61

I am nearly finished with a paper, as well as some accompanying open-source software. We are planning on submitting this to a prominent statistics journal. One of the more promising methods that we discuss in the paper deals with the normal-inverse Gaussian distribution, commonly referred to as the NIG distribution. This is usually pronounced “en-eye-gee”, but it’s understandable that some people say it phonetically (it’s a relatively obscure distribution).

In case it is not yet clear, my concern is that NIG is somewhat reminiscent of the “N word”. I had some reservations about writing NIG so many times in the paper and in the software (by convention, there would be many functions with NIG pre-fixes, e.g. nig_foo()). I thought maybe I was overthinking things, but I’ve had a few people (including a co-author) comment about the potentially inappropriate nature of this name.

Thus I feel like it might be worthwhile to refer to this distribution by a different name. It is not uncommon for distributions in statistics to have multiple names (e.g., normal vs. Gaussian), and we would make it clear that it is the same distribution. The problem is, I feel very unqualified to just change the name of distribution that has been around for 40 years.

If we do change the name, here’s a short list of some of the things I’ve considered.

  1. The Normal-Wald Distribution (Wald is another name for inverse-Gaussian)
  2. The Gauss-Wald Distribution (Same idea)
  3. The Barndorff-Nielsen distribution (Ole Barndorff-Nielsen was one of the first researchers looking into this distribution)
  4. The Barndorff distribution (Less wordy, but I don’t want to remove a surname without asking him)

Questions

Is it a bad idea to change the name? I’m slightly uncomfortable with it, but I might be overthinking it.

If I do change the name, what’s the most reasonable choice? (I realize this is very subjective, but feedback is welcome)

I don't think that the name of the distribution is intentionally or overtly "racist". The main point is that (i) this distribution is relatively obscure and (ii) many distributions have multiple names so it might make sense to have an alternative that has no chance of being problematic.

4

12 Answers 12

85

If you are at all unsure about causing offense, you could always use the full name in your paper "Normal-Inverse-Gaussian distribution". It won't be any less readable for that.

In your code pick a different prefix that still clearly identifies the distribution, e.g. ninvg_foo().

4
  • 24
    I am accepting this answer, because I think it provides a simple non-invasive solution. The prefix NINVG seems perfectly reasonable to me. – knrumsey Dec 18 '20 at 20:02
  • 13
    The NeurIPS rebrand is another good example of extending the prefix. – Neal Fultz Dec 19 '20 at 6:36
  • 12
    If you really need to save characters, you could still have "Normal-IG" (I've definitely seen "Poisson-IG" for "Poission-Inverse-Gaussian"), similarly for code normal_ig could be a sensible variant. – Martin Modrák Dec 19 '20 at 13:31
  • 4
    For the code, I'd add that longer variables names are usually better as they lead to less ambiguity. As someone who deals with safety-critical systems, I'm often in the position of reviewing others' code or specifications. When I come across 3 (or 4, etc.) letter abbreviations, it often forces me to interrupt domain experts. Sometimes 2 (or more) different domain experts will have 2 (or more) widely different guesses as to what the abbreviations stand for. – Ben Hocking Dec 20 '20 at 11:56
106

Actually changing the name because someone might be offended by a sound that comes from a completely different domain and is completely separate from the history of racism would be condescending, in my view.

And, as you say, the term is in common use. Changing it would also confuse some people.

Leave it be.

There was no intention to do harm and the acronym is natural. The language police are unlikely to chase you down. It would be different of course if the acronym were chosen for racist purpose, but that is clearly not the case here.

2
  • Comments are not for extended discussion; this conversation has been moved to chat. Please read this FAQ before posting another comment. – Massimo Ortolano Dec 18 '20 at 19:48
  • 22
    Condenscending to whom precisely? – Kai Dec 18 '20 at 21:03
20

Change the name if you like. Point it out prominently in your paper and code, possibly more than once (e.g. first mention in the main text and in the methods / supplement, of course also in the function documentation, etc.). The names of even common objects are changed for a variety of reasons. It happens, and people should be aware that names are usually pretty arbitrary.

Example: I once read a book where the discrete-time Fourier transform (DTFT) (by all means the standard name) was called DSFT (Fourier transform in discrete spaces), because, technically, it can be computed over other domains than time. It was weird for a moment, and then I got used to it, and know I'm kind of glad that the authors stressed this point because it made me think about something important.

Conventions exist until someone begins to challenge them. I'm scared of the language police (see another answer here) doesn't even need to be part of the justification. It can just be that another name sounds better to you. Of course, standard names shouldn't be changed on a whim, but they can be (and are).

1
  • Eh, the example you give provides a domain-specific justification (avoiding confusion about the implied meaning). OP's problem doesn't really have this justification, and the perceived similarity of NIG() and that "N-word" is relatively silly and distracting to point out. – Chieron Dec 20 '20 at 10:54
15

This is a really interesting question, especially since some other terminology (e.g., master/slave replication) has somewhat recently come under heavy criticism despite not being seen as controversial for quite some time.

Ultimately, it's your choice to make as an author (or as a group of authors). Using alternative terminology is not generally seen as a reason to reject a manuscript as long as you make sufficiently clear what you mean with your terms (even if the reviewer does not agree with your reasoning for not using a more standard term). On the other hand, there is neither a practical nor (at least in my opinion) moral obligation to preemptively stop using a term despite there currently not being any substantial debate about this specific piece of terminology. That said, if one person in the author group feels strongly about this I would be tempted to go with their wishes on this one.


Unlike some other answers, I do not want to prescribe how you should decide on this matter. Personally I do not see the term as problematic, but I'm neither an English native speaker nor of African descend. However, I do want to point that (a) a term not being "meant" racist and (b) a term having being used so far are both not sufficient arguments that further usage is appropriate (as the master/slave example above nicely demonstrates). Times change, and ultimately it's the affected group (people of African descend, in that case) that need to decide whether the term is appropriate or not.

2
  • 4
    I don't think "people of African descent" describes the affected group completely or particularly accurately. – Vaelus Dec 18 '20 at 13:29
  • 6
    @Vaelus If you have a better proposal I'm happy to edit it in. – xLeitix Dec 18 '20 at 15:33
13

As suggested in some other comments and answers, it might be a good thing to spell it out more fully (easy enough to do with modern word-processors).

I think there's no motivation to pretend to determine whether anyone would be "justified" in being offended by the old acronym. The fact is that some might easily be, AND it's easy to avoid this.

(Telling people in traditionally-abused groups that they "shouldn't be offended" by things that sound like references to their traditional abuse is really not supportable...)

Yes, it's kinda too bad that scientific acronyms and terminology seem to have a bit of an obligation to acknowledge ambient social things. :) But, yeah, they do, I think.

2
  • 5
    Especially so in a discipline that has historically had a strong association with eugenics and white supremacy - cf. people like Galton, Pearson, and Fisher, who built a lot of the foundations of modern statistics in the service of abhorrent eugenic/racist ideas that helped inspire Nazi "cleansing" programs. There's enough baggage here that it seems sensible to be considerate of how language choices might come across. – Geoffrey Brent Dec 18 '20 at 22:59
  • 2
    @GeoffreyBrent, good points. Just to repeat, although I was (even more) foolish when younger, it seems completely ridiculous to me (now) to tell people that language that reminds them of their (or family's, or ...) prior abuse should just be overlooked. – paul garrett Dec 18 '20 at 23:03
10

You could always abbreviate Distribution as D and call it the NIGD. This is different enough from that other word that I think it should be acceptable.

(Now we need to do something about SMB, that inter-computer connection tool in Linux.)

3
  • 3
    Upvoted because there are too many TLAs in this world already. There is more room for FLA. – MarkHu Dec 18 '20 at 20:37
  • 7
    What's wrong with SMB? – Azor Ahai -him- Dec 19 '20 at 21:25
  • SMB is usually pronounced Samba, as in Little Black Samba, a racist symbol. I was ready to overlook it as being just a meaningless coincidence, until I saw the icon -- a little black boy. – Jennifer Dec 20 '20 at 19:24
10

One possibility that doesn't seem to have been addressed yet is the following:

Ask members of the group most likely to be offended by the term---in this case Black people---what they think.

Obviously you probably don't have the resources to do a full, representative survey of Black people, and obviously there are going to be differences in opinion between members of any group. But if you ask a few people and they say they find it very offensive, it would likely be a good idea to try and use something different. Likewise, if every Black person you ask says that they think it is fine this is an indication (though not a guarantee) that the term does not need to be changed.

Language is complex, and while there is no intention to cause harm by using the acronym, perhaps the potential to cause harm outweighs the benefits of familiarity for your readers. You won't get a definitive answer because there is none, but in my view the people whose opinion should carry the most weight on this are the people who are most likely to be harmed in the first place.

It is also worth noting that Black people are underrepresented in academia, and therefore likely to be underrepresented on academia.stackexchange, which will affect the answers you get here.

Edit: I guess its important to note as well that it may not be well recieved to ask any Black person what they think. The people you ask should be those who have previously indicated that they are happy to discuss such topics, either in person or online.

5
  • 8
    That wouldn't be really representative of everyone, would it? Other people always tell me it's a very small amount of people (<10%) that has a very big voice and much impact (90+%). You can't weight the "votes" (objectively) to take this into account. – Gizmo Dec 18 '20 at 14:00
  • 6
    If I have a medical problem, I'll weight the opinion of a medical doctor more than that of someone who is not an expert in medicine. I think people who are part of the potentially offended minority group are more likely to have relevant experience and knowledge as to whether the term is offensive. – DavidH Dec 18 '20 at 14:21
  • 2
    "It is also worth noting that Black people are underrepresented in academia, and therefore likely to be underrepresented on academia.stackexchange, which will affect the answers you get here." +1 – user347489 Dec 19 '20 at 20:55
  • The problem with this answer is that it requires research! Is it really worth starting a completely new research project for the sake of one abbreviation used in another unrelated field. Simply asking a few random people from the public won't give a reliable result. How would you select your sample? How do you know when you can reach a conclusion with a reasonable degree of certainty? What do you do with outliers? – chasly - supports Monica Dec 20 '20 at 21:48
  • @Gizmo Even it's not going to be representative of the group as a whole, it's not a bad idea to ask some of them to at least have an image of the case and its severity in your mind. – iBug Dec 21 '20 at 9:35
7

I want to add one thing that's not been said explicitly in other answers: since, as you've said, it's relatively obscure, you absolutely can change it. People use different names for the same thing all the time. The name of a thing is the name that people use, and in this case, that's you. Changing it will not confuse anyone at all, not even slightly.

The fact that you are uncomfortable with the acronym is a good enough reason by itself, and a desire to make your field more welcoming to minorities is a very good reason for this kind of thing. It's a positive thing to do and you shouldn't feel any hesitation.

6

I am French and I used for sometime a data analysis library that was full of anal_in(), anal_deep() and similar functions.

Someone added in the code

# For fuck's sake, stop using anal as a variable or in a name!!

And that was pretty much it. Some were probably offended but nobody bothered changing the names.

It is your code so if you feel that nig is offensive, just know that eel is offensive to me, and that Mike will find god horrendous as a variable name.

If you called your variables/functions deliberately insulting then it would be bad taste, or douchery, or something like that. I would not bother in your case but if that raised the need to ask the question then for your own good (as in "so that you do not think too much abou that") just change it to NOINGA and be done.

5
  • Comments are not for extended discussion; this conversation has been moved to chat. – cag51 Dec 20 '20 at 21:17
  • 3
    Mod's Note: There was some discussion (see the chat) about the expletive in the code snippet; this has now become a post on meta. Please do not add further comments about this debate here; instead, take it to the meta post (or the chat, but the meta post is what will decide whether our policy changes). – cag51 Dec 20 '20 at 21:19
  • 1
    Is "eel" actually offensive? Or is that that an argument ab asurdum? – Azor Ahai -him- Dec 27 '20 at 5:01
  • @AzorAhai-him-: yes, this was an argument ad absurdum - I typed a few random letters and only later realized that this is actually a proper word. – WoJ Dec 27 '20 at 12:19
  • 1
    @woj Oh, OK I don't know French swears – Azor Ahai -him- Dec 27 '20 at 17:49
5

my concern is that NIG is too reminiscent of the “N word”.

It is not. This is in no way comparable to the "whitelist/blacklist" terminology that is currently widely criticised.

9
  • 28
    This answer would be more useful if it expanded a bit. What is the "whitelist/blacklist" terminology that is currently widely criticised? Why is this not comparable? I may have a different filter bubble and am not aware of widespread criticism, or you may be referring to something else than I think. People reading this in other parts of the world or in 10 years may be even less aware what thing is currently widely criticised. – gerrit Dec 18 '20 at 14:11
  • 4
    It is not comparable because while we can argue over the original source of "white" and "black" in whitelist/blacklist and whether or not it does now or did in the past have any racist meaning, etc. the words actually exist - "black" and "white" are the same exact words in "blacklist" and "whitelist" as in referring to people by the color of their skin. That is not the case with this acronym being discussed - it is simply an acronym that happens to match the beginning of another word. – manassehkatz-Moving 2 Codidact Dec 18 '20 at 16:01
  • 3
    @manassehkatz-Moving2Codidact There is an argument that the term blacklist is connected with race, and there's also one against it: en.wikipedia.org/wiki/Blacklist_(computing) – Kimball Dec 18 '20 at 17:13
  • 1
    @gerrit This was very widely covered in international media for software developers. I am confident the asker and some of the other answerers knew that. – Anonymous Physicist Dec 19 '20 at 1:13
  • 3
    Downvoted b/c I believe saying "it is not comparable to..." reduces the whole premise of the question to an opinion. – user347489 Dec 19 '20 at 20:42
5

I would like to add a different perspective to your question. Fortunately (or unfortunately, it depends) for the Americans, English become the most widely used language in science and business. That means, that it is used by many people who are not Americans and who are not really troubled by some secondary/offensive meanings of some words or some cultural concepts typical for the US.

NIG is an abbreviation of a name of a distribution; it is understandable. It would be confusing to many people, why you would want to change it, just because it resembles a word which in your cultural context is not appropriate to use. I think your concerns might be valid when writing a novel, in English, for an English-speaking audience, where someone could start looking for hidden meanings. Since we are talking about a scientific paper read by non-American readers as well, I don't think your concerns are relevant. Use what is standard and don't worry.

1
  • 6
    This seems a bit like arguing "why should we worry about people being allergic to our food when most customers won't be allergic?" As for reader confusion - like OP says, it's not a well-known distribution, so odds are most readers wouldn't be familiar with the other version in the first place. – Geoffrey Brent Dec 18 '20 at 22:19
-5

Niggardly, niggling, and Niger are a few "obviously inoffensive" words that people have found offensive (source). The Mandarin word 那个 has also been found offensive; offensive enough that the teacher using it was removed from the class where it was used (source). I suggest that you use an acronym other than NIG.

9
  • 4
    Your background information does not support your conclusion. – Anonymous Physicist Dec 18 '20 at 7:52
  • 3
    listverse.com is not a credible source. While it's headline agrees with your claim, the actual text gives no evidence or reasoning that "niggardly" is not offensive. Instead it gives some anecdotal examples of people being offended. Furthermore, none of the examples you give are related in any way to the acronym "NIG." – Anonymous Physicist Dec 18 '20 at 8:05
  • 9
    "Night" and "Nigeria" also start with "nig" but "NIG" starts with "Nor" as specified in the question. None of this is relevant. – Anonymous Physicist Dec 18 '20 at 8:11
  • 5
    As every fule kno, all people named Nigel are racists. If Adrian Mole said so, it must be true. – alephzero Dec 18 '20 at 15:01
  • 3
    Wikipedia has a whole article about controversies surrounding niggardly en.wikipedia.org/wiki/Controversies_about_the_word_niggardly – Flydog57 Dec 18 '20 at 21:43

Not the answer you're looking for? Browse other questions tagged or ask your own question.