The Research Doctrine

Research at Anabasis exists to raise the floor and extend the ceiling of human capability. It is not marketing, not thought leadership, and not content. If a piece of work would be embarrassing to submit to an unfriendly examiner, it is not research.

We publish because publication is the discipline that produces the work, not because publication is the goal.

A claim advanced by Anabasis is a claim we are willing to defend. Evidence is cited, methodology is described, and the strength of the finding is stated with the same precision as the finding itself.

The absence of evidence is stated as clearly as its presence. We prefer a well-mapped ignorance to a decorated guess.

Data is not adjusted to fit a narrative. Hypotheses that failed are reported alongside those that succeeded. Conflicts of interest are disclosed in the same document as the finding, not in a linked footer.

A single act of intellectual dishonesty extinguishes the credibility of a decade of honest work. We proceed accordingly.

Where methods can be released, they are. Where data can be released, it is. Where either cannot, the reason is stated.

A reader should be able to reconstruct our argument from the materials we provide. If they cannot, we have not finished the work.

Nothing of significance is published without external review by a reader who is qualified to disagree. The reviewer is named where the reviewer permits.

The purpose of review is not endorsement. It is the refinement of the claim under adversarial reading.

Anabasis publishes on the timescale of the question, not the timescale of the platform. A finding that requires a decade to establish will be given a decade.

We regard the disappearing publication — the post that was true this quarter and gone the next — as the opposite of research.