Chapter
    
    11
    
    Ore, Fire, Hammer, Sickle: Iron Production in Viking Age and Early Medieval Iceland*
    Kevin P. Smith
    
    Iron production may be used as a window through which to view, in part, the economic structure of Icelandic society during the Viking Age (c. AD 870-1000) and Early Medieval (AD 1000-1264) periods. kon was a critical resource for
    maintaining and reproducing medieval Icelandic society, yet while several medieval sagas and related sources' mention iron smelters and smiths, documenting their presence within the society, they provide insufficient information to reconstruct the iron industry's technological basis, organization or role in the larger economy.2 Recent archaeological research, combined with information gleaned from medievai texts, provides opportunities for addressing these issues.
    
    * Research on i¡on production at the site of Háls was undertaken in 1996 and 2000 with generous grants from the National Geographic Society and Norðurál (Nordic
    Aluminum/Columbia Ventures), in collaboration with the National Museum of Iceland and with the permission of the Kollslækur farm landowners. Additional support was provided by Eimskip, Sparisjóöur Mlraslslu, Ungmennafélag Reykdæla, Kvenfélag Reykdæla, Ungmennafélagið Brúin, the communities of Reykholtsdalu¡ and Hvítrárslsla, the Buffalo Museum of Science and the Haffenreffer Museum of Anthropology, Brown University. The author wishes to thank Peter Bush, Eiizabeth Hamilton, Kristín Sigurðardóttir, Alexande¡ T. D. Dixon and aI1 crew members who participated in the excavations for tt¡eir expenise and contributions; he also thanks Michèle Hayeur Smith fo¡ her invaluable insights and loving support. Finally, advice from Steven A. Walton, Shepard Krech III, Rachel Scott and four anonymous reviewers has greatly strengthened this presentation; any remaining mistakes or misinterpretations of the data are the sole responsibility of the autho¡. 1 The term saga,as used here, follows general usage in lcelandic studies and refers to literary-historical works produced indigenously, in the vernacular, from the mid-twelfth century onwa¡d that incorporate historical themes, events, and personages presented by their authors as real, rather than mythic or heroic, figures in Icelandic, Scandinavian and related early medieval societies. See Jónas Kristjánsson, Eddas and Sagas - Icelanì's Medieval Literature (Reykjavík, 1997),for a more detailed discussion. In addition to the sagas, limited information on the roles and products of smelters and smiths can be gleaned f¡om Iceland's medieval law code, Grdgás (see notes 8 and 9, below). 2 See Mark Hall, "Viking Age Ironworking: The Evidence from O1d Norse Literature," ín The Written and the Wrought: Complementary Sources in Historical Anthropology, KroeberAnthropological Society Papers 79 (1995), pp. 195-203.
    
    r84
    
    KEVIN
    
    P.
    
    SMITH
    
    IRON PRODUCTION IN VIKING AGE AND EARLY MEDIEVAL ICELAND 185
    
    sources of ore suitable for smelting into useable iron. The presence of a smehing facility at the Norse t:xploration base of L Anse aux Meadows, Newfoundland,
    
    The expansion of the Norse across the North Atlantic and their establishment of sustainable colonies were predicated, in part, on their ability to find and exploit
    
    Independent Households and Economic Networks in a New Society
    The mixed ethnic population that colonized Iceland during the late ninth and tenth century was drawn primarily from mainland Scandinavia and the British Isles.
    
    forged into billets and bars, and werr: then transformed into the essential equipment required to sustain the colony's expansion. This equipment included agncultural instruments such as the scythes, sickles and pack saddles required for the summer hay harvests that were the foun,"*" *ort trtety uetiäws shields from Viking Age forges-' !9191oeous
    
    ofa
    
    small,
    
    and changing Landuse ui FIet, Sourhwest
    
    .'torge stone" from Snaptun, Denmark (P. V. Glob, "Avlsten: Nye Typer tuåo.lr, ("essesteine") "ng.uuea lemaJder:, ümt,isse,p.ol,rig.2),theceramicforgesrones fraDanmarks Ribe (He1gé B¡inch Madsen, "Die Eisenschlacken von äoÀ smitrries in eariy Viking Age ceramic examples from Early lron Ribe,,, Metalta, no"n"*, lõõS,ip. S-f f ), arid simila¡ Hået, Germany, and Viking Age forging deposits at Àgï'tÀiinitg facilitiei at Mountains and the Sea: The HaithabuÆIedeUy,lenmãrt (see Frank Nikukä, "Between the in Northwestem Germanyi' in Prehis-toric an'd' Medieval Development of Iron Production Nørbach, Aa¡hus' 2003' Direct lron Smelting in scandínøviaanã. Europe,ed. Lars christian that flai, perforated forge stones, both ceramic and pp. 63_6). These examples suggesr Nordic Viking Age smithies, but their stone, may have taken ttie pfu.àä""tu*ic tuyéies in knowlédg9, unprovel' fo .1uclr "forge stones" or use in smelúng to*u."t'i., to my lcelandic forges, such as those fragments therJof have been'recováed from Háls. Ittlut"t (and preserved-as open-air museums in maintained in nineteenrhãntury turf-houses employedl the blast from the northern Iceland), ." ;;h forgÉ stones or shields were directly through long iron tubes, discharged through bellows reached t¡" ,rnitfri"t'trJartfts the stone-built walls forming the backs of the forges themselves'
    
    192
    
    KEVIN
    
    P.
    
    SMITH
    
    IRON PRODUCTION IN VIKING AGE AND EARLY MEDIEVAL ICELAND 193
    
    glacial clay.r? Given the smail diameters of the furnace bases exposed in 2000, it is unlikery that the furnace stack was higher than 70 cm. The wooden-framed construction surrounding the furnace stack may also have, therefore, served as a pratform for loading, charging, ancl emptying the furnaces after each smelting episode. Figure 11.5 provides a tentative reconstruction ofthis furnace,s appearance during use. As in other parts of the site, the superimposed series of fo*u"" bases exposed in 2000 preserves si-lggestions; of episodic use and change. For example, the ea¡liest furnace base identified at this location appea$ to have been almost twice as large as those that were buirt over it. stratigraphic cross-sections adjacent to
    26 see Irmerin Martens, "Bråsterjern og Ferlujern: Noen synspunkter på en Lite påaktet Funngruppe." in (Jniversiteteß ordsaksa;ti"g IiT år, Jubileumsårbok 1979 (osro, 1979), pp. i90-97; idem, "The Norwegian Bioomery"Furnales and rhei¡ Relation to rhe European Finds," offa 40 (1983), Ir92a; and idem,.iron in soutrr"urt"rn No.rìy in tir" pr"¿i"u¿ period.".in Bloomery lronmaking. "ol. II. pp. 55_6g. r/ Although the construction of the fúrnace at Háls differs in significant details from thei¡ tentative reconstruction, the possibility t¡ui jàau""o tu.r-¡uilt
    fce.land
    
    debris are generaily consistent with untapped, row-shaft broomery furnaces known from early viking Age Norway. rirÅe have been well documented by Irmelin Martens and others, and are thought to have produced blooms of raw iron 8-10 kg in mass du'ing good smerts.'u what clearly oiffers from the Norwegian furnaces is the absence of a thick ceramic furnace stack - a point that may be directly related to a paucity of substantial deposits of good clay in lceland,s young geologicar landscape. Smat deposits oi mixed glacialry áeposited ctay and silt are found along the banks of some glacial rivers, but these have never proven suitable for producing ceramic vessels and rnay have been unsatisfactory for building either serf-supporting or slab-framed cray furnace stacks. Instead, the evidence from lláls suggests that these furnaces had stacks buitt of turf blocks, lined internally with pieces of basart cemented into a thin veneer of silty
    
    that surrounded and supported the furnace's core and would have prevented air from entering the shaft through its turf wail. A weil-defined, verticiedge between this silt deposit and the surroundi'g trampred surface of the iron-pro¿uãtion work area suggests that the silt and turf body of the furnace had been encrosed in a rectangular crib-like strucfure, presumabry of wood, with a stone-flanked opening to the east through which beilows could have been inserted and through which slag may have been removed. overall, the size, shape, and construction of the fumaces and their associated
    
    around the
    
    surrounding the furnace bases was a matrix of dense silt-based turf. In crosssection it courd been seen that this matrix, surrounding the uppermost and ratest furnace base, included turf brocks inco¡porating the Landnám iepr'a stacted at an angle against the inner core of the furnace to form the furnace stack itself. Heaped
    
    this base suggest that it may have been rebuilt at least three times before being abandoned. The smaller-diameter furnace bases described above appeal to sit offcentel to this earlier construction, yet within the same general area. suggesting re-
    
    turf brocks was an extensive deposit of roose, sterile silt
    
    facility after a period of disuse. Although none of the furnace bases have yet been individually dated, or fully excavated, the observed differences in the sizes and orientations of the furnace bases raise questions about
    use of the frame and
    
    whether more than one type of furnace was employed at the site during the course of its use. No slag-tapping basins were found in the vicinity of the furnace bases exposed 2000 and virfually no tap slag has been recovered from the site - 97 per cent of in the estimated 5000 kg of slag present in the main slag heap appears to be amorphous furnace slag typical of non-tapped shaft furnaces." The main slag heap itself covers an area of at least 45 mz, blanketing a shallow slope between the production area and the margins of the adjacent bog. Between the furnace and this slag heap was a zone in which charcoal was abundant but large concentrations of smelting slag were relatively infrequent. At the eastern edge of this zone lay a
    
    battered boulder, beyond which plano-convex smithing hearth bottoms were clustered. Miniscule spheroids of slag and hammer scale, typical products of secondary smithing during bloom consolidation, were common in this area.'e Several broken or fragmentary pieces of semi-fabricated iron tools were also recovered from the same general area. The size of the spheroids, the number of smithing hea¡th bottoms and the relative concentration of semi-processed iron objects strongly suggest that raw blooms (btásturiárn) were processed here into compressed blooms (fettujárn),and perhaps further into semi-fabricated forms. During the 1996 season a broken iron shaft with a fragmentary, flattened expanding end was recovered from the fîll of the earlier of the two pit-houses identified at the site, and in 2000 two flattened spatulate fragments of forged iron with rounded ends were recoveted from the site (Fig. 11.6). One came from the
    
    production area near the boulder; the other was Íecovered from the fill of the same pit-hoor" that produced the expanded shaft in 1996. In size and shape, these three óU.¡""tt match remarkably well the spatulate ends and shafts of Norwegian Viking Ale ..cunency bars" - forged iron bars of consistent size and mass generally thought to have been produced as standardized units of exchange, in which the
    28 Dr Elizabeth Hamilton, notes (files, Haffenreffer Museum of Anthropology, Brown University); see also note 31, below. However, fragments of tap s1ag, including
    
    both runnels and fractured tap-basin slags, were recovered in 1989 from another eroding slag concentration located td the southeast of the slag heap excav¿ted in the 1996 and 20õ0 seasons at Háls. This suggests the possibility that additional furnaces, with slagtapping capabilities, were usedãi the site. If so, those furnaces have yet to be located or
    exposed. 29 On the identification of archaeologically ¡ecoverable by-products from forging and smithing, see J. G. McDonnell, "A Modei for the Formation of smithing Slags," Materialy Archeoioigiczne 26 (199r),pp.23-7; and Peter Crew, "Bloom Refining and Smithing Slags (1996)' and other Residuesl' Archnàology Datasheet No.6,Historical Metallurgy Society
    
    was_previously suggested
    
    in lceland" (see nores 43.44,and.a7. below).
    
    bi Fridriks;on
    
    unà H".rnunns_Arõ
    
    f"d;; ;;
    
    in ãIìonmating
    
    "rd¿tti;
    
    194
    
    KEVIN
    
    P.
    
    SMITI{
    
    IRON PRODUCTON IN VIKING AGE AND EARLY MEDIEVAL ICELAND 195 at Háls was later forged into objects for household use at such coastal sites' The
    
    ironworkers gathered at H¿írs brought with them a diversified range of skils, some quite specialized und, others less so, which enabled them to undertake nearly the entire iron-production process from ore and fuel gathering to the production of standardized iron cbjects for exchange (Fig. 11.7). At the same tirne, trre rack of evidence for comparably intensive smithing on-site suggests that most of the finar fabrication of tools made from the iron produced at Háls was done elsewhere. As noted above. there is no evidence for an inhabited farm-site at Háls contemporary with this iron-production phase, rendering it impossible to suggest where the next stages of fabrication were undertaken. However, the majority ãf welr-dated early viking Age habitation sites in Icerand are located in the coastar and near_coastar ìowlands.32 some, but not a[, of these sites have smithies, yet few have produced evidence of on-site smelting facilities.3'Thus it is possible that the iron produced
    
    and limited non-ferro's metarworking. clipped ìails impry the repair of objects brought to the site. A ¡iveted bucket puli, ,ogg"rts some work fabricating or repairing compound iron otrjects. Two small iron carving knives, discarded near the pit-houses, hir:t at craftwork or domestic activities undertaken in support of iron production. Srnal pieces of copper alroy scrap and possible flecks of silver in one block of slag examined by scanning erect¡on microscopy also suggest some level of on-site wo¡k in non-ferrous *"àlr, concurrent with iron production. Based on the evidence at hand, it is difficurt not to conclude that the
    
    undertaken at the site. incrucling generarized repair of iron and composiíe tools
    
    final shape demonsrrared both the iron's qualities and perhaps also its regionar origin''. The presence of these fragments at Hárs suglest, ìr,ut tr,"f represent waste from bars broken during forging, perhaps because-ofpoor iron [uahty.r, A small numb,':r of artifãcts found at Háli arso testify io ancilrary activities
    
    recovery
    
    of seal and cod remains in the pit-house deposits at Háls suggests contact, at least, with the coastal zone, and may provide tentative support for a model of coastal-interior economic interaction during the first century of Norse
    settlement in Iceland.3o
    
    Scale of Production Based on data collected at Háls in 1996 and 2000, it is possible to estimate roughly the scale of production at Háls, and from that basis to consider how it fit into the regional economic System and whether it was unique or representative of a larger
    
    class of sites. Controlled volumetric sampling within the main slag heap and surrounding production features indicates that they contain at least 5000 kg of smelting slag.3' Experimental work by Peter crew and Arne Espelund, among others, along with published laboratory work and limited ethnohistoric data from
    
    ed'' sites at Reykjavft, see No¡dahl, pp. 55-9 and 90-99, and Howell M' Roberts' ó Lfiunum/Arcñ.aeological Excavations at Æalstræti 14-18' 2001: A ior.ntelfuranisókn preümínary Report/Framvind.uskyrslur @õmleifastofnun Íslands, Reykjavft,2001), p.71' For compiabie evidence at Herjólfsdalur, vestmannaeyjar, see Mmqryt HermannsAuða¡dóitir, Islands Tidiga Bosciitning, Studier med Utgångspunkt i MerovingertídaVikingatida Gårdskimnin{ar i HerjóIfsdatur, Vestmannaeyjar, IsLand, Studia Archaeologica Universitatis Umensis I (Umeå, 1989)'pp' 118-20'
    
    "currency Bars wirh
    
    9l currency bars, se^elerer crew, - '1. Hßtoric Metar!,urgv.25 (r'r91), pp. "The Experimental production kon," peter
    .welded
    
    in Economic Asplctl o! vking Age,Bnrish Museum o."^i"""i paper 30, ed' David M. wilson and Marjorie-L. cayginllå'náon, 1981), pp. +l-z,li.-rl; and Hans Georg Resi, "The Norwesian Iron Bar Deþãsits: Have They uoiito reu aLout pro¿uction, Disrriburion. or Consumritio n?,,, uarin (öst"jãô^iiéssl" pp. 13146.
    
    in irudaren,ed. Ame Esperund 
    
    

        

          
 
          
            Add Comment
          
          
    
          

              
                
     
                 
    
                
                    
                                                                              
                    
                    
              
                    
          
          
          
            To add a comment, you need to 
                login or 
    
                signup (you can signup without leaving this page)