Three articles this month and all of them are from the softer side of science. One is about how doctors choose treatments for patients. The next one is a groundbreaking new study on the origins of language. The last one is a study confirming the widespread hunch that more education leads to less religion.

The first article covers a recent survey by researchers from Duke University which asked doctors how they would choose treatments for themselves and compared the results against what the doctors would choose for other patients. The survey found that when choosing treatments for other people, doctors tried to minimize the risk of death as much as possible, even if the treatment involved the risk of long-term complications. When it came to themselves however, doctors were much more willing to risk death if it meant avoiding medications with side-effects.

Apart from being a good illustration of one of the problems inherent in the medical professions, in which the layman needs to be guided by the expert, but both parties don’t quite have the same interests in mind, I think that a reformulated version of this survey could serve as a useful measure of just how common defensive medicine could be, particularly in rich countries that are facing rapidly escalating healthcare costs.

The next finding is so significant that nearly every news publication around the world has an article on it but this one is from the New York Times. The discovery is that just as all humans originated from Africa, language only emerged once in Africa and thereafter spread across the world. This claim is made by Quentin Atkinson of the University of Auckland who performed a mathematical analysis of the phonemes of some 500 languages in use today. Dr. Atkinson finds that just as the genetic diversity of humans declines in line with increasing distance from Africa, so does the diversity of phonemes in a given language decline with distance.

As such, the article notes that some of the languages in Africa which use distinctive clicking sounds found nowhere else in the world have more than 100 phonemes while English has around 45. The Hawaiian language by contrast, thought to originate from the far end of human migration out of Africa, only has around 13 phonemes. Dr. Atkinson further finds that according to his model of language, humans emerged somewhere in southwestern Africa, which matches findings from analysis of genetic diversity.

His model is quite a bit more complex than this simple summary of course. Part of it is that the number of phonemes gradually increases as the population of people who speak it increases. But as small groups break off from the main group to migrate elsewhere, the diversity would decrease again. Needless to say, all this is quite tentative and many linguists still view this with a skeptical eye, but if everything checks out, this type of analysis would be a powerful new tool with which to study the spread of human culture.

Finally, we have a blog post from Freakonomics that covers a new paper by Daniel Hungerman of the University of Notre Dame in the United States. Based on Canadian survey data, the paper confirms the long-held suspicion that the more education someone has, the less likely that person is to be religious. The paper finds that for each year of completed schooling a person has, the likelihood of that person identifying with any religious affiliation declines by roughly 4 percent.