Introduction
Sociolegal scholars have long been interested in legal knowledges. Or as the title of a well-known collection puts it: How Law Knows (Sarat, Douglas, and Umphrey Reference Sarat, Douglas and Umphrey2006). However, other adjacent fields like science and technology studies (STS) have moved beyond these important but classically epistemological questions by taking an “ontological turn” in their analytical focus. Woolgar and Lezaun (Reference Woolgar and Lezaun2013, 322) describe the ontological turn taken in STS as another “attempt to apply its longstanding core slogan—‘it could be otherwise’—this time to the realm of the ontological” (see also Kohn Reference Kohn2013; Pickering Reference Pickering2017; Lancaster and Rhodes Reference Lancaster and Rhodes2020). In other words, these scholars show how things typically taken for granted as innate, natural, or immutable are not actually set in stone; many of these things are contingent, if not explicitly multiple, meaning they can be enacted in different and equally true ways (see Mol Reference Mol2002).
Despite sociolegal studies’ parallel pursuits, it often continues to focus predominantly on sites of knowledge and knowledge production, such as the processes that turn knowledges into legal evidence or legal facts, but not so much on the status of the objects that law claims to know. In this paper, we begin to take the ontological turn into the courtroom, shifting our gaze away from strictly epistemological inquiries, such as which/whose knowledge matters, and instead outlining the legal importance and effects of which/whose reality matters. In so doing, we see that much like the plurality of legal knowledges previously uncovered and accounted for, there exists a plurality of ontologies for many legally-relevant objects. At the same time, we begin to document how criminal courts deal with this multiplicity when rendering decisions, for example, by opining on a particular ontology or by finding ways to avoid the ontological question altogether.
By taking the multiplicity of legal objects seriously, we show how an ontological turn can lead to fruitful sites of sociolegal inquiry, including in areas that might have previously been thought to be picked clean of novel insights. More specifically, we highlight how this led us to a yet untheorized tool of judicial discretion: what we are calling, broadly, adjudicators’ “ontological discretion.” Below, we use three examples from our own research – sleep, death, and intoxication – to document their multiple realities and the effects of this ontological discretion in the criminal law.Footnote 1
Framed as another tool in the adjudicator’s discretionary toolkit, we argue that the act of wrestling with the multiple ontologies of “an” object, at least in some cases, becomes a necessary task if a coherent verdict, judgment, or decision is to be rendered. In addition to documenting the effects of this judicial tool, the examples we provide indicate that it is often not the case that the “true” reality of a relevant object is sought or chosen by the courts. Rather, it appears judges implicitly recognize the multiplicity of objects, and in turn prefer to keep their ontological status open, generally, even if they might have to close it in the context of a single case. This allows adjudicators to maintain their discretion for future decisions, and thus their ability to pick and choose the reality that best suits the case facts—or their preferred verdict. While we recognize our chosen examples may not be objects in the conventional sense that tables and chairs are objects, part of what we show is that courts do treat (and perhaps have to treat) even liminal conditions or concepts as things to be reified. However, here we can only begin the task of accounting for ontological discretion at the adjudication phase of criminal law.
To state the obvious, discretion in its most literal form is inherent in any contested trial decision. Whether a case has been made out beyond a reasonable doubt is always a judgement call, almost by definition—no calculus or objective measure of reasonable doubt has ever survived the common law courts, even with the supposed actuarial turn documented by scholars of the new penology. Our analysis is therefore relevant even to legal scholars of traditional discretion: if we want to understand how judges come to and then justify their discretionary decisions, it may require us to contend with the various ontological games they play, which we begin to outline below.Footnote 2
As a result of the preliminary nature of this inquiry, our aim is primarily empirical; we prefer depth over breadth in our present analysis, documenting the presence and effects of ontological discretion in individual criminal cases over opining on its overall legal relevance or widespread generalizability. In favouring site-specific inquiry, we maintain both our commitment to the ontological turn and our preferred sociolegal method by “avoid[ing] the reification that is inherent in the use of the noun ‘law’” (Valverde, Levi, and Moore Reference Valverde, Levi and Moore2005, 91), instead confining our analysis to one “legal complex”—the Canadian criminal courtroom (on legal complexes, see Valverde Reference Valverde, Sarat, Douglas and Martha2007). In keeping with those commitments, our analysis is also more descriptive than prescriptive; we are purposefully silent on whether ontological discretion is “good” or “bad,” but—and in line with past sociolegal analyses of judicial discretion—we presume it can and will be both (for example, Provine Reference Provine1998).
Taking a (Sociolegal) ontological turn
In the most famous work on the ontological multiplicity of supposedly singular objects, anthropologist Annemarie Mol (Reference Mol2002) convincingly showed, through detailed ethnography in a Dutch hospital, how different medical actors enact different yet equally real atherosclerosis.Footnote 3 There is not one universal atherosclerosis to which medical actors have differing perspectives; the particular ontology of a particular atherosclerosis depended on a variety of factors, but mainly on who and/or what was looking at “it” (see also Barad Reference Barad2006; Law and Mol Reference Law, Mol, Knappett and Malafouris2008; Hadders and Hambro Alnæs Reference Hadders and Hamrbo2013; Van Oorschot Reference Van Oorschot2021).Footnote 4 The key revelation was that objects once thought to be singular—like diseases with a supposedly fundamental etiology—are really multiple, or that different objects with the same name come into existence via different sets of practices. In a similar vein, we intend to show how conditions that matter a great deal for criminal/legal purposes—sleep(s), death(s), and intoxication(s)—can appear in the criminal courtroom as multiple.
However, our goal is not to simply “do more multiples,” a la Mol, for example, by showcasing all the different ways these objects are enacted. Nor is our project solely aimed at documenting how law and legal procedure can themselves shape an object’s ontology (for example, Sudai Reference Sudai2022).Footnote 5 Given the well-documented multiplicity of objects in STS, we are more interested in what happens after specific, locally contingent, and perhaps implicitly political ontologies have been enacted and transported. That is, we are interested in the effects precipitated once different multiples with the same name meet in the courtroom. It is this meeting between multiples that may necessitate an exercise of ontological discretion, and that process ends up having real-world effects (it is not just an ontological game). We therefore prefer to highlight some of those downstream effects, and less so the processes and relations that led to the multiplication itself, as an actor-network theorist might (for example, Cloatre and Cowan Reference Cloatre and Cowan2024).
Contributing to sociolegal work on discretion using the analytical framework of ontological multiplicity, our research program can now be stated simply: rather than doing multiples, we try to uncover what courts and other crucial legal actors do in cases where competing ontologies have been made available, either in the court process or in the wider cultural field reproduced in judicial arenas. Classically, this becomes another way of approaching “law in action.”
Just as the traditional judicial discretion studied by legal scholars has certain effects, such as deciding what weight to give to various mitigating or aggravating circumstances in criminal sentencing, so too the game of ontological discretion has real consequences.Footnote 6 To highlight these effects, we build on previous theories and research programs related to the ontological turn and multiplicity (Mol Reference Mol2002; see also Mol Reference Mol1999; Law and Urry Reference Law and Urry2004; Woolgar and Lezaun Reference Woolgar and Lezaun2013; Lancaster and Rhodes Reference Lancaster and Rhodes2020; Flacks Reference Flacks2024), including research that has been used for sociolegal analysis (Quinlan Reference Quinlan2020; Van Oorschot Reference Van Oorschot2021). Even sociolegal scholars primarily interested in processes of legal knowledge production—especially those borrowing from the analytical toolkit of actor-network theory (for example, Valverde Reference Valverde2003; King, Shaw, and Kennedy Reference King, Shaw and Kennedy2023)—implicitly recognize ontological multiplicity when claiming that a legal object or piece of evidence helps “produce the world it purports to describe” (Best and Walters Reference Best and Walters2013, 348; see also Law and Urry Reference Law and Urry2004).Footnote 7 Two decades ago, Levi and Valverde (Reference Levi and Valverde2001) were already arguing that since sociolegal scholars had “available to them quite a number of studies documenting these different forms of knowledge and their relationships to the legal complex,” it “may be time to focus on the ways in which the competition between these forms of knowledge is dealt with in law” (820). Given the success of the ontological turn in other disciplines, we feel it is high time to examine how these competitions between what we might now call “forms of reality” are dealt with by legal actors.
While speaking of multiple realities may seem strange to some, even the most hardened physicist claiming to study the fundamental nature of reality must deal with ontological multiplicity in their daily life.Footnote 8 In its most banal form, the table in the physicist’s office is both really a table (a piece of furniture) and really a collection of individual particles and atoms that might be said to constitute it; whether in any given moment the table is enacted as a piece of furniture or as the atomic particles will depend on the context, and in turn demand or produce different ways of knowing. We can talk rationally about the table being objectively solid, as being a certain color, and as being made of a specific material. However, such statements may lose their relevance and possibly cease to make sense at the level of the individual particles, even if the physicist might say that there is no contradiction since the particles are what make the table up (and its solidity, color, and texture).
Our point is: the discursive possibilities that different enactments allow will also have legal effects. In fact, as we show below, criminal case outcomes may hinge on this often ontological—and often discretionary—exercise. Continuing the analogy, if the table is an entity in an estate and there is a dispute regarding the will, then a certain reality, the “common sense” (or macroscopic) one, will likely be adopted. Expert evidence showing that what we think of as a solid table is actually almost all empty space between particles or, microscopically and ontologically-deeper still, some form of quantum probability cloud, will not be deemed relevant; such is a case where the fundamental multiplicity of the table would cause little issue. But no judge in an estate dispute would say the solidity of the object that is part of the estate refutes the physicist’s view of the (no longer-) “same” object. Even after being enacted as a piece of furniture, the table remains multiple and can be acknowledged as multiple if discussions arise between different ways of talking about—yet again in the now tired but necessary scare quotes—“it.” Again, the key point: “it” remains “both.”
In another sense, however, the existence of ontological discretion in our context may seem obvious to other legal scholars. After all, the very foundation of the common law—that parliament expressly grants judges an extensive right to interpret law—requires the court process to instill certain legal concepts with meaning. But what is less obvious, and what requires the empirical research we begin to present, is that even despite any purported deference to precedent and doctrine, courts rarely appear to choose one ontology over another, once and for all. Instead, they often find ways to keep the ontologies of key legal objects and concepts open and undetermined, maintaining this discretion and uncertainty for future cases. Therefore, it is likely more accurate for us to say that courts “collect” relevant, plausible ontologies and then draw on specific ones as needed. Taking a largely pragmatic view of law and legal processes (see Valverde Reference Valverde2005), courts need not decide once and for all how many angels can dance on the head of a pin; they can and do in fact choose one angelic ontology for a certain purpose in a particular case but leave room for other possibilities, including quite contradictory ones. In short, the qualifying verb we prefer is “use”: the courts use ontologies.
Extending this type of inquiry into the courtroom allows us to better understand how conflicting ontologies create different conditions of possibility, facilitating the documentation of how and when legal actors directly deploy explicit or implicit ontological arguments, often (but not always) in support of a particular outcome or objective; the table can remain a table in the estate dispute, but the particle-theory of tables can still be enacted if adjudicating on a patent for some new material/method relevant to table-making. While this hypothetical example is more uncontroversial, partially because neither party need contest the ontology and since the different objects fit neatly into different doctrinal boxes, our empirical examples below are chosen precisely to show that the courts’ ontological maneuvers can vary even within the same legal complex and with the same object (for example, two murder cases discussing the time of death, and occurring in the same general jurisdiction at the same general time).
However, having to choose between different possible enactments is not the only relevant discretionary move illuminated by our ontological inquiry. Other effects and possibilities also become evident if the courts are able to avoid the ontological question altogether. One example not covered here but familiar to many legal scholars concerns sexual orientation: in the US and Canada, courts quickly moved away from the legally fruitless inquiry into the biological reality of sexual orientation; without any fanfare, sexual orientation became a legal object with a meaning that was taken as “within judicial notice.” Importantly, this does not negate the fact that sexual orientation can have multiple ontologies. Rather, it allows the court to avoid specifying any singular ontology, and, thus, they need not even close the ontology in a single case.Footnote 9 The act of sifting through multiple, possible ontologies and choosing one, or finding a way of refusing to choose among alternatives, is precisely what we call ontological discretion.Footnote 10
With the above clarifications and caveats in place, we can now turn to documenting our real-world examples from criminal cases involving three legally-relevant objects: sleep, death, and intoxication—again, cases and objects that have been selected so one can readily see the workings and effects of ontological discretion at play.
Asleep and awake: The legal effects of sleep(s)’s multiplicity
In this section, we show how the seemingly stable, well-known thing called sleep, which tends to be taken for granted in everyday contexts where sleep is regularly contrasted with wakefulness, is actually multiple, and that its multiplicity can demand the exercise of ontological discretion within the courts. This example stems from one of the author’s past research into the invention of the electroencephalograph (EEG)—a still-prominent brain scanning device that graphically records electrical activity in the brain—and the technology’s immediate merger with understanding the neurological processes related to sleep with criminal responsibility. In part, what we show here is that in a case where an accused claims they were sleepwalking during the criminal event, an EEG recording is only legally relevant if sleep can be ontologically enacted or represented as a specific cycle of brain wave patterns, and that this enactment was itself only possible after the EEG made it visible and thus knowable.
After the emergence of the EEG in the mid-twentieth century, doctors and medical researchers quickly began to reify sleep as a pattern/cycle of electrical activity in the brain (along with other physiological cycles) (Dynes and Finley Reference Dynes and Finley1941; Kohler and Kurz Reference Kohler and Kurz2017). Although we may still talk about sleep in different real yet more unscientific ways, that we now think of sleep as really—or fundamentally—a heterogeneous biophysiological cycle that includes various stages like REM (Rapid Eye Movement) and delta wave (deep) sleep, named for the long, slow waves that appear on an EEG record, has become almost common knowledge, especially to anyone who as ever undergone a sleep study or who suffers from sleep apnea.
As outlined by legal historians, whether an accused was sleeping/sleepwalking during the crime has been a relevant question for centuries, long before the advent of the EEG (Eigen Reference Eigen2003). However, in the years immediately preceding the EEG’s reconstitution of sleep via the brainwave, the problem of proving someone was actually asleep was articulated in the 1951 case R v. Kasperek.Footnote 11 As the judge explained: “The only conceivable defence to the presumption mentioned, other than insanity or drunkenness, might be that at the time of the commission of the offence the prisoner was walking in his sleep or otherwise acting with his mind a total blank. It is difficult to imagine how such a defence could be proved” (9).
In one sense, this was a curious remark, since, as mentioned, there was precedent for sleepwalking-related defenses throughout the history of the common law. Yet in another sense, the judge was also correct; it was only after the EEG had turned the question of “is someone actually sleeping” into one ostensibly knowable via an EEG recording that a certain kind of quasi-scientific, biomedical proof could be offered. And although it may sound like we are reverting back to old-fashioned epistemological questions here, it is precisely these issues that warrant our ontological turn: it was only possible for sleep to become knowable in this way when it was multiplied (enacted) to include something else entirely, a totally different thing yet a thing also called sleep.
In R v. Parks, our first empirical case example documenting the legal effects of ontological multiplicity and one of the most infamous, hotly debated cases in contemporary Canadian jurisprudence, the accused was acquitted of the murder of his mother-in-law and the attempted murder of his father-in-law despite there being no question he was the one who brutally beat and stabbed them.Footnote 12 Even though he drove over twenty kilometres to his in-laws’ apartment before parking his car, entering the home, and procuring a knife from their kitchen, and despite the Crown prosecutor pleading to the members of the jury to use their “common sense” in rendering a verdict, the jury accepted the defense’s theory that Parks was sleepwalking throughout the entire ordeal and thus was not conscious of what he was doing. Hence, they found him not criminally liable and entitled to a full acquittal, a verdict which the Crown unsurprisingly appealed.
As described in the appellate proceedings and extensive media coverage, Parks’ defence seemed to succeed largely on the evidence of various sleep studies where his brain waves were recorded via an EEG, and which apparently showed he was physiologically prone to sleepwalking. One newspaper summarized the medical evidence (and, unknowingly, the various ontologies of sleep) as follows: “A sleepwalker’s brain tracings in deep sleep reveal blips not seen in normal subjects. Sleep specialists call these anomalies ‘arousals,’ but the term is misleading. An observer would see only someone sleeping quietly” (emphasis added).Footnote 13 The subtext was that our commonsense notions of sleep could now mislead us. If sleep were defined as a more commonsense, subjective state of rest/consciousness, then it would be hard to fathom any person performing such tasks while asleep—driving on a highway, parking in a garage, entering a person’s home, etc. In that instance, the EEG recordings would likely also become irrelevant, given that the ways of knowing if a person was asleep would rely on different logics (for example, eye-witness observation versus expert evidence).
Thus, given the alleged abnormalities in the accused’s EEG, the defence attempted to enact sleep and sleepwalking as a specific electro-physiological brain wave cycle. At trial, the defense’s lead sleep expert was even permitted to give a makeshift lecture on the science of sleep, complete with slideshow images of EEG recordings, including those of the accused that were taken during his various sleep studies. When describing one exemplary slide, the doctor stated the brain wave patterns represented a two-minute sleepwalking episode where the person “got up in slow-wave sleep, wandered around, got to the end of the electrode cable, yanked it, went back to bed, and remained in non-REM sleep.” In case there was any doubt, they were really asleep, the expert assured the courtroom: “You can see that in the brain waves” (Callwood Reference Callwood1990, 264). He then described an episode where a sleeping child—that is, a child whose brain waves indicated they were asleep—was able to verbally respond to questioning. As a result of this framing, the use of EEG recordings was almost tautological in the defense’s arguments: persons showed particular brain wave patterns because they were asleep, and they were said to be asleep because they showed those particular brain wave patterns.
As an Ontario court explicitly noted in a later case (discussed below), even an EEG study that said the person was prone to sleepwalking could not tell the court if the person was actually sleepwalking at the relevant time of the crime. It is of course true that EEG evidence alone could never be logically sufficient to justify a verdict; however, and related to our commitment to documenting the effects of multiplicity, it did create new possible arguments that the defence could then mobilize. For example, since Parks’ brain wave tracings classified him as a sleepwalker, it became more plausible he was sleepwalking during the attack, despite commonsense assumptions about the implausibility of driving while asleep (which is obviously different than falling asleep while driving). This argument would likely not have been possible if sleep had never been multiplied via the brain wave, nor if Parks had not shown the supposedly characteristic brain wave pattern of sleepwalkers. Regardless of one’s own stance, the medical evidence clearly persuaded the jury, given their “not guilty” verdict.
Since this was a trial by jury, we do not pretend to know the ontological preference of the trial judge in reference to the verdict. Therefore, the appellate casebook more clearly shows the effects of the multiplicity of sleep among adjudicators, including the most obvious effect within the appeal: since Canadian appellate courts do not re-adjudicate cases or hear new evidence (they only hear claims of judicial error), the three justices on the bench had to accept the medical definition of sleep since no expert evidence directly countered it at trial, even if they were clearly uncomfortable with the legal conclusions it led to. Thus, the justices had to accept statements like “Wakefulness, REM sleep, and non-REM sleep are the three basic electro-chemical configurations of the brain, and the two forms of sleep are as different from each other as they are different from wakefulness.” They also expanded on this in their notes regarding the defence experts’ testimony about deep sleep, which was said to be the stage of sleep where sleepwalking events predominantly occur: “Sleep is periodic, going through cycles of approximately 90-100 minutes duration. Most deep sleep normally occurs in the first two cycles of sleep.”
This particular ontological enactment—that sleep is a cycle of brain wave patterns—informed the types of legal arguments that could be made on appeal. The question of whether Parks was asleep could not be reopened given the jury verdict. Thus, the Crown’s argument was that the trial judge erred in not putting the question of insanity to the jury, despite this going against centuries of common-law precedent. In Canada and most jurisdictions that rely on the infamous M’Naghten rules of insanity, an insanity claim (now known as “not criminally responsible on account of mental disorder”)—whether put forth by the Crown or defence—requires that a person possess a “disease of the mind,” and that at the time of the crime this disease eliminated their capacity to know what they were doing or that it was wrong. In support of sleepwalking being a type of legal insanity, the Crown also had to mobilize the medical ontology, arguing that the common-law precedent that a successful sleepwalking defence necessitated an acquittal was based on a past naïve and unscientific understanding of sleep.
The Crown’s legal argument was that since Parks had an abnormal brain wave pattern he also had an “abnormal mind,” and since he was asleep during the attack and could not have been conscious of what he was doing (now by definition), he should have been found legally insane and placed in a psychiatric institution. However, the defence countered not only that sleepwalking was not a disease in the medical sense of the term, but that it also did not mean his mind was impaired, legally; rather, it meant his mind was totally “absent.”
Whether a sleepwalker’s mind is impaired or totally absent (for example, “turned off” or non-existent), therefore became the focal question for the appellate justices, who went so far as to demand a “definition of the mind,” which both parties struggled to provide. Yet in support of the initial acquittal, the defence tried to equate the mind directly with the brain by citing medical evidence that “The brain does not function in deep sleep and sleepwalking. So that’s why there is no mind” (emphasis added). Even when the justices quickly pointed out that some part of Parks’ brain must have been functioning since he could breathe and apparently even drive a car, the defense again countered that the evidence indicated the part of his brain that was functioning was not the “thinking” brain, which they then equated with the mind (Callwood Reference Callwood1990, 339). In the end, and based on the medical ontology of sleep accepted at trial, the acquittal had to stand.
At this point, we must again emphasize the appellate court’s ontological discretion may have been limited by the way sleep was enacted at trial, and given the widespread acceptance of the medical evidence, one might even think of the ontology of sleep as being “closed” by this case. However, in testament to the notion that courts seem to desire to maintain the ontological openness of their objects—in part to avoid the issues faced in Parks (for example, in the interest of public safety, the courts did not want a person prone to violent episodes while sleepwalking roaming the streets)—this was quickly addressed by the Supreme Court of Canada (SCC).
In the very next sleepwalking case to make its way to the SCC, the court ruled that whether a given condition was a disease of the mind (mental impairment) for the purposes of insanity should be strictly predicated on questions about public safety, and not at all on the ontological status of a given condition and/or its association to the mind (however defined). If the condition was abnormal and likely to reoccur or otherwise put the public in danger, then it should be considered a disease of the mind even if, just like with sleepwalking, every doctor who testified did not view it as disease in any medical sense of the term. Thus, lower courts could now avoid the seemingly necessary legal effects that came with having to adhere to a particular ontology.
As we see in R v. Luedecke, another sleepwalking case that made its way to the Ontario Court of Appeal almost two decades after Parks, the court was able to state that the “neurological evidence demonstrated conclusively that the respondent was among the small group of adult persons who experience parasomnia [sleep disorder]” while at the same time quashing the initial acquittal in favour of an insanity verdict, since, in part:
The trial judge also failed to appreciate that Dr. Shapiro’s medical opinion that parasomnia did not constitute a mental disorder was largely irrelevant to the determination of whether, for policy reasons, the condition should be classified legally as a disease of the mind. These errors led to a failure to apply the proper legal standard when characterizing the respondent’s automatism. (para. 103).Footnote 14
Furthermore, despite the judges in Luedecke also referencing the medical ontology, albeit (unlike Parks) to now support an insanity verdict, here we stress that the brain wave and sleep cycle did not replace all older, more traditional ontologies of sleep; rather, and crucial to the existence of ontological discretion, it multiplied the object (now read: objects) in question: sleep… or the mind… or perhaps a lack of rational consciousness. This is clearly seen in R v. Teepell, when the trial judge appears to accept the multiple ontological framings of sleep while still only drawing from the one that seemed to best justify their verdict.Footnote 15 Although they accept the expert EEG evidence proffered by the defence that showed Teepell had neurological tendencies found among sleepwalkers, they were able to convict the accused (even rejecting insanity) by adhering to the more traditional ontology of sleep, which was based on direct, commonsense evidence that indicated he was not asleep during the relevant time. While not dismissing the reality of the electro-physiological ontology, the judge cited the Crown’s sleep expert when agreeing that “sleep studies done several years later cannot tell you whether or not someone was sleepwalking on the night of the incident. The only direct evidence of whether someone is sleepwalking is his behaviour, observed or inferred, during the episode” (para. 188). Going a step further, in their written decision the judge even found that “the use of the sleep studies is meaningless. Actually, it is something distracting from the direct evidence” (para. 210).
A similar move was made in the more recent case of R v. Antonius.Footnote 16 After hearing expert evidence and accepting the defence’s diagnostic opinion that the accused suffered from “sexsomnia” – a condition where a person performs sexual acts while asleep – the accused was convicted since the specific sexual event in question could be discounted as not being a product of his condition. Seemingly once again accepting the multiple realities of sleep while still invoking the reality that best justified a particular verdict, the judge again based their decision on the eyewitness evidence and commonsense assumptions about how awake and asleep people behave. Their reasoning is worth quoting at length:
In my view, the accused’s interactions with Ms. K. [the alleged victim], unlike those with Ms. L. [a previous partner of the accused, with whom the judge accepted the accused engaged in unconscious, sexsomnic acts], appear somewhat stealthy. Undressing Ms. K. and initiating sexual intercourse with her without moving her and without awakening her implies a cautious approach and a conscious purpose, quite unlike the essentially unorganized interaction described by Ms. L.
I accept that those experiencing parasomnia episodes can perform relatively complex tasks. However, the accused’s interaction with Ms. K. seems more than simply complex. It appears to involve consciousness of the context in which it was occurring. (paras. 98-99).
In justifying the conviction, the judge did not reject the defence’s medical evidence or their ontological framing of sleep. They simply invoked arguments based on that other ontology of sleep, the one of observed behaviour and common experience. As the judge accepted, various experts testified that extremely complex events, even ones that appear purposeful, can be performed by persons in parasomnic states. But in a rather brash exercise of their discretion, the judge selected the rules of their ontological game that led to the verdict, rules that meant they did not even have to address or invoke the Crown’s sleep experts who refuted the defence’s claims from within the more scientific ontological frame:
My conclusion that the accused has not rebutted the presumption of voluntariness does not rely primarily on the expert evidence of Dr. Pressman, Dr. Ceresney, or Dr. Fleming. While all three experts provided helpful insights into the nature and characteristics of parasomniac behaviour, my conclusion that this was not a parasomnic event rests on finding that the accused’s evidence is not credible in light of the evidence considered as a whole. (para. 151).
When multiples do not hang together (Mol Reference Mol2002) – for example, when evidence given in one ontological frame contradicts evidence given in another – a judge can instill the rules for their ontological games by claiming one ontology’s evidence is stronger, while still not rejecting that the other ontology and evidence is true and real.
Before we conclude this section, we should clarify one final point: ontological discretion is not the same as ontological relativism. Courts do not seem able to construct or affirm just any old ontology. For an ontology to be accepted, it still has to be consistent with a general logic or description of the world that is accepted as the truth in some “community of practice” (see Bowker and Star Reference Bowker and Star1999). For example, in Parks, the appeal justices did not have (or we should say, no longer had) the discretion to enact sleep as “the soul leaving the body,” which is what the Crown prosecutor claimed sleep used to be defined as throughout the common-law world. Perhaps paradoxically, the prosecutor referenced that older, no longer applicable ontology for the sole purpose of arguing that our increased scientific knowledge indicated sleepwalking was in fact a disease/disorder of the mind, and that it could be sufficient for insanity. At any rate, in Canada in the 1990s, it is clear souls were not valid legal objects the way they were in the days of the Domesday book. But at the same time, the court did have the discretion to enact the mind as the cortex (a physiological entity) or – as the other cases make apparent – as a set of psychological faculties, and thus sleep itself as a specific electrical state in the brain or the subjective phenomenon predicated on an only inferable lack of consciousness.
The choice of object, each produced by a different community of practice, had direct implications on whether sleepwalking was or was not an impairment of the mind, in the sense that such impairments are enacted for purposes of criminal responsibility. Whether this or that legal actor happens to be an ontological relativist in their spare time, the courts act, and perhaps have to act, as if reality were multiple. Dealing with the ontological question may therefore be a necessary step in certain cases before a coherent judgment can be rendered.
Dead and alive: The legal effects of death(s)’s multiplicity
Our capacity to modify and control vital aspects of human biology has led some scholars to claim we are living through a “politics of life itself” (for example, Rose Reference Rose2007). This form of biopolitical power is “shifting the boundaries between life and death,” leading to a lack of “clear-cut distinctions between living and dying” (Braidotti Reference Braidotti, Coole and Frost2010, 201, 203). Thus, oft-taken-for-granted questions like if someone is dead or alive are rendered more overtly contextually contingent, political, and capable of being answered in different yet equally true ways.
As we will now show, the multiplicity of death also creates ontological decisions for the criminal courts. In the civil law, this has already been empirically documented by STS-adjacent scholars in the medical context of “brain death,” which some argue was a form of death enacted with a specific historical goal in mind: to close the myriad of practical, political and legal concerns related to the development of organ donation and transplantation (Lock Reference Lock2002; Hadders and Hambro Alnæs Reference Hadders and Hamrbo2013). If death remained solely or even predominantly defined in health contexts by more traditional means, such as the cessation of the heartbeat and/or respiration, patients and organ donors being kept “alive” by machines such as ventilators or iron lungs would logically have to be “killed” by doctors so that their organs could be harvested and transplanted to another (more alive?) patient.
For practical medical purposes, including to avoid legal liability, doctors seem to want a definition of death that allows them to literally pull the plug on patients who are still breathing, even if only (but not always) artificially, namely patients whose bodies still look very much alive on the hospital bed to their friends and families (warm to the touch, stomachs rising and falling with every breath, etc) (see Hadders and Hambro Alnæs Reference Hadders and Hamrbo2013). The simplest way around this now medically untenable yet perhaps more visceral life-and-death distinction is to enact a new form of death located entirely within the brain, one that becomes less obvious to the friends and families of the deceased but grants more authority to medical professionals.
Just like the modern sleep cycle discussed above, the possibility of enacting a brain-based death was aided by the invention and proliferation of the EEG. The representational simplicity of the new death that this technology allowed – in theory, a flat line on a brain wave graph – meant doctors could appeal to the same logics, tools, and symbols as more traditional ontologies, no doubt increasing the success of the brain death criterion. Similar to asystole death – the technical term for the cessation of the heart but likely more recognizable as the flat-line of the electrocardiogram seen and heard (beep… beep… beeeee---) in every modern medical drama – after the EEG’s invention it also became possible for the brain to flat-line.
A brain-based definition of death was recently adopted by various Canadian medical professionals and associations, where death is defined as “the permanent cessation of brain function, observable by the absence of consciousness and brainstem reflexes including the ability to breathe independently” (Shemie et al Reference Shemie, Wilson and Hornby2023). They are also continuing their attempt to make the brain-based definition the sole national standard across a variety of contexts, at least in their official rhetoric; in a recent interview, Doctor Sam Shemie, the lead author of the new brain death criterion, stated this definition should apply “to all people in all circumstances” (Landry Reference Landry2023). However, before this definition of death was agreed upon, Shemie recognized that even “medical practice [referred] to two distinct forms of death – brain death and cardiac death […].” For Shemie, the need for a universal brain-based standard “comes down to one simple truth: the only organ that we can neither artificially support nor replace is the brain. And when the brain completely stops working, the person is dead” (Landry Reference Landry2023).
After the release of the new medical guidelines for brain death, Chandler and Pope (Reference Chandler and Pope2023) reviewed all the Canadian legislation pertaining to death, noting how “the law is quite flexible and can tolerate varying approaches to the legal definition of death for different purposes” (660). In medico-legal contexts, particularly within civil or administrative law, the individual Canadian provinces which have jurisdiction over healthcare matters continue to operate with different definitions, with some provincial statutes locating death solely and completely within the brain and others merely including this as one possible way a person can be declared dead. Yet some provinces still refuse to include any statutory definition of death, leaving those decisions to specific medical or legal actors. The most recent Canadian civil appellate court to deal explicitly with the definition of death seemed to also overtly accept death’s multiplicity, declaring that the “current state of the common law is that a person is considered dead where there is either the irreversible cessation of cardiorespiratory function or the irreversible cessation of all brain function.”Footnote 17
In this example, the criminal courts – and the criminal law itself, as handed down by Parliament – operate more like the civil law and those latter provinces in that they continue to reject any singular, stable definition of death. Even within the criminal context, death remains multiple. And just like the enactment of the sleep cycle discussed above, the ability to enact different deaths has serious legal effects.
One example depicting the existence and effects of death’s multiplicity and the criminal courts’ ontological discretion can be seen in the 1988 British Columbia murder case R v. Green. Footnote 18 Green’s defence was that his co-accused was solely responsible for the victim’s death since the co-accused was the one who fired the first gunshot at the victim’s head, at which point the defence claimed the victim would have instantly become brain dead even if his heart continued beating. If it could be said/proved that the victim was dead immediately after the first gunshot, any subsequent shots fired by Green logically could not have contributed to the victim’s death. Various doctors did in fact testify in support of Green’s theory that the victim would have met the medical criteria for brain death after the first shot, while simultaneously testifying that it was medically possible for the victim’s heart to continue beating for up to 20 minutes after they “died,” as the Crown evidence claimed.
Although this was another trial by jury, in this instance the judge had to directly address the potential problems posed by competing definitions of death if the jury was to render a coherent verdict: “If [the victim] was already dead when the first of these two shots was fired then Green cannot be convicted of murder except by virtue of the application of Section 21(1), which the Crown relies upon in its theory that Green aided and abetted Harrison in the killing. Thus it becomes crucial for the jury to know, as a matter of law, when death can be said to occur” (2). In other words, if brain death were accepted then Green could only be guilty of murder by proving he aided or abetted the co-accused, and not that his gunshots helped cause the victim’s death.
In this case, the definition of death was a question for the trial judge, not the jury. The defence therefore asked the judge to accept “the wisdom of both the medical profession and the Law Reform Commission on this subject and to instruct the jury that death takes place when a person becomes brain dead in the medical sense of the term” (2). Not surprisingly, the prosecution countered that “the traditional approach to the question should prevail, namely that life continues until all vital functions of the human body have ceased to operate” (2). Thus, if these two definitions were both really death and were both really at odds, then a person could legally, if not literally, be alive and dead at the same time – much like Schrodinger’s famous cat.
In exercising their ontological discretion, the judge in this case “adopted” (their word) the traditional definition offered by the Crown, allowing Green’s subsequent shots to be a contributing cause in the death of the victim. While the court’s ultimate decision is itself worth noting, what we want to stress here is the fact that it was up to the court to decide this question in the first place, even in spite of past precedent already constituting death as brain death (discussed below). Importantly, they were not answering this question for all times or all cases. The need to justify their definitional decision while also protecting the court’s future ability to enact different ontologies was implied in the judge’s written reasons, which we quote at length:
The suggestion that brain death or the irreversible cessation of brain function be the legal standard for determining when death occurs may be suitable in the medical context and even in the civil law context, but in my view it is a completely impractical standard to apply in the criminal law. The present case is a good example of why it is not feasible to apply such a concept to the criminal law. If the onus is on the Crown to satisfy this jury beyond a reasonable doubt that Mr. Frie [the victim] was still alive – that is to say that his brain function had not yet irreversibly ceased – when the Accused Green ‘pumped’ two bullets into him, it can be seen that such an onus would be impossible to discharge – unless someone had happened along with an EEG monitor and applied same to Mr. Frie either before or immediately after the two shots allegedly fired by Green. (2).
Perhaps one might read this statement as indicating the judge had no discretion to enact the brain death criterion and only the ability to say death is not brain death, in turn forcing them to adhere to the more traditional common law notion. Furthermore, and again seemingly contrary to our preliminary observation that criminal courts appear to prefer to protect their ontological discretion, the judge states, “the criminal law should, whenever possible, strive for certainty and simplicity in its approach to definitional problems” (2). However, we note this certainty and simplicity might only make sense or be possible after an actual exercise of ontological discretion within the case.
Indeed, years prior to Green, the Manitoba Court of Appeal had already rejected the traditional approach of heart death, instead defining death as brain death when upholding a manslaughter conviction in R v. Kitching.Footnote 19 After being carried out of a bar and dropped on the concrete by the two accused, the victim suffered a skull fracture and was quickly declared brain dead. After prolonged resuscitation efforts by emergency personnel, the victim began breathing sporadically, although doctors were unanimous that full recovery was impossible, even limited respiration would not continue unassisted, and (therefore) the victim was dead. Since the victim was an organ donor, he was placed on a ventilator until his kidneys could be removed, after which all life support was discontinued. According to medical records, the official death certificate was not signed until after the kidneys were removed, even though once again the victim’s heart continued beating for thirteen minutes after the removal. Kitching, therefore, argued that since common-law death required the cessation of all vital functions, it was solely the actions of the doctors that caused the victim’s death.
Saving the appellate court a further legal (and ontological) headache, the trial judge in Kitching was able to accept a brain-based definition of death that appeared to be more in line with the general assumptions of individual culpability raised by these particular case facts. In their charge to the jury, the judge agreed with the medical evidence that a person could be declared dead if there was no possible way to reverse or recover from a brain-dead state. In adopting this ontological framing, and in testament to the trickle-down effects of the court’s ontological game, the judge was then forced to also agree with the medical evidence that “It appears that death is a process, not an event, as I stated earlier” (para. 61).
While the Manitoba Court of Appeal upheld the trial judge’s ontological decision and legal justification, they evidently took the “definitional issue” seriously, rearticulating the possible legal and ethical paradoxes it can create:
The problem facing medicine and society in recent years is how do you get a living heart out of a dead body? If a person is not dead until after his heart ceases to function, it is practically impossible to imagine a successful heart transplant except by taking a living heart out of a still living person. But this is abhorrent to the conscience of many people. (Para. 56).
One appellate justice’s reasons for upholding the conviction read entirely as a pragmatic defence of the brain death criterion, if not a defence of their own ontological discretion. The justice explains his reasoning by noting, “This opinion does not proceed on the principle that if a man will surely die, then it is safe to call him already dead. It operates on the principle that, if a man’s brain is dead, he is dead.” (Para. 66).Footnote 20
Curiously, after opining on the definitional question, the appeal justices plainly state that the question was legally moot. As would have been clear to any first-year criminal law student, a person’s death can have multiple contributing causes, and for an offence to be a culpable homicide in Canada, all that matters is that an accused’s actions were a significant contributing cause. But what this case and reasoning in part show is that whether an action is considered causal can indeed depend on a certain ontological framing. Additionally, if brain death was not enacted and accepted in Kitching, and the courts instead took the view that the actions of the accused were simply a contributing cause to the cessation of the victim’s heartbeat, then they would simultaneously be asserting that the doctors also caused the victim’s death. Given their reasoning and the time they spent outlining the definitional question, this was clearly an implication the justices did not want to make.
Once again hinting at our focus on the pragmatic legal benefits of ontological discretion within the criminal court, it is now perhaps less surprising that Parliament has never enshrined a definition of death within the federal Criminal Code, even if some provinces have enacted specific definitions for the purposes of medical practice. Death in the criminal law continues to be ontologically open despite a Canadian Law Commission recommending in 1981 that the Code adopt a definition that stated, “For all purposes within the jurisdiction of the Parliament of Canada, a person is dead when an irreversible cessation of all that person’s brain functions has occurred.”Footnote 21 But if this definition became law, all the potential problems in Green could not have been avoided by the judge reverting back to the more traditional common-law ontology of death.
Still, after Green, various medico-legal commentators continued to press Parliament to adopt a purely brain-based definition; they wanted a coherent solution to the implication that doctors would be causing a victim’s death if the victim were removed from life support and the traditional definition was applied. Given that the brain death criterion was by then widely used in medical settings, even if it was up to the discretion of local medical associations, they also believed a clear definition that stated death was brain death for criminal purposes would help “bring the Code up-to-date.”Footnote 22 However, just like the criminal law’s traditional heart/respiratory death that only became standard via judge-made common law, brain death has never been officially enacted in the Code, nor have any definitions of any deaths.
Sober and drunk: The legal effects of alcohol intoxication(s)’s multiplicity
We spend less time with our final example since we recognize alcohol intoxication (colloquially, “drunkenness” or “impairment”) is less likely to be a controversial “multiple” to scholars familiar with ontological multiplicity. However, it is still a worthwhile example to include given its ubiquity within various legal complexes, and considering that despite their best efforts, medical and scientific experts have been much less successful in controlling its ontological status in any setting. As Levi and Valverde (Reference Levi and Valverde2001) note, since the SCC’s 1982 decision in Graat v. R, Canadian common law has recognized that determining if someone is intoxicated can be within the purview of everyday experience and common knowledge.Footnote 23 Thus, courts allow lay witnesses to provide opinion evidence on whether a person was intoxicated based on what they directly observed, even if they simultaneously allow expert witnesses to provide opinions based on indirect evidence and the physiological effects of alcohol on the body. In the words of the SCC, “intoxication is not such an exceptional condition as would require a medical expert to diagnose it” and therefore it “is not a matter where scientific, technical, or specialized testimony is necessary” (as cited in Levi and Valverde Reference Levi and Valverde2001, 838). Although this goes against general legal principles delineating expert witness opinion evidence and lay witness fact-based evidence, it is consistent with the historical observation that drunkenness, and even the more technical-sounding “alcoholism,” has never been fully medicalized (see Valverde Reference Valverde1998).
Regardless of what the SCC said in Graat, deference to any single ontology of intoxication, including the commonsense one, is only partial; adjudicators can still invoke different and often competing ontologies within a given case. While expert evidence is not technically necessary in deciding if a person was impaired, it can nevertheless be considered sufficient, even when it contradicts the direct evidence. This is precisely the legal paradox the SCC saw the need to address: within the same case facts, it is entirely possible for someone to be both drunk and sober, simultaneously, due to the ontological multiplicity of drunkenness. For example, in a single trial one might hear an expert witness testify that “based on their height, weight, and the number of drinks consumed, the accused would have a blood-alcohol content indicating they were intoxicated at the time of the car crash,” while also hearing lay witness testimony that they “didn’t think the accused was intoxicated when they left the bar because they were talking and walking normally.” Both of those statements can, of course, be true, but their relative truth in deciding if the person was “actually” intoxicated depends upon a certain ontological framing, which could go either way. It would seem that within a given case, the courts could have to explicitly determine: what is intoxication?
Just like with sleep and death, the SCC has not decreed one overarching definition for intoxication (or similar terms) and has instead found ways to avoid opining on a universal ontology. However, unlike our sleep and death examples, for the context of drunk driving, Parliament itself provided the courts with an ontological tool: it creatively (and as we argue, pragmatically) enshrined multiple ontologies of intoxication within the Code, essentially stating that either expert ontologies or lay ontologies can be sufficient for a conviction, even if the other ontology indicates the person was not intoxicated. Section 320.14(1) of the Code reads:
Everyone commits an offence who
(a) operates a conveyance while the person’s ability to operate it is impaired to any degree by alcohol or a drug or by a combination of alcohol and a drug;
[or]
(b) subject to subsection (5), has, within two hours after ceasing to operate a conveyance, a blood alcohol concentration that is equal to or exceeds 80 mg of alcohol in 100 mL of blood; … (emphasis added).
Clearly, for the purposes of drunk driving law, alcohol impairment can be enacted either as a specific blood alcohol content and/or as a cognitive-behavioural decline caused by any amount of alcohol.
Even if these multiple ontologies once again do not hang together or lead to contradicting logics within a single fact-pattern, from a pragmatic policy standpoint codifying both of them into the law makes good sense: you want to be able to hold people responsible who (for example,) blow under the legal limit on a breathalyzer but whose driving was affected by smaller amounts of alcohol, but you also want to hold people responsible who (for example,) had high quantities of alcohol in their system but did not show any outward, visible signs of impairment. If intoxication were enacted as only a decline in cognitive and behavioural function or as only a specific blood alcohol content, this pragmatic move would not be possible.Footnote 24
Thus, while the Code does provide two potentially competing ways a person could be deemed impaired, it is important that it explicitly stops short of proclaiming a singular, universal definition, instead including different possible yet still real definitions. By writing these multiple ontologies into the criminal law, Parliament and adjudicators are able to avoid ever having to declare what intoxication really is, for all time, instead allowing it to be different things, simultaneously and as needed.
Conclusion
As we hope to have shown, the criminal law must apply objects like sleep, death, and intoxication in certain cases to arrive at decisions about culpability. While the aim of the often-discretionary exercise of defining these objects is not ontological in and of itself, clearly ontological claims may have to be made along the way. Thus, in this paper, we have drawn attention to the relevance of STS literature on ontology and multiplicity to sociolegal inquiries related to common-law adjudication and discretion.
What our ontological turn shows is that once we take multiplicity seriously, new lines of inquiry emerge, lines that should be addressed if we are to better understand even practical legal processes, including how judges exercise and justify discretionary decisions. Furthermore, decisions that may not appear (or no longer appear) inherently discretionary—such as if a successful sleepwalking defence necessitates an insanity verdict or acquittal—can rely on historical and locally contingent ontological games. By examining the multiplicity of three objects in the specific legal complex of the Canadian criminal courts, we began to also document this game’s historical and locally contingent effects.
However, we of course recognize the preliminary nature of our inquiry and thus prefer to view ontological discretion as just one possible site of analysis within the ontological turn. We claim that this can and does help explain how judges come to and justify their decisions, but just how much ontology/ontological discretion one needs to account for in a variety of legal sites remains to be seen. We therefore once again call on sociolegal scholars, especially those interested in traditional notions of discretion in all its forms, to also take an ontological turn to see if this or similar discretionary tools are pervasive among a variety of legal decision makers, and, if so, how we can better understand its mechanisms and effects. This means expanding beyond the criminal law and the role of judges into all facets of “law” and legal actors where discretion has an important role, including police discretion, prosecutorial discretion, etc. (for example, Bowman, Lowrey-Kinberg, and Gould Reference Bowman, Lowrey-Kinberg and Gould2024).
By analytically moving beyond epistemology and towards ontology, much like the multiple atherosclerosis of Mol’s hospital, we see that the intoxication of the expert toxicologist and the intoxication of the patron on the barstool, the death of the brain and the death of the heart, and the sleep of the EEG and the sleep in the literal and proverbial eyes of the tired driver are not simply different opinions on the same underlying phenomena. Instead, we are required to see them as different real and potentially competing ontologies—of intoxications, of deaths, of sleeps. Additionally, it appears that judges might often implicitly see these objects as multiple, and perhaps must see them as multiple in some cases; whether judges are able to invoke their discretion to choose between these and other multiples in a given context remains an empirical question.
In making these often unstated and unnoticed ontological games our focus, sociolegal scholars in particular are reminded that few things have just one true definition, and that a particular definition is likely to be local, contingent, and mutable. Furthermore, by documenting the discretion adjudicators possess in stating what things are/will be, we not only see that many presumptively singular things are actually multiple, but also that the ontological games played out in criminal courts have serious and widespread effects. In fact, it can literally determine whether the unfortunate victim of an accused is taken to be alive or dead, and, the corollary, whether there is any murderer to be tried at all.