Findlay Stark’s book on the culpability of recklessness and negligence is rigorous, thorough and thought provoking. It is persuasively argued, and engages in stimulating ways with both criminal law theory and caselaw. At times one wants to hear even more about his interesting positive accounts, but the book offers valuable and original contributions throughout, which will make it a rewarding read for criminal law scholars and practitioners alike. In this review, I sketch Stark’s main moves, and then raise a worry about his view of risk awareness, which is central to recklessness.

Chapters 1–3 set the stage. While Chapter 1 introduces his project, Chapter 2 mounts a comparative argument that a consensus is coalescing in Anglo-American jurisdictions around a “Standard Account” of culpable carelessness. The Standard Account understands “recklessness in terms of the unjustified taking of a risk of which the defendant was aware, and negligence as … concerned with inadvertent taking of an unjustified risk”—i.e., risks of which one is unaware (26).Footnote 1 Chapter 3 then argues from the principle of fair warning that recklessness and negligence should be given uniform definitions, rather than being defined differently for different crimes.

The rest of the book is mainly devoted to defending accounts of recklessness and negligence as they figure into the Standard Account. Chapters 4–5 deal with recklessness, and unpack the notion of risk awareness on which it rests. Chapters 6–8 consider negligence, and aim to explain why inadvertent risk-taking, in which awareness of the risk is absent, is culpable. These chapters constitute the core of the book.

More precisely, Chapter 4 tackles the notion of awareness, as it figures into recklessness. Stark contends that “to be aware of a particular risk associated with φ-ing requires a belief that the relevant risk exists” (91), and then proceeds to defend a dispositional account of belief. Drawing on Eric Schwitzgebel’s work (99–111; Schwitzgebel 2002), Stark’s account takes it that believing that p consists in having a wide range of associated dispositions (e.g., to assent to p if prompted, to behave as if p were true, etc.). Notably, many of these can be present even when not consciously attending to p. The belief that p can be legitimately ascribed to someone iff she possesses enough of the relevant dispositions associated with believing that p (103). Plugging this account into the concept of recklessness, Stark contends that a reckless actor is culpable when he “manifests a sufficient number of dispositions to merit the ascription to him of a belief that there is a specific, substantial risk attendant upon φ-ing and is unmoved by the belief in circumstances that show him to be insufficiently motivated by the interests of others” (122).

Chapter 5 goes on to argue that belief is not only necessary for awareness, but also that nothing more than belief is required. Accordingly, Stark takes aim at Douglas Husak’s view that awareness that p just is knowledge that p. Instead, Stark argues that, on the best account of recklessness, “a defendant is aware of a risk when she believes that it exists” (140). I raise some questions about this view below.

Chapter 6 then shifts to negligence and begins considering the culpability of inadvertent risks. Stark clarifies the relation between choice and culpability, and argues that there are no persuasive grounds for thinking that “choice is a necessary component of culpability” (142). If choice were necessary for culpability, it would prevent negligence from being culpable, since one can’t choose to take risks of which one is unaware. By refuting the arguments for seeing choice as necessary for culpability (though it remains one possible route to culpability), Stark opens the door to regarding negligence as culpable.

The next two chapters continue to build the case for the culpability of negligence. Negligence involves non-awareness of risk, so we can’t establish the culpability of negligence without a theory of how culpability could exist without awareness. Thus, Chapter 7 examines several such accounts. However, Stark concludes that even the best ones—e.g., tracing accounts (192–201) and counterfactual accounts of the culpability of inadvertent risks (201–208)—face substantial difficulties.Footnote 2

Chapter 8 is where Stark aims to clinch the argument for the culpability of negligence by defending a positive account of culpability without awareness. On Stark’s view, negligence involves a failure of belief about the risks one imposes. Beliefs “are formed through a process of combining information gleaned from perception and the defendant’s ‘background beliefs’” (229). Thus, perception and background beliefs are “necessary (but not sufficient) conditions for negligence as failure of belief” (230). When the relevant perceptions and background beliefs are present and one fails to form the requisite beliefs about the risks of one’s conduct, this can manifest insufficient concern for others and thus be culpable.

Stark addresses numerous worries about his account. For one, if we had no control over what we believed, failures of belief might not be culpable. But Stark responds that, while people usually can’t directly “control the content of their beliefs,” they often can control “the process of encouraging the formation of appropriate beliefs—of looking for evidence and weighing it, etc.” (235). Since we don’t fully lack control over our beliefs—i.e., our belief-forming processes—it remains possible to be culpable for failures thereof.

Furthermore, one might worry that more explanation is needed of precisely how failures of belief can manifest insufficient regard. To flesh out his account, Stark relies on Stephen Garvey’s non-tracing account (244–247). Garvey’s insight is that sometimes our desires can distort our belief-forming processes so we fail to become aware of the risks we should have noticed, though often we can be expected to control the distorting effects of our desires—i.e., exercise “doxastic self-control” (244; Garvey 2006: 337–338). When one fails to exercise such self-control due to insufficient concern for others, it renders one culpable (245). Stark is sympathetic to Garvey’s account, but expands it—e.g., to go beyond controlling one’s bad desires (247). On Stark’s wider account, culpability can also arise from failing to control pernicious character traits that are reflective of one’s agency (247–249). Accordingly, he concludes that culpability for negligence arises “when a defendant was possessed of the background beliefs and perceptions that could have led to the belief that a risk was present, but failed to form that belief—in circumstances demonstrating insufficient concern for others—because of an accepted facet of her character” (271–272).

Finally, Chapter 9 discusses the practical question of how Stark’s views can be translated into workable jury instructions or legislative guidance. This reflects Stark’s commendable interest in keeping his book relevant to practitioners.

There is much to admire in this thorough and insightful book. In the remainder of my review, I’ll raise concerns about one aspect of Stark’s project. While his sophisticated account of culpability without awareness in Chapter 8 deserves close scrutiny,Footnote 3 I’ll focus on something more straightforward: recklessness.

Recall Stark’s belief-based account of awareness, a component of recklessness. He’s right that awareness requires belief (perhaps merely a latent one) (91). However, Stark goes further and contends that “[t]he better account of awareness of risk … is in terms of the defendant’s beliefs about risk alone” (123, emphasis added). Stark thinks that belief is not only necessary for awareness, but that “simple belief that a particular risk exists … ought to be sufficient for a finding of awareness of risk” (129, emphasis added; see also 140).

However, this view seems problematic because there is a fundamental difference between awareness and belief: the former is factive, while the latter is not. It is widely agreed that “being aware that p” is a factive expression (i.e., entails the truth of the relevant proposition), while “believing that p” is not (see, e.g., Nagel forthcoming). A simple argument shows this. It sounds odd—self-contradictory, even—to say: “Alex is aware that p, but in fact p is false.” More to the point in discussing recklessness, it sounds odd to say “Alex is aware of a 30% risk that his conduct will harm someone, but in fact there is no such risk.” The best explanation of this oddity seems to be that awareness is factive. By contrast, it does not sound similarly odd to say “Alex believes there is 30% risk that his conduct will harm someone, but in fact there is no such risk.” This suggests that belief is not factive.

Thus, awareness has a property that belief lacks: factivity. This makes trouble for Stark’s view that “‘[b]eing aware that there is a particular risk attendant upon φ-ing’ is … co-extensive with ‘believing that there is a particular risk attendant upon φ-ing’” (92). Instead, it seems more plausible that awareness that p must also require, in addition to belief in p, the actual truth of p. This is the easiest way to ensure the factivity of the belief-based account of awareness.

Stark is sensitive to this problem, and suggests a way to address it. He proposes separating the subjective component of awareness (belief) from its objective component (truth). That is, there must be “a belief that p (a matter of mens rea), and, to result in a conviction, it must coincide with the fact that p (a matter of actus reus)” (131). He thinks it’s “far simpler, semantically, to refer to the ‘subjective’ mens rea element of belief, not knowledge, and leave the [issue] of whether the matter is true as an ‘objective’ element of the actus reus” (131). Distinguishing the mens rea and actus reus components of awareness might help preserve a belief-only account of awareness. As he summarizes it, the account of awareness that is best “in keeping with Anglo-American criminal law’s distinction between actus reus and mens rea is to require simply that the defendant believes … there is a risk attendant upon φ-ing. It is only where there actually is such a (unjustified) risk … that an actus reus will have occurred” (137).

This is an intriguing way to deal with the factivity of awareness, but two problems remain. First, it does not save the account of awareness in its own right. Stark often states that a belief that p is itself sufficient for being aware that p (92, 123, 129, 140). However, because of the factivity problem, this seems incorrect as a claim about the intuitive notion of awareness. Suppose I subjectively believe that serving coffee to my guests after 4 p.m. is dangerous to their health, but I do so anyway. Because my belief is false, it sounds odd to say I’m aware that serving coffee to my guests after 4 p.m. is dangerous. Thus, belief that a risk exists doesn’t seem sufficient for being aware in the intuitive sense that this risk exists. Stark’s account still seems to depart from the intuitive notion of awareness.

This is especially important because Stark was at pains in Chapter 3 to argue that, if the criminal law is to succeed in providing fair warning, “[t]he language of the criminal law should be accessible to those lay actors who [want] to avoid the pain of punishment” (72). The law should be such “that non-specialists can utilize it effectively at the stage of planning their conduct” (73). As Stark explains, the “need for clarity [in the language of the criminal law] is particularly vital … as misunderstandings can have severe consequences” (73). This suggests that there is a weighty reason for terms used in the criminal law to be given their common sense meanings where possible. Thus, Stark is committed to regarding it as a non-trivial cost if his account departs from the common sense, factive meaning of “awareness.”

This objection may not yet show that Stark’s account yields implausible results in particular cases. But the second problem might. The worry is that Stark’s view succeeds only if the statute defining the crime independently makes the truth of one’s beliefs about risk an element. Stark thinks awareness should “require simply that the defendant believes … there is a risk attendant upon φ-ing,” and we can leave it to the actus reus elements to ensure that the defendant is convicted only “where there actually is such a (unjustified) risk” (137). However, this won’t work if we are faced with a poorly drafted statute, which happens not to make the existence of the relevant risk an independent element of the crime.

Consider the Model Penal Code crime of making terroristic threats: “A person is guilty of a felony … if he threatens to commit any crime of violence … in reckless disregard of the risk of causing … terror or [serious public] inconvenience” (MPC § 211.3). Suppose Falecia has a terrible sense of humor and thinks it would be hilarious to call Warner Brothers Studios and threaten to “drug everyone there to death.” She does not intend to terrorize or inconvenience people, but genuinely believes there is a substantial risk that this will result from her prank. However, suppose her belief is simply false. Falecia’s voice sounds exactly like the adorable cartoon character Tweety Bird from the Warner Brothers TV show Looney Tunes. Thus, no one hearing her would understand her to be menacing in the slightest. Indeed, any listener on the phone would misunderstand her as saying that she wants to hug everyone there to death. Accordingly, there actually exists no risk that terror or serious public inconvenience will result from Falecia’s prank. Because recklessness requires awareness of the risk, if awareness is factive, then there would be insufficient grounds for convicting Falecia.

But Stark’s belief-only account of awareness goes the other way on this. On his view, Falecia faces conviction because she believed she was creating a substantial risk of terrorizing or seriously inconveniencing people. Moreover, and this is the crucial point, the statute does not obviously make the truth of Falecia’s belief an independent element of the crime. The statute only requires threatening violence “in reckless disregard of the risk of causing … terror or [serious public] inconvenience.” However, the phrase “acting in reckless disregard of a risk” is plausibly interpreted as a pure mens rea term. It is not obvious that “disregarding” some risk requires that it actually exists—only that one believes it does. If this is how to interpret the statute, then Falecia would face conviction on Stark’s belief-only account of awareness. The statute is not unambiguously worded to make the truth of Falecia’s belief about the risk a separate actus reus element. Accordingly, for sub-optimally drafted statutes like this one, Stark’s account threatens to impose liability on those with false beliefs about the risks they are creating. There is at least a case to be made that this is overly harsh.

Of course, this result could be avoided if the statute were drafted differently. Suppose it said: “A person is guilty of a felony if she threatens to commit any crime of violence, and this threat creates a substantial and unjustified risk of causing terror or serious public inconvenience, but the person recklessly disregarded that risk.” This version of the statute would unambiguously prevent Falecia’s conviction, since it requires as a separate element that her beliefs about risk are true. To succeed across the board, Stark’s account requires that all statutes be drafted in this way—i.e., as making the truth of one’s beliefs about risk an independent actus reus element. But this can’t be taken for granted in all statutes.

Note one point in closing. Whether one’s beliefs about the risks of one’s conduct are true very likely does not impact the criminal culpability of one’s conduct. On many views, culpability is a subjective notion that tracks how one believes the world to be when acting, not whether those beliefs are true.Footnote 4 Very likely, Falecia who honestly but falsely believes she is creating a risk of terrorizing people is just as culpable as her otherwise identical counter-part, Trudy, who has true beliefs that she imposes such a risk.

Thus, it plausibly makes little difference to culpability whether Stark is right that recklessness requires only believing that one imposes substantial risks (which happen to be unjustified), or whether recklessness requires the true belief that one imposes such risks. Nonetheless, it makes a difference to how the legislature should draft statutes. Even if the legislature would be entitled on culpability grounds to criminalize not only the Trudys of the world (with true beliefs about risk), but also the Falecias of the world (with false beliefs about risk), the legislature very well might have good reasons not to go as far as they could (e.g., because of compassion, generosity or political prudence). Instead, the legislature might reasonably want to criminalize only the Trudys, but not the Falecias. However, if Stark’s belief-only account of awareness is correct, then to accomplish this result, the legislature would have to be much more careful in drafting criminal statutes. Specifically, it would have to formulate statutes to unequivocally make it an independent actus reus element that the defendant’s beliefs about risk are true. If Stark’s belief-only account of awareness is correct, the legislature cannot rely on the mens rea of recklessness to ensure that only the Trudys of the world, but not the Falecias, are convicted of recklessness crimes.

Stark’s belief-only theory of awareness would have real-world effects if adopted. Therefore, to avoid punishing Falecia as a reckless actor, it seems preferable to construe awareness of a risk as the true belief that the risk exists.Footnote 5 The factivity problem thus gives reason to rethink Stark’s belief-only account of awareness.