William M. Briggs

Statistician to the Stars!

Ritual Sacrifice Promoted Evolution Of Stratified Societies: A Statistical Fallacy

r12

His Majesty

Every human society except one is stratified. The exception is the empire of William I, a small sand island that comes and goes by a remote and isolated bend in the Chippewa River, depending on the rainfall. This kingdom is the only known truly egalitarian culture in existence.

Because why? Because this Island Empire has only one permanent resident, your Majestic Host, hence all inhabitants are necessarily equal.

We hesitate to tell you this because, given the mania for Equality, the solution we have hit upon might appeal to the zealous, and we have grown fond of many of you.

In any case, every other society and culture is stratified and unequal. At the least, in the countries which approach or claim egalitarianism, parents preside over their children, and treat their own progeny differently than the children of others, and the majority who vote or act one way establish dominance over the minority who oppose them.

Off with their heads

This is important to understand because some scientists, led by Joseph Watts, in a peer-reviewed paper in Nature, tell us that ritual human sacrifice produced stratified societies. This isn’t opinion, it’s scientific theory: “According to the social control hypothesis, human sacrifice legitimizes political authority and social class systems, functioning to stabilize such social stratification.”

Let’s scour the memory banks on this one and see if we agree. Well, in a society those that kill are at least more important than those who are axed, plus, as discussed above, every society with at least two members is stratified, so Watts’s theory appears to be on solid ground. We agree that there is “strong support for models in which human sacrifice stabilizes social stratification”, but we disagree that stabilization occurs only “once stratification has arisen” as Watts says, because stratification is ever present. What’s true, we suppose, is that the type of stratification would be set in blood.

“The methods of sacrifice [they studied] included burning, drowning, strangulation, bludgeoning, burial, being crushed under a newly built canoe, being cut to pieces, as well as being rolled off the roof of a house and then decapitated.”

That last one’s a little harsh, no? You’ve already suffered the indignity of being rolled off a roof, but then to have your head cut off, too? Even the killing methods are egalitarian. There is more dignity being crushed under a new canoe instead of a soiled one.

Have we lost track, here, dear reader? Ah, yes, we need more if we’re going to make this commonsense observation into science. We need quantification.

Watts looked to “93 traditional Austronesian cultures from the Pulotu database” and classified societies thusly:

Cultures that lacked inherited differences in wealth and status were defined as lacking social stratification, and were coded as egalitarian. Cultures were coded as moderately stratified if there were inherited differences in wealth and social position with the potential for status change within a generation, and highly stratified if there were inherited difference in wealth and social position with little or no possibility of status change within a generation.

We’ll give them this, but it’s not to be believed because no society is truly egalitarian. Even the great egalitarian society of the once United States gives preference to inherited status; just ask Jeb Bush; just ask your own children.

Watts et al. saw that ritual sacrifice “was practiced in 5 of the 20 egalitarian societies” and at greater rates in “moderately stratified” and still greater rates in “highly stratified” societies.

This accords with the commonsense model that the more killing there is, the more the killers secure their and their families’ positions, lest they find themselves under the canoe or rolled off rooftops.

Now there might be some interest in predicting, à la Hari Seldon, the rate at which priests in such (carefully defined) societies become hereditary. Once somebody hits upon the swell idea of killing to appease the gods, the society in which this happens surely changes. The nature of that change is an excellent question.

All this is fine, but the authors—and in this they are in no way unique—had to booger it up with a statistical (Bayesian!) model. What they had to do but did not was to examine, society by society, how the type of stratifications which existed before sacrifices began solidified or morphed after the heads came off.

Instead they used their models to “prove” that stratification and sacrifice were “linked” causally. But statistical models are silent on cause, so the authors committed a formal fallacy. The conclusion of their fallacy—societies in which ritual killings exist are more rigidly stratified—is true. And we understand it is true because we understand human nature. Yet the authors said the conclusion was true because some Bayes factor exceeded some number, which is a fallacy.

For a concrete example, they said: “The results from our second series of analyses indicates that human sacrifice increased the rate at which cultures with human sacrifice gain high social stratification, but did not function to stabilize high social stratification once it had arisen.”

They could have had that same result without a model, just by looking and counting and analyzing the societies—no model needed!—and therefore the result would have been on firm ground. It is only a coincidence that that statistical model agreed with reality. Just think: the rigidity of stratification could have occurred because of sacrifice in only one society, but because it was only one, the statistical model would have said the phenomenon didn’t happen!

Plus, they (and we) would have learned more if they examined why which societies fell into sacrifice and which did not. That would have taken real work; certainly more than an afternoon playing with numbers.

Statistical analysis is so often a tool of the lazy that it’s unremarkable that it occurred here. And then you wonder why they bothered at all, when they say at the very end this: “Throughout human history the practice of human sacrifice was often used by social elites as a display of power, intended to instil [sic] fear of the secular and supernatural consequences of transgressing ruling authority.”

Seems they already knew the answer before they began. Which indicates they thought the statistics would give formal proof of what they knew, which is false. Statistics should only be used when we don’t know what’s going on and need to quantify uncertainty of what might happen.

Allure of theory

Finally came this nugget, the last sentence of the article “Unpalatable as it might be, our results suggest that ritual killing helped humans transition from the small egalitarian groups of our ancestors, to the large stratified societies we live in today.”

Thwwwppbbt. This is too far, too much. There is no warrant to extrapolate from the some dumb model to such a gross simplification of human history as this.

Thanks to the many readers who brought this article to my attention. Since these were on Twitter, I’ve lost who they were!

19 Comments

  1. What’s true, we suppose, is that the type of stratification would be set in blood.

    If the victim is stoned, the ” stratification would be set in stone “?

  2. Surely this kind of paper must take into account the works of René Girard. Is he cited?

  3. “Peer-reviewed paper” sounds very funny after I spent time yesterday reading Retraction Watch.

    Isn’t this picking on people who were not like the enlightened researchers? I think it sounds bigoted and mean….

    Killing to appease the gods should increase church attendance. Or the following of tribal rules, etc. Death is a pretty big motivator. I wonder if voluntary sacrifice was done. Did they address this? Works with suicide bombers…..

    Speaking of things we know are true but science didn’t confirm, Fox News had a study from Reuters saying people don’t sleep well in a different place because half the brain stays “awake” watching for danger. I really thought that was fairly self-evident, but without scientific proof……

  4. Anybody who uses “transition” as a verb has lost my vote anyway.

  5. I would think that sacrifice would be used primarily to change the nature of the stratification. The French reign of terror was the model used by every communist revolution since. Of course the Roman Empire was not immune from this sort of thing either.

  6. “Watts et al. saw that ritual sacrifice “was practiced in 5 of the 20 egalitarian societies” and at greater rates in “moderately stratified” and still greater rates in “highly stratified” societies.

    This accords with the commonsense model that the more killing there is, the more the killers secure their and their families’ positions, lest they find themselves under the canoe or rolled off rooftops.”

    Mr. Briggs,

    So, the first paragraph is part of the evidence presented in the paper. The second paragraph is your conclusion or commonsense. I don’t see why your commonsense agrees with the evidence.

    Why? The first paragraph says nothing about the number of killing, and a higher level of social stratification is not equivalent to a higher position/status of the killer. Societies or cultures are the units of analysis in the study. Failure to recognize the unit can result in serious problems, e.g, in making conclusions.

    Anyway, is it really a commonsense? I did learn in my high school history class that emperors with higher status were often offered with a higher number of human sacrifices. I am not sure if vice versa is true.

    Plus, they (and we) would have learned more if they examined why which societies fell into sacrifice and which did not.

    You might find answers to this questions in the literature in the paper.

  7. Briggs

    April 25, 2016 at 12:57 pm

    JH,

    I read your comment twice carefully. It doesn’t make any sense. But I’d say, yes, it’s common sense that communities which engage in ritual killings are stratified.

    The last point you make, quoting me, is not true. Those answers are not in the paper. Which you could discover by reading it.

  8. The conclusion of their fallacy—societies in which ritual killings exist are more rigidly stratified—is true. And we understand it is true because we understand human nature.

    This is not their conclusion, is it? Ritual killings also exist in some of the egalitarian societies observed in the paper, and some of the societies with the highest level of social stratification didn’t practice human sacrifice. The conclusion that “societies in which ritual killings exist are more rigidly stratified” is not true.

    What “human nature” allows you to understand the said truth? Is it really true? Do I need to have some basic knowledge of the Austronesian culture to understand the aforementioned human nature? Well, I really don’t know much about this particular culture.

    I once read somewhere that it’s part of human nature to be scientific.

    Well, unlike you, Mr. Briggs, I admit that I lack the ability to “just-know.”

  9. This accords with the commonsense model that the more killing there is, the more the killers secure their and their families’ positions.
    To make such conclusion, one needs to observe the number of killing and the killer’s position/status. The evidence collected in the paper includes the practice of sacrifice killing and the social stratification. Nothing about the number of killings in the evidence or any killer’s status. You commonsense does not accord with the observed evidence.

    Right, those answers are not in the paper, and I didn’t say you cout find the answers in this particular paper. They might be in the literature in the paper. the references (literature) numbered 2, 13, 14, 1, 8… in the paper.

  10. Briggs

    April 25, 2016 at 2:20 pm

    Hey JH, Bayes factors are just as useful in ascribing cause as p-values. Which is to say, they are not useful at all.

  11. Briggs, moving posts? Bayes factors and p-values are more useful than you’d like them to be when coupled with subject knowledge.

  12. Briggs

    April 25, 2016 at 3:02 pm

    JH,

    Nope.

  13. Examining the entrails of chickens is also more useful when coupled with subject knowledge.

  14. Study Conclusion, in relevant part: “…ritual human sacrifice produced stratified societies…”

    WTF?!!?!

    A society comprised of groups of individuals, some of whom have the power & authority to SACRIFICE others — isn’t that arrangment an indication that that society was ALREADY stratified!

    If such a society wasn’t already stratified, how could some powerful select few obtain the power & authority to sacrifice other humans??

    Seems like a sort of chicken/egg issue the study participants got backwards — societies got stratified, and because they were, those in the topmost strata could maintain their position by force to include human sacrifice.

    Kinda like ‘absolute power corrupts absolutely’ — the extreme corrupt behavior begins in force AFTER the individual gets the power to abuse (typically getting the power, creating the stratification, by persuading the minions giving it to him [historically usually a him] because he would be nice & noble).

    Here’s a link to the paper: http://www.nature.com/articles/nature17159.epdf?referrer_access_token=e8FHGbetarfav_r3qexIKdRgN0jAjWel9jnR3ZoTv0P-VjvVXFNrjcb83nDNmh2QrbyT_mLXOCiwRBtfbkwkz3nXjlkCQH0DaMsPXtoFNwrA0DsFYL-mix6DJUzpDJVVBhWXe_LJ-i5LnE_aI1l48OpncNY2MSpYpKFjbZRVL21C7O9x91z1_wFqy8cNEjktAvM1Lfe6RowbTEk9pvWynQ%3D%3D&tracking_referrer=www.npr.org

  15. Heh heh!

    Couldn’t we say that these internet blog thingys are the ultimate anti-egalitarian society? The owner/moderator has complete, arbitrary power over who in the society gets axed (sacrificed) on the altar of his ideals… he doesn’t even need sycophants, lackeys, henchmen to assist him.

    Now, I’m not saying that it is necessarily a bad thing… bad apples need to be removed or they will spread an infection that destroys the whole society.

  16. Well, speaking as someone with more of a history background, these studies are trying to answer the old historians question, “Why did some aboriginal American cultures form large, complex societies while most remained hunter-gatherers?”

    That said, it shouldn’t have taken much study to arrive at that conclusion. If asked before, I would have said, “I presume they did it to keep whatever class-system they had in place.” Then there’s just the whole dreary spectacle of the thing. The problem these historians have is the belief that the more large and complex the culture, the more civilized it is. I would contend complexity and civility are rather elastically attached.

    JMJ

  17. Scotian,

    So what subject knowledge would be required in examining the entrails of chickens?

    Two of my uncles picked up Chinese astrology reading as a hobby
    (https://en.m.wikipedia.org/wiki/Four_Pillars_of_Destiny) in college. One moved to Melbourne, became a Christian and gave up the hobby. The other attempted to recruit me after I started college. I studied some books he gave me, and worked as his assistant to his fortune telling hobby for a couple of years for fun. I can tell you that the majority of the rules in those books are not fact based, but I guess some people, such as my two very lovable uncles, would see them as valid knowledge.

    I imagine some biologists might agree with you.

  18. JH,
    Why knowledge of the subject at hand, of course. Isn’t this what you meant as well? This is shown clearly in a scene from The Wisard of Oz:

    https://m.youtube.com/watch?v=PJaxd1BQr9g

    Here a crystal ball is used rather than chicken entrails but the principle is the same. Notice how cleverly Professor Marvel combines the technique of the crystal ball with knowledge of the subject at hand.

  19. Scotian,

    Isn’t this what you meant as well?

    No. I don’t wish to waste more of my time on this. If you wish to elaborate on Bayes factors and p-values, I’d love to respond.

Leave a Reply

Your email address will not be published.

*

© 2016 William M. Briggs

Theme by Anders NorenUp ↑