I represented the manufacturer.

The judge gave us thirty minutes for voir dire. Thirty minutes, in a wrongful-death case against a car company, in the dead man’s hometown, with his widow in the front row.

About ten minutes in, I asked the panel whether anybody had ever boycotted a business or a product because of something the company did or stood for. A few hands went up. We talked about it. And then I said—I don’t remember the exact words—something like: “My wife boycotts Target. Has for a while now. And I’ll tell you what, it’s saved me a lot of money.”

The room laughed. Everybody knew exactly what I meant. And from that moment until the verdict came back, I had a different jury than the one I’d walked in with.

For a long time I would have described what happened as a connection. A moment. Good instincts and a little luck. Most trial lawyers describe their best voir dire moments the same way: something clicked. They liked me.

I’ve come to think that’s the wrong frame, and that the wrong frame is not a small mistake. What happened in those thirty seconds was a transaction. I asked that jury for something specific, and they gave it to me. Everything I did for the rest of the trial either spent what I’d been given or earned more of it. To explain, I need to start with what jurors are actually doing while the lawyers are putting on evidence. It is not what any of us were taught.

The decision comes first

Here is the claim, as plainly as I can state it: jurors decide before the evidence. Not completely, and not consciously. But long before a case is submitted, most jurors have a working answer to the question that ends up controlling the verdict—whose story is this trial testing—and every piece of evidence that arrives afterward gets bent, quietly, toward that answer.

That is not courthouse cynicism. It is about as close to consensus as the empirical study of juries has produced.

Start in the mid-1980s, when psychologists Nancy Pennington and Reid Hastie—she at the University of Chicago, he at Northwestern—recruited people from real jury pools, showed them a realistic filmed murder trial, had each one reach a verdict, and then interviewed them at length. What they found became known as the Story Model. Jurors do not sit with a mental score card, weighing each witness, tallying each exhibit, applying the instructions at the end. That is the lawyer’s model of a juror, and it is wrong. From the first information they receive, jurors build a narrative—what happened, who these people are, who is trustworthy—and jurors who built different stories from the same evidence reached different verdicts. The story is not a summary of the decision. The story is the decision. Their follow-up work sharpened the point: the easier one side makes it for jurors to assemble its story, the more verdicts it wins. Jurors don’t total a ledger at the end. They ask which verdict fits the story they have already built.

Which raises the question that trial lawyers should continue: what happens to evidence that arrives after the story has taken shape? Kurt Carlson and Edward Russo, at Cornell, answered it in the study I would make every litigator read. They gave people a civil case one piece of evidence at a time and asked, after each piece, who’s ahead. Once a juror had a leader—a tentatively winning side—she stopped evaluating new evidence on its own terms and started bending it toward the leader: jurors with the defense ahead read an affidavit that mildly favored the plaintiff as neutral, or even as helping the defense. They named it predecisional distortion, and it grew with the juror’s confidence. Then the detail that should make you sit up. They ran the study on students, then on real prospective jurors fresh from courthouse orientation. The jurors distorted the evidence twice as much as the students—and despite an explicit instruction to set aside prior beliefs and decide strictly on the evidence, their general attitudes about plaintiffs and defendants leaked into their verdicts anyway.

Dan Simon at USC has documented the same machinery under another name—coherence shifts: as a verdict starts to form, the read on each piece of evidence shifts to cohere with it, supporting evidence firming up, contrary evidence going soft, until the decision feels easy. In his studies it happens within a single hour, and people don’t know they’re doing it; looking back, they misremember their earlier views as always matching where they ended up. Nobody feels themselves sliding.

Then add the filter running underneath it all. Dan Kahan and his colleagues at Yale showed people the same protest footage—half told it was an anti-abortion protest outside a clinic, half a protest against the military’s don’t-ask-don’t-tell policy outside a recruitment center. Asked plain factual questions—did the protesters block pedestrians, did they threaten anyone—people of opposing cultural outlooks, watching identical footage, gave sharply different answers. Each side saw the facts its identity needed it to see. Jurors carry those filters into the courtroom: before they have consciously decided anything, their gut has sized up who in the room belongs to their world and who doesn’t.

And if you are counting on deliberation to sort all this out, the field data is brutal: in Kalven and Zeisel’s classic study of American juries, replicated in the field since, the verdict favored by the majority on the first ballot becomes the jury’s final verdict over ninety percent of the time. Deliberation doesn’t re-try the case. Deliberation ratifies it.

You may have heard the CLE-circuit version of all this—that “eighty percent of jurors decide by the end of opening statements.” That number is folklore, a garbled cousin of the first-ballot finding. What the research actually shows is stranger and more useful: jurors don’t finish deciding at opening. They pick a leader early, then bend each new exhibit toward it, compounding, without ever feeling it happen. Trial is not a scale, where every piece of evidence lands with its own fixed weight. Trial is compound interest.

If that’s right, then the three most familiar lessons of trial training are backwards. Take them in order.

Permission

We teach voir dire as a removal exercise: identify the jurors who can’t be fair, take your cause challenges, spend your peremptories intelligently and, if your trial-advocacy program was feeling generous, “build rapport.” That is the whole curriculum.

But if jurors start building the story the moment trial starts, voir dire is not the paperwork before the trial. It is the casting call. Somebody in that courtroom is going to be cast as the trustworthy narrator, and somebody as the one with something to hide. Each juror’s gut is answering a question she isn’t consciously asking: Is this lawyer the kind of person whose side of the story I’m willing to take seriously? I call the yes answer permission. It isn’t agreement—nobody has heard evidence yet. It isn’t likability, exactly; likability is downstream of it. It is the pre-rational grant of standing to be heard. And once it’s granted, Carlson and Russo’s machinery starts running in your favor: your evidence lands softer, your objections sound more reasonable, the ambiguous exhibits bend your way.

Go back to the Target moment and look at what it actually did. It was a real question—I wanted the answer and had no idea who would raise a hand—in a room conditioned to expect speeches disguised as questions. It surfaced the most dangerous attitude in the venire, the impulse to punish a company for what it stands for, and got the panel describing that impulse out loud, in front of each other, before opposing counsel could dress it up in opening as civic courage. An attitude spoken aloud loses a great deal of its operational power: the juror who told the room she’d boycotted a company is the juror who, days later in deliberation, hears “we need to send a message” and thinks—isn’t that the thing the defense lawyer asked us about?

And the line about my wife told the panel something true about my household without claiming membership in theirs. I never said I was one of them; I gave them enough to place me themselves, and the conclusion a juror reaches on her own is ten times more durable than one a lawyer tries to install. Remember the protest video—jurors see what their identities prepare them to see. You want their identities working for you. Even the laughter wasn’t decoration: a juror who laughs with a lawyer in front of eleven strangers has just publicly affiliated with him, and affiliation is the precursor of permission.

One caution, because it is the difference between the move and the con: the line worked because it was true. My wife really did boycott Target, and the register matched the venue because I matched the venue. Manufacture a moment like that and the jury will feel the manufacture, and it will cost you double. The skill is not inventing the line. It is recognizing, when a true thing from your own life is also the right thing to say in this room, that the moment is in front of you.

An unanswered story

The second backwards lesson is the cautious opening. Defense lawyers especially are trained to hedge—reserve opening, or deliver a short, careful one about burdens of proof and keeping an open mind, and wait to see the other side’s case before committing to a story of their own. In a story-model world, that is malpractice in slow motion. Jurors will have a working draft by the end of openings, because human beings cannot sit in a chair for four days without a narrative to organize what they’re seeing. If only one side offers a story, the jury adopts the one on offer—and predecisional distortion spends the rest of the trial bending every exhibit toward it. The plaintiff’s bar, to its credit, has internalized this for a generation; much of the defense bar still treats it as optional. The lawyer who finally tells his story midway through his own case is no longer narrating. He is contradicting. And contradiction asks jurors to admit they have been wrong for three days—which the coherence research says is nearly the one thing minds will not do.

In the band director case, the plaintiff’s opening was complete: a beloved community member, a devoted family, a machine that was supposed to protect him and didn’t, a corporation that chose profit over safety. Hero, victim, villain, demand for justice. The temptation was to spend my opening rebutting it—no defect, the system performed as designed—all true, all necessary to prove, and all terrible as the spine of an opening, because every rebuttal re-centers the story it contradicts. What we did instead was tell a different story: about what cars can and cannot do, about engineers who spent a decade building a system that prevents thousands of collisions a year, and about a moment on a road that no system in the world could have prevented. The villain in our story was not the family, and it was certainly not the band director. The villain was the false expectation that technology can save us from physics. That story absorbed the loss as real and located its cause somewhere other than the defendant. And it handed the jury a frame—so that when our expert explained the limits of automatic emergency braking, the jury did not hear a contradiction of the plaintiff’s story. They heard confirmation of mine.

The preacher

The third backwards lesson is that cross-examination is for scoring points. Here is the trial that broke me of that thought.

A slip-and-fall, some years back. The plaintiff claimed the fall had left her permanently injured—life changed, pain constant. Her lawyer called her preacher as a fact witness: a man of God who had known her for years and could describe, from personal observation, how the fall had changed her. Months earlier, at deposition, the preacher had testified under oath that she’d had some earlier health issues but was, by then, fine. Back to normal. On the stand, in front of the jury, he testified that she was devastated. Declining. Not what he had said before—not close.

Every trial lawyer is taught exactly what to do with that: impeach. Walk him through the prior testimony, read it into the record, sit down with the contradiction hanging in the air. It is the cleanest, most satisfying move in the playbook. And on that day it would have lost me the case. The jury did not yet know the preacher had contradicted himself. All they knew was that a man of God had given his honest observations about a woman in his congregation—and if the next thing they saw was a defense lawyer attacking him, they were not going to update their story about the preacher. They were going to update their story about me. The lawyer who attacks preachers is a character who does not get to win the case.

So I gave him outs. Gently. Did he remember our conversation a few months back? Did he remember being asked how she was doing? Might his impression have been different then? Every opportunity to soften, to allow that his observations had changed—anything other than making me pull out the deposition. Not because I was afraid of him. Because the jury was watching me decide how to treat him, and the only version of the next sixty seconds in which I won was the one where they concluded I had bent over backwards to be fair to a preacher who would not bend back.

He took none of the outs. So I walked to counsel table, picked up the deposition, asked if he remembered giving it, read the question, read the answer, and asked whether that had been his testimony. He had no choice but to say yes. I sat down.

The jury needed to impeach the preacher, not me. I just handed them the document and got out of the way. That is what cross-examination actually is—not a duel between lawyer and witness, but a negotiation between lawyer and jury in which the witness is the subject matter. Every cross updates two running calculations, and only two: can I trust this witness, and can I trust this lawyer. The two are linked. Go hard at a witness the jury already distrusts—the hired-gun expert on his two-hundredth case—and you gain credibility by association. Go hard at one they are still inclined to believe—the treating physician, the coworker, the preacher—and you will not change their mind about the witness. You will change their mind about you. And once the jury’s story recasts you as the bully, every ambiguous moment for the rest of the trial gets read against you.

The objection

The honest objection to all of this is that it sounds like jury manipulation dressed up in psychology.

Here is what I’d say back. Everything the research describes is already happening, in every courtroom, every day, whether the lawyers participate or not. Jurors build stories. They bend evidence toward the early leader. They read cultural signals before they read exhibits. That is not something lawyers do to jurors. It is something human beings do, and the studies say they cannot turn it off and mostly don’t know it is running. The only choice a trial lawyer has is whether the story the jury builds gets constructed with his client’s side of it or without it.

And notice what the moves actually require. Ask a real question and listen to the answer. Say true things about your own life, or say nothing. Acknowledge the loss, and look the family in the eye while you do it. Locate the cause of the injury somewhere honest. Impeach with the document, no theatrics, and only after the outs are refused. Every one of those works precisely because it is not manipulation—jurors have world-class detectors for the other kind, and the lawyer who lies to the tribe gets caught by the tribe. None of this is how to fool twelve people. It is how to stop accidentally convincing them not to listen to you.

We won the band director case. Not because our evidence was better than anyone expected. Because by the time the evidence arrived, the jury had already decided whose story it was testing.

So here is the discipline I would hand any lawyer headed to trial, on either side of the v. Stop asking “am I winning”—the evidence phase will not tell you. Ask instead: am I being granted permission, or am I spending it? If it is being granted, deepen it: be a recognizable human being, tell a coherent story, handle witnesses with restraint, let the jury reach its conclusions on its own. If it is being spent, slow down and ask what the spend is buying.

Jurors decide before the evidence. They start in the first minutes of voir dire, they refine the decision through opening and the early witnesses, and everything after bends toward the side that is ahead. By the time twelve people walk into the jury room, the verdict is usually just waiting to be counted. The real trial happens in the first hour. Everything after is execution.

Sources: Nancy Pennington & Reid Hastie, “Evidence Evaluation in Complex Decision Making,” 51 J. Personality & Soc. Psych. 242 (1986), and “Explaining the Evidence: Tests of the Story Model for Juror Decision Making,” 62 J. Personality & Soc. Psych. 189 (1992); Kurt A. Carlson & J. Edward Russo, “Biased Interpretation of Evidence by Mock Jurors,” 7 J. Experimental Psych.: Applied 91 (2001); Dan Simon, “A Third View of the Black Box: Cognitive Coherence in Legal Decision Making,” 71 U. Chi. L. Rev. 511 (2004); Dan M. Kahan, David A. Hoffman, Donald Braman, Danieli Evans & Jeffrey J. Rachlinski, “‘They Saw a Protest’: Cognitive Illiberalism and the Speech-Conduct Distinction,” 64 Stan. L. Rev. 851 (2012); Harry Kalven, Jr. & Hans Zeisel, The American Jury (1966); Dennis J. Devine et al., “Jury Decision Making: 45 Years of Empirical Research on Deliberating Groups,” 7 Psych., Pub. Pol’y & L. 622 (2001).

Adapted from a presentation to the Arkansas Association of Defense Counsel.

This is the list of books that actually changed how I practice. Eleven of them. Some are obvious. Some are not on most defense lawyer reading lists. Two are written by plaintiff lawyers I have spent years learning to defeat. The principle is consistent across all eleven: every book on this list rewired something specific about how I prepare a case or stand in a courtroom. If I cannot point to the rewiring, the book did not make the list.

A note on structure. The eleven books fall into five categories, ordered roughly by how foundational each category is to the rest of trial work. The Cognitive Foundation comes first because every other category depends on understanding how the jury’s brain actually works. The Empirical Foundation comes last because it is the hardest reading and the slowest payoff. The Persuasion Sciences, The Plaintiff Bar’s Playbook, and The Craft of Trial sit in between.

A note on links. Every book in this post links to Amazon with my affiliate tag. If you buy through these links, I earn a small commission at no cost to you. I would have recommended these books either way; the tag is not why they made the list. If you would rather buy from your local bookstore, most titles are widely stocked.

I. The Cognitive Foundation

Every other category in this list depends on understanding how the brain that decides the case actually works. These two books are the foundation.

Thinking, Fast and Slow—Daniel Kahneman (2011) [Amazon]

The operating system underneath everything else I read on trial work. The book changed how I prepare in two ways. First, it made visible something most defense lawyers will not admit: our strategy is pitched to System 2 while plaintiff’s is pitched to System 1. We make careful, evidence-driven arguments to analytical brains the jury is not actually using. Plaintiff’s lawyers, especially Reptile-trained plaintiff’s lawyers, are doing the opposite—building fast, emotional, fear-shaped narratives that operate beneath conscious analysis. The result is that on the level the jury is actually using, the defense is not in the conversation. Second, it gave me the question that now drives most of my trial preparation: where in this case do I need the jury to slow down and engage System 2, and how do I get them there? You cannot shout the jury into careful thinking. Telling them to “weigh the evidence carefully” does not make them weigh the evidence carefully. You have to create the conditions—pacing that forces engagement, questions they have to work to answer, specific moments where the answer is not obvious until the jury has done the cognitive work themselves. Plaintiff wins when the jury stays on autopilot. Defense often wins when, at the right moment, the jury slows down and looks at the actual evidence. I now prepare every trial with two parallel maps: where I am engaging System 1, and where I am explicitly inviting System 2.

Noise: A Flaw in Human Judgment—Daniel Kahneman, Olivier Sibony, Cass Sunstein (2021) [Amazon]

The follow-up to Thinking, Fast and Slow that fewer lawyers have read. The lesson I took: two jurors hearing identical evidence will reach different verdicts, and the difference is not signal, it is noise. That changed how I prepare voir dire. I stopped trying to identify the juror who would vote for my client and started trying to identify the variance—the jurors whose verdicts are least predictable from the evidence—and strike them. Kahneman, Sibony, and Sunstein draw a distinction defense lawyers rarely think about: bias is predictable error, noise is random error, and the two require different countermeasures. Most defense preparation addresses bias. Almost none addresses noise. The implication for case theme is that themes which feel narrow and specific actually reduce variance, while themes that feel emotionally broad—“this is about doing the right thing,” “this is about responsibility”—invite each juror to fill in their own interpretation, which is exactly what produces the runaway verdict. You cannot eliminate noise. You can refuse to seat it, and you can refuse to leave room for it in your theme.

II. The Persuasion Sciences

Three books on how persuasion actually operates—not how lawyers think it operates, but how psychologists have documented it. The first two are by the same author. The third is from a different field entirely and translates into trial work better than most books written for trial lawyers.

Influence: The Psychology of Persuasion—Robert Cialdini (2021) [Amazon]

The bible of persuasion psychology. Every defense lawyer should read it once and re-read it before trial. The principle I think about most: commitment and consistency. Once a juror commits to a position, even a small one in voir dire, they work to remain consistent with it through the rest of the trial. That is why I never end voir dire without asking the panel to commit to something specific that helps my case. Cialdini lists six principles in total—reciprocity, commitment, social proof, authority, liking, scarcity—and a defense lawyer who understands all of them is operating with tools most plaintiff lawyers are using without naming. Authority is what credible expert witnesses create, but it is also what deliberate restraint in cross-examination creates. Liking is why the lawyer who is genuinely curious about the case wins more than the lawyer who is performing certainty. The book is short. Re-reading it the week before trial almost always surfaces a move I had forgotten to plan for.

Pre-Suasion: A Revolutionary Way to Influence and Persuade—Robert Cialdini (2016) [Amazon]

The follow-up to Influence, and in my view the more useful book for trial work. Cialdini’s argument is that what happens before a message is delivered controls how the message is received. For trial, this means voir dire is not just jury selection—it is the period when the jury’s attention is being primed for the case to come. I started writing voir dire with one question in mind: what is the jury already paying attention to when I start opening, and did I put it there or did plaintiff? The entire phase improved when I started asking it. Cialdini’s central concept is what he calls the channel of attention—the narrow band of focus a person has in any given moment, and how dramatically the meaning of a message changes depending on what filled that channel just before the message arrived. For trial, this means the first thirty seconds of opening statement do not deliver your case; they configure the jury’s attention so that the next forty minutes can deliver it. If I open by reading the safety rule plaintiff is going to invoke, I have already lost. If I open with a question that primes the jury to look for what the defendant was thinking in real time, the rest of the opening lands on a different cognitive surface.

Don’t Think of an Elephant!—George Lakoff (2004) [Amazon]

Lakoff was writing about political communication, but the framing theory translates directly to trial. The most useful sentence I have read on opening statement: do not negate your opponent’s frame; replace it. Telling the jury “this case is not about safety rules” reinforces the safety-rule frame even as you deny it. The defense move is to give the jury a different frame entirely—not a rebuttal of plaintiff’s, a substitute. Lakoff’s harder lesson is that frames operate unconsciously. The defense lawyer who responds to plaintiff’s frame, even to argue against it, has already accepted it, and the jury hears the acceptance even if the lawyer thinks they are dissenting. The discipline this requires is identifying plaintiff’s frame before trial, building a structurally different frame, and committing to the replacement from voir dire through closing without ever speaking on plaintiff’s terms. The book is short and uneven; read the framing chapters and skip the political ones, which have aged badly. I owe most of what I have written on the difference between framing and rebutting to this book.

III. The Plaintiff Bar’s Playbook (Know Your Enemy)

Two books written by plaintiff lawyers for plaintiff lawyers. You cannot counter what you have not studied from the inside, and the defense bar has spent fifteen years pretending these methods do not exist. They do.

Reptile: The 2009 Manual of the Plaintiff’s Revolution—David Ball and Don Keenan [Amazon]

You cannot counter a methodology you have not understood from the inside. Ball and Keenan wrote the operating manual for the most successful plaintiff’s strategy of the last twenty years, and every defense lawyer who is not pretending owes them a careful read. What they codified is a method for translating any defense conduct into a Safety Rule violation in the jury’s mind. The architecture is simple: the rule is universal, the rule is about safety, the rule is absolute, and the defendant either followed it or did not. Once that architecture is in the jury’s head, the defense fight has been moved to ground the defense cannot win. The book is openly taught at plaintiff seminars and openly studied at plaintiff firms; it is barely read by the defense bar, which still treats it as something to dismiss. Dismissing it is how you lose. I have written four installments of my own series on how to defeat the Reptile, and none of them would have been possible if I had not read this one first. Don Keenan has continued publishing in the Edge series in the years since; I have not read those volumes yet, but they are on my list, and any defense lawyer who has read past the original should send me what they took from them.

Rules of the Road: A Plaintiff Lawyer’s Guide to Proving Liability—Rick Friedman and Patrick Malone [Amazon]

The companion volume to Reptile, and arguably more elegant in its method. Friedman and Malone showed plaintiff lawyers how to reduce complex defense conduct to a small number of simple rules that the defendant either followed or violated. The defense lesson is structural: once a “rule” has been established, the case is no longer about judgment. It is about compliance. Friedman and Malone’s method is depositions, depositions, depositions. The rules they want at trial are extracted from your witnesses through carefully sequenced questions designed to feel like reasonable inquiry until the rule is locked in. Most defense witnesses agree to these rules in the moment because the questions sound like common sense, until plaintiff’s counsel reads them back in closing as binding commitments. Reading this book taught me to fight rule-establishment in deposition, where it is built, rather than at trial, where it is too late. The defense work this requires is brutal: prepare every witness for the rule-establishing question pattern, identify the specific rules plaintiff will try to extract on the facts of this case, and rehearse the witness on how to engage the question without ratifying the rule.

IV. The Craft of Trial

Three books on the mechanics of trial work—the moment-to-moment craft that the cognitive foundation enables. The first is about who the jury is deciding to trust. The second is about how you actually move and stand in a courtroom. The third is the technical canon of cross-examination.

Win Your Case—Gerry Spence [Amazon]

Spence’s late-career synthesis of trial advocacy. I read it every couple of years. The lesson I keep coming back to: the jury is deciding whether to trust me, and they decide that long before they decide whether to trust my case. Spence taught me that the lawyer’s credibility is a thing you build, not a thing you possess, and that you build it through small choices, eye contact, candor, restraint, the willingness to concede what you should concede. Spence’s deeper argument is that the lawyer who wins is the lawyer who allows the jury to see them, not perform for them, see them. The book is dated in places and Spence’s bombast can grate, but underneath the prose is a real lesson: the jury is constantly evaluating whether you are showing them the lawyer or showing them a performance, and the moment they detect a performance, you have lost them. The discipline this requires, speaking simply, conceding what should be conceded, refusing to overstate even the points that genuinely favor your client, is the hardest part of trial work to learn, and the part most defense lawyers skip. I have probably stopped more bad questions at the moment of asking them because of Spence than because of any technical training I ever received.

Theater Tips and Strategies for Jury Trials—David Ball [Amazon]

The book that taught me trial is, in significant part, a theatrical performance. Not a fake one, but trial requires the same craft of pacing, presence, and audience attention that a stage performance requires. Ball gives concrete advice on things lawyers never get trained on: where to stand for different witnesses, when to move and when to hold position, when to speed up and when to slow down. Where you stand affects how the jury hears the question. The pace at which you ask affects whether the answer registers. Whether you move during a critical moment or hold position affects what the jury remembers. Most defense lawyers think this is theater-school territory and dismiss it as beneath them. The dismissive lawyers are losing performances they do not realize they are giving. After this book, I started writing physical blocking into my trial outlines the way directors write blocking into scripts. Most defense lawyers do this work by instinct. After Ball, you do it on purpose.

Cross-Examination: Science and Techniques—Larry Pozner and Roger Dodd [Amazon]

The technical bible of cross-examination, and the book I cited as the foundation for my own piece on agreement-based cross. Pozner and Dodd’s chapter method—short, declarative, leading questions in a sequence the witness cannot escape—is the most disciplined treatment of cross technique in print. Their three rules (use only leading questions, ask only questions you know the answer to, ensure every question advances your theory) are the floor. The book’s distinctive contribution is the chapter method itself—the discipline of organizing cross-examination into self-contained units, each driving toward a single, specific factual concession. Most cross-examinations sprawl because the lawyer is trying to do too many things at once. The chapter discipline forces clarity: each chapter has one point, the point is achieved through a sequence of leading questions the witness cannot escape, and the chapter ends before momentum can leak. I still rehearse cross-examinations chapter by chapter the way Pozner and Dodd describe, which is how I keep them tight at trial. Once you have internalized the chapter method, the question becomes how you structure the sequence to deliver a concession the witness does not see coming.

V. The Empirical Foundation

The hardest reading on this list, and the one book I would put on a serious defense lawyer’s shelf even if they only ever opened it once. The methodology has not been improved on in nearly sixty years.

The American Jury—Harry Kalven Jr. and Hans Zeisel (1966) [Amazon]

A 1966 monograph that has not been bettered as an empirical study of how juries actually work. Kalven and Zeisel surveyed hundreds of trials and found, among other things, that the jury’s pre-deliberation lean predicted the final verdict in roughly 80 percent of cases. Their methodology was simple and brutal: they had judges record their own verdict and then compared it to the jury’s verdict in 3,576 trials. The two agreed in roughly 78 percent of cases, and where they disagreed, the disagreement followed predictable patterns. The lesson: deliberation does not change minds at the rate we tell ourselves it does. By the time the jury walks into the deliberation room, the case is largely decided. What has stayed with me is not just that juries decide early, but that the structure of decision is set by what the jury has already filtered in voir dire and opening. The book has not been replicated at meaningful scale in the decades since, partly because the modern academy treats it as outdated and partly because the methodology is hard to fund. The substitute literature is thinner than it should be. Slow reading, dated language, but worth the work.

Where to Start

Eleven books. Two from cognitive science, three from persuasion theory, two from the plaintiff bar, three from the craft of trial, one from empirical jury research. If you read only one this year, make it Thinking, Fast and Slow. If you have already read that one, read the Cialdini pair next. If you have read Cialdini, read Lakoff. If you have read all of those, and most defense lawyers have not, you are operating at a level of craft preparation that puts you ahead of most of the trial bar.

If you would add a book to this list, send it to me with a sentence on what it changed about how you practice. The strongest follow-up posts I write are built from reader letters. If you would tell me a book I included does not deserve its place, send that too—I will read your argument and may revise the list.

I publish weekly on the psychology of trial work, AI in litigation, and the craft of building a serious practice. Subscribe if you have not.

It is also the version that loses cases.

I have tried around twenty-five cases and help with even more. The cross-examinations that changed outcomes were almost never the dramatic ones. They were the quiet ones. The ones where the opposing witness, question by question, agreed with my theory of the case. Not because I tricked them. Not because I bullied them. Because I asked questions they could not honestly deny, in a sequence that led to a conclusion they did not see coming until it was too late.

If you have followed the Psychology of Defense series, you already know the cognitive framework that makes this work. The same Kahneman research that explains why the Reptile Theory is so effective also explains why destruction-style cross so often backfires and why a quieter, agreement-based approach is more powerful than the theatrical alternative the profession has been romanticizing for decades.

The goal of cross-examination should not be to destroy the witness. It should be to get the witness to agree with you.

The Quiet Cross

Here is what nobody tells you in CLE: the best cross-examinations can seem boring to watch. There is no drama. The witness does not crumble. The lawyer does not raise their voice. To an untrained observer, nothing much is happening. But to the jury, something very specific is happening. The witness is building the lawyer’s case for them, one undeniable concession at a time, and the conclusion is arriving without anyone having to argue for it.

Understand why that is more effective than the theatrical version, and you understand the rest of this piece.

What the Jury Actually Sees

When a lawyer attacks a witness aggressively, the jury’s System 1 does not process the content of the questions. It processes the dynamic. System 1 sees a powerful person attacking a less powerful person. It sees aggression. It sees someone being hurt. And it generates an immediate, instinctive response: sympathy for the person under attack.

This is the same protective instinct that drives the Reptile Theory. In a Reptile case, plaintiff’s counsel directs that instinct against the defendant. In a destruction-style cross, the defense lawyer directs it against themselves. You are creating a Reptile dynamic in reverse, and the person the jury wants to protect is the witness you are trying to discredit.

It does not matter that the witness is lying. It does not matter that the facts support your questions. System 1 does not evaluate substance. It evaluates dynamics. And the dynamic of an aggressive cross-examination favors the witness every time.

Even when destruction-style cross works factually, when you actually catch the witness in a lie, the jury may not process it the way you intend. The moment may register as cruelty rather than truth. The juror remembers how the witness looked, how they felt watching it happen, more than they remember the specific concession you extracted. You won the exchange and lost the jury.

Why Agreement Works Differently

Agreement-based cross produces the opposite System 1 response. When a witness agrees with the cross-examining lawyer, the jury does not have to choose sides. There is no conflict to process. There is simply a fact that both sides apparently accept.

This is powerful because of how System 1 handles consensus. When two people who should disagree actually agree, the brain treats that information as highly credible. The agreement carries more weight precisely because it comes from an adverse source. A defense expert telling the jury the defendant’s conduct was reasonable is expected. The plaintiff’s own witness conceding a fact that supports the defense theory is unexpected, and System 1 gives unexpected consensus more cognitive weight than any amount of argued evidence.

Larry Pozner and Roger Dodd formalized the mechanics in Cross-Examination: Science and Techniques. Their three rules, use only leading questions, ask only questions you know the answer to, and ensure every question advances your theory, remain foundational. What I want to add here is the why: the cognitive science that explains why their method is so much more effective than the alternative.

What Agreement-Based Cross Actually Looks Like

The following is drawn from a case I tried; details have been changed to protect client confidentiality.

A products liability case. A two-vehicle collision on a two-lane interstate. The central dispute was where on the roadway the vehicles made contact, because the point of contact determined fault. Plaintiff’s accident reconstruction expert had produced a report with a clear opinion: the vehicles overlapped by a specific amount at a specific location. His report included the overlap measurement, the width of both vehicles, and the starting position for one of them. A confident opinion backed by a detailed report.

A destruction-style cross would have started with his methodology, his credentials, his fee. He would have handled each one comfortably. He had testified dozens of times. He had answers prepared. The jury would have watched a defense lawyer try and fail to shake a confident witness, and they would have left the exchange trusting him more than before.

Here is what I actually did.

I started with his numbers. Not mine. His. I asked him to confirm the overlap measurement from his report. He agreed. The width of Vehicle A. He agreed. The width of Vehicle B. He agreed. The starting position he had identified for Vehicle A. He agreed. Each answer was calm, easy, taken directly from his own work. There was nothing to argue with.

Then I started doing the arithmetic with him. If Vehicle A started at this position, and the overlap was this measurement, then Vehicle B would need to be located at this position. He agreed, because that was what his own analysis required. I asked him to confirm the actual, measured width of Vehicle B. Seven feet. He agreed. Then I asked him to walk the jury through the math: given the starting position, given the overlap, how wide would Vehicle B need to be for his point-of-contact opinion to work? The answer, derived entirely from his own numbers, was ten feet.

His opinion required a seven-foot-wide vehicle to occupy ten feet of roadway.

I did not raise my voice. I did not call him incompetent. I asked him to explain how a seven-foot vehicle could be ten feet wide. He tried. He offered alternative explanations. He attempted to reframe the measurements. None of it worked, because the math was the math, and every number had come from his own report. The jury watched a credentialed expert struggle to explain away a contradiction they could verify with a pencil and a piece of paper.

The jury returned a defense verdict.

I never attacked him. I never questioned his integrity. The jury did not have to choose between believing me or believing the expert. His own numbers made the choice for them.

When the Witness Cannot Be Cornered by Math

A fair objection at this point: that example works because the contradiction is mathematical. The numbers do not care how confident the witness is. What about a witness who is not constrained by arithmetic, a standard-of-care expert in a med-mal case, a sympathetic plaintiff, a polished corporate representative?

The technique still works. The raw material just shifts.

For an expert whose opinion is judgment-based rather than measurement-based, the agreement target is not the conclusion. It is the methodology. Ask the expert to agree with each step in the reasoning their discipline requires. Get them to commit to the inputs that discipline demands. Then walk through which of those inputs are absent from the record they reviewed. They cannot deny what their field requires without losing professional credibility. They cannot deny what the record shows without contradicting documents already in evidence. Either way, the conclusion the jury draws is that the opinion was reached without the inputs the field itself demands.

For a sympathetic plaintiff, you almost never want extended cross. But the same principle controls the cross you do conduct. You target specific factual concessions, the date of a treatment, the contents of a record, an admission already made in deposition, and you stop. You do not argue. You do not press. You confirm three or four facts that matter and you sit down. The jury reads restraint as confidence.

For a corporate representative, the agreement target is usually the company’s own documents, policies, training materials, safety standards. The witness will not contradict the company’s own paper. Build the concession out of the paper, not out of the witness.

The pattern is the same in every case: identify the factual material the witness cannot deny without damaging themselves, and build the conclusion out of that material.

When Destruction Is the Right Move

The honest version of this piece has to acknowledge that destruction-style cross occasionally has a place. There are three circumstances where I will use it.

The first is when a witness has been impeached on a prior inconsistent statement on a dispositive point. At that moment the witness has handed you their credibility, and you have an obligation to your client to use it. You still do not do it theatrically. You do it surgically. Read the prior statement. Ask the witness to confirm it. Read the current testimony. Ask the witness to confirm it. Do not editorialize. Let the contradiction sit.

The second is when the witness is a paid expert whose entire history is built on testifying for one side. The relevant facts, the number of times testified, the percentage of plaintiff versus defense work, prior contradictory opinions in other cases, are themselves agreement-based. You are not destroying the witness. You are asking them to agree with their own résumé.

The third is when credibility is the entire case. In some fraud cases, in some matters turning on a single witness’s account, the jury cannot decide unless they decide whether to believe one person. In those cases the cross has to engage the witness’s account directly. Even then, the most effective version stays calm. The witness should hang themselves on inconsistencies in their own story, not on the lawyer’s aggression.

The common thread: even when the work of cross is confrontational, the manner should not be. The jury still reads the dynamic. The lawyer who needs to expose a paid expert is more effective doing it quietly than loudly.

How to Build an Agreement-Based Cross

The accident reconstruction cross was not spontaneous. It was the product of a preparation method I use in every case where I intend to cross an expert or a key fact witness. Four steps.

One: identify the conclusion you need the jury to reach. Not a vague goal like “undermine the expert’s credibility.” A specific factual conclusion, stated in one sentence. If you cannot state it before you begin preparing, you are not ready to cross.

Two: work backward from the conclusion to the individual facts that compel it. Each fact must be independently true and independently undeniable. The strongest agreement-based cross uses the witness’s own documents, own testimony, and own data as the raw material. You are not arguing with the witness. You are holding the witness to what they have already said.

Three: test every fact for deniability. For each fact in your sequence, ask: can the witness deny this without looking dishonest or incompetent? If the answer is yes, the fact is not strong enough. Remove it or replace it. A safe question is one where the witness’s agreement advances your case and the witness’s denial damages the witness’s credibility. When every question in your sequence meets that standard, you cannot lose the exchange.

Four: sequence the facts so the conclusion emerges at the end, not the beginning. Most lawyers want to lead with the punchline. That gives the witness the chance to explain, reframe, and prepare the jury before they have seen the predicate facts. The agreement-based approach inverts that. You establish every predicate first, calmly, so that by the time the contradiction arrives, the jury has already accepted every element that makes it undeniable.

The Discipline It Requires

Agreement-based cross is harder than destruction-based cross. It requires more preparation. It requires the discipline to stop the moment you have the concession you need. No not ask the witness to draw the conclusion for you; that is what closing argument is for. And it requires suppressing your ego.

The destruction approach feels good. It feels like winning. You get the rush of confrontation, the satisfaction of exposing a lie, the courtroom drama that makes you feel like a trial lawyer. The agreement approach does not feel like anything. It is quiet. It is methodical. It is invisible to anyone who does not understand what is happening. But when you sit down, the witness has built your case for you, and the jury does not even realize it happened.

That is the point.

The Connection to Everything Else

If you have been reading this Substack, you will recognize the through-line. The Psychology of Defense series argued that defense lawyers lose when they fight on plaintiff’s psychological terrain and win when they build their own. The same principle controls cross-examination.

Destruction-style cross is fighting on the witness’s terrain. You are engaging in conflict, and conflict activates the jury’s protective instincts. Agreement-based cross is building your own terrain. You are constructing a narrative out of facts the witness cannot deny, and the jury absorbs that narrative without resistance because there is no conflict to trigger a defensive response.

The best cross-examinations I have conducted did not feel like cross-examinations to the jury. They felt like conversations in which the witness happened to confirm everything I needed. That is a technique, not an accident. It is grounded in the same cognitive science that governs every other aspect of trial advocacy: the brain that decides the case is not the rational, analytical brain the defense has traditionally relied on. It is the fast, intuitive brain that processes dynamics, narratives, and moral judgments before the evidence is even fully presented.

Work with that brain instead of against it, and cross-examination becomes the most powerful tool in the defense lawyer’s arsenal. Not because you destroyed the witness. Because the witness agreed with you.

If you found this useful, do one thing with it: pick a cross you have coming up, and try the four-step preparation method. Then tell me how it went. Reader cases make the strongest follow-up posts I can write.

I publish weekly on the psychology of trial work, AI in litigation, and the craft of building a law practice. Subscribe if you have not.

If you have followed this series, you understand three things by now. Jurors decide cases through fast, intuitive System 1 processing, not through the slow analytical reasoning the defense relies on. The Reptile Theory is the most effective System 1 exploitation method in modern litigation. And the deposition is where the Reptile builds its case, one carefully constructed question at a time.

But depositions are preparation. Trial is performance. And the trial is where the Reptile either takes the case or gets beaten.

Here is the uncomfortable truth: most defense lawyers who understand the Reptile intellectually still lose to it at trial. They lose because they fight it on plaintiff’s terms. They spend their trial responding to plaintiff’s frames, explaining why the Safety Rule was not violated, why the conduct was reasonable, and why damages should be limited. Every one of those arguments accepts the Reptile’s premise that the case is about a rule and whether it was broken.

The defense that beats the Reptile does not respond to plaintiff’s psychology. It builds its own.

The Four Pillars of Defense Psychology

In Part 2, I introduced the four elements of an effective counter-narrative. Now I am going to show you how to execute each one, not as abstract principles, but as specific tactical decisions in voir dire, opening, witness examination, and closing.

Pillar 1: Restore Complexity. The Reptile works by simplifying professional decisions into binary pass/fail judgments. The defense must show the jury what the decision actually looked like in real time: the incomplete information, the competing considerations, the time pressure, the judgment that experience and training demanded. The Safety Rule framework says there was a rule and it was broken. The defense framework says there was a decision, and it was made by a trained professional weighing factors that cannot be reduced to a checklist.

Pillar 2: Distinguish Mistake from Wrong. The Reptile collapses the moral distinction between a professional who tried to do the right thing and erred and one who did not care. The defense must restore that distinction and make the jury feel it. A surgeon who makes an error in a difficult procedure is not morally equivalent to a surgeon who does not bother checking. A trucking company whose compliance system missed a signal is not morally equivalent to one that has no compliance system. The difference matters, and the defense must make the jury believe it matters.

Pillar 3: Establish Genuine Safety Commitment. The Reptile narrative depends on painting the defendant as indifferent to safety. The defense must affirmatively establish, through evidence that was true before the lawsuit existed, that the defendant invested in safety, trained on safety, built systems for safety, and cared about the outcome. This is not a rebuttal. It is a narrative that stands on its own.

Pillar 4: Name the Injustice of Hindsight. Plaintiff’s counsel is asking the jury to judge a real-time decision with the benefit of information that only became available later. That is not the applicable legal standard, and more importantly, it violates the jury’s own sense of fairness. Every juror has made a decision under uncertainty that turned out badly. The defense must make the jury feel the injustice of hindsight judgment, not as an excuse, but as a principle they would want applied to their own decisions.

In Kahneman’s terms, each pillar is designed to activate a competing System 1 response. Pillar 1 activates the intuition that “real decisions are complicated.” Pillar 2 activates the moral intuition that “trying and failing is not the same as not caring.” Pillar 3 activates the narrative of “this is a responsible actor.” Pillar 4 activates the fairness intuition that “you cannot judge someone by what they could not have known.” Together, they give the jury a System 1 framework that competes directly with the Reptile.

These four pillars must be consistent across every phase of trial. They begin in voir dire and they do not stop until the last word of closing argument.

Voir Dire: Start Before They Do

Plaintiff’s counsel will use voir dire to begin executing the Reptile strategy, priming jurors with community safety concerns, identifying jurors who are most susceptible to fear-driven decision-making, and seating a jury predisposed to punish. If you are passive in voir dire, you have already lost the first battle.

Identify the dangerous juror profile. Experienced trial psychologists call them “injustice collectors,” jurors with a history of perceived mistreatment by institutions who carry unresolved anger about those experiences. For the injustice collector, the Reptile case is an opportunity for vicarious revenge. They will decide the case before the evidence is in and hold firm in deliberations. Surface this profile with questions like: “Have you or a close family member had an experience with a large company or institution where you felt treated unfairly?” and “If you felt strongly about a verdict amount, would you be comfortable holding that position even if other jurors disagreed?” The answers tell you who your highest-priority strikes are.

Plant your counter-narrative seeds. Voir dire is not too early to begin establishing the defense psychology. Ask questions that create cognitive frameworks: “Would you agree that it is important to hear all the evidence before deciding what happened?” “Would you agree that a professional making a decision in a difficult situation in real time is not the same as someone judging that decision years later with full information?” “Can you distinguish between a mistake and a deliberate wrong?” These questions do not argue the case. They establish the mental architecture within which jurors will later process your evidence.

Watch for the Reptile priming. Pay attention to plaintiff’s voir dire. If counsel is asking about community safety, about whether jurors believe companies should be held accountable for rule violations, about whether jurors would be willing to return a large verdict to “send a message,” the Reptile is already running. Note these moments. You may need to address them directly in your own voir dire or in opening.

Opening Statement: Reframe, Do Not Rebut

The worst opening statement in a Reptile case is one that responds to plaintiff’s frames on plaintiff’s terms. “Ladies and gentlemen, the plaintiff told you Dr. Smith violated a safety rule. We disagree.” You have just conceded that the case is about a safety rule and invited the jury to decide who is right about it. You lose that argument because the Reptile has already primed them.

Instead, establish your own frame before addressing plaintiff’s.

Lead with the decision, not the rule. “This case is about what actually happened in that operating room on December 14th, not about a rule that has been simplified until it no longer resembles medicine. Dr. Smith made a decision based on thirty years of experience and the specific findings he observed. The question you will answer is whether that decision, judged by what he knew in that moment, not by what we learned afterward, was within the standard that medical professionals apply every day.”

Humanize through decision-making, not biography. Do not tell the jury your client is a good person. Show them how your client thinks. “Dr. Smith has performed over 2,000 spinal surgeries. He will tell you, in his own words, what he saw, what he decided, and why. And when you hear him, you will understand why what plaintiff’s counsel is calling a ‘violation’ was a clinical decision made in good faith with his patient’s welfare as his only concern.” This humanization is far more powerful than character evidence because it addresses the actual basis for liability.

Do not wait. In jurisdictions that permit defense opening immediately after plaintiff’s opening, and most do, give your full opening immediately. In System 1 terms, the first frame anchors all subsequent processing. The Reptile framing must not sit with the jury unopposed through plaintiff’s case-in-chief. If you reserve opening, you are letting the Reptile establish itself unchallenged for days. Every hour that passes without a counter-narrative makes the Reptile harder to dislodge.

Witness Examination: Make the Pillars Visible

Your witnesses, the defendant, your experts, your fact witnesses, are the vehicles for the four pillars. Every direct examination should advance at least one pillar, and ideally more than one.

The defendant’s testimony must embody Pillar 1 (Complexity) and Pillar 2 (Mistake vs. Wrong). Walk the defendant through the decision in real time. What information did they have? What were the options? What were the competing considerations? Why did they choose what they chose? This is not an apology. It is a demonstration of professional engagement with a difficult problem. The jury should finish the defendant’s testimony understanding that this was a thinking, caring professional who faced a hard call and made it with their patient’s or customer’s welfare as their primary concern.

Your experts must dismantle the Safety Rule without denying safety. An expert who says “there is no absolute rule” sounds evasive. An expert who says “the standard requires X, and here is the specific context in which the defendant’s decision met that standard” is credible. Your experts should walk the jury through what the professional standard actually says, not the simplified version plaintiff’s counsel presented, but the real standard that practicing professionals apply. The nuance is the defense.

Cross-examination of plaintiff’s witnesses should target the rule, not the plaintiff. Aggressive cross-examination of a sympathetic plaintiff activates exactly the protective instincts the Reptile has engaged. Cross the plaintiff only on genuinely dispositive issues where you have hard evidence. Instead, direct your cross-examination energy at plaintiff’s experts. Depose them extensively on the limits of their “absolute rules,” and use their own concessions at trial to show the jury that even plaintiff’s experts agree the standard requires judgment.

Consider naming the Reptile. In cases with sophisticated jurors, it can be effective to identify what plaintiff’s counsel is doing. “Doctor, isn’t it true that the way this examination has been conducted, asking whether each safety principle is ‘absolute’ with ‘no exceptions,’ is a litigation technique designed to simplify complex medical standards?” Done carefully, this creates a moment of meta-awareness that may cause some jurors to examine their own emotional responses with more skepticism. Done poorly, it sounds paranoid. Reserve this tactic for cases where you have a jury that will respond to intellectual candor.

Closing Argument: Return to Your Frame

Closing argument in a Reptile case requires discipline above everything else. The temptation is to spend the entire closing responding to plaintiff’s arguments. Resist it. You are fighting on their territory with their frames.

Start with your story, not theirs. “When you go into that deliberation room, you are going to be asked to decide what actually happened. Not a simplified version designed to produce a particular outcome. Not a story where complex decisions become obvious in hindsight. What actually happened, evaluated against the standard that actual professionals actually apply. That is what our legal system requires of you.”

Address the community safety argument directly. Do not ignore it. If you leave it unchallenged, it will dominate the jury room. But address it on your terms: “Plaintiff’s counsel told you this verdict is about protecting the community. I agree the community deserves protection. But the community is also served by a medical system in which doctors can make difficult decisions without fear that a single outcome, judged years later with full information, will destroy their careers. The community is served by professionals who exercise judgment. A verdict that punishes judgment will not make anyone safer. It will make careful, experienced professionals afraid to act.”

That argument, that punishing professional judgment makes the community less safe, is the single most powerful counter-Reptile argument available to the defense. It takes the community safety frame and turns it back on plaintiff. The jury is no longer choosing between protecting the community and protecting the defendant. They are being asked to consider which verdict actually protects the community: one that punishes a professional for making a difficult call, or one that preserves the system of professional judgment that everyone in the community depends on.

Reestablish the compensatory framework on damages. “The purpose of civil damages is to compensate this plaintiff for her actual losses, not to punish, not to send a message, not to change an industry. If you find liability, the question is: what does she need to be made as whole as she can be? That is the question the law asks. Not: what number would change this defendant’s behavior.”

The Mindset Behind All of It

This series began with Kahneman’s research proving that human beings are not the rational decision-makers we imagine ourselves to be. It ends with the practical implications of that finding for trial lawyers.

The defense lawyer who beats the Reptile is not the one who best understands negligence law. It is the one who best understands people: why they make decisions, what they are afraid of, what they need to hear to render a just verdict rather than a fearful one.

Your task is to make the jury safe again. Not by denying that the world contains risk, but by showing them the world contains judgment. Professionals who exercise judgment are not the community’s enemies. They are its servants. A system that punishes judgment destroys the capacity for judgment, and that harms everyone.

Frame every case around that truth. Prepare every witness to embody it. Build every exhibit around it. Argue it from voir dire through closing. That is how you defend a Reptile case: not by fighting the Reptile on its own ground, but by giving the jury a different, truer, and more compelling story about what happened and what justice requires.

This concludes the four-part Psychology of Defense series. If you found it useful, I would ask two things: subscribe if you have not, and share the series with a defense lawyer who needs it. I publish weekly on the psychology of trial work, AI in litigation, and the craft of building a serious practice.

If the Reptile Theory has a hunting ground, it is the deposition.

Not the courtroom. Not the opening statement. The deposition. It is the one phase where plaintiff’s counsel has uninterrupted, one-on-one access to your witnesses, before the jury is seated, before the rules of evidence apply with full force, and before the witness has the benefit of the narrative you will construct at trial. It is where Safety Rules are established, where admissions are extracted, and where the building blocks of plaintiff’s trial presentation are assembled one question at a time.

And it is where most defense lawyers lose the case without realizing it.

I have defended against Reptile depositions across products liability, medical malpractice, and commercial cases. From those, I have learned that the problem is not that defense witnesses give bad answers. The problem is that they are unprepared for what is actually happening. They are prepared for a normal deposition. They are not prepared for a psychologically engineered interrogation designed to construct a Safety Rule out of their own words.

Today I am going to walk you through the question patterns Reptile practitioners use and show you exactly how to defend against them. Not with evasion. Not with coached non-answers. With responses that are honest, that acknowledge the importance of safety, and that prevent the Safety Rule from being built on the record.

How a Reptile Deposition Is Structured

Before the specific patterns, you need to understand the architecture. A Reptile deposition is not a random search for helpful facts. It is a four-stage construction project.

Stage One, Rule Construction. Plaintiff’s counsel establishes abstract principles of safety so broad and obviously correct that any disagreement would seem bizarre. The witness is conditioned to agree.

Stage Two, Rule Absolutism. Counsel narrows those principles into absolute rules: “always,” “never,” “no exceptions.” This is where the Safety Rule is built.

Stage Three, Knowledge Attribution. Counsel establishes that the defendant knew the rule and knew the consequences of violating it. This transforms negligence into something that sounds like recklessness.

Stage Four, Violation and Harm. Counsel connects the rule to the specific facts and the plaintiff’s injury. By this point, every element of the Safety Rule has been conceded through the witness’s own testimony.

In the System 1 framework from Parts 1 and 2: Stages One and Two build the binary frame. Stage Three activates moral outrage. Stage Four locks the narrative. The defense cannot wait until Stage Four to respond. By then, the trap is closed. You intervene at Stages One and Two, or you lose.

The Question Patterns

I have cataloged ten recurring patterns that Reptile practitioners deploy in depositions. I will walk through the six most dangerous in detail, with the psychological mechanism behind each, examples across case types, and model witness responses. The remaining four I will describe more briefly.

Pattern 1: The Abstract Safety Agreement

What is happening: Counsel opens with questions so broad that disagreement would sound insane. “Is patient safety important to you?” “Should trucking companies ensure their drivers are not fatigued?” The witness says yes, because the answer is yes. But each agreement makes the next, slightly more specific question easier to accept. By the time the questions reach the actual Safety Rule, the witness has been trained to say yes and feels social pressure to continue.

What it sounds like: “Doctor, would you agree that patient safety is the most important consideration in any medical procedure?” “Yes.” “And that a surgeon has a duty to take every reasonable precaution to protect a patient?” “Yes.” “And that confirming the correct spinal level is one of those precautions?” “Yes.” “And there is never a good reason to proceed without confirming, correct?”

The defense: The witness can and should agree with the first few abstract questions. They are true. But the witness must be trained to recognize when the questions transition from genuinely abstract to tactically specific. The critical moment is the final, absolute question. It must not receive a simple yes.

Model response: “Confirming the operative level is standard practice and I follow it. But how that confirmation is accomplished, whether through fluoroscopy, imaging, anatomical landmarks, or a combination, is a clinical judgment based on the specific patient, the specific anatomy, and the specific operative findings. In this case, I used [specific method] because [specific reason]. That was appropriate and medically sound.”

This response does three things: it acknowledges the safety principle, which prevents the witness from sounding callous. It introduces context, which prevents the absolute framing from taking hold. And it pivots to the specific facts, which forces plaintiff’s counsel to engage with the evidence rather than the abstract rule.

Pattern 2: The Absolute Rule Demand

What is happening: Having established the general principle, counsel now demands that the witness confirm it is absolute: “always,” “never,” “no exceptions.” This is the core Safety Rule construction question. The goal is an unqualified commitment that can later be juxtaposed with the defendant’s actual conduct.

What it sounds like: “There is never, never, a situation where it is acceptable for a commercial truck driver to operate a vehicle after falsifying hours-of-service logs. Never. Correct?”

The defense: When a standard is genuinely absolute, such as a legal reporting requirement, the witness can agree. When professional judgment is involved, the witness must resist the absolutism without sounding like someone who minimizes safety.

Model response: “A driver falsifying logs is a serious violation. But I need to be precise about what happened here. The evidence shows that the logs in question contained a recording error, not intentional falsification, and the distinction matters. Intentional falsification of hours-of-service logs is never acceptable. An inadvertent recording error is a different matter governed by a different standard, and that is what the evidence shows occurred here.”

Pattern 3: The Knowledge-of-Danger Attribution

What is happening: Having established the Safety Rule, counsel now establishes that the defendant knew the rule and knew the consequences of violating it. In System 1 terms, this transforms the narrative from “the defendant made a mistake” to “the defendant knew and chose to do it anyway.” That is the difference between sympathy and outrage.

What it sounds like: “You were trained in medical school that operating at the wrong spinal level can cause permanent paralysis?” “You understood, the morning of this surgery, that the consequence could be that your patient would never walk again?” “And you knew the verification protocol existed specifically to prevent that outcome?”

The defense: Witnesses generally must agree with the general danger awareness questions. Denying knowledge of known risks is dishonest and destroys credibility. The key is the third question, where the witness must acknowledge the protocol while making clear it was followed.

Model response: “Yes, I know that protocol, I trained on it, and I followed it. What I think you are implying, that I skipped the verification step, is not what happened. I will tell you exactly what I did to confirm the operative level in this case, and you can evaluate whether it met the standard. I believe it did.”

Pattern 4: The Community Endangerment Broadening

What is happening: Counsel expands the risk beyond the plaintiff to establish that the defendant’s conduct endangered everyone. In a products case it is every purchaser. In a trucking case it is every driver on the highway. This is the “community” element that converts jurors from neutral arbiters into self-interested guardians.

What it sounds like: “Your company manufactured and sold 80,000 of these car seats, correct? And every one has the same latch mechanism? So the same latch that failed in this case is in 80,000 car seats currently in use in the United States?”

The defense: Answer the factual predicate (yes, we sold 80,000 units) while rejecting the implied conclusion that they are all dangerous.

Model response: “We manufactured and sold approximately 80,000 units, yes. Each unit has the same latch design. But I cannot agree with the implication that the latch is defective in any of those seats. Our testing showed the latch meets all applicable safety standards under the crash conditions it is designed to address. I would not characterize our product as dangerous to 80,000 families based on one accident without knowing what the evidence shows about why this failure occurred.”

Pattern 5: The Policy-as-Admission Trap

What is happening: Counsel uses the defendant’s own policies and training materials as the source of the Safety Rule, then uses any failure to follow them as a self-inflicted admission. “Your company wrote this rule. Your company trained on it. And you violated it.”

What it sounds like: “I am showing you Exhibit 14, your company’s Driver Safety Manual. It says, and I quote, ‘Drivers must not operate a commercial vehicle when impaired by fatigue.’ That is your company’s own rule, correct? And this driver was fatigued that morning, correct? So by your own manual, he should not have been driving?”

The defense: Acknowledge the policy without qualification. Then directly contest the factual premise. Was the driver actually fatigued? Do not let plaintiff’s counsel establish a contested fact as a given.

Model response: “That is our policy, and I stand behind it completely. The question of whether the driver was impaired by fatigue is what I disagree with. His hours-of-service records showed he had [X] hours of rest. His pre-trip inspection was completed. There was nothing in the information available to dispatch that morning indicating impairment. I cannot accept the premise because the evidence does not support it, and I would like to explain what the evidence actually shows.”

Pattern 6: The Hindsight Remorse Question

What is happening: Near the end of the deposition, counsel asks whether, knowing the outcome now, the witness wishes the defendant had done something differently. This exploits genuine human compassion. Agreeing is a near-admission. Disagreeing sounds callous. Both paths damage the defense.

What it sounds like: “Looking back now, don’t you wish someone had caught the issue with that latch before 50,000 units shipped?”

The defense: Reject the hindsight frame explicitly. Express genuine compassion without conceding liability.

Model response: “I am deeply sorry about what happened to this family. That is a tragedy. But I have to answer honestly, and honestly means answering based on what we knew at the time, not what we know now. Based on the records we had, the decision was sound. I can explain exactly what those records showed. If you are asking whether I wish the outcome had been different, of course I do. But wishing the outcome had been different is not the same as saying the decision was wrong.”

The Remaining Four Patterns

Pattern 7: The Rule-to-Violation Pivot. After building the abstract rule, counsel rapidly applies it to the specific facts: “You said the rule is X. The evidence shows you did Y. So you violated the rule.” Defense: the witness must break the chain by clarifying that the earlier abstract answer was more nuanced than recorded, or by contesting the factual predicate.

Pattern 8: The “Wouldn’t You Agree” Endorsement. Counsel asks the witness to endorse plaintiff’s framing in the first person: “Wouldn’t you agree that a responsible company would have...?” Defense: accept the general proposition and immediately apply it to what actually happened. “Yes, and we did exactly that.”

Pattern 9: The Document Ambush. Counsel produces an internal document the witness has not seen, a memo, an email, a complaint log, and asks the witness to explain it in the most damaging framing. Defense: this pattern is defeated entirely by preparation. Every potentially damaging document must be reviewed with the witness before the deposition. A witness who is surprised by a damaging document will improvise. Improvised answers in Reptile depositions are almost always bad answers.

Pattern 10: The Profit Motive Insinuation. Counsel establishes that the decision leading to injury also saved the defendant money, implying profit over safety. In System 1 terms, this activates the “greed” frame discussed in Part 1. Defense: never accept the premise that cost was the reason for a safety-related decision. Reframe: the decision was made on the merits, and the merits showed the product was safe or the conduct was appropriate.

The Rules That Govern Every Answer

Beyond the specific patterns, there are principles that should govern every answer your witness gives in a Reptile deposition.

Answer the question asked, not the question you wish they had asked. Volunteering information provides plaintiff with additional material and signals nervousness.

You are allowed to explain. If a question cannot be answered accurately with a yes or no, say so: “I cannot answer that with a yes or no. Let me explain.” Counsel’s insistence on a binary answer does not legally require one if the accurate answer is more nuanced.

Persistence is not proof you are wrong. If plaintiff’s counsel asks the same question six times, it is because they did not get the answer they wanted. That means your answer was good. Do not change a correct answer because of social pressure.

Emotion is appropriate; concession is not. Witnesses can and should express genuine compassion without conceding liability. “What happened to this family is tragic” is not the same as “we caused it.”

Know your documents before opposing counsel shows them to you. The document ambush is one of the most effective Reptile weapons. No witness should encounter a damaging document for the first time under oath.

The deposition is not the place to win the case. It is the place to ensure the case can be won at trial. A witness who survives a Reptile deposition without providing the foundational admissions plaintiff’s counsel is seeking has accomplished something critical: they have denied plaintiff the ability to construct the Safety Rule narrative from defense testimony. That is the goal. Not brilliance. Protection.

In Part 4, I will cover the counter-narrative: the four-pillar defense framework from voir dire through closing argument, and how to give the jury a psychological alternative to the Reptile that is equally compelling and more true. Subscribe if you have not.

Every defense lawyer in America has heard of the Reptile Theory. Most of them think they understand it. Almost none of them do.

Here is the proof: in the years since David Ball and Don Keenan published Reptile: The 2009 Manual of the Plaintiff’s Revolution, the theory has been used to produce hundreds of millions of dollars in verdicts that would have been unthinkable a generation ago. Medical malpractice verdicts exceeding fifty million dollars. Trucking verdicts north of thirty million. Products liability awards that dwarf the underlying economic damages. These are not anomalies. They are the predictable result of a method that the defense bar has been consistently failing to counter, not because the method is unbeatable, but because the defense bar’s response is built on a fundamental misunderstanding of what they are fighting.

In Part 1 of this series, I laid out Kahneman’s research on how jurors actually process information: through fast, intuitive System 1 thinking rather than the slow, analytical System 2 thinking that defense strategies are designed to engage. The Reptile Theory is the most sophisticated application of that insight in modern trial practice. It is not a litigation trick. It is a System 1 operating system for the courtroom.

How the Reptile Operationalizes System 1

The Reptile method does not ask jurors to sympathize with the plaintiff. Sympathy is an emotional response that activates what Kahneman would call a System 1 judgment, but it is a weak one. Sympathy can be countered with evidence that the plaintiff is less sympathetic than portrayed. The Reptile goes deeper. It asks jurors to protect themselves.

The theory converts the juror from a neutral arbiter into what Ball and Keenan call a “guardian,” a protector of community safety. A guardian does not need to like the plaintiff. A guardian does not even need to believe the plaintiff is a good person. A guardian only needs to believe three things: that the defendant’s conduct was dangerous, that the defendant knew it was dangerous, and that without a strong verdict, the defendant will endanger the community again.

The psychological engine driving this is self-interest, the most powerful System 1 motivator there is. A juror deciding whether to compensate a stranger for injuries that will never affect the juror personally is capable of detachment. A juror deciding whether to protect herself, her family, and her neighbors from a danger that could harm any of them is incapable of detachment. Once self-interest activates, System 2 analysis becomes nearly impossible. The juror is no longer weighing evidence. The juror is responding to a perceived threat.

This is why the damages in Reptile cases are so disproportionate to the plaintiff’s actual losses. The jury is not compensating. It is deterring. It is asking: what number will make this company follow its own rules? That question has no ceiling rooted in compensatory calculation, which is why traditional remittitur arguments so often fail.

The Safety Rule: System 1’s Binary Switch

The mechanism by which plaintiff’s counsel activates the Reptile is the “Safety Rule,” an absolute standard of conduct that the defendant allegedly violated and that, if followed, would have prevented the harm.

A Safety Rule has three characteristics. It is framed in absolute terms: no exceptions, no qualifications. It is tied to protecting people from serious harm. And it is something the defendant cannot credibly deny without sounding like someone who believes there are circumstances in which it is acceptable to endanger people.

This is where Kahneman’s “what you see is all there is” principle becomes a weapon. Once a Safety Rule is established (“a surgeon must always confirm the operative level before proceeding,” “a trucking company must never allow a fatigued driver to operate a commercial vehicle,” “a manufacturer must never sell a product it knows is unsafe”), the entire case organizes around a binary that System 1 processes effortlessly: the defendant either followed the rule or broke it. If they broke it, the community is in danger. If the community is in danger, the jury must stop it.

Every piece of nuance, context, and complexity that the defense tries to introduce is System 2 information. And as we established in Part 1, System 2 will not engage when System 1 has already provided a coherent, satisfying narrative. The Safety Rule gives System 1 everything it needs. Nuance becomes noise.

Why Motions in Limine Cannot Stop This

Now you can see why the standard defense responses fail.

A motion to exclude “safety rule” questioning assumes that the Reptile is a discrete tactic that can be identified and blocked. But a skilled plaintiff’s lawyer does not need to use the phrase “safety rule.” They do not need to ask the magic words. They can establish the identical psychological framework through standard liability questioning that no judge will exclude. “Doctor, is patient safety important to you?” “Is confirming the operative level part of the standard of care?” “Is there a good reason not to confirm?” These are legitimate, foundational liability questions. They also construct the Safety Rule step by step. No motion in limine will keep them out, because they are not objectionable. They are just effective.

The same problem undermines the standard witness preparation approach. Telling a witness “don’t agree to absolute rules” is a surface-level fix for a structural problem. A witness who refuses to agree that patient safety is paramount sounds callous. A witness who hedges on whether there is ever a good reason to skip a verification protocol sounds evasive. And a witness who offers the nuance that “it depends on the clinical circumstances” has just told the jury that sometimes, in this defendant’s view, it is acceptable to take shortcuts with people’s safety.

The witness preparation problem is not about coaching better answers. It is about the fact that the Reptile creates a lose-lose dynamic that cannot be solved at the level of individual questions. Either the witness agrees with the Safety Rule and concedes the violation, or the witness pushes back on the Safety Rule and sounds like someone who does not care about safety. The binary is baked into the psychology, not the question.

What Actually Works: Fighting System 1 with System 1

If the Reptile cannot be defeated by a motion or a coached witness answer, what does work?

You cannot fight psychology with procedure. You have to fight psychology with psychology. As I argued in Part 1, the defense must compete at the System 1 level. Against the Reptile specifically, that means giving the jury a counter-frame that is equally intuitive, equally grounded in moral instinct, and equally coherent.

That counter-frame has four elements.

First, it restores the complexity that the Reptile stripped away. It shows the jury that professional judgment in real time, under pressure, with incomplete information, is not the same as rule-following on a checklist. This reintroduces the System 2 considerations that the Reptile’s binary framework excluded, but it does so through a System 1 frame: the story of a professional making a difficult decision.

Second, it draws a moral distinction between a mistake and a wrong, between a professional who tried to do the right thing and erred and one who did not care. The Reptile collapses this distinction. The defense must restore it. This is a System 1 argument because every juror intuitively understands the difference between trying and failing versus not caring enough to try.

Third, it affirmatively establishes the defendant’s commitment to safety, not as a rebuttal but as a narrative that was true before the lawsuit existed. Evidence of safety investments, training programs, compliance records, and quality improvement processes must be part of the case’s fabric from the beginning.

Fourth, it names the injustice of hindsight. Plaintiff’s counsel is asking the jury to judge a real-time decision with the benefit of information that only became available later. Every juror has made a decision under uncertainty that turned out badly. Making them feel that injustice activates a System 1 response of fairness that competes directly with the System 1 response of fear that the Reptile has engaged.

In Part 3, I will get tactical. I will walk through the ten specific question patterns that Reptile practitioners use in depositions and show you exactly how to prepare your witnesses to respond. Not with evasion. Not with coached non-answers. With responses that are honest, that acknowledge the importance of safety, and that prevent the Safety Rule from being constructed on the record.

In Part 4, I will cover the counter-narrative from voir dire through closing argument: how to build the defense psychology before plaintiff’s counsel establishes theirs, and how to maintain it through every phase of trial.

The Reptile Theory is the most effective plaintiff’s methodology developed in the last two decades. It deserves to be taken seriously, not as a trick to be blocked, but as a System 1 framework to be understood and countered with equal psychological sophistication.

Subscribe for the full series. Next up: the deposition playbook.

In 2002, Daniel Kahneman won the Nobel Prize in Economics for research he conducted with Amos Tversky on how human beings actually make decisions. Not how we think we make decisions. Not how rational-choice models say we should make decisions. How we actually do it.

Their central finding, now foundational in behavioral economics, psychology, and public policy, is that human cognition operates through two distinct systems. Kahneman called them System 1 and System 2. Understanding what that means for defending high-stakes cases may be the most important strategic insight available to trial lawyers today.

Two Systems, One Juror

System 1 is fast, automatic, and intuitive. It operates below conscious awareness. It is the part of your brain that flinches before you realize something was thrown at you, that reads facial expressions without effort, that makes snap moral judgments based on how a story feels rather than what the evidence says. System 1 runs on pattern recognition, emotional association, and narrative coherence. It does not analyze. It reacts.

System 2 is slow, deliberate, and analytical. It is the part of your brain that works through a complex math problem, evaluates competing arguments, and weighs evidence against standards of proof. System 2 requires effort. It is cognitively expensive. And critically, it is lazy: it will not engage unless something forces it to.

Here is the problem for defense lawyers: jurors make moral judgments about corporate defendants through System 1. They reach conclusions about responsibility, blame, and punishment through fast, intuitive, emotionally driven processing. And the entire architecture of the modern plaintiff’s trial strategy is designed to keep them there.

How the Plaintiff’s Bar Exploits System 1

Consider the most common plaintiff’s framing device in a product liability suit: “They chose profits over your safety.” Six words. No evidence cited. No factual predicate established. But those six words activate a cascade of System 1 processing that can determine the outcome of a case.

System 1 hears “profits over safety” and immediately generates a moral judgment: greed. It does not ask whether the company actually faced a binary choice between profit and safety. It does not consider that pharmaceutical companies invest billions in safety research precisely because unsafe products destroy shareholder value. It does not evaluate the regulatory framework that governs how safety decisions are actually made. System 1 does not evaluate. It reacts. And the reaction to “profits over safety” is visceral moral disgust.

This is not an accident. The plaintiff’s bar has built its trial methodology around keeping jurors in System 1 for as long as possible. Reptile Theory is fundamentally a System 1 exploitation method. It uses fear-based language to activate jurors’ protective instincts. It frames every case as a community danger that the jury must eliminate. It bypasses analytical reasoning entirely and targets the survival-oriented, threat-detecting machinery of the fast brain. I will cover Reptile Theory in depth in Part 2 of this series.

Kahneman’s research explains why this works. System 1 is governed by what he called “what you see is all there is,” the brain’s tendency to build the most coherent story possible from whatever information is immediately available, without searching for what is missing. A plaintiff’s opening that presents a sympathetic victim, a wealthy corporation, and a simple moral narrative gives System 1 everything it needs to render a verdict. The fact that the story is incomplete, that it omits regulatory context, engineering complexity, and distributed responsibility: System 1 does not notice. It has a coherent narrative. That is enough.

The Framing Effect: Same Facts, Different Verdicts

Perhaps the most directly applicable finding from Kahneman and Tversky’s research is the framing effect, the demonstration that how information is presented determines the decisions people make, even when the underlying facts are identical.

In their classic experiment, subjects were asked about a medical treatment. One group was told the treatment had a “90% survival rate.” Another group was told it had a “10% mortality rate.” Same treatment. Same statistics. Dramatically different responses. The survival frame produced confidence. The mortality frame produced fear.

Now apply this to a product liability trial. “This warning label was inadequate” activates a frame of failure and insufficiency. “This product included comprehensive safety information” activates a frame of thoroughness and responsibility. The factual content, the same label and the same words, is identical. But the frame through which a juror processes that content determines the moral judgment that follows.

This is not about spin. This is not about deceiving jurors or misrepresenting evidence. This is about recognizing a fundamental truth of human cognition: the structure of language shapes the moral conclusions that listeners draw from identical facts. If you ignore that truth, you are ceding the most powerful weapon in trial to opposing counsel.

Why the Traditional Defense Strategy Fails

The standard defense playbook is, at its core, an appeal to System 2. Present the scientific evidence. Call the expert witnesses. Walk the jury through the clinical trial data, the FDA approval process, the engineering specifications. Trust that if you give jurors enough information, they will reason their way to the correct conclusion.

This strategy assumes that jurors are operating in System 2, that they are carefully weighing evidence, comparing competing expert opinions, and applying legal standards with analytical rigor. Kahneman’s research tells us that this assumption is, more often than not, wrong.

System 2 engagement requires cognitive effort, and people naturally avoid cognitive effort when System 1 has already provided a satisfying answer. If a juror has already formed a moral intuition (“this company was greedy and someone got hurt”), System 2 is not going to spontaneously activate to re-evaluate that intuition. The juror will instead use System 2 selectively, to find reasons that support the conclusion System 1 already reached. Kahneman called this confirmatory bias, and it is the silent killer of evidence-heavy defense strategies.

This explains a phenomenon every experienced defense lawyer has encountered: you put on a technically flawless case with strong experts, solid standard-of-care compliance, and favorable medical literature, and you still lose. The science was right. The evidence was right. The jury was not persuaded. The usual post-mortem attributes the loss to a bad venue, a sympathetic plaintiff, or anti-corporate bias. Kahneman’s framework suggests a different explanation: you were speaking to System 2 while the plaintiff was speaking to System 1, and System 1 had already decided the case before your first expert took the stand.

Competing at the System 1 Level

The answer is not to abandon evidence and expertise. The answer is to stop relying on evidence alone. Defense counsel must learn to compete at the System 1 level, to build moral frames that are at least as intuitive, as emotionally resonant, and as narratively coherent as the frames the plaintiff deploys.

This means several things in practice. It means understanding that the first moral frame a juror encounters tends to anchor all subsequent processing, so getting your frame in first, during voir dire, is not a luxury but a necessity. It means recognizing that denying a plaintiff’s frame (“we did not put profits over people”) actually reinforces it, because negation requires cognitive activation of the concept being denied. It means building defense narratives that answer jurors’ moral intuitions directly rather than hoping that enough data will override those intuitions.

What does a defense-aligned System 1 frame look like? Consider the difference between these two opening statements:

“The evidence will show that our product complied with all FDA requirements and that the warning label contained adequate safety information.”

Versus:

“This case is about how we make medicine safer together: the clinical researchers who test it, the FDA doctors who review it, the physicians who prescribe it, and the patients who use it.”

The first statement is a System 2 appeal: factual, regulatory, evidence-based. It asks jurors to analyze. The second statement is a System 1 appeal: it creates a moral frame of shared responsibility, distributed agency, and collaborative safety. It gives System 1 a coherent narrative in which the manufacturer is a partner, not a villain. Both are truthful. Both are ethical. But only one competes with “they chose profits over your safety” at the level where jurors actually make decisions.

What Comes Next

Kahneman’s research does not tell defense lawyers to abandon science, dismiss evidence, or manipulate jurors. It tells us something more important: presenting excellent evidence is necessary but not sufficient. Evidence must be delivered within a moral frame that System 1 can process, or it will be filtered through whatever frame the plaintiff has already established.

The plaintiff’s bar understood this before we did. Reptile Theory, nuclear verdict strategies, and anchoring techniques are all applications of behavioral psychology to trial practice. They are System 1 strategies. And they are winning.

In Part 2 of this series, I will take this framework and apply it to the single most effective plaintiff’s methodology in modern litigation: the Reptile Theory. I will show you exactly how it operationalizes the System 1 principles we have discussed here, why the standard defense responses fail, and what actually works. In Parts 3 and 4, I will get tactical: the specific deposition question patterns Reptile practitioners use and the counter-narrative framework from voir dire through closing.

The question for defense lawyers is whether we are willing to accept what Kahneman proved and adapt our trial strategies accordingly, or whether we will keep presenting facts to System 2 and wondering why we lose.

Subscribe for the full series. I publish weekly on the psychology of trial work, AI in litigation, and the craft of building a serious trial practice.

They’re right that the output was useless. They’re wrong about why.

The problem isn’t the technology. The problem is the prompt. A generic instruction produces generic output. But a prompt that’s been engineered for a specific litigation task, one that understands what a lawyer actually needs from a deposition, produces something that will fundamentally change how you prepare.

I’ve spent the last year building an AI-litigation operations system using Claude. That system now includes 66 documents, over 2,800 checklist tasks across 13 phases of a case lifecycle, and a suite of purpose-built AI prompts designed for specific litigation workflows. Today I’m going to walk you through three of the prompts I use for deposition preparation, the phase of litigation where I’ve seen AI provide the most immediate, tangible value.

One of these I’ll give you in full. The other two I’ll describe at a level that lets you understand the methodology and start building your own versions. All three are tools I use in active cases, not experiments I’m theorizing about.

Why Depositions Are the Highest-Value AI Target

Before I get into the prompts themselves, it’s worth understanding why deposition preparation is where AI is most valuable.

Depositions sit at the intersection of massive information intake and strategic decision-making. To prepare well, you need to analyze document productions that might run tens of thousands of pages, cross-reference prior testimony, analyze expert reports, identify contradictions, map relationships between witnesses, and then distill all of that into a question framework that advances your case theory while locking the witness into positions you can exploit at trial.

That synthesis work is exactly what large language models are built to do. Not the strategic decisions, those are yours. But the mechanical labor of processing, organizing, cross-referencing, and surfacing relevant information from massive volumes of material? AI handles that in minutes instead of hours. And when the mechanical work is done faster, you get something more valuable than time savings: you get the mental bandwidth to actually think about strategy before you walk into the room.

Here are the three prompts that made the biggest difference in how I prepare.

Prompt #1: The Discovery-to-Deposition Question Pipeline

This is the prompt I’m giving you in full, because it’s the one that will produce an immediately useful result the first time you run it.

The problem it solves: You’ve completed document review on a case. You have thousands of pages of production, contracts, emails, internal communications, incident reports, maintenance records, whatever the case involves. Somewhere in those documents are the facts that need to become deposition questions. But the translation from “I’ve reviewed the documents” to “here is a targeted deposition outline organized by topic with specific document references” is a brutal, time-consuming process. Most lawyers do it with a legal pad, a highlighter, and six hours they don’t have.

This prompt does it in minutes.

Here’s the full prompt. I use it with Claude or Abacus.AI, but the structure will work with any capable large language model. Replace the bracketed sections with your case-specific information.

PROMPT: Discovery-to-Deposition Question Pipeline

You are a senior litigation attorney preparing for the deposition of [WITNESS NAME AND ROLE]. This is a [CASE TYPE] case involving [ONE-SENTENCE CASE SUMMARY]. The deposition is being taken [for discovery / to preserve testimony / as a Rule 30(b)(6) designee on the following topics: LIST TOPICS].

I am going to provide you with documents produced in discovery that are relevant to this witness. Your task is to:

Step 1 — Document Analysis: For each document, identify (a) the key facts relevant to the claims or defenses, (b) any statements, admissions, or inconsistencies that could be exploited, (c) any facts that suggest the witness has knowledge beyond what the document states on its face, and (d) any gaps — things the document raises but does not answer.

Step 2 — Topic Clustering: Organize the facts, admissions, inconsistencies, and gaps you identified into logical topic clusters. Each cluster should represent a line of questioning that hangs together as a coherent deposition chapter. Name each cluster with a clear topic heading.

Step 3 — Question Generation: For each topic cluster, generate specific deposition questions. Each question should: (a) be designed to establish a single fact, (b) reference the specific document and page/paragraph that supports or motivates the question, (c) be phrased in plain, non-leading language appropriate for a discovery deposition [OR: be phrased as leading questions appropriate for an adverse witness], and (d) be sequenced to build logically — start with foundational facts, then move to the details, then to the inconsistencies or admissions.

Step 4 — Flag Strategic Opportunities: After generating the questions, separately flag: (a) any document contradictions that could support impeachment at trial, (b) any areas where the witness is likely to claim lack of knowledge (and the documents that would challenge that claim), (c) any topics where the documents suggest other witnesses may have more knowledge, and (d) any privilege or work product issues that the documents raise but do not resolve.

Output format: Organize your output as a structured deposition outline with topic headings, numbered questions under each topic, document references in brackets after each question, and a separate “Strategic Flags” section at the end.

[PASTE DOCUMENTS OR DOCUMENT SUMMARIES BELOW]

Why this prompt works: It doesn’t just ask AI to “help with deposition prep.” It mirrors the actual cognitive workflow of an experienced lawyer preparing for a deposition, document analysis, topic organization, question sequencing, and strategic assessment, but breaks each step into explicit instructions that force the model to work systematically rather than generating a superficial summary.

The Step 4 strategic flags are where the real value lives. I’ve had this prompt surface document contradictions I missed in manual review, identify witnesses I hadn’t considered deposing, and flag privilege issues that would have become problems if I’d asked the wrong question in the room.

How I actually use it: I don’t paste raw documents into the prompt cold. I run my documents through an initial summarization pass first, a separate prompt that extracts the key facts, dates, and actors from each document into a condensed format. Then I feed those condensed summaries into the pipeline prompt. This lets me process much larger document sets without hitting context limits, and the structured summaries give the model cleaner inputs to work with.

After the prompt generates the outline, I spend 30 to 60 minutes editing, reordering questions, adding case-specific follow-ups I know I’ll need, and cutting questions that don’t advance my theory. What used to be a six-to-eight-hour preparation process is now 90 minutes, and the quality of the outline is consistently better because the AI doesn’t forget documents the way I do when I’m reviewing my fourteenth exhibit at 10 PM.

Prompt #2: Witness Sentiment Analysis

This one I’ll describe at a higher level, because the prompt itself is more complex and the value is harder to capture without showing you the output on a real transcript.

The problem it solves: When you read a deposition transcript, you’re reading words on a page. But depositions aren’t just words. They’re performances. A witness who answers “I don’t recall” seventeen times is telling you something different from a witness who answers “I don’t recall” twice. A witness whose answers get shorter and more clipped when you approach a particular topic is signaling something. A witness who volunteers information unprompted in one area but gives monosyllabic responses in another is drawing you a map of where they’re comfortable and where they’re not.

Experienced lawyers read these patterns intuitively. But intuition is inconsistent, and it’s limited by the fact that you can only hold so much of a transcript in your active memory at once.

My sentiment analysis prompt processes an entire transcript and maps the emotional and behavioral patterns across the full deposition. It identifies sections where the witness was defensive, evasive, rehearsed, confident, or uncertain. It flags changes in response length, hedging language, and qualification patterns. It tracks how the witness’s demeanor shifted across topics and notes the specific transitions where the shift occurred.

The output isn’t a replacement for your own read of the transcript. It’s a second set of eyes that processes the entire document with a consistency your memory can’t match. I use it primarily for two things: identifying the topics I need to revisit in follow-up depositions or at trial, and flagging the moments that will resonate with a jury if I need to play video clips. The prompt essentially tells me where the witness was most vulnerable and where they were most prepared.

The methodology behind it draws on research in linguistic analysis and deception detection. It’s not a lie detector. But it’s remarkably good at identifying the patterns that experienced litigators learn to recognize over years of taking depositions — it just does it faster and more systematically.

Prompt #3: Expert Report Deconstruction for Deposition Prep

The third prompt is specifically designed for deposing expert witnesses, which is a different animal from fact witness depositions. The problem it addresses is one every lawyer knows: an expert report arrives, it’s dense, it’s technical, it’s written in the expert’s domain-specific language, and you need to deconstruct it into a deposition strategy in a compressed timeframe.

My prompt for this task does three things. First, it breaks the expert’s report into its component logical steps, every assumption, every data input, every methodological choice, and every inferential leap the expert made to reach their conclusions. This is the architecture of the opinion, and once you can see it clearly, you can see where it’s weakest.

Second, it identifies the pressure points, the places where the expert made a choice between competing methodologies, relied on incomplete data, made assumptions that aren’t supported by the record, or reached conclusions that don’t follow from their own analysis. These become the foundation of your cross-examination outline.

Third, it generates a set of Daubert or Rule 702 challenges keyed to the specific methodology the expert used. If the expert relied on a differential diagnosis, the prompt identifies the steps in the differential etiology framework and flags where the expert deviated from the accepted methodology. If the expert used a regression analysis, it identifies the variables, the assumptions, and the limitations. The output isn’t a motion to exclude. It’s the factual foundation you need to build one.

What makes this prompt different from asking AI to “summarize an expert report” is that it’s built around the adversarial task. It doesn’t just tell you what the expert said. It tells you where the expert is exposed. That’s the difference between an AI tool that saves you time and an AI tool that makes you more effective.

The System Behind the Prompts

These three prompts don’t operate in isolation. They’re components of a larger system I’ve built, one where the output from document review feeds into deposition preparation, which feeds into trial strategy, which feeds into cross-examination outlines, and so on across 13 phases of a case lifecycle.

That’s the part most lawyers miss when they experiment with AI. A single clever prompt is useful. A system of prompts that chain together across the entire lifecycle of a case is transformative. The discovery-to-deposition pipeline doesn’t just help me prepare a better deposition. It creates structured output that I can feed into my trial preparation prompts months later, so the work compounds instead of being done once and filed away.

I’ll be writing more about that system in future posts. How it’s structured, how the pieces connect, and what it’s meant for the way I practice. If you’re a lawyer who tried AI once and dismissed it, I’d encourage you to try the pipeline prompt above on your next deposition. Run it exactly as written. See what it produces. Then ask yourself what your practice would look like if every phase of your litigation workflow had a tool like that.

That’s the question that started all of this for me. The answers are what I’ll keep sharing here.

If you found this useful, subscribe. I publish at least twice a month on AI in litigation, the psychology of trial work, and the craft of building a trial practice. Next up: a deep dive into Reptile Theory from the defense perspective.

There are two kinds of lawyers right now.

The first kind is pretending AI doesn’t exist. They’re billing the same way, preparing the same way, thinking the same way they did five years ago. Some of them are very good lawyers. But they’re standing still, and the ground is moving under them.

The second kind has downloaded ChatGPT, asked it to write a motion to dismiss, gotten back something mediocre, and concluded that AI is overhyped. They tried it. It didn’t replace them. They moved on.

Both groups are wrong. And both groups are going to lose ground to a third kind of lawyer — one that barely exists yet but is coming fast.

I call them AI-Enhanced lawyers. Not lawyers replaced by AI. Not lawyers dabbling in AI. Lawyers who have fundamentally restructured how they think, prepare, and execute by integrating AI into the deep architecture of their practice.

I’m one of them. And this Substack is where I’m going to show you what that looks like from the inside.

The Shift

I’m a trial attorney. I handle commercial litigation and catastrophic personal injury defense from Arkansas and around the country — eight-figure cases where the stakes are real and the margin for error is zero. The kind of work where you live inside document productions, depositions, expert reports, and courtrooms. Miss a deadline, overlook a document, or walk into a deposition underprepared, and your client pays the price — sometimes eight figures worth of it.

About a year ago, something shifted in how I work. Not gradually. A fundamental break.

I had been using AI tools — primarily Claude, built by Anthropic — for discrete tasks. Summarize this deposition. Draft a discovery response. Outline arguments for a motion. Useful, but incremental. The kind of thing that saves you thirty minutes here and there but doesn’t change the game.

Then I started asking a different question. Not “what can AI do for me on this task?” but “what would my entire practice look like if AI were embedded in every phase of litigation, from case intake to post-trial?”

That question changed everything.

I stopped treating AI as a tool I occasionally consulted and started treating it as an infrastructure I was building. I began constructing a system — document by document, prompt by prompt, workflow by workflow — that would fundamentally alter how I prepare cases, develop strategy, take depositions, evaluate settlements, and try lawsuits.

Today, that system is 66 documents deep. It includes a master litigation checklist spanning over 2,800 tasks across 13 phases of a case lifecycle. It includes a seven-prompt AI intelligence suite that handles everything from cross-deposition impeachment mining to jury simulation to witness sentiment analysis. It includes standard operating procedures for every phase of litigation — not the kind of SOPs that live in a binder and collect dust, but living documents that integrate directly with AI to produce real-time, case-specific guidance.

I built all of this while maintaining a full caseload. I built it because I realized something that most of the AI-in-law conversation is missing entirely.

What the AI Conversation Gets Wrong

Here’s the discourse right now: Will AI replace lawyers? The answer is no, and the question is boring. AI is not going to sit at counsel table and cross-examine a witness. It’s not going to read a jury. It’s not going to stand in a hallway at mediation and tell opposing counsel that their demand is delusional with the precise blend of conviction and diplomacy that makes them believe you’ll try the case.

But here’s what AI will do — what it’s already doing in my practice: it will annihilate the rote work that consumes 60 to 70 percent of a litigator’s time. The document review that takes forty hours. The deposition preparation that involves manually cross-referencing five prior transcripts. The discovery responses that require restating the same objections across two hundred requests. The chronology-building that means reading a thousand pages of medical records with a yellow legal pad.

AI handles all of that now. Not perfectly. Not without oversight. But fast enough and well enough that the time I used to spend on mechanical tasks is now available for the work that actually wins cases: strategic thinking, creative problem-solving, psychological analysis of witnesses and jurors, and the kind of deep case knowledge that only comes from having the bandwidth to actually think.

This is the part nobody is talking about. The lawyers who will dominate the next decade aren’t the ones who use AI to write mediocre first drafts faster. They’re the ones who use AI to remove the noise so they can do the deep work that separates competent lawyers from elite ones.

The rote tasks aren’t just time-consuming. They’re cognitively deadening. They fill your day with low-level processing that leaves you mentally depleted before you ever get to the hard questions — the questions that actually matter to your client’s outcome. AI doesn’t just save you time. It gives you back your mind.

The Psychology Problem

This brings me to the second thing I write about: the psychology of litigation.

Trial work is a psychological enterprise from start to finish. Jury selection is applied cognitive psychology. Witness preparation is behavioral conditioning. Cross-examination is strategic destabilization. Mediation is negotiation psychology under time pressure with money on the table. And the lawyer’s own mental state — their confidence, their emotional regulation, their capacity for sustained attention — is the invisible variable that affects every single one of these activities.

Yet almost nobody writes about this with practitioner depth. There’s academic literature on jury decision-making. There are CLE presentations on Reptile Theory. But there’s a vast gap between the scholarly research and what a trial lawyer actually experiences in a courtroom, and almost no one is bridging it.

I think about this constantly. When I prepare a witness, I’m thinking about cognitive load theory and how memory works under stress. When I plan a cross-examination, I’m thinking about the psychological principles that make certain question sequences more devastating than others. When I evaluate a case for settlement, I’m accounting for the cognitive biases that affect how jurors process risk, causation, and damages — and how opposing counsel’s awareness or ignorance of those same biases changes the negotiation landscape.

AI amplifies all of this. One of the tools in my system is an AI-driven jury simulation that takes the facts, the jurisdiction, the demographics, and the case theory and pressure-tests it against modeled juror psychology. Another is a witness sentiment analysis that helps me identify not just what a deponent said, but the emotional patterns underneath their testimony — where they were defensive, where they were rehearsed, where they were genuinely uncertain. These aren’t novelties. They’re becoming core components of how I prepare.

The lawyers who will thrive in this environment are the ones who understand both the technology and the human psychology it’s designed to analyze. That intersection is where I live professionally, and it’s what I’ll be writing about here.

What I’ve Built

Let me be specific about what exists, because specificity is what separates this Substack from the hundreds of “AI will transform law” thought pieces you’ve already scrolled past.

My Litigation Operations System currently includes:

A Master Litigation Checklist — over 2,800 tasks organized across 13 phases of a case lifecycle, from initial case assessment through post-trial motions and appeals. Every phase has decision trees, compliance deadlines, and AI-integrated prompts that generate case-specific action plans. This isn’t a static checklist. It’s a dynamic system that tells me what to do today, what’s coming next week, and what I’m at risk of missing.

A 66-Document Standard Operating Procedure Library — covering everything from lien management to social media investigation protocols, from co-defendant joint defense agreements to medical records management, from trial budgeting to witness preparation. Each document is designed to be both a standalone reference and a component of the larger system.

A Seven-Prompt AI Intelligence Suite — purpose-built prompts for cross-deposition impeachment mining, witness sentiment analysis, jury simulation, theme extraction, chronology building, discovery-to-deposition question pipelines, and work product conversion. These prompts chain together across the case lifecycle, so the output from one phase becomes the input for the next.

A Case Management Prototype — a working application that encodes the entire system into an automated daily briefing tool with phase determination, compliance tracking, and AI-generated morning action plans.

I didn’t build this to sell it. I built it because I needed it. I built it because I was tired of the gap between how I knew litigation should be managed and how it actually was being managed — in my practice and in every practice I’d ever seen. The fact that it turned out to be something other lawyers want is a secondary discovery.

What You’ll Get Here

This Substack has three lanes.

The Trial Mind — the psychology of litigation. How cases are won and lost in the minds of jurors, judges, witnesses, and lawyers. Reptile Theory. Cognitive bias. Witness preparation psychology. The emotional architecture of a trial. If you’re a lawyer, this will make you sharper. If you’re a potential client, this will show you how serious trial lawyers actually think about your case.

AI in the Arena — practical AI integration in active litigation. Not the vendor pitch. Not the conference panel version. The real thing: what works, what doesn’t, what I’m building, and how it performs under the pressure of actual deadlines and actual courtrooms. Prompt engineering. System architecture. Workflow design. Implementation failures and what I learned from them.

The Craft — the art and discipline of trial practice itself. Case theory development. Deposition strategy. Mediation tactics. What it takes to build an elite trial team. The habits and systems that separate preparation from mastery.

I’ll publish at least twice a month, often weekly. Every post will be specific, tactical, and drawn from active practice — not theory, not aspiration, not what I read someone else did. If I’m writing about a deposition technique, it’s because I used it last week. If I’m writing about an AI workflow, it’s because it’s running in my system right now.

Who This Is For

If you’re a trial lawyer who wants to understand how AI actually fits into litigation practice — not the marketing version, the operational version — this is for you.

If you’re a lawyer at any level who’s interested in the psychology of advocacy and wants to think more rigorously about how human cognition affects case outcomes, this is for you.

If you’re someone facing a legal challenge and you want to understand what a trial lawyer who takes this work seriously actually does to prepare your case, this is for you.

And if you’re someone who senses that the legal profession is about to change dramatically and wants to hear from a practitioner who’s already living inside that change — not commenting on it from the sidelines — then subscribe.

I’ll be here every week.

Let’s get to work.

The Hidden Science of Winning Minds at Trial

Baxter Drennon

Partner, Hall Booth Smith, P.C.

Based on a presentation delivered at the
Product Liability Advisory Council — Spring Meeting, March 2026

The Invisible Weapon You’re Facing

Two pharmaceutical cases. Same drug—a blood thinner prescribed to millions of patients. Same alleged injury—a bleeding event. Same medical science—identical clinical trial data, FDA approval process, and warning labels. Both tried to verdict in the same year.

Case one: defense verdict. Zero dollars. Jury deliberated four hours. Case two: forty-seven million dollars for the plaintiff. Jury deliberated three hours.

The difference? Six words in the opening statement.

The plaintiff’s attorney in case two told the jury: “They chose profits over your safety.” Six words that created a moral frame so powerful that 14,000 pages of clinical data, FDA approval, and expert testimony from three leading cardiologists couldn’t overcome it. The defense attorney in case one used a different construction: “This case is about how we make medicine safer together.” Seven words that created a completely different moral universe—one where responsibility was shared, where the manufacturer was a partner in safety rather than an enemy of it.

Same facts. Different language structure. Opposite outcomes.

This disparity is not an anomaly. It reflects a fundamental truth about how juries process information and render verdicts—a truth that the plaintiff’s bar has been exploiting for years while the defense bar has largely failed to recognize it, let alone respond.

The thesis of this article is straightforward: language structure—not content—is the invisible weapon being used against corporate defendants in high-stakes product liability trials. The plaintiff’s bar has weaponized psycholinguistics, deploying techniques drawn from cognitive linguistics, conceptual metaphor theory, and behavioral psychology. Defense counsel must learn to recognize these techniques, neutralize them, and deploy competing frames. Those who do will transform their trial outcomes. Those who don’t will continue losing cases they should win.

I. The Science: How Language Structure Hijacks Juror Reasoning

The foundation of psycholinguistic framing rests on a principle established through decades of peer-reviewed research in cognitive linguistics and behavioral psychology: the structure of language influences and shapes thought. Not merely reflects it—shapes it. When you change the grammatical structure of a sentence without altering its factual content, you activate different neural pathways, trigger different emotional responses, and create different moral judgments in the listener’s brain. This is measurable neuroscience, and it has profound implications for trial practice.

Conceptual Metaphors and Moral Reasoning

George Lakoff, a cognitive linguist at UC Berkeley, spent forty years studying how metaphors structure thought. His central finding upends the common assumption that metaphors are merely decorative language: metaphors are the fundamental mechanism through which human beings understand abstract concepts.

Consider the concept of “corporate responsibility” in product liability. It is inherently abstract—you cannot see, touch, or directly measure it. Our brains therefore understand it through metaphor. When a plaintiff’s attorney uses the phrase “corporate shield,” the listener’s brain activates associations of protection, barrier, and concealment. The implicit message: the corporation is hiding, evading accountability. When that same attorney invokes an “accountability gap,” the brain activates spatial metaphors—holes, openings, dangerous spaces where things fall through. The implicit message: someone should be held responsible but isn’t.

These metaphors do not operate in isolation. In a skilled plaintiff’s opening, they cluster: “This corporation has hidden behind a legal shield for too long. There’s been an accountability gap—a gap between what they knew and what they told consumers. Today, you get to close that gap.” Three metaphors—shield, gap, closing—working in concert to produce a single moral conclusion before a single fact has been presented. That is not argument in the traditional sense. It is linguistic architecture that pre-structures how jurors will evaluate the evidence.

Framing Effects: Same Facts, Different Realities

The work of Daniel Kahneman and Amos Tversky on framing effects earned Kahneman the Nobel Prize in Economics in 2002. Their research demonstrated with experimental rigor that identical information presented in different frames produces different decisions.

In their classic experiment, one group of subjects was told a medical treatment had a “90% survival rate” while another group heard the same treatment described as having a “10% mortality rate.” Same statistics. Opposite frames. The result: subjects overwhelmingly chose the treatment when it was framed as survival and were far more hesitant when it was framed as mortality. “Survival” activates concepts of success, life, and safety. “Mortality” activates death, risk, and danger. The frame determines the emotional response, and the emotional response drives the decision.

The application to product liability is direct. When a plaintiff’s attorney says “this warning label was inadequate,” the frame activates concepts of insufficiency and failure. When a defense attorney says “this product included comprehensive safety information,” the frame activates completeness and responsibility. Same label. Same words printed on it. Completely different moral judgments triggered by the linguistic frame surrounding it.

Syntax and Causality: Word Order Determines Responsibility

Perhaps the most underappreciated mechanism of psycholinguistic framing is syntactic positioning of agency—the way that word order within a sentence determines who the listener’s brain assigns causal responsibility to.

Consider two descriptions of the same event: “The plaintiff was injured when the product failed” versus “The product was being used when the plaintiff was injured.” In the first construction, “product” occupies the agent position and “failed” is the active verb. The brain assigns responsibility to the product. In the second, the focus falls on active use, and the implicit question becomes: how was the product being used? Same event. Same injury. Different syntactic structure. Different assignment of moral responsibility.

Plaintiff’s attorneys exploit this mechanism systematically. Every key sentence places the corporate defendant in the grammatical subject position: “The company decided.” “The manufacturer knew.” “They chose.” “The corporation failed to.” Always active voice. Always the defendant as the causal agent. The result is that jurors absorb a narrative of corporate agency and responsibility through sentence structure alone, independent of the factual content of the testimony.

II. The Plaintiff’s Psycholinguistic Playbook

Understanding the science is necessary but insufficient. Defense counsel must also recognize how these principles are operationalized in plaintiff’s trial strategy. Three core frames appear in virtually every product liability case.

“Profits Over People”

This frame creates a false moral binary: either the company cared about safety or it cared about money. If the company earned a profit, it must have sacrificed safety. The linguistic markers are phrases like “chose to,” “decided not to,” “put profits ahead of,” and “prioritized earnings over.” Once this frame is established, every business decision becomes evidence of greed. The frame works because it activates what moral psychologist Jonathan Haidt identifies as the purity/degradation axis of moral reasoning: making money from something that hurt someone feels morally contaminating, triggering a visceral response that overrides analytical evaluation.

“Corporate Indifference”

This frame asserts that the company knew about the danger and did nothing. Its linguistic trick is the use of passive voice to erase context. “Warnings were inadequate” eliminates the question of who determined adequacy, by what standard, and when. “The risk was known” conflates all corporate knowledge—one engineer’s email, a preliminary internal review, a CEO’s briefing—into a single malicious entity that “knew.” Passive construction creates a strawman of unified corporate consciousness, as though fifty thousand employees all possessed the same information and collectively decided to ignore it.

“Dangerous Product”

This frame uses nominalization—the conversion of a complex process into a simple noun. “The defect” takes a complicated engineering question (Was the design reasonable given the state of science, intended use, regulatory requirements, and material limitations at the time?) and reduces it to a two-word label. Once something is nominalized as “defective,” the brain stops asking questions about it. Defective things are broken. Broken things should not have been sold. The engineering context vanishes, replaced by a simple moral judgment.

The Reptile Connection

These techniques are not theoretical concerns. Reptile Theory—the dominant plaintiff trial strategy for the past fifteen years—is fundamentally a psycholinguistic method. It uses fear-based language to activate jurors’ protective instincts and frames every case as a community danger that the jury must eliminate through its verdict. Reptile is a System 1 exploitation strategy disguised as trial methodology, and the plaintiff’s bar teaches it openly in their continuing legal education programs. Meanwhile, most defense lawyers remain unaware of the methodology they are facing, let alone how to counter it.

III. Recognize: Spotting Linguistic Traps in Real Time

The first skill defense counsel must develop is frame recognition. Frames are most powerful when they are invisible. Once you can identify them, you can counter them. Frame recognition applies across every stage of trial: voir dire, opening statement, witness examination, and closing argument.

Voir Dire: Where Frames Are Planted

Voir dire is where frames are most dangerous because they are disguised as neutral questions. The jury’s defenses are not yet engaged; prospective jurors believe they are simply answering questions about their ability to be fair.

Consider a common plaintiff’s voir dire question: “Can you be fair to a corporation that made millions of dollars from this product?” The question sounds neutral. It embeds three frames. First, it links corporate identity with profit—not employment, not innovation, not research, but “made millions.” Second, “from this product” creates a causal connection between the profit and the product that allegedly injured the plaintiff. Third, the question format itself—“Can you be fair to”—presupposes that fairness toward this defendant requires conscious effort, implying that the default instinct is skepticism.

Defense counsel should listen for loaded nominalizations (“corporate greed,” “safety shortcuts”), false causality (“because they wanted to save money”), and moral weighting in hypotheticals (“If you learned that the CEO knew about complaints, could you still be fair?”). Each represents a frame being embedded before the trial begins.

Opening Statement: Information Sequence as Moral Causality

In opening statement, frames crystallize into narrative structure. The critical insight is that information sequence determines causal perception. A typical plaintiff’s opening follows a specific progression: sympathetic injured plaintiff, dangerous product, greedy corporation, profit motive. That sequence is not chronological. It is structured to create moral causality in jurors’ minds—a causal chain constructed through information ordering rather than evidence.

Defense counsel should ask four questions when analyzing opposing counsel’s opening: Who is the grammatical subject of key sentences? What metaphors appear repeatedly? What causal sequence is implied by information order? And what is being nominalized that should not be?

Witness Examination: Embedded-Frame Questions

During cross-examination and depositions, plaintiff’s counsel embeds frames in the structure of questions themselves. A question like “You’re aware that 47 people died before you changed the design, correct?” presupposes that the company was aware of a causal link, could have changed the design earlier, chose not to, and that people died because of that choice. None of this is stated explicitly. All of it is embedded in the syntax. The witness cannot challenge these embedded presuppositions without appearing evasive.

Defense counsel should recognize presupposition triggers (“before X,” “when you decided X”), loaded verbs (“concede,” “admit,” “acknowledge”), and compound questions that glue together separate facts to create false causality.

IV. Neutralize: Defusing Plaintiff Frames Without Triggering Backlash

The central paradox of linguistic framing is that you cannot defeat a frame by denying it. As Lakoff demonstrated, if someone says “Don’t think about a blue elephant,” your brain must activate the concept before it can negate it. The activation persists; the negation does not. In trial terms, if a plaintiff says “They put profits over people” and the defense responds “We did not put profits over people,” the jury has now heard “profits over people” twice—once from the plaintiff and once, reinforced, from the defense.

Effective neutralization requires reframing, not denial. Six techniques, deployed across the stages of trial, form a comprehensive neutralization strategy.

Inoculation Through Prediction

In voir dire, defense counsel names plaintiff’s frame before it is deployed, creating cognitive distance. By telling jurors they will hear phrases like “profits over people” and labeling such language as a “slogan, not evidence,” you create a filter. When jurors subsequently hear the phrase from plaintiff’s counsel, they recognize it as the rhetorical technique they were warned about rather than accepting it as established truth.

Preemptive Reframing

Before plaintiff’s counsel can establish their frame in voir dire, defense counsel plants a competing frame. If plaintiff intends to frame the case around corporate greed, defense counsel frames it around shared responsibility: “This case is about whether everyone involved—the manufacturers, the regulators, the physicians, the users—did what they were supposed to do. Can you evaluate all of those parties, not just focus on one?” With this frame anchored, plaintiff’s subsequent focus on a single corporate defendant feels narrow and incomplete.

Frame-Breaking Questions

When plaintiff’s counsel uses abstract corporate language, defense counsel deconstructs the abstraction into concrete human reality: “This product was designed by 47 engineers—real people with families, degrees from top universities, expertise in safety. They spent eight years developing it. Can you evaluate what those 47 engineers actually did, based on evidence, before deciding if the product was reasonably safe?” Replacing “the corporation” with specific human beings makes it far more difficult for jurors to apply frames of corporate indifference.

Sequence Disruption in Opening

Rather than following plaintiff’s information sequence (injury, product, company, greed), effective defense openings disrupt the sequence entirely. Start with safety innovation, not injury. Introduce human actors, not “the corporation.” Address the specific case last, positioning the incident as one data point within a larger context of responsible product development. This reordering breaks the moral causality that plaintiff’s sequence is designed to create.

Frame Shifting on Redirect

After plaintiff’s counsel embeds a frame through cross-examination, redirect offers the opportunity to shift to a different frame entirely. If the cross-examination question was about awareness (“You were aware 47 people died”), the redirect reframes around action: the safety review board convened within 48 hours, the investigation that involved ten specialists over four months, the peer-reviewed findings, the FDA submission. The cross was about passive awareness. The redirect is about active, responsible investigation. Different frame.

Explicit Frame Dismantlement in Closing

Closing argument is the one stage where defense counsel can name plaintiff’s frames explicitly and dismantle them. By identifying the “profits over people” frame as a false dichotomy and replacing it with the reality of engineering tradeoffs—“That’s not a moral choice between good and evil. That’s engineering. That’s science”—defense counsel gives jurors permission to reject the plaintiff’s moral frame and adopt the defense’s alternative narrative.

V. Reframe: Building Defense-Aligned Moral Narratives

The ultimate objective is not merely to neutralize plaintiff’s frames but to offer superior alternatives. A defense frame must answer jurors’ moral questions about fairness and responsibility, align with their real-world experience and common sense, account for the plaintiff’s injury without assigning inappropriate blame, and elevate complexity without creating confusion. Three frames, drawn from trial experience and grounded in cognitive science, accomplish these objectives across the range of product liability cases.

Frame 1: Shared Safety Responsibility

Core principle: Safety is a shared ecosystem, not a corporate monopoly.

In the plaintiff’s frame, a single entity—the manufacturer—bears sole responsibility for safety. The shared responsibility frame establishes that safety is collaborative. Manufacturers design and warn. Regulators review and approve. Physicians prescribe and monitor. Patients follow instructions and report problems. Each actor has a defined role, and the question for the jury becomes whether everyone in the system fulfilled their responsibility—not whether one company failed.

The linguistic structure requires placing multiple agents in the subject position throughout the trial narrative: “Doctors prescribe based on patient history. Patients follow instructions and report side effects. Manufacturers design, test, and warn. Regulators review data and approve.” Each sentence has a different grammatical subject, distributing agency syntactically—the opposite of plaintiff’s strategy, which places the defendant in the subject position of every key sentence.

This frame is most effective in pharmaceutical and medical device cases where multiple actors are genuinely involved in the safety chain. In practice, juries respond to this frame because it aligns with their intuitive understanding that no single entity controls everything about how a product is used and experienced.

Frame 2: Innovation Integrity vs. Impossible Perfection

Core principle: Progress requires accepting the known limits of technology.

Plaintiff’s frame implies an unstated standard of perfection: if someone was hurt, the product must have been defective. The innovation integrity frame makes the implicit standard explicit and then challenges it. Engineers do not ask “Can we eliminate all risk?” They ask “How do we maximize safety given the laws of physics, current materials science, and intended use?”

The linguistic structure uses temporal language to establish the state of knowledge at the time of design (“In 2015, the state of materials science allowed for X but not Y”) and contrasts good-faith engineering with hindsight perfection (“Real-world tradeoffs” versus “theoretically ideal but practically impossible”).

This frame works best in design defect cases where plaintiff’s expert proposes an alternative design. Jurors understand tradeoffs from their own lives—they choose between safety features and cost, between fuel efficiency and vehicle size. The innovation integrity frame aligns with that lived experience.

Frame 3: Complexity Acknowledgment vs. Oversimplification

Core principle: The real world is complicated, and oversimplifying leads to wrong answers.

Plaintiff’s frame reduces complex product development to simple moral binaries: safe versus dangerous, responsible versus negligent, caring versus greedy. The complexity frame embraces nuance as a virtue. It names complexity explicitly (“This isn’t simple, and anyone who tells you it is isn’t respecting your intelligence”) and contrasts the depth of the actual record with plaintiff’s selective presentation (“Plaintiff’s counsel showed you one email. We’re going to show you 3,000 pages of safety documentation”).

This frame is most effective in cases with significant technical nuance, “smoking gun” documents that require context, or complex regulatory histories. It treats the jury as capable of handling complexity—a posture that jurors find respectful and credible.

Choosing and Deploying Your Frame

Effective frame deployment requires discipline. Choose one primary frame per case, matched to the case type and plaintiff’s likely strategy. You may add one supporting frame, but no more—too many frames create confusion rather than clarity. Write a frame statement of two to three sentences that captures the moral narrative. Develop five variations of that statement for repetition without redundancy. Deploy the frame eight to twelve times across trial—voir dire, opening, key witness examinations, and closing—using slightly different language each time while maintaining the same core concept.

VI. The 48-Hour Protocol

Theory without implementation is academic. The following protocol translates the methodology into concrete trial preparation steps that can be executed in 48 hours.

Step One: Audit your current case language (two hours). Read your draft opening statement aloud. Identify embedded frames, passive voice that weakens your position, defensive language, and instances where you are following plaintiff’s causal sequence. Then review opposing counsel’s discovery requests, deposition questions, and draft motions, highlighting every instance of loaded nominalizations, false causality, and presupposition triggers. This creates your map of the linguistic battlefield.

Step Two: Choose your primary defense frame (one hour). Match your case type and plaintiff’s likely strategy to one of the three frames. Pharmaceutical and medical device cases typically call for shared safety responsibility. Design defect cases with alternative design claims call for innovation integrity. Cases with technical complexity or document-heavy records call for complexity acknowledgment.

Step Three: Translate your frame into concrete language (three hours). Write your frame statement—two to three sentences maximum. Write five variations for deployment across trial stages. Map each variation to a specific trial moment: voir dire, opening, direct examination of your key witness, redirect after plaintiff’s cross, and closing. Audit your draft opening one final time, ensuring every sentence serves your frame rather than reacting to plaintiff’s.

VII.The Competitive Advantage

The plaintiff’s bar is already using these techniques. Their continuing legal education programs teach linguistic framing, metaphor theory, and syntactic positioning of agency by name. Defense is behind the curve. Most defense lawyers are still trying cases on the assumption that presenting enough facts, hiring the right expert, and explaining the science clearly enough will carry the day. That assumption is producing preventable losses.

But the asymmetry also represents an opportunity. Psycholinguistic framing is learnable, reproducible, and scientifically grounded. It is not trial magic or innate talent. It is methodology applied to language—methodology that any defense lawyer can master with study and practice.

The two pharmaceutical cases that opened this article illustrate the stakes. The attorney who won the defense verdict did not have better experts, better facts, or a friendlier jury. She had a better frame. Seven words that created a moral universe where her client was a partner in safety, not an enemy of it. The attorney who lost had all the science in the world—but science does not win cases. Frames do.

The lawyer who controls the frame controls the verdict. That is the competitive advantage available to every defense lawyer willing to learn this methodology and deploy it with discipline. Start with one technique. Start with the recognition checklist. Spot one embedded frame in your next voir dire. Neutralize it. See what happens. Then build from there.

Further Reading

George Lakoff, Don’t Think of an Elephant! (2004; revised 2014). The accessible foundation of framing theory, demonstrating that negating a frame reinforces it.

Daniel Kahneman, Thinking, Fast and Slow (2011). The definitive text on dual-process cognition and framing effects. Priority: chapters on framing effects and loss aversion.

Gerry Spence, Win Your Case (2005). The narrative structures that have been winning cases against corporate defendants for decades—essential reading for defense counsel who want to understand the architecture they are facing.

Jonathan Haidt, The Righteous Mind (2012). Moral foundations theory explaining the specific moral dimensions jurors use to evaluate corporate behavior.

David Ball & Don Keenan, Reptile: The 2009 Manual of the Plaintiff’s Revolution (2009). The plaintiff’s bar’s operational manual. Read it through the lens of Kahneman and Lakoff to see the psycholinguistic machinery that makes it effective.

About the Author

Baxter Drennon is a partner at Hall Booth Smith, P.C., where he focuses on the defense of catastrophic injury cases. This article is based on his presentation to the Product Liability Advisory Council at its Spring 2026 meeting.