Interesting and thought-provoking analysis -- thanks for posting it!

Posted by: Jed at March 16, 2009 07:13 PM

The things you categorize as Things You've Learned seem to be statistical observations about stories you've written in the past. Is there some Thing You've Learned that's more of a prescription for something you want to do/try/change/emphasize in the future?

Posted by: Jim Moskowitz at March 16, 2009 11:33 PM

Good things to keep in mind in the future, but for what it's worth, I didn't read Kadath-Naan in that way.

Kadath-Naan did appear to have distinct motivation from the main character, and wasn't participating in the story solely to aid the main character. He even enters and leaves the narrative on his own terms.

Moreover, the main character joins Kadath-Naan's kindred at the end of the story, which I'm not certain would line up with the pigeon hole you mentioned.

It's a difficult problem, in writing sci-fi or similar tales: how can one avoid previous stereotypes of the "Other", when part of the appeal is the unknown, the undiscovered, the (literal) alien?

Though I loved "Siege of Cranes", it did contain the single case of potential "racefail" that actually occurred to me as I read the collection: the white witch/Maghd as a potential Romani fail.

Posted by: jamesG at March 17, 2009 07:26 AM

Jim, I think the interesting part of doing statistics is to see the aggregate picture emerge. Because in any given case you may say "well, I'm doing this for such-and-such a reason" and it may be entirely reasonable; but the aggregate picture can show something different.

Writing is, for me, an intuitive, not-entirely-conscious process not entirely amenable to conscious steering. So I don't think I can sit down in front of a story and say "my goodness, I am under quota for characters of color, let's put some in this story!" First, because how characters emerge is a subtle, organic process: they teach me about who they are as they emerge onto the page. Second, because I'd be a little suspicious of that as a motive. To write about a character is to fall in love with them a little -- even the horrible ones. No one wants to be fallen in love with on purpose because of their demographic characteristics.

But what I think you can do is steep yourself in certain kinds of knowledge, and develop your instincts, your intuitive feeling for things. By confronting yourself with realities, you can get a little farther past that within you which obscures it.

My goal here is to be a better writer -- not to fall into easy cliches, or to depict things a certain way simply because that's what I've absorbed, or to unthinkingly perpetuate the lies and taboos of my culture, or to fail to get across to the reader the way things are in my head. To dream more vivid and fresher dreams and to convey them more fully. You know?

So my hope is that my observing the pattern, I'll be more likely in the future to question it. I'll be more likely to feel "something is wrong with this picture..." if all my characters are homogenous. I'll be less likely to stumble into a cliche, or a depiction which has emotional power for historical reasons which, if I stopped to analyze them, I would find sordid or horrific. I'll be more likely, as part of the creative process as opposed to on top of it, to say "wait, what if she were Asian?" or "hold on, isn't this kind of suspect here?"

I don't know if that's a very good answer to your question.

Posted by: Benjamin Rosenbaum at March 18, 2009 02:28 PM

jamesG, I'm glad Kadath-Naan didn't come across for you that way. You're right about some of the mitigating factors -- K-N does have his own, distinct, agenda, and the fact that Marish ends up joining up with his people, and on the grave-digger level, does give it a twist. Nonetheless, on balance, I have to say I don't think it entirely avoids the cliche: but everyone gets to have their own reading. :-)

Wow, I didn't get Maghd as Romani at all -- though that may be my ignorance of Roma stuff. I assumed she was the same ethnicity as the other inhabitants of Ilmak Dale. Do you know what cues contributed to this identification for you? Is "Maghd" a Romani name?

I did actually think a lot about Maghd's arc in terms of sexist (and maybe mental-health) fail -- the crazy, persecuted girl who lives on the edge of the village does in fact turn out to be a dangerous witch! This was something I wrestled a lot with in revision, trying to give Maghd more levels (and trying to make Marish more complicit in her downfall, which somehow helped, I found).

Posted by: Benjamin Rosenbaum at March 18, 2009 02:34 PM

I'm not so sure about the idea that being "unmarked" means that white is the obvious assumption. Doesn't/shouldn't not mentioning race mean that it is ambiguous? I bet some people simply project their own race on to "unmarked" characters.

Posted by: Ethan at March 18, 2009 04:45 PM

That is not what any people of color I've talked to report, Ethan. The assumption of whiteness is pretty universal, and powerful. As Mary Anne says in her first post at Scalzi's place, "Go on, test this. Pick up your favorite older science fiction and fantasy off the shelves, with those unmarked main characters, and re-read the first chapter — did you actually picture any of them as a bisexual Asian woman? I’d be very surprised." (The whole post, which expounds on the point at length, is worth reading). And she isn't just addressing white folks here -- no American, of any ethnicity, just naturally pictures Hari Seldon or Susan Calvin as black or Korean.

This is also powerfully attested to by the first reaction of so many people of color from previous generations when they did first encounter a character marked explicitly as nonwhite. Samuel Delany writes about how much it meant to him when a character in Starship Troopers looks in the mirror mid-book and turns out to be black. It blew his mind. He was not just assuming black because otherwise unspecified.

Now, I expect if you are a member of an ethnic majority in a relatively large country, and consume mostly media created there, and you're reading about something totally unmarked, like in the far future, then maybe. So maybe if you're Japanese, or Han Chinese, or Hindi-speaking Indian, and you consume mostly movies and books and TV made in-country, and you read Foundation in translation -- then, yes, maybe you'd imagine Hari Seldon as Japanese or Han Chinese or Hindi-speaking Indian. Especially if somehow it hadn't been made clear to you that it was originally an American book.

But that's hardly the minority experience. Trust me, I don't watch Star Trek and assume all the characters are Jewish -- unless they make the Cohen blessing hand move (and probably not even then)... or are big-eared, big-nosed, greedy, craven, untrustworthy merchants.

Posted by: Benjamin Rosenbaum at March 18, 2009 06:06 PM

Well, I'll grant you if you have a character named John Baker, it's probably a safe assumption that he's not Japanese or Indian or a woman. And I am not arguing against the idea of having characters whose race, etc. are explicit.

But it feels a bit wrong to me to say something along the lines of, "Unless otherwise specified, it should be assumed all non-specified characters are white & straight."

Posted by: Ethan Ham at March 18, 2009 08:30 PM

I'm far from an expert on Roma (or for that matter Roma stereotypes), but here are the things I picked up on.

Maghd is an outsider to the village.
Maghd is a thief.
Maghd uses magic.
Maghd does small crafts.
Maghd is associated with rags.
Maghd angers quickly and uses magic to "curse" those who drew her anger.
Maghd steals children.

Anyway, I found this site, which corroborates most of those as Roma stereotypes:
https://deviousdiva.com/2007/12/13/roma-stereotypes/

It even potentially adds "dirtiness" and "seduction" as portions of Maghd's behavior that can be interpreted that way. It lacks, however, the association with rags. Maybe I was just wrong about that, or maybe it's just not listed.

That's not to say there aren't mitigating factors:

Maghd doesn't appear to be transient, at least until she goes on her rampage. Though I thought the townspeople treated her a bit harsh for her to be natural-born, I don't think anything in the story rules it out.

Maghd doesn't travel with a group of her own people (if you assume she's not from the village herself).

And, as you mentioned, the "storybook witch" archetype includes many of the same traits. Though my mind did go to "Roma" as I was reading, this mitigation also occurred to me before I finished the story. After all, any storybook witch is going to be an outcast, is going to use magic. It's certainly not uncharacteristic for a witch to steal children. As I don't consider storybook witches to be a slander to any particular people, this is the largest mitigating factor to me.

Posted by: jamesG at March 18, 2009 09:16 PM

A couple more thoughts... I don't want to give the impression that I don't think it's worthwhile to examine your writing in this way. The fact of all explicit LGBTQ characters are comic relief is a good thing to be aware of.

But I think its problematic to define white as being normal. I bet you're careful not to exclusively use male pronouns in situations where the gender is uncertain... You'd hesitate to write something like, "I don't know who the pilot is, but he's doing a great job!" I think the same should go for race...

If black characters need to be specified as black, then it would probably be worthwhile to (at least occasionally) specify white characters as white. Or perhaps it worth trying imagining a character as black or gay or what have you without making it a stated (or implied) part of the character (in the same way white and straight does not necessarily need to be spelled out) and see if it makes a difference in the writing.

Posted by: Ethan Ham at March 18, 2009 09:19 PM

Ethan, I think you are conflating my suspicion about how readers do read with a proposal about how readers should read.

You know that Dilbert cartoon where step three is "PROFIT!" and step two is missing? There is a common tendency in discussions about racism to perform a similar trick of prestidigitation, where step three is a colorblind society and step two is missing. It's sort of like a fear of sympathetic magic: if I assert aloud that readers read unmarked characters as white, it either makes it more likely that they will, or is as if I am suggesting that ought to.

But actually the opposite is the case. I think it would be great if readers did not read unmarked characters as white. And in some cases they may not. But it is an empirical observation that they in fact do.

I said nothing about any John Baker. I mentioned Hari Seldon. That name does not sound particularly white. "Hari" is in fact one of the names of Vishnu, and I bet that if you were at work in an IT company and they said in the daily stand-up meeting "a new employee named Hari Seldon will be joining us Wednesday" your mind would not naturally flash to white guy. So it is notable, as an empirical fact about literature, that when I read Foundation as a kid I imagined Hari Seldon as white, and I will wager so did you (and if you "didn't imagine his race", again, imagine someone had asked you to draw the scene), and I will bet you five dollars, if she's read Foundation, so did Janet: ask her.

Similarly, "Susan Calvin" is a completely plausible African-American woman's name (and John Baker actually is a similarly plausible African-American man's name; most African-Americans descended from slaves in fact have English surnames, so a Jewish or Polish or Dutch surname would be a much clearer marker of whiteness in America than "Calvin" or "Baker"!).

You can tell by introspection that you, as a white reader, do this. It is an interesting hypothesis that perhaps people of other races imagine unmarked characters (for instance those with wholly invented names of no known ethnicity, such as "Zaphod Beeblebrox") as their own color. It is not, however, empirically the case, at least from those people of color -- raised in the U.S. -- I have read or spoken with on the topic. Mary Anne Mohanraj does not. Samuel Delany does not. Tempest Bradford does not. Octavia Butler did not. In fact the common experience reported, at least back in the day when most characters were relatively unmarked, is of bursting into tears upon finally encountering a character specifically marked as nonwhite. That sure does not sound a whole hell of a lot like they were complacently imagining that all the characters were, e.g., black.

If is profoundly problematic to define white as normal. It is in fact the root of tremendous iniquity and evil. It is shooting the messenger, however, to accuse someone pointing out that that is how we have been indoctrinated, of doing the "defining". The defining is already done -- I am trying to resist it. But seeing it is a necessary step in resisting it.

Now, specifying white characters as white is an interesting proposal, and indeed Mary Anne (whose article on this you should really read) notes that you should go beyond that, and in fact mark your white characters as what they are -- half-German half-Lithuanian, or of poor Scotch-Irish hog farming stock from the Tennessee river valley, or second generation Greeks who went from a fishing village to the Bronx, or whatever. Because the assumption of generic whiteness also erases white ethnicity.

And when Mary Anne was preparing the essay which you really should read, and asked for my feedback in email, one of the few things that I quibbled with was the word "mention" in the suggestion that you should "mention" the character's race or ethnicity, because that's often, as a matter of writing technique, clunky. She revised the relevant paragraph to read:

Give your white characters an ethnic and cultural history, even if it ends up barely mentioned in your story. But be sure you do indicate it somehow — it’s not enough for you, the author, to know their history. Unless the reader gets an intimation of it as well, that character might as well be generic white. Names, clothes, hair, foods, diction, holidays, memories, can all signal ethnicity to the reader without a blinking neon sign over their heads saying “fifth-generation Irish-Croat”.

You don't need to state ethnicity to break the empirically observed phenomenon of default whiteness -- but it does need to be on the page somehow, because the page is actually the only medium of communication between the author and the reader. Perhaps this is what you mean -- that if the author imagines the character's race, without ever intentionally signalling it, subtle signals will creep in that will influence the reader to imagine properly. Sure, fine. That's a suggestion of a writing technique. The fact remains though that something -- even if it's some subtlety of diction or dress the author is not fully conscious of -- must have gotten to the page.

To sum up: I am not proposing that authors who wish to have characters read as white comfortably refrain from giving any signal of race. Rather, I am warning that that is how unmarked characters get read. If I was concerned that my stories had perhaps too few white characters in them, then I might have counted differently!

Posted by: Benjamin Rosenbaum at March 19, 2009 12:02 PM

jamesG, thanks for your list. You're right: I think I did blunder straight into that stereotype. I particularly hadn't considered the influence of the fact that Maghd steals children. I mean actually Maghd stole everybody -- but the fact that she handles the children differently, and specifically her conversation about them with Marish at the end, pushes it toward children-stealing as a separate category. Thanks!

(I have been thinking at some point I might write a sequel to "A Siege of Cranes" from Maghd's perspective...)

Posted by: Benjamin Rosenbaum at March 19, 2009 12:16 PM

It seemed quite natural to me to read Maghd as not othered from the village by anything but her mental health, but a mention of some family tie would clearly mark her as an outcast rather than a foreigner. How much did you think about her back story? Perhaps there's a writing-process clue there.

Posted by: David Moles at March 19, 2009 02:41 PM

(Bearing in mind our discussions about our different approaches to characterization, it's tempting to guess that a Ben Rosenbaum character without personal ties also has no back story, much as a David Moles character without markers of ethnicity, geographic origin and class might have no back story...)

Posted by: David Moles at March 19, 2009 02:43 PM

Oh, the backstory in my head was definitely that she was local, ethnically of the same group, orphaned early, "different" mental-health wise, and then made the tactical mistake of trying to win connection and status by being sexually available to the village's men beyond acceptable local norms, which won her only the intense hatred of the women, and the scorn and exploitation of the men ("Bag-Maghd's good for holding one thing only!") and thus ended up totally outcast and in a mood for vengeance.

But as I argued to Ethan above, in my head doesn't count. :-)

Posted by: Benjamin Rosenbaum at March 20, 2009 08:16 AM

(She does have personal ties -- just not such pleasant ones -- and the process of revision we went through after you accepted the manuscript was largely a process of deepening the texture of those relationships, as I recall -- in particular, her mutual jealousy with Tamar and her ambiguous relationship with Marish -- particularly, intensifying the implication that Marish was complicit in exploiting her, not without some conflictedly tender feelings on his part...)

Posted by: Benjamin Rosenbaum at March 20, 2009 10:05 AM

Benjamin, thank you for a very clear demonstration of the issues we white allies are trying to face in ourselves-- and thank you so very much, for your long answer to Ethan Ham, which should generate a post all by itself!

it's really remarkable how visible the fingers-in-the-ears approach can be, when it is not oneself making the mistake... heheh...

Posted by: Stella Omega at March 26, 2009 01:55 AM

Belatedly, this is fascinating--thanks, I've linked it at fight_derailing.

Posted by: Kate Nepveu at April 12, 2009 03:12 PM

Glad you found it useful, Stella & Kate.

Posted by: Benjamin Rosenbaum at November 23, 2009 12:59 PM