I have an accepted paper which is now in the proof stage. Throughout the paper we define various acronyms, since we use these terms lots of times. For example:

These two acronyms are introduced in the first section:

A Bayesian network (BN) is a well known machine learning technique for (...) The conditional probability distribution (CPT) requires the specification (...)

Afterwards we generally used these acronyms, but not always. For example, at the start of a new section. We felt it is better to start the section with

Bayesian networks were introduced by (...)

than with

BNs were introduced by (...)

I think it is much easier to read the first way. Afterwards we do use the previously defined acronyms.

They have even changed these in the section titles. For example, a section is called

Estimating the Conditional Probability Tables of a Bayesian Network.

They have changed this to

Estimating the CPTs of a BN.

To me this does not look right. The reader who just wants to quickly grasp the contents of the paper will likely be confused.

Is there any convention about this? Do you think I am right to suggest the complete titles?

  • 4
    I agree with you and you can certainly suggest it. But ultimately, their journal = their style.
    – user9482
    Commented Jun 15, 2020 at 11:59
  • 3
    I agree with you. If the journal insists on using the acronyms once you've introduced them, I'd suggest that you consider not introducing the acronyms at all. (Your paper would get a little longer, but perhaps careful rephrasing could mitigate that effect.) In my experience, though, journals have usually reverted copy editors' changes if I objected to them. Commented Jun 15, 2020 at 15:44
  • I also agree with you, and I would go even further and write "Bayesian networks (BNs) were introduced by" at the beginning of a section rather than "Bayesian networks were introduced by". I'd also probably leave off the 's' on "BNs" in the revised version I gave, but my guess is that an editor will put it there no matter what you say or do. Commented Jun 15, 2020 at 17:08
  • I also agree with you, and you can also argue about readability and reader (un)familiarity with acronyms. If the suggested acronyms are not common knowledge in that area (e.g. PDF, CDF, MCMC etc, even ML for Machine Learning is debatable) you could ask they be avoided.
    – user117109
    Commented Jun 15, 2020 at 17:54
  • 5
    My cynical interpretation would be that it's 'just' a lazy find/replace on the part of the journal after interpreting your deliberate choices as an inconsistency!
    – jovisg
    Commented Jun 15, 2020 at 19:51

2 Answers 2


You should not assume that all changes introduced by the typesetters are the result of enforcing some house style. You should also not assume that these people are particularly competent. If they modify your text in some nonsensical way, you have two choices:

  • Page 7, line 15: Replace "Estimating the CPTs of a BN" by "Estimating the Conditional Probability Tables of a Bayesian Network".

  • Page 7, line 15: Please replace "Estimating the CPTs of a BN" by "Estimating the Conditional Probability Tables of a Bayesian Network".

Personally, I'd choose the first variant. This is not a request, this is an order. If they don't comply, escalate to their supervisor.

  • I think the copyeditors usually do not tell you their supervisor's contact information. Commented Jun 16, 2020 at 0:25
  • 1
    I kind of agree with @AnonymousPhysicist but for a slightly different reason which is that you're probably already doing your correspondence with the (academic) editor. They may not be a supervisor of a copyeditor/typesetter, per se, but they are the right person to work through if you have disagreements with copyeditors.
    – Bryan Krause
    Commented Jun 16, 2020 at 1:10
  • 4
    Please is polite. You also say "please" in a restaurant despite it's you that pay for your "order". It does not change the fact that you as author can insist on the paper being written in a readable way. Commented Jun 16, 2020 at 8:30
  • @CaptainEmacs I'm German. (Yes, I know that the English-speaking world uses "please" more liberally.)
    – Uwe
    Commented Jun 16, 2020 at 9:48

It could be that they find your titles overly long and verbose (and they may have a point - they do this work for a living). But your point about giving people a fresh start in a new major section is valid too.

An in-between option would be identifying the major/minor terms in your titles and using acronyms only for the minor terms. For example:

Estimating the CPTs of a Bayesian Network

If you're already thinking about Bayesian Networks (and know what they are) then "CPTs" is easily recognized. More easily than

Estimating the CPTs of a BN

But it's less ponderous than

Estimating the Conditional Probability Tables of a Bayesian Network

It's a matter of style, and since it's your text in their journal, you should both have a say in it. Aesthetics and clarity should be balanced against each other as well. Treat this as a negotiating position, and consider the question from both viewpoints.

  • 7
    "they do this work for a living" From several experiences with copyeditors (which, as it seems, are often cheap outsourced labor forces), I can assure that this doesn't mean much in this context. Commented Jun 16, 2020 at 8:01

You must log in to answer this question.

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