Grants talk:IdeaLab/Improve Wiktionary's coverage of women honoured in scientific names (and encourage new editors)

Latest comment: 9 years ago by Thepwnco in topic initial thoughts and feedback

Questions edit

  • Could this be driven off Wiki-species, Wikidata or Wikipedia?
  • Could this be extended to non-female epithets?

Rich Farmbrough 22:18 10 March 2015 (GMT).

Wikispecies, Wikidata and Wikipedia are all relatively incomplete and difficult to access programmatically. I've been using data (species names) from the Catalogue of Life (COL), which by no means complete, but is one of the better sources and allows download of their database. I've also used Google's ngram data, which comes from Google Books, to identify which species are most commonly found in publications. I'd like to use the eol.org's data for species but while have APIs to access individual items, they don't have any easy way to access a broader index of their species data. I've started working on integrating the Global Names Index (GNI), which includes data from COL, EOL, Wikipedia, and other sources, but their data is very noisy and some parts of it are difficult to access due to technical issues. In the end, incorporating these data sources would only lead to marginal improvements, so has been a low priority. Wikipedia does have some etymologies, so has been useful in research, but there's no systematic way to find them easily.
Wiktionary is not yet integrated with Wikidata, so even accessing Wikidata from templates is not yet possible (although should be in a few months). I would like to see better integration with Wikidata to reduce duplication of effort between projects, especially of the multiple possible views of phylogenetic trees, but for now it's a bluesky idea.
  • "Could this be extended to non-female epithets?"
Since 2006, I've been working on identifying epithets that are most needed on Wiktionary, starting with those of endangered and threatened species, my main area of interest. Mostly they are not named for women or men, but are simply Latin adjectives or nouns. My understanding of Latin has only marginally improved since then so I've largely relied on other Wiktionary editors to create new entries. The most recent "to do" list is here: wikt:User:Pengo/common_epithets/missing. I originally planned to merely start with specific epithets and then move onto the genera and other orders of classification, but there are always more epithets needing entries.
Entries for eponymous epithets are some of the easiest to create as Wiktionary considers them "Translingual" rather than "Latin", so there are no declension tables to work out, etc. We already have many for men, so focusing on those named for women really is a good starting point for new editors. Genus entries are also relatively straightforward, and finding ones named for women adds a (non-technical) challenge. Pengo (talk) 21:50, 11 March 2015 (UTC)Reply
Thanks for the answers. I have dabbled wit GNI, 17 million entries IIRC, Looks like you've done some great work here. Rich Farmbrough 16:24 12 March 2015 (GMT).

Does this drive female participation? edit

Has anyone investigated whether increasing the coverage of women in science (or something else) affects how inclined women are to edit? Because it's not obvious to me that it would have an effect either way. Going from "Mostly men editing male scientist biographies" to "Mostly men editing male & female scientist biographies" isn't the mission. — The preceding unsigned comment was added by Dingsuntil (talk)

Let's be clear, this is not a civil rights issue. The goal is to remove systemic bias to improve the projects. There is evidence (albeit weak) that there is gender based systemic bias in Wikipedia. Part of this evidence demonstrates that as an additional paradigm, on Wikipedia, increasing the number of female editors will improve the quality of items of predominantly female interest, and to a lesser extent, items of interest to both genders. This is a good reason to want more female editors.
Wiktionary is a different proposition, it is not about biographies. Creating the missing Wiktionary entries is not a gendered output activity in the same way that a biographical article might be.
This proposal then, addresses a proposed pro-female discrimination in Wiktionary entries, that is achievable by editors of any gender.
What are the advantages?
  1. Increase in Wiktionary coverage - a good thing on its own, and one which I welcome, regardless of the systemic bias it arguably introduces (just as I welcome coverage of the latest video game on Wikipedia, even though it arguably exacerbates the systemic bias on Wikipedia).
  2. Provides a potential jumping off point for creating Wikipedia biographies. This may encourage female editors in greater proportion than Wiktionary entries on males honoured in scientific naming. But it certainly contributes (although not in a currently quantified way) to increasing coverage of females on English Wikipedia.
Disadvantages
  1. Latest research indicates that balance on F/M biographies of notable people is not a significant issue on en:WP
  2. Narrowing the work by gender risks prioritising less commonly used epithets
Rich Farmbrough 14:06 11 March 2015 (GMT).
Rich, I am interested in which research you refer to for "F/M biographies of notable people", or what you mean by "not a significant issue"? Pengo (talk) 00:05, 12 March 2015 (UTC)Reply
It's a man's Wikipedia? concludes that Wikipeia over-represents women compared with several other compendia. Other exercise have produces similar results. There is a slightly subtle comparison with ADNB (Reagle and Ruhe I think) which effectively states that, while we cover more women from ADNB than Britannica, we cover an even bigger proportion of men from ADNB - or we have a bigger proportion of women missing. When we are talking about fairly obscure people the impact is very low, e.g. w:Augusta_Stanley which I created in November, generally gets 2-4 visits per day, compared with W:Je Suis Charlie which had nearly 3/4 million visits in its first month.
The other conclusions of It's a man's Wikipedia are also interesting, but there are issues, which I intend to write up at some point, probably on W:WP:GGTF talk page.
Rich Farmbrough 16:16 12 March 2015 (GMT).
I don't know why Rich brings up civil rights issues, since this had nothing to do with what I was thinking or--I thought--what I was expressing. For the purposes of the discussion, I take it as axiomatic that more women is a good thing. My questions were directed to the practical question: will this strategy bring them in? It's not obvious to me that it would. Some possibilities:
  1. Women interested in increasing coverage of sciname-honored women are already well-represented in Wikipedia, when compared with women interested in increasing coverage of other things.
  2. Increasing the coverage of sciname-honored women is already so heavily pursued by well-meaning men who earnestly believe this will drive female participation that women who might otherwise be attracted to this opportunity are crowded out, driving down our score.
  3. "Increase coverage of women honored in scientific names" is an example of what people think ought to engage women, rather than what actually engages women.
I don't know what topics actually engage women in general, and women likely to edit wikipedia in particular, and without specific evidence that this category rather than that one is a key source of engagement, I wouldn't want to say it's the project here proposed. Maybe it turns out we'd get the biggest win from something that's not associated with a particular gender or gender stereotype, like Local Landmarks of Interest in Major Cities or something (or does that already have a gender stereotype?) Dingsuntil (talk) 10:28, 12 March 2015 (UTC)Reply
@Dingsuntil: "Has anyone investigated whether increasing the coverage of women in science (or something else) affects how inclined women are to edit?" I don't know, but I am currently researching the degree to which presenting a once-off opportunity for women to create and edit women-in-science entries on Wiktionary (not Wikipedia), presented through the forum of an IdeaLabs proposal, will drive women to edit Wiktionary (not Wikipedia). You can read about it here. I will also be researching the degree to which discussing the participants in third person, together with speculation on their participation rates and editing preferences, on the talk page of the proposal, increases their participation in the project. I will let you know if I receive enough data to make any meaningful conclusions. Pengo (talk) 12:07, 12 March 2015 (UTC)Reply
I wanted to lay down the reasons that are given for wanting a more neutral gender balance (on Wikipedia, and by extension Wiktionary). There are different ways of reacting to the current gender balance, none of which is fringe, for example:
  1. Women are being oppressed here
  2. We are missing out on a lot of editors
  3. Wikipedia is a thing men like to do more than women
  4. Our content has to be sexist
(There are also a fair number of people, regardless of gender, who say "so what?")
None of these positions is sustainable without research, nor indeed are they (if true) easy to address without hard data.
Currently we are in a position to say (although with caveats) that increasing female (women's and girls') participation will bring improvements in subjects of female interest, smaller improvements in subjects of mixed interest (and I suspect, smaller still improvements in subjects of male interest). That is assuming the new editors are of the same calibre and general inclination as existing editors, which is not a given.
What we have not looked at, in depth, is systemic ways of addressing systemic gender bias other than recruiting females, partly because our descriptions of systemic gender bias are weak and flawed.
Nor have we looked at other questions of ethics and practicality. Most recruitment of female editors will also recruit male editors thus perpetuating the current gender balance, but increasing the absolute number of female editors. Is this a good thing? If someone offered us 200,000 well written male biographies under CC-BY-SA3 it would increase the systemic gender bias, would we accept them? Suppose some predominantly male institution wanted to train its members to edit Wikipedia, would we deny them training?
So for me attracting more (quality) editors is a good thing, and attracting more female editors it follows is also a good thing. Attracting more female editors will probably also help address one factor of systemic bias. For that reason also it is a good thing. It is not an axiom for me, it is consequence of wanting to improve content.
Rich Farmbrough 16:58 12 March 2015 (GMT).
The context here is grant proposals for getting more female editors. Arguing that the grants support a worthy goal is moot. The grants are there. Now we ask: can we find and recruit large blocks of women to edit wikipedia, and if so, how? Dingsuntil (talk) 03:29, 13 March 2015 (UTC)Reply

initial thoughts and feedback edit

@Pengo: thanks for this idea! It's nice to see a well-scoped project put forth from outside of Wikipedia. I also learned quite a lot just from reading through your outline! I note that your work on "kingsleyae" required some research on your own behalf to link the term back to Mary Henrietta Kingsley - are there any particular strategies you would recommend for tracking down notable women (who may or may not have biographies on Wikipedia)? Are there particular resources or research guides that could be developed to specifically support this type of work? In my opinion, all of the above would be valuable processes to document, refine, and promote to others. It would also be nice to have a way of coordinating with both Wikipedia and Wikidata communities and tapping into their respective efforts to improve coverage in this area.

I would also like to encourage you to consider further developing this idea as a grant proposal (by clicking on the "Expand your idea into a grant proposal" button on the grant page) - this would require outlining a budget for the proposed work, as well as articulating some deliverables and/or measures of success that demonstrate the sustainability of the project (for example, documentation for extending the work in other, non-epithets areas of Wiktionary and other languages). -Thepwnco (talk) 05:17, 18 March 2015 (UTC)Reply

@Thepwnco: Thanks for the feedback. The entry wikt:kingsleyae does contain a link back to w:Mary Henrietta Kingsley, although it is a little hidden in the etymology section (it's not obvious that it's a Wikipedia link, but that's how's the formatting is on Wiktionary).
The main strategy for tracking down notable women for this proposal is to find ones who are named in bionomial names and researching who they are. Another strategy, not explored in this proposal, would be to go the other way around -- taking a list of notable biologists/naturalists, such as from a Wikipedia category or from a list of author citations and seeing if their names appear in binomial names (e.g. by adding -ae or -ia to their surname and searching a species database). Even if the gender is not known, it can be guessed to be female if there is a species which ends with their surname suffixed with -ae. Finding notable female botanists who are missing from Wikipedia might also be possible by automatically guessing the gender by the first name of authors, which are listed in a the Plant List database. I don't know of an equivalent database (which includes author first name) for animals or other kingdoms, however there are print books which contain biographies of eponymous species for some groups of animals, which could be helpful. I'd love to look deeper into these other strategies if there was more overall interest in the proposal.
I don't know of any particular research guides. I make random stabs at Google, but other Wiktionarians appear to have more refined methods, and use other reference materials. I don't know of any attempts to formally document research methodologies, or how much they have been discussed on talk pages, but it would be no doubt be useful to create a guide. Similarly, documentation for editing Wiktionary can be sparse despite some good attempts to improve it. One user has taken up the task of writing and maintaining the guide to editing taxonomic names, but it would be helpful if it were expanded to include help for users who are just getting started. (It could also use more information for old timers too). But it seems to be a bit of a chicken and egg problem, as there is little incentive for Wiktionary's editors to write and maintain the documentation when there are so few new editors, and it's difficult for new editors to join without more documentation.
Anyway, for whatever reasons this proposal doesn't appear to have gained much traction outside of the existing Wiktionary editorship, so I'm not looking to continue with trying to turn it into a participatory project. I didn't think it was highly likely to gain much interest, perhaps being a little too technical and lateral in approach, but it seemed like a good match for the incubator, so I wrote it up. If there is a sudden surge of interest, time-permitting I'd be happy to continue, but for now it seems very productive to do so. Pengo (talk) 06:34, 24 March 2015 (UTC)Reply
@Pengo: hi again. I understand and support your decision not to invest more time in the proposal given the lack of endorsements and commitment from other participants/volunteers, however I wanted to again thank you for submitting an idea to the Inspire campaign in the first place - regardless of whether an idea gains much traction, I think it's always valuable for other grantees and community organizers to be exposed to new ideas and approaches that seek to address the gender gap on WMF projects. All the best -Thepwnco (talk) 21:58, 26 March 2015 (UTC)Reply
Return to "IdeaLab/Improve Wiktionary's coverage of women honoured in scientific names (and encourage new editors)" page.