Comment by Barrera | Hacker News Reader

Comment by Barrera | Hacker News Reader

A lot hinges on the dating of this site. How was it done?

According to this article[1] (which has a nice wide-angle view of t-pillars in context), it was radiocarbon dating[1]. You find bits of stuff at the site and the isotopic composition of the carbon-containing material tells you the age based on known rates of carbon-14 decay.

The actual isotopic analysis seems pretty solid.

The problem is that this isn't any ordinary site. The article notes that the site appears to have been deliberately buried. This raises the question of where the samples that were dated actually came from.

This critical review suggests major problems with the older-than-everything-else hypothesis for the site.[2] It notes that at least some of the samples were dated from "fill," or the stuff that was used by someone at some point to bury the site. And that stuff could have itself come from sites much older than Gobeklitepe:

> We already discussed the problem with dating “fills” as opposed to dating “structures”. A fill’s date (no matter how confident we may fill about its actual date) in no way dates structures, as it simply can be coming from soil deposits that are either older or younger than the structure itself. You can fill your home with dirt from your yard, which could be from various geologic strata, some containing fossils from the Pleistocene. This will not make your home a Pleistocene Epoch home. Or you can currently fill a 4th century BC Temple with soil from riverbanks containing live exoskeletons; this will not render the Temple a 2000 AD structure.

Even if the site wasn't deliberately buried, everything hinges on where the fill came from. The base assumption of radiocarbon dating is that no foreign material was brought in. The shakier that assumption, the shakier the claim to the ages being quoted.

[1] https://www.asahi.com/ajw/articles/14487942

[2] https://www.researchgate.net/publication/317433791_Dating_Go...

Things rarely depend on any single form of evidence. The prevailing interpretations are usually based multiple forms of corroborating evidence. For example, there are plenty of other neolithic sites in the same region, and many of them share the same art style as Göbekli Tepe. The overall timeline is rather well established based on evidence such as human / animal / plant remains, art, tools, and genetics.

Also, [2] seems to be a self-published article written by someone with no background, publication record, or known collaborations in archaeology or related fields. I would not put much weight on it, especially because I have no background in archaeology either and I'm therefore unable to interpret its reliablity.

Things rarely depend on any single form of evidence.

The article deals at length with those other forms of evidence. In particular, the site does not fit (architecturally, the author's specialty) with what was happening around it at the alleged time of building.

> Also, [2] seems to be a self-published article written by someone with no background, publication record, or known collaborations in archaeology or related fields.

Perhaps true, but irrelevant. Einstein was a patent clerk, a standing that prevented a great many people from even bothering to read the work.

The article is packed with citations that can be followed up on. It methodically takes apart Schmidt's original work. If you doubt the credibility of the analysis, perhaps start there. If he's being unfair to Schmidt's evidence, then that's a different matter.

Einstein had already published several papers in physics and he was about to receive his PhD when he made his major discoveries in 1905. The idea that he was just some outsider patent clerk is an urban legend that needs to die.

As an active researcher, I often find papers in distant fields easier to understand than those in my field. Papers on topics related to my research but not that close to it are particularly difficult. When I'm unfamiliar with the topic, I just see some text written in everyday language. And while most researchers are not great at writing, they are not particularly bad either, because they write so much. When I read such papers, I have absolutely no idea what I'm missing or misinterpreting. I may interpret technical terms with specific definitions as everyday words, and there may be shared context, implicit assumptions, and basic background knowledge I'm simply unaware of.

Hence my default assumption that if you read a research paper and think you understood it, you probably didn't. Research is so specialized these days that you should require something better than your own beliefs to convince yourself that you understood what you read. Relying on external validation ("an expert is someone other experts on the topic recognize as an expert") is pretty useful when you are not an expert yourself.

I think you're half right.

You might have read a specialized article and understood it, i.e. created a meaningful whole of all the parts. But you cannot verify it from prior experience with the domain, so it is an internally unconfirmed reading. In addition, subtle key terms might have been replaced with their mundane meanings.

So while you understood the article, you don't have enough information (yet) to confidently say that you understood it correctly.

Over time you gain domain knowledge, and it becomes easier to verify assumptions of new things you read, or to question and correct your own apparatus of understanding.

There's an important observation to be made here, though, with regards to this behavior. Whenever subject field outsiders want to discuss a matter in my field of study, they invariably simplifies what they've read to childish levels and relate some personal property or recent event in their own life. I recognize they're simply being social, and appreciate the interaction for what it is, rather than getting annoyed by what it isn't.

Human beings are social beings, and many internet comments (including of own) are just expressions of individuality and basic human need to partake rather than thoughtful arguments in an actual effort to discover something new.

That said, I do believe e.g. Einstein's Relativität was written for the larger non-expert audience, so I would expect it to safeguard against most of the interpretative traps that would entail.

>Einstein was a patent clerk.....

Picking a case you think supports your claim without checking the rate at which unknown author claims turn out to be true likely only undermines your claims.

The proper metric is: is the rate at which unknown authors making big claims that stand up to further scrutiny higher than the rate for well established experts in a field?

The answer is no. Crackpots litter all fields, and rarely is anyone at that level correct when the experts are not.

As to Einstein, he was published in top physics journals before he became a patent clerk, and he published at least 16 papers (some incorrect) before his big 1905 paper, which was a few years into his patent clerk job. He was certainly not some unknown. Your case is nothing at all like Einstein's case.

Dendrinos is a moderately well-known guy by ancient near-eastern peeps and he's not regarded well. He doesn't really know what he's talking about and his paper makes that pretty obvious. Here are some random issues I can quickly point out from a skim:

1. Doesn't understand how typological classification works

2. Kind of offhandedly acknowledges, but doesn't actually cite the PPNA plaster dating that Dietrich and Schmidt did 6 years prior, negating the entire fill argument.

3. Starts comparing things to Natufian and other levant material cultures.

4. Just randomly starts assuming his urban planning hypotheses, which are neither well-accepted in the field nor well-regarded by people who specialize in that stuff as it applies to the ancient near-east.

All of this is fine because he's essentially retired and simply getting into internet fights to avoid boredom. Nothing wrong with that, but let's avoid bringing him up as an authority.

"doesn't fit architecturly" is not a good argument. We have never built temples in the same style as houses. The dating is rejected only because it doesn't fit the author's worldview. I laughed out loud at the caption of a photo of a much younger site; "structure more primitive than Gobekli Tepe, thus older."

According to this publication [1] by the "rediscover-er" of Gobeklitepe, they found and dated an animal tooth that confirmed their original dating of the site to somewhere around 9000 BCE. It doesn't look like the critical review that you linked in your comment addresses that.

[1] https://www.researchgate.net/publication/257961716_Establish...

"The last intrusions in the big enclosures can be dated by a charcoal sample taken from under a fallen pillar fragment in Enclosure A to the middle of the 9th millennium."

https://www.researchgate.net/publication/258182967_Radiocarb...

I'm confused, why wouldn't they date some tiny piece from the structure itself? It seems like an obvious thing to do, is it not?

There are probably no organic parts of the structure left, just rocks.

didn't they also find skulls and bones and stuff? anyway, I'm sure they will figure it all out eventually.

According to Sweatman (a year ago), 'three bodies' have been found in 'a burial' 'under the floor of houses'. Start at about 5:10 [https://www.youtube.com/watch?v=9F3qZQRzzA4]

That sort of assumes you have a structure containing elements that were a) made by hand that b) survived. A pile of stone laid by hand doesn't help, for example.

The review I linked has a remarkably readable discussion of this. In a nutshell, the filling of the site rendered everything that has been tested liable to contamination by the fill.

Apparently, the plaster on the walls (which contain lime, which I believe is thought to have come from fires) was one of the main pieces of evidence presented by Schmidt. But plaster can absorb carbon from the fill. This is plausible given both age (at least thousands of years) and the fact that plaster only occurs on the surface in direct contact with the contaminating fill.

AFAICT, there as been no enclosed space from which samples have been dated. It all comes from areas that were directly exposed to fill, or the fill itself.

The author of the review also notes something that should be obvious to those outside the field, but for some reason doesn't get discussed much inside:

> Thus, in questioning the claim about Gobekli Tepe’s date (of PPNB, possibly earlier) one might think that the one who does the questioning must have extraordinary and abundant as well as “almost beyond reasonable doubt” convincing evidence to counter what the archeological establishment has claimed about Gobekli Tepe. In effect, it seems it is no longer asked that the agency who makes the extraordinary claims about Gobekli Tepe provides the extraordinary evidence. But instead, the burden of proof has shifted to those who tend to counter the claims. Be that as it may, the paper will proceed as if the burden of proof is on the counterclaim. The paper will be setting the most stringent of all arguments and criteria in an attempt to support the counter arguments, although it doesn’t have to do so.

As Sagan famously said "extraordinary claims require extraordinary evidence." But the evidence on the dating of this site seems pretty weak.

Note that the infill dates to the Black Sea Deluge period (8000 years ago), when there would be massive population upheavals in the area.

As someone who knows nothing about the field of archeology, it would interst me to know, what are other methods of dating they coud use to improve the estimate, or address to some degree the concerns you raise?

When other known and nearby civilizations are around at the same time, you also date by the foreign goods you find -- Greek coins, or pieces of Bronze-age armor, for instance. In this case, I'm not sure whether that stuff exists or is easily dated.

We have an unbroken timeline of tree rings that goes back a few millennia.

https://en.m.wikipedia.org/wiki/Dendrochronology for those interested in learning more.

this goes much older than that.

Depending on the site, the situation can be quite different. The curves are specific for a certain type of tree in a certain region. Very long unbroken curves are still quite rare. The "better" ones go back about 5,000 yrs. The longest one is the Hohenheimer Jahrringkalender, based on oak and pine trees from Southern and Eastern Germany. It goes back about 12,500 yrs (researchers work to extent it to 14,400 yrs before now).[1] To my knowledge, this is the only curve that goes back to the Gobekli Tepe area, but it is from a different region and therefore not helpful here.

[1] Project Web-site (in German): https://botanik.uni-hohenheim.de/dendro_hoh-jahrringkalender

Reply on news.ycombinator.com