Saturday, January 16, 2016

Walrus Ranching: Early Settlement of the Norse High Atlantic?

Well, finally dragged myself into the library at 3PM the other day, and that's just not going to cut it when I'm working morning shifts. I am pleased to report that I've got probably the most egregious single defence of high unemployment yet uttered by Geoffrey Crowther* of The Economist lined up for you when December, I, goes live, but that'll be next week, hopefully.

After the jump, the headline graphic is a boring table, which is boring, so I'm going to introduce this piece

 --Wait. Actually, I'm going to introduce it with an apology to Christian Keller of the University of Oslo, who has been pushing the "walrus hunters settled Iceland, Greenland and Vinland" angle since at least 2008--

Okay, I'm going to introduce this piece, as far as Google search image results goes, with some nice scenery very tenuously linked to the actual subject. Which, in case you are wondering, is the possibility of pre-871 settlement of Iceland, either a live issue in Icelandic archaeology, per the indefatigable Margret Hermanns-Audardottir, or something almost too gauche to be discussed, per everyone else, although the taboo may be breaking down

Look! It's the Pemberton Valley, courtesy of Pemberton Valley Lodge. It's twenty minutes past Whistler, Vancouver's favourite snow-and-snow destination (think Aspen, only more provincial), and cheaper to stay in.

The relevance here is --Okay, let's see if I can make this brief. Back in the day, people wanted to build railways across North America from the East Coast, where all the people were, to the West Coast, which is a good place to catch a boat to China. Meanwhile, on the West Coast,  people were alert to the fact that there weren't all that many good places in the mountain country that faces the North American Pacific where you can build cities and there supporting hinterlands. Yet, nevertheless, there were more prospects than railways likely to be built. So how do you make a killing on real estate? by influencing the route so that it goes through, say, Reno instead of Spokane, then buying land, not only in Reno versus Spokane, but San Francisco versus Tacoma. This is an American example, of course: for the Pemberton Valley, the issue is Vancouver on the one hand, and, ultimately, Regina versus Saskatoon --towns in Saskatchewan, if you were wondering-- on the other. (No, I'm not going to explain here: brevity.) 

So here's Google Maps.

The route traced here shows the Cariboo Road, built by Governor James Douglas of British Columbia, to service a gold rush off the map to the north. It later became the route of the first two Canadian intercontinental railways. However, I've centred the map to show an alternate route, which follows the course of the Lillooet River from the head of Harrison Lake --actually a widening of that river-- and over a mountain pass to the upper Fraser River, passing through the Pemberton Valley portion of the Lillooet on the way. This was actually Douglas' originally preferred route. built in 1862, it was quickly abandoned as impractical and replaced with the route shown, which goes up the stupendous gorge cut through the coastal mountains by the Fraser River. 

So, uhm, you gently prod me --relevance? It's simply that Douglas may have been less nuts than he seems. The early history of the Pemberton Valley is a bit obscure, but we do know that Squamish medicine women brought potatoes to the Valley, perhaps in the 1820s, and established farms there that were flourishing in 1860. Given that it was the cost of provisions in the gold fields that was driving agitation for a wagon road, it might have seemed logical to Douglas that a road connecting them with the coast should run through the largest active farming area of the period on the mainland. I'm saying "may" because, of course, this runs into the race issue. We generally don't like our First Nations actors jumping out of the box marked "noble savage" and into the one marked "early settler," a running theme on this blog that I won't pursue here because the same issues are going to come up after the jump. Douglas, embarrassed though his descendants may be to admit it, was mixed-race himself, and married a high status First Nations woman. She wasn't a Squamish --that would have been too on-the-nose-- but we shouldn't entirely exclude the possibility that someone in Douglas' circle had an interest in the Pemberton Valley. 

The question remains, on this theory, "Why Pemberton?" The answer, I think, is clear enough. Go to Google Maps yourself and zoom in, and you'll see that the valley's farmland was formed by the internal estuary of the Lillooet at the head of Lillooet Lake. This means that it was probably meadow land, and did not need to be cleared before it could be farmed. Later generations would realise that the amount of farmland in the Valley was utterly dwarfed by that in the Fraser Valley along the later route: but most of that land had to be logged, first. It's this logic, the difference between the value of forested farm land once cleared, and marginal, flooding land kept naturally unforested, which is going to play a part in the settlement-of-Iceland question. 
Looking for something better than a strawman, my all-too-casual eye falls on The Cambridge History of Scandinavia (2003), which has the disadvantage of length, allowing Magnus Stefansson to be a bit diffuse, but: "Most of the Norse settlers [of Iceland] came directly from western Norway." (210; forget accurate citation, gotta go to work in forty minutes.)

Given that ABO blood type groups are highly hereditable, the fact that Iceland's profile is virtually identical to Ireland's, and very different from that of Norway, the majority of the early "Norse" settlers of Iceland were, in fact, Irish. This is not, however, what some kid, somewhere, may non-ironically calls a "drop the mike" moment. We've known this for almost a century, and there was apparently a similar intervention by old time skull-measuring physical anthropologists a century before that. Stefansson, in fact, goes on to discuss this evidence in the next para, waving his hand at magic recent research showing that smallpox epidemics somehow made Icelandic blood look Irish, vice Norwegian. 

One assumes that the argument is over now that mitochondrial DNA studies have shown that about two-thirds of Icelandic female DNA, and one-third of male, is "Irish." Apparently, the Icelanders are descended from chieftainly Vikings and their Irish slave girls. I'd drop a Franzetta graphic if I wanted to get even creepier . . . Distasteful as this is, while you can argue with old science, but new science is off the table. We can get back to defending the honour of old-time Icelanders once the glow is off mitochondrial DNA studies. 

Anyone interested in further details can refer to Ann Humprey's Rutger's seniors honours essay (2009), linked not least because of the title: "They accuse us of being descended from slaves!" I suspect that this is just the tip of the kind of iceberg that became more common in the North Atlantic in the fifteenth century, endangering voyages west, because of this. "Language, Society and Identity in Early Iceland." I'll let you know if Stephen Pax Leonard has successfully deployed historical linguistics to destroy the Viking Age when my copy arrives next week. 

So that's what's potentially lurking in Icelandeic family trees to make everybody all uncomfortable. About Pemberton, you prod?

Well, okay: as we all know (it's been all up on the Google lately), it has long been said that Irish monks preceded the Norse on Iceland. Artificial caves containing rock-cut crosses have been, for centuries, pointed to as hermitages along the south coast of Iceland. What I didn't know, until I got un-lazy enough to look it up on Wikipedia (so still pretty lazy), is that this is one of those square-cornered facts that have had to be fitted into the round holes of Icelandic history from the beginning. Our source is one Dicuil, an Irish monk active in Charlemagne's court, and the best, or at least most widely read geographers of his day. In the Thirteenth Century just as much as the Twenty-First, and for that matter the Tenth, we have had to start with Dicuil's report that Irish monks had made at least one summertime voyage to "Thule," in 795. There is room to argue that "Thule" isn't Iceland, but not a lot. Now, this isn't the fabulous tale of Irish saints voyaging in coracles, and it isn't vague references to a "Greater Ireland" some days sail west of that island. It's facts, printed in the best geography of the day, disseminating from the centre of the intellectual/literary side of the "Carolingian Renaissance." Take that for what it's worth. The more Irish, and the earlier, the more dishonourably slavey, and the more creepily fetishistic, the early family history of Iceland.

Moving on, we have the actual early archaeology of Iceland, hard-to-date artificial caves aside. Traditionally, the bulwark against early settlement is a layer of volcanic ash laid down on Iceland by an eruption datable from Greenland ice cores to 871. Only structures with foundations below the layer will meet the strict criteria of proving early settlement, and there are none.

 It's good to have a scientific bulwark. More importantly, said the first and most optimistic generation of post "settlement layer" archaeologists, many structures were found immediately above the settlement layer. It looked like not only was the first settlement happened exactly when it was said to have happened, but that it was a regular land rush! Even more reason to take the traditional account, which I should probably  note here as "The Book of Settlement," as scrupulously accurate, in spite of being based on three-centuries-old oral history.

The tephnam layer doesn't actually cover the whole of Iceland, as the extreme north and west are excluded. The graphic above indicates the border by the areas where there aren't early archaeological sites, excepting a dig in Reykjavik and the midden sites on the western peninsula, which are dated independently of the tephnam layer. And, yes, it is a pretty indirect way of showing the pattern of ash deposit, but it'll have to do. (Kevin P. Smith, "Landnam: Patterns of Early Settlement in Iceland.")

Now notice that blob in the middle, which is the much cited abandoned hall at Hofstadir near Lake Myvatn (diacriticals and special Icelandic characters omitted because Blogger), high on the run of the Laxa River, a rich trout stream running into the Arctic Ocean. 

From McGovern, Vesteissnsen, et al, "Landscapes of Settlement"
Archaeology on Iceland, unlike Greenland, is dependent on a mix of rescue archaeology, rare in a country where construction isn't exactly booming, and failed sites. It's not surprising to find a failed site forty miles inland from the north coast at 200 meters elevation, now surrounded by heavily eroded upland plains of gravel, although these northern rivers have an old history of being famine resorts because of their trout runs, and a more recent one as vacation fishing destinations. Needless to say, the territory around Hofstadir wasn't always so bleak, but it is still the third choice for a chieftain's estate in the area, which is probably why it was abandoned. 

The immediate salient point about Hofstadir is that it is an area where carbon dating of animal bones (a number of notorious horse and cow remains giving a confidence interval of 770--900AD) and sites adjacent to the tephnam layer argue for very early settlement. In other words, instead of being settled relatively late, it was attractive to what are, by the accepted view, the first wave of settlers. This is where the Pemberton Valley analogy comes in, although to make it make sense, I need to add that archaeobotany has established that pre-Settlement Iceland was covered by a gentle, fluffy comforter of birch forest to 400m elevation, covering all but the central highlands. Hofstadir, and the other nearby "chiefly" estates, are located on finternal alluvium, interpreted as flooding meadows. It is assumed that they were  unforested, making them immediately attractive as cow pasture. While our presumed first settlers must have known the interior of Iceland well to be aware that Hofstadir existed at all, given this, it is a likely location for early settlement. So much for that.

There are three eye-opening aspects to excavations in this area. The first is the discovery of very large quantities of birch charcoal from middens in general, and a specific highland site in particular. The stratification of the midden deposits shows that there were substantial exploitable woodlands as much as two centuries after initial settlement. Additionally, circular deposits on ridgelines in the area yield substantial quantities of iron slag and charcoal, showing significant iron production, and raising the question, in the authors' words, of whether swidden farming was being practiced in early Iceland. 

That one's kind of a cat-amongst-the-pigeons point. Swidden farming might easily be low enough profile archaeologically as to be invisible in below-tephnam layers, and it is a common-enough lifestyle in the forested marginal districts of Iron Age Scandinavia that one could easily imagine it being transplanted to Scandinavia. In ethnogenetic terms, it would also seem very likely that Iron Age swidden farmers from the Baltic basin would be more readily identifiable as Slavs or Lapps than as Nordics, but that's just an ass-pull. 

The third interesting aspect of the excavation is the discovery of walrus bone in the remains. But not just any walrus bone. Complementing discoveries of cached tusks in other early structures, these finds are of the dense maxillofacial tooth bone matrix in which walrus tusks are embedded. Per recent work,**  it has been observed that extracting tusk from matrix without chipping it is hard, skilled work. The finds in Iceland are tentative evidence of a "putting-out" system in which farmers took walrus jawbones home to work them in the winter. Very tentative of course: there's not nearly the amount of evidence necessary to make such a bold conclusion. This is not, however, true of Greenland, where  maxillofacial bone is found on remote farms, and in great quantities at every level of the "chiefly" estate of Sandeness in the Western Settlement, and probably at the bishop's seat of Gardar in the Eastern Settlement, although the 1930s excavation did not twig to the importance of this and document the finds adequately. So in the Greenland case, it is clear that there was such a putting-out system (as well as direct renders of walrus jawbones(?) to chiefs and bishops) through the entire period of the settlement, with Perdikaris and McGovern even believing that they can detect an increasing level of skill in the extraction process. 

Iceland is also different from Greenland in a more obvious way: the walrus breeding colonies were wiped out within a few generations of the settlement. This is interesting in the light of continuing work on early marine archaeology and on Icelandic archaeobotany and -zoology. In the former, it is an open question as to just how heavily "traditional" societies relied on marine resources; and, assuming that they did, whether there were extinction events in the marine ecosystem comparable to the Pleistocene extinctions of big land game, and of the Modern extinctions due to commercial hunting of various marine mammals. In the latter, it has long been argued that the Vikings arrived in Iceland with an inappropirate farming-technical "package," and rapidly depleted its resources, leaving the highly eroded, treeless landscape of today. 

If we are to explain the fact that there were not marine extinctions comparable to the Pleistocene land extinctions, and that the Icelandic forests lasted into the Twelfth Century, we would take a turn towards modern anthropological interpretations, and see our ancestors as being capable of regulating their impact on the local ecosystem and preserving essential resources. That is certainly the broad outline of human history on Iceland, after all: but walrusses are the exception. 

Taking one step further back, into the question of early use of marine resources, we have the problem of the North Atlantic middens. Clearly, humans around the basin have been using fish extensively for very long time periods. Just how long, and for what, once eluded archaeological science. This is no longer the case. Ongoing archaeology in the Lofoten Islands and parts of the coast immediately to the north, show rich chieftainships, inferred from troves of imported silver, gold, glassware and impressive boat houses. Walrus and ivory hunting are posited as early reasons for this, but analysis of codfish remains suggest that they were being preseved for transport by a suboptimal technique (beheaded and dried flat as "clipfish," detectable by retention of the gills in the piece removed from the processing site, unlike stockfish, which has the gills removed as part of drying in the round). This, it is inferred, was part of a "pre-commercial" distribution of Lofoten resources to unknown places. For northern Norway, no more can be said: but the fish remains around Hofstadir are clipfish, telling us that Hofstadir was the destination for cod processed elsewhere, perhaps on the western peninsula. Hofstadir was part of a "precommercial" exchange network, receiving cod from the coast, and perhaps walrus to work, although I wouldn't push that one too far. 

Looking again far afield, excavations at the "Pictish" elite site of the Brough of Bramsy in the Orkneys suggests the storage and redistribution, not of dired cod, but of fish oil taken from cod livers and herring. The inland distribution of fish oil probably predates that of whole fish. It certainlyu does on the West Coast, for the fairly obvious reason that hunter-gatherers, especiallly in winter, have access to protein or nothing from winter-starved animals. The last thing they want in their diet is even more protein: but fat is very welcome! The story may not be so clearcut in agricultural Europe, but if not as a dietary supplement, then certainly as a fuel and an illuminant, fish oil and rendered blubber would have been welcome. 

The salient point about the Second Century Lofoten and Eighth Century Orkney finds is that they substantially predate the "Medieval Fishing Event Horizon" of c. 1050. This is why so much emphasis is put on the idea of a "precommercial" distribution of fish products. Whether the preference for clipfish over stockfish actually supports this, I have no idea, but it is clear enough that people were receiving fish from coastal fishing spots before this became a largescale commercial industry that supported a massive expansion of coastal fishing. 

Okay, so much for that. Now, as a final line of attack, we have patterns of settlement. The received view takes very largely from the story of one "Skallagrim," an aristocrat from Norway who claimed a very large portion of Iceland, built a central estate, and, over time, settled his freed slaves and other dependents on specialise out farms to supply his central estate with various resources. Much recent archaeology has been an effort to refute the Skallagrim story in detail, showing that a much more egalitarian story is likely. The focus of the archaeologists tends to be on groups of families, which is plausible enough given the Atlantic experience of partible lineage land ownership, but it is not the only possible model. In Greenland, where all settlements are open to investigation, we find a pattern compatible with the "Skallagrim" model at Brattahild, with a large estate surrounded by small farms, including ones in marginal landscapes settled very quickly after the intitial phase. However, at Gardar, we find an even larger centre surrounded by a wide "buffer zone," bereft of farms. It is so large, and so well captialised, that Gardar even has an ingenious irrigation system to maintain its meadows. Gardar was, it seems, a production centre on a scale larger than a single  household, or even a chiefly one. This is not surprising for a cathedral site, which would have had an attached canon chapter of clergy who would presumably have worked for a living most of the time. 

Now, Perdikaras and McGovern infer that the landscape at Gardar was remade in 1147 to make an estate fit for a bishop, which is certainly something that could have happened. I am probably telegraphing my preferred explanation, however, that Gardar was founded pristinely as a large estate. That is, that it was a monastic settlement, probably a distant satellite of one of the large monasteries of the Atlantic Northwest --perhaps an unknown "Pictish" church settlement on the Orkneys? Dicuil's monks might or might not have been going to Iceland in search of walrus, two centuries earlier. While that doesn't sound like a properly monastic pursuit, the latest evidence is that someone was, and that they were involved in a precommercial exploitation of this luxury resource, so widely used in the artistic production of the Atlantic monasteries. Who better than monks to secure these resources?

The final piece of the puzzle, then, is the settlement itself. It would seem that, about 880, Iceland very abruptly ceased to be prized as a source of walrus, and came instead to be used as a source of food, iron, and perhaps other goods. The economical explanation is of a westward shift in walrus production --Greenland drove Iceland out of the market. It will, anyway, do for now. 

*When it is a seller's market for labour, only the high paying industries will get enough labour. This will lead to "distortions" which will inevitably cause a "slump." So the only way to avoid high unemployment is to have high unemployment! Or to use state coercion to force labour into "vital industries" which cannot pay high wages. It's hard to tell for sure, but I think Crowther prefers the "state coercion" angle. The editorial office of The Economist was an odd, odd place in the winter of 1945. What's really surprising is that if he'd just brought up inflation, he'd be close to modern economic orthodoxy. Perhaps he's just too rushed to express himself clearly? 

**Sophia Perdikaris and Thomas H. McGovern, “Codfish and Kings, Seals and Subsistence,” in Torben C. Rick and John H. Erlandson, eds., Human Impacts on Ancient Marine Ecosystems: A Global Perspective (Berkeley: University of California Press, 2008) 187—214. 


  1. Hey, I read Language, Society and Identity in Early Iceland a few years back. Don't remember there being much in it that would be useful for your thesis (though that wasn't what I was looking for), but it had a very interesting explanation for why Icelandic vocabulary and grammar have been so improbably conservative.

  2. The oh-so-speculative Farley Mowat's "Farfarers" I remember as an interesting case for the importance of walrus ivory in medieval Europe - I don't know if it was used in late Antiquity or not - and the idea (likely enough mistaken in the specifics) of a North Atlantic seagoing culture/s.

    Though this man has an alternative view and here

    Though you'd be a better judge on all this.

  3. Walrus ivory is a prominent early Medieval luxury good.Further, it was used for cover panels on illuminated manuscripts, which brings us into the heart of monastic production.

    However, it is also a signifier for a less visible trade in rendered blubber, highly appreciated as an illuminant. This is more of a bulk trade good, and it's a bit hard to imagine it being exported from Greenland, or even driving voyages to Iceland. It would account for the early wealth of the Nordmark "chieftanships," prior to the stockfish trade, though.

    As for the pan-Arctic confederation of the skin boat peoples --well, duh. Except that they didn't cross the Atlantic. The migration was pretty clearly counter-clockwise, not clockwise, and, if recorded examples are anything to go on, was pretty fast when people had reason to move. I highly doubt that the Mesolithic peoples of the British Isles needed to be taught how to take seals from skin boats by Finnish migrants, who, given all that we know about the rebound of the Scandinavian peninsula (and the prehistory of Finno-Ugric, to touch on a sensitive point), probably didn't yet exist. Don't forget that the use of marine resources fell off considerably in Neolithic times. Again, Neolithic peoples had all the protein they needed, and the seals were already being depleted. (Probably.)

    Walrusses (and narwhal) are another matter, never mind right whales. Taking these larger animals virtually requires cooperative hunting from boats. We're pretty clear that this isn't a subsistence hunt. It's unreliable, and it is dangerous.

    So when we see it, it is already entangled in commercial exchange --and that exchange can't be in "trinkets," but rather goods that can substitute for subsistence activities. Iron (and ideally blacksmiths, too), and blankets, thank you very much. That stuff isn't crossing the Atlantic in curraghs --nor, probably, even knorrs.