Should all authors on a paper be comfortable explaining every aspect of the paper?

Question

Say I am person A who has built up some degree of expertise in mathematical topic X. Person B works on mathematical topic Y and realizes she needs a result from topic X. She asks me if I know how to prove this result. I work for a bit and prove this result, which person B works into their paper.

Now, I'm confident about my work on topic X, and am confident that person B is not manhandling its use for topic Y - i.e., they're very clear about the assumptions my result uses and what it implies. However, my knowledge of topic Y is not very deep, and it would take a significant amount of time for me to comfortably grasp exactly how the paper's use of topic Y works.

Is it then dishonest to proceed as an author? Is it expected that every author knows every aspect of the paper?

Answer 1

By most standards, authors of a scientific paper need to have made substantial intellectual contributions to the paper (e.g., coming up with the idea behind it, proving a required result, etc), made contributions to the write-up of the paper, and ultimately approve of the paper draft when it is submitted to a journal. As such, I think both you and your collaborator meet authorship criteria even though your grasp of topic Y is not as deep as your collaborator (and presumably your collaborator's grasp of topic X is not as deep as yours). At the very least, it is not dishonest to proceed as an author.

As for whether an author should be able to explain every aspect of a paper, I would think not necessarily. My field is very interdisciplinary, and it is not uncommon for clinicians to work with statisticians on papers. They are all experts in their respective fields, but their knowledge of the other person's field is almost guaranteed to be less than their collaborators. Yet, it is still possible for them to write papers together because their knowledge bases are complementary.

Answer 2

Should all authors on a paper be comfortable explaining every aspect of the paper?

No. And if you replace the word "should" by "can" then the answer would still be (for some collaborations): No.

Quite often the reason papers get written by multiple authors in the first place is that two or more people pooling together their expertise are able to say something more novel and profound than a single author. That means, almost tautologically, that each of the collaborators possesses some knowledge and expertise that the other does not. Sometimes this knowledge gap shrinks during the writing of the paper as one or both of the collaborators go to the trouble of learning and internalizing the ideas of their coauthors, but often the gap is so large that this is simply not possible. So if you don't feel like you understand everything that your coauthors are doing or saying in your joint papers, don't worry too much about it. Of course, you do need to have some basic level of trust in them to do a competent job, otherwise you may not want to put your scientific credibility on the line by being a coauthor on a paper containing their ideas that you don't fully understand.

Personally, several of my coauthored papers contain some ideas that I don't fully understand and wouldn't necessarily feel comfortable explaining. I see this as normal, unavoidable and not dishonest in any way.

Answer 3

Authors of a paper, at least in my field (mathematical physics), are not always capable of explaining every aspect of the paper. There is a notable paper that sticks out in my mind which presents a rather tour de force calculation by one of the authors but it is responsible for a tiny result necessary for a much larger calculation spanning the entire paper.

However, this particular co-author is known among physicists for his prowess in this specific type of calculation and was specifically sought after to perform it - he did not make other contributions and though probably aware of other content of the paper, his work could be done in isolation.

This provides an example demonstrating that authors needn't be comfortable with the entirety of a paper and simply being listed as a co-author does not imply contributions that span the entire paper.

Answer 4

It is simply not realistic to expect this. For instance, the paper on detection of gravitational waves published last year has 100s of authors, some of which specialize in instrumentation, some of which specialize in numerical relativity, many of which fall in-between those expertise. It is inconceivable that people who designed and ran the detectors can have a deep knowledge of the kind of signal resulting from binary black holes collision.

Answer 5

On the contrary, every author needs to know when to be honest about the limits of their knowledge, and to pass the question on as appropriate (assuming a question is asked just as an example). An overview understanding of fellow authors' contributions is wise, especially for the lead author or one who presents the work.

A corollary could be that if the work based on two distinct fields is being presented to an audience made up of members of one of those fields, the author from that field would be best placed to present it.

Answer 6

Think about the famous Large Hadron Collider paper which has more than 5,000 authors. Should every single member of the research team be able to explain every single part of the research process? Of course not! That is literally beyong the capabilities of any human being.

So don't worry and proceed with your research.