The Third Field

Cases · evidence in context

What reached the decision?
What survived into action?

The decisive information is not always missing. Sometimes it is known, but never heard, answered or carried into the work.

Read seven documented cases for the way people handled evidence, disagreement and changing conditions. Then explore four fictional situations that show how a continuing decision process can work today.

Science-led. Context matters.

History gives us questions.
Research helps us test them.

The accounts below reveal patterns worth examining. They are not product trials, customer testimonials or proof that a tool would have changed an outcome.

Our design draws on these questions alongside research on judgment and group work. Explore the evidence and its limits

01Independent perspectives

What people know, what they say and what leaders hear can differ.

02Evidence and authority

A deadline, a vote and a written warning do not settle whether the evidence is sufficient.

03Follow through and review

A concern needs someone to answer it—and a route back to the decision it affects.

Documented historical cases

Follow the evidence through the decision.

Official investigations, participants’ accounts and original research are linked in each case. Our interpretation is identified separately.

01

The picture reaching the decision

What people know, what they say and what leaders hear can differ.

  1. 01

    Bay of Pigs → Cuban Missile Crisis · 1961 – 1962

    The same leaders, a different way of deciding

    Eighteen months apart, one administration faced two crises over Cuba. The record shows how the way it reached decisions changed between them.

    Process lens Prepare → Proposals

    Read the case
  2. 02

    Nokia · 2007 – 2013

    The threat was seen. The news did not travel.

    Nokia led the phone industry when the smartphone arrived. Research with its own managers examined how fear shaped the information that reached senior leaders.

    Process lens Prepare → Proposals

    Read the case
02

The conditions for proceeding

A deadline, a vote and a written warning do not settle whether the evidence is sufficient.

  1. 03

    HealthCare.gov · 2011 – 2013

    The warnings were written down. The date did not move.

    Before the 2013 launch, the agency building the federal marketplace received repeated written warnings. The inspector general found that delay was never formally discussed.

    Process lens Decide → Accept

    Read the case
  2. 04

    Aduhelm at the FDA · 2020 – 2024

    The evidence was disputed. The route to approval changed.

    An Alzheimer’s drug was approved after the FDA’s outside experts voted against the main evidence for it. The agency’s own records show where the disagreement went.

    Process lens Proposals → Decide

    Read the case
  3. 06

    Frances Kelsey at the FDA · 1960 – 1962

    One reviewer kept asking for the evidence

    A newly arrived FDA medical officer found the application for a popular European sedative incomplete, and kept asking for evidence for eighteen months. That application was withdrawn without approval.

    Process lens Proposals → Decide

    Read the case
03

The signal coming back from the work

A concern needs someone to answer it—and a route back to the decision it affects.

  1. 05

    Mars Climate Orbiter · 1998 – 1999

    The discrepancy was noticed. It stayed informal.

    A spacecraft was lost at Mars because one file used the wrong units. The investigation board looked past the arithmetic to how the warning signs were handled.

    Process lens Follow through → Review

    Read the case
  2. 07

    Toyota andon at NUMMI · 1984 onward

    An obligation to stop the line

    When Toyota and General Motors reopened a closed California plant, the workforce largely returned. What changed was who was expected to raise a problem, and what happened when they did.

    Process lens Follow through

    Read the case

Fictional worked examples

One goal. Several questions.
Each advances when it is ready.

See independent preparation, proposals, human decisions and accepted work in four different settings. These illustrate the method; they are not customer results.

01 · Manufacturing

A supplier changes. Can production proceed?

Qualification and customer planning advance while production release stays open.

Evidence · readiness · release authority

Follow the example

02 · Customer onboarding

The sale is signed. Can delivery begin?

Scope, capacity and customer responsibilities become a phased plan people can accept.

Scope · capacity · accepted work

Follow the example

03 · Life-science evaluation

Will this evaluation justify a purchase?

Agree the test, the support and the evidence to review before a commercial decision.

Criteria · resources · purchase decision

Follow the example

04 · A family goal

A shared goal. What can each person carry?

Turn a shared ambition into voluntary commitments, with room for different needs.

Choice · flexibility · responsibility

Follow the example

In the business examples, two questions advance into accepted work while another stays open. Beginning an investigation is distinct from authorizing production, a launch or a purchase.

The Third Field · Edition 2

The deeper inquiry
continues in the book.

The expanded hardcover is in preparation. It will develop the research, its limits and the worked cases—including manufacturing, customer onboarding and life-science evaluation—in greater depth.

Explore the book and the forthcoming edition