What CRediT is
The Contributor Roles Taxonomy is a standard vocabulary of fourteen roles describing contributions to scholarly output. It covers conceptualisation, methodology, software, validation, formal analysis, investigation, resources, data curation, writing (original draft), writing (review and editing), visualisation, supervision, project administration and funding acquisition.
Each author is assigned the roles that genuinely apply, and multiple authors can share a role. The output is a short statement naming each author and their contributions.
It exists because author order carries different meanings in different fields and communicates almost nothing reliably. CRediT states the contribution instead of encoding it in a position.
Author order means different things in different fields. CRediT states the contribution rather than encoding it in a position.
Writing one
Go role by role rather than author by author. For each of the fourteen, ask who genuinely did it — this is faster and produces a more honest result than asking each author what they contributed.
Be accurate rather than generous. If someone did not write any of the draft, do not assign them the writing role. The value of the statement lies entirely in its accuracy, and a statement assigning everyone everything conveys nothing.
The finished statement reads simply: "A.B.: Conceptualisation, Methodology, Writing — original draft. C.D.: Formal analysis, Software. E.F.: Supervision, Funding acquisition."
Do this before submission and circulate it to all authors. Disagreement is far easier to resolve at that point than after a decision letter arrives.
Why it prevents disputes
Most authorship disputes are about credit, and most become intractable because there is no record of what anyone actually did — only an order that different people interpret differently.
A contribution statement drafted at the start of a project forces the conversation early, when contributions are still prospective and nobody is defending a position. It also surfaces the mismatch between what people expect to do and what the paper will need.
It has a second benefit for early-career researchers specifically: it makes a substantial contribution visible even from a middle-author position. That is exactly the evidence a hiring panel, funder or evidence-based assessment cannot infer from a citation string.