What does first authorship really mean?

Question

In my field (theoretical computer science), authors of any paper are always listed alphabetically; our papers don't have "first authors". (Well... hardly ever.) In most other disciplines, at least within science and engineering, the ordering of authors is a signal about their relative contributions to the paper, with the first author indicating the most significant contributor. Hiring and promotion committees do give extra weight to "first-author papers" (and sometimes have to be reminded that not all areas have them). As an outsider, I find this practice confusing.

What does first authorship actually mean in your discipline? I understand vaguely that the first author is supposed to be the one who "did the most work", but what counts as "work" in this comparison? Does "most" mean "more than all the other coauthors together" or just "more than any other coauthor"? What happens when the comparison is unclear? How often is "did the most work" the actual truth, versus a cover story for a more complex political decision?

I realize that the precise answer is different for every paper. I'm looking for general guidelines for how an outsider (like me) should interpret first authorship in your field. Pointers to guidelines from journals or professional societies would be especially helpful.

Please give only one answer per discipline.

Answer 1

Pure Mathematics: All authors are assumed to have contributed equally and are listed alphabetically. The American Mathematical Society has put out a statement to this effect.

Answer 2

Cognitive neuroscience. The first author (most of the time a PhD student or a post-doc) is typically involved in designing the experiment, running it, analyzing data and writing up. The other authors are mostly involved in some but not all of these steps. They will usually help out but not do all of the work (e.g. they might show the first author how to do some analyses, or they might make many useful comments on a draft of the manuscript). In the institute where I'm studying (in The Netherlands), all the papers where I am first author will also be chapters in my thesis, and all the unpublished chapters in my thesis could potentially become papers where I am first author. Those where I am only partly involved will be a chapter in someone else's thesis, and I will not be first author.

The last author is as important as the first one. It is typically the supervisor, and ideally the supervisor is heavily involved. In labs that grow too big, a post-doc (once trained by the supervisor) might take over this role, but the last authorship still goes to the supervisor. It's like a brand name, it tells you whose lab the work is coming from. If you know a bit about the field, you will know the general ideas the paper will revolve around. If two supervisors are involved, they have to work out whose name will be last. I know of one situation where the two believed in different outcomes of the experiment, and decided beforehand that the person who turns out to be right will get last authorship. Most of the time, though, the decision is based on who did more supervision, which is ideally agreed on beforehand.

Answer 3

Applied mathematics. The first author is usually the one who contributed the most. However, sometimes the pure mathematics convention of alphabetical ordering is used; this may be expressly declared in a footnote. There are no official guidelines from SIAM.

There is no significance to being the last author, and only those who contribute substantially are listed as authors. If the supervisor is not directly involved in performing the research and writing the paper, he is typically only listed in the acknowledgments.

Answer 4

Computer Science This really depends on the institution and the group. In one group I've been working in the ordering was always alphabetical and doing otherwise would have been considered impolite. In another group, the PhD-first-boss-last principle was used.

Answer 5

Computer Science. Computer science varies by field:

• Theoretical computer science generally follows the same conventions as mathematics: the ordering of authors is alphabetical. Cryptography follows the same conventions.

• In computer systems (e.g., operating systems, databases, computer security, etc.) and other applied fields, the order of the authors is significant. The authors are often listed in order of decreasing contribution; faculty or senior folks are typically listed last. The first author often has led the design, implementation, and experiments presented in the paper or has contributed the msot to these elements. Other authors may have contributed more in total, or even individually, to these components, but sometimes at the direction of the lead author. The lead author may also have been considered to be primarily responsible for the writing of the paper, though not always.

  In cases where the lead is shared between several people, papers can have multiple "first" authors, listed alphabetically, followed by an alphabetical listing of the other junior authors, followed by the senior authors. I have seen this fact listed explicitly on a CV. Usually, a PI comes last even if he or she provided the bulk of the leadership of the project; a PI coming first is indicative of an unusually high level of contribution from the PI / low level of contribution from the junior authors.

  Overall, the meaning of first authorship ends up being vague enough that usually you have to explain the level of contribution explicitly in reference letters as such.

• HCI (human-computer interaction) follows similar conventions to those in computer systems. The authors are listed in decreasing order of contribution. The first author is generally the person who both had the "main idea" and led the effort to ensure that the efforts to carry out the research and write the paper occurred properly. The authors are generally then decreasing in order of their contribution.

Answer 6

Engineering: A first author is usually the lead student or worker on the particular project from which the paper originates. If there are multiple people working on a common project, then the authorship goes to the person whose results are most prominently featured, and who has done the most work in preparing the manuscript for publication.

A significant exception might be in multi-part papers, in which the first authorships may be shared among different people to recognize equality of contributions throughout the combined work.

Answer 7

Microbiology: Similar to cognitive neuroscience: Ph.D student is first author by virtue of having done most of the work, and the PI is the last author. If it is agreed upon that more than one person did "first-author-level" work, then the authors are listed alphabetically with a footnote noting this fact on the title page.

Answer 8

Chemical Engineering:

First author is generally considered as the main contributor. In case there are multiple students who made equal contributions, then it is specified as such in the list of authors. (Mostly by adding an asterisk on the names and a footnote explaining the asterisk). Certain groups follow policy of Adviser first and then rest. Though it is considered as arrogant (This is prevalent in mostly Chemistry related sub fields).

In Machine Learning/Applied Computer Science the policy is again similar as Chemical Engineering with Student first and Adviser last, if there are multiple advisers then the advisers tend to rotate between different papers from the same project.

Finally in Medicine especially in General Medical journals, There is a detailed statement of contributions.

E.g. Author Contributions: Dr De Wals had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: De Wals, Deceuninck, Toth, Boulianne, Landry, De Serres. Acquisition of data: Deceuninck, Toth, Boulianne.

Analysis and interpretation of data: De Wals, Deceuninck, Brunet, Boucher, De Serres.

Drafting of the manuscript: De Wals, Deceuninck.

Critical revision of the manuscript for important intellectual content: De Wals, Deceuninck, Toth, Boulianne, Brunet, Boucher, Landry, De Serres.

Statistical analysis: Deceuninck. Obtained funding: De Wals, Boulianne, De Serres. Administrative, technical, or material support: De Wals, Deceuninck, Toth, Boulianne, Landry.

Study supervision: De Wals, De Serres.

Another important point is how are paper cited, from what I remember esp. in chemical engineering. A paper is generally mentioned as Last_name et. al. and if there are only two authors or two equally contributing authors then it is mentioned as Last_name_1 & Last_name_2 et. al or just Last_name_1 & Last_name_2.

Answer 9

Medicine:

The first author is the author. He is credited with the bulk of the work, and some even consider first authorship to be the only authorship of value. This may be partly due to the fact that a medicinal paper often has many authors, with some having done next to nothing for the paper (maybe read it). Although journals would like to discourage this, people write their colleagues' names on their papers, so maybe their colleagues will do the same for them and both get a more impressive publication record.

Sometimes you see asterisks above the first two authors' names, indicating that "both authors contributed equally", although it seems to me that this is in general not well recognized. I've been told journals want one main author. Also, for many academic positions a given number of publications is required, with some minimum number first authorships.

Answer 10

Chemistry: similar to cognitive neuroscience and microbiology and many other fields - the first author is usually the individual who put most of the labor into the work. The PI, usually the last author, may have come up with the idea, but the first author usually does most of the following work: designing the experiments, synthesizing and purifying the compounds, collecting and analyzing the data, and writing the paper. The other authors might be: A student in a collaborating group that conducts an important, specialized experiment for the first author, a junior student in group who prepared some of the intermediates and collected routine data to help the first author and to learn the workings of the group, or a consulting professor offering expertise in an area that the first author and PI are weak in. The last author is usually the PI.

Some journals are beginning to ask for specific descriptions of the contributions of each author to combat vanity authorships. You also occasionally see the note that two or more authors may have contributed equally, but some journals discourage this practice also. As an example of this sort of declaration, the following statement was attached to the final article from my thesis:

Author Contributions

B.N.N. and S.Z. were coequal in their contributions and should both be considered first authors. B.N.N., T.Y.M., and G.R.H. proposed the project and designed the experiments. B.N.N., S.Z., J.T.A., and P.C.M. performed the synthesis and characterization. C.M.C. and G.R.H carried out the calculations. B.N.N., S.Z., T.Y.M., and G.R.H. assembled the data and wrote the manuscript.

Historically, the practice of putting the PI last is relatively recent. The PI used to be listed first, so that the PI was easier to identify and the collective works of the PI were easier to find in printed catalogs systems (where articles wee often indexed by the first listed author only). The historic order would have thus been: PI, first author, second author, etc. The advent of electronic databases removes the need for the PI to go first, though there are some who still do it that way.

Answer 11

Epidemiology: First author generally means the author who did the bulk of the writing, and is likely directly responsible for the analysis of the data. The last author is (often but not always) the project's PI, a senior member if its a multi-site collaboration, or a place where someone who contributed heavily in some aspect, but not as much as the first author, goes.

Generally speaking, first authorship is considered the most important, last authorship has some benefit in terms of establishing a mentor role or the concept of the author as a senior researcher.

The exception for this is a small number of "pairs" of methodologists who tend to write papers together, which end up getting seen as a kind of equal contributors.

Answer 12

In the computer science/software engineering and human-computer interaction, the first author is generally the person who both had the "main idea" and led the effort to ensure that the efforts to carry out the research and write the paper occurred properly. The authors are generally then decreasing in order of their contribution. There is generally no consistent policy of putting the PI last in SE or HCI.

Thus, the policy is very similar to the Applied Mathematics answer posted above and quite different from the theoretical Computer Science areas.

Answer 13

Earth Sciences (Physical Geography, Geology etc.). Authors are listed according to their intellectual contribution to a paper. The first author named on the paper is thus the person who has contributed intelectually the most to the paper. The second, third etc, names have decreasing importance (contribution). If more than one person can be considered first author, those names are listed alphabetically and a note to the fact is made in the acknowledgement.

Only persons who have contributed intellectually to the paper are included. Lab assistants, techncians etc, are thus not included (although it still happens).

If a leading scientist, project leader etc. is not first author the lead role may be indicated by refering to that person as "Corresponding Author". This is common when first authors are junior contributors (students).

Answer 14

From what I've seen:

If the order is not alphabetical, I assume there is a reason and the most probable reason is the work importance.

If the order is alphabetical, I doesn't assume there is a reason (As work order could or couldn't match work order). But I try to make it clear.

You can usually get importants hints with informations about authors:

If one works for an applied math lab and the other is a doctor you can guess who got the idea and who coded the method. It usually also works when one is working for a private company and the other for a public one.

Keep in mind this is not an exact science. For exemple, when there is a hierachical order between two people, you can't conclude the chief had the idea and the subaltern applied it. This could be a manager/scientist relation.

You can also find hints in the text (depending on the language), for exemple in france the article could be written with "I" or "WE".

My opinion is that a good article shoold be clear about that. My advice would be that you should take a look at every author's resume and list of publications, you will have an idea of the quality and importance of the article, the competence of the author, other articles of the author on the topic, the importance of the question adressed in the article, the importance of other relative questions... etc After this (quick) look you should be able to answer the order of work question. If it is still not clear and the order of work matters, the only solution I see is to ask.