Wikipedia talk:Requests for comment/Abortion article titles

Requests

Request of closing admins

HJ Mitchell, Black Kite, and EyeSerene: As has come up in discussion repeatedly, one of the greatest challenges in this whole process has been the pattern of editors starting with the idea that these two articles are "our abortion articles", whatever that means to that individual, and seeing the titles as just labels to be placed on a vague presumed scope that varies dramatically from person to person. Contrariwise, I and others have been at pains in this RFC to bring home the point that, per WP:TITLE, title sets topic and scope, and that we are choosing these articles' scope here, not just labeling a fuzzy pre-set scope. Nonetheless, it is easy to foresee that, however this RFC is resolved, people who begin implementing it by editing the articles in a fashion consistent with their new scope (because their current content is not consistent with any of the possible scopes) will face a lot of mind-numbing resistance from editors who believe that the articles' scope is whatever notion they have in their heads, not that set by the articles' titles. Therefore, I beg you, please, please include language in your closing statement that specifically reinforces WP:TITLE's mandate that title determines scope, so that editors have something specifically, clearly and authoritatively about this situation to point to for backup. Thanks for your consideration. —chaos5023 (talk) 15:43, 27 March 2012 (UTC)

Well said, and seconded. (It would also be good to make this clear at this stage, by formulating an explicit voting question that "improves on" the question as originally phrased by ArbCom.)--Victor Yus (talk) 15:53, 27 March 2012 (UTC)

@Chaos, If this is a problem, it isn't helped by the watchlist notice, which says "The RFC on the abortion article names as requested by the Arbitration Committee is now open for voting by the community. Please consider providing your opinion." Perhaps it should say "The RFC on the title of two abortion-related articles, as requested by the Arbitration Committee, is now open for voting by the community. Please consider providing your opinion." Yaris678 (talk) 14:46, 28 March 2012 (UTC)

Ugh, yes, good point. I wish I had the faintest idea who to talk to about that. —chaos5023 (talk) 14:51, 28 March 2012 (UTC)

I've just made a request at MediaWiki talk:Watchlist-details#"Abortion article" watchlist notice. There's a bit more info on the whole watchlist notice thing at Wikipedia:Watchlist notices. Yaris678 (talk) 17:52, 28 March 2012 (UTC)

Thank bunches! —chaos5023 (talk) 15:13, 29 March 2012 (UTC)

Tally

I'm working on a raw tally. https://docs.google.com/spreadsheet/ccc?key=0Am3BsUGKovVvdDN4d2M2dVlVWnVpZ2lNWUNDZkhhM2c ... more to follow. Homunq (talk) 17:13, 24 April 2012 (UTC)

OK, I have the raw borda results in the form suggested by yaris678. I do NOT ENDORSE using this counting method, but it was the one which was announced, so here it is (option:points):

1:1380

2:1225.5

3:999

4:973

5:1084.5

6:1088

7:967

8:870.5

9:959

10:1002

11:874

12:1183

13:829.5

14:920.5

15:706

I took the liberty of tallying an additional option 15, which is votes I read as endorsing 12+4+1 as suggested by 20040302.

Again, this is a VERY BAD voting system but it's what was announced. Further tallies of saner voting systems in a few minutes. Homunq (talk) 19:03, 24 April 2012 (UTC)

OK, here's the Condorcet matrix. The Condorcet winner is option 2, which beats option 1 by a (statistically insignificant) total of two votes. The runners up are options 1, then 12, and then a tie between 4 and 5.

net (row beats column - column beats row) totals:
00	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15
1	0	-2	20	17	25	15	24	27	22	22	22	13	27	21	34
2	2	0	18	21	19	23	28	27	26	27	29	4	27	24	30
3	-20	-18	0	-5	-2	-4	4	5	2	0	5	-18	7	4	16
4	-17	-21	5	0	0	-2	8	11	6	4	11	-16	13	8	18
5	-25	-19	2	0	0	-5	6	8	7	5	8	-11	11	9	19
6	-15	-23	4	2	5	0	15	18	15	13	18	-7	17	13	21
7	-24	-28	-4	-8	-6	-15	0	12	1	3	10	-13	11	5	15
8	-27	-27	-5	-11	-8	-18	-12	0	-9	-11	0	-17	3	-2	16
9	-22	-26	-2	-6	-7	-15	-1	9	0	-4	9	-16	8	2	15
10	-22	-27	0	-4	-5	-13	-3	11	4	0	9	-15	8	2	15
11	-22	-29	-5	-11	-8	-18	-10	0	-9	-9	0	-17	7	-1	14
12	-13	-4	18	16	11	7	13	17	16	15	17	0	23	16	28
13	-27	-27	-7	-13	-11	-17	-11	-3	-8	-8	-7	-23	0	-7	15
14	-21	-24	-4	-8	-9	-13	-5	2	-2	-2	1	-16	7	0	17
15	-34	-30	-16	-18	-19	-21	-15	-16	-15	-15	-14	-28	-15	-17	0

As you can see, this is a bit confusing, but if you can understand how to read it, it is MUCH more informative than the Borda count. Basically, the Borda count was off because more of the "option 1" voters bullet voted and buried option 2, which is the more strategic effective thing to do. But overall there were more supporters of option 2. Homunq (talk) 20:01, 24 April 2012 (UTC)

Homunq, do you have any evidence that supporters of 1 buried 2 for strategic reasons (i.e., not as an expression of their honest opinion) more commonly than the reverse? Personally, I ranked 2 last because I believe it to be actively unfair, while the other non-1 options are simply unwise. I really do appreciate all the work you've done here, but I'm finding it hard to square this particular comment with WP:AGF. Best, --BlueMoonlet (t/c) 03:06, 25 April 2012 (UTC)

I'm sorry, I wasn't clear. The fact that 1-supporters often voted 2 explicitly solo last, while 2-supporters often voted it after another option such as 3,4,5, or 12, is the "more strategic" thing to do and results in what appears to me to be a distorted Borda count, even if those sentiments are perfectly honest. Insofar as this is my bias showing through, it's my bias against the Borda count, not among the options. Homunq (talk) 11:05, 25 April 2012 (UTC)

Thanks, I appreciate that. --BlueMoonlet (t/c) 11:46, 25 April 2012 (UTC)

(If anyone doubts that my main interest in this debate is the voting system itself: I could show you a web page from 1998, an FAQ in support of IRV and Condorcet which I made using a custom python script, where I talk about the strategic problems with Borda. That's older than wikipedia itself!) Homunq (talk) 13:33, 25 April 2012 (UTC)

You can see the calculations involved at the spreadsheet link above. Of course, we should all remember that this is NOT A PURE VOTE and that the administrators involved have discretion to consider factors other than raw totals. In particular, they should consider user 20040302's suggestion of 1+4+12 (or as I've called it, option 15), which most voters did not explicitly vote on either way, and which therefore makes an artificially poor showing in the numbers above. However, as I and others said before I saw the totals, Borda count is not the way to go here, and I'd consider the Condorcet numbers more meaningful. Homunq (talk) 20:09, 24 April 2012 (UTC)

Also, working out those tallies was a lot of work, so I'd appreciate it if someone gave me a barnstar of some kind. (But don't give me a bunch of them, that would just be silly.) Homunq (talk) 20:17, 24 April 2012 (UTC)

You have my hearty thanks. Mixed in, of course, with a little more advocacy for the idea that if options 1 and 2 are statistically equal (or if 1 should rank negligibly higher based on Borda and Jezebel advocacy), we should have 1+2+12, or 1+4+12, or (better) 1+2+4+12+.... See Chaos's essay linked above. JJB 20:45, 24 April 2012 (UTC)

What happens if you don't 'take the liberty of tallying an additional option 15', but treat these as actually votes for option 12? Do we add up 1183 + 706 = 1889 in the Borda results? If so, there is a clear majority there. --Nigelj (talk) 21:24, 24 April 2012 (UTC)

There are only 12 votes which I counted as mentioning 15; for all others, I counted 15 as equal-last. About half of those 12 votes already explicitly mention option 12 as well. If I add option 12 on the remaning ~6 votes, then option 12 goes from 1183 to 1233-1239 Borda count points (depending on whether I add it after 15 or before it.) In the Condorcet matrix, I haven't checked (my spreadsheet crashes when I calculate the whole condorcet result at once, so I have to do it in two parts, which is annoying) but just eyeballing the matrix it's entirely possible (likely) that the result becomes a three-way Condorcet cycle of 12>2>1>12. (Though I expect that's not an honest cycle; I suspect that the 12>2 result is a result of 1-voters strategically burying 2. If even two of those burials were dishonest, from voters who actually equally disliked 2 and 12 but saw 2 as the greater threat, then 2 would still be the honest winner.)

But let me say that one of those votes which I counted for 15 and not for 12 was my own, and I would not be happy with 12 alone. So from my perspective, this isn't so much "counting votes for 15 as votes for 12", but rather vice versa. Homunq (talk) 22:00, 24 April 2012 (UTC)

ps. Total number of voters is 158, including JoshuaZ's late vote (which I'm counting because I suspect it's a time zone issue). Homunq (talk) 22:18, 24 April 2012 (UTC)

pps. A lot of voters divided the options up into categories (two or more out of "strongly support", "support", "neutral", "oppose", and "strongly oppose"). To me, this shows that thinking of voting in such categorical terms is natural to people. If the voting hadn't been preannounced as Borda, and all voters had been encouraged to vote in this way, then we could have counted the results using Majority Judgment. This would have given a median grade to each option (for instance, "support-", or "neutral+"), which I think would have been useful in making a final determination. If I were a closing admin, I'd see a difference between a near-tie between two options which are supported by over half of voters ("support-" or better), and one between two options which are opposed by over half ("oppose+" or worse).

In fact, exhortations like HelviticaBold's, to simply vote a ranked ballot without extra arguments or commentary, were actually counterproductive in this sense. HelviticaBold is correct that it is marginally easier to cut-and-paste a simple ranked list; but it is impossible to infer the difference between levels of support or opposition from such a ballot. On balance, I think that's a bad trade-off.

So, if this kind of thing ever happens again, I would strongly suggest using majority judgment or Condorcet voting, perhaps augmented by some formal way to delegate the votes (a la SODA voting). Other good systems, such as Range voting or Approval voting seem to me less practical in this case, because the inevitable problems of how to interpret votes which didn't explicitly rate every option would be more serious than under the systems I suggested. Homunq (talk) 22:36, 24 April 2012 (UTC)

Homunq, what if you do a Condorcet Borda tally of only the four broad categories, as I suggested? --BlueMoonlet (t/c) 02:57, 25 April 2012 (UTC)

I'm not exactly sure what that would mean in practice. What do you do if someone voted something like 5,2,1,4,6? And insofar as people didn't do that, the answers you seek are already in the Condorcet matrix above. Basically, choose the "best" row for each of the categories - that would be 1,2,5, and 12; though 6 is a close second to 5. Homunq (talk) 03:41, 25 April 2012 (UTC)

I'm sorry, I should have said Borda, not Condorcet. My suggestion was an attempt to mitigate some of the problems with the Borda method (though, in the event, things perhaps went in the opposite direction to what I had feared). --BlueMoonlet (t/c) 11:46, 25 April 2012 (UTC)

Yes, it's interesting. Voting system says that Borda is vulnerable to teaming; I did not know that, if the "team" is unstrategic, it can work the opposite way and be vulnerable to spoilers. Homunq (talk) 13:27, 25 April 2012 (UTC)

But by the way: IRV could also have shown pathology in this case. Vote-splitting between 12 and 2 could easily have led the Condorcet winner 2 to have been prematurely eliminated. Thus votes for 12>2>1 could have helped 1 beat 2. Homunq (talk) 13:41, 25 April 2012 (UTC)

This is exceptionally silly. If the closing admins take any of this voting system nonsense into account, we'll be back at arbcom in a week. Stop it. Hipocrite (talk) 13:42, 25 April 2012 (UTC)

I understand what you're saying, but I disagree. I feel that a good voting system (not Borda), used wisely (ie, remembering WP:VOTE) can (or at least could have) help(ed) find a stronger consensus. Homunq (talk) 13:47, 25 April 2012 (UTC)

The Arrow voting theorem assures us that there is no 'best' voting system. There seems to be a pretty strong pointer to option 2 from the above results (which, I should mention, was my preference). Picking a voting system after tabulating the results--esp. if it requires second-guessing what voters meant-- is what will hamper consensus. I'm sure a system could be found to make many of the options come out on top. The closing admins aren't bound to take the top vote-getter, unless I misunderstand the closing method to be used here, so using more systems strikes me as just muddying the water. Thanks for your effort in tallying this, but I'd be inclined to let the closing admins take it from here. JJL (talk) 17:22, 25 April 2012 (UTC)

Black Kite, at least, is on record as considering the "voting" tally highly advisory. Which seems great to me; I'm really looking to the closing admins to actually think about this in terms of policy and encyclopedic coherence, since the community has done spectacularly poorly at that to my way of thinking. (Every bit of support for option 2, which does not unambiguously identify a topic, is part of that IMO, but whatever.) —chaos5023 (talk) 17:58, 25 April 2012 (UTC)

Arrow's Theorem doesn't tell us to give up trying to find a best system; it just puts some constraints on how good that can be (if you want a ranked system; and there are similar constraints for rated or delegated systems). Anyway, I was on record as opposing Borda and preferring Condorcet from before the tally. And the message of Condorcet is: there's no consensus, and 1,2, and 12 all get notable support; so the closers should try to find something more subtle than just "call them X". I think chaos5023's ideas are promising but this section is not for making that argument it's just for presenting the tally. (And even if you think the tally is useless, someone still had to do it, or I guarantee people would have squealed.) Homunq (talk) 18:24, 25 April 2012 (UTC)

Homunq, both methods show that 1 and 2 are close enough to be a statistical tie. Since the voting isn't binding anyway there is no need to just let the one with a slightly better showing win. I say this as someone who voted for the winner under the announced system (1). The differences in the systems are trivial to us since they both show the statistical tie. If all we want is a clear decision at this point a coin flip would serve as well as anything.

However, if the point of this exercise was to hash out ideas then let me say I am intrigued by 20040302s suggestion. I don't like putting the prefix 'anti' in either topic because of the negative spin. I do like the idea of creating pages with a clear and manageable focus. I want unbiased titles on stand alone pages but I think it's valid for two opposing pages to have bias in their own favor. It's like an op. ed. (opposing editorial) in a newspaper. But here's the sticky bit: How do we fairly manage two opposing pages in a controversial subject when anyone on either side can edit either page? If resolving that problem is the real point of all this page refactoring I may want to change my vote. I like the idea of op. ed. but not if we can't make it work. Gaelhalee (talk) 16:14, 5 May 2012 (UTC)

British administrators?

• It's mildly interesting that all three closers are British, which may have an impact on how Option 1 is assessed. I think most Brits would recognise "pro-choice" and "pro-life" as largely American language from a largely American debate. Whether this means that those who !voted option 1 because it's the most common term that they encounter will receive any less weight is anyone's guess, of course.—S Marshall T/C 11:40, 25 April 2012 (UTC)

  Shouldn't matter unless they're working from an unquestioned assumption that these articles are inherently scoped as each about one side of abortion issues, ignoring WP:TITLE and evaluating those terms as ways to talk about abortion advocacy globally. Hopefully we can expect better of a batch of ArbCom-hand-picked admins. —chaos5023 (talk) 13:06, 25 April 2012 (UTC)

Regarding S Marshall's comment "I think most Brits etc", above: Surely verifiable sources (eg. Oxford and Chambers online dictionaries) always trump opinions? Petecarney (talk) 14:08, 25 April 2012 (UTC)

Heh. Okay, if you like, it's my opinion that "pro-life" and "pro-choice" are largely American language from a largely American debate. In my opinion if it wasn't largely about the American right wing, "pro-life" would mean things like "in favour of banning the death penalty". But that's only my opinion.  ;)—S Marshall T/C 22:07, 25 April 2012 (UTC)

I added some additional information here, as it doesn't directly relate to the discussion closure. Petecarney (talk) 09:12, 2 May 2012 (UTC)

Questions about closure

What is the best thing to do with additional options (such as 12+4+1/12+4/12+1, or "activism" instead of "movement"/"movements") which came up after voting had started and so didn't get enough consideration/discussion by many voters?

1. Not consider them; too late, too bad.
2. Closing administrators decide on them yes or no, without giving the full community a chance to comment.
3. Use WP:IAR to extend the closing deadline enough for a brief comment period on these extra options.
4. Any of the above, at the discretion of the closing administrators.

Obviously, in practice, the answer is #4, but I'd like to hear comments on the other options. Personally, I'd favor #3 over #2 over #1. Homunq (talk) 13:23, 28 April 2012 (UTC)

Certainly we want the best outcome, but given the highly contentious nature of the subject matter which led to this unusual way of handling it I think that anything other than #1 will fail to achieve the sense of consensus and closure that this whole exercise was about. JJL (talk) 15:01, 28 April 2012 (UTC)

As for the question of 12+1 or 12+4 (I do not understand how 12+4+1 is coherent), people have been saying all along that 12 is not incompatible with either 1 or 2 or 4. On the contrary, the only sensible way to go with 12 is for it to focus on the issues while you still have two other pages on the opposing movements, and the latter must follow 1 or 2 or 4. I don't think it would be any problem for the closing admins to decide on this course of action, if that is their conclusion based on the discussion that was just closed.

As for "activism" rather than "movements", that is decidedly a secondary issue, behind the question of whether to go with 1 or 2 or 4. I would suggest we choose 1 or 2 or 4 and use "movements" for now. But if, after that decision is made, there is another !vote resulting in a large consensus for "activism" rather than "movements", I should think that would be non-contentious enough for some relevant authority to IAR enough to let it be enacted. --BlueMoonlet (t/c) 19:41, 28 April 2012 (UTC)

The question is on the title(s) to be used for this content and it is clear that !voters considered options (like 12) where the content did not remain in exactly two articles. To continue my advocacy and gratified that the admins are watching this option closely, I submit that if the final decision is 1+?+12, namely, to spread this content around five or more titles, this has the benefit of allowing the admins to kick part of the question back to the community as to how much content should go in each title. If WP:SUMMARY style is observed cautiously in a proper hierarchy with "abortion debate" at the top but with exceptions such as excluding content about say capital punishment from the top article, this allows the admins to give a recommended outline for each topic that gives the community enough space to return to harmonious editing. I have prepared such outlines for contentious areas successfully before and may return to this page. JJB 19:54, 28 April 2012 (UTC) On second thought, my flurry of creative writing may be way too hot to handle right now. Let it simmer. JJB 02:25, 29 April 2012 (UTC)

Comments by Closing Admin

In the closing section, the following comment is made: "This obviously isn't a counting exercise, but there are only three options with any serious amount of support, which are 1,2 and 12 (and its variants). Proponents of 1 are often virulently opposed to 2 and vice-versa. These numbers are about equal."

I'm not sure what "These numbers are about equal" means. The number of votes separating the #1 choice (option 1) and the #2 (option 2) is greater than the amount of votes separating the #2 choice and the #6 choice. That's not close. That's a blow out.LedRush (talk) 13:26, 30 April 2012 (UTC)

• I thought it was quite clear. The number of "1" voters who clearly oppose "2", and the number of "2" supporters who clearly oppose "1" is roughly equal. A number of voters are ok with "1" OR "2" or at least do not strongly oppose the "opposite"choice. Black Kite (talk) 16:00, 30 April 2012 (UTC)

  Unfortunately option 2 is useless from an encyclopedic and policy standpoint. It only even makes sense as an option if you assume that these articles' scope magically arrives out of nowhere and merely calls for some kind of label, instead of their scope being defined by their titles per WP:TITLE. —chaos5023 (talk) 16:44, 30 April 2012 (UTC)

LedRush's comment appears to be referring to Borda count points rather than votes. In this case it is unclear what the concrete meaning of the Borda count points is/should be, so I strongly endorse Black Kite's response as more pertinent. Specifically, the gap between first place (option 2) and second place (option 1) is two voters, while the gap between first and third place (option 12) is four voters. Homunq (talk) 16:53, 30 April 2012 (UTC)

Where are these numbers of voters? Are they tabulated anywhere? I guess when you say, "the gap between first and third place (option 12) is four voters" you're looking at people's first choice !votes? I may be being thick, but I can't make head nor tail of the 'Condorcet matrix' above - are these figures hidden in there somewhere? --Nigelj (talk) 18:16, 30 April 2012 (UTC)

Yes. Look at the row which starts with "2" and the column "1": the number 2 says that there are 2 more voters which have 2>1 than those which have 1>2. Similarly, row 2 column 12 has the number 4. Homunq (talk) 19:03, 30 April 2012 (UTC) ps. see Condorcet_method#Pairwise_counting_and_matrices, except that the matrix above has ((votes for) - (votes against)) in each cell, instead of just (votes for). Homunq (talk) 19:09, 30 April 2012 (UTC)

The quote seems to me to directly relate back to the concept that people in favor of 1 are virulently opposed to 2, and vice versa. However, the Borda count demonstrates that this is not true. If 2 more people prefer 2 to 1, but the difference between option 1 and 2 in borda count is greater than the difference between option 2 and the next 4 options together, it is quite clear that people who support 1 virulently oppose 2, but that the people who support 1 don't hate 2 2 don't hate 1 nearly as much. To me, that is quite instructive as to what is more likely to be an acceptable outcome for all.LedRush (talk) 19:58, 1 May 2012 (UTC)

(I think you mixed up 1 and 2 in the last clause there) You are correct that the Borda result shows an assymmetry, and that part of that is the strong opposition to 2 by supporters of 1. But another part of it is that many who strongly oppose 1, do support 2, but not as their favorite choice; many chose 3 or 4 (and/or in some cases 5,6, or 12) over 2. Since the Borda count, unlike Range voting or Majority Judgment, only counts preference order and not preference strength, it is not really fair to assume that this means that 1 would be "more acceptable" to the electorate in any meaningful way. Homunq (talk) 22:59, 1 May 2012 (UTC)

(your point about my mix up in the last clause above is correct) Your point about people who prefer 2 to 1 not necessarily having 2 as their first choice is quite revealing. Counting up first-place votes and Borda votes strongly indicates that choice 1 is the favorite choice by a wide margin. And I disagree that the count does not show preference strength. It does, and quite clearly. Sure, it's not perfect, but ranking items 1-14 is inherently a listing of the strength of your opinion on the matters.LedRush (talk) 23:15, 1 May 2012 (UTC)

Tabulating votes

Just noting here that I've come up with my own Heath Robinson-style results table (transcluded into Wikipedia:Requests_for_comment/Abortion_article_titles#Admin discussion on the project page). I mean no disrespect to those who've worked on the matrix above, but I wanted something that preserved the nature of the preferences expressed and there was some disagreement about formal methods of vote tallying and weighting. Best, EyeSerene^talk 19:25, 2 May 2012 (UTC)

Now that I understand. Thank you, EyeSerene. What I see now is that 46 people support or strongly support ('SoSS') option 1; 16 SoSS option 2; and 28 SoSS option 12 or some combination involving it (1+4+12, 1+12). I don't suppose anyone is interested in my opinion - I'm just one of the previously uninvolved !voters here - but I see the level of support for option 1 as the problem, and that for option 12 as the most viable solution. --Nigelj (talk) 21:16, 2 May 2012 (UTC)

I like the fact that you include support levels in your table... but I don't fully understand it. Is there an overlap between the two halves of the table? That is, are there some voters who show up as both "strong support" and "first choice", as well as others who show up as only one or the other? Homunq (talk) 21:25, 2 May 2012 (UTC)

No vote has been double-counted. It's complicated because some respondents recorded their vote along the lines of "Support 1 & 12, then 3, 5, 2, 9; strongly oppose 14". In that case I recorded 1 & 12 in the "Support" column, ranked 3, 5, 2 & 9 in the "Ranking" section (counting the supports for 1 & 12 as though they were in 1st and 2nd place, so starting from 3rd), and recorded 14 in the "Strong oppose" column. Really the two halves of the table have to be read together. It would almost be possible to move everything in the "Support" column into "1st" in the "Rankings" section, but I tried as far as possible to keep the format a reflection of the way the votes were offered. I also wasn't sure how to appropriately weight strong support/oppose votes in the ranking system. As I said, it's a bit of a kludge but hopefully representative of the discussion. I'll post up another table showing how I parsed the votes as soon as I can sort out the spreadsheet>wikimarkup conversion. EyeSerene^talk 07:27, 3 May 2012 (UTC)

Changed my mind and tried to work out a scoring scheme (see main page). It's perhaps not entirely worthless but it probably shouldn't be treated with too much respect either... EyeSerene^talk 11:34, 3 May 2012 (UTC)

I believe the scoring system you've made is a good try, but fundamentally unfair; it gives more power to support/oppose than to rankings. Point-systems are always going to have various problems (that's why I prefer Condorcet here) but if you are going to do it, you should at least be as fair as possible. Strong support or support should both get 14 points, neutral should get 7, oppose should get 1, and strongly oppose should get 0. (I'd say 13 for support, but there were some people who showed support without strong support, and that would be unfair to them.) Homunq (talk) 15:46, 3 May 2012 (UTC)

I agree, though I wanted something where an oppose effectively cancelled out a support :) Tbh I was really just playing on the spreadsheet and thought I'd post up the results, but I might remove that table if it's causing misconceptions about the level of attention I think we should be giving to scoring. If you want the data table I used I'll be happy to supply it; you can then weight it more appropriately if you like. As I said earlier though, it's not much more than an exercise in playing with statistics and won't loom large in any final decision. EyeSerene^talk 16:07, 3 May 2012 (UTC)

Fair enough. I've played with my own spreadsheet enough to know that these numbers are close enough that you can make them come out however you want them. So, judge on the arguments, not on the votes, and I'll be happy. Homunq (talk) 18:18, 3 May 2012 (UTC)

I'm bothered by these comments that somehow these results are close. Any way you look at all the numbers and analysis, it's clear that the voting produced a total blowout. First place votes? Option 1 doubles option 12 and triples option 2. Borda? Option 1 crushes option 2 by a margin wide enough to encompass the difference between option 2 and the next 4 closest options. Strength of opinion? Option 1 garners far more top level votes and fewer bottom ones than option 2. Only in Condorcet do the numbers get close, and that is a statistical tie. So, when every way of looking at the numbers shows a blow-out, but one way shows a tie, what's the obvious outcome?LedRush (talk) 18:42, 3 May 2012 (UTC)

Because, as I said before I knew which results would point which way, all those metrics you're talking about are deeply flawed, and Condorcet is in my opinion the only fair way to use the ballots we have in an asymmetric situation like this (that is, one unified side against a diffuse group which is dispersed through vote-splitting). In fact, since strategically Borda tends to favor teaming, I actually suspected Borda would favor option 2 (and 3 and 4), which I personally favor over 1, and I still warned that it was a bad system. And you're wrong to say "there's many results which point to 1 and only one which points to 2"; in fact, the "numerous" metrics you're citing are all point-systems, and thus essentially the same at heart. By the same token, I could say that all of the many distinct Condorcet methods point to 2; but you'd be right to point out that, since they all use the same data in essentially the same way, of course they agree. Homunq (talk) 19:24, 3 May 2012 (UTC)

The idea that the people who supported option 2 were not a unified side is contradicted by all evidence of the history of the articles. This unified group of editors tried unsuccessfully to change article titles on multiple occassions, pushed through changes a mere month after their last proposal was defeated, and then used that one victory to push for a change in the other article's title (after previous defeats there). What has happened here is that many new faces have entered into the discussion, and many people that were merely tired of having to fight the same fight over and over again returned for one last discussion. That results in letting us know that option 1 has far more support than option 2, and that new editors voting here either prefer option 2 to 1 but that opinion is not strongly expressed, or prefer option 1 to 2 and that opinion is strongly expressed.LedRush (talk) 20:22, 3 May 2012 (UTC)

Lets not assume that a vote for 12 was a vote against 1, especially since 12 voters had the chance to explicitly voice an opinion against 1 and most chose not to. The only two options that are close are 2 and 12 for 2nd place, 1 garnered as many votes are 2 and 12 put together while receiving about half the number of no votes. Juno (talk) 21:31, 3 May 2012 (UTC)

I'm not assuming anything. There are two more voters who clearly stated a preference for 2 over 1 (including votes like "support 2 over all the rest") than support 1 over 2 (ditto). That is a fact. Since several of these votes were expressed as rankings, not ratings, it is impossible to reliably infer preference strength from these votes, and therefore impossible to be sure that the 1>2 preference is stronger overall than the 2>1 preference. That is another fact; a fundamental weakness of strictly ranked systems like Borda, and a reason not to use this system in the future. As to whether some people were too "tired of having to fight the same fight over and over again" to vote or whatever, I have no opinion on that. Homunq (talk) 22:02, 3 May 2012 (UTC)

Thank you for pointing out that I am actually capable of reading the various tabulations. Black Kite (talk) 00:19, 4 May 2012 (UTC)

Alleged US-centricism of pro-x titles

I made a comment in the section "British administrators?" above which I though was somewhat off-topic to the discussion closure. Having belatedly read EyeSerene's comments on the project page I realise that it is actually very relevant. In EyeSerene's comment 1 (Site policy, etc.) he says "The point is repeatedly made that ... option 1 is not compatible with a world-wide viewpoint". It it true? What is the evidence? It seems undisputed that the two pro-x phrases were coined in the US but it is worth asking the question whether they have now become international or remain regional. There is good evidence that they are familiar in Britain but I have no knowledge of other global flavours of English. Surely it would be a simple exercise, and helpful to the debate, to find reliable sources to answer this key question of fact? Petecarney (talk) 16:30, 4 May 2012 (UTC)

They remain regional, and people repeatedly raising that as if it were some kind of problem mostly displays how broken the popular way of approaching the RFC is. If we use those titles, it scopes the articles as about the US movements concerned, and that is a good thing because 1) that's what the articles were originally about and WP:PRESERVE would have us not go around gratuitously trashing article history and whatnot, and 2) it is completely stupid to have two separate opposing issue-oriented articles as people seem to think is some kind of mandate from heaven. —chaos5023 (talk) 16:52, 4 May 2012 (UTC)

Three mile island and watergate are familiar in Britain, but that doesn't make them international or British. --Nigelj (talk) 16:59, 4 May 2012 (UTC)

Chaos5023, several editors have asserted that these terms are regional. The point is can you provide any evidence? We don't accept bare assertions in articles, we insist on verfiable sources. That's what I'm asking for here. Petecarney (talk) 17:31, 4 May 2012 (UTC)

No, I am way over my dealing-with-bullshit threshold on this RFC and seriously cannot be bothered. —chaos5023 (talk) 17:33, 4 May 2012 (UTC)

I'd also suggest that, as the one editor so far making the exceptional claim that they are not regional, that the burden would be on you to produce verifiable sources saying so. —chaos5023 (talk) 17:35, 4 May 2012 (UTC)

It's not an exceptional claim. Just search any English language news source outside the US (examples: [1][2][3][4][5][6][7][8][9][10][11][12][13]) Formerip (talk) 20:10, 4 May 2012 (UTC)

Okay. So we do in fact have people referring to pro-life and pro-choice movements outside the US. Do you figure that there's enough of that that the US movements aren't WP:PRIMARYTOPIC for the terms? Because if so, option 1 is ambiguous and what we would actually need is United States pro-choice movement and United States pro-life movement. Which would be kinda unfortunate to not have been considered in the RFC. —chaos5023 (talk) 20:58, 4 May 2012 (UTC)

*Pro-choice movement in the United States, surely. Quick, everyone, let's restart the RFC!  :-)—S Marshall T/C 00:53, 5 May 2012 (UTC)

Eww, S Marshall. That's two entire unnecessary words. Gross. —chaos5023 (talk) 04:54, 5 May 2012 (UTC)

Going back to EyeSerene's comments on the project page I'd note that so far no sources have been identified to support his statement that "option 1 is not compatible with a world-wide viewpoint", while plenty of reliable sources do contradict it. If subsidiary articles are required then titles such as "Pro-x movement in the United States", "Pro-x movement in the United Kingdom", and "Pro-x movement in Australia" etc. would be perfectly reasonable. Petecarney (talk) 10:51, 5 May 2012 (UTC)

• I don't think it's a serious concern, actually, Petecarney. On reading of the closers' deliberations my view is that we're more likely to get a compromise than a decision, so the abortion article titles won't be "pro-choice" or "pro-life" anyway.—S Marshall T/C 11:15, 5 May 2012 (UTC)
• Well, EyeSerene is only noting the "worldwide viewpoint" issue as relevant to the case where we're interpreting these articles as inherently scoped to worldwide one-sided issue coverage, which of course tons of !voters did but is actually completely nonsensical. —chaos5023 (talk) 14:55, 5 May 2012 (UTC)

Really all I intended in my summary was to acknowledge that a 'global view' argument had been put forward by voters depending on what scope they thought the articles should have and/or what they thought the RfC was about. I make no comment on the weight of the argument other than that it's one additional point for consideration among many. EyeSerene^talk 07:49, 8 May 2012 (UTC)

HJ Mitchell conclusions

Affirm there is no consensus and stronger consensus should be demonstrated. While putting in another advertisement for Chaos5023's analysis, I believe consensus would best be demonstrated by having more than three articles rather than limiting options to one or another of the pairs or the monad. I have tested this theory indirectly by rebuilding the dabs pro-death and anti-life, which after initial challenge are now quite stable; a strong consensus (such as this debate needs) was demonstrated on this minor question at DRV. (The dabs are not part of the debate because they have other meanings than abortion, but they illustrate the point that a clear topic scoping upfront removes the need for contention.)

In short, if there is one article that gives all the meanings of pro-life, including those unrelated to the abortion debate, then one subsection of it can be a summary of a second article on right-to-life or anti-abortion or both; other terms are handled similarly. Those who object that "pro-life" shouldn't point to anti-abortion get their way; those who object that "pro-life" should have an article that includes anti-abortion details get their way; and clear scoping is laid out to prevent move wars or fork wars. In particular, if any solution of three or fewer articles exists, by definition there will exist some term that will redirect to some other term that some significant group of Wikipedians believes is inaccurate or misleading. This is so obvious mathematically that I would hope it would inform consensus-building in a second RFC.

In my experience if you can get people to agree on clear scoping the rest is easy. Thank you again for the opportunity to present this proposal. JJB 17:05, 31 May 2012 (UTC)

Thank you JJB, though I'd have appreciated seeing that support expressed closer to 10:57 pm, 23 February 2012 (UTC) when I raised the issue of the false dichotomy. The community ignored it in droves, so I must infer an unfortunate preference for black-and-white thinking predominates. LeadSongDog come howl! 19:16, 31 May 2012 (UTC)

Comment on the initital HJM response

I reserve comment on the rest of HJ Mitchell's response except to note that I agree strongly that whoever proposed voting should be fired, and then to comment on his request that "the same three admins ... should be the ones to close this second phase of the RfC."

In a word? "No."

These three admins were derelict in their duty to close this RFC. Why anyone would give them another bite at the apple after they took more than a month to close this is beyond me. No, sorry, you three are fired. If you want not to be fired, set deadlines, and set consequences for breaching them - I suggest desysoping - for abject failures like this month long wait. No excuses - I don't care. Hipocrite (talk) 12:41, 1 June 2012 (UTC)

The delay is entirely my fault, and EyeSerene and Black Kite shouldn't be considered responsible. I've been inactive on WP and busy in real life of late, which explains but does not excuse the length of time it took me to write my conclusions. Assuming there is a second discussion, you are quite welcome to suggest to ArbCom that they dismiss me (if AGK or a majority of arbs agree with you, I will step down). You're equally at liberty to suggest they replace all three of us, though finding three other impartial admins in good standing might be difficult, so I would recommend you suggest possible replacements. Please accept my apologies for the delay. HJ Mitchell | Penny for your thoughts? 19:05, 2 June 2012 (UTC)

A second (or extended) RFC would certainly be demoralizing. On the other hand, getting it as close to right as possible is important--this has to be settled. It takes too much energy away from actual article improvement. I hope a second RFC can be avoided. The process to stop the interminable edit-warring should not itself be interminable. A decision and finality on the matter is needed. Incidentally, the comment about "discouraging voting" is meaningless. Every comment will be a vote. If you want no voting, you need to do it courtroom style (I am not advocating this)--select an advocate for each view and limit the discussion to those individuals. To think that this RFC could be handled in a way that would keep editors from voting with their comments, by whatever verbal subterfuge is needed, is not so much wonderfully idealistic as it is shockingly wiki-naive. JJL (talk) 03:58, 2 June 2012 (UTC)

This issue will never be resolvable by counting tallies because the community appears, by and large, incapable of desisting from believing this is a referendum on terminology long enough to think about the implications of any of these choices in an encyclopedic context. Everybody likes to go by numbers in tough situations like this because it gives people a nice democratic feeling, but it is not going to work here because the tallies are of people answering six different wrong questions. Hell, to all visible evidence, most of the people registering opinions never read the prepared arguments (and fascinatingly, the closing admins have had almost nothing to say about them either). The consensus you need to be reading to resolve this is the consensus that starts with "Wikipedia is an encyclopedia" and goes from there. You need to actually interpret WP:TITLE and WP:COMMONNAME and WP:PRIMARYTOPIC and WP:PRESERVE and apply the established global consensus as to how this project works. In short, you need to be administrators, not tabulators. And if you're going to disregard my essay on how this process is going wrong, you ought to have a hell of a lot better reason for doing so than that it's inconvenient to think about. —chaos5023 (talk) 07:41, 2 June 2012 (UTC)

Actually, I read the prepared arguments and your essay with great interest. In fact, I think many of your contributions to this RfC have been extremely valuable, but there is only such much administrators can do. Our job is to sum up the arguments and determine what the consensus is (if there is one at all). We could state that the consensus arrived at here wasn't compatible with site policy (if that were the case), but we don't have the power to impose a solution that doesn't have consensus (and people would, not unduly, be calling for our heads if we attempted to). HJ Mitchell | Penny for your thoughts? 19:05, 2 June 2012 (UTC)

Indeed; or we could have closed it one day flat by saying "No consensus, sorry". I think the discussion about option 12 certainly does take the idea of scope into question, so it certainly isn't inconvenient to think about. Ignoring Hipocrite's silly rant, if there is a second RfC I do not mind whether we use the same 3 admins or not; there is always the chance that whoever agrees to take the task on may be delayed by real world issues, as HJ and to an extent myself both were. Black Kite (talk) 19:15, 2 June 2012 (UTC)

In regards to your essay, it is difficult for administrators to base any decision on it because editors sadly did not respond to it. I would actually welcome a second RFC in the hopes that the same issue could be brought up. Ryan Vesey Review me! 19:26, 2 June 2012 (UTC)

Okay. Maybe the best thing that could happen here would be a new RFC that attempts to ask the right question. My minimal standard for the question being right would be that resolutions to the effect of this RFC's option 12, and hybrid resolutions like "1+4+12", are clearly valid answers to it. Wikipedia:Requests for comment/Abortion-related topic coverage, perhaps. We can probably get a more thoughtful reaction out of the community if our initial structure doesn't look like a referendum on terminology. —chaos5023 (talk) 00:51, 3 June 2012 (UTC)

Affirm Chaos, in hopes of a fair hearing for 1+4+12 and ilk. JJB 01:24, 3 June 2012 (UTC)

Black Kite conclusions

Black Kite first seemed to allow the likes of 1+4+12 as a contender, then seems to disallow it. The best response to nonconsensus is not instant runoff. (The best response is to figure out concerns and address them.) The idea that this is about finding the best titles for 2 (of 3) extant articles is demonstrably false, as one or more of the options were for the 2 articles not to continue existing as 2 articles. This should not be a runoff between the highest-grossing candidates (or pairs) to pick one winner and runningmate. This is about organizing extant content into a widely agreeable framework: the current framework is the problem!

The community's concerns are best addressed by a multipolar rather than (um) bipolar solution. If your concern is that a term is too loaded, have an article on that term that expresses the degree of loadedness. If your concern is that a term has the wrong scope, have an article on that term where the scope is defined by the RS. If your concern is that one term should not redirect to another because they aren't synonyms (none of them are), have interlinks to show the compare and contrast between terms.

In the bipolar "solution", every term in the same group gets lumped together. The editors who are favorable to one group object that the nuances are lost, the editors who are opposed to that group similarly (!!!) object that mention of one term imbalances the propriety of another. I don't know that I will take time to review the comments and demonstrate that the nonconsensus arises from the straitjacket of (1) a candidate-pair list and (2) voting, and that consensus would favor a proposal like mine. But the fact is that limiting ourselves to 2 articles (among 4 million) has now been shown to be a loser from the start.

Maybe the best path forward is for me to get my hands dirty and reorganize the current text (along with new text) of the 2 articles towards what I identified as the highest-level terms (pro-choice and right-to-life), with subsections on what I call the lower-level terms, insofar as consensus permits. If it starts to look like I anticipate, we might actually get ripe for splitting before a second RFC can conclude. But that's optimistic both about my time and my reception. At any rate, the takeaway is that bipolar voting doesn't work, and my next attempts will be to demonstrate alternatives. JJB 21:18, 3 June 2012 (UTC)

• Apologies if my second comment appeared to disallow 12+4+1 etc - I have refactored them to add this option. As I suggested originally, I think it is one of the better options given the amount of polarity between options 1 and 2. Black Kite (talk) 09:35, 4 June 2012 (UTC)

I agree that multiple articles (such as 12+4+1) is probably the best way to go here. It does, however, present a problem; while the focus of any such "sub" article is of course the scope given by its title (in general, historical information about a movement), all such articles will have to include a summary-style section on the overall debate. The more such sections there are, the more inevitable (intentional or unintentional) POV-forking will happen.

Therefore, if we do go this route, I suggest that we use a single common template (with v/t/e links) for all the summary-style sections on "abortion debate". I know this solution has its own downsides (linguistic awkwardness of not customizing the section to each article, interface awkwardness of two "edit" buttons with only the inner one working, issues with re-wikilinking terms that may occur in the lede) but I absolutely think they're worth it. These problems are much less serious than POV-forking; and no matter how vigilant and conscientious we are, if there are 4 or more individual "one side" articles, POV-forking is completely unavoidable. Homunq (talk) 12:06, 4 June 2012 (UTC)

Thanks both. I'm not certain that every article requires much summary on the debate, as they are spinouts of the debate and so the debate article is expected to summarize them. There would only be enough mention of the overall debate to ground the reader (such as in the lead); the ordinary weighting issues in each section (i.e., pro and con statements each duly weighted) would still arise but would be expected to be resolved locally according to scope. Nobody overloads one side's article with the other side's POV; that's not a naming issue, that's ordinary disruption. Article forking is valid, but POV forking would not arise because that only happens when two titles are obviously the same subject but one is unwatched, and in this case we have consensus that no two titles are really the same subject, and plenty of watching.

A common boxed template containing debate description would be contraindicated because its contents would be encyclopedic rather than navigational and would be expected to be inline. An inline template containing debate description is possible but these have been rarely successful allegedly because of editability confusion. I like them myself and one such as you describe is possible (feel free, I'll support), but the questions of maintenance, complexity, contingency, and necessity will make it an uphill climb for consensus. JJB 12:40, 4 June 2012 (UTC)

What we need is an RFC that asks the question of what abortion-related topics are we going to cover and under what titles, with a specification that Support for the legalization of abortion and Opposition to the legalization of abortion are considered particularly eligible for new titles, if appropriate, and other articles should probably not be moved unless a particularly compelling reason for doing so is found. Apply the original three-year binding to whatever titles those two articles have afterward, if they still exist, and specify that no prejudice is indicated against the creation of new articles on notable abortion-related topics unless the RFC specifically generates such prejudice (such as by concluding, idiotically, that we will not cover the United States pro-life, pro-choice, and right-to-life movements as distinct topics). Said RFC should not be limited to debate among "winning" results from this RFC because debate on this RFC centered around a fundamentally flawed question and people should be given the opportunity to think outside its boxes. —chaos5023 (talk) 14:27, 4 June 2012 (UTC)

Affirm again. JJB 14:58, 4 June 2012 (UTC)

What happens next?

Whose turn is it to respond? (That's the problem with a triumvirate, it's enough for bystander effect to kick in.) Homunq (talk) 17:42, 9 June 2012 (UTC)

I am feeling as if I should mock up the RFC I have in my head for people's consideration moving forward. I have not so far overcome the mental exhaustion the concept induces far enough to actually do so. —chaos5023 (talk) 21:31, 10 June 2012 (UTC)

I too have A Path Forward that can be sparked by my own contributions and realistic hopes of harmonious editing, but that tempts one to the mental exhaustion that finds other work to do first IRL. I like the admins' latest RFC suggestions asking about scope, and the expansion of the net to other articles is indicated. But the degree to which I can contribute is limited by those other factors that Chaos has alluded to. JJB 17:32, 11 June 2012 (UTC)

Possibilities

If we are to move forward, could people chip in here on EyeSerene's latest comment? Are those three options enough? Are there others that should be considered? What are they? etc. Thanks, Black Kite (talk) 08:40, 17 June 2012 (UTC)

I said this in my !vote, but it seems to me that the obvious solution would be to have the "Pro-life movement" and "Pro-choice movement" articles focus only on the history of the U.S. movements, with a very brief summary of the issues in debate and a {{seealso}} hat to "Abortion debate in the United States". Conversely, "Abortion debate in the United States" should focus only on the debate, with very brief summaries of the movements and {{seealso}} hats directing to "Pro-life movement" and "Pro-choice movement". --BlueMoonlet (t/c) 00:30, 18 June 2012 (UTC)

Followup RFC draft started

Bullet bitten; I have started drafting my idea of what a followup RFC should look like at User:Chaos5023/Abortion coverage. —chaos5023 (talk) 22:10, 21 June 2012 (UTC)

Thank you very much for your ongoing work on this. I wonder though if it might be feasible to ask those two questions separately rather than in the same RfC? As I mentioned on the main page, I think until the issue of the topic structure is decided, discussing individual article names might be premature. It could be, for example, that editors decide we should have a single article on the US abortion debate rather than separate articles for each viewpoint. Obviously in that case discussing the separate article names would be moot.

I also wonder if, rather than pre-defining a set of options, it might be worthwhile leaving it to respondents to suggest them (or to predefine some possibilities in fairly vague terms and invite other suggestions)? Admittedly the last RfC has thrown up some options that could form the basis of the next, but given some of the comments hereabouts and bearing in mind the eventual need to judge consensus as fairly and accurately as possible, I'd like to be sure that everyone understood exactly what they were discussing! In my experience the best results on WP often come from asking brief, tightly focused and unambiguous questions with open answers that invite explanation (eg "How should Wikipedia organise its coverage of the abortion debate?"). On the other hand, maybe I'm just being too cautious... :) Finally on a personal note, I'm not really sure about the propriety, as a closer, of my commenting on the framing of a new RfC. If there are any objections—perhaps some might feel it calls my status as a disinterested outsider into question—I'll happily step back. EyeSerene^talk 10:15, 22 June 2012 (UTC)