The Importance of Siblings

[melannen]

So, as part of Hugos, I have read a lot more different original-worlds in a very short time than I have for awhile, so here is a worldbuilding rant aimed at Most Of The People Who Wrote Secondary-World Stories I Read:

If you are writing a world with a high death rate among young people - whether it's due to disease, war, accidents, space vampires, cage-fighting, spectacularly failing the magical initiation, your paper-and-balsa-wood wings failing at a terrible moment, being dragged away by Mysterious Masked Figures, whatever - your characters should not all be only children.

Even in a world with a low death rate among the young, you need a birthrate of slightly-more-than-2-per-uterus-having-person for a stable population of humans. With a high death rate, you need a birthrate substantially higher than 2-per. (Also, dear stories I read for the Hugos, if your base family unit is, say, triads of women, your average family size does in fact need to be six kids. If it's two men and one women, then a, what are you doing to all your girlchildren to get that gender ratio, and b, your average family size should be at minimum three kids. Or possibly two boys, a girl, and a baby girl you exposed at birth.)

If you're picking off a significant number of your young people before they breed, and none of them have siblings to carry on the line, your society is not going to last very long. With a high-death-rate society, two kids should be a small family. And they should have a lot of first cousins. A lot of them will be dead at a young age, but hey, you're the one who decided to build a high-mortality society.

(You also need your characters to have lots of siblings and cousins if your society is actively expanding into new land; if your society has a lot of outward migration; or if rampant population growth is a plot point of your story. Also, if a lot of your people of childbearing biology don't ever have children, the ones who do have kids need to have more kids to cover for the childless ones, so people will still have siblings even if they don't have kids.)

You can get away with fewer siblings if your culture is getting a lot of inward migration - if it's a city that's growing due to migration from the countryside, for example; or if it's an elite group of warrior-mages who bring in talented people from the outside population. But in that case, your society needs to show evidence that it's structured to take in a lot of people raised outside its culture, and your people from the outside population need to have siblings, since their culture is supporting both their own population growth and the elite group's.

Other than that, if all your characters are only children of only children, a rapidly shrinking population had better be a plot point.

(obviously if you go all-out to, idk, growing your characters in vats from gene-banks, or if they are left on doorsteps by storks, or if you have 0 mortality, all of this will apply very differently, but it'll still apply somehow.)

I get the reasons why authors write so many characters from small families. Most people getting published in English these days are from societies where small families are the norm, and especially in YA, are writing for an audience who are presumed to be from smaller families. (Although this is less true than you might think: a lot of modern low-birth-rate populations are doing the 'a lot of women don't have kids, the ones who do have large families' thing instead of 'everyone's an only child'.)

And it also greatly simplifies your storytelling not to have all those extraneous relations hanging around. And having your main character being the Last of Their Line is built-in Drama. So make your main character in a high-mortality society an only child with no cousins if you must. But they had better be an exceptional case, and your secondary characters better have siblings and cousins and aunts and uncles. And there had better be in-story justifications for why they are an exception.

And 'one of their parents died or disappeared when they were a baby' is not a good enough justification; in a high-mortality society there will be strong incentives for a single parent of reproductive age to remarry and give them half-siblings.

Bujold's Barrayar is a good example of how to do this well: Barrayar, at the time of Miles' birth, is a fairly high-mortality society between war and poor health care, and it's also actively settling a new continent and about to colonize a new planet and expanding militarily, and yet her three main characters in that generation, Miles, Ivan, and Gregor, are all only children with no first cousins. But there are story-relevant reasons for this: Ivan and Miles' fathers did have siblings, but they were all dead before marriage; Miles' parents want more kids and are under family pressure to have more, but know it would screw things up for Miles and Gregor if they did, so they resist the pressure; Gregor's mom actually was going to remarry and have more kids but got murdered first; Ivan's mother is under social pressure to remarry but resists it and is considered exceptional. Plus, Ivan's mother and Miles' mother do both start new relationships at an age when they could still have kids, though I haven't read the newest book so don't spoil me on whether they have yet in canon. And Miles actually does have maternal cousins, they just aren't a major part of his life for political reasons. And pretty much all of the other Barrayaran characters whose families we know about do have multiple siblings, and people constantly remark on the fact that it's weird that Miles and Ivan and Gregor don't. So Bujold gets away with it.

Bad example: Naruto. Okay I'm only about 1/4 of the way in, so maybe it gets better as Kishimoto expands the world, but so far we have a society where most people assume they will be violently dead by age thirty, and where there's a cultural emphasis on large lineal clans carrying forward certain bloodlines, and yet nearly all the main characters are only children, and the ones who aren't have only a single sibling. There is only one family so far with as many as three siblings (and the third sibling is treated as an unusual extra), and nobody has more than one aunt or uncle, either. And most of the characters who are in their twenties have no kids. Main character Naruto gets away with having no siblings because his parents both died the day he was born. But the fact that he has no aunts, uncles, or first cousins, and nobody finds this unusual, and the fact that nearly everyone he knows, even the ones who are supposed to part of large powerful sprawling clans, have at most one sibling and generally none, is bothering me more and more the longer I read.

So, conclusion: If you are killing a lot of characters, make sure most of them have siblings, and at least some of them have lots of siblings. You can have most of the siblings be dead or in another country if you need them out of the way, but they should at least exist.

Comments

[alexseanchai]

Your rants are always A+ educational. :)

[melannen]

If one fewer person accidentally writes their society out of mathematical existence, I have done good. ^_^

[beccaelizabeth]

good thoughts

[melannen]

thanks!

[watersword]

This drives me bonkers.

Also, it's more-or-less commonplace in societies with a high infant mortality ratio to re-use names from deceased older siblings -- which I don't think I've ever seen in fiction.

[the_rck]

I've only seen it once-- in the Flora Segunda series by Ysabeau Wilce.

I think that it's not likely to be used frequently in modern, English language fiction because readers get confused by names that are too similar. I read a book once with a large cast of characters, and half of them had names that started with A. I got to the end of the book and still had an awful time tracking which A name went with which bits of the story. (It was a terrible book for other reasons which can't have helped, but I particularly noticed the names as a problem.)

[melannen]

Most of the stuff I've read have higher young adult/adolescent mortality than baby/toddler mortality, which, the realism there is a whole nother rant.

Although it is really hard to write a high actual-infant mortality-rate society in a way that feels emotionally real to modern readers, and it is a lot easier to fanwank an explanation for why your low-tech society has low infant mortality that than it is to get around the inexorable math of not enough siblings to cover the vampire deaths, so I let that pass usually.

[jjhunter]

When you read Bujold's latest book, I would be interested in what you think of it very much in lens of your points here.

[melannen]

Ok, now you've got me worried that she ruined it on the new book and I am all wrong. :p I will cover my ass and note that the Barrayar of miles' adulthood has a very different mortality profile from the one he was born in, so all this stuff is changing for modern barrayar.

(My library holds were all set to reactivate after hugo noms were over so I should be getting it very soon)

[brownbetty]

If you have an all female society that reproduces by selfing, or parthenogenesis or something, you could probably get by with a rate of about 1.5?

I mean, yes. Good point.

[melannen]

If you have a society where everybody is equally capable of bearing childen (lesbian separatist utopia, mpreg verse, Imperial Radch, etc.) then you only need 1-and-a-bit children per person who can bear children (and the and-a-bit is only there as a fudge factor; if you've got really low mortality you don't even need that.)

Basically, the replacement rate is actually 1:1 in pretty much every population: everybody dies exactly once, so everybody needs 1 replacement. With a high-youth-mortality rate, some people are dying before they can create their own replacement, so the survivors need to cover their replacements too, thus: lots of siblings for everyone.

(If you change the rule that everybody dies exactly once, the 1:1 ratio changes, but what happens to population dynamics when you mess with mortality existing AT ALL is a whole nother rant. And something people screw up even more.)

Population demographers complicate things by counting the birthrate as per "woman" (that is, adult-person-hypothetically-capable-of-bearing-children) rather than per person, which simplifies the stats-gathering in some ways, and also presumably stops people feeling funny about talking about men birthing. But that means you have to double your replacement number since you're only counting half the population as your base. (The slightly-more-than-2 rule is a fudge factor for messy reality, like the fact that the sex binary is not actually binary, and because they usually don't include girls who died in childhood in the averages for practical reasons; in hypothetical worldbuilding you can mostly ignore it.)

If you mess with the 'one man one woman required to make a baby' rule, the replacement birthrate becomes the inverse of the proportion of people who can birth children to the total population.

Which sounds complicated but really isn't: if 1/2 of your population can birth children, it's 2-children-per-birthing-parent. If 1/1 of your people can birth children, even if it requires six genetic donors to get them pregnant, it's 1-child-per-birthing-parent. If only 1/1000 people are queen mothers, they each have to have 1000 children. If you have three genders and two of them can birth children, but the gender proportion is 2:2:1, then 3/5 of your population can birth children, and the replacement rate per childbearing person is 5/3.

If you are in an a/b/o verse and have six sexes of which four can bear children, and all six are equally common in the population, then the average number of children per person who can bear children needs to be at least 6/4, or 2/3. But you can't just then say that the average number of children per couple should be 2/3, because some of your couples will have two child-bearing-capable parents in them, and that brings the overall average kids per couple back to 2.

So in a society organized around subunits of adults who are raising children - i.e., some parents and all their kids, where the vast majority of adults are parenting as part of a nuclear family - the shortcut is that for a stable population, the average number of kids being raised by a family should be the same as the number of adults parenting them, regardless of which, if any, of the adults were genetic or wombspace contributors.

So in your lesbian separatist parthenogetic society, the replacement birthrate is one child per woman, but if most women are living in couples, for a stable population, most couples should have two kids. And if most women are living in polyamorous communal farms of about a dozen parent-age women, there should be about a dozen kids in them too.

(If lots of adults aren't part of the childraising structure - say, your lesbian separatists aren't parthenogenetic, they just exile all the boychildren into the wildlands and then sex up the wildmen once a year at the festival - then your families-with-kids need to have extra kids to be replacements for the men out in the wildlands as well, so a non-parthenogenetic lesbian separatist commune of 12 parental-generation women should have about 12 girlchildren for their replacements, *and* twelve boychildren which they eventually exile.)

[staranise]

I've also noticed that in fantasy there's often an aversion to established relationships. Someone has to Go Somewhere New and Meet New People! It's a lot rarer to have a fantasy story that happens, say, within a mile of where someone was born, full of people they've known since childhood. Even though I think that could be super cool.

[melannen]

I am in favor of more established relationships in general.

And yes! More fantasy that is entirely set in one village. That is definitely a genre that is unexplored (and yes, a large part of the reason why family relationships often get shafted in the worldbuilding.)

[luzula]

You could read Jo Walton's Lifelode! That's pretty much domestic fantasy with established relationships all set in one village. From what I remember, someone comes in from outside, but most people are from the village.

[umadoshi]

*so much love for this post*

[melannen]

Thanks!

[beatrice_otter]

I have a couple of world-building rants in me, about other things, more the stuff that tends to pop up in Regency-Victorian-early20thCentury fics and stories, because that's what I tend to read more of. Like, class relations. No, if there are class-based social strata, most people are NOT going to find respectful forms of address based on those strata to be offputting or weird. And, in fact, they existed for a darn good reason, and the lower classes had a REALLY GOOD reason for preferring them, because it was PROTECTION for them against the captriciousness of their "betters." (Yes, sure, they want you to be all chummy now--but if you say something they don't like ...) I want to read a story where the hero insists on/is quite comfortable with the titles and rank and stuff, and the villain is the one who (some of the time) goes for more egalitarian we're-all-friends stuff.

Also clothing. Good heavens, clothing. Acceptable clothing for different circumstances has changed SO MUCH in the last 50 years, and what it all MEANS, you can't just project current attitudes backwards!

And Regency slash-fic. I know that arranged marriage for financial reasons is fun in Regency-fic, and if the slashers want to have fun with it they should, but if so they need MPREG or some other way for this m/m couple to bear kids, because heir-and-a-spare WAS A THING. If you have a fortune, you need someone to pass it on to! Your family is depending on you! Your ancestors are looking down at you disapprovingly from their portraits in the family gallery if you don't!

But, anyway, thank you for this, I love your worldbuilding rants.

[melannen]

I would read those rants!

Although historical/cultural stuff is always really hard to make positive statements about, because in no time and no place was culture ever monolithic, so what an upstairs maid at Downton Abbey in 1890 prefers (for good reason!) is not going to be at all the same as what a hotel maid in Deadwood, Nevada the same year expects. And may not even be the same thing a stableboy at Downton Abbey considers respectful, because stableboys are frequently anarchists. Because class is super-complicated. And it's just as easy to go way too far down the paternalism road as it is to be anachronistically egalitarian. Generally I want a hero who is willing to go with whatever the less-powerful person expects, and a villain who expects everybody to go with what they want regardless.

Clothing is even more complicated. This is why I'm sticking to a rant about basic math. :P

[petra]

I'm glad I'm not the only one who finds m/m Regency stories deeply weird because of the lack of heirs.

[untonuggan]

this is a timely reminder, thanks.

*adds a scene about childhood fever for the character I forgot to do so for*

[melannen]

Always glad to provide a service! Especially if it results in dead fictional children!

[vass]

If it's two men and one women, then a, what are you doing to all your girlchildren to get that gender ratio, and b, your average family size should be at minimum three kids. Or possibly two boys, a girl, and a baby girl you exposed at birth.

Nope, you didn't succeed in making Baru Cormorant any more depressing than it already was. :DDD

[melannen]

Oh, darn, I was hoping there was a non-depressing explanation for that later in the book after I gave up on it.

(I mean, if literally half the women were pairing off as lesbians or sth, it would work, but there was no evidence of that up to the point when Baru sailed out, and you'd think tiny Baru would've noticed it at some point, given the givens.)

(And she still needed at least two siblings, yes even if one of her dads died.)

[dhampyresa]

This is a very interesting post.

I'm wondering about the number of families with X children, if Y is the average/replacement rate. For example, if my replacement rate is 2.1, how many children are only children and how many have two or more siblings?

[melannen]

It depends. With a replacement rate of 2 (I'm doing that just 'cause the math comes out cleaner, but it doesn't make a huge difference) you could have a society where everybody has two kids, or a society where half the population has 2 kids, and 1/4 have 0 kids, and 1/4 have 4 kids. Or a society where 3/4 the population has 1 kid, and 1/4 has five kids.

But your average still has to be 2, and so for every family with less than 2 kids, there has to be a family with more than 2 kids. And the families with more than 2 kids will have more kids in them, so it actually comes out that the more families you have that are below the average birthrate, the more kids are from families with more than the average number of kids.

This may seem paradoxical, so here's an example with numbers:
.25% of women have 1 kid.
.5% of women have 2 kids.
.25% of women have 3 kids.

This comes out right - .25 + .25 + .5 = 1 family, and .25 * 1 + .5 * 2 + .25 * 3 = 2 kids.

But because the large families have more kids per family, when you look at number of siblings per kid instead of kid per mom, you get:
.125 of kids have no siblings
.5 of kids have 1 sibling
.375 of kids have 2 siblings.
--so the average number of siblings per kid is actually 1.25, even though the average kids per family is still 2.

If your % of hypothetically-childbearing people with less than two kids is greater than 50%, then over 75% of your children need to come from families with 3+ kids in order to make an average of two kids per family. In the example where 3/4 of families have 1 kid and 1/4 have five, the average siblings per kid is actually 2, and about 2/3 of kids are from large families.

This is why fantasy worlds with mostly only children and no large families worry me.


Basically, birthrate = average number of kids per childbearing person, so for every family with less than the average, there needs to be more kids born in a family with more than average, to balance out to the average.

But the larger-than-average families have more kids per family, so siblings from those families get counted more than once in the sibling-per-kids count, since they are siblings to more than one other kid, so the more kids in a large families, the more siblings-per kids. And the more small families, the more kids in large families are needed to balance them, and the larger the imbalance in average kids per family vs. average siblings per kid. This generalizes to any birthrate or distribution of family sizes.

...I would write out the general equations but I don't have the math for it today, and yet would rather try to work out the math from scratch than try to remember enough to google it, because it's that kind of day...