Co-Authorship

Co-authorship exists whenever two or more people jointly meet the criteria for authorship on the same output. It does not add a new definitional test on top of authorship itself — every co-author still has to independently satisfy whatever authorship standard applies (most commonly the ICMJE’s four-part test: substantial contribution to the work; drafting or critically reviewing it; final approval of the version to be published; and agreement to be accountable for it). What co-authorship adds is a second, separate layer of practice: how a group of qualifying authors organizes itself — who is listed first and last, how equal contributions are signaled, who serves as corresponding author, and how disputes over any of that get resolved.

Co-authorship vs. sole authorship

A sole-authored work has one person who meets the authorship criteria and no allocation questions to resolve. A co-authored work has two or more, which immediately raises questions a sole-authored paper never has to answer: in what order do the names appear, does order signal relative contribution or something else (alphabetical convention, seniority, chronology of joining the project), and who takes on the administrative and accountability responsibilities that a single author would otherwise hold alone. None of this changes who qualifies as an author — it only governs how qualifying authors are presented and how responsibility is distributed among them.

Author order: what governs it

Author order is a genuinely underspecified area by design. The ICMJE’s own recommendations state directly that “the criteria used to determine the order in which authors are listed on the byline may vary, and are to be decided collectively by the author group and not by editors” — ICMJE deliberately does not prescribe an ordering scheme, and journal editors are not supposed to adjudicate it. In practice, several conventions coexist across fields:

• Contribution-ranked order — common in biomedicine and the life sciences: first author is typically the person who did the most hands-on work (often the lead researcher or trainee), and the last author position is often reserved for the senior investigator or lab head who supervised the work, sometimes distinguished as the corresponding author.
• Alphabetical order — the norm in mathematics and economics, and common in some areas of physics, on the reasoning that co-authors contributed as equals and that ranking them would misrepresent that.
• Chronological or task-based order — used in some engineering and computer-science subfields, ordering by when a contributor joined the project or by the specific task they led.

None of these is a formal standard imposed externally; each is a disciplinary norm the author group is expected to know and apply, and to be able to explain if asked. Because CASRAI’s Dictionary is coordinated with the CRediT taxonomy work at NISO, it’s worth being explicit that CRediT (ANSI/NISO Z39.104-2022) does not solve the ordering question either — it documents what each named contributor did, using 14 standardized roles, but a CRediT statement is commonly published alongside byline order, not as a substitute for it. See CASRAI’s dictionary entry on the “degree of contribution” qualifier for how CRediT layers lead/equal/supporting distinctions onto individual roles.

Signaling equal contribution

When two or more co-authors did comparably substantial work and the group does not want the byline order alone to imply a ranking, the common convention is an explicit equal-contribution note — typically a footnote or statement such as “these authors contributed equally to this work,” attached to the relevant names, sometimes reinforced with a shared symbol (an asterisk or dagger) next to each name. This is a publisher- and journal-level convention rather than a single universal rule; authors should check the specific journal’s instructions for how it wants equal contribution formatted, and some journals cap how many authors can be designated as joint first (or joint senior/last) authors. An equal-contribution note addresses order, not the underlying authorship test — every named person still has to independently meet the full authorship criteria regardless of where they sit in the list.

Co-corresponding authors

A related but distinct co-authorship convention is naming more than one corresponding author — the person (or people) responsible for handling correspondence about the work after publication, including data requests, corrections, and retraction inquiries. See CASRAI’s dedicated entry on the co-corresponding author convention for when this is used and what responsibilities it actually carries; it should not be granted as an honorific, since it comes with real, ongoing obligations.

Cross-institutional co-authorship

Co-authorship increasingly spans institutions, and sometimes countries — a collaboration between a university lab, an industry partner, and a clinical site, for example. This introduces considerations that single-institution authorship rarely faces:

• Affiliation listing. Each author’s institutional affiliation is listed independently under their name (and increasingly, alongside a persistent identifier such as an ORCID iD and, at the institution level, a ROR identifier, where the journal or repository supports it) — co-authorship does not require a shared institution or a shared funder.
• Who can mediate a dispute. When co-authors disagree about order, inclusion, or credit, COPE’s own guidance is explicit that journal editors are not positioned to adjudicate an authorship dispute; COPE instead points toward the authors’ home institution(s). When co-authors are affiliated with different institutions, this can mean more than one institutional research-integrity office is involved, which is a genuine practical complication cross-institutional collaborations should plan for before it becomes a live dispute, not after.
• Data and material ownership. Cross-institutional co-authorship often intersects with data-sharing and material-transfer arrangements that are governed separately from authorship itself — agreeing on authorship does not by itself resolve who controls or can redistribute the underlying data or materials after publication.

When co-authorship goes wrong

Two recognized integrity failures sit at the co-authorship boundary specifically: gift (or honorary) authorship, where someone is added to the byline without meeting the authorship criteria (often as a courtesy to a senior colleague or to strengthen a paper’s perceived credibility), and ghost authorship, where someone who did qualifying work is left off the byline entirely (historically associated with undisclosed industry-sponsored medical writing). Disagreements over author order or inclusion that don’t rise to either of those are still common enough that CASRAI’s Dictionary tracks them separately — see authorship dispute for how COPE frames resolution when co-authors can’t agree.

Worked examples

Example 1. A five-person biomedical research team publishes a clinical study. The two graduate researchers who ran the experiments and analyzed the data are listed first and second, marked with a footnote stating they contributed equally; the principal investigator, who designed the study and supervised the work, is listed last and designated as corresponding author. All five meet the ICMJE’s four criteria independently; the order reflects the lab’s contribution-ranked convention, not a separate authorship test.

Example 2. Three mathematicians at three different universities co-author a proof. Because their field’s convention is alphabetical order regardless of relative contribution, the byline lists them by surname with no equal-contribution note needed — the convention itself already communicates that order does not imply ranking. Each author’s own institution is listed under their name; no single institution is treated as the paper’s “home.”

Counter-example. A department head is added to a manuscript’s byline because it is customary in the lab to include the head on every paper it produces, despite not having contributed to conception, drafting, or review of this specific work and not having reviewed the final submitted version. This is not co-authorship — the department head does not meet the authorship criteria at all, regardless of position in the author list, and including them is gift authorship, a recognized integrity violation rather than a variant of legitimate co-authorship practice.