Comparing Levels of Subsistence Stress amongst Norse Settlers in lceland and Greenland using Levels of Bone Fat Exploitation as

Enaironmcnfol Archncology 8, 2003; pp. 119-128

Comparing Levels of Subsistence Stress amongstNorse Settlers in lceland and Greenland using Levelsof Bone Fat Exploitation as

Alan K. Outram

AbstrnctThe background to the Icelandic and Greenlandic sites under investigation is outlined and prior work on the Norseeconomies of the two islands is discussed. The importance of fat in the diet and the use of levels of bone marrow andgrease exploitation as an indicator of subsistence stress are explained. The methodology for establishing levels of bonefat exploitation is outlined. This methodology involves the detailed study of fragmentation levels of different types ofbone, study of bone fracture types and many other taphonomic indicators. The results of the study are described anddiscussed. On Greenland, the Norse inhabitants exploited almost all available fat from land mammal bones, leavingonly the ribs. lt is argued that this indicates a severe level of subsistence stress amongst the Greenlanders that is mostlikely related to a seasonal dearth in resources. On lceland, whilst a certain amount of bone marrow is almost certainlyexploited, the settlers appear to almost totally ignore the potential to exploit bone grease. This is likely to be indicativeof a much more healthy subsistence economy than on Greenland. These results are discussed in relation to differingclimate, availability of good soil, fishing practices and seasonal rounds.

Kevuords: Nor<sr:, lcr:r.ANo, Gr<rr:Nr-a,No, BoNL MAnltow, BoNE GREASE/ Far, SunsrsrENCE

an Indicator

Background

lrlorse lceland and Creenlnnd

The Vikings had discovered Iceland by no later thanAD 850 and the subsequent settlement of the island,the lnndndm, took place over the period c. AD 870-930 (Byock 2001, 9). This early society was one ofsmall farmstead units with little centralised govern-mental control (Byock 2001). The settlers quicklyadapted to the different environment presented bythe island and exploited both the marine and landresources that were available. The most commondomestic animals represented on farmsteads werecattle and sheep, with goat, pig, horse and dog alsopresent (Amorosi 7992, 123, 772; Tinsley in prep.).

This pastoral economy was supplemented by limi-ted exploitation of marine mammals such as com-mon and harp seals, the occasional whale (Amorosi1992, 124) and a reasonable amount of fishing,principally for gadids, with the presence of somesalmonids (Amorosi 7992, 124; Tinsley in prep.).Although limited by wood supplies to fishing fromrelat ively smal l boats, the r ich coastal f ishinggrounds could yield good catches (Byock 2007,47).A key economic feature of this activity is that fish,in particular cod, can be preserved by air-drying inwinter to produce stockfish, without the need forlarge amounts of salt (Perdikaris 7999,390) or fuel

Recciaed Junc 2002, rert iscd nnnuscript nccepted October 2002.Author's addrcss: Dapartment of Archaeology, Uniaersity of Exeter, Laaer Buildittg, North Park Rond, Exctu, EX4 4QE, UK,

Email : [email protected]

720 A. K. Outram

for smoking. The settlers therefore had access tostored fish and also stored milk products such asskyr (coaeulated milk) (Byock 2001, 47), a factorthat takes on more importance in discussions below.The Icelandic Norse had no arable crops, but thegrowing of sufficient fodder for livestock was ofgreat importance (Zutter 7992). The early settlershad the advantage of living in the Little ClimaticOptimum (c. AD 870-7170) (Byock 2001, 57),but,with the onset of colder weather leading into theLittle Ice Age and soil exhaustion, productivity fell(Zutter 7992,144). The init iallnndndnr saw c.60"/. ofnatural vegetation destroyed (Zutter 1992, I39)resul t ing in increasing erosion from the tenthcentury onwards (Byock 2001,56). In these later,harsher conditions, some farmsteads went out ofuse but others adapted. Later medieval and post-medieval Iceland saw a greater dependence uponmarine resources with some farmsteads seeingvastly greater exploitation of seals (Amorosi 7992,724). By the fourteenth century, Icelanders becamemore and more reliant upon the commercial exploi-tation of stockfish, trading with Norway, Germanyand England (Byock 2007, 44).

Creenland was reached by the Norse in c. AD985 and was settled in two main areas, now calledthe Eastern and Western Settlements (BuckIand etnl. 7996). The economy was very similar but notidentical to that on Iceland. It was based upon thepastoral exploitation of domestic cattle, sheep andgoat and the hunting of wild birds, seals and caribou(McGovern 1985; McGovern et aI.7996).It is worthhighlighting some of the differences, in comparisonto Iceland. Icelandic settlers did not have access tocaribou/reindeer. The few specimens of reindeerbone found on Iceland can probably be attributedto the importation of craft materials (Amorosi1992,723). The most important difference is that, whilstthe Creenlandic Norse did have access to fish, theirexploitation of it was negligible. Numbers of fishbones in Greenlandic middens are often limited toone or two per species and could easily have foundtheir way there in the bel l ies of hunted seals(McGovern 1985, B0; McGovern et al. 7996, 775;Buckland et sl. 7996, fig. 3).Furthermore, the absenceof such bones cannot be blamed upon recoverymethods (much fine sieving has been employed) orpreservation conditions (which are very favourable)(McGovern 1985, B0).

If availability of good pasture and the ability togrow enough winter fodder was a concern for theIcelanders, it was even more a matter of life anddeath for the Greenlandic Norse. The grazingseasonin Creenland was probably very short indeed, Juneto early September (McGovern 7985, fig.7), and theanimals spent most of their lives in byres dependantupon fodder. The climatic downturn leading into

the Little Ice Age resulted in repeated short andpoor growing seasons that may have resulted in thefinal abandonment of Greenland by the end of the15th Century (Buckland et al. 7996).

The Norse groups living on Iceland and Green-land were culturally similar and both attempted tooperate largely pastoral economies in diff icultenvironments. It is clear that the environmentaldifficulties in Greenland were more severe. It isalso clear that there were significant differences ineconomybetween the two, particularly with regardto fishing. This paper seeks to use the extent of bonefat exploitation as a proxy for comparing levels ofsubsistence stress in Iceland and Greenland inrelation to their overall level of marginality and totheir specific seasonal economies. The rationalebehind using exploitation of bone fats as a proxy isoutlined below.

Why bone fats are importnntThe acquisition of fat is of very great importance toany society that is dependent upon animal productsalone for survival. As already established, theNorse in both Iceland and Greenland did not growcrops and therefore did not have any source ofcarbohydrate. Protein is not easily metabolised toprovide energy and the intake of large amounts ofprotein (i.e. large amounts of lean meat) withoutaccompanying carbohydrate or fat can lead tosevere illness (Speth and Spielmann 7983; Speth7983; 7987; 1997). Fat is also valued for its highcalorie count, having a higher energy value thaneither carbohydrates or proteins by a ratio of 9:4(Mead et nl. 1986; Erasmus 7986). Given that theNorse peoples in question had no access to reliablesources of carbohydrate, the acquisit ion of fatswould have been of great importance. But why arebone fats so important?

Bones represent one of the most reliable sourcesof fat in land mammals. This is because they arelast in the fat-mobilisation sequence of starvinganimals (Cheatum 7949; Brookes et ql. 1977; Peter-son ef nl. 1982; Davis et sI. \987). This means thatwhen animals are lean and there meat contains verylittle fat, there will still be considerable amounts offat within their bones. This fat is held within themedullary cavities of long bones but also withinthe honeycomb of cancellous (spongy) bone thatcan be found in long bone epiphyses and most axialelements. Bone marrow can be recovered easily bybreaking into the shaft of a bone, but the grease incancellous bone can only be extracted by frag-menting the bone and boiling the fat out. Suchrendering for bone grease has been noted in manyethnographies (e.g. Binford 7978, 758; Leechman7957; 1954; Wilson 7924\.

V

Compnring Leuels of Subsistence Stress nmongst -Ay'orse Settlers in lcelsnd nnd Greenlnnd 127

The above explains the importance of bone fatsto the Norse, but they are important to the archae-ologist because one is actually left with physicalevidence for their exploitation. The bones surviveand bear the scars of fat exploitation in the form ofspecific patterns of fragmentation and fracture (seebelow for details). What is more, it is possible to seehow much effort went into obtaining fats. It isrelatively easy to break a bone for its marrow, butthe comminution of cancellous bone for greaserendering requires much more effort and is muchfurther along a clear sequence of diminishing returnsthat may be indicative of greater desperation.

Two Norse farm sites in Greenland have alreadybeen studied for levels of bone fat exploitation bythe current author (Outram 7998; 7999). Prior tothis work it had already been suggested that, fromthe lack of fat-loving diptera lavae in Greenlandicmiddens, most fat must have been removed frombones by rendering before deposition (Buckland efnl. 7996). The results of this author's work con-firmed this, showing that almost all sources of bonefat were exploi ted to a level suggest ing greatsubsistence stress. Furthermore, similar studies onPalaeoeskimo sites in Creenland, revealed that thePalaeoeskimos could afford to ignore many sourcesof fat (Outram 7998;7999).In this paper the resultsfrom the two previously studied farmsteads inCreenland are summarised and compared withresul ts f rom two newly studied farmsteads inIceland.

The sites being inuestigated

The two Greenlandic sites previously studied bythe author (Outram 7998;7999) are both situated inthe Western Settlement of Greenland. Sandnes (V51)is the largest farmstead in the Western Settlement,having a church, and was possibly the centre of thecommunity (McGovern 1985). The site was firstexcavated in the 1930s and the bones were reportedon by Degerbol (7936), but the sample analysedhere was taken from excavations carried out in the1980s (McGovern 1985; McGovern et s\.7996) as thematerial was properly sieved. According to Mc-Covern et al. (7996, Table 3) the assemblage consistedof (in terms of '/"NISP, Number of Identif iableSpecimens) 7,56"/" cattle, 77.72"/. sheep, 74.72"/"caribou and 31 .23"/,, seal. The second site wasNiaquussat (V4B), which was excavated in 7976/77.The site is of lower status than Sandnes and hasfewer cat t le 7.75-2.96% TNB (Total Number ofBones) (McGovern 1985, Table 6). Sheep / goataccount for 9.77-77.27"/" TNB and caribou for 4.75-6.56%. There was considerably more sealing at79.27%-84.99"/" TNB (ibid.). It should be noted,however, that Outram (7999,116) has argued that

the heavy levels of comminution of land mammalbones, that is not observed in the sea mammalassemblage, may have resulted in the overestimationof seal numbers.

The following dates are available for the Green-landic samples. At Sandnes (V51) the ear l iestcalibrated radiocarbon date range obtained frommidden layers is AD 7025-7275 and the latest isAD 7278-7387 (McGovern in prep.). The samplefrom Sandnes was from the upper part of themidden, so represents later end of this range. TheNiaquussat midden has yield an earliest date rangeof AD 960-1760 and a terminal layers have yieldeda date range of AD 7284-7405 (McGovern ob:et nl.7984). The sample, studied by this author, was acolumn taken through the deepest part of themidden and represents the entirety of that daterange. There was no apparent change in the frac-ture and fragmentation pattern in relation to depth.It has been noted that species representation in theWestern Settlement does not appear to changegreatly over time (McGovern 1985). The samplescan be taken as representative of the bulk of theoccupation period in Greenland.

The newly studied material from Iceland consistsof a large sample from Sveigakot (SVK) and a small,but still meaningful, sample from Hofstadir (HST).Both of these sites can be found in the Lake Myvatnarea of Northern Iceland. Excavations at Sveigakotbegan in 7999 and the most recent at Hofstadirstarted in 7995 (Tinsley in prep.). At both sites,excavated material was sieved using a 4mm mesh(Tinsley in prep.). The faunal assemblage in the 9thcentury at HST and SVK contains 50'lu domesticfauna (mainly caprine and cattle, but also pigs andhorses) with the rest of the assemblage made up ofvar ious f ish species ( t rout , char, salmon, cod,haddock, saith) (Tinsley in prep.). There is a trendfor an increase in wild species, including more birds,in the 10th century, which is reversed in the 11thcentury in favour of domesticates (Tinsley in prep.).The Hofstadir sample taken for the present studycame from HST9B Area G, layer 7, an early phase(layer 6 above it has AMS date range of AD790 -

1000) (Tinsley, in prep.) , and consists of 2602fragments of bone. The larger Sveigakot sample of70,996 fragments came from SVK00 Area T, context055 which has two AMS dates (AD 910-1030, AD980-1140) (Tinsley, in prep.). Both samples thereforedate to the period before climatic downturn.

Methods

Below, the methodology for establishing the extentand nature of bone marrow and grease exploitation,from the study of bone fracture and fragmentation,

722 A. K. Outrnnt

wil l be briefly outl ined. Detailed discussion ofvarious aspects of the methodology can be foundin Outram (1998; 2007; 2002) and the speci f icmethodology, as applied to the previous Creen-landic study, can be found in Outram (1999).

Pntterns of bone fnt exploitntion

The most obvious source of fat in bone is bonemarrow. The exploitation of bone marrow is very

simple. One has to break into the medullary cavityof the long bone. This will occur whilst the bone is

still relatively fresh and contains its organic com-ponent. As such, the fracture pattern created willbe helical and at the point of impact there will be adynamic impact scar (see Morlan 1984; Johnson7985; Outram 2002). The overall pattern resultingfrom this activity will be one of undamaged articu-lations and axial elements with helical splinters ofdiaphysis bone.

In order to exploit bone grease, one has to renderthe fat from the cancellous bone found in epiphysesand axial elements. In most ethnographic examples,e.g. the Nunamiut (Binford 7978), the Hidatsa(Wilson 7924) and the Loucheux (Leechman 7951;7954), this is carried out by storing up the ap-

propriate parts until there are sufficient to warrantprocessing, then comminuting them and renderingthe fragments in boiling water. As the water cools,or is cooled deliberately with snow or water, the fatcoagulates at the surface and can be skimmed off.The resulting pattern wil l be one of large numbers

of very small pieces of cancellous bone accomPaniedby larger, helical shaft splinters.

The extent of exploitation can be assessed byquantifying how much of the assemblage has beenfragmented in this way. One can assess the level ofmarrow exploitation through noting how many

shaft cylinders have been left unbroken and assess-ing the extent to which the fragmentation of cylin-ders was the result of deliberate, helical fracture. Inthe case of grease processing, one can assess hor,r,'much cancellous bone has been left unprocessed(i.e. not comminuted). In some cases of incompleteexploitation it will simply be that a random selectionof grease-bear ing elements have been ief t un-

processed. In other cases, a selection for particulartypes of grease wil l have occurred. The grease fromaxial and appendicular elements can be different innature (Binford 7978,159) and some peoples may

select to exploit one in preference to the other. It is

possible that particular elements wil l be avoided as

not being worth the effort, or because they offer

poor quality fat.In order to study these patterns archaeologically,

one must establish how fragmented the assemblageis and which types of bone (in terms of marrow

and grease ut i l i ty) are f ragmented and whichpotential sources of fat have been left unprocessed.One also needs to assess the evidence for deliberatefracture and record fracture types. Alongside this,one must keep a record of other taphonomic factorsthat can lead to attrition, so that one can know howmuch the fragmentat ion may be the resul t ofhuman action rather than post-depositional tapho-nomic processes. A11 the factors must be consideredtogether.

Fragttentation leaels

The method used for recording the levels of frag-mentation at both the Greenlandic and lcelandicsites is as follows. All fragments were included,whether identifiable or not. Whilst identification tospecies and element may not be possible, suchfragments still carry valuable information in theform of size, fracture patterns and bone type. It ispossible to tell cancellous bone from diaphysis boneon even very small fragments and such informationis very important in the context of this study. Theentire assemblage was divided into size classes (by

maximum dimension). The size classes used were<20mm, 20-30mm, 30-40mm, 40-50mm, 50-60mm,60-B0mm, B0-100mm, 100+mm and part and wholebones. Whole bones clearly have not been exploitedfor grease at all. Part bones include bones that arenot whole but represent whole units that could havebeen exploited for grease but were not broken up.Part bones include entire epiphyses and completevertebral centra. In this study the part and wholeclass have been presented as a single (P/W) class,representing unexploited sources of cancellousbone. One differencebetween the Greenlandic study(Outram 1999) and the Icelandic study is that forIceland the P/W class has been divided into small(<50mm) and large (50+mm).

Quantif ication of the size classes was by num-ber and mass. Whilst numerical data was collected,only the mass data is displayed in this paper. Themass data is more meaningful because it repre-sents actual amounts of bone present. Clearly oneunbroken large bone represents the same amountof potential fat as a similar element broken intomany pieces, yet the latter would be representedby many hundreds on a numerical count. By mass,both would be suitably equal. Brink (1997), in hisstudy of bison, concluded that dry bone mass wasan accurate predictoi of elements' bone greaseutil i ty.

For each size class, a distinction is made betweenwhether the bone is cancellous or cortical in nature.For large size classes, distinction was made betweenaxial cancellous bone (other than ribs), ribs, articularbone from appendicular elements and diaphysis

Compnring Leuels of Subsistence Stress amongst Norse Settlers in lcelnnri nnd Crce nlnnd 723

bone. This enables one to see, inuti l i ty, which types of bone hadand to what level.

terms of bone fatbeen fragmented

Fracture types

When dynamically fractured in a fresh state, densediaphysis bone creates a very distinctive fracturepattern. Such fractures are characterised by helicalf racture i ines radiat ing out f rom the point ofimpact. The fracture surface will form either anacute or obtuse angle to the cortical surface of thebone. This fracture surface wil l also tend to besmooth in texture (Johnson 1985; Morlan 7984;Outram 7998; 2007; 2002). As bones dry out theydevelop small cracks that interfere with the fractureline, creating roughness or steps, hence affectingthe fracture shape and surface texture. As bonesloose their organic content they react differently toforce. Loss of elasticity results in bones snappingin straight l ines that tend to be perpendicular tothe cortical surface. A largely mineralised bone wil lbreak with a straight, rough edge that is close tobeing at right-angles to the cortical surface.

The three criteria of fracture outl ine (shape),fracture angle (to cortical surface) and fracturetexture (smooth or rough) can be used to charac-terise large assemblages of fragments in terms ofthe amount clf deliberate fracturing of fresh bonesversus levels of post-depositional breakage of drybones (Vil la and Mah ieu 199I; Outram 7998; 2007;2002). The indexing method used in the study ofboth the Greenlandic and Icelandic material wasdeveloped and tested experimentally by Outram(1998;2002).

All diaphysis fragments of 30mm or more inlength were studied for fracture type providing thatpreservation was good enough. Occasionally, frag-ments have eroded edges and fracture features areindistinct. Such fragments were disregarded, butwere also very uncommon in al l assemblagesstudied. For each of the three criteria a score of 0, 1or 2 was awarded. In broad terms, a score of 0denotes that that criterion is consistent with frac-ture of a fresh bone, a score of one denotes amixture of fresh and unfresh features (but withfresh still dominating) and 2 denotes that unfreshfeatures are dominant. Much greater detai l isavai lable in Outram (2002). Shape, angle androughness are all estimated by eye. This is essentialto make assessments of large samples practical.Individuai misjudgements wil l be irrelevant as themethod is being employed to character ise theassemblage in general and sample sizes are large.The angle and outl ine characteristics are fairlyeasily defined, but assessment of roughness is moresubjective and relies upon the analyst having a

good mental template (like much zooarchaeologicalanalysis) of the possible range.

When the scores are added together one ends upwith a score from 0 to 6 for each fragment, calledthe Fracture Freshness Index (FFI) score (Outram2007;2002). Scores of 0, 1 and 2 will represent bonesbroken in a relatively fresh state. Scores of 3, 4,5represent either bones that were broken whenbecoming fatrly dry (unlikely to be for fat extraction)or bones which had some fresh fracture on thembut were further fragmented when unfresh. A scoreof 6 represents a bone with no evidence of freshfracture. The profile of scores and overall averagefor a sample can be displayed.

The FFI score is a very good indicator of thetaphonomic history of the assemblage, but otherindicators can be recorded that provide more detailand help to deal with potential problems of equifi-nality within the FFI. If one records whether or nota fragment has an example of an individual mineral-ised break it is possible to distinguish between bonesthat had fresh features, but then got broken whenmineralised, and bones that showed no fresh orcompletely mineralised features, but were dry whenbroken ( i .e. scores 4,5,6). Such completely mineral-ised breaks are easy to spot (on their own theywould score 6). New breaks (caused by excavationor storage) can aiso be recorded, as the fracturesurfaces wi l l be an obviously di f ferent colour.Dynamic impact scars, created at the point of impacton a fresh bone, can also be recorded as evidence ofdeliberate fracture, much like bulbs of percussionon flints. If the bone cylinder was broken on ananvil, there may be a rebound scar due to the op-posing force of the anvil (see Outr arn2002, fig. 6.3).

Other tnphonomic indicntors

Several other criteria are important for under-standing the taphonomic history of the assemblage.The above shaft fragments are also studied forevidence of animal gnawing and butchery (chops,cuts, polish and sawi.g). Numbers of burnt frag-ments are counted for the entire sample at the levelof size class and bone type. Butchery can clearlyadd to the overall level of deliberate breakage,however, the breakage of bones for butchery pur-poses wil l be restricted to particular elements for apart icular purposes, such as an al ternat ive todisarticulating a diff icult joint. There is also l ikelyto be a difference between the fractures producedby chopping through meat and bone and thosecreated by direct impact to the bone. It is essentialthat all indicators (fragmentation level, bone types,fracture patterns, gnawing, butchery and burning)are considered holistically to effect a successfulinterpretation and avoid pitfalls of equifinality.

724 A. K. Outrgm

A note regnrding sea mammnls

Sea mammal bone is very different in nature fromland mammal bone in terms of its physical make-up, nature of fat and fat utility and is generally notexploited for its fat content (see Outram 7999,115).It has been separated out in this study and will notbe included. However, reference to the state ofpreservation of seal bones will be made, since thisis a taphonomic indicator. Seal bones are not verydense, being composed almost entirely of cancellousbone with quite thin cortical shaft bone, and, assuch, their survival would indicate relatively lowlevels of post-depositional attrition. The relevanceof the exploitation of sea mammal fats will alsoform part of later discussion.

Results

Combined Greenland resLtlts (Figs. 1-3)

For the purposes of this paper, the results fromSandnes (V51) and Niaquussat (V48) have beencombined graphically to show the mass of fragments(Fig. 1) and FFI score distributions (Fig. 3). For bonetype (Fig. 2) only Sandnes results have been dis-played. At Niaquussat the results were almostidentical, but ribs were not separated from otheraxial bone, which made that data less informative.If one examines the masses of fragments in differentsize classes (Fig. 1), the most striking feature is thelack of bones that survive whole or as completebone parts (vertebral centra and epiphyses). Thereis in fact an almost equal mass of bone that is com-minuted to a size under 20mm! There is a fairamount of material surviving to over 100mm, butthe assemblage is generally quite broken up.

In order to understand this pattern properly oneneeds to view it in conjunction with a graph showingthe type of bone in each size category (Fig. 2). Thisshows that the bone in the larger size classes consistslargely of shaft and rib fragments. It is clear thatvery little cancellous bone other than that in ribssurvives in large pieces. Most of the cancellous bonecan be found in size classes under 50mm. This isvery important. The expected pattern for bonemarrow exploitation and intensive grease pro-duction is one where most cancellous material willhave been comminuted, leaving shaft splinters todominate the larger classes. This is what we seehere, with the exception of ribs that appear not tohave been broken up. If this pattern is the result ofhuman agency then i t looks l ike a pattern ofextensive bone fat exploitation where only the ribshave been disregarded in the rendering process.

The FFI data (Fig. 3) strongly supports thishypothesis. Fracture patterns consistent with the

Figure 3. Frncture Freshness Index scores for shaft

frngments nt Ssndnes (V5L) and Niaquussat (V4B).

fracture of bone in a fresh state very much dominatethe assemblage. Therc appears to be very little post-depositional damage to the bones, once they hadbegun drying out and loosing organic content. Atboth sites about 9"/. of shaft fragments bore directevidence of dynamic fracture in the form of scars.This is a good proportion, considering that suchscars are only found at the actual point of impact.

1 800

1 600'1400

1200

O roooo3 eoo=

600

400

200

0

Figure 1. Mttssesclnsses nt Sandnes

of bone fragments in different size(V51) nnd I'l iaquussat (V4B).

o o ) o ) o ) o o ) o ) v >$ q 9 Y ' r ) F - o , oA A A O d

N C D $ r . ( ) ( O @

Size Glass (mm)

100%

90%

80%

70%o

60To

50%

40%

30%

20o/o

10%

0%

Figure 2.clnsses at

O) O) O)N C O $

N C D $

O ) O ) O ) Vl r ) f - o ) o

r o @ c o

Size Class (mm)

Types of bone fragmentsSnndnes (V51.).

=tL

in different size

- 1 5 0

Ez 1 0 0

Comparing Leuels of Subsistence Stress nmongst Norse Settlers in Iceland ond Greenlond 725

There was evidence for only small amounts ofdamage from other causes. Carnivore gnawing wasnoted on 3.6"1' of specimens, mineralised fractureon 3.2"/o, new breaks on 7.2% and butchery markson 0.5%. Burning was very low at Sandnes, at lessthan 7"h of fragments in any class. At Niaquussatmore than 30'lo of the very small class was burnt,but levels were much lower in larger classes. It islikely that that assemblage contained a small amountof fire output. The general lack of post-depositionaldamage is also indicated by the good survival ofseal bone and land mammal ribs.

The interpretation therefore is that the Green-landic Norse were exploit ing bone marrow byfracturing long bone shafts, creating shaft splinters,and rendering almost all cancellous tissue exceptribs for bone grease, leaving a pattern of com-minuted appendicular and axial bone and large ribfragments. This would seem to be indicative of asociety who needed to exploit fully the resourcesavailable to them, even though this would requiremuch effort for a relatively small return. For a moredetailed consideration of these sites specifically, seeOutram (7999\.

Hofstadir nnd Sueignkot (Figs. 4-9)The masses of fragments in different size classes atHofstadir (Fig. 4) and Sveigakot (Fig. 5) are verydifferent from the Greenlandic sites. Whilst there isa certain degree of fragmentation the assemblagesare dominated by large whole or part bones andlarge fragments. The bone types in different sizeclasses are even more strikingly at odds with thepicture in Creenland. At both Hofstadir (Fig. 6) andSveigakot (Fig. 7) vast amounts of cancellous bonesurvive in large, unrendered pieces. Shaft bone ispoorly represented, largely because much of itsurvives attached to whole epiphyses or forms partof complete appendicular elements.

If one studies the FFI scores at Hofstadir (Fig.8) and Sveigakot (Fig. 9), one sees that, whilstthere is still significant evidence for the fracturingof some fresh bones, there is also much indicationof the breakage of bones when they were no longerfresh and the bone fats were no longer exploitable.The idea that some shaft breakage was deliberateis supported by the incidence of impact scars, 6.9%at Hofstadir and 9.7"/" at Sveigakot (two exampleshad rebound scars). At Hofstadir and Sveigakotrespectively, gnawing was 0% and 7.2/", mineral-ised breaks were 5.9Y,, and 4.0'%, new breaks were7.0% and 1.2"/" and butchery was 7'% and 7.6"/o.There was a fairly substantial level of burning inthe smal lest s ize c lass at both s i tes, 25.5"h atHofstadir and 25.7"/" at Sveigakot with levels inlarger size classes being generally low.

Figure 4. Masses of bone frngments in different sizeclnsses at Hofstndir.

1 000

1200

1 2000

1 0000

o Ann(g=

400

8000

a 6000(9E

4000

o o o ) o ) Q o ) o v . U c JN N c 9 $ ! O f . - o ) o - -v d a d a a e 9 € €

N ( r ) $ r o @ @ d o _

Size Class (mm)

o o ) o o ) a A A v U ) JI l 3 l e \ ? P = =

N - + 8 3 3 = f fSize Class (mm)

s q q q B E { 2 rA o 6 o A : 6N -

Size Class (mm)

Figure 5.clnsses nt

Masses of bone fragments in different sizeSaeigakot.

100o/o

90%

80%

70%

60%

50%

40%

30Yo

20%

10%i o /v l o

Figure 5. Types of bone fragments in different sizeclnsses nt Hofstadir.

The patterns at both Icelandic sites are verysimilar and show with great certainty that bonegrease was exploited, if at all, at very low levels.Incidences of low FFI scores and dynamic impactscars suggest that some long bones were de-liberately broken for their bone marrow, but theexploitation of this relatively worthwhile resourcewas not exhaustive.

II

126 A. K. Ott t rnnt

100%90%80%70%60%50%40%30%20%10o/o

o%

Figure 7. Types of bone frngments in different sizeclasses at Sueignkot.

0 1 2 3 4

lndex Score

Figure B. Frncture Freshness lndex

frngments nt Hofstndir.scores for shnft

0 1 2 3 4 5 6

lndex Score

Figure 9. Fracfurc Freshness Index sclres for shnft

frngnrcnts nt Sueignkot.

Discussion

Leaels of subsistence stress

It is clear that the Greenlandic Norse were exploit-ing bones for their grease content very intensively.Such intensive exploitation is not a very efficientway to gain food, but one suspects that they had

little option. The only type of cancellous bone thatseems to have been systematically ignored is rib. Itis interesting that Nunamiut informants told Binford(7978,32) that rib grease was not thought of as verydesirable and was only exploited in bad times andfor lighting.It canbe concluded that the exploitationof bone fats was very intense and probably indicatesconsiderable economic stress. Even though theremay have been sufficient stress to make the laborioustask of rendering worthwhile, it seems that, onaverage at least, the settlers could still exercise alitt le taste and avoid the use of ribs.

The situation on Iceland during the early part ofits settlement seems to have been very different. Itis clear that most cancellous bone was not beingrendered for grease. In fact, there is no really goodevidence for rendering at all. Fracture patternsshow that bone marrow was being exploited, butnot exhaustively. It seems that the Icelanders couldafford to let much bone fat go to waste, just usingeasily extracted marrow when they felt like it.

Cnuses of differentinl subsistence stress

It seems likely that there are three principal reasonsfor the apparent differences in bone fat exploitationpatterns and inferred economic stress levels betweenthese two communities. These relate to the overallproductivity of their environment, their seasonalrounds and the production of stockfish.

A simple explanation for the difference is thatthe Icelandic settlers were far less limited in thegrazrr.g land available to them. In the early periodof settlement the soils would not have become toodepleted or eroded and, as such, the pastoraleconomy could have been considerably more suc-cessful than on Greenland, where pasture was verylimited.

Another reason for the difference can perhaps befound in periods of seasonal dearth of resourcesand the need for stored foodstuffs. The Greenlanderssealed in the spring (McGovern 1985, 101). Thisactivity would have brought both meat and muchfat, in the form of blubber. The rendering of bonesat this time would have been entirely unnecessary.Mi lk ing would have been possible dur ing thesummer and it is likely that culling of some domesti-cates and the hunting of the odd caribou wouldhave taken place before the onset of winter. Winterwould see little in the way of new resources. Therewould be stored dairy produce, possibly some driedmeat and a stock of fat-bearing bones from theautumn kills. That is the likely time when intensivegrease exploitation became necessary.

The winter situation in Iceland was different,principally because of the abundance of stockfish.This important resource of stored gadids could also

O ) O ) O ) O ) O ) O ) V A

1 ? Y 9 r . - o ) o _ 'o o o o A : 5 = =N - + . r > 6 t s f o -

Size Class (mm)

120

't 00

E d Uz

OU

20

0

Compnring Lcuels of Subsistence Stress amongst l,Iorse Settlers in lcclnnd snd Greenlnnd 127

be supplemented year roundavailabil ity of freshwater trout2001,52). Without an equivalentstress, the Icelanders need notintensive bone rendering.

by the abundantand char (Byock

period of seasonalhave resorted to

Conclusion

The detailed examination of the nature and level ofbone marrow and grease exploitation, when con-sidered alongside other economic and environ-mental data, is a powerful tool for understandinglevels of subsistence stress in marginal environ-ments. This study has shown up clear and inter-esting differences in the patterns seen on Greenlandand amongst the early Icelandic settlers. It raisessome interesting questions. lt once again raises theconundrum of the lack of fishing on Greenland.Was this a purely cultural, if disastrous, decision oris there a clear practical reason why fishing couldnot be carried out? Such possible reasons couldinclude a lack of time, a lack of wood for suitableboats or even the wrong conditions for air-dryingfish. More work wil l be needed here. Anotherobvious quest ion, and one that can easi ly beaddressed in the future, is whether the nature ofbone fat exploitation changes over time withinIceland. What are the effects of soil degradation,erosion and the onset of the Little Ice Age and theeconomic changes related to it?

Acknowledgements

I particularly need to thank Tom McGovern, ClaytonTinsley, Sophia Perdikaris and Colin Amundsonfor allowing me to study the Icelandic material atthe City University of New York and for the valuableinformation they have provided me with. Thanksalso to Matt Brown for his hospitality during mystay in New York at a particularly difficult time inSeptember 2001. My gratitude to the CopenhagenZo oIo gical Museum, p articularly K. Aaris-Sorensenand J. Mohl, for their assistance while I studied theCreenlandic material. Thank you to Paul Bucklandand Peter Rowley-Conwy for useful discussions andmy parents, Derek and Helen Outram, for commentson earlier drafts. Anonymous referees made helpfulsuggestions. I am grateful to the British Academyfor funding my work.

References

Amorosi , T. 1992. Cl imart ic impact and human response inNortheast Iceland: archaeokrgical investigations at Sr.'albard,

1986 - 1988, pp. 103-38 in Morr is , C. D. and Rackham, D. J.(eds.), Norsc nnd Lnter Sattlcnrent nntl Sultsistt ncc irr t ln' l ' , lortlrAt lnt t t ic . Glasgt ' rw: Univers i ty of Glasgow, Department ofArchaeology.

Binford, L. l t . 1978. Nutrn. l . l iu t Etht t t t t rc l t t tarr lot r7. Ncw York:Academic Press.

Br ink , J .W.1997 . Fa t con ten t i n l eg bones o f B iso l b i son , andappl icat ions to archtreology. Iourrml of Archnaol t tg icul Scictrce24,259-74.

Brookes, I ' . M., Hanks, J. and Ludbrook, J.V.1977. Bone marrowas an index of condi t ion in Afr ican ungulates. Solr f / r Af i " ic t r t r

Jourrml of Wildhfe Re sanrch 7, 67-6.Buckland, P. C. , Amorosi , T. , Bar low, L. K. , Dugmore, A. J. ,

Mayewski , P. A. , McGovern, T. H. , Ogi lv ie, A. E. I . , Sadler ,

J. P. and Skidmore, P.1996. Bioarchaeological arncl c l imato-logical evidence for the fate of Norse farmers in MedicvalCreenland. Ant iqui t ry 70, 88-96.

Byock, J. L. 2001 . Vik i r tg Agc lc t , lnt td. London: I )enguin.Cheatum, E. L. 1949. Bone marrow as an index of malnutr i t ion

in deer. Ntru York Stnta Ctt r tscrant i t tu 3,79-22.Davis, J. L. , Valkenburg, P. and Reed, S. J. 1987. Correlar t ion

and deplet ion pat terns of marrow fat in carr ibou bones.

Jourrutl of Wi\dltfa MnnLtgunartt 51,365-71.Degerbol, M. 1936. Animal remains from the Western Settlement

in Creenland. Mct ldclc lscr ( ) t t t Cronlnnd 88(3).Erasmus, U. 1986. Fnts nnd Oi ls : t l rc Ct tnryt l r tL, Cuidc t t t Ft t ts und

Oi ls i r r Hcal th and Nutr i t iorz. Vancouver: Al ive Books.

fohnson, E. 1985. Current developments in bone technokrgy,pp.757-235 in Schiffer, M. B. (ed.), Atlrtotrct,s in ArclmeologicttlMct l rod and Theon18. New York: Academic Press.

Leechman, D. 1951. Bone grease. Anrcr icnn Ant iqui t t116,355_6.Leechman, D. 7954. The Vanta Kutchin. Nnt iorrn l Muscum t t f

Cnnndn Bul let in 130, 1-77.McCovern, T. H. 1985. Contr ibut ions tc l the palaecteconomy of

Nrrrse Creenland. Actn Archncologia 54, 73-722.McGovern, T. H. , Amorosi , T. , Perdikar is, S. arnd Wool let t , J .

1996. Vertebrate zooarchaeology of Sandnes V51: economicchange at a chief ta in 's farm in West Creenland . Arct icA nt I r r op Lt I o gt1 33, 9 4-727.

McGovern, T. H., Buckland, P. C., Savory, D., Sveinbjarnardottir:,C. , Andreason, C. and Skidmore, P. 1983. A study of thefaunal and f loral remains f rom two Norse farms in theWestern Settlement. Arctic AntlrroTtologrl 20, 93-720.

Mead, J. F. , Al f in-Slater , I l . B. , Howton, D. R. and Popjak, C.7986. Lipids: Clrcmistry, Biocltamistry Ltnd Nutrit it lr. New York:Plenum Press.

Mor lan, R. E. 1984. Toward the def in i t ion of cr i ter ia for therecogni t ion of ar t i f ic ia l bone al terat ions. Qlnf crnnrt l Rasnrcl r22,760-77.

Outram, A. K. 1998. Tltc ldentificLttion ttrtt l Palnaoecortorrric Cttntextof Prclistoric Bttne Mnrrout urtd Grensc Ex1tltt itttt iort UnpublishedPhD thesis, University of Durham.

Out ram, A . K . 1999 . A compar i son o f Pa leo -Esk imo anc iMedieval Norse bone fat explo i tat ion in Western Greenland.Arctic Anthropologtl 36, 103-77.

Ou t ram, A . K . 2001 . A new approach to i den t i f y ing bonemarrow and grease exploi tat ion: why the " indeterminate"fragments should not be ignored. Journttl ol ArclncologicnlSciencc 28, 401-10.

Outram, A. K. 2002. Bone f racture and wi th in-bone nutr ients:an exper imental ly based method for invest igat ing levelsof marrow extract ion, pp. 51-64 in Miracle, P. and Mi lner,N. (eds.) , Cctnsunr i r tg Pnssi t l l ls ntr t l Pnt tcrrrs oJ Ctt r rsutnpt i ( )n.Cambridge: McDonald lnst i tute for Archaeological l te-search.

Perdikar is, S. 1999. From chief ly provis ioning to commerciar lfishery: long-term economic change in Arctic Norway. WorltlArch aaol o gt1 30, 3BB-402.

-.-

r28 A. K. Outrnm

Peterson, l l . O. , Al len, D. L. and Dietz, J. M. 1982. Deplet ion ofbone marrow fat in moose and a correlation for dehydration.

Journal of Wi ld l i fe Manngemert t 46,547-57.Speth, J. D. 1983. Bison Ki l ls and Bone Counts: Decis ion Makingby

Ancier t t Htnt ters. Chicago: Univers i ty of Chicago Press.Speth, I . D. 7987. Ear ly hominid subsistence strategies in

seascrnal habitats. lournal of Arclmeologicnl Science 14,73-29.Speth, J. D. 1997. Nutr i t ional constra ints and late g lacia l

adapt ive t ransformat ions: the importance of non-proteinenergy sources, pp.769-78 in Barton, N. , Roberts, A. | . andI loe, D. A. (eds.) , T l rc Lntc Glncin l in Nort l t -West Europe:I l t tmnn Adnptat ior t nrd Enuironmentnl Clmnge at t l rc End of thePleistocene. York: Council for British Archaeology.

Speth, J. D. and Spielmann, K. A. 1983. Energy source, protein

metabol ism, and hunter-gatherer subsistence strategres.

Journal of Anthropological Research 2,7-37.Vi l la, P. and Mahieu ,E.7997. Breakage pat terns of human long

bones. lournal of Humnn Eaolut ion 2 '1. ,27-48.

Wilson, G. L. 7924. The Horse nnd tlrc Dog irr Hitlntsn Culturc(An th ropo log ica l Papers o f the Amer i can Museum o fNatural History 75,2) . New York: American Museum ofNatural History.

Zul ter , C. M. 7992. Icelandic p lant and land-use pat terns:archaeobotanical analysis of the Svalbarr) midden (6706-60),

northeast Iceland, pp. 139-48 in Morris, C. D. and Rackham,D. j. (eds.), Norse ctnd Later Settlenrcnt ud Subsistcncc in tlrcNortlr Atlanfic. Clasgow: University of Clasgow, Department

of Archaeology.

The Greek Mesolithic: Problems and Perspectivesedited by Nena Galanidou and Catherine Perlds

"This book is the first major review of the Mesolithic agc of Greece,"writes Colin Renfrew. "If the (Franchthi Cave publications) represent afirst milestone in mesolithic studies in the Aegean, this volume certainlyconstitutes a second. It establishes a perspective going beyond that ofthe single key site, allowing earlier finds such as those fiom Sidari onCorfu to be re-assessed, and including more rcccrlt discoveries at theTheopetra Cave, at the Cave of Cyclope on Youra, and fiom the I(isouraGorge. The book moreover develops a number of methodologicaladvances, along with a careful review of the chronology. It embarks alsoupon detailed investigations of the fauna and microfauna." 224 pngeq,68 b/w i.llus (British School nt Arhens Studies 10, 2003) hb 090488743X

Zooarchaeology in Greece: Recent Advancesedited by Eleni Ifutjabopolrlou) Yannis Hamilakis,, Paul Halstead,Clive Gamble and Paraskevi E,lefantiAnimals have always been integral to lifb and culture in Greece . Thisvolume considers recent advances in the study of zooarchaeology inGreece, from the Palaeolithic to the Medieval periods. Thematicsections fbcus on the environment and subsistence) consumption,ritual and other uses of animals, as well as ethnographic, textual andartistic or representational evience for animals. The interdisciplinaryand international character of zooarchaeology is confirmed and manynew avenues fbr research are suggested. 310 pnges, 111 b/w illus(British School nt Athens Studies 9, 2003) bb 0904887413