Re: Diversity and offensive terminology in RFCs

[Ted Lemon <[e-mail]>]

On Sep 20, 2018, at 3:10 PM, Carsten Bormann <[e-mail]> wrote:
> Just in case people don’t know, I’m (very mildly) offended if someone calls me (or any other human) a “person”, except of course for those who are just that.

I think that there's a tremendous lack of clarity about what's being discussed here that is worth stating more explicitly.

The issue is not that someone might get offended.   It's impossible to never give offense.   Nobody is proposing, as far as I know, that the IETF enforce some set of rules intended to produce this impossible result.   The issue is also not "politics," unless you think the wish not to have to work in a hostile work environment is political.   What we are talking about is something different.

Consider this: there are words in the modern vernacular the use of which is actually hurtful, and not merely offensive.   The reason it's hurtful is because it causes an individual sensitized to the use of that word to revisit an ongoing oppression under which that person is suffering.   People who enjoy engaging in such oppression may use these words in a professional context in the hopes of creating suffering for their co-worker who is part of an oppressed group.   An example of this would be the use of the "n-word" or the use of the "c-word."   (If you don't know what I'm talking about, please ask me privately and not here on the list).

There are contexts in which the use of these words is acceptable, but I think most if not all people here would agree that there is never a time where these words should be used either in IETF discussions or in IETF standards.   So that's the extreme.

Then there are words for which there isn't any such association.   For example, I don't think we are in any danger of having to redact the word "packet" from any RFCs.

So those are two extremes to think of.   Now of course, the set of terms that are on the bad extreme varies from country to country and from language to language.   RFCs are in English, so we can somewhat ignore the language, although there might be cases where a perfectly fine english word means something in another language that sends it over to the bad extreme.

I think, and of course I may be wrong, that terms like "blackballed," "blacklist," "whitelist," and so on, are probably not down at the bad end of the spectrum, although I'm certainly willing to entertain the possibility that I am wrong, since I am not an affected party.   But words like "master" and "slave," particularly when used together, definitely seem to be at least toward the bad end of the continuum, and definitely not at the okay end where words like "packet" live.

So the question we ought to be talking about is whether there are words that we habitually use in the IETF that might now, or might in the future, be considered to be as abusive as the n-word and the c-word, or might be used in that way.   If we have reason to believe that such words are in use, it would be wise for us to find different words to use.   This will save us trouble in the long run.

Arguing that we ought not to have to change our use of such words because they are understood to have some non-abusive meaning misses the point; there may have been a time when discrimination against the affected population was so pervasive that there was nobody to complain about the problem, or nobody who could safely complain about it.   And for those not in the affected population, it had no loaded meaning and therefore seemed like a perfectly reasonable metaphor to use.   That pre-existing and blameless motivation does not make it okay to continue using the term now, assuming that we care to address the known problems of non-inclusiveness in our profession.