Adarsh Vaibhav*
The Paradox Of Certainty
The Anglo-American criminal justice system is centred around a seemingly innocent piece of advice: the prosecution should have to prove guilt beyond a reasonable doubt. This standard has been considered by courts in all common law jurisdictions, whether it is the Old Bailey or the Supreme Court of India, as the utmost epistemic hurdle that a legal system can pose. A silent, unsettling threat to this underlying assumption is, however, posed by cognitive neuroscience and psychology: the human brain is fundamentally unable to attain the sort of certainty that the law implicitly requires.
The term itself is difficult to define legally. The United States Supreme Court, in Victor v. Nebraska (1994), recognised that the Constitution did not dictate that any specific form of words had to be employed in advising the jury of the burden of proving a particular claim on the part of the government but refused to provide anything universal. Indian courts, in the same spirit, have taken the standard to be intuitively as opposed to operationally defined, according to the formulations of the Shivaji Sahabrao Bobade v. State of Maharashtra (1973) case, where the Supreme Court noted that “it is preferable that ten guilty be acquitted than one innocent be convicted”, a principle that implicitly acknowledges the margin of error to be tolerated by the system.
The issue of epistemology is even bigger than legislative inaccuracy. The process of memory is constitutive and not reproductive. There are no binary perceptual judgements. And the subjective sense of certitude, the sense that psychologists refer to as noetic familiarity, can be proven to be unrelated to accuracy. That is, a juror can be absolutely certain of an empirically incorrect conclusion. The same measure the law relies on, internal confidence, is not a reliable proxy for truth.
“The subjective feeling of certainty is demonstrably dissociable from accuracy. A juror can feel absolutely certain of a conclusion that is empirically wrong.”
The Architecture Of Doubt: The way Juries Process Evidence
Nancy Pennington and Reid Hastie, cognitive psychologists, developed what has become the best empirically validated model of jury decision-making: the Story Model. Jurors do not weigh discrete pieces of evidence in some sort of pseudo-statistical fashion but rather create a narrative explanation of events. They filter testimony, physical evidence and attorney arguments into a coherent causal narrative and match the narrative with the categories of verdicts available. The verdict selected is the best story with the best goodness of fit. The implications of this narrative construction process for the BRD (Beyond Reasonable Doubt) standard are important. After a juror has picked a story, disconfirming evidence does not simply weigh the story but is read in, or proactively misconstrued by, the narrative frame picked. It is the process behind what cognitive psychologists refer to as narrative transport: a state in which an engaging story makes it less likely to be expected to critically evaluate evidence. It is the art of the skilled prosecutor not to need to demonstrate facts beyond reasonable doubt but to create a story that pre-empts doubt.
To make this worse, there is the design of human memory. Working memory is rigorously capacity-bound, that is, in lengthy, complicated trials, jurors cannot maintain all pieces of evidence concurrently in working memory.
Cognitive Impairments In The Courtroom
In the context of the BRD standard, two cognitive phenomena should be given special consideration: “Implicit bias” and the “CSI” (Crime Scene Investigation) Effect.
Juror research has also recorded implicit bias: automatic associations that are made unconsciously. Levinson and Young (2010) discovered that the implicit racial bias mock jurors had a higher chance to find ambiguous evidence that was incriminating when the defendant had shared the same racial traits with criminality in his simultaneous associations. More importantly, these jurors did not claim to be consciously prejudiced; they thought that they were being objective in their reasoning. This result indicates that the reasonable doubt can be structurally reduced among the defendants belonging to stigmatised groups not due to dishonest jurors but due to the fact that the line where the doubt is considered reasonable is itself predetermined by unconscious schemas.
The CSI Effect creates an alternate, yet no less disconcerting, deformation. The CSI Effect, the belief among jurors that forensic evidence, in the form of CSI DNA profiles, ballistics, digital forensics, etc., must be consistently available and conclusive, has been popularised by television procedural dramas and is part of the expectation of a specific degree of infallibility in evidence.
The Weight Of Witnessing: Witnessing And Fallibility
It is hard to imagine a topic in legal psychology that has created as much empirical data or created so much of a gap in scientific opinion and legal practice as the testimony of witnesses. Eyewitness identification has been given a great persuasive value by the legal system in the past. The eyewitness testimony is always rated as very credible by jurors, especially when the witness is confident.
The mental evidence is to the contrary. The pioneering studies by Elizabeth Loftus showed that post-event information could distort what witnesses thought they had seen, what she referred to as ‘misinformation effects’.
Memory is not a fixed record; it is susceptible to constant ongoing reconstruction: further interrogation, media coverage, and social discourse all have the potential to introduce errors, which can be integrated into the honest memory of the witness.
Neurobiology of stress also subverts the legal ideal of the confident witness. Huge amounts of cortisol at the time of a traumatic experience led to a constriction of the attentional focus, the so-called weapon focus effect, of which peripheral details (including, most often, facial features of assailants) are poorly encoded despite the increased emotional salience of the situation. This means that a witness can have a strong sense of certainty of an event and the ability to recollect only a small part of the event that can be used in court. The inability of the law to be more systematic in taking into consideration such a disconnect is a weakness in the structure.
“Eyewitness misidentification contributed to approximately 69% of wrongful convictions overturned by DNA evidence. Not failures of perjury, failures of human memory operating in good faith.”
Reimagining The Standard: Toward Psychologically Informed Justice
The above discussion is not a reason to drop the BRD standard; the standard is a principled fortress against state encroachment, but to change the institutional circumstances in which it is used.
Standardised and psychologically calibrated jury directions. National Research Council (2014) suggested that jury instructions should have clear information about the unreliability of eyewitness testimony, such as the counterintuitive connection between witness confidence and accuracy. Other jurisdictions, including New Jersey, which followed the leading case of State v. Henderson (2011), have updated their model instructions with research-based cautionary language, a practice that ought to guide reform in other common law traditions, such as the Indian context under Sections 59 and 60 of the Indian Evidence Act, 1872, which regulates oral evidence and implicitly favours eyewitness.
The combination of probabilistic inference models. Some critics such as Tribe (1971) warned of wholesale mathematisation of the evidence in law, but there is a golden mean, i.e., there is a middle path. Experimentally, structured deliberation protocols, which force jurors to explicitly give probability weight to the main disputed facts before making a verdict, have been shown to reduce both anchoring bias and the narrative transport effects of Pennington and Hastie. The protocols do not require jurors to be statisticians; they only require the deliberation process to externalise and interrogate, instead of conceal, the inferential leaps on which the specific verdict is depending.
Eyewitness evidence evaluation professionalisation. Expert psychological evidence on eyewitness reliability should be routinely admitted in the court as a regular practice rather than a rare one since cases involving contested identifications are common. The case R v. Turnbull (1977) gave guidance by the UK Court of Appeal on a judicial warning system of identification evidence, which has since been superseded by much more advanced research. An established protocol, based on ‘Signal Detection Theory’ to differentiate between discriminability (the capacity to differentiate between targets and non-targets) and response bias (the point at which identification is voluntarily offered), might offer a more analytical framework to courts than the mostly impressionistic determinations currently used.
All these reforms do not eradicate the psychological irreducibility of human judgement. Their thing is to introduce structural checks into the system which counterbalance, at least in part, the failures of unaided cognition which are recorded. In this regard, a legal innovation project is not so much about substituting the role of human decision-making as about creating institutional settings where the unavoidable constraints of that decision-making result in the minimal possible injustice.
* The author is a second-year Law student at National University of Study & Research in Law (NUSRL), Ranchi. The author may be contacted at adarsh.vaibhav@nusrlranchi.ac.in.
This blog reflects the personal views of the author and doesn’t necessarily represent the views of The Policy Chronicle.