You Have the Right that I’d Pretend that You Are not Silent

One objective of this blog is to act as an idea catcher. That is, it forces me to write down ideas and develop them beyond what I would otherwise do if I were just speaking to myself. But this implies that it will not be rare for me to bring up at least some ideas that would seem self-evident to other people or that have been previously addressed in the literature. And while I invite readers to comment and tell me—“How can you even be surprised by such a triviality?”—I will not limit my writing to only those ideas that are necessarily new.

With this caveat in mind, let me trace my reflections on the right against self-incrimination (the “Fifth” in Americanese) and a short argument I had over the issue.

My basic sense of this right was—and, mind you, I never practiced criminal law—that it is some variant of the right to remain silent. But this, on reflection, is not a good reason. The main rationale underlying the right to remain silent is that it is used as a shield against torture and some forms of coercion we deem illegitimate. But if we care so strongly about torture, our primary line of defense should be greater investigative transparency, financial liability of investigators for transgressions, free legal representation in allegations of this sort, use of statistical evidence to deduce criminal investigative patterns, greater scrutiny of police departments, etc. Applying this policy to all cases, even those where no coercion was alleged, is definitely a high price.

So I was wondering about the rationales of this strange right, and a few days ago I spoke with a friend who was a criminal attorney for a few years and who has some sense of the real world. She explained to me that such a right is imperative, because we respect people’s autonomy and their rights to express and not express themselves when they want to, and that we worry about coercion, which has been repeatedly proven to lead to false admissions by criminal suspects.

I argued, and I think that we were able to agree on this point, that coercion is irrelevant. Indeed, we might worry about coercion when someone confesses to something they did, but the invocation of the right against self-incrimination proves that the defendant has managed to resist coercion, not succumb to it.

That leaves the autonomy argument, which belongs to the class of arguments that I am strongly suspicious towards when they are used in legal discourse. Ronald Allen does a good job here of debunking many of these autonomy-based arguments. I just want to stress the analytical peculiarity involved here.

In essence, this privilege only indirectly interacts with autonomy. Even in a privilege-less world, a defendant who does not want to speak could be given the right to remain silent

What this privilege does is not to let people shut their mouths; this is a right they have as a virtue of the general prohibition on torture and assault.

What it does is to reduce substantially or entirely negate the inference we are allowed to draw from such silence.

So, in a privilege-less world an accused that would refuse to answer the question, “Were you present at the crime scene?” would still be allowed to remain silent, as we do not force people to speak, but it would lead to the logical inference that he was indeed present or that wherever else he was present, is supportive of the accusation to some degree (we would need to discount the inference by the probability we believe the suspect is refusing to speak because he is ashamed to admit that he was engaged in some socially unacceptable behavior, such as cheating on his partner). Instead, we just say, “Oh, you don’t want to tell us? This is fine, we will pretend the question was never asked.”

I focus here solely on the autonomy argument, so I try to imagine what it might be. Consider World A, where the right does not exist, and judges adequately discount alternative non-guilt-related reasons for refusing to speak. Now consider World B where this right exist, and if the suspect is silent, the judge is not allowed to draw any inference from his silence. The situation of those who are guilty is better in World B compared to World A. This right, in essence, gives preferential treatment to those who are guilty. But this is very, very weird—we are absolutely fine with the cruel, wrong, and terrible punishment of incarceration (with all that follows it) for those guilty of crimes, but we feel squeamish about treating them instrumentally at the interrogation stage? Recall, with proper discounting (which you are welcome to debate), this privilege only favors guilty defendants.

P.S. What do we know (and think) about the distribution of procedural criminal rights between suspects of different crimes? I do not mean just the effect of having a better lawyer by virtue of being richer, but more about how often do we see certain rights used in different types of crimes, different defendants, different contexts, etc. controlling for attorney’s costs.

3 thoughts on “You Have the Right that I’d Pretend that You Are not Silent

  1. “If judges adequately discount alternative non-guilt-related reasons for refusing to speak”

    Unless the judge *knows for a fact* that the suspect is innocent, remaining silent is Bayseian evidence for guilt. Discounting this evidence means that the judge does not automatically consider this evidence to imply guilt–rather it means that the judge increases his estimate of the probability of guilt based on the silence. This increases the chance of convicting an innocent person compared to if the judge doesn’t use the silence at all.

    You seem to be trying to imply that “discounting non-guilt-related reasons” means “the probability of convicting an innocent doesn’t go up”. This is not true.

    Suppose that half of all suspects are guilty. All the guilty suspects remain silent and 1/4 of the innocent ones remain silent. The Bayseian likelihood that someone is guilty, given that he is silent, is 80%. The judge, *even properly discounting* would (based on this 80% chance) then convict everyone who remains silent. Result: innocent people, even ones with non-guilt-related reasons for remaining silent, will go to jail if they remain silent.

    Also, you ignore the impact of incentives. If silence can be used against you, that makes it easier for the government to trick you into saying something that can be interpreted as a confession, and therefore incentivises the government to do that.

  2. Good points. I edited a few sentences to avoid some of the confusion due to the original wording.
    Now, to use your example, without the right to remain silent, there is 80% chance that a silent suspect is guilty. With the right, the judge can only use the base probability, so the odds of a suspect being guilty is 50%. If a judge needs to be 75% certain to convict, the guilty would be convicted only when the right does not exist. Hence my point that the right subsidizes the guilty at the expense of the innocent.

    The incentive story is much more complex and I can see it going in many different directions. With our Bayesian judge, account could be given to counteract some of this effect.

  3. Without the right to remain silent, there’s an 80% chance that a silent suspect is guilty, so the judge convicts them all, including both the innocent and guilty ones. With the right to remain silent, silence is not evidence, so the judge doesn’t convict any of the silent innocent people, but may convict some of the silent guilty people if there is other evidence.

    That’s not a right which subsidizes the guilty at the expense of the innocent; it’s a right which protects the innocent. Without the right, the silent innocent people get convicted; with the right, they don’t.

Leave a Reply

Your email address will not be published. Required fields are marked *