criminalisation. And these normative facts are, necessarily, created by the act of criminalising Φ—any prior normative (in this case, moral) consequences that Φing may have generated may be relevant, but are not identical to, the normative consequences which are entailed by criminalising Φ. It is the creation of these normative facts which allows for us to separate the legal consequences for Φing from the moral consequences for Φing (though, granted, they are not necessarily entirely separate, but they are, at the very least, distinct from each other). And it is in this sense that I argue that an obligation not to Φ is declared—it is created as a normative fact within the context of the law.

Finally, the fourth normative fact being declared is a new responsibility with regards to Φing. By responsibility, I mean a requirement of answerability³⁷—if someone Φs, they must answer with some kind of explanation or response (which, in the context of the criminal process, may be to remain silent) for the fact that they have Φed, and that they will have to do so within the context of the criminal process. As Duff puts it, “I am responsible for that for which I must answer, and I must answer for that which there was reason for me not to do”.³⁸ These reasons are precisely the ones being declared by the act of criminalising Φ—the obligation not to Φ and the liability for breaching that obligation—and it is in virtue of these reasons that anyone who is found Φing can be brought to answer for doing so in a criminal process. Again, this declared normative fact does not need to be explicitly stated in the speech-act, but it is also a constitutive part of it as a result of being declared in the context of the criminal law. That is, since we already have other rules that confer powers to legal officials to bring in people who are suspected of committing a criminal offence to answer, and this answering must be done within the criminal process, then in the act of making a conduct count as a criminal offence we are simultaneously declaring that anyone who is suspected of committing that offence will now have a responsibility to answer for it.

We are also declaring, through this fact of responsibility, that potential defendants will be subjected to the criminal process, and therefore will potentially have to provide their answer within the context of a criminal trial, or at the very least will potentially have to make some kind of plea about the charges against them. In other words, this new normative fact informs citizens of the appropriate arena in which to respond for their actions, which further entails that they will be under the normative context of that arena—they will have a presumption of innocence afforded to them, the charges against them will have to be proven beyond reasonable doubt, the decisions made regarding them will be based on evidence, and so on.

Duff’s distinction between liability and responsibility is also useful here, whereby “responsibility is a necessary but not a sufficient condition of liability. I am liable to conviction or blame for X only if I am responsible for X, but I can be responsible for X without being thus liable.”³⁹ I have mirrored it in distinguishing the liability being created from the responsibility being declared—the former only establishes what kind of consequence an offender will potentially face and the latter creates the normative facts that establishes the process by which an offender will be required to answer for their conduct (which may ultimately lead to facing conviction or not). Again, we might argue that there are also other kinds of responsibility at play, like moral responsibility for moral wrongdoing, but those are not the responsibilities that are being declared in the act of criminalisation—the answering for committing a crime being declared here is specifically letting people know that they will (or may, depending on the level of discretionary powers in the relevant jurisdiction) hereby be subjected to the criminal justice system if they are found (or suspected) to have Φed.

These are all normative facts which did not exist until the criminalisation claim was uttered. Even if we were to argue that the law is, in some important sense, only reporting purported moral obligations or that legal obligations are reducible to such moral obligations, it does not make sense to see the act of criminalising as a mere form of finger-pointing to moral obligations—it still needs to, on the basis of such obligations, create new ones that are triggered by putting the criminal law in motion when an offence occurs. And these new obligations, liabilities and responsibilities are declared—are performed in speech-act theory terms—by criminalising the conduct in question.

I have not yet spoken as to where wrongfulness could potentially appear in a criminalisation claim. This is because, I argue, wrongfulness of any kind (including moral) is better seen as an assertion—as a reporting of the fact that the criminalised conduct is wrong—rather than as a declaration. I will explore this possibility in the next section.

5 Criminalisation Claims: Asserting Wrongfulness

As I stated earlier, things are being both asserted and declared in the act of criminalising a conduct. Let us begin with what is being asserted. I propose that there is one assertion being made through a criminalisation claim—the claim asserts that the conduct is, in some important sense, wrongful. By the act of criminalising Φ, we are reporting as a matter of fact that Φ is wrongful. Thus, it commits the speaker (whoever that may be in practice) to expressing the belief that Φ is wrongful, and to the truth value of that assertion: it is true that Φing is wrongful.

A quick clarification is necessary at this point. The commitment a speaker takes on by making an assertion is regarding what the speech-act is expressing, not to actually holding the belief in what is being expressed. As Searle explains, when someone performs an illocutionary act with a propositional content, like an assertion, “the speaker expresses some attitude, state, etc., to that propositional content”,⁴⁰ and that this holds “even if he is insincere, even if he does not have the belief, desire, intention, regret or pleasure which he expresses, he nonetheless expresses a belief, desire, intention, regret or pleasure in the performance of the speech act.”⁴¹ This becomes obvious when we make explicit how this contrast works in a sentence. If I say “it is raining outside, but I do not believe it is raining outside”, the sentence seems linguistically unacceptable—something is off or infelicitous about that assertion. Of course, it is perfectly possible that, in reality, I do not actually believe that it is raining outside, and it is also possible that I insincerely assert that it is raining outside. The point, however, is that if I said, “it is raining outside, but I do not believe it is raining outside”, that statement can no longer perform its illocutionary point of asserting a propositional content correctly. Thus, if the speech-act states “it is raining outside”, that assertion commits the speaker to expressing a belief in the truth value of that statement, even if the speaker does not actually hold that belief.

Following Cornford,⁴² there are two potential kinds of arguments that can be made to propose that the act of criminalising a conduct communicates some kind of portrayal of the conduct as wrongful. The first is the familiar idea from legal theory that crimes are a kind of legal wrong,⁴³ which I have echoed in the declared obligation discussed in the previous section, and the Razian idea that the law claims legitimate authority to make compliance with said legal obligations morally obligatory.⁴⁴ The second kind of argument relies on looking at the institutions of the criminal law, particularly the liability to punishment I discussed in the declared normative facts, and argue that those institutions can only make sense if the conduct which makes one liable to punishment is being portrayed as a wrong. As Edwards puts it, the “fact that offenders are liable to punishment cannot but further imply that as far as the law is concerned, offenders should not so act.”⁴⁵ This same strategy is used by Husak in order to arrive at a wrongness constraint, where he looks at the doctrines of the general part of the criminal law to establish moral wrongness as a constraint on criminalisation.⁴⁶ But, as Cornford points out, since there is no express mention of the wrongness of the criminalised conduct, if this wrongness is being conveyed in criminalisation “it must be that criminalisation has a certain symbolic meaning: We share the understanding that criminalisation conveys a judgement that the targeted conduct is wrongful.”⁴⁷ This is, precisely, what I will argue for in this section.

One might object to this picture. One may think, as Allott and Shaer propose, that “the initial promulgation of a statute is a matter of enactment, not description.”⁴⁸ That is, when a conduct is criminalised through legislation, the act by which a conduct is described as a particular offence, as Marmor puts it, “is not a description of how things are in the world, but rather, a prescription that one ought not to Φ in C.”⁴⁹ That being the case, it may seem strange to think of the act of criminalisation to include some kind of description—even if it is an implied one—of the fact that a conduct is wrongful. There is a sense in which the ‘enacting’ rather than ‘describing’ point is trivially true—the enactment of a criminal offence creates the legal category of the offence being enacted, which was not part of the world previously (what I called the labelling fact). But could the same point be made about the possibility of including wrongfulness into a criminalisation claim? In other words, can there be a criminalisation claim that does not, at least implicitly, include some kind of assertion of the wrongfulness of the conduct being criminalised?

One way to answer this is to apply Grice’s ‘cancellability’ test, particularly with regards to contextual cancellability whereby, as Allott and Shaer put it, “a change in the context of the utterance effaces implied content but not the encoded content.”⁵⁰ The basic idea is that if we were to alter the context of the claim being made, the original implication of the utterance stops making sense. So, for an example relevant to this discussion and which I used earlier, to preclude wrongfulness from being a part of the criminalisation claim the following cancellation would need to make logical sense:

“(1) If a person (“A”), with A’s penis-

• (a) without another person (“B”) consenting, and

• (b) without any reasonable belief that B consents,

penetrates to any extent, either intending to do so or reckless as to whether there is penetration, the vagina, anus or mouth of B then A commits an offence, to be known as the offence of rape.

(2) But this does not mean that rape is wrong.”

If (2) seems counter-intuitive, then it seems strange to exclude the assertion of wrongfulness from the speech-act. It is, at the very least, an important implication given by the context in which the criminalisation claim is uttered, and taking it away from that context seems to give us an incomplete meaning for what is being said. It makes more sense to understand the utterance of the above enactment as it including the idea that raping is wrongful in some important way. And the way in which this is done, as I stated earlier, is indirectly. It is not by explicitly stating the wrongfulness of rape in the utterance of s.1, but rather it is understood from the fact that rape is being subjected to the declaration of all the normative facts I discussed in the previous section. And, importantly, it is not done by a separate illocutionary act, but rather in the original act of stating the contents of s.1 there is the indirect illocution of asserting that wrongfulness. Thus, the assertion of wrongfulness, I argue, is an implied assertion made as part of that indirect illocution.

How does this implied assertion work? What we need to establish is how the context in which the criminalisation claim is uttered allows for the hearer of it to rationally infer that an assertion of wrongfulness is being conveyed. The implicature here is a contextual one, so a change in the relevant aspect of the context would mean that the implicated content changes. A good way to figure this out is to look at what aspects of the context in which a criminalisation claim is uttered, if they were changed or eliminated, would negate the implicated content of the claim. As stated earlier, one possible argument in favour of the idea that wrongfulness of the conduct being criminalised is being asserted in that act is linked to the presence of criminal punishment. The fact that liability to punishment is being established for those who Φ signals to hearers that Φ is wrongful. Again, this signalling would be done through implicature as there is no explicit statement of the wrongfulness of Φ, so if we removed criminal punishment from the context of the utterance of a criminalisation claim—and the argument were correct—then the implicated content should change. Using the same example as before, this would look something like the following:

(1) If a person (“A”), with A’s penis-

• (a) without another person (“B”) consenting, and

• (b) without any reasonable belief that B consents,

penetrates to any extent, either intending to do so or reckless as to whether there is penetration, the vagina, anus or mouth of B then A commits an offence, to be known as the offence of rape.

(2a) But those who commit rape shall not be liable to criminal punishment of any kind.

If my argument that wrongfulness is being implicitly asserted and can be inferred from the presence of the declared normative facts is correct, the presence of (2a) should take away any contextual information that allows us to infer, from the criminalisation of rape by s.1, that we are being implicitly told that rape is wrongful. And this seems like a plausible conclusion to reach—if there is no possible punishment for rape, then at least in the eyes of the law, one could infer, it is not being portrayed as wrongful. I propose that the same is true about the other declared normative facts I argued for in the previous section—declaring a legal obligation not to Φ and that there is a responsibility to answer for Φing gives us enough background information or context to plausibly infer that Φ is, in some way, being portrayed as wrongful by the act of criminalising it. Note, however, that I am speaking here of the broad sense of wrongfulness I introduced earlier, because it is still an open question whether the particular wrongness being implicitly asserted is of a moral kind. That is, if criminalisation claims are implicitly asserting the wrongfulness of the criminalised conduct, how do we know if that wrongfulness is of a moral kind? I cannot answer that question here in full, but what I can do is point to where we might find an answer: figuring out whether the sense of wrongness being asserted is of a moral kind (i.e. wrongness in a narrow sense) requires answering metaethical questions on the conceptual, ontological, and epistemic nature of moral wrongness, which I have discussed elsewhere.⁵¹ From there, we will need to determine whether the declared normative facts of a criminalisation claim are enough contextual information to correctly infer from them that the implicit wrongness being asserted is, specifically, of a moral kind. In Searle’s terms, whether we can have an inferential strategy to determine that the illocutionary point of the criminalisation claim is precisely to convey the moral nature of the wrongness being implied. Whether this is possible remains unclear. What is clear, however, is that at least as things stand with criminalisation as a practice, it makes sense to understand the act of criminalisation as the conveying of the wrongfulness of the criminalised conduct in a broad sense.

6 Conclusion

Criminalisation claims, then, are doing both assertive and declarative work—or have both kinds of illocutionary force, in Searle’s terms—which are happening at the same time. However, the assertion of wrongfulness that I have argued for above is done implicitly, which means that its content needs to be reasonably inferred by any potential hearers from the context of utterance. Criminalisation claims can do so appropriately for a broad sense of wrongfulness, but the fact that the assertion is implicit creates problems if it is thought of as specifically an assertion of moral wrongfulness. This is because it is not immediately clear that hearers of the claim have the necessary context to infer the moral character of the wrongfulness involved from the context of utterance. This does not preclude the possibility of the wrongness involved being of a moral kind, but an argument needs to be made as to how the context of utterance of a criminalisation claim provides enough background information to draw the required inference. The upshot of this is that theorists can now identify a more specific target for arguing in favour of or against the moral nature of the wrongfulness involved in criminalised conduct. The target of the argument is no longer the fact that a conduct by itself is morally wrong or not, but instead it is whether the context of utterance of the criminalisation claim can provide the grounds for inferring that, for the purposes of the law at least, the criminalised conduct is a moral wrong specifically.

Additionally, there are both descriptive and normative facts being declared in a criminalisation claim. The descriptive fact is the labelling of Φ as a particular category of an offence, which may or may not be present in the criminalisation claim itself. The normative facts being declared—the obligation not to Φ, the liability for Φing and the responsibility for Φing—are not moral (pre-legal) in nature, but instead are new, legal normative facts. If there were any moral obligations, responsibilities or liabilities for Φing, they are not the ones being performed by the criminalisation claim and, therefore, are not the relevant normative facts that are engaged when the criminal law apparatus comes into action to criminalise Φing.

The objective in presenting this model for criminalisation is to provide a framework under which it becomes easier to place potential normative discussions fellow theorists may have with regard to different aspects of criminalisation, based on all the things that are being said and done through the act of criminalising. For example, discussions on fair labelling can be understood as either a distinctively descriptive analysis of language that may have normative implications, or it can be understood as a purely normative question of the link between the labelling fact being declared and the wrongfulness being asserted. The descriptive analysis asks whether the terminology used in the labelling fact accurately describes the properties of the conduct being criminalised and, if so, what normative connotations does that terminology have with regard to the wrongfulness being asserted in the same act of criminalisation. The normative analysis asks whether it is fair to assign a particular label, with particular potential normative meaning, regarding the wrong being asserted in the act of criminalisation. The framework can also make easier any potential debates around the wrongfulness included in criminalisation: if the wrongfulness of Φ is an assertion of a pre-existing normative fact, where does that fact come from? What is its factual nature and how do we have access to it? Importantly, how do we make sure that our assertion is correct? Hopefully, by having clarity as to where these issues occur in the context of what is being said when we criminalised, future debates in criminalisation theory can continue to flourish.