A man suffers from tinnitus and other hearing impairments caused when a tire that a custom auto shop had sold him exploded as he was checking its bolts. The man sued the auto shop in products liability. At trial, one of his expert witnesses, an audiologist, presented to the jury a digital sound file that, the audiologist testified, corresponded to the tinnitus sounds the man claimed to hear. The sound file was based on the data that the audiologist had obtained by testing the manâs hearing. Jurors took turns putting on headphones and listening to the sounds. Ultimately they awarded the plaintiff over a million dollars in damages for permanent acoustic injury and pain and suffering, largely for his tinnitus (Janson v. J.D.O.R.A.P., Inc. 2011).
Tinnitus is the subjective experience of sound without a corresponding external source, analogous to phantom limb pain. The sufferer is the only one who can hear it. In this case, no one doubted that the plaintiff had tinnitus. But how could anyone really know that the evidentiary exhibit, the sound file, accurately recreated the sound of his tinnitus? How can jurors know that this is what itâs like to hear what the litigant doesâthat these specific features of his subjective experience can be taken as fact for purposes of the law, and thus relied on in awarding pain and suffering damages?
The tinnitus case is not unique. In a handful of other cases, lawyers, clients, and experts have created exhibits that purport to simulate litigantsâ subjective perceptions for judges and jurors. For instance, one man contended that malpractice in the performance of LASIK surgery left his vision blurred and doubled, including âstarburstsâ around lights at night, making it very difficult for him to do his work as a pharmacist. The jury saw photographs of the plaintiffâs workplace, Photoshopped to appear as he claimed it looked to him (Devadas v. Niksarli 2009). In another case, a woman alleged that as a result of a negligent delay in diagnosing and treating her idiopathic intracranial hypertension, an increase in pressure on the brain that can impinge on the optic nerve, her field of vision contracted much more than it otherwise would have. To obtain a favorable settlement of the case, her lawyer showed defense counsel a video simulation of what various scenes in the plaintiffâs daily life looked like to her, with her greatly constricted visual field (Smith v. Jones 2012).1 In a third case, a police officer was prosecuted for shooting at a driver who was attempting to flee after being pulled over. The officer claimed that he fired his gun in justified self-defense because, in the heat of the moment, he thought he saw the driver heading straight towards him, even though the dashcam video from another police cruiser showed that perception to have been mistaken. At trial, the defendantâs lawyer played jurors a computer animation of what the critical events looked like from the defendantâs point of view, based on his testimony about what he claimed to have seen (State v. Murtha 2006).
How is it that demonstrative evidence can be understood to let jurors know, and not merely know about, what itâs like to be inside these litigantsâ minds? When and how do one personâs purely subjective perceptions, purportedly simulated by a sound file, a series of photos, a video, or a computer animation, become as much of a fact as any other on which jurors rely in reaching their judgments? I begin with some remarks about facts in general, facts in law, and the anomaly of regarding other peopleâs subjective experiences as âfactualâ in the customary sense. I explain why legal advocates might seek more than the usual testimonial descriptions of their clientsâ internal experiences by creating computer-generated exhibits, often scientifically based, that purport to simulate the experiences themselves.
Then, examining the Janson tinnitus case, I explore how the lawyers, the expert witness, and the judge talked about and used the sound file at trial so that Dennis Jansonâs auditory sensations could emerge as a scientifically plausible fact. We will see why, based partly on the sound file itself and partly on how it was framed for them in court, jurors might well have concluded that by listening to it, they could actually know what it was like to hear what the plaintiff heard. In conclusion, I suggest that litigantsâ subjectivity as made known through these sorts of simulations will, however, remain only equivocally factual in the law, in part because of the deep cultural tensions that digital simulations of personhood implicate.
Facts and facts at trial
What is meant by a âfactâ is, of course, highly contested terrain, depending in part on what âfactâ is being opposed to: artifact, fiction, theory, value, interpretation, and context being merely some of the possibilities. There are histories (e.g. Poovey 1998; Shapiro 2000) and sociologies (e.g. Fleck 1935/1979) of what counts as a fact. The very notion of fact cannot be understood apart from particular epistemologies; different conceptions of the subjects, objects, and processes of knowledge yield different ideas of factuality. Iâll proceed on the assumption that the lawâs dominant, if largely implicit, epistemology is one of naĂŻve realism (Sargent, this volume; see Scheppele 1998), with which the following common sense notion of factuality is consistent. A fact is a state of affairs in the world, capable of being established by publicly accessible evidence and reasoning. A fact is generally thought to be objective at least in the sense that it is the same for me as it is for you. When we say that x is a fact, we mean that the proposition âx is the caseâ is true, and we also imply that the evidence on which we rely for our belief accurately represents the way things really are.
Fact-finding in law is peculiar in several respects. Legal facts are sculpted by the need to resolve particular disputes when the litigants bring them to court and by substantive and procedural rules that limit the sorts of facts that matter and the means of proving them. Each proceeding needs to establish the facts only for the case at hand; legal actors, unlike scientists, do not generally think of themselves as adding to a body of general knowledge. Perhaps most importantly, as evidence law scholar Robert Burns (1999) has explained, jurors determine facts in the course of making an essentially normative judgment. Every verdict reflects a moral and political evaluation of the situation and the parties, and the facts that are found result from and support this exercise in practical wisdom.
Some might push the point further. Legal facts are not just peculiar or pragmatic, the argument goes. They arenât âfactsâ in the ordinary (or scientific) sense at all, but merely the referents on which legal arguments and legal rulings operate (e.g. Latour 2010b), and yet the common sense of facts as more or less objective truths in the world, and the belief that trial evidence can often get at those facts, do have a firm purchase in the law. Many of the practices of litigation and trial, from discovery to witness examination and most of the rules of evidence, would make no sense if legal actors perceived no need to find the facts as best as they can. As legal scholar Charles Nesson (1985) has argued, the publicâs belief in the validity of verdicts depends on accepting them as accurate statements about what happened in the world and not merely as the outcome of a dispute resolution process. In short, the fact of the matter really matters in law.
One of the sort of facts that sometimes matters in law is a visual, auditory, or other sensory experience known only to the person who experiences (or experienced) it. Here is where the puzzle begins.
The philosophical problem of other
peopleâs subjectivity
Given the common sense notion of what facts are, itâs apparent that what itâs like to sense or feel somethingâthe subjective, phenomenal nature of each personâs own experienceâis anomalous. Simply by introspection, we seem to know the qualia that make up our sense of conscious experience more immediately and more certainly than we can know anything else. To pick a classic example: although I may be mistaken about the source of a pain, I canât be mistaken about whether Iâm in pain, and although I may have trouble putting it into words, I canât be mistaken about what that pain feels like. The question of whether what one senses or feels accurately represents what itâs like to have those experiences doesnât arise because oneâs awareness of the feeling (or percept or thought) is the experience. These subjective states, therefore, seem to occupy a privileged epistemological status.
At the same time, no personâs conscious experience is directly available to anyone else. Each of us assumes that (except in extreme cases) other people also have mental states that we would recognize as conscious experience (e.g. Damasio 1999), but when it comes to their qualia there seems to be a threshold we canât cross. As William James (1890/1950) put it, we can know about but we canât know. To the extent that knowledge of what is outside our own consciousness must be in some sense objective to be truly factual, other peopleâs phenomenal experiences seem not to qualify. It seems impossible in principle, as Thomas Nagel (1974) (among others) has famously argued, to give an objective, third-person account of subjective, first-person experience. From the perspective of what each of us can know about the world outside ourselves, other peopleâs qualia are so far from being prototypical facts that we might wonder whether they should be considered facts at all. So how, if those qualia are relevant to a legal dispute, can they be made credibly available to legal decision makers?
A pragmatic approach and its limitations
for the law
In David Lodgeâs novel, Thinks⌠(2002), Ralph the neuroscientist, echoing Nagel, remarks to his friend Helen: ââThatâs the problem of consciousness in a nutshell: How to give an objective, third-person account of a subjective, first-person phenomenon.â âOh,â Helen replied airily, âbut novelists have been doing that for the last two hundred yearsââ (Lodge 2002: 42). Few of us can describe internal states as well as skilled novelists can. But in everyday life, when we need to know what itâs like for other people to be experiencing what theyâre experiencing, we generally rely on what they tell us. We trust, even if we donât fully grasp, what they say and if we know the person we can try to calibrate our belief. Sometimes we can also observe the other personâs behavior or learn about it from others. By interpreting these self-reports and observations in light of our beliefs about how people in general can be expected to behave and what we ourselves have experienced in what we think are similar situations, we can often know about othersâ subjective states well enough.
But there are often good reasons for judges and jurors to be more suspicious. Parties to lawsuits are prone to misstate their experiences, deliberately or subconsciously. Personal injury plaintiffs, to pick an obvious example, have an incentive to exaggerate their pain or incapacity; the defendant police officer in the case mentioned earlier might well have wanted to claim a perception that he actually didnât experience. Some of the subjective perceptions to which litigants may testifyâsevere tinnitus, for instanceâmay be so far beyond jurorsâ own experiences that they fail to grasp it. Where failing to describe oneâs subjective experience convincingly may mean a difference of hundreds of thousands of dollars in damages, or the difference between freedom and prison, mere words and circumstantial evidence may not be good enough.
Responding to doubt: simulating subjectivity
Some litigants and their lawyers have therefore turned to a potentially more convincing sort of evidence of subjective perceptions: digital simulations of the experiences themselves. Three types of simulations can be distinguished, each generated by a different method and thereby staking a distinct kind of epistemological claim to provide reliable knowledge of the subjective state it purports to reproduce. Brief descriptions of examples of each type will also show how particular features of the simulation mediaâsound file, photo sequence, video, animationâare deployed to achieve desired rhetorical effects, especially inducing jurors to empathize with the litigant and thus be more inclined to acquit (in the case of the criminal defendant) or to award damages (in the cases of personal injury plaintiffs).
The artistâs sketch
The first type of demonstrative is produced when the litigant collaborates with someone who has expertise in fashioning digital images or sounds. Much as an eyewitness to a crime might collaborate with a police investigator wielding an Identikit to work up a sketch of a suspect, so here the litigant and someone adept at manipulating photographic data or digital sound files go back and forth until the litigant says that the product corresponds to his subjective perception. The person working with the litigant to create the exhibit need not have any scientific expertise relevant to the litigantâs clinical condition. The claim that this picture or sound file accurately simulates the litigantâs subjective experience thus depends entirely on the plaintiffâs say-so.
Examples of this include the Photoshopped pictures in Devadas, the LASIK malpractice case, and the animation in Murtha, the police shooting case. In Devadas, jurors saw a series of photographs of shelves of medicine at the plaintiff pharmacistâs workplace and then a set of pictures of a traffic intersection at night, part of his evening commute. Each series began with a âreference photoâ taken by the plaintiffâs wife with an ordinary digital camera, intended to represent the scene as it would appear to someone with normal vision. This was followed by photos edited to show how that scene appeared to the plaintiff, with his impaired vision, first in one eye and then the other, with and without corrective lenses. In his testimony, the plaintiff made it clear enough that the edited photos were produced like a sketch artistâs drawing:
I work with a colleague who was savvy with the computer who could, who could describe how I see and he can like edit the photos so that I can explain to somebody else how I would see. ⌠[A]fter e-mailing the [reference] picture, I talked it over the telephone in describing how I see and the halos that I see, the double visions that I see, and I described to him how far apart that they are. And we went back and forth and we corrected it as we went.
Devadas v. Niksarli 2009: 2162
The presentation of each photo simulation, one after the other, in what appears to be the same medium and format as the initial, reference photograph, implicitly locates the simulations in the same order of representation as an ordinary, unedited photograph. Just as viewers adopt the perspective of the cameraâs lens and regard the ordinary photo as a trustworthy and accurate record of the external reality in front of the lens, so the photo simulations encourage viewers to believe that they are occupying the plaintiffâs position and seeing an equally trustworthy record of what itâs like for him to see as he does.
The visual simulation in the Murtha case is actually a hybrid of video and animation. It begins with a clip from the video taken by the dashboard camera of the trailing police cruiser. The camera approaches to within perhaps twenty feet of Officer Murthaâs cruiser, which is stopped with its lights flashing, and the suspectâs car, stopped in the snow on the side of the road. As the suspect begins to pull back onto the road, the action âfreezesâ and the video turns into an animation. The virtual cameraâs (and our) point of view rises up and rotates to the left around Murthaâs cruiser, coming to rest behind a digital Officer Murtha as he stands, outstretched arm holding his gun, confronting the car. The viewer is now just behind Murtha, sharing his point of view in the crucial moment: the bright headlights of the suspectâs car flash in our eyes, the left front headlight seemingly only a few feet away. Because the scene is frozen, the car could, for all that appears, be headed right at us. The animation holds this shot for a moment, then pulls away, up, back, and around again to the point of view from the surveillance camera, and we see the suspectâs car continue down the road, away from us.
No one would be likely to confuse the simplified, schematic animation with a dashcam video. Nevertheless, there are several reasons why viewers may have been inclined to believe that this video-plus-animation showed them what Officer Murtha really saw at that moment. The montage of animation and video clip bolsters the animationâs credibility by grounding it in the presumptive truthfulness of the video (see also Mopas and Curran, this volume). The transition from video to animation folds the two into the same level of represented reality: by starting with the dashboard cameraâs point of view and swinging around to the defendantâs, the animation implicitly claims that both the defendant and the camera were âlooking at the same reality.â Jurors have no trouble understanding what theyâre seeingâa lifetime of watching movies and television has accustomed them to identifying intuitively with the camera âeyeâ as it shifts from third-person omniscient to first-person subjective point of view. In addition, connecting the animation to the video in this way normalizes the more novel medium of animation by associating it with the very familiar (and presumptively reliable) medium of automatically recorded videoâan implicit version of what media theorists call remediation (Bolter and Grusin 1999). Finally, the video-animation hybrid persuades by associating Murthaâs point of view with that of the video taken by the dashboard camera in the other police cruiser, implicitly associating Murtha with those other âgoodâ officers rather than singling him out as a renegade shooter. All of these effects, like those created by the photo simulations of the plaintiffâs visual perception in Devadas, can be traced to the simulation media and to the audienceâs mediated viewing habits. The actual warrant for any claim that these simulations faithfully recreate the litigantsâ subjective perceptions remains solely the litigantsâ assertions that they...