Re: [rain1 at airmail dot cc] Delete abortion joke

[Joseph Myers <joseph at codesourcery dot com>]

On Sun, 6 May 2018, Richard Stallman wrote:

> I expect that readers who don't know about the gag rule mostly will not
> relate the joke to abortion -- that they will be puzzled by the joke.
> 
> There are people in this discussion who did not know about the gag
> rule but do know that the joke relates to abortion.  My theory is that
> others told them it relates to abortion, and that's how they know.
> 
> Is that correct?
> 
> Did any of you figure out _on your own_ that the joke had to do with
> abortion, despite not knowing about the gag rule?

That would need to be asked of people who had read that part of the manual 
without first seeing this thread.

I think it's much more obvious from being in documentation for "abort" 
that it's about abortion, than that it's about a US-specific gag rule 
("Federal" is the only cue I see there to it being something US-specific, 
but is hardly unique to one country and certainly doesn't identify the 
particular rule, or that it's referring to a specific real rule at all).

I think being so US-specific - and, beyond that, relying on knowledge of a 
particular US rule - is by itself sufficient justification for not having 
this joke in the manual; that would apply equally to a joke that didn't 
involve a controversial topic such as abortion at all, but still depended 
on some very country-specific background not widely understood by people 
outside a particular country or group of countries.  Carlos has found 
evidence 
<https://stackoverflow.com/questions/48445031/why-would-it-be-illegal-to-inform-about-abort> 
<https://www.reddit.com/r/linux/comments/d4783/federal_censorship_regulations_may_restrict/> 
that this joke does confuse readers in practice.

I do not have figures, but I think it very likely that the free software 
community is much more diverse than when this joke was added in 1992.  The 
manual should be written for the users we have now, not the users we had 
in 1992.  Even if in 1992 the expected readers of the manual were a small, 
culturally homogeneous group of free software hackers, with a common 
background that included an understanding of the gag rule, and a common 
understanding of what humour is appropriate in what contexts that 
indicated that joke to be appropriate in the manual, it is for today's 
readers that we must now maintain the manual.

*Even if we were all agreed* that the joke was appropriate, even if we all 
disagreed with the attitudes in certain communities disapproving of 
certain subjects of humour in certain contexts and considered the 1992 
sense of humour objectively better, it would *still* be our duty to our 
users to put aside our own personal preferences and judge what is best 
given the users we actually have and the background knowledge and 
attitudes to subjects and context of humour that they have.  This is just 
the same as the need to keep the technical content of the manual up to 
date with the context of today (for example, the correct emphasis for 
portability information and for information on where functions came from 
is very different for readers now than it is for readers in 1992; even if 
we think some part of the manual is clear with our backgrounds, if we find 
users misunderstanding it, that justifies changes to make it clearer for 
the people actually reading the manual).

As a development community we should judge a comment from 1992 saying some 
text should stay in the manual much the same as a comment from 1992 saying 
some code is needed - by considering the reasons behind it in today's 
context (if the code from 1992 was to do with K&R C, it isn't applicable 
any more).

Of course, as a part of a project with explicit goals for society, we 
cannot always give users what they want now - for example, we must not 
promote proprietary software even if sometimes users might find a 
recommendation to use some piece of proprietary software helpful; in such 
cases, we must work for the longer-term goals of building a free software 
world that ends up better for users even if less helpful to them now.  But 
I don't think that has any bearing on a culturally-specific joke about a 
non-software sensitive subject.  Similarly, when making technical 
decisions we need to use our technical expertise to judge what works best 
long term even if a user would prefer an interface specific to their 
problem right now - but again that has no bearing on this joke.

We as maintainers and developers have responsibilities to both the GNU 
Project and our users.  And the GNU Project has moral responsibilities to 
both the maintainers and developers and the users as well.  The 
responsibilities of all relevant people to the users include putting aside 
our own preferences when ensuring the content of the manual works best for 
the users we have now.

Now, if there are still serious doubts about the effects, interpretation 
and appropriateness of the joke in question, we could see if it's possible 
to survey an uninvolved, gender-balanced group of users to find their 
views (on the basis that the effects on users outweigh our personal 
preferences).

-- 
Joseph S. Myers
joseph@codesourcery.com