Companion tools

Intellectual Humility by Design

Tools for courts adapted from the appendix to Steve Leben’s June 2026 Arkansas Law Review article, Intellectual Humility by Design: How Courts Earn Legitimacy and Improve Decisions.

Official location note: The Procedural Justice Hub’s official website is proceduraljustice.org. This page is the official online location for the companion materials while expanded versions are being prepared. A corrected-URL copy of the article is linked here so readers are directed to proceduraljustice.org.

Bench card

Intellectual Humility by Design Bench Card

Recognize · Reflect · Respond · Reassess (R4)
A practical routine for trial and appellate judges.

1. Recognize fallibility

  • What would change my mind?
  • Consider the opposite: write two plausible counterpoints.
  • Premortem: if this decision is later criticized or reversed, why?
  • Outside view: how would a neutral colleague frame this?
  • Reminder: revision is progress; better reasons are the goal.
  • Trial: name one uncertainty or assumption you should test before ruling.
  • Appellate: note a plausible alternative reading and the limits of the holding.

2. Reflect before ruling

  • Voice first: “Here is what I hear - did I get that right?”
  • Ask each side: “What is your strongest point, and the best legal authority for it?”
  • Quick procedural-justice check: voice, neutrality, respectful treatment, and trustworthiness.
  • Trial: paraphrase each party’s key point; invite a one-sentence correction; pause or take the matter under advisement if tempers rise.
  • Appellate: in a bench memo, write one sentence stating each side’s best case; in conference, start with that before voting; ask what you might be missing.

3. Respond with reasons

  • Roadmap early when practical: issue/outcome → rule/standard → key facts → analysis → remedy.
  • Address the strongest opposing point as the party would state it.
  • Name constraints: standard of review, statutory bounds, record limits.
  • Add one sentence recognizing the human stakes; tone signals care, not bias.
  • Trial: give step-by-step oral reasons; make next steps clear.
  • Appellate: engage the best contrary authority respectfully; explain why your reading prevails; keep the holding bounded.

4. Reassess after decision

  • Second look: read as the losing party - did I fairly state and answer the best point?
  • Invite perspective; treat disagreement as data.
  • Optional: 15-minute quarterly R4 retro to review reversals, unclear passages, template tweaks, and one improvement to log for next time.
  • Trial: do a brief tone and clarity pass; update templates; log one improvement.
  • Appellate: ask a colleague to present the strongest contrary argument; revise where others expose gaps.

How to use

  • Pick one prompt from each R when the case is routine; use more prompts when the stakes, uncertainty, or time pressure increase.
  • Use in-chambers prompts before issuing a decision; use on-the-record prompts to make voice, neutrality, and constraint visible.

In chambers

  • Write the best argument for the outcome you are not inclined to reach, using the other side’s own framing.
  • Identify one assumption you are making and what evidence would test it.
  • Ask: “If I am wrong, where am I most likely wrong - facts, law, or procedure?”
  • Use a short premortem: “A reviewing court or the public criticizes this - why?”
  • When you disagree with a colleague’s view, treat it as information: “What are you seeing that I am not?”

Toolkit

Reasons and Explanations Toolkit

Goal: Make reasons do double duty - improving decision quality and making fairness observable, especially to a losing party.

Overriding requirements

  • Orient the reader or listener early: what decision is being made and what questions must be answered.
  • Engage the strongest opposing point in its best form.
  • Make the decision path transparent: rule/standard → key facts → application → conclusion.
  • Include a plain-language layer, especially in trial rulings and for self-represented court users.
  • State constraints and limits: what the court can and cannot decide; what is and is not being decided.
  • Provide next steps in trial courts or clarify what follows from the holding in appellate courts.
  • Use respectful acknowledgment: recognize what was at stake without signaling partiality.

Plain-language layer

Use this as a closing summary in an oral ruling or as a short segment in a written decision.

  • What is being decided today.
  • The rule or standard that controls.
  • The key fact or facts that mattered most.
  • The court’s conclusion.
  • The strongest opposing point and why it does not change the result.
  • Scope and limits.
  • Next steps or effect of the holding.

Trial oral ruling template

The question today is ____. The law requires me to decide it by applying ____. The facts I find most important are ____. For those reasons, I am ordering/denying ____. I have considered ____ (the strongest point you raised), but it does not change the outcome because ____. This ruling decides ____ and does not decide ____. The next steps are ____.

Reasons discipline checklist

  • Have I identified the real question and the decision standard?
  • Have I answered the strongest opposing point, not a straw-man?
  • Can I state the determinative facts in two sentences?
  • Did I explain the decisive point?
  • Should I add a short scope/limits sentence?
  • For trial rulings, did I add next steps in plain language?

Trial courts

Trial Oral-Ruling Sequencing Options

Goals: Improve understanding, reduce unnecessary escalation, and make fairness observable without adding needless length.

Default principles

  • Signpost your structure: “I’m going to do three things: my decision, the reasons, and what happens next.”
  • Orient first. State what is being decided and what questions the court must answer.
  • Use one closed-loop listening move: “Let me make sure I understood your key point: ____. Did I get that right?”
  • Keep reasons to the decision-controlling points.
  • End with a plain-language close: what I decided, why, and what happens next.
  • If emotion spikes, name the process move: “I’m going to pause so I can decide carefully.”

Option A: Path-first

Issue → Standard → Key facts → Decision → Next steps. Use when parties are self-represented, emotion is high, or you suspect people will stop processing once they hear the result.

Option B: Result-first with a promise

Result → Promise reasons/next steps → Issue framing → Reasons → Next steps → Plain-language close. Use when immediate clarity matters.

Option C: Two-pass hybrid

Forecast → Bottom line → Pause → Summarize reasons → Summarize next steps and key deadline → Brief structured reasons → Recap. Use for adverse, high-conflict, or high-stakes rulings.

Retention devices

  • Headline recap: “To summarize: the decision is ____. Two reasons control: ____ and ____. The next step is ____ by ____.”
  • Written or replayable takeaway.
  • Understanding check: “Tell me, in your own words, what you believe you need to do next and by what date.”
  • Bounded Q&A focused on what the order requires and the timeline.

Conference and administration

Constructive Disagreement Protocols

Goals: Improve decisions and written products by surfacing decision-controlling concerns early, protecting collegiality, and reducing late issues.

Quick-use checklist

  • Name the question and the tentative disposition.
  • Surface one decision-controlling counterargument.
  • Identify one assumption that must be true for the reasoning to hold.
  • Decide whether the issue is routine or hard/high-salience.
  • Close with: “Any concerns that would strengthen the opinion or the plan?”

Routine matters

  1. Disposition in one sentence.
  2. One-round issue-spot prompt: each judge offers “no concern” or one sentence naming the strongest decision-controlling concern.
  3. If a concern is raised, decide whether it can be handled now or should trigger the full protocol.

Hard or high-salience matters

  1. Independent-first tentative position: each judge writes the tentative disposition and one reason they might be wrong.
  2. Use the court’s normal speaking order, adding a structured challenge point at the end.
  3. Strongest opposing point round: one judge states the strongest decision-controlling counterargument; the tentative majority gives its best response; the panel asks what would strengthen the reasoning.
  4. Key assumptions check: name assumptions that must be true and ask what could break them.
  5. Micro premortem: “It’s two years later and this decision created serious problems or was reversed. Why?”

Rules of engagement

  • Critique reasons and assumptions, not motives.
  • No interruptions during the opposing-point round.
  • If tension rises, pause and restate the shared aim: accuracy, fairness, and institutional credibility.

Decision science

Decision Hygiene Starter Kit

Goals: Reduce unwanted variability across similar cases, make reasoning more reliable, catch mistakes early, and support legitimacy by showing disciplined, evenhanded decision-making.

Quick-use checklist

  • State the decision and the standard in one sentence.
  • Name the two or three decision-controlling facts.
  • Run one opposing-view check: what is the strongest reason the other side should win?
  • Confirm comparators: what similar cases am I implicitly comparing to?
  • Confirm the next step: written order language, conditions, or deadlines.

Sentencing workflow: anchor protocol

  1. Before hearing: write a tentative range or outcome and one sentence why; list mitigating and aggravating factors; identify one anchor risk; decide what you should hear to confirm or change your tentative view.
  2. During hearing: let each side state the single best point they want you to remember; ask one clarifying question tied to the standard; if a number seems to pull the sentence, pause and restate the governing reasons.
  3. Before announcing: state the standard and the decision-controlling considerations; run a symmetry check; if close, run a micro premortem; state next steps clearly.

Self-audit

The point is not to grade yourself. It is to make noise visible and catch drift early. Pick one recurring category, review a small sample monthly, look for unexplained outliers, and identify one guardrail to try next month.

Case No.DateCategoryOutcomeKey reasonsOutlier?