"Social Identities" field in CK

Edited: This plan has been cancelled. See https://www.librarything.com/topic/345540#7968437

We are thinking of adding a new, multi-line Common Knowledge field for "Social Identities."

The term itself is up for discussion. But the idea is to provide a place for members to record categories that often fall under the term "identities" or "social identities." Here, for example, is a page from the University of Wisconsin Library https://lo.library.wisc.edu/DEI_foundations/lesson_1.html , which lists Race, Ethnicity, Sexual orientation, Religion, Disability, and so forth. We considered bucketing this--have different fields for different types of identity, but this gets complicated and potentially contentious quickly. So, for now, only Gender will have a separate field, because such a field already exists.

The goal here will be to allow members to ask questions like "What are some great examples of science fiction written by Asian Americans?" and to search, sort and see graphs of their libraries by these terms.

As with the Gender field, this would be free text. LibraryThing staff will not be setting down a list of official terms. Members will have to work that out. As with the Gender field, we will be insisting that members act in good faith and consideration and, where possible, use the terms given by the authors themselves. (When in doubt, use the most recent terms given by the author.)

Intentional abuse of the system will, of course, be a violation of the Terms of Service. Gender has worked pretty well, with relatively few disagreements, so we have some hopes this can work as well.

Okay, what do you think? Is one field okay? Is "social identities" a good, embracing term? Any other comments or suggestions?

I have to say that when I saw the subject line, I thought "what the heck are 'social identities'"? I'm not sure of a better term, but this one is confusing to me.

It could be just "Identities." This feels too open-ended to me, and suggests identity-verification systems like VIAF, but I don't feel strongly.

"Identities" is probably worse.

It's really tough. The site you linked to defines "social identities" as . . . groups that are based on the physical, social, and mental characteristics of individuals, which is very nebulous, and their attempt to differentiate between "social" and "personal" identity, is not very helpful.

I am not sure that the term really makes sense and people won't use it to store the author's Facebook and Twitter ID for example or pseudonym names - because this is what I think of when you say "social identities". It also does not make it clear that it is about what the author calls themselves as opposed to what the world or some part of it call them (so is the American son of Japanese immigrants to USA an Asian American if he is calling himself just American everywhere)? This will also dilute the results a lot - I would like to know the heritage of my authors occasionally but if we rely on what they call themselves, it won't really get me what I am looking for. Plus in most countries outside of USA, you don't generally become hyphenated just because you belong to two cultures.

And then comes the issue with race and its connotations even in English, let alone when you go outside of English - the same person may be considered or even called and marketed as a different race across languages and sometimes across different markets of the same language to match better the local realities

Just thinking aloud...

>5 AnnieMod:

Everything you say is true, but I also think these are things people care about and use. Lots of real people want lists of African American mystery writers, or plays by disabled authors. These lists are often provisional, contentious, tricky, in the main or at the edges. This problem does not go away just because the lists is made on a blog, not on LibraryThing. The value doesn't go away either. This is data people want, want to play with and want to improve.

>6 timspalding: I don't disagree with you in general - if properly collected as information, these can be very useful and interesting (and I am a data junkie - so I like data). I am just not really optimistic at them being friction-free - gender, as loaded as it can be in some cases, is almost a non-issue compared to some of the other identifiers, especially in a global context. So as I said - just thinking aloud.

Are you planning to give any guidance as to what goes in the field? Any limitations on what should be entered?

So, my first thought was that the categories you have listed do not enable answering the question you have suggested as the goal. Not all people of Asian descent are Asian-American.

With gender, we tend to look for how the author self-identified. There's considerably more guesswork and audience opinion in an identifier such as "Asian American".

I think we're better off sticking to lists and tags, because I'm coming up with a slew of reasons this is going to be more dramatic than helpful.

As far as not knowing what "identities" refers to, don't we have to make these adjustments constantly? Forty years ago, who of us would have understood what "trans" referred to or "LBGTQIA"? Yet those are pretty essential concepts now. We can learn.

So, this would basically be "author tagging" as opposed to "work tagging"? I mean, with just a tiny bit more instruction to users?

I feel like there are requests for other work related fields in CK that are much less ambiguous and more likely to be used than this one.
This field just feels "fiddly" to use a term Tim likes to use.

>10 aspirit:

>7 AnnieMod: gender, as loaded as it can be in some cases, is almost a non-issue compared to some of the other identifiers, especially in a global context. So as I said - just thinking aloud

I was worried we'd have fights. I was pleasantly surprised, I have to say.

So, my first thought was that the categories you have listed do not enable answering the question you have suggested as the goal. Not all people of Asian descent are Asian-American.

Yeah, it's the problem with uncontrolled vocabulary. It can be partially solved with additive searches--Asian American plus Asian plus anything with "Asian" in it, etc. But there's no real alternative, short of providing members with a single racial and ethnic ontology. Shiver.

I think we're better off sticking to lists and tags, because I'm coming up with a slew of reasons this is going to be more dramatic than helpful.

Members don't tag books with author identities that often. We could provide tags for authors, but that would be the same problem. Except that tags are about quantity (something tagged "thriller" more times or a higher percent of the time is "more" a thriller, but that doesn't work with, say, being black). And tags are (currently) personal. So if someone tags an author wrong, you can't do anything about it.

So, this would basically be "author tagging" as opposed to "work tagging"? I mean, with just a tiny bit more instruction to users?

No, because tags are about degree and are not corrigible by other members. This would be a CK field, so any member could change it, add to it, etc. And, as CK, it would be "binary," in that sense that someone either is or is not Black--it's not a matter of degree.

I feel like there are requests for other work related fields in CK that are much less ambiguous and more likely to be used than this one.
This field just feels "fiddly" to use a term Tim likes to use.

I don't think fields should be pitted against each other.

My initial thoughts:

1. A social identity may change depending on who is doing the identifying.

I recently read a book by Aminatta Forna in which she talked about how in the UK and US she's considered black because she is darker than the majority, but in most of Africa she's considered white because she's lighter than the majority.

2. Claiming to be part of a group doesn't necessarily mean the group claims you back.

Rebecca Roanhorse says she is of Ohkay Owingeh Pueblo descent, but they dispute the claim.

3. Just because someone could be identified as part of a group doesn't mean they want to be.

As >5 AnnieMod: brought up.

>14 timspalding: Members don't tag books with author identities that often.

Might that indicate members aren't particularly interested in author identities?

>14 timspalding: someone either is or is not Black--it's not a matter of degree

I was an enumerator in the 2010 US Census. A person can definitely be "kind of" Black. I saw that over and over again when people talked themselves through how to answer the racial question for themselves or other household members.

I have a cousin who's "partially" Black. I've lived with multiple people who would only sometimes identify as "Black" and mostly only when they were tired of arguing about their own identities.

These are some of the simpler cases for racial or ethnic identity by degree. Seriously, when "Asian" is thrown into the mix, identity becomes a (ETA bigger) argument. I've heard this directly from authors who might or not be labeled that way.

In other words, >15 amanda4242: 100%.

>14 timspalding: So how dark you should be to be considered black? Or how much black blood do you need?

People pass for different races for various reasons. And a lightly skinned black person who is calling themselves black in USA can be considered white in Africa or in Brazil. Race is a social construct and depends on language and locality. Are we going to enforce the US definition to everyone?

tags are about degree and are not corrigible by other members

I see that as an advantage here. There's no question about who has the final say. Each member decides for themself.

Yet we can still search on the identity terms and create lists based on the tags when desired.

someone either is or is not Black--it's not a matter of degree.

Ack, no!

See, this is a huge problem. Is "black" being used in the sense of "color", or in the sense of "race" (without getting into the whole issue of whether "race" is a useful, or even valid, construct)? In neither case is someone "is or is not". Colorism is rampant all over the world (see "brown paper bag test", see "if you're white, you're all right, if you're brown, stick around, if you're black, get back"), and there are slews of bi- and multi-racial people.

>16 amanda4242: Might that indicate members aren't particularly interested in author identities?

Naw. I think tags bring out certain things and not others. Readers are interested in whether a book is fast-paced or believable or optimistic or depressing. But they don't use such tags that much.

I see that as an advantage here. There's no question about who has the final say. Each member decides for themself.

New rule: Every member is required to tag 15 million authors.

someone either is or is not Black--it's not a matter of degree.

Poorly expressed. I don't mean that it's obvious or simple. I mean that having two members tag someone black is not worse than having 16 do so. Tags are about numbers. When Wikipedia says someone is an Japanese American authors there isn't a little vote up or down next to it. It has to be there, or not there. That's not to say there aren't edge cases, but it's not matter of how many people take the trouble to apply the label.

Let me just note that while CK is binary it allows multiple entries. So if someone can be described in two ways, you can add both ways.

>22 timspalding: So if someone can be described in two ways, you can add both ways.

But that doesn't answer whether they should be or want to be. I don't want to make assumptions about an author's identity, or describe them based on cultural standards they may not share.

I like the term "social identity", but I'm not sure this is going to work without some options to choose from. To use an example from the existing gender field, there is not a consensus on the use of "non-binary" vs "nonbinary", let alone every possible combination of "non-binary/trans-masculine", "transmasculine/nonbinary", etc. The gender field works (to a certain extent) for searching/sorting/graphing only because the vast vast vast majority of entries are either male or female.

Once you get into questions of Latino vs. Latinx vs. Latin-American vs. Mexican-American vs. Hispanic, or disabled person vs. person with a disability, it's going to be useless for searching or sorting or graphs. It's interesting information to have e.g. in the biography field, but I don't think it's going to work as a CK field without some parameters.

>21 timspalding: New rule: Every member is required to tag 15 million authors

Tim, do you really expect us regular members to fill out this social identity field on millions of author pages?

That's interesting. Excuse me. The word I meant was disturbing. We're not bots, and we're not LT employees. CK is an unpaid volunteer effort relying on humans. We complete the fields we feel are worth our time. So far, I'm not seeing anyone eager to so much as copy author identity tags for 15,000 authors, nevermind do intensive research for a 1,000 times more pages than that.

>25 aspirit: All the gender fields had been filled by “regular members” together with all other fields in CK. The number is a tongue in cheek but yes, the idea is for the members of the site to fill the new field.

>19 aspirit: Bingo, I was about to say the same thing. Personal tags are the perfect answer for this problem.

>26 AnnieMod: You missed my point.

This is a different type of project than the gender field. It's more challenging.

Who currently wants to do this when it looks like more trouble than it's worth for us when we're already struggling to complete author CK?

Additionally, I'm fairly certain the authors with identity tags already will be those whose CK field is filled. The purpose would be to create a graph. See >24 norabelle414: again then imagine what our graphs would look like.

>28 aspirit: I was responding before your edit. :)

>28 aspirit: I'm sorry I'm not being clear to start with. I'm going to Ignore this thread for later and log out.

I don't think fields should be pitted against each other.

This isn't about pitting fields against each other. This is about making efficient use of developer time. This field will probably be minimally used, as seen by the numerous comments against it above. But something like say a Table of Contents field would get more use because more people actually want it and have openly asked for it.

No one is asking us to create a field that "forces" a "social identity" onto a person who may or may not want or agree with what we put into that field.
If I were an author who had a page here, and saw it filled in... I'd come erase the whole thing each and every time.

>31 gilroy: I am a (very obscure) LT author, and I will monitor this closely for my own page.
I'm of Asian descent, but I'm not Asian American (because I'm a European). And I'm part of the rainbow community, but I will definitely nuke any info calling me gay - that's inaccurate.

A Table of Contents field would be very useful.

Finding out how an author self-identifies will be difficult, even for current authors. Good grief, sometimes getting gender information is hard enough, never mind nationality. These are often just taken for granted on an author's website, or the blurb in the back of the book. Then add in how someone a few generations would have self-identified, and how that translates into today's vocabulary, especially in the cases where certain terms are no longer acceptable today.

So what do you enter for the American born child of a Bengali and a Punjabi immigrant? Or of a Polish and a Ghanian immigrant.

I just don't see this working.

>5 AnnieMod:

And there is the big disconnect that Asian in the UK means someone with South Asian heritage, whereas in the US to most people Asian still typically means East Asian heritage - despite the US Census categories.

>6 timspalding: This is data people want, want to play with and want to improve.

Yes, data people want so they can ban, discriminate against, and use to push back against these lists of authors, rather than just reading them

While I'm sure it would be interesting to have this kind of data, and I'd definitely use it if it were there, there is something about the whole idea that makes my skin crawl.

Everything I've ever been taught about data protection tells me that you should never put this kind of information about named, living individuals into an open, searchable database, especially when they haven't vetted it themselves and you don't know where it's come from. Even if the individual pieces of data seem to be in the public domain, the act of aggregating them takes you into a potentially very bad place. It's not quite on the level of handing out pink triangles and yellow stars, but it has a whiff of that about it.

>35 gilroy: >36 thorold: You both make a good point (and to anyone else who made this point earlier; I simply browsed over this thread). I sometimes forget how much evil can exist.

>25 aspirit:

Tim, do you really expect us regular members to fill out this social identity field on millions of author pages?

Over time effort builds up. Gender has been applied to 607,879 authors. While not all authors by any means, assignments are strongly clustered in the most popular authors. Authors held by 100 members are 91% gender-assigned. Authors held by 1,000 or more are 98.7% assigned. This is much better than saying every member should decide for themselves, with no shared data. Shared data is what makes the effort worthwhile.

>36 thorold:-37

Go to Wikipedia pages. Go to home pages and news articles.

Javier Zamora — "Javier Zamora (born 1990) is a Salvadoran American poet and activist" (Wikipedia)
R. F. Kuang — "Rebecca F. Kuang is a Chinese-American fantasy writer." (Wikipedia)
Neil deGrasse Tyson — "But Tyson's current stature is unlikely for another reason: He's a black astrophysicist, as elusive a phenomenon as the Higgs boson." (NPR)
Silvia Moreno-Garcia — "Silvia Moreno-Garcia (born 25 April 1981) is a Mexican Canadian novelist, short story writer, editor, and publisher." (Wikipedia)
Tiffany D. Jackson — Wikipedia category "African-American novelists21st-century African-American women21st-century African-American writersAfrican-American women writers"
Aiden Thomas — "Aiden Thomas is a trans, Latinx, New York Times Bestselling Author with an MFA in Creative Writing from Mills College" (their publisher, Macmillan)

How do people think libraries make lists of African American sci-fi authors? They look it up.

Now, if members like, we could require citations. There are some fiddlies about where to put that, but it could be done.

>35 gilroy: Correct. Book lists have been weaponized. This is a danger that professional book reviewers and librarians have been talking about, and it's been something that's been weighing on some of LT members who make public lists based on author or character identity markers.

The more I think about it, the more increasingly concerned I am about the reason this particular project is being brought up now. My guess is it suddenly looks profitable, especially with most of the work being done with free labor.

I'm also realizing how bad the consequences could be. I woke up thinking how this social identity project could endanger actual lives (which is why I returned to the topic).

All this for someone maybe to have an automated graph...?

>38 timspalding: Shared data is what makes the effort worthwhile.

What you're saying is that crowdsourcing real people's racial, ethnic, sexual, and social health identities is "worthwhile". And that's only if it's done in a way to make the data look authoritative, which tags don't. This type of effort has a history that is sounding off mental alarms.

Do you know anyone who will enter this data saying they believe this effort is worthwhile? If so, do you know what they intend to do with the information?

>39 timspalding: How do people think libraries make lists of African American sci-fi authors? They look it up.

Yes, for their own list, usually meant to be used for a short time by a specific audience, who usually given disclaimers about the terminology.

>39 timspalding: As I stated above, I do think this is interesting information to have on an individual basis (e.g. I can look up an author I like and read the biography field and think "oh, I didn't know this person is of Filipina and Indian background, I can see how that is reflected in her work" ) but any kind of automatic aggregation is completely useless.

Librarians do look up this information to make lists of books by particular social groups - as do users here. But they create a list first and then add books/authors to it based on their research and good-faith assumptions, they don't have one automatically-created list for every author whose website says "nonbinary" and a separate list for every author whose website says "non-binary"

What you're saying is that crowdsourcing real people's racial, ethnic, sexual, and social health identities is "worthwhile".

Do you use Wikipedia? Do you use authors websites? Publishers? This information is out there. Authors want it out there.

Yes, for their own list, usually meant to be used for a short time by a specific audience, who usually given disclaimers about the terminology.

Anyone can edit Wikipedia. Do you object to the existence of these pages?

https://en.wikipedia.org/wiki/List_of_African-American_writers
https://en.wikipedia.org/wiki/Category:Hispanic_and_Latino_American_writers

>39 timspalding: Javier Zamora's author page doesn't yet link to Wikipedia; the source of his nationality is actually Poetry.org, copied to Wikipedia; the term brings up the debate about whether all Salvadorans are "American"; and neither nationality nor place of residence have been added to the English language LibraryThing author page by a person.

I want to fill in fields because the page was linked as an example for CK but figure the reminder of how infrequently the existing nationality and residence forms are completed is good for this discussion.

These fields are frequently left blank when the gender field is completed.

>45 aspirit:

And Javier Zamora’s own website simply describes him as “born in El Salvador”.

Ordinarily, the term {nationality}-American is not used for immigrants.

Other than stirring up the controversy we see here, I don't understand the need for this. As Tim has recently stated above, there are other sources for this information other than LT, so why here? Is LT trying for some obscure SEO hit from Google?

Personally I think some of the other features we've been begging for for years would be a better use of time and I have no intention of filling in any of this information on authors. I don't care. I don't have a stake here. And before you call me selfish, I fill in CK data on other books, books I don't have and don't personally care about, but the information is useful and I think helpful for other people. This I can't see as being of primary bibliographical importance.

Here's my proposal: Every identity must be sourced to a link.

I understand your concerns, but I don't see that, if an author is listed as something on Wikipedia, their home page, their social media, publisher and so forth, we should avoid "compiling" data that is far more easily found and read elsewhere.

>44 timspalding:

Although I would argue that Olaudah Equiano should not appear on the list of African-American writers.

>48 timspalding: Maybe you need to add a different response here, Tim.

WHY? Why must this be warehoused HERE?
What is your purpose in pushing so hard for this field?

Is one of your contract partners demanding that we create this resource or they pull fundung?

Okay, let's take an example—trending authors in an upcoming "explore" tab feature.

Here is an unadjusted list of trending authors:(1)



It has one person of color, Celeste Ng.

Here is a list of trending authors with some consideration given to racial and ethnic diversity:



It has six people of color, or 25%, still far below US demographics.

My feeling is that the latter is more interesting and gives more opportunity for serendipitous discovery. Yes, Lee Child and J. K. Rowling lost their place in the top 24 trending authors, but something is gained in representation and variety.

My feeling is that a diverse set of authors in a trending list is rather like a diverse class at university—a good thing. But, indeed, there is the counter-argument, given recently by the conservative justices at the Supreme Court, that we should not consider diversity, but be color blind.

Which way, LibraryThing?

---

1. In fact, I'm already considering a sort of diversity. If you do it completely "straight" you get mostly long-time repeat winners, like Patterson and King. (Such lists are also more male.) The list is boosting authors whose current success comes from nothing or from a low base, and dinging authors who have hit after hit. Why? Because a list of authors everyone knows is boring. I tend to think a parade of white men is also a bit boring.
2. By the way, as mentioned before, LibraryThing recommendations consider a wide variety of "diversities." For example, we try not to recommend ten well known books, or ten obscure ones, but to create a mix of popular and obscure, because such lists are more interesting. We do the same for "recency." And, yes, we also use gender as a boost. This ensures that a book about American presidents isn't a parade of men, but might have a book by Doris Kearns Goodwin or Kate Andersen Brower. I think that improves recommendations. If members disagree, well, I can certainly bring back the parade of men. Or, since members like features, maybe we could make a parade-of-men toggle.

WHY? Why must this be warehoused HERE?

We record the location of the authors grave for Pete's Sake!

>52 timspalding: That doesn't answer the question

ETA: When it seemed like we had a possible literary Tourism side building with LT Local, it made a little sense to track grave location. But since that feature has really been allowed to grow dormant, maybe the grave site field needs to go away too.

>48 timspalding: What if they are listed as something different on Wikipedia, their home page, their social media, and their publisher? To use one of your own examples from >39 timspalding:, Neil deGrasse Tyson is on Wikipedia's list of "African-American scientists", but your quote from NPR says he's black. So which one should be entered into the field? And that's not even getting into the fact that he's on Wikipedia's list of "Puerto-Rican people of African descent" and "American people of Saint Kitts and Nevis descent". Tyson himself has stated that he does not want to publicly discuss race, and his website does not include any reference to his race so there's also an argument for listing him only as "American".

And what about listing people as "White" or "Caucasian"?* Some white people can be very adamant that white is the default and does not need to be mentioned, or that simply "American" means "white" (or "British" means "white", getting into non-US nationalities is a whole other can of worms). Some are very adamant that they are not white, they are "of Scottish, Irish, German, and Italian descent".

Some possible other solutions (since so far I've just been complaining):
- author tags (as mentioned above)
- encourage people to fill out the "short biography" CK field, as well as nationality, birthplace, places of residence
- pull more "author biographies" from other places (very few authors currently have them)
- searching of "author biographies"
- author lists (similar to book lists)

_______
* or "straight" or "able-bodied" or "neurotypical" or "Christian"

Tim, I think some context you may be forgetting is the fraught nature of the current environment, that makes this look like doxxing to some people. There's a difference between a list of books by trans authors compiled by a Pride organization, and a list of trans authors compiled by a right-wing hate group. It's the same information about identity in a very different context.

norabelle414 (#54):

Some white people can be very adamant that white is the default and does not need to be mentioned

Being adamantly wrong does not mean they're right or that their views need to be respected. You're undermining your entire point by saying what amounts to "Racists don't think they have a race, so we shouldn't list one for them."

Tim, I think some context you may be forgetting is the fraught nature of the current environment, that makes this look like doxxing to some people. There's a difference between a list of books by trans authors compiled by a Pride organization, and a list of trans authors compiled by a right-wing hate group. It's the same information about identity in a very different context.

Indeed. But if we source all information we can eliminate the small chance that some member uses a right-wing hate group for their information.

Again, these arguments all apply to gender. Members could be filling the gender fields with right-wing hate terms. They aren't. If they ever did, members would squash that data like a bug. If that theoretical concern were enough, we should take down gender too. Other fields too. Nationality, names (deadnaming), etc.

>52 timspalding: A burial location isn't a nebulous concept.

>56 lorax: I am not undermining my entire point, I did not ever say that we should not list a race for racists. Do not put words in my mouth.

Some possible other solutions (since so far I've just been complaining):

- author tags (as mentioned above)

This is author tags. The only difference is that tags about truth emerging from large quantities of data. And tags can't be changed by others. We aren't going to get large amounts of data. We're not going to be able to say "Looks like Neil deGrasse Tyson is black because he's been tagged black 50 times and white only once."(1)

- encourage people to fill out the "short biography" CK field, as well as nationality, birthplace, places of residence
- searching of "author biographies"

That's not going to get us any sort of analysis. There's a good case for getting a list of Mexican American mystery writers. If that requires searching author bios and then cross-checking them manually against tags, well, nobody's going to do that.

- pull more "author biographies" from other places (very few authors currently have them)

I think we've got close to all their is, without copyright problems. We can pull in the Bowker ones, which come from Bowker or (mostly) the publisher. We can pull in Wikipedia. That's about it.

- author lists (similar to book lists)

I don't see a real difference. If members think it wrong for us to record "Mexican American" for an author who calls themselves Mexican American, it doesn't matter if that's in a CK field or a list.

---

1. I use The Diary of Anne Frank for this. A lot of people have tagged it "Netherlands" because, well, it happens in the Netherlands. One person tagged it "Belgium." Are they misinformed? I actually asked, and they tagged it that way because their copy is in Belgium, so it's not wrong, just potentially misleading without context. Misleading or not, however, it doesn't matter, because the number of "Netherlands" tags is so great.

So it's virtue signaling so that LT's popular authors photo list looks right according to today's ideas of wokeness and anti-racism? Sounds like it to me.

>57 timspalding: So then change the context like any well read person would do:

There's a difference between a list of books by Black authors compiled by the NAACP, and a list of black authors compiled by a right-wing hate group.
There's a difference between a list of books by Asian authors compiled by a Pacific Rim organization, and a list of Asian authors compiled by a right-wing hate group.
There's a difference between a list of books by {Insert relevant term} authors compiled by {relevant} organization, and a list of {Insert relevant term} authors compiled by any hate group.

It's the same information about identity in a very different context.

ETA: Having the information all compiled in one location would make it EASIER for these hate groups. Not more difficult. Just because we have a store house, doesn't mean it is fire proof. All you need is ONE bad actor.

I think that when people are making an effort to read more books by authors of color this can be a very good thing to have.

>52 timspalding:
This seems like a politician's 'answer'. Can we know if there is any commercial advantage to this proposal?

The value, now at no monetary cost to me, of LT in many different ways is such that I can begrudge no expenditure of time in adding data to the site. I'd still like a proper answer to >50 gilroy:'s question.

>63 susanbooks: The challenge is that "of color" has a different definition for different people and cultures. Which is why author tags were proposed a few times - then all these different definitions can be there - and one can filter anyway they want...

Which is why author tags were proposed a few times - then all these different definitions can be there - and one can filter anyway they want...

We have have multiple terms in CK too, though. The difference between tags and CK entries is not this. It's that tags are personal, and can't be corrected or changed by others. And tags are statistical—five people tagging someone "black" is worth more than one person doing so.

>66 timspalding: But tags can only be added to from other members while CK terms can be replaced. I am not even talking about deliberate vandalism but about well meaning person whose definitions differ from another well meaning person's and who is trying to normalize their graphs (because this is the ultimate goal here, right? Pull that into graphs and tables and nice statistics) and replace different terms (which mean different things for someone else) with a single one (because they mean the same for them). Black and African American may be the same for one person and very different things for someone else. Same for Latino, Latina, Latinx and Latino-American. Or South-Asian and Asian. And so on.

If you've made up your mind to add this field, you will add it. Maybe you are right and everyone else is wrong and it will be enthusiastically embraced and filled in by the community. I doubt it but what do I know :) I still think it is a really bad idea and if added, I plan to stay away from it. But your site, your decisions.

>67 AnnieMod: Actually, I think you've hit on an interesting point.

Tim seems to be coming at this as a computer data scientist and programmer. All he is seeing is the graphs and data and the end results. All those of us arguing against it are coming at it from a humanity and psychological stand point. So where we are saying "This doesn't work because of this factor," someone just looking at it for the data won't see and won't care about that argument.

ETA >1 timspalding: IF this had been proposed many many years ago, like maybe 2010, 2012, even as late as 2014, you might not get as much of a push back as you are getting now. The world at large, outside of LT, is primed for an explosion related directly to the field you are trying to build. Maybe you think this is being hyperbolic or overstating the situation, but I don't think so. Things are trending the wrong way for this type of field RIGHT NOW. If things calm, I might not be as vehemently against it, but the cauldron that is the internet world is not exactly a good and safe place for such a concept, regardless of intent. Tweak an algorithm based on photos, encourage more photos of the actual author instead of graphic imitations, and do things that way.

>51 timspalding: If you're looking for more diversity in trending authors, why not try nationalities? Looking at the second set of pictures, it still looks like a hell of a lot of US authors. I know an author's nationality isn't always straight-forward, but I feel giving a bit more weight to authors from India or Latvia or Indonesia or anywhere else outside the US/UK would produce a more diverse kind of diversity (if that makes sense).

Tim seems to be coming at this as a computer data scientist and programmer. All he is seeing is the graphs and data and the end results. All those of us arguing against it are coming at it from a humanity and psychological stand point. So where we are saying "This doesn't work because of this factor," someone just looking at it for the data won't see and won't care about that argument.

"So it's virtue signaling so that LT's popular authors photo list looks right according to today's ideas of wokeness and anti-racism? Sounds like it to me."

Wow. Humanities has changed since grad school.

IF this had been proposed many many years ago, like maybe 2010, 2012, even as late as 2014, you might not get as much of a push back as you are getting now. The world at large, outside of LT, is primed for an explosion related directly to the field you are trying to build. Maybe you think this is being hyperbolic or overstating the situation, but I don't think so. Things are trending the wrong way for this type of field RIGHT NOW. If things calm, I might not be as vehemently against it, but the cauldron that is the internet world is not exactly a good and safe place for such a concept, regardless of intent. Tweak an algorithm based on photos, encourage more photos of the actual author instead of graphic imitations, and do things that way.

Same with gender, right? But gender worked on LT. We aren't overrun with people turning trans people's gender around. We haven't had lots of vandalism. We have some simple rules, and members who care. Such few disagreements as we had got worked out. It worked. It works.

As for "tweak an algorithm based on photos" are you seriously suggesting I make a machine-learning person-of-color spotter? And this would be better?

At best, I think a social identities field will tell us more about the person doing the identifying than it does about an author.

If it's added, I won't be contributing to it.

>71 timspalding: Wow. Humanities has changed since grad school.
Why yes, yes it has.

>61 Bookmarque: What's wrong with wokeness and anti-racism?

I don't know, jjwilson61, is there?

I have to say that I feel uncomfortable with making this a global data type editable by LT members. There are just too many loaded and subjective decisions related to the data. Any decisions made about whether or not a person is Black, Asian, African-American, etc. can ONLY be made by verifiable self-identification by an author, and we just don't get that very often. Asking or allowing members to extricate that information from public sources is going to be problematic at best.

What I mean by this is very few authors are going to publicly and officially self-identify as some group unless they have some reason to do so (maybe the success of the book depends on it or gives it some sort of caché?). The vast majority of authors just are what they are and give no thought to mentioning it in any official capacity.

And even if they did, at some point, mention some self-identification they might not want that information to be used to "identify" them in a more general sense. They might not WANT to be known as the "Asian" or "Black" or "Activist" or "Insert Category Here" author even if they did mention it once for some reason.

And to be clear: we are not trying to gather or use this information in any nefarious ways. We think it would be useful and quite interesting for various features, both new and old. If we could actually get valid data I believe that we could do some really interesting things with it. But I'm unconvinced about the validity of the data we could get.

While I don't disagree with others that there are security and abuse concerns, even in a better world this functionally would not work. Gender is not the same as other social identities, and analysis of gender on LT does not work for anything other than male and female. And even those two options only function because they are prescribed (by the example text below the field) so that users enter "male" and not "m" or "man" or "boy" or "he/him", none of which are incorrect or abuse of the system, but would make analysis impossible.

>60 timspalding:
This is author tags

This is not author tags, and the reasons author tags would be better are the reasons you list. Tags about truth emerge from large quantities of data. Tags can't be changed by others. We could be able to say "Looks like Neil deGrasse Tyson is black because he's been tagged as African-American 21 times, Black 19 times, BIPOC 8 times, Afro-Caribbean 2 times, and white 1 time" if we were able to tag authors.

That's not going to get us any sort of analysis

What you're proposing won't get us any kind of analysis either. What good is a list of Mexican-American writers if it doesn't include Rudolfo Anaya and Sandra Cisneros because he identified as Chicano and she identifies as Chicana?

We can pull in Wikipedia

That would be a significant improvement.

Re: author lists, I don't see a difference

There is not much of a difference when it comes to security and abuse concerns (a separate but important issue), but there is a big difference between automatic analysis and a hand-curated list when it comes to data accuracy. I, a human being acting in good faith, can include Rudolfo Anaya and Sandra Cisneros on my list of Mexican-American writers because I believe that "Mexican-American", "Chicano", and "Chicana" are similar enough for the purposes of this list. Debatably I might leave Diana Gabaldon off of my list, though other people might choose to include her.

>76 conceptDawg: My concern is that the well meaning gathering is used for nefarious means by OTHERS, not necessarily by LT itself.
Those with good intent are frequently used and abused by those with other intent.

>76 conceptDawg:

First off, absolutely self-identification should be critical and decisive. I'm not inclined to take it to a Rachel Dolezal level, but that's an edge case. At the same time, this data is hardly rare or hidden. Sites of every kind, from Wikipedia to author and publisher pages, regularly describe authors' identities. I don't really see why, if Wikipedia's first sentence is "Rebecca F. Kuang is a Chinese-American fantasy writer" we can't reasonably and provisionally infer that Rebecca F. Kuang is a Chinese-American fantasy writer. That anti-Asian Nazis are going to take this info from obscure LibraryThing not Wikipedia or the 1,000 other sites that say so seems far fetched to me.

I also think that lists like 36 Science Fiction And Fantasy Books From AAPI Authors That Are Impossible To Put Down (Buzzfeed), 32 YA Sci-Fi & Fantasy Books by Asian American Authors, 5 Asian & Asian-American Sci-Fi/Fantasy Authors You Should Know, and on and on testify to interest in exploring the world of books in this way. I think members would be interested in doing that, and in seeing the library broken down by various identities, generally and so they can choose to read authors with identities they don't usually read from.

>74 jjwilson61:

I find it darkly humorous that the "anti-anti-racist" and the "we can't stop the racists" have joined forces to kill the ability to promote authors from diverse backgrounds.

>77 norabelle414: and analysis of gender on LT does not work for anything other than male and female

You can see it all broken out for your library. Yes, you need to decide how and if to think about the categories, but that would always be true.

What good is a list of Mexican-American writers if it doesn't include Rudolfo Anaya and Sandra Cisneros because he identified as Chicano and she identifies as Chicana?

Multiple authoritative sites online, including author bios, use the term Mexican American for both them. Wikipedia and Wikidata put them in categories too. We can in any case make it possible to combine searches for terms, such as Chicano and Chicana (e.g., Chican*).

Firstly, I didn't make this about race. Tim did with his post up there and the photos. I asked about data, why here and the purpose. Instead I got evasions and name calling -

"I find it darkly humorous that the "anti-anti-racist" and the "we can't stop the racists" have joined forces to kill the ability to promote authors from diverse backgrounds."

If that's how you're going to argue your corner, that's your affair, but it isn't effective, although it is darkly humorous. I'm not a racist and calling me one, even obliquely, won't make it so or change my mind about that.

And now posts are disappearing. Wow.

Posts are disappearing? (Edit: This is Tim)

timspalding (#57):

But if we source all information we can eliminate the small chance that some member uses a right-wing hate group for their information.

I didn't make my point clear; what some people may be concerned about is hate groups, or hateful individuals, or oppressive governments, using LT as a source of information.

I didn't make my point clear; what some people may be concerned about is hate groups, or hateful individuals, or oppressive governments, using LT as a source of information.

If Wikipedia says "Rebecca F. Kuang is a Chinese-American fantasy writer" and we put down that Rebecca F. Kuang is a Chinese-American fantasy writer, you're concerned that right-wing hate groups will target Kuang based on LibraryThing?

I am proposing that everything needs to be sourced, and that we take some time to define what sort of sources are okay to use. Wikipedia doesn't allow "original research" either.

>79 timspalding:

Why again are you willing to deny Rachel Dolezal the right of self-identification?

>84 timspalding: So if Wikipedia says Chinese-American and her site says "American" or "Asian-American" or does not say anything at all, which one do we use for our authoritative list of identities? And what do you do when someone decides to clear up their own charts and changes Chinese-American to Asian-American (which is technically not wrong) instead of adding it (because they do not realize they can add or that they are changing everyone's data or because they think it is really the same)?

>84 timspalding: Are we really going to claim Wikipedia as an acceptable source for this?

The Wikipedia page for Javier Zamora, one of the authors pictured in >51 timspalding: describes him as Salvadoran American, but his official website bio does not use that term. I'm sure we would have no problem compiling a long list of other authors who do not identify themselves the same way Wikipedia does.

>86 AnnieMod:

There are various ways of handing equivalent and semi-equivalent terms. I would generally favor members working through that. I suspect they'll end up deciding that anyone who's Chinese American is also Asian American. But we could also handle it in software, allowing for searches of multiple identities to be combined.

As for "what happens when people screw up" the answer is the 15,437,568 edits CK has already had. People do screw up. And other people notice it, and fix it. If members don't think the tools are good enough for that, we can talk about improvements. But there's nothing here that isn't already true for all the other CK fields.

>84 timspalding: Are we really going to claim Wikipedia as an acceptable source for this?

Absent other information, I think we can indeed consider it acceptable. Members regularly use Wikipedia for information on authors.

As for the term, he was a fellow at Radcliffe 2018-2019. His bio is here: https://www.radcliffe.harvard.edu/people/javier-zamora and uses "Salvadoran American." I don't think he objects to his official Harvard bio. Indeed, he probably wrote it. However, if you feel that both Wikipedia and his official Harvard fellow biography is wrong, you could delete the term.

This topic has had 21 participants. Several of us are regular contributors to CK. Has so much as one participant said they'll directly contribute to this social identity field?

It seems like the one person who is (rather rudely*) pushing this project is the only participant I'm sure I haven't seen in the Helpers activity entering CK data.

* By the way, the site owner labeling some of the site members who are not only active in CK edits but among the most active in promoting diverse book recommendations "the 'anti-anti-racist' and the 'we can't stop the racists'" strengthened my feeling that this whole thing is a bad idea.

>79 timspalding: Should we understand from your answer to >74 jjwilson61: that this proposed data field is designed to promote authors from diverse backgrounds?

>91 abbottthomas:

Yes. The ability to see your library by identity and to see genres and so forth faceted by identity are good ways to promote authors from diverse backgrounds. I favor this as a goal, and there is clear demand out there for such lists.

>90 aspirit:

I'm sorry, but I'm not fond of "So it's virtue signaling so that LT's popular authors photo list looks right according to today's ideas of wokeness and anti-racism?" (How do you feel about that, may I ask?) I'm not doing this to "virtue signal"—a gross and rude term—but because I think promoting authors from diverse backgrounds is a good thing which, incidentally, members will actually use and enjoy. If this is "woke," well, then I and the site are "woke" indeed.

As for the notion that this can't work because of vandalism, I don't buy it. We haven't seen much of that in other fields—not even gender, which is a particular flash-point right now. As for the notion that right-wingers will use LibraryThing to target minorities, rather than Wikipedia, or publisher and author pages with the same information, not to mention a thousand BookRiot posts, I don't find it very convincing.

>92 timspalding: Tim, we've given you reasons other than potential vandalism why this idea could be anything from difficult to disastrous.

>92 timspalding: How do you feel about that, may I ask?

Conflicted.

>92 timspalding: this can't work because of vandalism

>93 amanda4242: Yes, we have.

Tim, I feel that you're neither comprehending what's being said here nor why we're saying it.

It's fairly clear by now that you expect LT helpers to give you want you want while you ignore our related efforts that have lower personal risks for more community gain.

Of course you're free to do what you want. We'll make our decisions, too.

>79 timspalding:

Hmm Aliette de Bodard is on the Buzzfeed list. Although she was born in New York I seriously doubt that she identifies as AAPI (Asian American and Pacific Islander). Her parents are French and Vietnamese and she has lived in France for most of her life and French is her first language.

Wikipedia states that she is French-American which also doesn't seem to be right.

Aliette has stated previously that she is Franco-Vietnamese

I think that this shows that miscategorisation is easy. Even when there is well sourced (panels and interviews) which document how the author sees their own identity.

I was skeptical as soon as I saw this topic and the first few comments yesterday -- my spontaneous reactions were "This is such an American feature" and "To use an American expression, I'm not touching this with a ten-foot pole."

But I held off on commenting until I could gauge the reception of this feature -- I'm a white European, so what do I know? My instinctive recoiling at the thought of a systematic, user-generated and fine-grained racial/ethnic/disability categorization scheme for minority authors may be my personal ignorance.

I've since decided to amend my stance to "This feature is wildly American" and "I'm not touching this with a three-metre pole."

Wildly American, because, as examples of the "authors from diverse backgrounds", Tim has mentioned: Chinese-American, Salvadoran-American, Asian-American, African-American, other X-American, Latinx, and a few others. Speaking as a non-US-ian, that's a very, well, American perspective.

Less facetiously, this is wildly American to me because the X-American-style identities used as examples in this thread may not translate well to other parts of the world. (This has been pointed out by others: >5 AnnieMod:, >18 AnnieMod:, >23 amanda4242:). Speaking as an EU citizen, official national and EU-wide statistics simply don't reflect racial and/or ethnic information the way North America does; in fact, in several EU member states, requesting such information on government forms is straight-up illegal. Instead, nationality or regional or linguistic groupings are used. (Whether that is useful or counter-productive is a separate discussion.) Of course, nationality/ies is already a CK field (Example), and regional and/or linguistic affiliation aren't necessarily part of an author's "social identities."

And that, I think, is why I feel such instinctive reluctance to assist in populating a "social identities" field -- even with a three-metre pole: the messy nature of the relationship between heritage and regional/linguistic affiliation on the one hand and identities on the other. Like, I'm sure that it would be nice and useful and diverse to compile lists of authors who are Chinese-American, or Berber, or Provençal, or Hungarian-speaking Romanians, or Yoruba, or Igbo. Or, say, disabled science-fiction authors. And sometimes, these labels may even be relatively unambiguous and uncontroversial in a biographical context, such as introductory paragraphs on Wikipedia. But to translate national or geographical labels directly into identity labels seems completely unwarranted to me. To what extent is any person that satisfies the criterion Chinese-American / Berber / Yoruba / disabled / ... a proponent or a spokesperson of that group? Is that criterion meaningful in talking about that author? Is Chinese-American a label for geography / nationality or for ethnicity? Does that make a difference for the "social identities" of the author in question?

Those are deliberations that I will not be making. If this feature is rolled out (and it sounds like Tim's mind is made up), I will not be contributing to it.

Promoting diversity through more or less carefully curated lists is one thing; large-scale volunteer-contributed data entry with, apparently, nested hierarchies (>88 timspalding: "I suspect they'll end up deciding that anyone who's Chinese American is also Asian American") and automated conflation ("searches of multiple identities to be combined") is something entirely different. I don't think you can substitute the latter for the former.

Okay, I call uncle. We will be skipping this feature.

Thank you for your candor. I remain convinced that it's a good idea. I think members would enjoy analyses and breakouts of their libraries by ethnicity and other identities, and faceting tags, genres and other headings by them as well. I also think it's good if trending lists and other such are conscious of diversity, because the alternative is all too often extremely non-diverse. I think all of this would serve to promote reading authors from disadvantaged backgrounds, or who otherwise differ from what we normally read. But clearly members don't agree, or don't think such things can or should be arrived at in a fielded wiki.

I reserve the right to reintroduce this idea at some point, or, more likely, to bring in data from Wikipedia categories, from Wikidata or to get staff members to do it, without member input. But for now the idea of recording author identities, apart from those already recorded in the other fields, is dead.

I really appreciate what you tried to do here, Tim.

Well thanks for that. Sigh.

>99 timspalding: I understand it was well meant. And I'm not drastically opposed. But this feature would also be wide open to abuse. And some of us need to be more careful than others.

I actually do like the idea behind this field but suspect that the data are ultimately too messy to make it work. Too many different things to put in a free text field, no practical way to search Common Knowledge for all the possible inputs.

I'm glad this proposal did not go through.

I don't use Wikipedia for anything but checking when some model of iPod was discontinued by Apple or something like that. I avoid it for biography, history, literature, philosophy, etc., because Wikipedia is by this time fully corrupted.

I'ld have no objection to free form tags for authors just as there are tags for books. I ignore the tags for books, I'm just not interested in those sorts of one-word opinions. I expect I would ignore the tags for authors as well.

What I do pay some attention to is recommendations by other LT users. I have often wondered why one can't do author and series recommendations as well as book recommendations.

Just saying.