Articles to 2018-08-30

Zum Seitenende      Übersicht Artikel      Home & Impressum

First the link to this week's complete list as HTML and as PDF.


Doctor et al.'s intervention has been shown to be effective, but they fail to even ask whether it's beneficial or detrimental. Granted, death by an inadvertent overdose is to be prevented, but then with rare exceptions none of these patients was looking forward to many happy and healthy years but nearly all were in the final stage of a fatal illness. Is one slightly premature death worth many cases of weeks of unnecessary pain? I do not know the answer but the authors have totally failed to even ask the question.


If an experiment fails abysmally after running for twenty years, that surely must be a blow but is no excuse for publishing nonsensical results anyway. Thankfully Wolf & Ziska have noticed the shortcomings and written a comment, although they too seem to have missed the main problem as outlined in my list of 2018-05-21.


There are quite a few issues with Roffet-Salque et al. The amplitude of their signal is less than the spread inside the individual phases and no point differs from any other by more than its own standard deviation. Even the noise from their measurement precision is of the same order as their whole signal. For their δ²H they use far fewer samples than the successfully extracted fatty acids they determined carbon isotopes on. By eliminating just one measurement each from their phases O and P, I can make the whole result disappear and those two important phases are the ones where they discarded the most samples. They don't tell us which ones or why.

In their data they find no evidence at all for dairy fats but they still postulate “herds [of] domesticated cattle”. I find that unsupported assumption highly improbable. As far as we know the start of dairying was exactly synchronous to the sudden arrival of imported domestic cattle and roughly contemporary to the 8.2 event and the move from the east to the west hill.

If their phases O and P really were to fall into the 8.2 cooling event, why would people have chosen just that time to give up their well sheltered former dwellings and live in “light shelters with large open spaces”? If anything the following phases Q and R with “multiroomed dwelling structures with central “living rooms” [...] surrounded by small, cell-like spaces” seem much better suited to colder seasons.

I also note two glaring errors not impacting their result. The final abandonment is after 7900 cal BP, not BC, and C16 not C18 (Palmitic not Stearic acid) is the one mainly derived from plants. Stearic acid is the more abundant (and thus easier to extract) one in animals and is synthesized by their own metabolism from all kinds of food including starches. It is thus more indicative of the whole environment and precipitation and correctly chosen for analysis. Both these errors can happen to anyone, especially the BC typo, but how can they have been missed by nine co-authors, several referees, and the editors? Have the supposed co-authors even read the article they allowed their names to be put to?


Blake et al. ask an important question about gender relations in modern society. Their main result – one expected correlation missing entirely and another one appearing quite strong – is sound. That said their proof leaves a lot to be desired. In the cargo-cult manner of the social sciences we are not given, much less shown, any data whatsoever but only the tabulated results of ill-explained mathematical shenanigans.

As Borgerhoff Mulder correctly points out there are several weaknesses on the interpretive side. First, looking only at regional aggregates of behaviour doesn't allow you to say anything about the subset of people actually doing it. She makes a second important point when she observes: “[T]here are places in the world where posting a sexy selfie would torpedo a woman's marriage chances or make any potentially high-investing man run a mile.” Contrary to her places I see that applying to the upper strata of all societies everywhere.

There is one glimpse at Blake et al.'s raw data in their supplementary figure S1. The obvious main driver is the economic development index, which they correctly interpret as mainly being a confounder of the underlying social process. The dependence on inequality is much weaker and even changes direction in some parts of the graph.

I had offered an alternative explanation of the data in the first version of this blog entry, but after a second look it turned out to be utter rubbish. I've deleted it now (i.e. same day, one hour later).

Zum Anfang      Übersicht Artikel      Home & Impressum

Creative Commons Attribution-Share Alike 3.0 Unported License Viewable With Any Browser Valid HTML 4.01! Valid CSS!