What are appropriate steps to take when the corresponding author of a paper one finds very interesting, and potentially very useful, but slightly puzzling (due to the enforced brevity of the publication format) in its nuts and bolts implementation doesn't correspond? (I do not mean, 'Does not provide instant and copiously detailed answers,' I mean, 'Complete radio silence.')

Clearly, one does not send e-mail after e-mail after e-mail. That is both rude and insane. (Expecting different results from the same action.)

Clearly, one does survey that author's other literature (including past theses or dissertations, often the best place to find more verbose treatments of early papers) to see if the nuts and bolts questions are resolved there; one also surveys the literature of the non-corresponding co-authors for the same reason. (In fact, one does this, in my opinion, before contacting anyone in the first place.)

Clearly one does verify that the e-mail is up to date, to the best of one's ability.

When these options fail, and one is still extremely interested in a working implementation of a tersely published technique, are there any other reasonable strategies to apply? Reasonable, as in, will not cause terrible reputation in the field.

  • 2
    Contact the co-authors starting at the upper-left end of the author list and working to the right and down? If there is only one author, you may be out of luck.
    – Bill Barth
    Jul 8, 2015 at 0:09

1 Answer 1


Authors, corresponding or otherwise, geneally do not respond for one of two reasons. The first is they miss the email. This could either be because of a wrong address, spam filtering, or a huge email backlog on their part. The other is that they do not think it is worth their while to respond. They could be so busy, that no matter how simple of a request, that they just are not going to respond.

Providing an explanation on how the nuts and bolts implementation works, is not an easy task. Providing a working implementation, can be even harder. Your want to make your request as easy as possible for them to (1) understand, and (2) address/answer. This means ideally asking yes/no questions where you provide enough background so they can see where you are coming from.

Not the answer you're looking for? Browse other questions tagged .