Neuroscientists in Court Owen D

Vanderbilt University Law School Scholarship@Vanderbilt Law

Vanderbilt Law School Faculty Publications Faculty Scholarship

2014 Neuroscientists in Court Owen D. Jones

Anthony D. Wagner Stanford University

David L. Faigman University of California, Hastings College of the Law

Marcus E. Raichle Washington University in St. Louis

Follow this and additional works at: https://scholarship.law.vanderbilt.edu/faculty-publications Part of the Courts Commons, Criminal Law Commons, Evidence Commons, and the Neuroscience and Neurobiology Commons

Recommended Citation Owen D. Jones, Anthony D. Wagner, David L. Faigman, and Marcus E. Raichle, Neuroscientists in Court, 14 Nature Reviews Neuroscience. 730 (2014) Available at: https://scholarship.law.vanderbilt.edu/faculty-publications/1081

This Article is brought to you for free and open access by the Faculty Scholarship at Scholarship@Vanderbilt Law. It has been accepted for inclusion in Vanderbilt Law School Faculty Publications by an authorized administrator of Scholarship@Vanderbilt Law. For more information, please contact [email protected]. PERSPECTIVES

Whether this will provide a net gain NEUROSCIENCE AND THE LAW — SCIENCE AND SOCIETY in the fair and effective administration of justice is a topic of current debate23–26. Neuroscientists in court But the new types and increasing quality of neuroscientific evidence — particularly brain imaging techniques, on which we pri- Owen D. Jones, Anthony D. Wagner, David L. Faigman and Marcus E. Raichle marily focus here — ensure that the interac- Abstract | Neuroscientific evidence is increasingly being offered in court cases. tion between law and neuroscience is both unavoidable and intensifying. This article Consequently, the legal system needs neuroscientists to act as expert witnesses explores some of the reasons why neuro- who can explain the limitations and interpretations of neuroscientific findings scientific evidence is being offered in legal so that judges and jurors can make informed and appropriate inferences. The proceedings, several key features of law growing role of neuroscientists in court means that neuroscientists should be that neuroscientists may wish to know and aware of important differences between the scientific and legal fields, and, several important clarifications about and limitations of neuroscience that the legal especially, how scientific facts can be easily misunderstood by non-scientists, system needs to learn from neuroscientists. including judges and jurors. Why neuroscience? In 2005, convicted child-rapist Grady Nelson of documentary neuropsychological reports Why is the legal system increasingly turning brutally murdered his wife Angelina. After and sometimes in the form of neuroscien- to neuroscientists? The answer is simple: it stabbing her 61 times, he left a butcher knife tists testifying in court proceedings. Some of does so in the hope that new technologies may embedded in her brain. Later, his own life these neuroscientists appear willingly, and help to satisfy many acute and long-lingering hung in the balance as the Florida jury that some reluctantly. It appears that sometimes needs, including the law’s need to answer convicted him of murder next had to decide their involvement in a case sparks a plea bar- questions such as: is this person responsible whether he would be executed or spend his gain or settlement before trial7,8. The princi- for his behaviour? What was this person’s life behind bars. Nelson’s attorney offered to pal importance of the example cases above is mental state at the time of the act? How much provide neuroscientific evidence — specifi- to raise this question: when and how should capacity did this person have to act differently? cally, quantitative electroencephalography neuroscientists participate in litigation? What are the effects of addiction, adolescence (QEEG), introduced through the testimony In barely a decade, a distinct field of ‘law or advanced age on one’s capacity to control of a neuroscientist — to suggest that Nelson and neuroscience’ has emerged4,6–20, accom- behaviour? How competent is this person? had potentially relevant brain abnormali- panied by a sharp rise in both conceptual and What does this person remember? How accu- ties. The jury should hear this evidence, the empirical scholarship (FIG. 1), conferences (see rate is this person’s memory? What are the attorney argued, because although it may not neurolaw conferences on The MacArthur effects of emotion on memory, behaviour and excuse Nelson’s behaviour, it should mitigate Foundation Research Network on Law and motivation? Is this person telling the truth? his punishment1. Neuroscience website), international neu- In how much pain is this person? How badly In a different case, in 2010, psychologist rolaw societies (see the external links page injured is this person’s brain? Lorne Semrau went on trial in federal court on The MacArthur Foundation Research Because society uses the legal system to for allegedly masterminding healthcare fraud Network on Law and Neuroscience website), help regulate the behaviour of its citizens, the in connection with psychiatric care that ‘law and neuroscience’ courses being taught legal system turns to disciplines (typically two of his companies provided for patients in law and other departments internationally, social science disciplines, such as psychology, in nursing homes. His attorney offered to a forthcoming coursebook21, new neurosci- economics and sociology) that claim to have introduce neuroscientific evidence — specifi- ence–law joint-degree programmes, and so special insights into the causes of human cally, the results of a functional MRI (fMRI) on. In light of the possibility that techno- behaviour, what patterns human behaviour test, introduced through the testimony of a logical advances might aid the legal system, manifests and how people are likely to react neuroscientist — to suggest that Semrau was and in view of how important it is for law to as law shapes incentive structures within truthful when he claimed that any overbill- separate neuroscientific wheat from chaff, social environments27. Neuroscience may in ings were accidental (rather than purposeful, the John D. and Catherine T. MacArthur part be ‘hot’ in law because its technological as the government would have to prove). Foundation has funded two consecutive sexiness may lend it persuasive power and Among the evidence2,3 he offered to intro- interdisciplinary, collaborative research ini- because legal advocates are, in turn, always duce was the neuroscientist’s conclusion that: tiatives in the United States — the Law and alert for potentially persuasive ways to aid “Dr. Semrau’s brain indicates he is telling the Neuroscience Project and The MacArthur their clients. However, in our view, neurosci- truth in regards to not cheating or defrauding Foundation Research Network on Law and ence has become attractive mainly because the government” (REF. 2). Neuroscience. These developments as well many legal professionals, courts and com- In these cases, and a steadily increasing as the rising number of references to neu- mentators hope or believe that it can provide number of similar cases in both criminal roscientific evidence in court opinions4,22 a tool that not only usefully supplements tra- and civil courts, neuroscientific evidence (N. Farahany, personal communication), ditional social science perspectives but that has been introduced to support a party’s suggest that neuroscientists may be called also may be, in some contexts, more objective legal claim as well as to argue its irrelevance upon with increasing frequency — and with and powerful. or invalidity (by the opposing party)4–6 implications and consequences yet unknown Not surprisingly, some substantial (N. Farahany, personal communication). — to lend their expertise to matters of fears accompany that hope. In our collec- That evidence comes sometimes in the form legal import. tive experience at the neuroscience–law

730 | OCTOBER 2013 | VOLUME 14 www.nature.com/reviews/neuro

1,200 not party to a lawsuit but who nonetheless believe they have information or perspectives that the Court should consider when deciding the case. For instance, in the past 1,000 decade, amicus briefs involving neuroscientists were filed in three prominent Supreme Court cases regarding criminal punishments 47–49 800 of juvenile offenders . The fourth way to become involved in litigation is as an expert witness50. If the judge in the case decides that a proposed 600 witness is qualified (on the basis of special- ized knowledge that has typically been acquired through education, training and experience) to be designated as an expert 400 witness, then that person can offer opinions

Cumulative number of publications about, or interpretations of, the facts in the case — something that fact witnesses are 200 ordinarily not allowed to do. In the remain- der of this article, we focus on issues that neuroscientists might encounter when acting as an expert witness. 0 Legal and scientific cultures 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 Suppose that, after much discussion, review Year and reflection, a neuroscientist agrees to be Figure 1 | Cumulative growth in the number of ‘law and neuroscience’Nature Reviews publications. | Neuroscience The figure an expert witness. The neuroscientist’s expe- shows a sharp rise in the number of publications in the ‘neurolaw’ field between 2003 and 2013. rience as an expert witness will depend on These publications include both conceptual and empirical scholarship in the neurolaw field. Figure (among other things) his or her understand- is reproduced from the Law and Neuroscience Bibliography on the website of The MacArthur ing of a number of things about the distinct Foundation Research Network on Law and Neuroscience. cultures and contexts of neuroscience and law. We will discuss six crucial matters of which neuroscientists should be aware when (neurolaw) intersection, it has become Four roles for neuroscientists acting as expert witnesses. clear that many people — both inside and There are four main ways in which a neu- outside of the legal professions — worry roscientist may become involved in litiga- Decisions under uncertainty. At the most that neuroscience is too complex and too tion. The first is as a so‑called ‘fact witness’ general (and therefore over-simplified) technical in nature for laypeople to under- (sometimes called ‘lay witness’). Fact wit- level, we can consider science to be about stand and apply, even when particular neu- nesses can testify about the underlying facts facts and law to be about values. More roscientific evidence could, if it is properly of a case on the basis of their own personal specifically, we could say that science aims understood, be useful to law’s purposes. knowledge, which is grounded in direct expe- to discover facts, and thus to increase our In addition, they worry about the risks of rience with the parties involved in the case or collective knowledge of reality through over-reductionism, the possible low explan- issues in the case. For example, the neurolo- various inductive and deductive means, atory power of neuroscientific evidence gist who first examined a plaintiff after his including hypothesis-driven experimenta- (that is, when the neuroscience evidence injury in an accident may be called upon to tion. By contrast, law aims to pursue the adds little beyond the behavioural evi- recount her examination and findings. ends of society’s values — with respect to dence) and — more importantly — about The second way is as a non-witness orderly, productive and just behaviour. It the general problem of drawing inferences consultant. In this role, they may help does this through various legislative, execu- about the consequences of brain states that attorneys to evaluate neuroscientific evi- tive and judicial means, including through are defensible both scientifically and within dence offered by the opposing side, suggest courts that exist for one single reason: to the specific legal context that each case questions an attorney should ask opposing resolve disputes. may present. These worries are fanned by witnesses or provide general, non-testimo- There are of course many other impor- concerns about, among others, the ecologi- nial advice about the strength of a claim, tant aspects of the legal system that distin- cal validity of laboratory-based studies, the about the significance of a finding or about guish it from the scientific system. One is challenges of drawing inferences relevant who else should be consulted as the case that trial courts typically must decide who to an individual from group-based stud- develops. wins and who loses a case. There are no ies and the potential over-persuasiveness The third way is to join an effort to ties, there are no maybes and there is no of neuroimages18,25,26,28–46. We believe that prepare a so‑called ‘amicus brief’ for cases tabling the issue for further court-managed neuroscientists can play crucial roles in before the US Supreme Court. In suitable study. A second key aspect is that courts addressing these concerns during legal circumstances, such briefs can be submit- cannot completely control when they must proceedings. ted by individuals or organizations that are decide. As a case progresses, there comes a

NATURE REVIEWS | NEUROSCIENCE VOLUME 14 | OCTOBER 2013 | 731

time when the decision must be made. At Put another way, science generally opinions. In the United States, the federal that moment, a civil plaintiff either wins approximates truths by hypothesis-testing, system and each of the constituent states can or loses, and a criminal defendant is either whereas the legal system frequently approxi- develop their own rules about, for example, freed or not. mates truths by evaluating what happens how to define ‘murder’ and can likewise A third and crucial aspect is a function of when two highly incentivized teams shoul- develop their own procedural rules about, the prior two: jurors and judges must almost der a legally imposed duty to gather evi- for example, how to decide whether to admit always make decisions under conditions of dence and to argue in favour of two directly the testimony of scientific experts. considerable uncertainty. The decisions they opposite propositions. This difference has Many state courts in the United States make depend not only on the level of uncer- major implications for the experience of continue to use the so‑called ‘Frye test’ tainty but also on the specific legal context. neuroscientists in court. (articulated in 1923 in the case Frye v. United Roughly speaking, the more consequential States)59 for determining the admissibil- the decision, the more certain the decision- Experts on the stand. One consequence of ity of scientific evidence. Under the Frye makers must be (that is, the higher the ‘bur- the legal system’s trying to grind truth from test, which is sometimes referred to as the den of proof’ must be) before they should between the abrasive surfaces of two oppos- ‘general acceptance test’, the opinions of decide that a claim or allegation is meritori- ing parties is the unpleasant phenomenon of scientific expert witnesses are admissible if ous. For example, when life or liberty is at cross-examination — the process by which they are based on principles or techniques issue in criminal cases, the US Constitution the other side tries to expose flaws in the that are generally accepted as reliable in the requires proof “beyond a reasonable doubt”. expert witness’s background, credentials and relevant scientific community. By contrast, to win a civil trial, at which only reasoning. This can come as a shock to the Since 1993, all US federal courts have amounts of money are at issue, a mere “pre- expert witness, especially if he or she fails to been required to apply a different test to ponderance of the evidence” (the US stand- anticipate it or fails to take it in stride. The determine admissibility, and many state ard, akin to the “balance of the probabilities” way this process works, systemically, is that courts have chosen to adopt this test as in the United Kingdom) is required. Of after the attempt at undermining is over, well. That test is reflected in Rule 702 of the course, research scientists can and do have the opposing attorney will undermine the Federal Rules of Evidence60, which instanti- to deal with uncertainties of various kinds undermining in an effort to expose the cross- ates the so‑called ‘Daubert standard’ (named in their own research. However, the nature, examination as misleading, irrelevant and after the 1993 US Supreme Court case range and contexts of that uncertainty in futile, and to show that the expert witness is Daubert v. Merrell Dow Pharmaceuticals61 science are considerably distinct from those indeed an expert in both the scientific and and further articulated in several subsequent in law. The key point is that courts cannot the legal sense. cases62,63). Under the Daubert standard, avoid — as scientists generally can (and often which is sometimes described as the ‘gate- should) by continuing their own research The role of experts. Neuroscientists may keeping standard’, the opinions of scientific — making high-impact decisions in the have been called upon because of their expert witnesses are admissible only if a face of imperfect information. Keeping this expertise, which a court may have recog- judge is satisfied that they are helpful and systemic constraint in mind helps to make nized by ‘qualifying’ them as expert wit- appropriately scientific and that they have sense of many otherwise puzzling features of nesses. They may therefore think (at least been correctly applied to the case at hand. litigation. the first time they appear in court) that they Unlike the Frye test, which calls upon judges are being asked to provide a little lecture, to inquire whether the science is generally How to approximate truth. Because science with their own preferred organizational accepted by the field, the Daubert stand- and law have different functions and therefore structure, about what they know that others ard requires that judges themselves assess different attitudes with respect to conditions do not. But that is not what being an expert whether the expert’s testimony is grounded of uncertainty, science and law often pursue witness entails. Rather, expert witnesses in valid science. Relevant (but emphatically truth quite differently51,52. Science engages in are typically required to answer specific non-exclusive) factors in making this assess- an iterative process of trial, error and refine- questions, which often emerge from prior ment include: first, whether the theory or ment in pursuit of generalizable knowledge; discussions with the lawyer. The lawyer uses technique can be tested and has been tested; indeterminate experiments can be followed this approach to elicit the relevant opinions second, whether the theory or technique by new and better experiments. By contrast, in such a way that they are understand- has been subjected to peer review and pub- at least one major component of the legal able to the judge and jury. The answers will lication; third, the known or potential rate system — the resolution of disputes in court ultimately be weighed by non-expert jurors of error of the method used; fourth, the — requires individual courts to repeatedly alongside other evidence presented by both existence and maintenance of standards engage in particularized, one-shot decision- sides. Depending on the circumstances of controlling the technique’s operation; and making that often has no generalizable bear- the trial, among the factors that usually fifth, whether the theory or method has been ing on other disputants or on the systemic influence jurors’ acceptance of expert testi- generally accepted by the relevant scientific accumulation of a greater body of knowl- mony are their perceptions of the witness’s community64,65. edge. Courts cannot conduct experiments, accomplishments, of any bias that may be Importantly, the decision of whether nor order any, in pursuit of the truth. Instead, revealed and of the clarity and accessibility a neuroscientist’s evidence has passed the they depend on an adversarial process that of the testimony53–58. applicable test (Frye or Daubert) does should, it is hoped, reveal the relevant truths not end the admissibility analysis. That is by harnessing, within an ultimately gladiato- Admissibility of expert testimony. Expert because the legal system not only requires rial arena, the competitive spirits, economic opinions must first be evaluated for admis- that scientific testimony be directly relevant interests and ethical obligations of each sibility — that is, the judge will decide to a decidable issue but also recognizes that party’s lawyers. whether to even allow the jury to hear the the value added by some kinds of evidence

732 | OCTOBER 2013 | VOLUME 14 www.nature.com/reviews/neuro

may be outweighed by the potentially preju- used in both fields but that mean completely Similarly, when one party plans to intro- dicial effect the evidence may have on jurors. different things to members of the two duce neuroscience-based evidence that has For example, an undisputedly accurate but disciplines. been improperly gathered, inappropriately extremely lurid and graphic photograph of The first fount of confusion is when analysed, misrepresented or is otherwise a murder victim’s maimed body may risk each discipline has a different but special- insufficient for the inferences that legal unfairly inflaming the jurors’ passions in ized meaning of the term in question. For decision-makers are being urged to draw such a way as to prevent a fair trial of the example, whereas in psychology the word from it, then again the legal system needs accused, who may actually be innocent. ‘normative’ quickly invokes the meaning neuroscientists who are willing to serve as Judges have the discretion to exclude such ‘representative of the group being studied’, experts, so that countervailing views of the evidence, and they often do. the same word in law is just as automatically evidence can be aired. The ability of judges to exclude relevant understood to be referring to an ‘ought’ prop- Neuroscientists can provide crucial evidence if its effect is very likely to be dis- osition. That is, in law, the word ‘normative’ information and perspectives for juries and proportionately and unfairly strong provides is used in reference to how something should judges, and in many cases this informa- an important check on the admissibility of be done. Other examples are terms such as tion will be highly case-specific. Below, we scientific evidence. For example, Rule 403 ‘theory’, ‘trial’, ‘threshold’, ‘representation,’ discuss a number of more general science- of the Federal Rules of Evidence empowers ‘evidence’ and even ‘fact’. related points on which jurors and judges judges to “exclude relevant evidence if its The second fount of confusion is when often need guidance. probative value is substantially outweighed one field uses a very common word in a very by a danger of … unfair prejudice, confusing technical way and the other does not. To a How technologies that acquire brain data the issues, misleading the jury…” (REF. 66). neuroscientist, the word ‘significant’ brings work — and do not. Jurors and judges do Some commentators worry that the visual p‑values to mind, whereas the use in law is not need in‑depth courses in neuroanatomy impact of brain images may be so great, typically synonymous with ‘important’. A neu- or need to learn, for example, how flip and the memory of them so vivid, that they roscientist may use the word ‘development’ to angles and T2 weightings work in fMRI. unfairly prejudice the jury in favour of the mean the process by which the brain matures, Nevertheless, no one can draw legitimate party offering them and that for that reason whereas a lawyer recognizes the word to mean inferences from data if they do not have a alone they should sometimes be excluded ‘a thing that happened’. Similar confusion good sense of how the data were obtained from evidence. For example, in Illinois — follows the use of terms such as ‘plastic’, ‘reli- and what they actually mean. For instance, which uses the Frye test — the judge presid- ability’ and many others. The same is true in with respect to fMRI evidence, it is essential ing over the trial of murderer Brian Dugan the reverse, when law imbues common words that legal decision-makers understand that allowed a neuroscientist to describe to the with technical meanings. For example, to a when they see an image of colours inside jury his methods and findings but prohibited neuroscientist, the question of whether a per- the skull, they are not looking at something him from showing the jurors any images son behaved ‘knowingly’ is largely an inquiry meaningfully analogous to an X‑ray of brain of Dugan’s brain itself 67. (Experiments with into what the person knew, whereas within activity in those locations but rather at the potential jurors suggest that brain images criminal law, in which not only a bad act but outcomes of statistical analyses performed sometimes have an outsized impact in actual also a culpable state of mind is required for on the data. court proceedings68 — by having a more per- someone to be convicted of a crime, ‘know- suasive effect on jurors than the facts warrant ingly’ means something quite different. It is Structural and functional images are dif- — and sometimes do not69.) one of four highly defined and specific cul- ferent. There are many contexts in which pable states of mind within the Model Penal neuroscientific testimony could be relevant Two founts of confusion. There is a wide Code80 — the four being purposefully, know- to law, such as interpreting neurotransmit- variety of resources on the subject of how ingly, recklessly and negligently. Each term ter deficits, explaining the memory deficits experts should handle questions (both on carries the baggage of hundreds of thousands of patients with Alzheimer’s disease or direct examination and under cross-exam- of cases and hundreds of scholarly articles explaining why someone seeking disabil- ination), how they should communicate in parsing its contextual nuances. ity benefits has impaired brain function ways that judges and jurors can understand, Consequently, when neuroscientists after a car accident, notwithstanding the and so on70–79. Our collective and exten- and lawyers talk with one another, it is absence of cranial penetration. Those dif- sive personal observations of interactions quite common that they focus on trying to ferent contexts and the different kinds of between lawyers and neuroscientists — understand the obvious jargon and therefore structural or functional evidence will carry both inside and outside of litigation con- miss and, as a result, misunderstand the different opportunities for neuroscientists texts — suggest there are some additional, non-obvious jargon. to help the legal system to avoid important important, terminology-centred issues of misunderstandings. communication between the two fields that What courts need from neuroscientists Take MRI and fMRI as examples. transcend the courtroom context. As mentioned earlier, neuroscience is Individuals who are unfamiliar with brain Each field has its jargon. A neuroscientist increasingly being offered as evidence in imaging can be forgiven for not knowing may mention the ‘dorsolateral prefrontal litigation. The legal system needs solid evi- that functional images are meaningfully cortex’ and a lawyer may use common Latin dence that can aid just decision-making, different from structural images. After all, legalese, such as ‘res ipsa loquitor’. In these and although neuroscience is not always both types of image may show structural fea- obvious cases of jargon use, the neuroscien- relevant, there are many cases in which it tures in high resolution and both may have tist and lawyer are probably aware that the can be. In these cases, the legal system needs embedded features, such as a carpenter’s terms need explanation. However, a more neuroscientists who are willing to serve as nail in a structural MRI scan, and a colour- insidious problem concerns words that are experts to enable the evidence to be heard. ized representation of varying statistical

NATURE REVIEWS | NEUROSCIENCE VOLUME 14 | OCTOBER 2013 | 733

significance in an fMRI scan. Neuroscientists violent acts. The neuroscientist could explain, scan that shows some functional abnormal- are in the best position to point out that for example, that although it is possible that a ity is indeterminate. It is often not possible scanners do not actually generate fMRI brain causal connection exists, there are other pos- to conclude that because the brain is func- images and that the images are instead gen- sibilities too. For example, it could be that tioning or malfunctioning in a particular erated through a series of (human) decisions the experience of being a repeated contract way now it was necessarily functioning or about how the data should be processed, killer results, over time, in this particular malfunctioning in the same way at the time what statistical comparisons should be made statistical abberation in brain function. Or it of the criminal act. In addition to the ques- and what statistical thresholds should be could be that the two things co-vary because tion of whether one can deduce any causal used in those comparisons. Relatedly, neuro- of something else entirely. Neuroscientists connections between the brain function and scientists can usefully note that fMRI enables are in the best position to help decision- the act on the basis of a scan, it is also by no inferences to be made about the mind that makers navigate the narrow path between means certain that any ‘abnormality’ in the are based on inferences about neural activity under-interpreting and over-interpreting brain or in brain function that is detected on that are based on the detection of physiologi- neuroscientific evidence. a scan caused the act that landed a person cal functions, which are thought to be reli- in jail rather than being a consequence of ably associated with brain activities. These Brains differ. Sometimes group-averaged being in jail. are the sorts of clarifications and caveats data about brain function are presented that the legal system will often need to hear. in court to help prove something about Conclusions Neuroscientists can also explain that other the brain of, for example, the accused. In order for wife-killer Grady Nelson to be types of neuroscience evidence are similarly Although this may in some cases provide sentenced to death, seven of the twelve jurors dependent on data acquisition and analysis useful information, it is far from certain (a simple majority) had to vote in favour of procedures. how often it does. In some cases, the executing him. Only six did, so his life was law needs neuroscientists to clarify that spared by the narrowest possible margin. Base rates are important and often although aspects of brain structure, brain Following the vote, it appeared that the neu- unknown. The third point can be illustrated activity and neurotransmitter function roscientific evidence had been crucial. Two with this example: Herbert Weinstein, a may be similar between subjects, there is of the jurors who voted against executing 65‑year-old advertising executive, strangled often a great deal of variation across indi- Nelson told the press that the neuroscientific his wife and threw her out of the apartment viduals. That is, scientific findings from QEEG evidence had changed their minds, window, apparently to make it look like studies of a group of individuals are not given that they had each initially favoured his suicide. It turned out that he had a large automatically or necessarily relevant to execution. One of them said: “It turned my subarachnoid cyst — highly visible on a individual cases82–85. decision all the way around. The technology positron emission tomography scan — the really swayed me… After seeing the brain growth of which had displaced and thereby Brains change. Today’s brain is not yester- scans, I was convinced this guy had some sort compressed brain tissue. day’s brain. People do not walk around with of brain problem.” (REF. 86) Connecting the location of the cyst with miniature brain scanners on their heads By contrast, in the case of Lorne Semrau, results of a number of academic studies just in case brain functioning at a particular the psychologist charged with fraud, his could give reason to believe that some of moment may turn out to be important later. fMRI ‘truth verification’ evidence was not the defendant’s cognitive capacities were Some factors that influence brain function admitted to the jury. After 2 days of intense impaired at the time of his violent act. enable reasonable guesses about prior brain testimony from two neuroscientists (S. Laken However, many brain regions are involved states. For example, knowing the growth rate and M.E.R.) and one statistician (P. Imrey), in a wide variety of functions, and this con- of a certain tumour detected now may render the court excluded the neuroscientific evi- siderably complicates any effort to directly it reasonable to believe that the tumour was dence because (among other things) it failed connect a particular and unusual brain already there, a bit smaller, 6 months ago. But two of the four recommended Daubert feature with a particular past behaviour81. other factors that influence brain function factors. Specifically, the error rates of using And, perhaps more importantly, we do not are more transient and dynamic. It can there- fMRI for lie detection were not known (third know the base rate of the phenomenon: fore be difficult to know today how the brain Daubert factor), and using fMRI for this how many people are walking around with was functioning 6 months (or more) ago, at particular purpose is not generally accepted similar cysts in their heads who do not the time of the legally relevant event. within the relevant scientific community strangle their wives and throw them out of For example, it has become extremely (fifth Daubert factor)87. Following the trial, windows? common in death penalty appeals for the Semrau was subsequently convicted of three defence counsel to offer the results of a counts (out of sixty) of healthcare fraud. Correlation is not causation. Suppose that recent brain scan of a client who has been Whatever the merits of these two results, the brains of nine out of ten killers-for-hire, on death row for many years. Interpreting they illustrate a growing intersection of neu- when scanned after being arrested, each have the relevance and meaning of such a scan is roscience with law. It is becoming increasingly the same and statistically significant abnor- not easy. It is possible that it reveals a late- common for lawyers to offer neuroscientific mality in brain function (compared with discovered structural or functional condi- evidence — particularly brain images — in law-abiding matched controls). A neurosci- tion of such a massive nature, about which both criminal and civil litigation. In our view, entist can help to point out that neither this so much is known from research studies, this development is both promising and per- statistic nor the functional abnormality — that it calls into question the prior conclu- ilous depending on whether and how well nor the combination of the two — can legiti- sion of the legal system about the person’s courts can come to distinguish, within the mately support a strong inference of causal culpability at the time of the criminal act. contours of distinctly adversarial proceed- connection between the abnormality and the But it is more often the case that a brain ings, between justifiable and unjustifiable

734 | OCTOBER 2013 | VOLUME 14 www.nature.com/reviews/neuro

inferences. Neuroscientists have crucial parts 16. Gazzaniga, M. S. The law and neuroscience. Neuron 50. National Research Council. The Age of Expert 60, 412–415 (2008). Testimony: Science in the Courtroom (The National to play in a legal system that needs to under- 17. Aharoni, E., Funk, C., Sinnott-Armstrong, W., Academies, 2002). stand and interpret neuroscientific evidence Gazzangia, M. Can neurological evidence help courts 51. Faigman, D. L. Legal Alchemy: the Use and Misuse of assess criminal responsibility? Lessons from law and Science in the Law (W. H. Freeman, 2000). and to separate wheat from chaff. neuroscience. Ann. NY Acad. Sci. 1124, 145–160 52. Faigman, D. L. et al. (eds) Modern Scientific Evidence: The ability of neuroscientific techniques to (2008). The Law and Science of Expert Testimony (West 18. Brown, T. & Murphy, E. Through a scanner darkly: Publishing Co., 2013). shed light on important aspects of human cog- functional neuroimaging as evidence of a criminal 53. Shuman, D. W., Champagne, A. & Whitaker, E. Juror nition has generated hope that neuroscience defendant’s past mental states. Stanford Law Rev. 62, assessments of the believability of expert witnesses: a 1119–1208 (2010). literature review. Jurimetrics J. 36, 371–382 (1996). can help to answer some perennial questions 19. Buckholtz, J. W. & Marois, R. The roots of modern 54. Shuman, D. W., Champagne, A. & Whitaker, E. in courts of law. However, one should keep in justice: cognitive and neural foundations of social Assessing the believability of expert witnesses: science norms and their enforcement. Nature Neurosci. 15, in the jurybox. Jurimetrics J. 37, 23–33 (1996). mind that it is easier to misunderstand or mis- 655–661 (2013). 55. Shuman, D. W., Champagne, A. & Whitaker, E. An apply neuroscience data than it is to under- 20. Jones, O. D. in Neurosciences and the Human Person: empirical examination of the use of expert witnesses New Perspectives on Human Activities (eds Battro, A., in the courts — part II: a three city study. stand and apply them correctly, and this is Dehaene, S. & Singer, W.) (Pontifical Academy of Jurimetrics J. 34, 193–208 (1994). crucially important when lives and livelihoods Sciences, 2013). 56. Ivkovic, S. K. & Hans, V. P. Jurors’ evaluations of 21. Jones, O. D., Schall, J. D. & Shen, F. X. Law and expert testimony: judging the messenger and the depend on it. Whether courts can successfully Neuroscience (Aspen, in the press). message. Law Soc. Inquiry 28, 441–482 (2003). navigate these challenging waters will depend 22. Farahany, N. A. Memories and lies in law 57. Robertson, C. T. & Yokum, D. V. The effect of blinded (Presentation). Colloquium Law Neurosci. Crim. experts on juror verdicts. J. Empir. Leg. Stud. 9, on the level of engagement by neuroscientists. Justice (2013). 765–794 (2012). 23. Greene, J. & Cohen, J. For the law, neuroscience 58. Sanders, J., Saks, M., & Schweitzer, N. in Modern Owen D. Jones is at the Law School and Department of changes nothing and everything. Phil. Trans. R Soc. Scientific Evidence: The Law and Science of Expert Biological Sciences, Vanderbilt University, Nashville, Lond. B 359, 1775–1785 (2004). Testimony (eds Faigman, D. L. et al.) 191–264 (West Tennessee 37203, USA. 24. Sapolsky, R. M. The frontal cortex and the criminal Publishing Co., 2013). justice system. Phil. Trans. R Soc. Lond. B 359, 59. Frye v. United States, 293 F1013 (D. C. Cir. 1923). Anthony D. Wagner is at the Department of 1787–1796 (2004). 60. Rule 702 of Federal Rules of Evidence. uscourts.gov 25. Morse, S. J. in Law and Neuroscience: Current Legal [online], http://www.uscourts.gov/uscourts/ Psychology and Neurosciences Program, Stanford Issues (ed. Freeman, M.) 529–562 (Oxford Univ. rulesandpolicies/rules/2010%20rules/evidence.pdf University, Stanford, California 94305, USA. Press, 2010). (2010). 26. Morse, S. J. Avoiding irrational neurolaw exuberance: 61. Daubert v. Merrell Dow Pharmaceuticals, 509 U.S. David L. Faigman is at the UCSF/UC Hastings a plea for neuromodesty. Mercer Law Rev. 62, 579 (1993). Consortium on Law, Science & Health Policy, Hastings 837–859 (2011). 62. General Electric Co. v. Joiner, 522 U.S. 136 (1997). College of the Law, University of California, San 27. Jones, O. D. & Goldsmith, T. H. Law and behavioral 63. Kumho Tire Co. v. Carmichael, 526 U.S. 137 (1999). biology. Columbia Law Rev. 105, 405–502 (2005). 64. Berger, M. A. in Reference Manual on Scientific Francisco, California 94102, USA. 28. Morse, S. J. Brain overclaim syndrome and criminal Evidence. 3rd edn 11–36 (The National Academies, responsibility: a diagnostic note. Ohio State J. Crim. 2011). Marcus E. Raichle is at the Departments of Radiology, Law 3, 397–412 (2006). 65. Faigman, D. L. The Daubert revolution and the birth of Neurology, Neurobiology, and Biomedical Engineering, 29. Morse, S. J. Determinism and the death of folk modernity: managing scientific evidence in the age of Washington University, St. Louis, Missouri 63110, psychology: two challenges to responsibility from science. UC Davis Law Rev. 46, 893–930 (2013). neuroscience. Minnesota J. Law Sci. Technol. 9, 1–36 66. Rule 403 of Federal Rules of Evidence (as amended in USA. (2008). 2011). federalevidence.com [online], http:// 30. Morse, S. J. New therapies, old problems, or, a plea federalevidence.com/pdf/2011/04-Apr/FRE. Correspondence to O.D.J. for neuromodesty. AJOB Neurosci. 3, 60–64 (2012). Restyling.4-26-11SCT.pdf (2011). e‑mail: [email protected] 31. Morse, S. J. in Neuroimaging in Forensic Psychiatry: 67. Hughes, V. Science in court: head case. Nature 464, From the Clinic to the Courtroom (ed. Simpson, J. R.) 340–342 (2010). doi:10.1038/nrn3585 341–357 (Wiley-Blackwell, 2012). 68. Saks, M. J., Schweitzer, N. J., Aharoni, E. & Kiehl, K. Published online 12 September 2013 32. Pustilnik, A. C. Violence on the brain: a critique of The impact of neuroimages in the sentencing phase of neuroscience in criminal law. Wake Forest Law Rev. capital trials. J. Empir. Leg. Stud. (in the press). 1. State of Florida v. Grady Nelson No. F05-00846 (11th 44, 183–237 (2009). 69. Schweitzer, N. J. & Saks, M. J. Neuroimage evidence Fla. Cir. Ct., 4 Dec 2010). 33. Opderbeck, D. W. The problem with neurolaw. Seton and the insanity defense. Behav. Sci. Law 29, 2. United States of America v. Lorne Allan Semrau Hall Public Law Research Paper No. 2214601 (2013). 592–607 (2011). 07–10074, opinion by Judge Pham. http://www.tnwd. 34. Pardo, M. S. & Patterson, D. Neuroscience evidence, 70. Brodsky, S. L. The Expert Expert Witness: More uscourts.gov/JudgePham/opinions/659.pdf (2010). legal culture, and criminal procedure. Am. J. Crim. Law Maxims and Guidelines for Testifying in Court 3. Shen, F. X. & Jones, O. D. Brain scans as evidence: 33, 301–337 (2006). (American Psychological Association, 1999). truth, proofs, lies and lessons. Mercer Law Rev. 62, 35. Pardo, M. S. & Patterson, D. Philosophical 71. Brodsky, S. L. Testifying in Court: Guidelines and 861–883 (2011). foundations of law and neuroscience. Univ. Illinois Law Maxims for the Expert Witness (American 4. Mackintosh, N. et al. Brain Waves 4: neuroscience and Rev. 2010, 1211–1250 (2010). Psychological Association, 1991). the law. The Royal Society [online], http://royalsociety. 36. Pardo, M. S. & Patterson, D. More on the conceptual 72. Couture, W. G. & Haynes, A. W. (eds) Litigators on org/uploadedFiles/Royal_Society_Content/policy/ and the empirical: misunderstandings, clarifications, Experts: Strategies for Managing Expert Witnesses projects/brain-waves/Brain-Waves‑4.pdf (2011). and replies. Neuroethics 4, 215–222 (2011). from Retention through Trial (American Bar 5. Farahany, N. A. The Impact of Behavioral Sciences on 37. Pardo, M. S. & Patterson, D. Minds, brains, and Association, 2010). the Criminal Law (Oxford Univ. Press, 2009). norms. Neuroethics 4, 179–190 (2011). 73. Gutheil, T. G. The Psychiatrist as Expert Witness 6. Jones, O. D. & Shen, F. X. in International Neurolaw: a 38. Pardo, M. S. & Patterson, D. in The Future of (American Psychiatric Press, 1998). Comparative Analysis (ed. Spranger, T.) 349–380 Punishment (ed. Nadelhoffer, T.) 133–154 (Oxford 74. Kuhne, C. C. A Litigator’s Guide to Expert Witnesses (Springer, 2011). Univ. Press, 2013). (American Bar Association, 2007). 7. Rosen, J. The brain on the stand. The New York Times 39. Weisberg, D. S., Keil, F. C., Goodstein, J., Rawson, E. & 75. Lubet, S. Expert Testimony: a Guide for Expert (11 March 2007). Gray, J. R. The seductive allure of neuroscience Witnesses and the Lawyers Who Examine Them 8. Davis, K. Brain trials: neuroscience is taking a stand in explanations. J. Cogn. Neurosci. 20, 470–477 (2008). (National Institute for Trial Advocacy, 1998). the courtroom. ABA J. (2012). 40. Discover [online], http://blogs.discovermagazine.com/ 76. Matson, J. V. Effective Expert Witnessing: a Handbook 9. Morse, S. J. & Roskies, A. L. (eds) A Primer on neuroskeptic/ (2013). for Technical Professionals (Lewis, 1990). Criminal Law and Neuroscience (Oxford Univ. Press, 41. Uttal, W. R. The New Phrenology: the Limits of 77. Neal, T. M. S. Expert witness preparation: what does 2013). Localizing Cognitive Processes in the Brain (MIT Press, the literature tell us? Jury Expert 21, 44–52 (2009). 10. Spranger, T. (ed). International Neurolaw: a 2003). 78. Warren, R. A. The Effective Expert Witness: Proven Comparative Analysis (Springer, 2011). 42. Legrenzi, P., Umilta, C. & Anderson, F. Neuromania: on Strategies for Successful Court Testimony (Gaynor, 11. Garland, B. Neuroscience and the Law: Brain, Mind, the Limits of Brain Science (Oxford Univ. Press, 2011). 1997). and The Scales of Justice (American Association for 43. Tallis, R. Aping Mankind: Neuromania, Darwinitis and 79. Scarrow, A. M. & Scarrow, M. R. Medicolegal issues: the Advancement of Science and Dana Press, 2004). the Misrepresentation of Humanity (Acumen, 2011). providing expert witness testimony. Surg. Neurol. 57, 12. Freeman, M. D. A. & Goodenough, O. R. (eds) Law, 44. Shen, F. X. Neuroscience, mental privacy, and the law. 278–283 (2002). Mind and Brain (Ashgate, 2009). Harvard J. Law Public Policy 36, 653–713 (2013). 80. Model Penal Code, Section 2.02, Uniform Laws Annotated 13. Zeki, S. & Goodenough, O. R. Law and the Brain 45. Shen, F. X. Mind, body, and the criminal law. (1962). (Oxford Univ. Press, 2006). Minnesota Law Rev. 97, 2036–2175 (2013). 81. Poldrack, R. A. Inferring mental states from 14. Jones, O. D., Buckholtz, J. W., Schall, J. D. & 46. Satel, S. & Lilienfeld, S. O. Brainwashed: the Seductive neuroimaging data: from reverse inference to large- Marois, R. Brain imaging for legal thinkers: a guide for Appeal of Mindless Neuroscience (Basic Books, scale decoding. Neuron. 72, 692–697 (2011). the perplexed. Stanford Technol. Law Rev. 5 (2009). 2013). 82. Faigman, D. L. Evidentiary incommensurability: a 15. Greely, H. T. & Wagner, A. D. in Reference Manual on 47. Roper v. Simmons, 543 U.S. 551 (2005). preliminary exploration of the problem of reasoning from Scientific Evidence 3rd edn 747–812 (The National 48. Graham v. Florida, 560 U.S. ___ (2010). general scientific data to individualized legal decision- Academies, 2011). 49. Miller v. Alabama, 567 U.S. ___ (2012). making. Brooklyn Law Rev. 75, 1115–1136 (2010).

NATURE REVIEWS | NEUROSCIENCE VOLUME 14 | OCTOBER 2013 | 735

83. Faigman, D. L., Monahan, J. & Slobogin, C. Group to 87. Wagner, A. in A Judge’s Guide to Neuroscience: a Neuroscience. We are grateful for helpful comments from individual (G2i) inference in scientific expert Concise Introduction (eds Gazzaniga, M. & Rakoff, J.) N. Farahany, D. Kaye, M. Saks and F. Shen. testimony. http://dx.doi.org/10.2139/ssrn.2298909 13–25 (SAGE Center for the Study of the Mind, 2010). (2013). Competing interests statement 84. Fienberg, S. E., Faigman, D. L. & Dawid, P. Fitting The authors declare no competing financial interests. science into legal contexts: assessing effects of causes Acknowledgements or causes of effects? Sociol. Meth. Res. (in the press). The authors are members of The MacArthur Foundation FURTHER INFORMATION Research Network on Law and Neuroscience, Vanderbilt Law 85. Sanders, J. Applying Daubert inconsistently? Proof of Law and Neuroscience Bibliography: http://www.lawneuro. School, Nashville, Tennessee, USA. Preparation of this article individual causation in toxic tort and forensic cases. org/bibliography.php was supported, in part, by a grant from the John D. and Brooklyn Law Rev. 75, 1367–1404 (2010). Neurolaw News: http://www.lawneuro.org/listserv.php Catherine T. MacArthur Foundation to Vanderbilt University, 86. Miller, G. Brain exam may have swayed jury in The MacArthur Foundation Research Network on Law and Nashville, Tennessee, USA. Its contents reflect the views of the sentencing convicted murderer. ScienceInsider [online], Neuroscience: www.lawneuro.org http://news.sciencemag.org/technology/2010/12/brain- authors and do not necessarily represent the official views of exam-may-have-swayed-jury-sentencing-convicted- either the John D. and Catherine T. MacArthur Foundation or ALL LINKS ARE ACTIVE IN THE ONLINE PDF murderer (2010). The MacArthur Foundation Research Network on Law and

736 | OCTOBER 2013 | VOLUME 14 www.nature.com/reviews/neuro