Agreeing authorship before the first result

Why this needs explicit discussion

Authorship conventions differ across fields and across research cultures — the meaning of first, last and corresponding positions is not universal, and neither are expectations about who is included. In a collaboration spanning two systems, two sets of unstated assumptions meet, and neither side realises the other holds a different one until a draft exists.

By then, positions have become personal. Before the work starts, they are procedural.

What to agree, and when

At the start of the project, write down:

1. Which journal authorship criteria you will apply. Most fields have published criteria; naming one removes most ambiguity.
2. What each position means in the papers you expect to produce.
3. Who is provisionally in which position for the anticipated outputs, and on what basis.
4. How this will be revised if contributions change — because they will.
5. Who decides if there is disagreement.
6. How students are handled. A student who does the substantive work should normally be first author on the paper from that work.
7. Acknowledgements for contributions that do not meet authorship criteria.

One page is enough. Revisit it when the first draft is planned.

Contributions that get under-counted

Some kinds of work are systematically under-recognised, and they are often the ones contributed by the local partner in an international collaboration:

• Access to a field site, a population, or an institutional setting.
• Data collection, including the substantial work of obtaining permissions and building local trust.
• Contextual interpretation — knowing why a result means something different here.
• Translation and adaptation of instruments, which is intellectual work, not clerical work.
• Navigating local ethics and regulatory approvals.

Whether these meet authorship criteria depends on the criteria you adopted. What is not acceptable is treating them as services that need no recognition at all.

The pattern to avoid

A collaboration in which one side consistently designs, analyses and writes while the other consistently collects and facilitates produces a predictable authorship pattern — and a partnership that does not reach a second project. If that is the shape of the work, the honest response is to change the shape: involve the local partner in design and analysis from the start, including training if needed.

This is not only an ethical point. Design decisions made without local knowledge are frequently wrong in ways that only become apparent during data collection.

Practical mechanics

• Use a contribution statement. Many journals require one; use it even where they do not. It makes each person's role explicit and reduces disputes.
• Keep a shared document recording who did what as the project proceeds. Reconstructing this a year later is unreliable.
• Circulate the author list with the first full draft, explicitly, and ask for agreement in writing.
• Every author must see and approve the final version. Adding someone's name without their knowledge is a serious breach, even when meant as a courtesy.

If a dispute arises

1. Return to the written understanding and the journal criteria — discuss against a standard, not against each other.
2. Involve a neutral senior colleague if direct discussion stalls.
3. Use the research integrity office of one or both institutions if it does not resolve.
4. Do not submit the manuscript while authorship is unresolved. Journals treat authorship disputes seriously and a submission made under dispute creates a far worse problem.

Note: requirements, procedures and institutional practices described here vary between countries, institutions and journals, and change over time. Always check the current rules of the specific institution or publisher you are dealing with.