Depositions and Rationality
Reading this guide on deposition strategy I was impressed by how many of the techniques lawyers use on uncooperative witnesses seem like good rationality techniques (a deposition is a sworn oral testimony taken out of court without a supervising judge).
Searching LessWrong, nobody seems to have noticed the parallel. Searching the CFAR participant handbook, a number of these techniques are fitting and novel.
Applicability
These techniques are designed to pull out answers from uncooperative witnesses that are legally compelled to answer truthfully. Uncooperative witnesses are often coached by a defense team in avoidance techniques.
Avoiding difficult or unpleasant questions is a common self-defeating impulse. This is true both individually, and in social interactions. Hesitation is always easy, rarely useful. Legal repercussions are normally absent, so social and group truth-seeking are not fully identical to a deposition, but the formal commitment to finding a truthful answer may be a sufficient common factor.
Summary of techniques
Technique zero is to recognize that it is likely that you will face an evasive witness, and commit explicitly to generate a written plan to deal with them.
“Marshal the facts and standards that you can use to build toward your ultimate question”. Before requesting a conclusion (“did the motor carrier you represent violate Federal Motor Carrier Safety Regulations?”) request individual facts that build up to your conclusion (“does FMCSR apply to your company?”, “does FMCSR requires that your company ensure compiance by drivers?”, “do you provide training and testing for your drivers on the contents of the FMCSR?”, “does FMCSR requires that drivers keep accurate logs of their trips?”, etc).
Exhaustion. Make the witness commit to their answer by asking “what else?”, “is that all?”, “is there any material you need to fully answer the question?”.
Avoiding rabbit trails. The correct response is to let go or write down the tempting line of questioning and proceed with exhaustion.
Example: “Q. Give me all the reasons you believe Dr. Smith conformed to the standard of care. A. He identified the median nerve at the time of the surgery which was abnormally small and congenitally demyelinated, he protected the median nerve with [...].”
Incorrect response: “Q. What do you mean abnormally small?”
Correct response: “Q. Is there any other reason you believe Dr. Smith confirmed to standard of care?”
Restate and summarize. “Exhaustion of a particular topic may require many questions and many pages in a transcript. If left in its raw form, the testimony maybe unmanageable and unusable with a jury or court.”
Boxing in, by bracketing[1]. People who claim to have no idea about a quantity will often give surprisingly tight ranges when explicitly interrogated.
Example: “Q. How far apart were your truck and Mrs. Agan’s car? A. I don’t know. Q. Could it have been at least 5 feet, say the distance between you and me? A. No, that is to close. Q. Could it have been 15 feet, say the distance between yourself and that wall? A. No, that is too far. Q. So it is fair to say that Mrs. Agan’s car was between 5 and 15 feet from your truck? A. Correct.”
Boxing in, facts-witnesses-documents. “A technique that forces the witness to commit to testimony and/or describe any and all possible circumstances that might allow their future testimony to change. Witnesses explain their change in testimony by using one or more of three broad categories of information that a witness didn’t have or consider during their deposition. These three broad categories are: facts they did not know or recollect, witnesses/individuals they had not spoken to at the time of the deposition, and/or documents they had not seen, recollected or considered. If none of those things exist, there is no basis for the witness’ testimony changing.”
Creating commitments from witnesses. Overlaps significantly with 1 (“marshall the facts”), focused on establishing the witness is/should be able to answer the questions. This is done by establishing that they are the person with most knowledge about the matter of the questions, that they are bound to answer and aware of this, that they do not need further preparation to provide answers. This line of questioning may reveal that the witness is not able to answer your questions to satisfaction, and you should be questioning someone else.
Dealing with “I don’t know”. Two broad strategies are outlined:
“Nail down the fact that this witness doesn’t know something of importance” using facts-witness-documents box-in.
“Convince the witness that while he/she ‘may not know’ or they are ‘not sure’, there is a plausible explanation/definition/standard that they will (inevititably) accept as true.”
Setting a timeout to think of answers (“Having thought about it for over a minute can you think of any other safety reasons for the ‘No Left Turn Rule’ for truck drivers?”). Arguably part of exahustion.
Dealing with witnesses that want you to define your terms. Three broad strategies are outlined:
“Ask for and Adopt the definition used by the witness”
“Use the person’s life experiences to create a reasonable, fair definition”
Use a regulation/dictionary/thesaurus to establish common use
Dealing with witnesses that want to avoid responsibility by quibbling with word choice (“That’s just a guideline, not a rule”). The response is to establish that the norm is expected of them, that they expect them of others, and for good reason.
Dealing with hedging answers (“for the most part”, “mostly”, “not necessarily”). Call out that the answer leaves room for doubt and interpretation, and ask the witness to make exceptions explicit until exhaustion (“that is all”).
Demolishing rationalizations. This is not explored in detail. An example is given. (see after “When a witness has obviously manufactured some bogus rationale”).
Interrupt witness rants.
Do not allow witnesses to interrupt your questions.
Points of coincidence with the CFAR Handbook
“Marshalling the facts” is similar in spirit to identifying cruxes, in that any disagreement is supposed to be reduced to an explicit minimal argument.
“Convincing a witness to accept a standard” uses the framework of policy level decisionmaking: making a witness accept that a norm applies to them is easier by having them endorse it first; it then becomes blatantly hypocritical for them to reject it.
Setting apart some time for a witness to actually try recalling further information is almost exactly the same as resolve cycles.
Points of disagreement with the CFAR Handbook
“False beliefs feel the same as true ones” means that witnesses are probably too willing to nail down their testimony and end up accidentally perjuring themselves (“Q. If you were provided facts about the position and description of the other cars, would that make a difference? A. No Q. If you had a conversation with one of your passengers about what they remember [...]? A. No”).
“Boxing in by bracketing” or “calibrated estimation” is missing from the handbook
- ↩︎
This technique appears almost verbatim under “calibrated estimation” in Douglas Hubbard’s How to measure anything.
Am lawyer. Have deposed 30+ people, defended 30+ depositions, and was a “potted plant” or active observer in another 20. All the items you mention from the article are good advice for learning how to depose someone.
I haven’t done any CFAR except insofar as it’s baked into the material here. I’m guessing CFAR assumes that both participants at least believe that they want the same thing—to model reality better, to reach and resolve a crux, something(s) along those lines. Certainly for crux-hunting, you need two people who are fundamentally collaborating.
Depositions are...really different. Background might help. 100 years ago, relative to today, everybody just showed up at trial with surprise witnesses and documents in what is known as “trial by ambush.” This was like Texas Hold ’Em.
The Federal Rules of Civil Procedure, first published in 1938, aimed to do things dramatically differently by requiring each side to give the other side all of its important documents and witness information, and made depositions a bigger and more useful thing. This is like Texas Hold ’Em except that your hole cards are face up and visible to the other players. The idea was to encourage settlement (and thereby reduce the number of cases on court dockets, and the attendant administrative burden) by forcing everyone to give the other side all the stuff that hurt their case, instead of just the stuff that helped their case, or nothing at all. Every State I know of has since followed suit with similar rules frameworks to govern litigation in State courts. It really works at incentivizing settlement.
Depositions serve two purposes:
Learn about the facts your side doesn’t already know (adversarial; as you can imagine, lawyers try not to disclose bad facts or documents, but it’s easier to get deponents to tell you that stuff in person for the usual savannah-monkey-brain reasons).
Tie down the witness’s trial testimony (adversarial; if Witness 1 says X at deposition, and Y at trial, you get to beat them up in front of the jury, and it’s really fun).
Notice that nobody’s incentive here is “expose the truth”—not the courts’, not the parties’. Not until a jury is impaneled is anyone sworn to do anything along those lines.
At bottom, depositions are a fight. Most witnesses have been prepared by the other side’s lawyer and are not going to make it easy on you. You have to use the techniques above to pin them down, and they fight like cornered animals as you do so. People sweat, get hangry, pit their own executive function against the lawyer’s, play stupid word games they would never play in front of a jury of strangers, answer evasively, interpose leading objections to hint to the witness how to answer the question. There are rules for when you can abort a deposition based on witness or attorney misconduct, because of course we need such a rule.
Most of the time, if I’m doing this to a witness, it’s because I’m trying to get them to say something they don’t want to disclose and I think that me staring at them in silence will trigger savannah-monkey-brain compliance instincts, not because I think they need more time to remember stuff. They probably met with their lawyer for at least an hour in advance. They probably didn’t sleep well the night before because they were thinking about the case.
This is vanishingly rare in my practice. If the witness’s lawyer did zir job, the witness is maximally unwilling to be nailed down.
I have started incorporating rationalist-adjacent concepts into my deposition strategies, but it’s really only effective against expert witnesses like engineers, surveyors, doctors, lawyers, for attacking the factual and inferential underpinnings of their opinion-for-hire. Most lay witnesses really, truly don’t understand what you’re talking about if you start getting into statistics or even really basic epistemics.
This doesn’t seem like what policy-level decisionmaking is about per the handbook. PLD is more like “it’s hard to decide edge cases consistently with your true values when you’re fatigued, so pre-commit to a decision policy matching your true values when you’re not fatigued”. I can see a connection to the standard-endorsing strategy, but I have to squint pretty hard.
Edited: to remove ableist language and replace it with the “trial by ambush” phrase.
It has been pointed out to me that therapy is analogous to depositions in a way relevant to your argument: in therapy both patient and therapist are there with the stated purpose of resolving emotional tensions in the patient, but the patient can prove unhelpful or actively oppose the therapist’s probes.
I think this is an example of an interaction that is collaborative in principle, but where techniques designed for adversarial interactions may do good.
I agree with that particular observation about therapy, having been therapized effectively myself (big ups to therapy!), but I fundamentally trusted that the therapist knew what ze was doing—based on diplomas on the wall, recommendations from trusted sources, and the first session with zir. The time that I did not feel that fundamental trust, I didn’t go back. Thus, when a therapist asked me a question that bothered me, I believed that 1) ze had a good professional reason for asking, so I should figure out the answer and 2) my discomfort about the question was worth investigating as well. And maybe even 3) the discomfort is worth pushing through because it’s a signal of a possible avenue for a breakthrough, and given that I value previous breakthroughs at $2,000 cash or maybe more if I thought hard about it, the discomfort is exciting in a way.
Contrast: If I’m a deponent, and the lawyer asks me a question that bothers me, my (sensible) belief is that the truthful answer is bad for my case, meaning that the amount of money I’m going to {win/lose} is going to go {down/up} when I answer. If you thought that answering a certain question truthfully would cost you $10,000 USD, you wouldn’t answer it if you didn’t have to.
Thus, another crucial distinguishing element of a deposition is that the deponent will face serious consequences if ze doesn’t sit for zir deposition: if the deponent is a party to the lawsuit, a judgment will probably be entered against zir, and if the deponent is a non-party who has been properly served with a valid deposition subpoena, ze will be held in contempt and theoretically jailed until ze does comply. Few people, by contrast, are compelled by sanction of law into therapy or crux-hunting.
Thank you for the comment!
Parts of legal training and advise can be applicable to rationality and truth-seeking—much of it is designed to reduce ambiguity and produce clear, direct propositional statments. But this doesn’t seem to be in that set.
It’s very much adversarial and starting with the presumption of deception and evasion, rather than simple ambiguity or misunderstanding. This isn’t truth-seeking, it’s result-seeking.
Whether or not you are truth seeking or not depends on your own goals and not whether you presume that the other side engages in deception and evasion.
Using strategies that still work when some people act adverserial with you and try to deceive you is in line with being rational.
I think this gets close to the insight that motivated my post: a part of ourselves often tries to curl into a ball and deny reality to avoid emotional stress, interacting with that part of you is kind of adversarial.
If truth helps one side, it probably hurts the other side. Well, not literally; there may be factual statements like “the crime happened on Tuesday” that neither side objects to. But the ultimate goal is, for one side to get the other side punished, for the other side to avoid punishment.
Not sure what this means in practice, but my first impression is quite bad. Saying “I don’t know” in proper context may be the right thing to do.
“I don’t know” can be a accurate. I think the advice is intended against people playing dumb, like Bill Clinton’s “depends on what the meaning of the word ‘is’ is” or this witness denying knowledge of what a photocopier is. I know I’ve pulled this bullshit on myself at least once.
And most of the time their original “no idea” will be more accurate than the stuff you made them make up.
I do think there’s a rationality skill implicit in the text: the “coaching” that witnesses undergo to avoid giving answers they don’t want to give. That’d be worth learning, as it’s literally defense against the dark arts. And the test for it could be an interrogation of the kind that you describe.
It seems to me that most of the techniques are absolutely great for attempting to prune evidence to fit a preordained conclusion. Knowledge of the techniques is probably fine, and very careful application of them away from a conclusion may be prudent sometimes, but as internal evidence collection and evaluation techniques they seem to be the very opposite of rationality.
The mindset being employed here is extremely insensitive to the evidence that, actually, things are complex and people aren’t just “being evasive”.
I’m not sure about this. Arguing with myself can get really hostile as it is, and a lot of the OP seems to encourage an adversarial mindset.
On the other hand, I think there’s definitely potential here. Generating ranges with upper and lower bound questions seems super useful, for instance.