Free Expression Scholars Conference part 6

Seana Shiffrin, Lying and the Law

Discussant: Jack Balkin

Chapter of book on sincerity and law. The big idea is that the pure lie, the lie whether or not it deceives, whether or not the thing stated is true, but merely a lie in the sense that the person believes/knows her statement to be false, should receive no protection under the free speech principle (which can be distinguished from the 1A).  It undermines a particular kind of social practice which is necessary to who we are as a people: a practice in which we say things to each other and believe that others believe them—the warrants each of us gives engaging in testimonial speech. Much social and moral life is based on relying on other people to tell the truth. An economy of warranted representations undermined when people lie, even when what they say is actually true or even when they don’t deceive.  Millian argument: lying helps us see the truth—but Mill wasn’t talking about deliberate lying, but rather error.

Coda: on a different set of political values, accommodation: reservations about legal regulation of pure lie that emanate from concerns about equality, fraternity and toleration. Roughly speaking, the notion that we understand ourselves to be fallible human beings who make mistakes. Given that, others are equally human, so accommodation and toleration mean that we don’t demand of others everything we might. There’s good reason to apply this idea generally, especially when people lie about who they are and where they came from. Might not apply when businesses/institutions lie because they don’t give rise to the same considerations.

Key idea: there is damage that lying creates to the collective testimonial framework. Thus the argument that lies don’t harm particular individuals is not relevant, given the structural harm. Not an empirical argument, so the demand for causal proof is inapposite.

So why is it not empirical?  If we looked for empiricism, we’d have something like Gresham’s law (bad money drives out good) or Akerlof’s market for lemons, where credibility decreases for everyone. But the nature of the transaction is different—it’s more about the used car salesmen than the cars; you trust everyone you deal with less because of lying.  But that’s an empirically verifiable hypothesis about asymmetric information, and there are some conditions under which it works and others in which it doesn’t. If this is instead a deontological claim about moral duties, then it’s puzzling because arguments like Shiffrin’s arguments about counterspeech appeal to effects.

Great argument why these bans aren’t content based under current doctrine. One could argue that defamation law is content based, but if defamation is like Shiffrin’s argument about truth-telling—public discourse is sustained by people’s ability to say things without being defamed (or even privacy is sustained by such an ability)—then it should look content-neutral too.  But the doctrine doesn’t work that way presently.  What she says about lying has broader implications for 1A law.

Shiffrin says suppressing pure lies don’t violate the free speech principle. But a general antilying rule would invite discretionary/arbitrary prosecution. Maybe civil fines/actions would be more appropriate. Higher proof principles (as in defamation of public figures).  Need for accommodation and tolerance too.  She produces a set of rules that looks very similar to what you’d get under 1A doctrine: lies generally protected unless there’s individualized harm.  Comes out close to doctrine today.  So maybe the 1A is not the free speech principle, but rather contains a number of other values along with that: accommodation and tolerance; we already know the 1A concerns state overreaching; etc.

Shiffrin: Not primarily an empirical argument because the core argument is that when people lie they give you a reason to stop believing testimonial warrants, so what is produced is a reason rather than an effect.  The damage is not the effect.  For the state to act to prevent/express its displeasure at a practice that reduces/eliminates reasons for believing testimonial warrants doesn’t have to show citizens become cynical and disbelieving. When the liar reduces the reasons you and I have to accept testimonial warrants, they put us in a position where we have to gather particularized knowledge about person and conditions under which she’ll misbehave—particular epistemic burden to engage in empirical research rather than accepting each other’s testimonial warrants. This happens whether we know it or not. Either we respond to it or we accept warrants irrationally.

Balkin: but couldn’t Akerlof say the same thing?  He’s discussing asymmetrical information.

Shiffrin: yes, it’s a model. The reasons we have to respond to models differ in different contexts. In the economic market, where we are interested in getting goods distributed well, it may be that actual reactions are more important than knowledge. I’m arguing about what epistemic reasons we’re entitled to have and not whether we manage to navigate through the world without adequate reasons because we have other empirical guides to what’s relevant/not relevant. Car market v. moral relations—treating people as other humans rather than anthropological units.

My main argument about counterspeech is that uncovering the truth doesn’t restore the reliability of the speaker and others who supply testimonial warrants, though as to whether the listener is at risk with respect to the accuracy of the content, that will depend on empirical facts. Counterspeech in general suggests we should move from relying on what people say to an empirical investigation of whether they are trustworthy, and that’s an epistemic loss.

Not as confident about extension to defamation law. Because defamation law requires that you show that the speech is false, that’s a form of identification by content and discrimination by content. We don’t seem bothered by that, so why so pressing in Alvarez?  She doesn’t think the regulation of the lie turns on actual falsity, but all deception law—which we have a lot of—is content-based if you think that having to show that proposition uttered is false is a form of content discrimination. She doesn’t actually think regulating pure lie is content based, but doesn’t care much unless it involves a certain kind of gov’t effort by intent, effect or targeting to restrict/punish people for engaging with ideas and attempting to engage them sincerely, and content based/content neutral distinction doesn’t always do that work.

1A v. free speech principle—less confident that worries about gov’t abuse are really outside the free speech principle, because they encompass concern for abuse that hinges upon attempts to stifle speech.  Accommodation principle is more about acknowledging shared defects in our moral character that happen in this case to be about what we say but don’t have to be.

Vincent Blasi: we have a system of epistemic reliance, but also public reliance on officials.  NYT v. Sullivan principle is costly in that way; perhaps the tradeoffs are worth it. But isn’t there a reason to think that the balance is not as lopsided as Sullivan said. Or is there a qualitative difference in the epistemic value of sincerity v. competence?

Shiffrin: duties of accuracy are different. Requirement not to deceive is founded in duty to take care w/r/t each other’s beliefs.  When appropriate, ensure what another walks away with is accurate. That’s at the bottom of a lot of defamation issues when they turn on what’s false rather than on sincere representation of what speaker believes to be true.  Doesn’t have a particular theory across contexts on how strong duty of accuracy is. Should be fairly strong for press, which represents itself as expert, and have some privileged access that common citizen doesn’t in scrutiny of gov’t.  Still comfortable with NYT privilege, though it would still be wrong for the press to abuse it.

Sincerity argument may well extend to political discourse.

Balkin: political lie is more important from the shoring up trust standpoint than the lie about the used car. Response is wariness about partisanship, favoritism, corruption and undermining of fairness. Value comes from democracy/rule of law than conception of free speech.  1A as overlapping set of political/moral/other considerations.

Shiffrin: Agrees that in order to assess whether regulating political lie would be good idea you’d have to look at culture produced by the regulation.

Balkin: you’re about what we owe to each other, I’m talking about culture.

S: we owe to one another a particular kind of culture that stimulates freedom of thought. But we do have to look at culture can have a well-crafted legal regime that won’t undermine free and robust discussion of gov’t. Whether we’re capable of that may vary according to time and place although the underlying principles don’t vary.

There are a variety of different arguments contributing to our understanding of free speech/1A, but less confident that they aren’t principled or capable of principled ordering.

Adam Kolber: why regulation of lying might be content neutral: as I understood it, you said lying was about the relationship to the expression, so two people could say the exact same thing and one is sincere and the other not.  The fact that only one is regulated would be a reason to find content neutrality.  How to apply to other kinds of attitudes: one person might speak ironically and the other not? 

Shiffrin: I’d want to know the motive for regulation.  A content neutral regulation can be assessed by target, purpose, and effect.  Right to say that the intrinsic relation between speaker and utterance need not be content based though it may be as when identity of speaker determines meaning of statement. Ability to speak ironically is important to express self.  But doubts that ability to represent self insincerely is essential aspect of ability to externalize mental content.  Might have a problem on effect side as well as purpose side if you’re regulating irony.  Why would we care about difference between content based and neutral regulation? Because content based captures main concerns about regulating speech. If I were able to show that those concerns were implicated by irony regulation, that should be enough; not interested in tracking doctrine.

Alexander Tsesis: When a physicist speaks about string theory, they know they’re fudging the numbers because the equations are too complex.  When we say the sky is blue, we should know that it’s actually a multiplicity of colors.  We say walls are solid but they’re not. Our communications are so inexact, and we trust inexact experts, and we tell our students there are three levels of scrutiny.  So what’s a lie? Sometimes we’re insincere even when we know.  People expect answers and so we answer despite our doubts.

Shiffrin: the lie is presentation of proposition to be taken as true objectively. But standards of truth with respect to precision vary quite a lot according to who you’re speaking to. Requisite level of precision to students varies from to colleagues, judges, child.  When we have that context fixed, may you permissibly represent what you say as true within the standards of the context even when you believe it’s not?  My topics are the last part, not the interesting questions of how precision is specified within a context and what happens when people attempt to alter the underlying standards through their speech.  If the string theorist speaks in the way she speaks to children to her colleagues, then something has gone terribly wrong.

Zephyr Teachout: Past election, when Paul Ryan repeated that Obama took $716 million out of Medicare. The story is similar to Shiffrin’s: repeating of lies has led to nihilism in citizens’ understanding and acceptance of information, so they make judgments not based on information.  If Paul Ryan repeats that in a speech, would you accept a law regulating candidates’ false knowing claims about facts?  What if it was specific to insurance companies?  That is, what are the implications of allowing context/identity to affect the rules?  Could we have different standards for SuperPACs? 

Agrees it’s dangerous to bring all values inside the 1A and make the 1A our Constitution, because it tends to put a 1A hurdle in front of anything.

Shiffrin: Doesn’t accept the instrumental reason for deciding whether something falls into the 1A or not. Accepts the idea that overly rigid const’l interp. interferes w/ our legislative freedoms. Thus there should be boundaries to the entire const’l structure. But is less confident that such reasoning applies to any particular provision, especially since she thinks application varies according to social context.  Let’s just see where the arguments go.

Putting aside the issue of the facts, let’s suppose Ryan said something he didn’t literally believe but that is subject to contextual better understanding.  (Balkin: Obama reduced payments to healthcare providers, and Ryan called this a cut.  Dispute over whether this is a cut or savings in services. Ryan also voted for these cuts twice; he liked it except that it was Obama.  But he said “it’s a cut.” Shows the molasses of these disputes.)  That’s a better example of deception than of a pure lie.  He’s culpable for believing it and for believing it’s the appropriate unit of information to transmit, but my argument isn’t about deception.  What matters to me is not its falsity but that he doesn’t believe it and represents it as true.  Free speech doesn’t bar regulation of that even by politicians; would certainly think business organizations were also covered. Gov’t abuse arguments have more sting w/r/t politicians but how that should go depends more on political culture/barriers we can create to abuse.

David Pozen: not whether we could regulate lies but whether we should. Would have to know a lot more about enforcing prohibitions on lying. Many contexts have strong moral, social, reputational sanctions against lying already. Would want more info on their efficacy before knowing about whether extra regulation would be helpful.  Paper deals with costs like potentially sapping intrinsic motivation to be truthful; would we lose anything by not having collective moral response calibrated on the fly about how to deal with lies?  Something more than accommodation and toleration may be compromized by legalizing response to lies—deliberative process of figuring out boundaries of acceptable behavior and appropriate punishment, if any, for lies. Something valuable happened in response to Paul Ryan that could be lost if it moved to prosecutorial decision.  (I think this weirdly puts lawmaking off the table as a result of a deliberative process.)

Shiffrin: I take my argument to be necessary but not sufficient and take no position on when we should regulate the lie. These conversations are useful too. 1A concerns may get in the way of having a larger discussion of whether we have a problem that the law can address.

Sociologically, we’re not good at taking lies seriously. We’re stunningly casual about accepting that others lie and that we can’t have a healthy political culture.  Whether legal regulation would make a difference, not certain that we have to understand legal regulation as taking over/monopolizing our social and moral attitudes rather than reinforcing them. Legal regulation can announce and affirm that we think this is a serious harm that is not merely interpersonal. We’ve seen laws play leadership roles rather than occupying morality and taking it over. Antidiscrimination laws helped to actually give a number of people the ability to make social claims of each other; illegality gave a reason to insist on compliance. But the law wasn’t enough. People had to find ways to manage problems, but the legal expression of the wrong made a difference in helping people accept that this was their responsibility.  Look more to that model than to that for pharmaceutical regulation, which is supposed to take off our agendas the question “what can I ingest?”

Leslie Kendrick: Structure of project overall—relationship between duty of sincerity and conception of freedom of speech.  Kendrick’s take: sincerity is a moral duty that we all have, and then you’re asking whether regulating lies in the service of this duty can be consistent w/freedom of speech.  Freedom of thought seems related to this sincerity requirement—can you say more?

Shiffrin: duty of sincerity is rooted at least in part in political duties to treat one another well, respond to one another as individuals; we need to cooperate to do so, and to cooperate we have to know what others think as they’re willing to share it. Duty of sincerity is essential to the functioning of the economy of epistemic sharing. One of the arguments for it is that our political obligations depend on it. So to freedom of speech depends on those values—to become the kinds of people we have the right and obligation to be, we have to develop as thinkers, and ability to engage in sincere speech is an essential component of that. What underlies the duty of sincerity underlies the duty of freedom of speech, so it’s not shocking that they’re fully compatible even if the duty of sincerity were legal rather than moral.

Would there be an obligation to have a legal duty of sincerity? There certainly are freedom of speech values that depend on sincerity, and we need a full culture of sincerity to meet our duties of justice well. I would turn to specific questions of how regulation functions in a sociological culture to answer whether we need legal regulation. If we can manage w/o legal regulation we have no obligation to legalize it anyway, but is open to argument that law would foster free speech values.

Tabatha Abu El-Haj: are you concerned we’re complacent about lying throughout?  In personal relations, we seem quite concerned or at least there are quite varied reactions. But in political culture we’ve reached a place where politicians lie all the time and don’t think they need reasons. Struggling with empirical/nonempirical divide.

Shiffrin: doesn’t mean for paper to rest on sociological claim.  One’s sociological views have an influence on whether you think this is a problem or not.  Not sure we’re doing well at interpersonal moral level—many people lie casually to friends, to corporations, when they think it’s trivial.  When else are they willing to lie?

Meant accommodation argument to apply to autobiographical speech—reasons to tolerate that weakness. But when people take on expertise, fiduciary duties, there’s no political or legal reason to tolerate that.

Q: you can be insincere by asserting something beyond your level of confidence: you can make a testimonial warrant and believe that it might or might not be true.  Wouldn’t your principle also apply to that?

Shiffrin: my position on what a lie is does extend to that—if I say “p” but believe “maybe p,” that’s a lie. But because for legal purposes it may be difficult to ascertain level of confidence of belief as well as level of confidence contained in the assertion, she’d make part of the standard of proof an active belief in the contrary.