Working method

Begin with the problem

Some questions do not fit neatly into one discipline. We follow the question and draw on the research it needs.

Explanatory alternatives matter.

A hypothesis must be compared with simpler or already established accounts.

Predictions matter more than descriptive resemblance.

Shared terminology or similar behaviour does not establish a shared mechanism.

Negative evidence matters.

An experiment that fails to distinguish the proposed mechanism from an alternative must not be presented as confirmation.

Collaboration is substantive.

Contributors should be able to change the question, evidence, interpretation or experimental design.

Research status must be explicit.

Distinguish conceptual work, exploratory experiments, verified observations and independently established results.

What we share

Public pages show our questions, published work and released results. Detailed working material remains private until there is a reason to publish it.

We share unfinished work with contributors where relevant. Confidential material stays private, and attribution and publication permissions are agreed together.