Related: The Psychological Diversity of Mankind, An African Folktale, many of the more interesting infanticide & abortion debates on this site

A fascinating post that however might need some background reading, most relevant material is linked in the article itself. I encourage reading up on the material.

Link to article.

Stories about changelings replacing babies and the recommended course of action being basically to expose the child is not a human universal, they are found only in European cultures. These rely more heavily on guilt and less on shame to regulate behavior than most other human societies. This may not be a coincidence. The stories look like they work as a ready made rationalization to reduce guilt from infanticide. Common problems often acquire common solutions like this.

Guilt and Shame Cultures

On his blog Evo and Proud, anthropologist Peter Frost recently wrote a highly interesting two-part article entitled The origins of Northwestern European guilt culture. In guilt cultures, social control is regulated more by guilt than by shame, as is the case in shame cultures that exist in most parts of the world. A crucial difference between these types of cultures is that while shame cultures require other people to shame the wrongdoer, guilt cultures do not. Instead, he or she will shame themselves by feeling guilty. This, according to Frost, is also linked to a stronger sense of empathy with others, not just with relatives but people in general.

The advantages of guilt over shame are many. People can go about their business without being supervised by others, and they can cooperate with people they’re not related to as long as both parties have the same view on right and wrong. And with this personal freedom come individualism, innovation and other forms of creativity as well as ideas of universal human rights etc. You could argue, as Frost appears to, that the increased sense of guilt in Northwestern Europe (NWE) is a major factor behind Western Civilization. While this sounds fairly plausible (in my ears at least), a fundamental question is whether there really is more guilt in the NWE sphere than elsewhere.

How to Measure Guilt

The idea of NWE countries as guilt cultures may seem obvious to some and dubious to others. The Protestant tradition is surely one indication of this, but some anthropologists argue that other cultures have other forms of guilt, not as easily recognized by Western scholars. For instance, Andrew Beatty mentions that the Javanese have no word for either shame or guilt but report uneasiness and a sense of haunting regarding certain political murders they’ve committed. So maybe they have just as much guilt as NWE Protestants?

This is one of the problems with soft science – you can argue about the meaning of terms and concepts back and forth until hell freezes over without coming to any useful conclusion. One way around this is to find some robust metric that most people would agree indicates guilt. One such measure, I believe, would be murder rate. If people in different cultures vary in the guilt they feel for committing murder, then this should hold them back and show up as a variation in the murder rate. I will here take the NWE region to mean the British Isles, the Nordic countries (excluding Finland), Germany, France and Belgium, Netherlands, Luxembourg, Australia, New Zealand and Canada for a total of 14 countries. According to UNODC/Wikipedia, the average murder rate in the NWE countries is exactly 1.0 murder per 100K inhabitants. To put this in perspective, only 20 other countries (and territories) of 207 listed are below this level and 70 percent of them have twice the murder rate or more.

Still, criminals are after all not a very representative group having more of the dark traits (psychopathy, narcissism, machiavellism) than the rest of the population. Corruption, on the other hand, as I’ve argued in an earlier post, seems relatively unrelated to regular personality traits, so it should tap into the mainstream population. Corruption is often about minor transgressions that many people engage in knowing that they can usually get away with it. They will not be shamed because no one will know about it and many will not care since it’s so common, but some will feel guilty and refrain from it.  Looking at the Corruptions Perceptions Index for 2013, the NWE countries are very dominant at the top of the ranking (meaning they lack corruption). There are seven NWEs in the top ten and two additional bordering countries (Finland and Switzerland).  The entire NWE region is within the top 24, of a 177 countries and territories.

But as I’ve argued before here, corruption appears to be linked to clannishness and tribalism (traits rarely discussed in psychology) and it’s reasonable to assume that it is a casual factor. How does this all add up? Well, the clannish and tribal cultures that I broadly refer to as traditional cultures are all based on the premise that the family, tribe or similar ingroup is that which should be everyone’s first concern. So while a member of a traditional culture may have personal feelings of guilt, this means little compared to the collective dislike – the shame – from the family or tribe. At the same time traditional cultures are indifferent or hostile towards other groups so if your corruption serves the family or tribe there will be no shame in it, the others will more likely praise you for being clever.

(In this context it’s also interesting to note that people who shame others often do this by expressing disgust, an emotion linked to a traditional dislike for various outgroups, such as homosexuals or people of other races. So disgust, which psychologist Jonathan Haidt connects with the moral foundation of sanctity/degradation, is perhaps equally important to the foundation loyalty/ingroup.)

When Did Modernity Begin?

One important question is whether this distinction between modern and traditional is to what extent it’s a matter of nature or nurture. There is evidence that it is caused by inbreeding and the accumulation of genes for familial altruism (that’s to say a concern for relatives and a corresponding dislike for non-relatives). Since studies on this are non-existent as far as I know – no doubt for political reasons – another form of evidence could be found in tracing this distinction back in time. The further we can do this, the more likely it’s a matter of genes rather than culture. And the better we can identify populations that are innately modern the better we can understanding the function and origin of this trait. Frost argues that guilt culture can be found as early as the Anglo-Saxon period (550-1066), based thing like the existence of looser family structures with a relatively late age of marriage and the notion of a shame before the spirits or God, which can be construed as guilt. This made me wonder if there is any similar historical evidence for NWE guilt that is old enough to make the case for this to be an inherited behavior (or at least the capacity for guilt-motivated behavior). And that’s how I came up with the changeling,

The Changeling

As Jung has argued, there is a striking similarity between myths and traditional storytelling over the world. People who have never been in contact with each other have certain recurring structures in their narratives, and, as I’ve argued before here, even modern people adhere to these unspoken rules of storytelling – the archetypes. The only reasonable explanation for archetypes is that they are a reflection of how humans are wired. But if archetypal stories reveal a universal human nature, what about stories found in some places but not in others? In some cases they may reflect differences in things like climate or geography, but if no such environmental explanation can be found I believe that the variation may be a case of human biodiversity.

I believe one such variation relevant to guilt culture is the genre of changeling tales. These folktales are invariably about how otherworldly creatures like fairies abduct newborn children and replace them with something in their likeness, a changeling. The changeling is sometimes a fairy, sometimes just an enchanted piece of wood that has been made to look like a child. It’s typically very hungry but sickly and fails to thrive. A woman who suspected that she had a changeling on her hands could find out by beating the changeling, throwing it in the water, leaving it in the woods overnight and so on. According to the folktales, this would prompt the fairies or whoever was responsible for the exchange to come to rescue their child and also return the child they had taken.

Infanticide Made Easy

Most scholars agree that the changeling tales was a way to justify killing sickly and deformed children. According to American folklorist D. L. Ashliman at the University of Pittsburgh, people firmly believed in changelings and did as the tales instructed,

""There is ample evidence that these legendary accounts do not misrepresent or exaggerate the actual abuse of suspected changelings. Court records between about 1850 and 1900 in Germany, Scandinavia, Great Britain, and Ireland reveal numerous proceedings against defendants accused of torturing and murdering suspected changelings.""

This all sounds pretty grisly but before modern medicine and social welfare institutions, a child of this kind was a disaster. Up until the 1900s, children were supposed to be relatively self-sufficient and help out around the house. A child that needed constant supervision without any prospect of ever being able contribute anything to the household was more than a burden; it jeopardized the future of the entire family.

Still, there is probably no stronger bond between two people than that between a mother and her newborn child. So how could a woman not feel guilty for killing her own child? Because it must be guilt we’re talking about here – you would never be shamed for doing it since it was according to custom. The belief in changelings expressed in the folktales gave the women (and men) a way out of this dilemma. (Ironically, Martin Luther, the icon of guilt culture, dismissed all the popular superstitions of his fellow countrymen with the sole exception of changelings which he firmly believed in.) Thus, the main purpose of these tales seems to have been to alleviate guilt.


If this is true then changeling stories should be more common in the NWE region than elsewhere, which also seems to be the case. There are numerous changeling tales found on the British Isles, in Scandinavia, Germany and France. It can be found elsewhere in Europe as well, in the Basque region and among Slavic people and even as far as North Africa, but at least according to folklorists I’ve found discussing these tales, they are imported from the NWE region. And if we look beyond regions bordering to Europe changelings seem to be virtually non-existent. Some folklorists have suggested that for instance the Nigerian Ogbanje can be thought of as a changeling, although at a closer inspection the similarity is very superficial. The Ogbanje is reborn into the same family over and over and to break the curse families consult medicine men after the child has died. When they consult a medicine man when the child is still alive it is for the purpose of severing the child’s connection to the spirit world and make it normal. So the belief in the Ogbanje never justifies infanticide. Another contender is the Filipino Aswang which is a creature that will attack children as well as adults and is never takes the place of a child but is more like a vampire. So it’s safe to say that the changeling belief is firmly rooted in the NWE region at least back to medieval times and perhaps earlier too.

Before There Were Changelings, There Was Exposure

Given how infanticide is such a good candidate for measuring guilt, we could go back further in time, before any evidence of changelings and look at potential differences in attitudes towards this act.

I doing so I think we can find, if not NWE guilt, so at least Western ditto. According this Wikipedia article, the ancient Greeks and Romans as well as Germanic tribes, killed infants by exposure rather than through a direct act. Here is a quote on the practice in Greece,

""Babies would often be rejected if they were illegitimate, unhealthy or deformed, the wrong sex, or too great a burden on the family. These babies would not be directly killed, but put in a clay pot or jar and deserted outside the front door or on the roadway. In ancient Greek religion, this practice took the responsibility away from the parents because the child would die of natural causes, for example hunger, asphyxiation or exposure to the elements.""

And the Archeology and Classical Research Magazine Roman Times quotes several classical sources suggesting that exposure was controversial even back then,

""Isocrates (436–338 BCE)  includes the exposure of infants in his catalog of horrendous crimes practiced in some  cities (other than Athens) in his work Panathenaicus.""

I also found this excerpt from the play Ion by Euripides, written at the end of the 400s BC. In it Kreusa talks with an old servant about having exposed an unwanted child,

Old Servant: Who cast him forth? – Not thou – O never thou!

Kreusa: Even I. My vesture darkling swaddled him.

Old Servant: Nor any knew the exposing of the child?

Kreusa: None – Misery and Secrecy alone.

Old Servant: How couldst thou leave they babe within the cave?

Kreusa: Ah how? – O pitiful farewells I moaned!

It seems to me that this play, by one of the most prominent playwrights of his time, would not make much sense to the audience unless exposure was something that weighed on many people’s hearts.

Compare this with historical accounts from other cultures, taken from the Wikipedia article mentioned above,

""Some authors believe that there is little evidence that infanticide was prevalent in pre-Islamic Arabia or early Muslim history, except for the case of the Tamim tribe, who practiced it during severe famine. Others state that “female infanticide was common all over Arabia during this period of time” (pre-Islamic Arabia), especially by burying alive a female newborn.

In Kamchatka, babies were killed and thrown to the dogs.

The Svans (a Georgian people) killed the newborn females by filling their mouths with hot ashes.

A typical method in Japan was smothering through wet paper on the baby’s mouth and nose. Mabiki persisted in the 19th century and early 20th century.

Female infanticide of newborn girls was systematic in feudatory Rajputs in South Asia for illegitimate female children during the Middle Ages. According to Firishta, as soon as the illegitimate female child was born she was held “in one hand, and a knife in the other, that any person who wanted a wife might take her now, otherwise she was immediately put to death”

Polar Inuit (Inughuit) killed the child by throwing him or her into the sea. There is even a legend in Inuit mythology, “The Unwanted Child”, where a mother throws her child into the fjord.""

It seems that while people in ancient Greece practiced exposure, something many were troubled by, the active killing was common in the rest of the world and persists to this day in many places. While people in other cultures may or may not feel guilt it doesn’t seem to affect them as much, and it’s sometimes even trumped by shame as psychiatrist Steven Pitts and clinical psychologist Erin Bale write in an article in The Bulletin of the American Academy of Psychiatry and the Law regarding the practice of drowning unwanted girls,

""In China, the birth of a daughter has traditionally been accompanied by disappointment and even shame.""

To summarize, the changeling lore provides evidence of a NWE guilt culture dating back at least to medieval times, and the practice and attitude towards exposure suggests that ancient Greece had an emerging guilt culture as early as the 400s BC which enabled a similar individualism and intellectual development that we’ve seen in the NWE in recent centuries. I’m not sure exactly how genetically related these populations are, but the geographical proximity makes it hard to ignore the possibility of gene variants for guilt proneness in Europe responsible for guilt cultures both in ancient Greece and the NWE region. Some branch of Indo-Europeans perhaps?


New Comment
19 comments, sorted by Click to highlight new comments since: Today at 7:06 PM

This is labelled [LINK] but I don't see the actual link, only a copy-and-paste of what may or may not be the whole article being referenced.

I think this is the article, and (admittedly on the basis of skimming only) I think what Konkvistador posted here is in fact the complete text.


Ah pardon! I've now added the link.

It's an interesting theory, but I'm hesitant to give much weight to weakly-supported hypotheses intended to explain very broad and inclusive phenomena, like "murders (or a lack thereof) occurring within these arbitrary geographical borders." This is especially true when there's no shortage of plausible theories and a lot of potentially-useful information is missing.

The changeling myths seem to serve the purpose of guilt-relief only insofar as they also aid shame-relief, so I''m not sure they're all that helpful. (Am I missing something?)

Basically, this reads to me like an interesting but not particularly credible just-so story.

I was following along just fine until:

One way around this is to find some robust metric that most people would agree indicates guilt. One such measure, I believe, would be murder rate. If people in different cultures vary in the guilt they feel for committing murder, then this should hold them back and show up as a variation in the murder rate.

What? Is this an exercise in confounders? Could no longer suspend disbelief after that.

The idea here is interesting, but I wonder if anyone has tried to actually put it to the test. Not out of any personal desire to replace reasoned argument with statistics, mind, but simply because it's pretty clear now that anything short of repeatedly replicated psychometric data will be dismissed without consideration if it disagrees with the doctrine of HNU.

Apparently there are such things as a Guilt Inventory, so assuming it's actually as reliable as it's supposed to be it seems to reason that one could take Guilt Inventories of various populations and see what shakes out.


(In case anyone else reading the parent is scratching the head, that's “human neurological uniformity”. I correctly guessed from the context what concept it referred to, but it took me a few minutes' intensive use of Google's minus operator to find out what word each letter stood for.)

Without saying this is true, I'm not sure how much this would disagree with Human Neurological Uniformity. It seems to me that guilt would be present everywhere but if the culture is one where shaming is used a lot, it wouldn't get as much exercise, which could lead to this.

Not to mention that we don't know for sure that there even is a significant population difference here. It could just as easily be one of the things which humans seem to be generally consistent on as a species.

The point I was making, albeit ineptly, is that good research on the topic would be interesting and any potential ideological fallout shouldn't deter people from it.

I am a molecular biologist and tend to see genetic at places, where others do not. However, in this case, i find it more simple to explain observed things by memes, not genes. The same way the comparatively unrelated languages tend to borrow words from each other, if there is geographical proximity, the changeling stories and general attitude to killing babies can be borrowed and propagated.


Did you read the related links in the article?

I did read one of them before and the other one after you suggested. What is the important message for me ?

This is a very poor post, full of almost comically bad arguments and rationalizations.

One way around this is to find some robust metric that most people would agree indicates guilt. One such measure, I believe, would be murder rate. If people in different cultures vary in the guilt they feel for committing murder, then this should hold them back and show up as a variation in the murder rate.

This borders on the idiotic. Since the author doesn't understand the obvious huge problem that it's impossible in this case to control for other factors, which are likely to be overwhelmingly important, none of their arguments should be taken seriously. It's the sort of cluelessness that makes it likely that in all other cases the author was merely fishing around for plausible correlations, rather than tried to judge the issue or build a convincing argument.

There is evidence that it [the distinction between modern and traditional] is caused by inbreeding and the accumulation of genes for familial altruism...

What evidence?

Since studies on this are non-existent as far as I know

Oh, OK.

– no doubt for political reasons

But political reasons didn't prevent people from publishing studies on race and gender and IQ, or the whole of evolutionary psychology from existing.

Really, the whole thing is very, very bad, and that even before they get to "gene variants for guilt proneness".

Downvoted for not providing a summary.

Here is my attempt at summarizing the quoted article:

  • Personal guilt is defined as shame aimed inward

  • Some cultures are based on guilt, others on shame

  • One way to measure guilt might be murder rate [the author provides little justification for this]

  • Another suggested metric is the corruption perception index. The author expects guilt-dominant cultures to have low corruption, and, unsurprisingly, finds the confirming evidence.

  • Changeling tales are conjectured to be a tool to relieve the guilt a family feels over abandoning a sickly newborn. The author gives a long and rambling comparison of changeling tales from different cultures and eras, from ancient Greece to modern Africa.

  • The author concludes with musings on genetic and cultural origins of the "guilt culture".

I must say that I have low confidence in amateurish research like this one, especially from a person proudly proclaiming "I actually believe in aliens and reincarnation". But it did make me think about guilt vs shame (and other cultural fictions controlling people behavior, like pride, honor etc.). Whether there is a reliable way to rate cultures by guilt prevalence and whether such a rating is an interesting metric remains to be seen. The author's case is certainly far from convincing.


Downvoted for not providing a summary.

Stories about changelings replacing babies and the recommended course of action being basically to expose the child is not a human universal, they are found only in European cultures. These rely more heavily on guilt and less on shame to regulate behavior than most other human societies. This may not be a coincidence. The stories may represent a ready made rationalization to reduce guilt from infanticide. Only common problems acquire common solutions like this.

Will you reverse your downvote? Also if I may ask when did writing a summary for a post, a link post even become a LW norm?

Sure, downvote removed. And, while writing a summary may not be a norm, it ought to be, and I try to do my part in nudging the forum in that direction.


Just wanted to check because I've been inactive for a while.

I don't know, but it kind of does seem to be a norm. Orthogonally, though, when did it become a norm that you should only downvote things that violate LW norms, as opposed to things that violate rules that you want to be norms?

An interesting thought, but I think this is one of those social science hypotheses that you're just not gonna be able to prove.

Your experimental measures are not strong enough to answer this question. And if we find some way of measuring society-wide guilt other than your subjective selection and interpretation of lines from ancient plays, and if we suppose hypothetically that this still backs up the conclusions of your incredibly subjective survey, there can very easily be a common cause (technological progress, increasing "liberal drift"), reverse causation (maybe cultural success allows for more guilt), or just random chance at work since the sample size is small.

In guilt cultures, social control is regulated more by guilt than by shame, as is the case in shame cultures that exist in most parts of the world. A crucial difference between these types of cultures is that while shame cultures require other people to shame the wrongdoer, guilt cultures do not.

I see a sentence like this, EXTREME OVERSIMPLIFICATION WARNING goes off in my head and I don't read further.