marshtide: (Too-ticki)
[personal profile] marshtide
Prompt: I've often wondered how archaeologists determine the sex of a skeleton - I know there are differences in the pelvis, but what else is taken to be a clue? I'm also curious about how accurate it is, and whether anyone's done any studies on that.

I'm afraid I'm focusing more on the general methods used than specific studies - I'm not the only one tackling this prompt so hopefully you'll get a different range of information from the other answer! Basically all my books are in the wrong country & I lack journal subscriptions, so I can't pull up as many specific figures and examples as I'd like for accuracy. But here is the as lay as possible guide to the kinds of techniques that get used in determining the sex of human remains in archaeology.

As a note before we start, I'm also one of those people who dislikes overemphasising sexual difference and I do see this whole thing as a sliding scale rather than a choice between two boxes. I do not automatically connect sex with gender either. That's the perspective I work from, though it's by no means shared by all archaeologists!


There are three things that are often used to determine the sex of skeletons, although they're of variable use and appropriateness. All of these have problems, although number two is the really what were you thinking option. More on this later! The methods are:

1. Looking at the bones
2. Looking at the grave goods
3. DNA testing

1. Looking at the bones

1a. There are a lot of different skeletal clues that can be used in determining the sex of a skeleton. The most well-know is the pelvis, as mentioned in the prompt. The differences here come from actual function, which makes them the best guide when it comes to dealing with the bones. Overall one is looking at things like the width of the pelvis (wider in female skeletons). Names of features which are often used: sciatic notch (wider angle = more female), ventral arc (more pronounced = more female), ischio-pubic ramus (wider = more female). I appreciate that these names don't generally mean much to anyone, but here: have the comparative diagrams from grey's anatomy, which show at least the general idea of what all this is getting at. Broader pelvis: more female. Narrow chunky pelvis: more male. That is to say, could a baby get out through there or not?

Gray242 Gray241

(These, being anatomical diagrams, represent the extremes. Most people's pelvises aren't so far along the spectrum in either direction as these ones!)

One further method, which, uh, doesn't work, is to look for a certain kind of pitting on the surface where the two halves of the pelvis meet. This pitting was thought to indicate childbirth (& therefore a female skeleton). You'll probably see this one around in older reports, but I would hope not so much in recent ones, as in at least one test of the method (spitalfields, see below) the results of this analysis indicated that almost none of the women could possibly have given birth. The church records, on the other hand, indicated that most of them had.

1b. After this the next place to look is the skull, which has a range of different clues; in general, the rule is that a lot of protrusions from the skull are more marked in males than females, although this is a problematic assumption. It's a problematic assumption because, unlike the pelvis, which has the aforementioned functional difference, what the size of some protrusions from the skull can also indicate is muscle mass. I have also seen judgement of sex from the skull described, although I'm afraid I forget where, as boiling down to "how butch does this skull look to you?" ... which doesn't seem completely devoid of truth as a summary, to be honest.

Features which are used on the skull include supraorbital ridges (the ridge of bone above the eye-sockets), which is to say, does this skull have a jutting brow? Then there's the squareness of the jaw - the idea is that more square is more male, in rugged spaghetti western fashion. The mastoid process, a bony structure which sits just behind the ear canal (the bit of bone sticking downward in the diagram below), is similarly judged to be more male if more pronounced. And then there's general size of the skull and, well, to put it bluntly, the lumpiness.


Oh, and you see the way the bone angles up at the back of the jaw? You can use that too, with the idea that a sharper angle is more male.

As mentioned, this isn't as reliable a set of characteristics as the pelvic ones, and I'd definitely prefer to use the pelvis and cross-reference with the skull. Naturally one doesn't always have a complete skeleton to work with, though; bones can be found out of context and burials are very often disturbed by animals or by later burials/ditches/housing foundations being dug through the site. You work with what you have, but the ideal is getting as many of these measurements as possible.

And a final note on skulls - they're fragile. You might well not be able to see many of the traits you're looking for at all.

1c. Another measurement which can be taken is height, which is estimated from the length of a complete longbone, generally the femur (thigh bone). There's a formula for this, but it doesn't give a real living height for the individual, just some kind of approximation. Average heights vary wildly across time periods and areas - for example the early medieval anglo saxons (in England) seem to have been fairly similar in height to modern populations, while later medieval populations became much shorter, because they didn't have as much food - so to use a height estimate to assign sex, you need a range for the time period and population in question to compare it to. Again, there are always tall women and short men, with a considerable overlap where nothing can be learnt.

1d. Then there's measuring bone thickness (thicker = more male). The problem with the thickness of bones is that it varies considerably based on muscle mass, and I don't think we're buying the idea that male individuals are always more muscular than females, are we, so I think you can probably take it from there. (This was illustrated to me in my bones 101 class way back when with x-rays of the right and left arms of a right-handed tennis player who basically had one big beefy arm and one which was visibly less so. I suspect the player in question did not have one male arm and one female arm.) There is however a range to which males tend more towards one of and females more towards the other, as with everything else, so it's a potential one, but I'm not that fond of it.

1e. Anyway, the measurements for all of these types of differentiation, in any case, are recorded as points on a scale. The scale can look something like this:

may be male
may be female

As many of the identifying features as possible should be used and recorded separately, and an overall score can be taken from them. This gives a probable sex for the skeleton overall, although in some cases you may have to squint and in others no amount of squinting will produce a convincing answer.

1f. (it just goes on!) And all of that is for adult skeletons. For non-adult skeletons it's even more complicated! Basically:

Children: oh, forget it. A child is a child; there is basically no difference until one hits adolescence.

Adolescents: You can measure the crown size of the adult teeth in an adolescent skull, although this varies by the group of people you're looking at - the Spitalfields study for example showed very, very little distinction between sexes in this respect. Also, you need to have found teeth that were actually still set fast in the jaw or skull, which, depending on the site, you may have loads of or almost none. They need to be well-preserved too - and this is actually true of all specific width measurements - because if the surface is worn away then the result isn't terribly meaningful (how much surface have you lost? You have no idea). As with height you need to know what the averages for the specific population are.

The other measurement that I know gets used is the pelvis, although this isn't done in the same way as for adults, and I've never actually been called upon to do it, so here I am unfortunately hazy. I believe that the part of the pelvis called the ilium (which you can find on the diagram above) is the most useful for measurement in non-adult skeletons.

1g. So then, let's have a proper list of the potential problems before we get to accuracy:

- to be able to compare traits from different bones you need to be working with a very clearly articulated skeleton, that is to say, one that's been lying in the ground so that you can see it's all one person.

- bones can get knocked around by animal digging, human digging, or high acid content in the soil, to name but a few of many; even if you've got a whole set of bones they might be too damaged for you to be able to say much about them. Sometimes a skeleton is beautiful whole bones with their surfaces intact and sometimes you're trying to put a skull back together like the world's most complicated jigsaw puzzle. With bits missing.

- This sexual dimorphism business has grey zones, so you're not going to get an answer for all the skeletons, even the perfectly preserved ones.

Finally: Accuracy of this method: Well, that really depends, doesn't it. I've seen figures quoted between 60% and 97%, depending on the study and the interpretation of the study, as well as the part of the world - since these things are by no means the same everywhere, just to add some extra confustion. General consensus is that it's Pretty Good Considering, though, especially in cases where you can get a full set of data and fit it into the context of a really well-studied population.

On which note, some of the big studies on accuracy, as far as I remember, have been done on modern or early modern populations, in several cases north American, and how well these results translate around the world is not something I'm completely clear on. In some places there are local studies and in some there aren't, but using standards from anglo-saxon remains around the world? I wouldn't. But I've only worked with anglo-saxon remains, so this is a bit outside my area.

Also, will obviously be more accurate the more experienced the person doing the work is. So a question to keep in mind when thinking about the results from any given dig and how much weight to place on them is: who was doing the analysis?

Now, studies that I know of include the aforementioned Spitalfields excavation, where skeletons were identifiable by name and age at death in some cases (via inscribed metal plates), allowing an blind test of methods for sexing skeletons, with varied results - see above for the unfortunate "what childbirth" incident - but overall showing that this stuff gives good but not perfect demographic data. (Here is the report on the skeletal remains from the site.)

The word "demographic" is significant here. The purpose of these techniques, and what they are good at, is giving a broad picture of a population. The questions archaeology is primarily interested in are more demographic than individual, and while the answers it gets may be good ones on the level of an entire cemetery, when you're looking at a single person, you're already outside of the realm that archaeology is primarily interested in. I've noted again and again that these techniques can be problematic on the single-person level, and this is basically why they're being used as valid archaeological techniques anyway, and why they can be used to great effect for generating patterns, i.e. for understanding the population as a whole. If the person using the techniques isn't being steered too strongly by their expectations to really look at the evidence, that is.

2. Looking at the grave goods

First: DON'T DO IT.

Looking at the grave goods is something that I've seen done in modern archaeology as a decider when a skeleton is indeterminate. Looking at the bones doesn't give you a clue, but the individual was buried with keys, a selection of pots, clothing pins which indicate a dress to your eye, and you're working on an early medieval British site: female. It's entirely possible that the alternatives don't even occur to many. There are definitely a good number of archaeologists out there who would let male grave goods override a fairly strong probability of a female skeleton, as well, by the way. I have worked with them.

And historically grave goods have been among the chief ways of assigning sex to skeletons.

But seriously, guys, DON'T DO IT.

This is basically a conflation of sex with gender, and an assumption that gender is being assigned along biological lines. Obviously neither of those assumptions are going to stand up all the time, and to assume that they do is to a) simplify the picture and b) hide away even more of the complexity of people's lives than is already necessarily lost. The reality is: we don't always know who is male and who is female, or who is a man and who is a woman, and if man and male/woman and female go together in this case.

Don't doooooooooooooooooooo it.

But of course people do do it.

NB: Grave goods are of course wonderful archaeological evidence and can tell one a lot about construction of gender & all kinds of other aspects of culture when used in combination with one of the techniques which actually has something to say about biological sex. I'm not knocking grave goods.

Except when people think a spear means it must be a male burial.

Accuracy of this method: :|

3. DNA testing

DNA testing is probably the ideal way of getting an individual result, to be honest. OK, I am not a biologist, so all I know about it really is that samples go to lab, answers come back! But I'm told it's good stuff.

Some provisos:

- preservation. It certainly isn't every skeleton you can get a good DNA sample from, and this depends on the kind of soil it's been buried in, how old the bone material is, climate...

- contamination. OK, you got a female result back, but how has the bone that you tested been handled by site staff and lab staff? Do it properly or don't do it.

- expense. My main project certainly didn't have the money to do this more than a couple of times, and while I wasn't directly involved I suspect that was pushing it.

Ta-da! I do hope that was at least somewhat helpful.

Date: 2011-04-27 10:43 pm (UTC)
dancing_moon: Jadeite / DM / Me (Default)
From: [personal profile] dancing_moon
Interesting and educational!

I actually thought that the bone judgement was more solid than you describe it, but perhaps I'm mixing it up with forensic sexing of skeletons, which should include much more modern samples?

Date: 2011-04-28 01:15 am (UTC)
From: [personal profile] rho
This is fascinating. Thanks for posting it!

As regards the third method, I'm not a biologist either, but I'll add that DNA testing is also not infallible, even when done properly on a well-preserved sample. The existence of XY females, XX males and many other varieties of intersex makes that impossible.

Date: 2011-04-28 06:22 am (UTC)
sollers: me in morris kit (Default)
From: [personal profile] sollers
There's another problem with bones other than the pelvis, and that's population factors. When I was studying "Human Skeletal Remains in Archaeology" we were told a cautionary tale.

Dr M had been examining subjects from an Anglo-Saxon cemetery, then moved on to a Medieval one. She needed to go back and check on one of the AS subjects, glanced at the skull and the robustness of the bones and thought, "Oh, yes, definitely male." She then looked at the notes she had made and saw "Definitely female". She wondered, "Why did I ever think that?" and fished out the pelvis. Female it was indeed. The problem was that the AS community - note, of exactly the same genetic stock as the Medieval village - were all taller and more robust, and the AS females were as beefy as the Medieval males.

Skulls, she said, are particularly dodgy, because bone is dynamic; the skulls of women in positions of authority develop "masculine" characteristics, and men in subservient positions lose them. This has in the past caused confusion with religious communities, but once she was clear about that we found we could use it for determining the status of monks, for example, within the community.

There's one exception to not being able to sex children, and that's neonates. For some reason their pelvises do show a difference. As we were interested in lifestyles, this was particularly interesting (were there more male neonates or more females? If the former, deaths were almost certainly natural, but if the latter it could be iffy).

DNA: yes, expensive and likely to be contaminated, so rarely used. Best source is from inside teeth.

Date: 2011-04-28 09:10 am (UTC)
vacillating: picture of Dana Scully (The X-Files); text: please explain the scientific nature of the 'whammy' (Default)
From: [personal profile] vacillating
Thank you, this was fascinating. I was especially interested to learn that the Spitalfields work disproved an older method.

Date: 2011-04-28 04:59 pm (UTC)
synecdochic: torso of a man wearing jeans, hands bound with belt (Default)
From: [personal profile] synecdochic
If you are interested in learning more about the forensic details, my wife [personal profile] sarah is a forensic DNA analyst and has written lots of awesome essays about how it works. (Some of them are over on [ profile] sarahq and I don't know how much she imported, but still.) She plans on making some more posts about it, so you can ask her questions!

Date: 2011-04-28 05:06 pm (UTC)
synecdochic: torso of a man wearing jeans, hands bound with belt (Default)
From: [personal profile] synecdochic
Be glad you didn't! The field is so small there's like two lab positions opening up per year. *g*

Date: 2011-04-28 05:09 pm (UTC)
synecdochic: torso of a man wearing jeans, hands bound with belt (Default)
From: [personal profile] synecdochic
Yeah, I think that's pretty common. *g*

Date: 2011-04-28 06:35 pm (UTC)
pauamma: Cartooney crab holding drink (Default)
From: [personal profile] pauamma
Why do I have a feeling this is related to the big hoopla described in ?

Date: 2011-04-28 08:35 pm (UTC)
pauamma: Cartooney crab holding drink (Default)
From: [personal profile] pauamma
Yeah. :-(

Date: 2011-04-29 06:44 am (UTC)
vacillating: picture of Dana Scully (The X-Files); text: please explain the scientific nature of the 'whammy' (Default)
From: [personal profile] vacillating
I did have that case in mind when I wrote the question, yes, although it wasn't the first and I sadly suspect that it won't be the last.

Date: 2011-04-28 08:52 pm (UTC)
melannen: Commander Valentine of Alpha Squad Seven, a red-haired female Nick Fury in space, smoking contemplatively (Default)
From: [personal profile] melannen
Oh, this is great! And thank you for putting it the way you did: after about the third time I read an account where grave goods were used to sex a skeleton, and then the skeleton's sex was used to gender the same the grave goods, I started to get really, really disappointed.

Q: So is the pelvic pitting=childbirth thing thoroughly debunked in terms of positive results as well as negatives? Because I just went to the temporary forensic anthropology exhibit at the Smithsonian (which is admittedly almost two years old) and it was talked about there, without any caveats. And that exhibit is spoken of very highly by lots of forensic anthropologists, so D: if so.

Date: 2011-05-01 12:14 pm (UTC)
eggcrack: Icon based on the painting "Kullervon kirous ja sotaanlahto" (Default)
From: [personal profile] eggcrack
So much new and fascinating information in this post! Thank you for writing it. <3

Date: 2011-05-06 07:00 pm (UTC)
From: [personal profile] foxfinial
(found via [community profile] history)

This is really interesting! Archaeology has always interested me, at a thoroughly lay level; it's cool to have this explained. :> And I LOL at the idea of grave goods being used to determine sex.

Do you mind if I add you? From this and from your info post, you seem interesting, and I'm trying to be more active on DW and stuff.


marshtide: (Default)

December 2012

30 31     

Style Credit

Page generated Mar. 28th, 2017 06:20 am
Powered by Dreamwidth Studios

Expand Cut Tags

No cut tags

Most Popular Tags