Awareness and the Recklessness/Negligence Distinction

Abstract

The distinction between the criminal fault elements of recklessness and negligence is one of Anglo-American criminal law’s key distinctions. It is a distinction with practical significance, as many serious crimes require at least recklessness and cannot be committed negligently. The distinction is standardly marked by awareness. Recklessness requires awareness that one’s conduct carries a risk of harm. Negligence only requires that one ought to have been aware that one’s conduct carried such a risk, even if one was in fact unaware of this. But should the recklessness/negligence distinction be marked by awareness of risk, or by something else? Does a defendant’s awareness of risk really have the normative significance to mark such a distinction? In this paper, I answer these questions by discussing a challenge to this ‘standard account’ of the recklessness/negligence distinction raised by the work of Antony Duff, who defends an alternative, non-awareness-based model of the recklessness/negligence distinction. I will argue that, although Duff’s alternative model fails, seeing how it goes wrong helps us see how awareness genuinely does have the right kind of normative significance to mark the distinction between recklessness and negligence.

1 Introduction

Recklessness and negligence are two of Anglo-American criminal law’s key mens rea or fault elements, i.e. conditions that defendants must meet, in addition to carrying out a proscribed act (the actus reus), in order to commit a criminal offence. Recklessness and negligence are different kinds of culpable unjustified risk-taking. Standardly, the difference between recklessness and negligence is marked by awareness. Recklessness requires a defendant to be aware of a risk, a risk which makes her course of action an unreasonable one to take.¹ Negligence only requires that a defendant ought to have been aware of such a risk – and ought to have taken precautions against it – even if she was unaware of the risk.²

In addition, the recklessness/negligence distinction, framed in these terms, is standardly accorded normative significance in two key respects. Firstly, recklessness is understood to be more culpable, all else being equal, than negligence.³ Secondly, although it’s accepted that negligence can sometimes be culpable, it’s taken to be a defeasible presumption that recklessness is the minimum fault element required for criminal offences.⁴

This ‘standard account’ of the recklessness/negligence distinction has recently been objected to by Antony Duff. In essence, Duff’s work raises the question of why, if awareness isn’t necessary for criminal culpability, should it be regarded as more culpable and presumptively necessary for criminal sanction. Duff himself favours a sceptical answer to this question, arguing that awareness doesn’t in fact have the normative significance the standard account ascribes to it. He instead puts forward an alternative model of the recklessness/negligence distinction which does not frame the distinction in terms of awareness of risk, but in terms of the agent’s ‘practical attitude’ in relation to the risk. While I will argue that this alternative model fails, seeing how it goes wrong allows us to better understand the normative significance of awareness. Duff’s model goes wrong because awareness plays a necessary role in responding to a reason. And the role of awareness in responding to reasons explains its normative significance. A defendant’s awareness of risk is normatively significant because it puts him in a better position to respond appropriately to that risk than a corresponding negligent defendant. This provides us with a justification of the standard account of the recklessness/negligence distinction, one which improves upon previous suggestions about the significance of choice made by A.P. Simester and Findlay Stark.

2 Duff’s Alternative Model

On Duff’s alternative model, which draws on the German law concept of dolus eventualis, the recklessness/negligence distinction is framed not in terms of awareness, but in terms of how the risk figures in the structure of the agent’s practical reasoning. Recklessness involves a ‘practical attitude’ of ‘endorsement’ or ‘acceptance’ of the risk, whereas negligence does not:

A reckless agent is one who ‘endorses’ the risk that he takes or creates: … he endorses or accepts the risk as one that is worth taking for the sake of the enterprise in which he is engaged. A (merely) negligent agent, by contrast, does not display this practical attitude of endorsement towards the risk she takes: she takes the risk not because she endorses it, or thinks it worth taking, but because she does not notice it; she is ‘pained’ when she realises the risk (even if the threatened harm does not ensue); it was not a risk that she endorsed, or would have endorsed, as one worth taking for the sake of her goal.⁵

Duff claims his alternative understanding of recklessness comes in both advertent and inadvertent varieties because ‘sometimes, an inadvertent risk-taker displays, in his very failure to notice the risk that he creates, the kind of acceptance or endorsement of the risk that characterises recklessness’.⁶

Duff supports this claim by appealing to the laws of rape and murder. In English law, rape can be committed if a defendant has sex with a non-consenting person without reasonably believing that the other person consents.⁷ In relation to murder, Duff points to the English law of murder, which requires an intention to kill or to cause grievous bodily harm,⁸ and so allows for a murder conviction in some cases where the defendant is unaware that she might kill her victim. In cases like these, the law allows (justifiably, Duff thinks) for convictions of inadvertent defendants for serious offences. Duff claims this is more naturally explained by his alternative non-awareness-based framing of the recklessness/negligence distinction. Duff asks us to consider one defendant who ‘persists in sexually penetrating a person who provably does not consent to it, without a reasonable belief that she consents’, and another defendant who ‘attacks another violently, in a way that creates an obvious risk of death’.⁹ Duff claims these defendants manifest the kind of ‘acceptance’ or ‘endorsement’ he claims characterises recklessness. This is because the relevant risk – of non-consent or of death – is so intimately connected to the intended act – sexual penetration or the violent attack – that the defendant’s failure to notice this risk shows that the other person’s consent or life ‘figures in the structure of [the defendant’s] practical reasoning as a somewhat unimportant consideration, as to which one could easily be inadvertent’.¹⁰

The role of these cases is not to provide straightforward counterexamples to the standard account, i.e. cases which straightforwardly show that inadvertent defendants can be just as culpable as advertent defendants. If it were, proponents of the standard account could respond that they only claim that recklessness (defined in terms of awareness of risk) is more culpable than negligence all else being equal.¹¹ And it’s not clear that Duff’s cases are examples of inadvertent risk-takers for whom it’s not the case that advertent risk-taking in an otherwise identical case is more culpable. We should instead understand Duff as claiming his alternative model provides a better and more elegant explanation of these cases than the standard account does.¹² The crucial claim for this argument to succeed is Duff’s claim that some inadvertent risk-takers, such as his inadvertent rapist and murderer, display ‘the kind of acceptance or endorsement of the risk that characterises recklessness’. This claim would imply that some inadvertent wrongdoing involves the same kind, and not merely the same degree, of fault as advertent cases of recklessness.

In the next section, I will argue that this key claim of Duff’s – that someone can count as ‘accepting’ or ‘endorsing’ a risk he is unaware of – is not a claim we can make coherent sense of. For this reason, Duff’s alternative model should be rejected. But I first want to highlight how Duff’s alternative model still raises an important overlooked challenge to the standard account. Specifically, the suggestion that we might not mark the recklessness/negligence distinction in terms of awareness challenges proponents of the standard account to say why awareness is normatively significant. This challenge is especially pressing for proponents of the standard account who accept that criminal liability without awareness – the standard account’s negligence – can sometimes be justified. To such non-sceptics, Duff’s argument for his alternative model poses a key overlooked question: if awareness isn’t necessary for criminal culpability, why should we still regard culpability grounded in awareness as more culpable and presumptively necessary for criminal sanction?

This question has been overlooked because of the prominent role sceptical arguments against criminal negligence liability have played in the literature. Attempts to explain the normative significance of awareness for criminal culpability have almost always been in the service of sceptical arguments against criminal negligence liability. In such arguments, a defendant’s awareness of what she is doing has been claimed to be necessary for culpability because it it’s necessary for, say, choice, control, or the ability to do otherwise, conditions which are then argued to be necessary for criminal culpability.¹³ It has been largely overlooked that if one thinks such sceptical arguments can be answered – as Duff does,¹⁴ and as do I¹⁵ – one then needs explain why awareness still has the more moderate normative significance the standard account ascribes to it. One needs to explain why, if awareness isn’t necessary for criminal culpability, culpability grounded in awareness is nevertheless more culpable and presumptively necessary for criminal sanction. While Duff’s alternative model should be rejected, his insight that the defenders of the standard account face an explanatory demand here is entirely correct. But it’s an explanatory demand which can be met by seeing how Duff’s alternative model goes wrong.

3 Assessing Duff’s Alternative Model

What should we think of Duff’s claim that some inadvertent risk-taking displays ‘the kind of acceptance or endorsement of the risk that characterises recklessness’? Is this claim correct? What I will suggest now is that it’s not a claim we can make coherent sense of, because we cannot make sense of someone’s action as manifesting ‘acceptance’ or ‘endorsement’ of a risk if she is unaware of that risk. This is because of two related points. Firstly, ‘accepting’ or ‘endorsing’ a risk looks like a specific way of responding to a reason, namely the reason constituted by the fact that there’s a risk. Secondly, responding to a reason requires awareness of it.

It will be easiest to explain the second point first, by stepping back and considering quite generally what’s involved in responding to a reason. Responding to a reason is a certain way of responding to a fact. Reasons are facts that count in favour or against certain courses of action, attitudes, emotional reactions, and so on.

The key claim I will rely on is that someone only counts as responding to a reason if he is aware of the fact that constitutes it. For example, let’s say I’m doing some hammering. The fact that my hammering could smash through your wall constitutes a reason to stop hammering. I can only count as stopping my hammering in response to the fact that it risks smashing through your wall if I’m aware of this fact. Of course, I could take the course of action this fact recommends – ceasing to hammer – without being aware of this fact. I could stop hammering because I get too tired or because the doorbell rings. But my ceasing to hammer only counts as a response to the fact that it risks smashing through your wall if I’m aware of this fact. More generally, I only count as responding to a reason – in thought, feeling, or deed – if I’m aware of the fact that constitutes that reason.

This claim should not be confused with the claim that responding to a reason requires awareness of the fact’s status as a reason. There seem to be a variety of cases in which one can respond to a reason – in thought, feeling, or deed – without awareness of the fact’s status as a reason. A young child might respond to the fact that it’s her birthday tomorrow by getting excited about it. A dog might respond to the fact that there’s a cat in the garden by chasing it. Both of these cases look like responses to reasons, but they are cases in which the agents lack awareness of the relevant fact’s status as a reason, because they lack the requisite conceptual sophistication. But the child is still aware that it’s her birthday tomorrow and the dog is still aware that there’s a cat in the garden, and we can only understand their behaviour as a response to those facts on the condition that they have such awareness.

What sense of ‘awareness’ is required for responding to a reason? ‘Awareness’ can mean many things in different contexts, and some things we call ‘awareness’ are not required for responding to a reason. For example, ‘awareness’ sometimes refers to events in the stream of consciousness. If I were to describe myself as suddenly becoming aware, while on my way to work, that I forgot to lock the front door, that would naturally be taken to refer to an event in my stream of consciousness, such as a conscious judgement that I forgot to lock my door, or my consciously thinking to myself ‘I forgot to lock the door’. If I were to describe myself as becoming aware, out of the corner of my eye, that someone else had entered the room, that would naturally indicate another kind of event in the stream of consciousness, namely perceptual awareness.

Awareness understood as an event in the stream of consciousness is clearly not required for responding to a reason. We often respond to reasons without perceiving or making conscious judgements about the facts which constitute them. For example, consider someone who, during routine driving, changes down gear because she is approaching a roundabout. Such an action is a response to reasons. It is a response to, inter alia, the reason constituted by the fact that she needs to change down gear. But in routine cases, a driver wouldn’t normally consciously judge, or think to herself, ‘I need to change down’.¹⁶

The sense of ‘awareness’ which is necessary for responding to a reason is instead a standing state of awareness. This is a dispositional state, like knowledge or belief. Now some have identified awareness, when it plays a role in the criminal law, with knowledge¹⁷ or belief.¹⁸ The claim I am defending doesn’t necessitate such an identification though, as awareness could also be a sui generis standing state. The important thing is that such a state of awareness is not an event in the stream of consciousness, though it might be manifested in the stream of consciousness in various ways. My awareness that my train leaves in two minutes might be manifested in anxious feelings, or my consciously thinking to myself ‘My train leaves in two minutes – I’m not going to make it’. But the standing state of awareness itself can pre-date and outlast such manifestations in the stream of consciousness. In addition, such states of awareness can affect behaviour without being manifested in the stream of consciousness. Indeed, this is exactly what happens in our case of the driver. Her awareness that she needs to change down gear leads her to change down as she approaches the roundabout. And the key claim I am making is that in order for her behaviour to count as a response to the reason constituted by the fact that she needs to change down, she must be aware of this fact, in this standing-state sense of ‘awareness’.

This precise claim – that responding to a reason requires awareness of it – is widely accepted in philosophy of action. It’s not a claim that is often explicitly defended.¹⁹ This is because it’s a claim that is assumed or presupposed by both parties in the key debates in philosophy of action. For example, the claim is implicit in the causal theory of action, according to which acting for a reason is for one’s bodily movement’s to be caused by certain mental states – such as a belief–desire pair – mental states which will include awareness of the reason for which one acts.²⁰ It is also implicit in the main alternative to the causal theory of action, G.E.M. Anscombe’s view, according to which intentional actions (a slightly broader category than actions done for reasons) are those actions to which an agent grants application to the question ‘Why did you do that?’.²¹ Granting application to such a question presupposes awareness of the reasons for which one acts.

Similarly, the claim is accepted by both parties in the psychologism/anti-psychologism dispute about motivating reasons. This is the dispute as to whether motivating reasons – i.e. the reasons in the light of which we act – are an agent’s mental states or facts those mental states are about. Those who claim that motivating reasons are an agent’s mental states include awareness of normative reasons among those mental states. Those who claim that motivating reasons are the facts our mental states are about nevertheless accept that awareness of those facts is a necessary condition for acting for a reason (even though they reject the identification of awareness with the motivating reason).²²

With this first point – the claim that awareness is necessary for responding to a reason – defended, I can now move on to my other key point. This point concerns Duff’s characterisation of recklessness, on his alternative model, as involving ‘acceptance’ or ‘endorsement’ of a risk. The reckless agent, according to Duff, accepts a risk as a price worth paying for the sake of the enterprise he is engaged in. The key point I want to make here is that ‘accepting’ or ‘endorsing’ a risk is a certain kind of response to a reason, namely the reason constituted by the fact that there’s a risk.

I grant that accepting or endorsing a risk is not exactly the same kind of response to a fact as was involved in my previous example of responding to the fact that my hammering risks smashing through your wall by ceasing to hammer. In that case, I responded to that fact by doing what the reason recommended. Someone who accepts or endorses a risk in the sense Duff has in mind – i.e. someone who accepts the risk as a price worth paying for the sake of the enterprise he is engaged in – does not respond to the fact that there’s a risk by doing what this fact recommends. That’s precisely the problem. He responds to the fact that there’s a risk attendant on his planned course of action – something which counts against this course of action – by carrying on anyway with the course of action. But such acceptance or endorsement of a risk still counts as a kind of response to the reason constituted by the fact that there’s a risk.

If this point is correct, it creates a problem for Duff’s alternative model if we combine it with the previous point that responding to a reason requires awareness of the fact that constitutes that reason. In particular, it creates a problem for Duff’s claim that some inadvertent defendants manifest the kind of acceptance or endorsement of the risk that characterises recklessness. If what I’ve said so far is correct, we can’t make coherent sense of this claim. This is because if accepting or endorsing a risk is a response to a reason, namely the reason constituted by the fact that there’s a risk, then it cannot occur without awareness of the fact that constitutes this reason, i.e. the fact that there’s a risk. In short, acceptance or endorsement of a risk cannot be inadvertent.

Now Duff does say something which may speak to this objection. With the examples of the inadvertent defendants he discusses, Duff claims that the relevant risk ‘figures in their practical reasoning’. These two defendants, recall, were one who ‘persists in sexually penetrating a person who provably does not consent to it, without a reasonable belief that she consents’ and another who ‘attacks another violently, in a way that creates an obvious risk of death’. Because the risks of non-consent and death are so intimately bound up with what these defendants are up to, Duff claimed that their failures to notice these risks show that the other person’s consent or life ‘figures in the structure of [the defendant’s] practical reasoning as a somewhat unimportant consideration, as to which one could easily be inadvertent’.²³

This does not answer the objection. This is because Duff equivocates here between two very different ideas. The first idea is that someone’s failure to be aware of a particular fact can be a manifestation of the fact that he doesn’t care much about facts of that general kind. My failing to notice my partner’s discomfort at my telling an embarrassing story about her in front of my friends can express the fact that I don’t care much – and don’t care enough – about facts of that general kind, i.e. facts about her discomfort. The second idea is that someone can count as ‘endorsing’ or ‘accepting’ a particular fact – or a particular risk – by failing to notice it. These are very different ideas.

This equivocation matters because only the first idea is supported by the examples of inadvertent agents Duff appeals to, i.e. the inadvertent rapist and the murderer who is unaware of the risk of death. Such agents’ failures to notice the particular risks attendant on their conduct do plausibly show that they don’t care enough about certain general kinds of facts, i.e. facts about the victim’s consent and facts about the victim’s life. But the second idea is what is needed to undermine the standard account of the recklessness/negligence distinction, because only the second idea would support the claim that some cases of inadvertent risk-taking involve the same kind, and not merely the same degree, of fault as cases of deliberate risk-taking. But Duff’s examples don’t support this second idea, as these agents’ failures to notice, respectively, the particular fact that there is a risk of non-consent and the particular fact that there is a risk of death don’t plausibly amount to them accepting or endorsing these facts about these risks. This is because, as I have argued, we can’t make coherent sense of the idea that one can accept or endorse a fact one is unaware of. Duff’s characterisation of recklessness should therefore be rejected.

I mentioned above in passing that Duff takes his alternative model of recklessness to be inspired by the German law concept of dolus eventualis, sometimes also referred to as conditional intent. If we accept my claim that we can’t make coherent sense of Duff’s claim that his alternative characterisation of recklessness can be inadvertent, must we conclude that we can’t make coherent sense of dolus eventualis? I think the answer is no. Being no expert on German law, I am forced to rely on other scholars here, but there seems to be agreement that dolus eventualis necessarily requires a defendant to be aware that his conduct may have some result.²⁴ There is debate is about whether there is an additional volitional element that is also necessary (and jointly sufficient) for dolus eventualis, an element constituted by, e.g., ‘approval’ or ‘acceptance’ of the result in question.²⁵ It is this additional volitional element that Duff is drawing on, in his account of recklessness, and which he claims can be manifested in inadvertent risk-taking. It is this latter claim – which is not part of the concept of dolus eventualis – which I am claiming we cannot make coherent sense of. So no wholescale rejection of the notion of dolus eventualis follows from my argument.

At this stage, an opponent of the standard account may argue that all I’ve done is show that some of the details of Duff’s alternative non-awareness-based model of the recklessness/negligence distinction don’t work, but maintain that Duff is still correct that it is best not to frame that distinction in terms of awareness. Now this might be a reasonable move to make, but for it to be sustained, it will be have to be backed up with some alternative non-awareness-based framing of the recklessness/negligence distinction that doesn’t face the problems I have raised for Duff’s. So, as a defence of Duff, this move is incomplete.

In fact, one possible route an opponent of the standard account might take here is to appeal to some of Duff’s earlier work, in which he characterises recklessness as differing from negligence because it involves a kind of ‘practical indifference’:

What makes a reckless agent more culpable, more fully responsible for the risk she creates, is that she displays a gross indifference to that particular risk or to the particular interests which she threatens: negligence, however, involves a less specific kind of carelessness which does not relate the agent so closely, as an agent, to the risk which she creates. To show that I recklessly endangered someone’s life it must be shown that I manifested a culpable indifference to her life: but negligently endangering her life need involve only a lack of attention to what I am doing – not a specific indifference to that particular risk.²⁶

This characterisation is helpful for our purposes because it notably does not feature the concepts of ‘acceptance’ or ‘endorsement’, concepts which created problems for Duff’s more recent characterisation of recklessness, which we’ve been focusing on. Nonetheless, it’s a contrast which Duff intends to mark a ‘categorial’ distinction between two kinds of risk-taking, either of which, Duff claims, can be inadvertent.

Can this contrast thus provide an alternative non-awareness-based characterisation of the recklessness/negligence distinction?²⁷ I struggle to see that there’s a genuine contrast here between a more serious form of culpable indifference and mere lack of attention. It’s not clear that this marks a key distinction in culpability. Now it is plausibly correct that there are some cases where we think mere lack of attention is insufficiently culpable for criminal sanction. But I suspect that such cases will be ones which the defender of standard account will agree shouldn’t be criminalised for other, independently plausible reasons.

For example, one kind of case which we might describe as one of mere lack of attention, rather than culpable indifference, would be a case in which someone lacks the capacity to pay sufficient attention. For example, consider the English case of Elliott v C, in which the defendant was a 14-year-old girl of low intelligence, who set fire to a shed by igniting white spirit and was convicted of criminal damage. This is despite the fact that the court found that she ‘had given no thought’ to the risk she was causing, and that the risk ‘would not have been obvious to her or appreciated by her if she had given thought to the matter’.²⁸ It’s natural to describe such a case as involving lack of attention, but not culpable indifference, as Duff himself notes in his discussion of the case.²⁹

However, a number of defenders of the standard account allow for such cases not to be criminalised, typically by arguing that negligence assessments should be relativised to the defendant’s cognitive capacities.³⁰ Moreover, there’s also the option of arguing that the defendant in Elliott v C shouldn’t face criminal sanction because the interests at stake, property interests, are not important enough for negligent harm to them to be criminalised.³¹ The standard account, after all, claims that negligence is presumptively insufficient for criminal sanction, and one of the considerations feeding in to when that presumption is defeated will be the importance of the relevant interests at stake.³²

On the other hand, if we try and imagine a case for which neither of these two moves can be made in defence of the standard account – i.e. a case in which mere ‘lack of attention’ is shown by someone with the capacity to pay attention and which also concerns a sufficiently important interest – it’s no longer clear that there’s a plausible distinction between mere lack of attention and culpable indifference. If something sufficiently serious is at stake – if not paying attention to what I am doing might cost someone’s life, for example – and if I have the capacity to pay attention, then it seems that a mere lack of attention to what I’m doing is just as serious as being inadvertently ‘indifferent’ to someone else’s life. In such a case, Duff’s contrast between mere lack of attention and culpable indifference seems to evaporate. In such a case, they do not mark a key distinction in culpability.

For these reasons, Duff’s earlier characterisation of recklessness as ‘practical indifference’ also doesn’t provide a workable non-awareness-based framing of the recklessness/negligence distinction. Therefore, if it’s still going to be claimed that the recklessness/negligence distinction should not be drawn in terms of awareness, the onus is on those making such a claim to provide a plausible alternative way of drawing that distinction.

4 Why Awareness Matters

We have seen that Duff’s alternative non-awareness-based framing of recklessness/negligence distinction fails. However, it is still the case, as I said earlier, that Duff highlights an overlooked explanatory demand faced by proponents of the standard account of that distinction. The idea that there might be a non-awareness-based framing of the recklessness/negligence distinction raises the question of why awareness has the normative significance the standard account ascribes to it. Why, if awareness isn’t necessary for criminal culpability, should culpability grounded in awareness still be thought of as more serious, and as presumptively necessary for criminal culpability?

If we don’t provide an answer to this challenge, then even if I’m correct that Duff’s alternative model fails, the answer to Duff is somewhat unsatisfactory. This is because we won’t have fully explained why the standard account is correct to grant awareness the significance that it does. And while my objection to Duff’s alternative model doesn’t by itself provide an answer to this explanatory demand, we can find the beginnings of that answer in my previous discussion of responding to reasons. Specifically, the necessary role awareness plays in responding to reasons can provide an explanation of why awareness genuinely does have the normative significance the standard account ascribes to it. This is an explanation which, I will argue, improves on existing choice-based explanations put forward by Stark and Simester.

It is plausible that there is a connection between culpability and responding to reasons. Indeed, some have put forward analyses of culpability in exactly these terms. Simester, for example, claims that culpability ‘derives from [a defendant’s] engagement with the reasons why she should not φ’ and that a defendant ‘is culpable when her engagement with those reasons is defective in a manner that reflects a moral vice on [her] part’.³³ But whether or not we want to follow Simester in defining culpability in these terms, everyone should agree that the way in which a defendant engages with the reasons there are against φing – that is, the manner in which she responds or fails to respond to those reasons – can tell us something about her culpability for φing.

Once this is accepted, we are then able to provide an explanation of why awareness genuinely does have the normative significance that the standard account of the recklessness/negligence distinction ascribes to it.

The key point is that a reckless agent, in virtue of his awareness of the relevant risk, is in a better position to respond to that risk than the negligent agent is. We can explain this idea by comparing parallel negligent and reckless agents. Take a reckless agent who smashes someone else’s window by performing a martial arts move in close vicinity of it in order impress his friends.³⁴ Compare him to a negligent agent who smashes her neighbour’s window because she fails to notice, but should have noticed, that the tree she was cutting down would smash into it. For both agents, the fact that there’s a risk constitutes a reason not to proceed with their respective courses of action. And both agents fail to respond to this reason in the way that they should. But the reckless agent is in a better position to respond to this reason appropriately because – in virtue of his awareness of it – he meets a necessary condition for responding to it appropriately that the negligent agent does not meet. We can put this point by saying that the reckless agent, in his awareness of the reason, has an opportunity to respond to it appropriately which the negligent agent lacks.

Now the negligent agent would have had this opportunity if she had paid more attention, which she could and should have done. This is why her very failure to have this opportunity can make her culpable for her inadvertent risk-taking. But as things happened, the negligent agent didn’t have the opportunity to respond to this reason which the reckless agent did.

This difference between the reckless and negligent agent, I suggest, is what allows us to explain the two key respects in which the recklessness/negligence distinction is standardly given significance. These were, firstly, the claim that recklessness is more culpable than negligence, all else being equal, and secondly, the claim that recklessness is the presumptive minimum culpability required for criminal sanction.

Let’s start with the first of these claims. The fact that the reckless agent meets a necessary condition on responding to the risk, a condition which the negligent agent does not meet, plausibly makes him more culpable. The reckless agent recognised his course of action carried a risk of harm – and in so doing had an opportunity to respond to this risk which the negligent agent didn’t – and he still went ahead and took the risk. This plausibly makes the reckless agent more culpable, all else being equal, than the negligent agent.

This is consistent with saying that, all things considered, the negligent defendant did have an opportunity to respond to the risk, because she had the opportunity to become aware of it. But the claim I am appealing to is the claim that the reckless defendant, in virtue of his awareness, had an extra opportunity to respond to the risk which the negligent defendant lacks. And it is this extra opportunity which makes the reckless defendant more culpable, all else being equal, than a negligent defendant.

This explanation of why recklessness is more culpable than negligence may sound similar to existing explanations appealing to the notion of choice. But I want to suggest that my account in fact improves upon such explanations. Stark and Simester, for instance, have both claimed that awareness makes a difference because it is necessary for choice.³⁵ They claim that chosen wrongdoing – i.e. intended and reckless wrongdoing – matters because an agent’s choosing to cause or risk harm means that he ‘identifies’ or ‘aligns’ himself with the harm. For example, Stark writes:

Choosing to risk a certain consequence … could … be said to align the actor with that consequence … in a special way. This implicit identification with the desire to act, despite the risks of which the defendant is aware, would provide a potential basis upon which to reflect that person’s agency in A-ing and open him up to appropriate reactive attitudes with regard to A-ing (and the consequences of the attached risks), such as the condemnation attendant upon a criminal conviction.³⁶

And we find a similar sentiment expressed by Simester:

When D chooses to act for bad reasons, she accepts and aligns herself with those reasons, and in turn with her wrongdoing. That action, being chosen, expresses her own values and dispositions, her knowing preference for bad reasons over good ones.³⁷

My suggestion is that insofar as these claims are plausible, they are claims that are helpfully elucidated and clarified by my account. Without elaboration, there is a danger of it being unclear why a reckless agent counts as ‘aligned’ or ‘identified’ with a risk he takes (and the risk’s consequences) more than a corresponding negligent agent. The reckless agent will not be aiming at the risk or else he would count as intending it. He may instead bring about the risk grudgingly or reluctantly. A foreman of a building company may be aware of unsafe practices that create risk to the lives and safety of his employees, and reluctantly continue with such unsafe practices under the pressure of getting the job done on time. How then does it make sense to say that the reckless agent is aligned with the risk – and in particular more aligned with it than a negligent agent – given that he doesn’t aim to create it?

What I’ve said provides an answer to this question. The reckless agent, in virtue of his awareness of the risk, is in a better position to respond appropriately to the reason against his course of action constituted by that risk than the parallel negligent defendant is. This is why the reckless agent’s choice to go ahead and take the risk makes him more aligned with the risk than the parallel negligent defendant. Although the reckless agent doesn’t aim to create the risk, he does have an opportunity to refrain from taking the risk the negligent agent lacks. It is because he fails to take this opportunity, but instead goes ahead and takes the risk, that he is more aligned with risk than the parallel negligent defendant who lacks this opportunity. Again, this is not to say that the negligent agent lacks an opportunity to respond appropriately to the risk all things considered. The negligent agent will have this opportunity if she could have been aware of the risk in question. The point is just that the reckless agent has an extra opportunity, in virtue of his awareness of the risk, to respond appropriately to the risk which the negligent defendant lacks. And it is this extra opportunity which explains why the reckless agent counts as more aligned with the risk than the negligent agent, even though neither agent aims at creating the risk. Thus a choice-based explanation of why recklessness is more culpable than negligence, of the kind put forward by Stark and Simester, is improved upon if it’s unpacked in terms of the account put forward here.

So far I’ve explained how the necessity of awareness for responding to reasons explains why recklessness is more culpable than negligence. If what I’ve said so far is correct, this also provides part – though only part, I believe – of an explanation of the second key respect in which the recklessness/negligence distinction is standardly given normative significance, i.e. the defeasible presumption that recklessness is required for criminal offences. If recklessness is more culpable than negligence for the reasons I have just given, that could partly explain why recklessness is a presumptive minimum requirement for criminal sanction. Though this presumption, of course, might be defeated.

In what kinds of case might it be defeated? It’s natural to think it might be defeated if the interests at stake are particularly important. For example, criminalising negligent killing and negligent non-consensual sexual intercourse look like prohibitions which protect particularly important interests – namely, the victim’s life and the victim’s sexual autonomy and integrity – and so in such cases the presumption that recklessness is required for criminal sanction is plausibly defeated.³⁸ Furthermore, offences of negligence can sometimes require the defendant to carry out associated intentional acts, which by themselves aren’t necessarily wrongful, but are closely associated to a particular risk. Rape, although it can be committed through negligence, is such an offence. In English law, although rape can be committed if the defendant unreasonably believes another person consents, the offence still requires intentional sexual penetration. Given that this intentional act is one which is so closely associated with matters of consent, we can expect a person who carries out such an intentional act to take care that the other person is consenting. This kind of consideration may also be part of the reason why the presumption that recklessness is required for criminal sanction is defeated in the case of rape.³⁹

However, I think the account I’ve given here will only ever be part of the explanation of why recklessness is the presumptive minimum bar for criminal sanction. This is because this presumption is, in effect, a principle of criminalisation. It provides us with a (defeasible) principle about the extent to which negligence should be criminalised. And the discussion so far has exclusively concerned the culpability attaching to recklessness and negligence. Culpability is one input into questions of criminalisation, but it is by no means the only one. Wrongness is another. But more important in this context, I suggest, are considerations of the loss of individual freedom that a coercive system like the criminal law inevitably has, and how this loss of freedom should be weighed against the gains the criminal law provides (whether we think of those gains in terms of the deterrence of future wrongs or the punishment of existing wrongdoers). One role played by mens rea elements is in determining the extent of the loss of individual freedom criminalisation enacts. Making a criminal offence one of negligence places a greater restriction on freedom – because it incentivises people to be more careful and thus to take fewer risks – than making it an offence that requires recklessness.⁴⁰ Therefore, if we want to balance this loss of freedom against the other aims of the criminal law, the best way to accomplish this may well be to have recklessness, rather than negligence, as the presumptive minimum culpability required for criminal sanction.

5 Conclusion

The fundamental difference between reckless and negligent agents is the manner in which they fail to respond to reasons. And this can explain, at least in part, why awareness genuinely does have the normative significance that the standard account of the recklessness/negligence distinction ascribes to it. The fact that the reckless agent is in a better position to respond to the risk he takes explains why he is more culpable than a parallel negligent agent. It also provides part of the explanation of why recklessness should be the presumptive minimum bar for criminal sanction, though fully defending this second claim would require a more detailed discussion of criminalisation than I can provide here. If what I have argued here is correct, awareness does have the normative significance the standard account ascribes to it. Some may find this explanation of the normative significance of awareness somewhat underwhelming, especially when compared with criminal negligence sceptics’ claims that inadvertence rules out control or the ability to do otherwise. But if we reject the criminal negligence sceptics’ claim that awareness is necessary for culpability, as I think we should, then I suspect we can only regard awareness as having this more moderate significance. The lesson of Duff’s challenge is that non-sceptics about criminal negligence still have to say something about why awareness has the normative significance the standard account ascribes to it. I hope to have shown that they can indeed say something about this.