From Reasonable Doubt to Undue Scepticism
Interdisciplinary conference
London, 4-6 May 2017
Organisers:
Marion Vorms & Ulrike Hahn
(Birkbeck College, London)
KEYNOTE PAPERS
Secrets and lies: The Psychology of Conspiracy Theories
Karen DOUGLAS
Professor of Social Psychology, School of Psychology, University of Kent
Was 9/11 an inside job? Is climate change a hoax? Was Princess Diana murdered? Millions of people appear to think so, disbelieving official explanations for significant events in favour of alternative accounts that are often called ‘conspiracy theories’. In recent years, psychologists have begun to investigate what makes conspiracy theories appealing to so many people. In this talk, Karen will broadly overview what is known so far, and will outline her own programme of research on the causes and consequences of belief in conspiracy theories.
Knowledge-centred epistemic consequentialism
Branden FITELSON
Professor of Philosophy, Northeastern University
Julien DUTANT
Lecturer in Philosophy, King's College London
Abstract. The idea that knowledge plays an important role in norms for belief can be spelled out in two ways. The deontologist way straightforwardly lays out categorical principles such as “one must: believe only what one knows”. The consequentialist lays out a notion of knowledge-value and recommends maximizing its expectation. The latter approach has been overlooked so far. Knowledge-firsters have mostly adopted deontologist outlooks. Epistemic consequentialists have mostly focussed on accuracy—distance from the truth—as the fundamental epistemic good. This paper explores a version of epistemic consequentalism that is knowledge-centered instead: distance from knowledge is treated as the fundamental good. The result for categorical belief is a threshold view in terms of expected knowledge: believe p iff your expectation of knowing p is high enough. The threshold yields an attractive package of answers to the lottery, preface and gate-crasher paradoxes.
Law as Inquiry: A Lack of Epistemic Foundations?
Larry LAUDAN
Professor of Philosophy, Lecturer in Law, University of Texas
It is widely believed that the criminal law is a form of empirical inquiry, utilizing a wide array of evidence to elaborate a narrative about a putative criminal act. Higher courts in most countries characterize the principal aim of a criminal trial as a ‘search for the truth’. The distressing fact is that the law utterly fails, and repeatedly, to exhibit many of the key features of a search for the truth. So much so that legal systems invest virtually no resources in exploring how often they err in terms of convicting the factually innocent and acquitting the factually guilty. Put crudely, appellate courts are fixated on procedural errors but scarcely at all on errors of fact.
My arguments will focus on the general absence in the law not only of how often it reaches false verdicts but in identifying and learning from its mistakes. Above all, proper inquiry has to be self-corrective; capable and eager to identify mistakes of fact and to use those mistakes to correct prevailing rules of inquiry. Put simply: the law doesn’t (and can’t) learn from its mistakes because it doesn’t bother to find out what its factual mistakes are.
From reasonable doubt to undue skepticism and demagoguery: The age of “post-truth” politics
Stephan LEWANDOWSKY
Professor of Cognitive Psychology, University of Bristol
Imagine a world that has had enough of experts. That considers knowledge to be "elitist". Imagine a world in which it is not expert knowledge but an "opinion market" on Twitter that determines whether a newly emergent strain of avian flu is really contagious to humans, or whether greenhouse gas emissions do in fact cause global warming, as 97% of domain experts say it does. In this world, experts are derided as untrustworthy or “elitist” and people are exhorted to be “skeptical” and to doubt knowledge, science, expertise, the media, educational institutions, and the government. We may not have reached this dystopian future yet—but there are reasons to be concerned that we are heading in the direction of a “post-truth” world in which misinformation abounds and manufactured emotion trumps expertise. I analyze a few societal mega-trends that might help us understand the current malaise in public discourse, for example the decline in civic capital and trust, the increasing but asymmetric political polarization, and the fractionation of the media landscape into separate epistemological bubble. I conclude by providing some tentative pointers to possible solutions that involve “technocognition”—that is, cognitively-informed information technology that seeks to create a more epistemically grounded platform for public discourse.
Epistemic Standards in English Criminal Jurisprudence
Paul ROBERTS
Professor of Criminal Jurisprudence, University of Nottingham, UK
Adjunct Professor, UNSW Faculty of Law, Sydney
Adjunct Professor of Law, CUPL, Beijing
References to standards of proof in – and beyond – Anglophone legal jurisdictions are almost automatically associated with the iconic phrase ‘beyond reasonable doubt’ (BRD). Legal scholars and philosophers have long debated the merits of the BRD standard of proof and many critics have questioned its meaning, efficacy and practical intelligibility in routine criminal adjudication. There are equally forceful defences of common law orthodoxy, encapsulated by Blackstone, Woolmington and Winship, resulting in theoretical stalemate. Revisiting these familiar debates from a broader methodological perspective, this paper suggests that pronounced focus on BRD has skewed and misinformed scholarly examination of epistemic standards in criminal adjudication, with particular regard to the following five considerations:
1) BRD is a normative jurisprudential standard rather than a straightforwardly epistemic threshold;
2) BRD should not be assimilated to or equated with quantified probabilistic values (not even the comparative utilities employed in decision theory);
3) All common law trial systems employ a range of epistemic standards; BRD is a – not the – criminal standard of proof;
4) Trial-centric approaches to legal epistemology fixated on BRD fail to investigate a range of interesting, and important, epistemic standards in operation in criminal investigations, prosecutions and pre-trial proceedings;
5) English criminal jurisprudence abandoned BRD more than a decade ago.
Explaining how, why and with what effect (5) was brought about in England and Wales, and reflecting further on the fact that few people seem to have noticed, will further elucidate and illustrate the paper’s central themes.
Evidence Integration and Reasonable Doubt
Dan SIMON
Professor of Law and Psychology, University of Southern California Gould School of Law
Legal fact finding can be a most difficult undertaking. Fact-finders are routinely expected to make discrete decisions from massive arrays of pieces of evidence, many of which are ambiguous, uncertain, contradictory, and incommensurable. This process inevitably entails performing a pair of cognitive tasks: drawing inferences from the facts and integrating those numerous inferences into discrete conclusions. Conventional approaches to legal fact-finding (e.g., Bayesian updating) propose that these two functions are performed sequentially and separately. This talk will propose instead that the interpretation of the facts, the inferences drawn from them, and the emerging decisions are all integrated in a cyclic process of bi-directional reasoning. In this view, legal decision making (not unlike a host of other cognitive processes) is best understood as represented in connectionist networks and processed by means of parallel constraint satisfaction mechanisms that drive the representation towards high levels of cognitive coherence. Coherence is defined as the state at which the facts and inferences that support the winning decision are boosted, and the facts and inferences associated with the rejected choice are relegated and dismissed. The construction of a coherent representation is adaptive in that it enables the making of confident decisions even from vexing decisional conflicts. Yet is must not be overlooked that this functional and epistemological advantage is made possible by means of a reconstruction—actually, a distortion—of the underlying facts.
This view of the process has two implications for the subject of this conference. First, bi-directional inferential processing means that just as the fact-finder’s evaluation of the facts will impact her decision, so will the emerging decision influence her interpretation of those facts. Second, one must not lose sight of the fact that, however adaptive, the high level of confidence afforded by the construction of coherence is essentially an artefact of the cognitive process, rather than a true reflection of the probativeness of the facts of the case. It follows that the coherence-maximizing processes will tend to inflate the fact-finder’s confidence, and perhaps even bring her to experience a contestable decision to be true beyond a reasonable doubt.
CONTRIBUTED PAPERS
Reconstructing Reasonable Doubt
Yuval ABRAMS & Cosim SAYID
CUNY Graduate Center – Philosophy
Proof of guilt beyond a reasonable doubt (BARD) is notoriously difficult to quantify. We hold that trying to cash BARD out as a simple threshold probability or credence is mistaken. Clearly, the standard (correctly) allows for some doubt, but how much? When is such doubt reasonable? We offer two guiding insights to answer the latter question, both of which make more precise the prima facie qualitative notion of reasonableness. The first is that BARD requires (i) higher-order evidence concerning the evidence adduced at trial and (ii) that formation of a trier-of-fact’s belief states was (epistemically) safe. The second insight engages directly with credences: while recognizing that doubt is always potentially rational given some prior assumptions, doubt is unreasonable when the priors one would have needed to rationally arrive at doubt are themselves unreasonable, so the BARD test asks: of all the possible starting points (priors) that lead to doubt, is there at least one that is reasonable?
Reasonable doubt, presumption of innocence and standard of proof: how does criminal law’s “Holy Trinity” impacts international adjudication?
Teresa BRAVO
Judge at the UNDT/ Geneva
PHD in European and International Criminal Law
Researcher in CEDIS/ Faculdade de Direito da Universidade Nova de Lisboa
International criminal adjudication constitutes a fertile ground for discussions regarding legal principles and its implementation at a supranational level. Also, the differences between legal cultures judge’s different backgrounds in international courts and the lack of a coherent legal body as well as a deficient dogmatic approach, contribute to the emergence of a reluctant jurisprudence that lacks consistency and is an easy target for critic scholars.
The most consistent and systematic jurisprudence held at a regional and supranational level is the one that emerges from the ECHR, which has established the minimum standards of compliance with defence rights and served as an inspiration for other international jurisdiction mainly through cross fertilization.
Standard of proof and reasonable doubt are the “both sides of the same coin“ that allow the presumption of innocence to operate while keeping a balance between human rights and punishment. Presumption of innocence defines the threshold of reasonable doubt.
The leading role of the Strasbourg Court, the interpretation of the ECHR regarding legal principles as the presumption of innocence and operative rules like “reasonable doubt” and “standards of proof” can be challenging in multicultural settings, particularly those who have few in common with the western legal systems.
Evidence-Based Policies for Communicating the Reasonable Doubt Standard
Mandeep K. DHAMI
PhD, Professor of Decision Psychology, Department of Psychology, Middlesex University, UK
Samantha LUNDRIGAN
PhD, Department of Humanities and Social Sciences, Anglia Ruskin University, UK
Katrin MUELLER-JOHNSON
PhD, Institute of Criminology, University of Cambridge, UK
Court decisions are typically made under conditions of uncertainty. Decision-makers cannot, and are not required to, be 100 percent certain of the decision they are making. Rather, their degree of certainty must pass a given legal threshold (e.g., ‘beyond a reasonable doubt’ is the standard of proof used in criminal cases). Although legal standards of proof lie along a continuum (i.e., a scale of probabilities from 0-1), they are not expressed numerically. Neither do they reflect common parlance. It is therefore unsurprising that court decisions are sometimes reversed because of beliefs that decision-makers misunderstood and misapplied the relevant legal standard of proof (see Power, 1999). There is empirical evidence that judges and jurors do not always interpret these standards as intended (e.g., Kagehiro, 1990; Montgomery, 1998; Zander, 2000). Several jurisdictions have developed and implemented official policies (i.e., judicial instructions) on defining these standards for decision-makers (see Hemmens, Scarborough & Del Carmen, 1997). In addition, in some jurisdictions individual courts may define the standards as they desire. Unfortunately, these policies and practices are not evidence-based. In this paper, we will review psychological theory and methods that can be used to develop an evidence-base for informing policies on communicating legal standards of proof to court decision-makers. We summarise our empirical research on individual differences in people’s interpretations of reasonable doubt (RD), and the effectiveness of different definitions of/instructions on RD. Finally, we discuss the implications of the growing body of research on RD for under-researched standards of proof used in civil trials.
A Two-Dimensional Theory of Proof Beyond a Reasonable Doubt
Marcello DI BELLO
Fellow, School of Social Science
Institute for Advanced Study, Princeton
Assistant Professor, Department of Philosophy
Lehman College, City University of New York
I defend a two-dimensional theory of proof beyond a reasonable doubt, consisting of (1) a threshold requirement that the guilt of the defendant be established with an appropriately high probability and (2) a weight requirement that the evidence presented at trial be as inclusive as reasonably possible. In order to assess the adequacy of this theory, I identify a set of functions which proof beyond a reasonable doubt is expected to serve, such as, protecting the innocent, weighing costs and reducing the risks of mistaken verdicts. I then show that the proposed two-dimensional theory can adequately accommodate these functions, or at least, can do it better than a mono-dimensional theory can.
How to deal with unreliable witnesses
Leah HENDERSON
Faculty of Philosophy, University of Groningen
In criminal trials, there are few pieces of evidence more convincing than a confident eye-witness who claims they saw the accused commit the crime. Typically such an eye-witness will have identified the suspect by picking them out of a police line-up. However, there is now substantive empirical evidence to suggest that eye-witnesses are not very reliable in producing correct identifications in line-ups. This has been shown in a number of cases involving DNA-based exoneration of convicted suspects, as well as in psychological studies of the factors that influence eye-witness reliability. The puzzle then is: how should the information that eye-witnesses are not as reliable as often thought be factored into the overall judgment about whether the accused is guilty?
The answer to this puzzle depends in part on how the uncertainty involved is represented. In this paper, I will compare two possible ways of taking such reliability information into consideration, which depend on two different types of representation of uncertainty.
Question everything: Critical open-mindedness and reasonable doubt
John LAMBIE
Reader in Psychology, Anglia Ruskin University, UK
One solution to the problem of how to distinguish between reasonable and unreasonable doubt may lie with the notion of “critical open-mindedness” (Lambie, 2014). Critical open-mindedness is an attitude in which all points of view are subject to critical testing, all are provisional, but some have more (provisional) credibility than others, based on how well they have survived critical testing.
The notion comes from the view of rationality as criticism (e.g. Socrates; Mill, 1859; Russell, 1928; Popper, 1959). No view is ever completely “justified” and points of view are rational to the extent that they have been subjected to critical testing.
Thus any reasonable point of view – including a point of view that is sceptical (“the Moon landings were faked”) – has to be reflexively corrigible, i.e. it needs to outline what would critically test it and what critical tests it has already survived. Such a view constitutes “unreasonable doubt” if it is uncritical of itself.
But is science itself becoming less critical and hence less rational? Fanelli (2012) found an increasing trend between 1990 and 2007 for scientific journals to publish results supporting the hypothesis being tested. Instead science needs to be the art of critical open-mindedness.
Statistical Evidence and Epistemic Anxiety
Liat LEVANON
Lecturer in Law, Brunel Law School, Brunel University London
The paper will examine the status of ‘naked statistical evidence’ in legal proceedings. It will focus on the evidence’s inability to satisfy legal standards of proof, including the beyond-reasonable-doubt standard. The paper will advance the argument that naked statistical evidence stimulates helplessness in subjects trying to know the world; it produces epistemic anxiety similar to that which motivated Pyrrhonian sceptics. These helplessness, anxiety, and scepticism clash with fundamental legal assumptions about fact-finding – assumptions that fulfil basic objectives of the legal system. The paper will further suggest that by excluding the evidence, the legal system seems to reject scepticism altogether. Yet exclusion implies acceptance of two premises: that we have a duty to seek for the truth, and that at the same time we are unable to know the truth. In accepting these premises, the legal system’s response to naked statistical evidence actually bears resemblance to Pyrrhonian scepticism. Last, it will be proposed that a more refined approach to naked statistical evidence should be developed. To advance these arguments the paper will pull together several analyses of naked statistical evidence, including those by Martin Smith (2016); by David Enoch, Levi Spectre, and Talia Fisher (2012); and by Alex Stein (2005).
Crime and Punishment, Too: This Time, It's Personal
Clayton LITTLEJOHN
King’s College London
PRESENTATION HERE
Abstract: Is it appropriate to convict and punish defendants using only statistical evidence? In this paper, I argue that it is not and try to explain why it is not. This is difficult to do because there is a powerful argument for thinking that we should convict and punish using statistical evidence. It looks as if the relevant cases are cases of decision under risk and it seems pretty clear that we should act to maximize expected value in such cases. Given some standard assumptions about the values at stake, the case for convicting and punishing using statistical evidence seems solid. In trying to show where this argument goes wrong. I shall argue (against Lockeans, reliabilists, and others) that beliefs supported only by statistical evidence are epistemically defective and (against Enoch, Fisher, and Spectre) that these epistemic considerations should matter to the law. The key to solving the puzzle about the role of statistical evidence in the law is to revise some commonly held views about epistemic value and to defend the relevance of epistemology to this practical question.
Reasonable Agnosticism
Thomas RALEIGH
University of Vienna
When a proposition’s truth or falsity is not beyond reasonable doubt, then one ought to suspend judgement. But what is the nature of this agnostic mental state? Evidence serves not only to ground beliefs, it can also ground the attitude of suspending judgement (a.k.a. ‘withholding judgement’), making it more or less justified, more or less rational. Taking this epistemic feature of suspension seriously has, I contend, implications for the metaphysics of the mental state. I will argue that the attitude of suspending judgement whether p constitutively involves having a belief – more specifically, a belief about one’s evidence for p – for it is only if suspension of judgement is a species of belief that we can account for how/why it is subject to evidential norms. If there is time, I will also discuss some of the tricky issues that arise from considering (Higher-Order) evidence one one's evidence.
OPENING TALK (MARION VORMS) AUDIO PODCAST HERE
CLOSING TALK (ULRIKE HAHN) AUDIO PODCAST HERE