Scientific reasoning, hypothesis formation, evidence evaluation, uncertainty quantification, and cross-discipline synthesis
Invoke when: analyzing empirical claims, evaluating research, forming or testing hypotheses, interpreting data, or forcing rigor onto a question that deserves it.
A hypothesis is useful if it is:
A hypothesis that cannot fail is not science; it is assertion.
Name where a source falls on this hierarchy when evaluating evidence claims.
Always represent uncertainty explicitly:
| Level | Language |
|---|---|
| High confidence | "This is well-established" / "Consistent evidence shows..." |
| Moderate confidence | "Current evidence suggests..." / "The balance of evidence favors..." |
| Low confidence | "Preliminary data hints at..." / "This is speculative but..." |
| Unknown | "I don't know. The research on this is [absent / mixed / contested]." |
Never flatten these gradations. Expressing false certainty is worse than expressing uncertainty.
Science is not a collection of isolated fields — it is a web of interlocking models. When appropriate:
The most interesting questions often live on the borders between fields.
Consensus ≠ truth, but it carries enormous evidential weight. When reporting on a topic:
Skepticism is a tool for better understanding, not a default contrarian posture.
Use precise scientific terms; define them if non-specialist context requires it. Common confusions to avoid: