In a 2013 opinion piece, Professor Pratiksha Baxi wrote about the injustice that victims of sexual assault have historically suffered at the hands of the criminal process in India, reminding us that even those cases which forced our laws to change were stories of sexual assaults never proven before the eyes of law. That opinion piece was written in the wake of allegations in the case registered as State v. Tarun Tejpal, where on 21.05.2021, the Court of the Additional Sessions Judge at Panaji acquitted the accused on all charges, i.e. for alleged commission of offences under 376(2)(f), 376(2)(k), 354, 354A, 354B, 341, and 342 of the Indian Penal Code 1860.
The judgment has been critiqued on the court’s consideration of the victim’s testimony [see, for instance, here, here and here]. It appears that an appeal has been filed by the state challenging the acquittal, where the High Court has initially directed that sections of the judgment ought to be redacted as they reveal the identity of the victim.
This post does not attempt a microscopic review of the merits of the case, not only because an appeal is pending, but also because the judgment does not give a clear conspectus of the entire evidence on record to allow for such an exercise. Instead, while making some broad observations on the judgment (to the extent possible based on the evidence extracted) it brings up three issues that the judgment throws into sharp relief: (i) appreciating evidence, with a focus on witness credibility and the handling of inadmissible evidence at trial; (ii) consideration of digital evidence from victims in sexual assault cases, and; (iii) consequences of “bad” orders on the system itself.
On the “Sterling Quality” of Sole Testimony
7 years, 2 months, and 25 days: that is how long it took for this trial to finish from the date when the chargesheet was filed, which itself happened four months after the allegations first surfaced. Such delays are endemic to the system because of its structure [see here for a long take on this] — litigants can challenge almost every decision taken by the trial court during the process all the way to the Supreme Court, and this happened here as well.
A very real consequence of delay is reducing the quality of oral testimony on offer before a court. The importance of oral testimony is a central feature of the criminal process almost everywhere. It is also one of its central fallacies because, despite overwhelming science to show the instability of memory, criminal trial rules still demand that we recall events to an unreal standard. This is something the Serial podcast has played with very well to demonstrate that not only do we forget things often, but we do so very quickly. In spite of this, trial rules play up contradictions in witnesses’ recollections by comparing their oral testimony in court with what they said earlier during investigation. The more there are contradictions in versions, the more untruthful the witness supposedly is. Even in a setup like India, where trials take place years after investigations are done, and memories have faded to the point of oblivion, this is the yardstick adopted to judge the credibility of a witness. Sure, courts agree that some leeway is needed because of this, but the more contradictions you can point out, the worse the credibility of a witness.
Which brings us to the specific context of rape trials and their rules of evidence. Unlike other cases, some special rules of evidence govern these cases in India, such as a rule that the “sole testimony” of the victim is enough to mandatorily create a presumption about the non-existence of consent where the alleged physical act is proven; with it being possible for the accused to rebut that presumption. Alongside this rule is a set of general judicially-ordained directions that ask courts to ordinarily trust the victim’s version and not necessarily ask for corroborative evidence. This leads to a position, in practice, where all facts may be proven and a conviction may be based on the sole testimony of the victim. The result of such a situation which permits convictions based on the testimony of a single witness is to amp up the existing yardsticks for judging credibility: The court demands a “sterling witness” and not an ordinary one.
Take a look at how the Supreme Court has, on one occasion, described it:
“In our considered opinion, the ‘sterling witness’ should be of a very high quality and caliber whose version should, therefore, be unassailable. The Court considering the version of such witness should be in a position to accept it for its face value without any hesitation. To test the quality of such a witness, the status of the witness would be immaterial and what would be relevant is the truthfulness of the statement made by such a witness. What would be more relevant would be the consistency of the statement right from the starting point till the end, namely, at the time when the witness makes the initial statement and ultimately before the Court. It should be natural and consistent with the case of the prosecution qua the accused. There should not be any prevarication in the version of such a witness. The witness should be in a position to withstand the cross-examination of any length and howsoever strenuous it may be and under no circumstance should give room for any doubt as to the factum of the occurrence, the persons involved, as well as, the sequence of it. Such a version should have co-relation with each and everyone of other supporting material such as the recoveries made, the weapons used, the manner of offence committed, the scientific evidence and the expert opinion. …” (Emphasis mine)
It’s hard to fault this reaction which asks for more when it comes to this one witness which can decide a case, but it should be obvious by now that this reaction is illogical and sets an impossibly high standard for any witness to meet, let alone a witness who is the victim of sexual assault and is testifying about it. If anything, the judgment in Tejpal demonstrates the double-edged nature of this process of establishing the quality of a witness. Opening a victim to “cross-examination of any length and howsoever strenuous it may be” in a situation where it isn’t enough to merely doubt the witness’ narrative but also damage her credibility is an invitation to place the victim on trial rather than the facts. This appears to have been a tactic for the defence in Tejpal and the court obliged: it took the bait that the victim ought to behave in a certain way by setting up false ideals of credibility.
So, the court found it incredible that the victim could have been assaulted because she was found smiling in the presence of the accused after a few days of the alleged incident. That the victim spoke to persons and legal professionals before moving a complaint, that she was educated, a journalist, and knew more than ordinary persons about cases of sexual assault owing to her professional expertise — all of this appears to have driven the court to expect an impossible level of consistency in narrative, and rendered every contradiction or omission in testimony fatal to her credibility. When the victim testified to having placed her professional responsibilities before other concerns, even her own well-being, it rendered her testimony unbelievable to the court [more on proving consent here]. The extent to which the court engaged in such an exercise is such that, in some places (“the prosecutrix typically gave highly evasive answers”), the judgment appears to be a contest for proving that the victim is an unbelievable liar rather than determining the innocence or guilt of an accused person.
“Un-Ringing the Bell” and Character Evidence
Returning to the idea of special rules of evidence for rape trials, one such rule renders evidence of past sexual history irrelevant on the issue of proving consent and quality of consent. Another rule also renders it impermissible to pose questions to the witness on such lines during cross-examination. The consequence of terming something irrelevant is to render it inadmissible in evidence — such information cannot be the basis for arriving at any conclusions [more on conflating relevance and admissibility here].
Both of these rules can be seen as the legal system’s technique of curbing tendencies to badger a witness and, in turn, help the trial stay rooted to the facts and not extraneous circumstances. Tejpal offers a stark reminder of the truth that these rules are only as good as the context they operate in — the victim went to the High Court to get an order for curbing irrelevant questioning, presumably because the trial court was not doing enough to screen irrelevant questions or the defence counsel was clever enough to frame the line of questioning in a way which made the character-questions not so obvious.
Possibly poor implementation of trial rules which are meant to screen bad questions is a problem. But it should not distract us from a much bigger problem afflicting criminal trials which renders the special rules of evidence for rape cases anodyne. This is the manner in which evidence recording ordinarily takes place. Normally, irrelevant questions are curbed by way of raising objections when evidence is being recorded, and a court decides that objection before letting the testimony go on. However, in a bid to curb delays, criminal trials in India normally do not address objections this way. While objections are noted during recording of evidence they are decided only at the stage of argument. Using Tejpal as an example, the end result looks something like this: a judge still looks at all of the irrelevant evidence, and then in the end, says that it has to be “glossed over”.
Think about what this means for a second: not only does a witness still have to answer the irrelevant questions but the answers, possibly inadmissible in evidence, get recorded as well and form part of the record. Judges are expected to “un-ring the bell” if they uphold the objection at the stage of arguments and not let the inadmissible evidence influence their decision. This is great on paper but, frankly, impossible to either implement practically or to account for subsequently. Decision-making is a complex phenomenon and while judges might not ascribe formal reasons in the judgment to inadmissible evidence, that material can undoubtedly still influence their thinking at a sub-conscious level. Are the special rules offering little more than a placebo, then, to deal with the problems at hand?
The Victim as a Source of Evidence
A significant feature of the defence strategy in Tejpal arose out of scouring the victim’s life for facts which could be used in evidence. This came as a result of the police having seized her mobile phone and also having sought access to her email etc. for purposes of the investigation. As per the rules of disclosure in a criminal trial in India, an accused is required to be provided copies of all materials that are proposed to be relied upon by the prosecution to establish its case at trial. Courts have interpreted these rules broadly of late, and have required sharing even material which the prosecution might not plan to rely upon but had seized during the course of the investigation, especially where this might prove beneficial to an accused. In line with this interpretation, the accused was provided clone copies of the mobile phone of the victim.
Exploring digital devices in the hope to find material useful as evidence is not extraordinary in any sense for criminal investigations. But, as we know by now in this post, proving sexual assault cases at trial is different from going about this job in other cases. Here, it is unfortunately quite common for the victim to be on trial herself instead of only the facts. Broad rules of disclosure can create unique harms for victims in such cases and I need not elaborate upon the many kinds of dangerous consequences which can arise. In fact, the Supreme Court recognised this to some extent in 2019, ruling that a “copy” ought not to be given to the accused and only an inspection be allowed of some sensitive material as it agreed that giving copies could lead to an accused person duplicating the material to, amongst other things, defame the victim. The Supreme Court did not direct that an accused ought not to be given material because this would be unfair, creating a situation where accused persons must only look at what the prosecution thinks is relevant for a case. Allowing for such a rule in one set of cases is likely to seep across contexts and should be avoided no matter what.
What might be a way out? Better regulation of how investigations handle digital evidence is a good place to start. These are difficult questions as on one hand the police ought to have room to investigate all lines of inquiry but on the other hand a victim ought not to be humiliated via a “digital strip search” of her entire life on the pretext of establishing whether a specific set of allegations are credible or not [see here and here about how the UK was struggling with this problem]. By reducing the amount of access that can be demanded by law enforcement, unwanted intrusion into the private lives of victims can be minimised (a factor often cited as being a hindrance to reporting of sexual offences), while also automatically limiting the access given to an accused at trial. Victims ought to be informed at the outset about how their information can become part of the case and eventually land up in the hands of accused persons as well, giving them an option of not doing so without exposing them to censure while making them aware of how this might also hurt the investigation by closing down lines of inquiry. If anything, such an approach will help make informed decisions about how victims participate in the process, allowing them to retain agency. It might also help police in remaining focused on securing the available independent evidence — a feature which appears notably absent in Tejpal and contributed to castigating remarks from the court.
Of course, it is impossible to change or hope for change in investigative practices without thinking about the trial as a whole. Investigations are all about collecting material which may be used as evidence — if what the trial wants as evidence focuses upon creating unrealistic “sterling witnesses” then that is what an investigation will also look for, going beyond the allegations and down the rabbit hole of credibility and character.
Being Wary of the Ripple Effect of “Bad” Orders
Public opinion has decried the judgment in Tejpal as a travesty. Beyond the debate about whether the court reached the right result, the criticism has been about how the court went about its task. Naturally, there has also been talk about doing something to prevent such episodes from repeating again. It is earmarked to go down in history as a “bad” order. Bad orders have, historically, carried serious repercussions for the criminal process in India, especially in the realm of sexual assault laws. As Professor Baxi’s 2013 piece outlined, most of the notable changes to the legal system were a result of particular instances of injustice being done to women. In that way, Tejpal offers a chance to build consensus around problems afflicting the investigation and prosecution of sexual assault cases in India. That is the only silver lining around this dark cloud.
At the same time, one must be wary about the ripple effects of bad orders. The more recent examples of a consensus did see a response from the legislature, but hardly the kind of response which anybody except vote-seeking politicians supported. In 2013, parliament rushed amendments to the criminal law by picking and choosing parts of a Committee Report, and the outcome was insensible and illogical. In 2018, the outcome to public outrage was simply increasing the severity of crimes and adding the death penalty for some crimes, which many argued was counterproductive to the issue itself. The underlying commonality in these responses is their cost-effectiveness — legislators create new crimes or evidentiary rules but they hesitate in taking steps which cost money, such as improving state capacity to investigate and prosecute a case.
This post suggests that there are some serious problems as well as red herrings out there when it comes to thinking about how the system deals with sexual assault cases. To make things better it is imperative that the conversation going ahead is rooted in an approach which considers the problems holistically — looking at the entire life of a sexual assault case and its consequences on stakeholders while recognising the unique harms to victims — instead of adopting band-aid measures which only worsen the malaise.