Göbekli Tepe: Houses of the Holy?

In the series introduction, I asked whether Göbekli Tepe was (as the excavator Klaus Schmidt suggests) an archaeological or metaphorical Stairway to Heaven. Continuing the Led Zeppelin riff, a better question might be whether Göbekli’s megalithic structures were Houses of the Holy.

E.B. Banning suggests something along these lines in “So Fair a House: Göbekli Tepe and the Identification of Temples in the Pre-Pottery Neolithic of the Near East” (Current Anthropology 2011). Banning exhaustively reviews the Göbekli evidence and challenges the prevailing interpretation of the site. This is precisely what was needed and it shows archaeology working as a science.

In his very first site report from 1998, Schmidt had already concluded that Göbekli was a ritual center and claimed the “archaeological evidence is overwhelming, as the function of two partially excavated pillar buildings irrefutably prove.” As Banning’s article shows, the evidence is not overwhelming and the claims that have been made about Göbekli are refutable.

After surveying the evidence and various claims made by Schmidt, Banning offers an alternative view. It is a view informed by lessons learned from history and theory. The history comes from another famous Neolithic site, Çatalhöyük, which was first excavated in the 1960s by James Mellart. Mellart interpreted richly decorated structures as ritual “shrines” and claimed they were not residential. It was later found that the so-called “shrines” were in fact houses.

The theory comes from the ethnographically informed realization that binaries such as sacred/profane and secular/religious are post-Enlightenment Western constructs rather than human universals. By extension and association, this means that the ritual/domestic binary is either suspect or provincial. None of these binaries can be projected uncritically back in time and mapped onto 11,000 year old ruins. Historically situated and modern conceptions are not reliable guides to ancient cosmologies. And given what we know about most non-Western cosmologies, it seems unlikely that the Göbekli world was constructed or perceived through these binaries.

It is more likely that the sacred/profane existed on a continuum and were conjoined, as were ritual and domestic activities. With these things in mind, Banning observes: “The point is not that specialized shrines are incompatible with domestic ritual but that evidence for ritual or conspicuous symbolism does not automatically imply specialized temples.”

These salutary reminders out of the way, Banning turns to the nub of the Göbekli issue: “The question is whether the evidence justifies the site’s interpretation, as its excavator argues, as a hunter-gatherer cult center with no domestic occupation at all.” To answer it, Banning examines several aspects of the site: (a) the famous T-shaped pillars which Schmidt asserts were free-standing, open-air monoliths (similar to those at Stonehenge), (b) the supposed lack of evidence for household or domestic activities, (c) the alleged lack of access to water and (d) the ostensible absence of domesticated plants-animals.

For each, Banning points to contradictory evidence and suggests looking for additional corroborating or refuting evidence. In some cases this involves nothing more than looking at the existing evidence differently, more closely, or without preconceptions. In all cases, Banning finds the evidence or lack thereof equivocal.

If the Göbekli structures were in fact unroofed, it surely follows they were not houses. Beginning with a structural examination of the pillars, Banning suggests they are placed and buttressed in a manner that would have supported overhead wooden beams, which in turn would have been thatched. There are several hints (ranging from grooves and notches to wood) that this may in fact have been the case, and Banning has sketched one possible layout:

Göbekli’s T-shaped pillars are arranged in the round and may seem completely unique (which they are in terms of size alone), yet it turns out that similar pillars and arrangements are found at other Neolithic sites in the area, and in several cases these structures are residential.

Aside from the structures themselves, the most remarkable feature of Göbekli is that it was discovered virtually intact. Fortuitously for archaeologists, Göbekli’s users (whether occupants or visitors) periodically filled earlier and older structures with surrounding debris and built on top of them. After its final use, the site was again filled. This explains why the site went unrecognized for so long; it looked like just another hill.

Banning is particularly interested in the huge amounts of fill material that were used and which he suspects was created on site as a result of occupation:

Notably, the site’s deep deposits also exhibit high densities of lithics—including a variety of points, scrapers, burins, and sickle blades—as well as evidence for “all stages of production.” One might expect to find stone tools related to the quarrying and manufacture of limestone monoliths and debris from the tools’ manufacture, but those in the fills, at least, are not noticeably different from what one might expect to find in a domestic deposit.

There is also abundant animal bone while dark earth found in the soil horizons may be anthropogenic, probably associated with the high density of bone fragments and other organic materials. Plant remains are not well preserved in these deposits but include a broad suite of edible wild seeds and the charcoal of trees such as ash, almond, poplar, and Brant’s oak that could have furnished both fuel and roof timbers.

In addition, Banning identifies possible hearths or hearth rings and mortars that would have been used to process grain. Some of these are bedrock mortars, which I happened to notice — near what appears to be a large cistern — in privately taken pictures of the site. The cistern is interesting because it speaks to the issue of water: Schmidt asserts that Göbekli had no easy or reliable access to this essential resource.

It would be unwise to assume that the lack of water at the site today is indicative of the situation 11,000 years ago:

[D]uring Göbekli Tepe’s occupation around 8000 cal BC, during the early Boreal period, the climate was considerably more humid than the current 450 mm of mean annual precipitation would suggest, and the water table was likely rather higher, potentially with springs closer to the site that no longer exist. Deforestation and modern irrigation projects have also had serious impacts on local water tables and streamflow, making the present distribution of water a poor indicator of Neolithic water sources.

Indeed, the many different kinds of (moisture-loving) plant remains found at the site suggest that water fell or flowed in amounts sufficient to nourish them.

In Schmidt’s estimation, these plant remains are — like the abundant animal remains — significant because they do not show signs of domestication. Aside from the difficulties of identifying domestication on the basis of morphology (with domesticated seeds being larger and domesticated animals smaller), Göbekli is a transitional Neolithic site. Hunting and gathering did not simply stop when people began planting seeds and controlling animals or domesticating them.

During this transitional period, plants and animals on the way to domestication may not look like their wild counterparts or may be “tweeners.” Of course some of these plants and animals were never domesticated; their presence is best explained by a mixed economy: there was some hunting and gathering of non-domesticates while at the same time others were being selected for domestication.

Where does this leave us? Banning has an idea and states it forcefully, though not in precisely this order:

While there is no doubt that Göbekli Tepe is an important site and that aspects of its structures were symbolically loaded, the claim that the site had no residential occupation is simply not credible.

Most likely, either the famous “temples” are actually houses or houses lie elsewhere on the site and are simply not represented or not yet identified in the excavated sample.

In short, there is no strong reason to assume that the people who used the buildings at Göbekli Tepe, in any stratum, were not Neolithic villagers.

Ignoring even the possibility that some of the claimed shrines and temples at Neolithic sites may have been houses or other types of buildings, however, could distort our interpretations not only of Neolithic religion but of nonreligious aspects of the communities that inhabited or used those sites.

So fair a holy house indeed. In the next and final post in the Göbekli series, we will synthesize the materials from the previous ones and take stock of the whole.


Banning, E.B. (2011). So Fair a House: Gobekli Tepe and the Identification of Temples in the Pre-Pottery Neolithic Current Anthropology, 52 (5), 619-660 : 10.1086/661207


Did you like this? Share it:

12 thoughts on “Göbekli Tepe: Houses of the Holy?

  1. Maureen Lycaon

    You’ve been in my RSS feed for a few weeks now. I just want to thank you for doing this series; I’ve been interested in Göbekli Tepe since reading about it in Mithen’s “After the Ice: A Global Human History 20,000-5000 BC”.

    I begin to wonder if the people of cultures such as Çatalhöyük and Göbekli Tepe were mentally very different from people of any living culture, as Julian Jaynes once suggested, which led them to embed so much rich symbolism in their living space — whether the difference was having bicameral minds or whatever.

    In any case, I think you and Banning are right: it’s too easy for us to project our own concepts and thinking into cultures of the distant past.

  2. Cris Post author

    I’m glad you are following and enjoying! When you say “mentally” different do you mean an actual difference in brain structure or function? My educated guess would be that there have not been very many neurophysiological changes to human brains in the last 40,000 years, which is about the time we begin seeing widespread evidence of symbolic thinking and behavior. While there surely has been some ongoing brain evolution in the last 40,000 years, we don’t see any archaeological or other evidence that might suggest people living 11,000 years ago were in any way different from people alive today. What has changed is culture. I think this alone is sufficient to account for the chasm that seems to cleave people from that time and people from our own. I would bet, however, that if you or I could time travel back to Gobekli, it wouldn’t take us long to get into the swing of things.

  3. Tom Rees

    For sure there haven’t been meaningful genetic changes, but Julian Jaynes suggested that there was a cultural shift that radically reprogrammed the way our brains work. He talks though some of the archaelogical evidence. I really love Jaynes’ thesis – even though I reckon he’s almost certainly completely wrong!

    Anyway, getting back to the point. Isn’t it the case there there is pretty much no evidence for settled communities at the time of Gobekli? Catalhoyuk is later, right?

  4. Cris Post author

    Sounds like I need to read some Jaynes because it sounds interesting, a bit like Merlin Donald’s contention that external storage in the form of marks, glyphs, and eventually writing radically altered our conception of the world and interaction with it.

    On point, it is funny you should ask this because today I just read a series of papers on Neolithicization in the Levant, and there is no doubt that Natufian hunter-gatherers in the Levant (near Gobekli) were becoming sedentary thousands of years before Gobekli. Natufian settlements begin appearing approximately 15,000 years ago!

    So there is a long history of intermittent and variable sedentism in that region leading up to larger sites such as Gobekli. Some of that sedentism surely involved incipient agriculture and wild animal management or control, which in turn led to domestication.

    And there are more traditional looking “villages” that are contemporaneous with Gobekli, but which lack megaliths. It is the megalithics that make Gobekli special; they are currently the oldest known.

    And yes, Catalhoyuk is later; it gets started about the time that Gobekli is filled and abandoned, circa 7,500 BCE.

    Did you get all those Rodney Stark articles I sent? It was a large-bit email and I’m not sure it went through.

  5. Maureen Lycaon

    There’s no evidence for *physical* changes in brain structure over the past 40,000 years, I agree. Neurological changes are another thing entirely, and this is what Jaynes, at least, claimed. To use a computer analogy — you can run several different operating systems (Windows 7, Ubuntu Linux, even Macintosh) on the same piece of hardware. So in humans, the “hardware” shows no clear signs of change between Gobekli and now, but the “operating system” and the “software” might have changed radically.

    A history film of my youth claimed that if we were stranded in the medieval European world, we would probably *never* fully be able to adjust, because the mindset of that time was so different from today’s. A much larger time gap lies between our time and that of cultures like Catalhoyuk or Gobekli. We can’t be sure they were just like agricultural societies of today, because even the most remote cultures have at least some contact with more complex cultures and may have had more complex pasts. These were among humanity’s early experiments in sedentary town living. This is why I think their inhabitants might have mentally been very different from any people living now.

    Jaynes would have asserted that their minds were what he called “bicameral” — that consciousness is actually a recent thing dating from circa 1200 BC or so. You’d have to read “The Origin of Consciousness in the Breakdown of the Bicameral Mind” to understand fully what that means; I highly recommend it despite its age.

    My own pet idea is that people then were not yet mentally well-adapted to living together in large numbers: crowded on top of each other yet cut off from each others’ view much of the time, in too great numbers for everyone to know each other, and so on. They had not yet had millenia and countless generations to adapt to the stresses of urban living. They put huge quantities of ritual imagery in their dwelling places as part of a mass of ritual and conformity that dominated their daily lives, and they *needed* that ritual and conformity for their society to stay coherent and not dissolve into madness.

    (And no, I’m not an anthropologist or archaeologist, just a non-scientist with a headful of notions.)

  6. Cris Post author

    First, I love the comments and ideas. Second, aren’t neurological changes physical changes? Surely so. Neurons are physical, as are the chemicals involved in neural activity. Unless Jaynes is suggesting quantum changes that have never been verified, I don’t see how any kind of neurological change wouldn’t be a physical change.

    Third, this seems like a less than parsimonious explanation. Human behavioral changes are usually due to culture and not neurology, except for cases of pathology. Humans and human brains are incredibly plastic, and there probably has been intense selection pressure for learning variability or behavioral plasticity. Humans are so successful, in part, because we can adapt to just about anything. We live in almost all environments under all conditions imaginable.

    On your suggestion, I just downloaded a bunch of Jaynes’ scientific publications, including one on the bicameral mind, so I’ll write about it after I’ve read. In the meantime, you might be interested in my article on the human brain and brain evolution (under my Bio tab as a pdf). I have a section on consciousness which is clearly at odds with Jaynes’ suggestions.

    I like your pet idea except that early Neolithic people weren’t living in what we consider large groups. Population estimates for places like Catalhoyuk are in the several hundreds, which doesn’t seem like such a big deal. Not many of these early villages ever exceeded 500 people, and humans can track up to 500 people in a social network. What you suggest may have been a problem later, when villages became cities. That would have been around 5,000 years ago in Mesopotamia.

  7. John Bennett

    I wondered if it was an art museum, and they would bury the art to keep it from being destroyed when they were invaded. Then when the coast was clear again, they would decide it was too much trouble to dig up the old stuff and they’d just make new art instead.

  8. Cris Post author

    This idea could of course be framed as an hypothesis and then tested. My guess is that the hypothesis would be falsified.

  9. Per-Allan

    Of course there is a probability that hunter-gatherers lived permanently in one place for thousands of years in huge megalithic structures that were both residential and places for worship. But I don’t see the logic in it.

    The foraging lifestyle is a nomadic lifestyle, maybe not always in areas especially rich in game and wild plants, but sooner or later the group of people will have to move. To build permanent monumental structures to live in, under those life circumstances, is not a very smart thing to do. I think the hunter-gatherers would agree with me.

    But suppose that the area in question was immensely rich in food resources, which allowed for that long period of permanent settlement:
    The structures at Göbekli Tepe took a long time to build, with a large workforce of people. During the building period, where did they live? I mean, their house wasn’t finished yet? If they lived in temporary dwellings nearby, why did they abandoned them to move into the big stone house once it was finished? It doesn’t make sense.

    Banning proposes a theory that Göbekli Tepe’s structures were roofed, and provides a sketch of pillars with a sloped roof on top. The problem here, as I see it, is that the T-shaped pillars of limestone are flat on top. Why, if the roof is sloped? The ability to shape the stone was apparently not the builders biggest challenge during the construction. One could suppose here, that Banning wanted to see a roof on top of those pillars, no matter what shape they actually had.

    But suppose there was a roof. And the monumental structures were intended for combined religious/residential use. If there actually was a permanent settlement for hunter-gatherers for thousands of years, with sustainable access to water all that time and immensely rich resources of food that for the largest time of Göbekli Tepes existence didn’t require domestication of plants or animals, then there is a probability. But as I have said, it doesn’t make much sense.

Leave a Reply