Journal of Archaeological Science 40 (2013) 1845e1865

Contents lists available at SciVerse ScienceDirect

Journal of Archaeological Science

journal homepage: http://www.elsevier.com/locate/jas

Integrated geoarchaeological methods for the determination of site activity areas:

a study of a Viking Age house in Reykjavik, Iceland
Karen B. Milek a, *, Howell M. Roberts b
a
Department of Archaeology, School of Geosciences, University of Aberdeen, St. Marys, Elphinstone Road, Aberdeen AB24 3UF, United Kingdom
b
Institute of Archaeology, Iceland, Brugata 3, 101 Reykjavk, Iceland

a r t i c l e i n f o a b s t r a c t

Article history: For over a decade, geoarchaeological methods such as multi-element analysis and soil micromorphology
Received 13 August 2012 have been used to identify and interpret activity areas on archaeological sites. However, these tech-
Received in revised form niques, along with others such as magnetic susceptibility, loss on ignition, and microrefuse, artefact and
9 October 2012
bone distribution analyses are rarely integrated in the study of a single site, even though they provide
Accepted 26 October 2012
very different and potentially complementary data. This paper presents a comparative study of a wide
range of geoarchaeological methods that were applied to the oors sediments of a Viking Age house at
Keywords:
the site of Aalstrti 16, in central Reykjavk, Iceland, along with more traditional artefact and bone
Activity areas
Soil micromorphology
distribution analyses, and a spatial study of oor layer boundaries and features in the building. In this
Loss on ignition study, the spatial distributions of artefacts and bones could only be understood in the light of the pH
Electrical conductivity distributions, and on their own they provided limited insight into the use of space in the building. Each of
Magnetic susceptibility the sediment analyses provided unique and valuable information about possible activity areas, with soil
ICPeAES micromorphology proving to have the greatest interpretive power on its own. However, the interpre-
Viking Age houses tation potential of the geochemical methods was dramatically enhanced if they were integrated into
a multi-method dataset.
2012 Elsevier Ltd. All rights reserved.

1. Introduction
The understanding of the spatial organisation of activity areas is
of prime importance to the archaeological interpretation of
settlement sites. It provides information about how individuals,
households and communities organised the wide range of social
and economic practices that constituted daily life, how they
perceived and managed different types of waste products, and
what living conditions were like in and around their dwellings and
work places. To identify activity areas archaeologists not only use
features such as hearths, cooking pits, storage pits and middens,
and the spatial distributions of artefacts and bones, but increasingly
they are making use of the most minute residues of human and
animal activities: microrefuse (bones and artefacts under 1e2 mm
in size), plant phytoliths, organic residues and associated elements
and isotopes that accumulated on presumed occupation surfaces
(e.g. Sampietro and Vattuone, 2005; Shahack-Gross et al., 2008;
* Corresponding author. Tel.: 44 1224 273693 (ofce), 44 7951 612151
(mobile); fax: 44 1224 272331.
E-mail addresses: k.milek@abdn.ac.uk (K.B. Milek), howell@instarch.is
(H.M. Roberts).
Smith et al., 2001; Sullivan and Kealhofer, 2004; Terry et al., 2004;
Vizcano and Caabate, 1999; Vyncke et al., 2011). Samples for these
micro-residue studies are normally in the form of loose bulk
samples in which the occupation deposits are homogenised, even
though it has long been recognised that occupation surfaces are
usually palimpsests, comprising the residues of multiple, super-
imposed events (Malinsky-Buller et al., 2011).
The interpretation of artefact, microrefuse, and geochemical
distributions on archaeological sites is dependent on a clear
understanding of the complex depositional and post-depositional
processes that created and subsequently impacted the occupation
deposits under study (Carr, 1984; LaMotta and Schiffer, 1999;
Wandsnider, 1996). Human actions frequently result in the depo-
sition and/or removal of particular artefacts and residues, especially
objects over 1e2 cm in size, which are commonly kicked aside,
removed during cleaning, or dumped or cached during site aban-
donment (Lange and Rydberg, 1972; Stevenson, 1982; Tani, 1995;
Wilk and Schiffer, 1979). There is also a wide range of natural
processes that alter the composition of occupation deposits over
time as they become subject to the same physical, chemical, and
biological processes affecting local landforms and soils (e.g.
Johnson and Hansen, 1974; Rolfsen, 1980; Schiffer, 1996; Stein,
1983). It is therefore essential to develop a framework for

0305-4403/$ e see front matter 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.jas.2012.10.031
1846 K.B. Milek, H.M. Roberts / Journal of Archaeological Science 40 (2013) 1845e1865

interpreting activity areas that incorporates an assessment of
cultural and natural processes that may have affected the formation
of occupation surfaces. The ability of soil micromorphology to
resolve minute lenses representing super-imposed events and to
identify post-depositional processes has been well attested (e.g.
Macphail and Crowther, 2007; Matthews et al., 1997; Milek, 2012;
Milek and French, 2007; Shahack-Gross et al., 2005; Shillito et al.,
2011), but the method continues to be underused in comparison
to geochemical methods.
In order to assess the relative contributions that artefact and
bone distributions and different geoarchaeological analyses can
make to the interpretation of site activity areas, an interdisciplinary
study was conducted on a house dated to the late 9th and 10th
century AD, which was excavated in central Reykjavik at Aalstrti
16 (formerly 14e18) (Fig. 1). The house was well preserved, and its
turf walls, internal features (hearth, post holes), and 25 distinct
oor layers located in different parts of the house were readily
identied in the eld (Figs. 2 and 3; Roberts et al., 2003; Snsdttir,
2004). The distributions of artefacts and bone fragments, organic
matter and carbonates (loss on ignition), pH, soluble salt content
(electrical conductivity), magnetic susceptibility, and multiple
elements (ICPeAES), were compared to each other and to the
results of soil micromorphology, in order to evaluate the relative
contribution that each technique made individually, and as part of
an integrated dataset, to the interpretation of the use of space in the
Viking Age house.
2. Study area
Aalstrti 16 is situated 1.95e2.15 m above sea level, at the base
of a moderately steep slope that rises to the west. The climate in
Reykjavik is cool and wet, with an annual mean temperature of 5  C
(0.4  C in January, 11.2  C in July) and an average of 805 mm of
rainfall per year (rarinsson, 1987, 8). The site was well drained,

Fig. 1. The location of the Viking Age house (a) in Iceland, (b) in Reykjavk, and (c) on Aalstrti (drawing by skar G. Sveinbjarnarson).
 K.B. Milek, H.M. Roberts / Journal of Archaeological Science 40 (2013) 1845e1865 1847

Fig. 2. The Viking Age house at Aalstrti 16 (formerly 14e18) under excavation, showing (a) the oor layers exposed in 2001 (facing north) and (b) the porch structure discovered
when the excavation area was extended in 2003 (facing southeast). The photo scales are 2 m long, with white and red increments of 0.5 m. (For colour images, the reader is referred
to the web version of this article.)

however, because it was located on an ancient pebble beach on
which a thin andosol (2e4 cm) had developed, derived from
aeolian silt and ne sand of volcanic origin. Both the in situ andosol
below the house and its turf walls contained the so-called landnm
tephra layer, which had erupted from the Veiivotn system in
either AD 871  2 (GRIP ice core; Grnvold et al., 1995) or AD
877  4 (GISP2 ice core; Zielinski et al., 1997), providing a terminus
post quem for the site. AMS dates on seven charred barely grains
(Hordeum Sativum) from the hearth lls and adjacent oor deposits
indicate that the site could have been rst occupied no later than
AD 890 (Sveinbjrnsdttir et al., 2004), and the datable artefacts
from the oor of the house, including a polychrome glass bead
(Callmer Type B6100) and a glass vessel fragment with grape
decoration, support a late 9the10th century date (Hreiarsdttir,
2005; Mehler, 2002). The house used in this study had suffered
only minor damage of its walls from the foundation trenches of
19th-century factory buildings, but an annexe that abutted the
southern end of the building was severely truncated by later
building activity and only fragments of its walls and its central
hearth had survived (Nordahl, 1988).
3. Methods and materials
3.1. Excavation and sampling
Each of the 25 occupation deposits found below the post-
abandonment turf roof/wall collapse and above the underlying
in situ soils was recorded separately and sampled on a 1 m grid
(Table 1; see Fig. 3 for the locations of the different occupation
deposits). Soil samples were also taken from the andosol that had
accumulated outside the building, but since the sampled oor
layers were anthropogenic sediments, with clear, sharp lower
boundaries with the underlying soils, the natural andosol that
had accumulated around the house could not be used to provide
background levels of chemical or magnetic properties, and the
oors were sampled in a way that would permit detailed inter-
comparisons. Small bulk samples (c. 200 ml) for geochemical
and magnetic analyses were taken from each grid square, while
the remainder of the sediment in each square was taken for
oatation and wet sieving with 1 mm mesh. All bone material
over 1 mm in size was counted and identied if possible (Tinsley
1848 K.B. Milek, H.M. Roberts / Journal of Archaeological Science 40 (2013) 1845e1865

Fig. 3. Plan of the house at Aalstrti, showing the internal features of the house, the occupation deposits used for geochemical analysis (each numbered and represented in
a different colour) and the locations of micromorphology samples (drawing by Howell M. Roberts, Karen B. Milek and skar G. Sveinbjarnarson). (For interpretation of the
references to colour in this gure legend, the reader is referred to the web version of this article.)

and McGovern, 2002; see Supplementary data Table 1). Artefacts
found during excavation were 3D recorded, and those found
during wet sieving were given the coordinate of the centre of the
grid square. The house was excavated in a checkerboard pattern
using sextants, and eight undisturbed block samples for micro-
morphological analysis were taken from the exposed sections
using 12  6  5 cm aluminium tins (ff. Courty et al., 1989)
(Fig. 3).
3.2. Sediment analyses
Bulk samples were air dried, gently powdered with a mortar and
pestle, and sieved to remove inclusions larger than 2 mm. Loss on
ignition was conducted at 550  C and 950  C following Nelson and
Sommers (1996). Magnetic susceptibility was tested in 10 ml plastic
pots using a Bartington MS2 magnetic susceptibility meter with
a low frequency sensor. Electrical conductivity and pH were tested
 K.B. Milek, H.M. Roberts / Journal of Archaeological Science 40 (2013) 1845e1865 1849

Table 1
Descriptions of the occupation deposits used in this study with associated bulk samples for geochemical analyses and block samples for micromorphological analysis.

Context Location Field description Artefacts Bulk samples Block samples

747 Above oors Light red-brown turf containing patches and lenses of e e 67, 79, 80
landnm tephra (turf collapse)
763 SW entrance Medium brown clayey silt mottled with LNL tephra with e e 94
occasional charcoal and small stones
792 Central hearth Soft medium to dark brown silt (ash) containing frequent Iron nail 54 74, 75
small lenses of black ash and charcoal
793 Central hearth Soft black ash containing frequent lenses of orangeebrown Spindle whorl fragment 60.1-4, 82.1-3 74, 75
silt and re-cracked stones
795 Central hearth Friable black, brown, and brownish-white ash containing e 62.1-8, 84.1-8 74, 75
charcoal, small fragments of burnt bone and occasional
pebbles
802 Central hearth Soft mixed brown, white and grey ash containing pieces of e 66-3-7 86.1-8 74, 75
charcoal, occasional pebbles and stones
824 SW entrance Friable brown silt containing patches of purple and red, e 78, 83 94
ecks of wood ash and charcoal
831 Central hearth Dark greyebrown and black ash with small lenses of e 87 74
pinkish-white ash
844 NW sextant Firm, friable dark greyish-black clayey silt with greasy 2 clench nails with wood 88.1-12 80
organic content, containing frequent small rounded stones attached; 2 loomweights,
and charcoal one with drilled perforation
849 SW sextant Soft black silt with charcoal ecks Clear glass vessel frag with 90.1-6 e
grape dec; three loomweights
851 SW sextant Soft black silt e 91 79
852 SW sextant Soft medium brown silt containing pebbles e 95.1-6 79
858 NW sextant Friable orange and black silt Iron nail 102.1-6 71
859 SW sextant Compact medium brown clayey silt containing occasional e 101.1-3 e
charcoal ecks, frequent small stones, and patches of ash
861 SW sextant Firm mottled medium brown and black clayey silt with e 131.1-5 e
occasional charcoal ecks and small pebbles
862 NW sextant Friable mixed light brown and grey silt containing some e 132.1-3 e
charcoal fragments
864 ME and MW Friable black and orangeebrown silt and ash around the Blue glass bead with wavy 110.1-17 71
sextants central hearth, containing charcoal, sand, small stones, purple lines (Callmer B6100);
shells, and burnt bones slag with burnt bone; 3 nails
866 SE and SW Soft light redebrown silt in and around beach pebbles e 109.1-2 e
sextants
868 ME sextant Friable brown, black and grey mixed silt and ash Iron knife fragment; green 116.1-16 68
containing rare charcoal and burnt bone rhyolite spindle whorl
871 SE sextant Loose black and dark brown clayey silt with powdery e 113.1-7 e
charcoal ecks and frequent charcoal
873 MW sextant Friable black silt deposit around the central hearth e 115.1-3 e
containing small charcoal pieces, burnt bones, and small
stones
890 NE sextant Compact light brown silt containing charcoal, ash, and Schist whetstone fragment 118 e
turf fragments
894 NE sextant Medium dark brown silt containing frequent small stones e 122 e
and rare charcoal ecks
901 ME sextant Friable black silt containing charcoal, burnt bones and e 124.1-4 e
small stones
904 NE sextant Friable brown, grey and black silt and gravel containing e 127.1-3 e
30e40% charcoal
907 NE sextant Friable light brown silt with 40e50% charcoal Small fragment of copper 128.1-3 e
910 Under oors In situ landnm tephra layer below house e e 71, 79, 92, 94
and walls
913 Under 910 Light brown soil underlying landnm tephra e e 71, 92, 94

in triplicate on three separate subsamples using DiST WP3 and
pHep 3 meters immersed in 10:25 ml soil:water suspensions and
the mean values were accepted as representative of that grid
square. Inductively coupled plasma atomic emission spectroscopy
(ICPeAES) was conducted on the <180 mm fraction digested with
nitric acideaqua regia. This near-total digestion regime was chosen
because Icelandic andosols have very strong P xation, necessi-
tating the analysis of total P (Arnalds et al., 1995). The geochemical
data are summarised in Table 2, and the complete dataset,
including grid coordinates, is provided in Supplementary data
Table 1.
Statistical analyses using SPSS were employed to examine the
probability distributions of the geochemical data, correlations
between different element concentrations, and correlations
between element and soluble salt (EC) concentrations (Table 3).
Spearmans rank correlation coefcient (rs) for non-parametric data
was chosen as the most reliable correlation statistic because only
some of the probability distributions approximated the normal
curve while others were positively skewed. Interpretations of
geochemical data were aided by a survey of the physico-chemical
properties and possible sources of elements, as well as a survey
of the elements that had previously been identied in modern
reference materials (Supplementary data Tables 3 and 4).
Micromorphology samples were dried using acetone replace-
ment of water, impregnated with crystic polyester resin and thin
sectioned following the method of Murphy (1986). Thin sections
were rst scanned on a atbed scanner and then analysed with
petrographic microscopes at magnications ranging from x4-250
with plane-polarized light (PPL), cross-polarized light (XPL) and
oblique-incident light (OIL) following Bullock et al. (1985) and
1850 K.B. Milek, H.M. Roberts / Journal of Archaeological Science 40 (2013) 1845e1865

Table 2
Geochemical, loss-on-ignition, and magnetic susceptibility data.

Context Sample pH EC LOI at LOI at Magnetic Ba Ca Cu K Mg Na P Sr Zn

(mS/cm) 550  C 950  C susceptibility (ppm) (%) (ppm) (%) (%) (%) (%) (ppm) (ppm)
(%) (%) (108 m3 kg1)
795 62.2 7.1 142 13.2 2.9 1374.2 e e e e e e e e e
795 62.3 7.9 120 15.6 3.9 1416.9 e e e e e e e e e
795 62.4 7.8 130 14.1 4.3 1605.6 e e e e e e e e e
795 62.5 7.8 157 10.6 3.1 81.5 1450 7.02 266 0.73 3.59 0.50 2.63 1075 8250
795 84.1 6.7 137 21.0 1.6 602.9 e e e e e e e e e
795 84.4 8.1 110 16.4 3.7 1337.0 750 6.39 193 0.31 2.53 0.38 2.25 854 5530
795 84.6 7.1 174 13.7 4.1 1617.1 e e e e e e e e e
795 84.7 6.4 174 24.7 1.7 1063.5 e e e e e e e e e
844 88.1 7.6 160 13.7 0.9 153.4 150 2.26 79 0.09 1.05 0.50 0.28 117 308
844 88.2 7.8 183 17.8 1.4 150.0 510 2.09 81 0.10 0.74 0.36 1.05 321 252
844 88.3 7.5 167 16.5 1.6 138.2 160 1.62 75 0.08 0.77 0.35 0.28 84 176
844 88.4 7.5 191 19.5 2.4 107.3 230 1.37 81 0.08 0.59 0.38 0.37 93 238
844 88.5 7.3 169 17.2 1.5 134.7 350 1.77 79 0.09 0.72 0.36 0.60 140 184
844 88.6 6.0 195 17.2 2.2 161.6 90 1.19 72 0.05 0.60 0.26 0.12 47 118
844 88.7 6.7 209 19.8 2.2 107.4 150 0.88 75 0.06 0.29 0.19 0.33 62 92
844 88.8 7.0 160 18.2 2.0 173.1 80 1.09 58 0.04 0.40 0.17 0.13 58 142
844 88.9 7.3 170 16.5 1.3 166.1 150 1.61 65 0.07 0.68 0.26 0.26 90 190
844 88.10 6.6 203 17.0 1.3 154.7 90 1.66 73 0.08 0.67 0.30 0.20 77 166
844 88.11 7.3 165 13.6 0.6 147.2 160 2.35 81 0.12 1.02 0.45 0.25 123 154
844 88.12 5.2 191 15.4 1.3 180.3 80 1.50 84 0.07 0.79 0.34 0.24 51 130
849 90.1 5.3 432 17.1 1.4 168.3 80 1.62 101 0.07 0.77 0.33 0.26 59 198
849 90.2 4.5 514 16.7 1.3 201.5 100 1.83 106 0.09 1.00 0.40 0.35 59 196
849 90.3 4.9 595 15.8 1.5 322.5 140 2.10 133 0.10 1.09 0.40 0.64 105 548
849 90.4 6.5 494 17.9 0.9 229.2 110 2.32 113 0.11 0.93 0.41 0.31 112 244
849 90.5 5.8 312 15.1 1.4 259.7 90 1.69 95 0.08 0.91 0.37 0.30 60 198
849 90.6 5.3 488 17.0 2.1 256.2 100 1.79 103 0.09 0.94 0.40 0.54 64 232
852 95.1 4.6 259 12.3 0.9 142.5 180 2.93 93 0.12 1.50 0.69 0.31 86 128
852 95.2 4.8 470 12.1 1.1 250.7 100 1.56 115 0.09 1.21 0.35 0.52 54 174
852 95.3 5.3 663 13.9 1.0 240.9 90 1.60 115 0.08 1.00 0.36 0.49 54 172
859 101.1 6.8 366 14.2 1.7 199.3 330 2.41 88 0.14 0.86 0.36 0.68 202 422
859 101.2 6.3 395 12.7 1.3 163.8 110 2.34 76 0.10 1.05 0.45 0.30 109 166
859 101.3 6.5 131 15.3 2.3 219.3 150 1.85 84 0.09 0.75 0.32 0.44 126 298
861 131.1 6.7 415 11.5 1.3 222.7 380 3.66 103 0.20 0.97 0.50 0.83 308 478
861 131.2 6.8 344 10.0 0.7 200.0 210 3.05 83 0.13 1.11 0.58 0.46 143 234
861 131.3 6.7 430 18.1 1.9 185.1 320 2.53 88 0.11 0.83 0.41 0.64 167 324
861 131.5 6.7 137 18.4 2.8 167.1 110 1.47 83 0.05 0.53 0.22 0.23 79 206
862 132.1 6.9 142 17.9 2.4 324.1 270 2.92 121 0.10 0.78 0.38 0.63 195 490
862 132.3 6.3 222 18.2 2.3 205.1 210 1.60 92 0.06 0.64 0.28 0.52 87 206
862 132.4 7.5 127 16.9 1.5 103.6 480 2.24 85 0.10 0.72 0.44 0.75 187 182
864 110.1 7.0 244 21.5 1.4 310.1 490 3.71 110 0.13 0.81 0.33 1.18 353 614
864 110.2 7.4 169 16.8 1.1 336.2 500 3.81 115 0.14 1.06 0.44 1.08 311 612
864 110.3 8.0 224 19.2 2.3 431.3 330 3.39 119 0.11 0.83 0.31 0.85 312 658
864 110.4 6.1 215 20.3 1.0 481.1 340 2.55 131 0.10 0.94 0.38 0.94 201 708
864 110.5 7.9 185 20.0 1.4 566.3 390 3.44 154 0.15 0.88 0.35 1.11 310 1115
864 110.6 7.8 154 18.5 1.8 533.7 240 3.23 134 0.14 0.91 0.35 0.79 290 1005
864 110.7 5.5 386 21.2 1.5 1079.6 540 2.95 180 0.16 0.99 0.29 1.60 304 1460
864 110.8 4.8 361 21.8 1.4 742.0 450 2.69 158 0.15 1.06 0.36 1.19 229 992
864 110.9 5.4 122 24.6 1.6 649.0 490 2.34 154 0.15 0.89 0.31 1.39 182 1070
864 110.10 6.7 118 17.5 0.5 564.5 360 2.64 161 0.14 1.03 0.43 1.16 158 690
864 110.11 6.1 177 21.0 0.8 462.3 280 2.19 116 0.09 0.72 0.29 1.00 181 700
864 110.12 7.9 202 18.6 1.7 417.2 280 3.15 115 0.12 0.82 0.29 0.85 296 824
864 110.13 6.5 132 21.0 1.4 380.3 450 2.79 119 0.10 0.74 0.33 1.45 269 664
864 110.14 7.7 141 19.8 2.0 384.2 360 3.43 117 0.10 0.78 0.32 0.85 295 570
864 110.15 7.7 140 16.4 1.5 221.8 140 2.33 117 0.08 0.79 0.35 0.36 135 312
864 110.16 6.3 435 16.8 0.3 311.0 340 2.32 116 0.09 0.82 0.34 0.81 177 434
864 110.17 5.8 302 15.8 1.2 232.0 170 1.78 102 0.07 0.76 0.33 0.52 100 220
866 109.1 6.9 273 13.0 1.5 150.4 170 2.32 79 0.10 0.93 0.41 0.49 130 154
866 109.2 6.2 480 12.5 1.2 178.2 120 2.37 76 0.11 1.07 0.46 0.39 112 166
868 116.1 6.2 95 16.0 1.4 750.0 430 2.09 153 0.20 0.86 0.37 0.89 138 624
868 116.2 5.7 234 19.7 2.2 222.7 370 1.54 121 0.07 0.64 0.33 0.72 83 522
868 116.3 6.7 253 22.2 0.2 280.4 440 2.07 147 0.10 0.73 0.46 0.95 124 614
868 116.4 7.7 143 24.6 2.4 241.7 430 4.10 137 0.13 0.77 0.41 1.01 317 570
868 116.5 7.5 152 17.3 1.3 191.7 320 2.92 110 0.11 0.94 0.39 0.60 177 414
868 116.6 7.2 303 17.0 0.6 165.7 320 2.86 117 0.09 0.84 0.43 0.64 177 454
868 116.7 6.9 162 17.3 1.1 306.9 360 2.72 152 0.12 0.96 0.45 0.73 153 530
868 116.8 7.1 116 16.7 0.7 361.6 440 2.96 183 0.14 1.09 0.48 0.91 168 1740
868 116.9 7.6 173 20.9 1.5 550.2 380 3.21 176 0.14 1.05 0.41 1.20 202 914
868 116.10 7.9 167 18.4 0.5 296.3 440 3.79 117 0.13 0.87 0.40 0.86 285 688
868 116.11 7.9 184 21.3 1.1 338.5 360 3.62 148 0.11 0.86 0.41 0.82 256 640
868 116.12 7.6 161 14.6 0.3 257.1 230 3.63 114 0.10 1.20 0.62 0.49 194 378
868 116.14 5.0 137 22.4 2.9 561.2 610 1.73 246 0.17 0.73 0.27 1.42 137 1150
868 116.15 6.9 191 17.2 1.8 155.9 120 2.14 104 0.06 0.77 0.41 0.25 77 184
 K.B. Milek, H.M. Roberts / Journal of Archaeological Science 40 (2013) 1845e1865 1851

Table 2 (continued )

Context Sample pH EC LOI at LOI at Magnetic Ba Ca Cu K Mg Na P Sr Zn

(mS/cm) 550  C 950  C susceptibility (ppm) (%) (ppm) (%) (%) (%) (%) (ppm) (ppm)
(%) (%) (108 m3 kg1)
868 116.16 6.8 170 16.5 1.4 175.7 150 2.44 113 0.08 0.88 0.45 0.35 109 278
871 113.1 6.0 370 19.9 1.1 378.1 210 2.23 147 0.15 0.92 0.37 0.75 151 570
871 113.2 6.3 760 16.6 1.7 605.9 240 3.00 115 0.17 1.03 0.41 0.75 304 582
871 113.3 6.0 530 20.4 2.4 367.0 370 2.37 151 0.15 0.82 0.30 1.04 247 782
871 113.4 6.6 1088 13.8 1.3 257.9 240 3.26 100 0.13 1.06 0.48 0.73 256 626
871 113.5 6.8 954 15.1 1.0 147.4 190 2.91 95 0.12 1.05 0.51 0.41 176 238
871 113.6 6.8 1159 15.9 0.5 184.1 170 2.79 100 0.12 1.02 0.49 0.42 172 252
871 113.7 6.7 1068 15.0 1.0 164.9 160 2.57 89 0.11 0.96 0.46 0.38 148 216
890 118 5.1 239 16.1 1.1 323.5 570 1.87 123 0.08 0.87 0.35 0.95 126 456
894 122 7.3 148 18.6 2.4 163.8 140 1.85 80 0.06 0.64 0.37 0.26 72 176
904 127.1 7.6 179 17.7 1.7 146.4 110 1.99 86 0.06 0.70 0.37 0.24 78 144
904 127.2 6.9 156 18.3 1.7 142.6 130 1.81 96 0.06 0.71 0.36 0.30 73 144
904 127.3 6.7 86 16.9 1.5 147.4 130 1.89 100 0.08 0.80 0.45 0.33 65 166

Stoops (2003). Key micromorphology descriptions are summar- 4. Results

ised in Table 4, and full descriptions of all micromorphology
samples are provided in Supplementary data Table 2. The identi-
cation of anthropogenic materials such as bone, dung, and ash
was aided by modern reference collections (Supplementary data
Table 5).
3.3. Data presentation
Geochemical data and artefact locations were plotted using
ArcGIS. Fourteen of the 36 elements determined by ICPeAES had
clustered, non-random patterning (Al, Ba, Ca, Cu, Fe, K, Mg, Mn, Na,
Ni, P, S, Sr, Zn), of which Ba, Ca, Cu, Mg, P, K, Na, Sr and Zn were
selected for presentation here due to their potential contribution to
the interpretation of activity areas. A few oor layers overlapped,
and in these cases, for ease of presentation, the uppermost sample
from the affected grid square was chosen for display. Although
geochemical data are often graphically represented by surface
contours based on data interpolations, this study presents point
data because numerous oor contexts are displayed at the same
time, and data from different contexts cannot be interpolated. After
experimenting with different ways of binning geochemical values
in the GIS, it was noted that the standard deviations from the mean,
rather than the raw values, showed more pronounced patterning
(greater differences in symbol sizes), and this was therefore the
preferred method for graphical presentation of the ICPeAES data in
Figs. 6e8. The raw data are presented in Table 2 and the complete
original dataset is available in Supplementary data Table 5.
4.1. Field evidence for activity areas
The house at Aalstrti 14e18 was 16.70 m long and 3.74e
5.81 m wide, with 1.27e1.72 m thick turf walls faced with stones
(Figs. 2 and 3). There was one entrance towards the northern end of
the eastern long wall and a narrower one in the southwest corner of
the house. The northeast entrance was stone-paved, and an ante-
chamber was later added to it that extended the length of the
entrance passage to 4.8 m. Within the entrances were alignments
of post holes, and the fact that context boundaries respected these
alignments suggests that the posts supported partition walls
separating small entrance rooms e possibly acting as wind breaks.
Context boundaries also respected an area in the northwest corner
where there were three rows of stake and post holes oriented
perpendicular to the western long wall, and an area in the south-
east corner where there was a row of post holes perpendicular to
the southern end wall; both corners therefore appear to be discrete
spaces surrounded by partition walls. The large central hearth was
lined with curb stones and contained a at stone slab that had been
blackened, reddened, and cracked by heat. There was a rectangular
setting of four stake holes towards the northern end of the hearth,
which may represent a wooden superstructure that had been
suspended over the re, perhaps for a grill or spit. The post hole
cluster south of the central hearth probably represents a series of
post replacements, while most posts were in rows parallel to the
long walls of the house, dividing the space into three aisles (Fig. 3).

Table 3
Correspondence analysis based on Spearmans rho (rs) for electrical conductivity (EC) and multi-element values.

Al Ba Ca Co Cu EC K Mg Na P S Sr Zn
Al 1 0.320** 0.183 0.253* 0.064 0.181 0.107 0.408** 0.022 0.130 0.090 0.299** 0.040
Ba 0.320** 1 0.583** 0.206 0.637** 0.277* 0.670** 0.144 0.064 0.900** 0.175 0.808** 0.778**
Ca 0.183 0.583** 1 0.209 0.501** 0.054 0.751** 0.578** 0.475** 0.583** 0.621** 0.852** 0.635**
Co 0.253* 0.206 0.209 1 0.451** 0.039 0.253* 0.147 0.035 0.292** 0.154 0.218 0.463**
Cu 0.064 0.637** 0.501** 0.451** 1 0.155 0.597** 0.341** 0.035 0.787** 0.157 0.532** 0.848**
EC 0.181 0.277* 0.054 0.039 0.155 1 0.025 0.295** 0.143 0.131 0.176 0.120 0.170
K 0.107 0.670** 0.751** 0.253* 0.597** 0.025 1 0.620** 0.344** 0.718** 0.432** 0.761** 0.716**
Mg 0.408** 0.144 0.578** 0.147 0.341** 0.295** 0.620** 1 0.655** 0.265* 0.457** 0.307** 0.294**
Na 0.022 0.064 0.475** 0.035 0.035 0.143 0.344** 0.655** 1 0.013 0.541** 0.157 0.007
P 0.130 0.900** 0.583** 0.292** 0.787** 0.131 0.718** 0.265* 0.013 1 0.133 0.797** 0.875**
S 0.090 0.175 0.621** 0.154 0.157 0.176 0.432** 0.457** 0.541** 0.133 1 0.413** 0.281*
Sr 0.299** 0.808** 0.852** 0.218 0.532** 0.120 0.761** 0.307** 0.157 0.797** 0.413** 1 0.780**
Zn 0.040 0.778** 0.635** 0.463** 0.848** 0.170 0.716** 0.294** 0.007 0.875** 0.281* 0.780** 1

N 79.
*Correlation is signicant at the 0.05 level (2-tailed).
**Correlation is signicant at the 0.01 level (2-tailed).
1852 K.B. Milek, H.M. Roberts / Journal of Archaeological Science 40 (2013) 1845e1865

Table 4
Summary micromorphology descriptions of the occupation deposits, with representative descriptions of the overlying turf collapse layers (units 747, 792, 858) and underlying
soil and tephra layers (units 824, 910, 913) shaded grey.

Sample Context number and Structure Porosity Groundmass

Number microstratigraphic unit
Microstructure Orientation of Planar Simple Compound Vesicles Channels Texture
basic components voids packing packing voids and classication
voids vughs
68 747 UfGr, Ch R - --- ---- Zl
74 792 UfGr, Ch R -- ---- zCl
71 858.1 UfGr, Ch R --- ---- Zl
67 n/a Cr Min: R; Org: H ----- --- sZl
68 868 M H - -- Zl
71 864 Cr, Ch, local UfGr R ---- ---- Zl
74 793/795 Sp H -- ----- sZl
74 802 Ch H - ---- Zl
74 831 UfGr, Ch H -- - -- ---- sZl
75 795.1 Sp H ----- sZl
75 795.2 Ch H - - ---- sZl
75 795.3 Ch H - ---- sZl
75 795.4 Ch H - --- Zl
75 802 Ch H -- - ---- Zl
79 851 Cr; Sp in top 10 mm Min: R; Org: H, local R ----- --- sZl
80 844.1 Cr, local Sp R; Org: local H --- ----- Zl
80 844.2 UfGr, with Ch Min: R; Org: H -- --- Zl
80 844.3 Cr R; Org: local H --- ---- zCl
80 844.4 Pl, local Cr Org: H; local R -- -- ---- zCl
94 763.1 sBl,Ch H -- --- Zl
94 763.2 sBl Ch R - -- Zl
94 763.3 sBl;local Sp, Pl H -- ---- Zl
94 763.4 sBl, Ch; local Sp, Pl H -- ---- Zl
94 824 sBl, Ch H --- - ---- S
94 910 sBl, Ch, local Sgr Min: R; Org: H -- -- ---- sZl; localfS
94 913 sBl, UfGr, Ch Min: R; Org: H -- --- ---- Zl

Present in trace amounts, - <2%, -- 2e5%, --- 5e0%, ---- 10e20%, ----- 20e30%, ------ 30e40%, ------- 40e50%, --------
50e60%, --------- 60e70%.
Microstructure: local: localised; Ch: channel, Cr: crumb, Gr: granular, M: massive, Pl: platey, Sgr: single grain, Sp: spongy, sBl: subangular blocky, UfGr: ultrane granular.
Dominant orientation/distribution of basic components: Min: Mineral material, Org: Organic matter; H: horizontal, sH: subhorizontal, R: random.
Texture classication: C: Clay, zCl: silty clay loam, S: sand, fS: ne sand, vfS: very ne sand, Z: silt, Zl: Silt loam, sZl: sandy silt loam.
Coarse/ne related distribution: En: enaulic, Mon: monic, Por: porphyric.
Nature of ne mineral material (PPL): Br: brown, pBr: pale brown, Gr: grey, Or: orange, OrBr: orangey brown, RBr: reddish brown, D: dotted, Sp: speckled. Listed in
descending order of frequency.
Birefringence fabric (XPL): Cr: crystallitic, mCr: microcrystallitic, Un: undifferentiated.

The occupation deposits varied considerably in colour, texture
and inclusions, enabling 25 to be distinguished (Table 1). The layers
in the central aisle were very compacted and rich in charcoal and
ash, suggesting the spreading and trampling of hearth waste in this
space. Context 844, in the northwest corner of the building, was
noted to be particularly organic. In the middle of the western side
aisle the occupation deposit was so thin and patchy that no context
was recorded, and the excavators suggested that this area might
have been covered by a wooden platform.
4.2. Artefact and bone evidence for activity areas
The distribution of artefacts and bones in these occupation
layers provided additional information about the locations of
activity areas (Fig. 4). There was a concentration of artefacts such as
beads, nails, a spindle whorl and a fragment of vessel glass (rare in
the Viking Age) around the central hearth, suggesting this was
a focus for daily activities. The quartz stones east of the hearth
might have been used as gaming pieces for the Viking Age board
game hnefata, which is known from medieval written sources (e.g.
Frijlfs Saga, Hervrs Saga, Saga of Gunnlaug Serpent-Tongue).
Clusters of jasper akes and strike-a-lights around the northern
half of the hearth provide a good indication of where the re was lit.
The piece of pumice, the knife and the spindle whorl in the eastern
side aisle suggest that this may have been a sitting and working
area, while the four loomweights clustered in the southern part of
the western side aisle suggest the location of an upright loom. The
cluster of quartz stones in a post hole in the northeast corner can be
interpreted as a foundation or closing deposit rather than an every-
day activity area.
In the eastern side aisle northeast of the hearth there was
a cluster of unburnt bones, which could have been the location of
a butchery area, but they could also have been placed there for
storage or disposal after the meat had been consumed. Every part
of the house except for the western side aisle appears to have
received the deposition of hearth waste and burnt bones. The
largest cluster of burnt bones outside the hearth was at the
southwest entrance, indicating that hearth waste was deliber-
ately dumped or swept there. A pile of reddened, blackened and
heat cracked cobbles excavated just beyond the southwest
entrance reinforces the interpretation of this as a route for waste
disposal.
4.3. Geochemical evidence for activity areas
4.3.1. pH
Conditions in the central hearth and the deposits around it were
alkaline (pH 7e8), and most of the oor deposits were neutral to
alkaline (Fig. 5a). Since pure water in equilibrium with atmospheric
CO2 has pH 5.6, the hearth and oor deposits with pH >5.6 acted as
a base, contributing alkali salts such as Ca2, Na, Mg2, or K to the
soil solution. A common source of alkali salts in domestic contexts
 K.B. Milek, H.M. Roberts / Journal of Archaeological Science 40 (2013) 1845e1865 1853

Biomin. Organic Matter

Coarse: Coarse/ Nature of Birefringence Phytoliths Diatoms Amorphous Amorphous Fungal Fungal
ne ratio ne related ne mineral fabric (XPL) (dark brown (light brown to sclerotia spores
(10 mm) distribution material (PPL) to black) light reddish brown)
30:70 Por OrBr, rBr, Sp Un -- - --
20:80 Por OrBr, rBr, Sp Un -- - -- -- - -
30:70 Por OrBr, rBr, Sp Un -- - --
30:70 Por Br, OrBr, D Un -- --- --- -
30:70 Por Br, rBr, D Un ----- -- ------ -
30:70 Por Br, OrBr, D Un ------- --- ---- - ---
50:50 Por pBr, rBr, D Un -- --- ----
5:95 Por pBr, Gr, rBr, Sp Un; local mCr - -
20:80 Por, En Br, pale pBr, Sp; Un; local mCr - -
25:75 Por Br, OrBr, D Un - --- -- -
25:75 Por pBr, Gr, rBr, Sp Un; local mCr -
25:75 Por Br, OrBr, D Un - ---- -
5:95 Por pBr, Gr, rBr, Sp Un; local mCr -
5:95 Por pBr, rBr, Gr, Sp Un; local mCr - -
90:10 Por Br, OrBr, D Un -- ---- ---
30:70 Por Br, RBr, D Un --- ----- --- -
25:75 Por OrBr, RBr, Sp Un -- - ---
20:80 Por pBr, OrBr, D Un ----- -- ---- - --
20:80 Por Br, RBr, D Un ----- ----- ---- - -
25:75 Por pBr, OrBr, Sp Un --- -- ----
30:70 Por pBr Un - - -- -
25:75 Por pBr, OrBr, Sp Un --- -- ---- - -
25:75 Por pBr Br, RBr, D Un --- --- --- - -
5:95 Por OrBr, Sp Un - - -
90:10 Por; local Mon, En Br, D Un - --
40:60 Por OrBr, RBr, Sp Un - - --

is wood ash (Evans and Tylecote, 1967; Pierce et al., 1998), and
inclusions of wood charcoal were also found in most of the oor
deposits in variable quantities, suggesting that wood ash was
spread from the hearth throughout the building. The distribution of
alkaline pH values, like the distribution of burnt bones, which
travelled with the hearth waste, is therefore not indicative of
specic activity areas in the building but of a oor maintenance
practice that involved the intentional spreading of ash e a practice
used until the early 20th century in Iceland to keep oors dry and
smooth (Milek, 2012).
While most grid squares contained sediment that was neutral or
alkaline, there were several grid squares with extremely low pH
values (4.5e5.0), notably a 1 m wide strip south of the hearth. This
strip was located below a 19th-century foundation trench that had
contained wet, organic sediment, and its pH can be attributed to the
accumulation of organic acids. Any bone originally deposited in this
area had little chance of surviving (Fig. 4c).
4.3.2. Electrical conductivity
The electrical conductivity of the oor sediments was generally
very low, but context 871, in the southeast corner, had electrical
conductivity values ten times higher than most other contexts,
indicating very high nutrient or soluble salt levels (Fig. 5b). Since
soluble salts are susceptible to leaching, the original concentrations
must have been even higher. Identication of the salts present
required elemental analysis (Section 4.3.5).
4.3.3. Loss on ignition
Organic matter content estimated by loss on ignition at 550  C
was signicant throughout the occupation deposits (Fig. 5c).
Concentrations were notably high in the central aisle around and
north of the hearth (12e25%), in the northwest corner of the
building (19e22%) and especially in the eastern side aisle (22e25%),
indicating the locations of activities, ooring materials or organic
furnishings (e.g. bedding material) that resulted in accumulations
of organic matter. Since ignition at 550  C combusts both charred
and uncharred organic remains, and both plant- and animal-
derived materials, interpretations about specic activities or
furnishings required additional data from micromorphological
analysis (Section 4.4).
Carbonate content measured by loss on ignition at 950  C was
highest in the hearth (c. 4%), where it was probably derived from
CaCO3 in wood, peat, and turf ash (Canti, 2003). Individual grid
squares in the eastern and southern edges of the building where
carbonates were also in the region of 3e4% (Fig. 5d) are probably also
places where wood ash and calcined bones were concentrated, an
interpretation supported by the distribution of burnt bones (Fig. 4d).
4.3.4. Magnetic susceptibility
High magnetic susceptibility values were limited to sediments
in and adjacent to the hearth, with slightly elevated values also
present in the eastern side aisle next to the hearth (Fig. 6a). Since
wood ash, charcoal and bone are not magnetic, the magnetic
enhancement of the hearth ash indicates the presence of soil
particles, pebbles and/or iron nodules that were magnetically
enhanced by heating. This suggests that peat and/or turf was used
as fuel in addition to wood or that heated soil material from the
base of the hearth was mixed with the ash residues sampled in and
near the hearth.
4.3.5. Multi-element analysis
Levels of P, Ca, K, Mg, Zn, Ba, and Sr were highest in the central
hearth deposits at three or more standard deviations above the
mean (Figs. 6 and 7). However, all of these elements also had
elevated concentrations in the central isle around and north of the
hearth, and in the eastern side aisle east of the hearth, in areas that
also have signicant organic matter content. These elements are
either plant macronutrients or are trace elements commonly
1854 K.B. Milek, H.M. Roberts / Journal of Archaeological Science 40 (2013) 1845e1865

Table 4 (continued)

Inclusions Pedofeatures

Charcoal Bone Burnt bone Rubied ne Microcrystalltic Aggregates/ Fe nodules Fe plant Silty clay Excrements
mineral granules lenses of pseudomorphs intercalations
material tephra
------ ----- --
-- --- ------ -
--- ----- ----
--- - - ---- ---
- --
-- -- -- ----
-- --- - --
----- -- -- - - - -
- --- ------ --- - - -
- -- ---- -- - --
--- - ------- -- - -- -
------ --- - - - - -
--- -- ----- --- - - -
- --- ------ -- - -
----- - -- - - - ---
--- --
------- ----- -
-- -- -- ---
- -- --
- - - -- -
----- - - ------ --- - -
- - - - - -- -
-- - - - - --- --
- --- --
-------- --
----- -- -

present in ground water (Sr), which are taken up by plants and pass
through the food chain to animals (Supplementary data Table 3).
They are incorporated into hard and soft organic tissues and are
present in elevated levels on archaeological sites wherever plant or
animal tissues or their ashes are deposited (Wilson et al., 2008;
Misarti et al., 2011). The metals Cu and Ni, which had the same
distribution as organic matter and its associated elements, had
clearly followed the same depositional pathways (Fig. 8bec).
The multi-element data were interrogated by the visual
comparison of distribution maps and statistical correlations in
order to determine which soluble salts might be responsible for the
high electrical conductivity values in the southeast corner of the
house (context 871). None of the element distribution plots were
identical to the EC distribution, but Ca and Na both showed slight
elevations in this area, and Mg was higher overall in the southern
end of the house than in the northern (Figs. 6bed and 8a). Spear-
mans rank correlation coefcient (rs) showed a strong positive
correlation between EC and Mg, which was statistically signicant
at the 0.01 level (Table 3), and which points towards the presence of
Mg2 salts. However, some common salts, such as chloride (Cl),
bicarbonate (HCO3), ammonium (NH 
4 ), nitrate (NO3 ), and nitrite
(NO 2 ), cannot be detected by ICPeAES.
4.4. Micromorphological evidence for activity areas
4.4.1. North end
Analysis of thin section AST01-80 conrmed the high organic
content noted in context 844 in the eld and in the loss on ignition
data, and provided more information about its origin. Four sepa-
rate lenses were identied in thin section (Supplementary data
Table 2.14), and although these had been heavily reworked by
soil fauna, it was possible to see the original horizontal bedding of
the organic matter in localised areas (Fig. 9a). The uppermost and
lowermost lenses of the occupation deposit, 844.1 and 844.4,
were stained dark brown by organic acid pigmentation and con-
tained 20e30% dark brown, partially decomposed plant matter
and 10e30% phytoliths (Table 4). The phytoliths included short,
broken strands of articulated silica skeletons that are distinctive of
grasses that have been chewed by animals, suggesting the pres-
ence of herbivore dung. It is also possible that elongated strands of
vegetal matter have been broken up by bioturbation, and that this
area was used for hay storage or was a grass-bedded sitting/
sleeping area for the human occupants of the house (Fig. 9a, b).
Any faecal spherulites, if produced, did not survive. Faecal
spherulites have not been found in any reference dung samples or
stabling deposits in Iceland, and they are either not produced in
Iceland or were post-depositionally dissolved by rainwater (Canti,
1999).
Lens 844.3 was composed primarily of phytoliths and very
pale brown amorphous organic matter, and also contained 2e5%
charcoal, 2e5% small bone fragments and an unusually high
concentration of fungal spores (2e5%) (Fig. 9a). The bone frag-
ments were smaller than 5 mm and highly weathered, with
abundant pits and cracks and class 0e1 weathering rims (cf.
Bullock et al., 1985). The presence of minute, chemically weath-
ered bone fragments, the yellowish/pale brown colour of the ne
organo-mineral matrix in which they were embedded, and the
abundance of fungal spores and phytoliths (some in discrete
aggregates) suggest that this layer contained a mixture of
omnivore and herbivore dung, possibly mixed with vegetal
matter such as hay, while the charcoal indicates that wood ash
was occasionally sprinkled here. Lens 844.2 was a thin lens of
turf containing the landnm tephra layer. This lens had not been
identied in the eld, and may have been a localised inclusion. If
the turf was intentionally deposited, it may have served as
bedding material.
 K.B. Milek, H.M. Roberts / Journal of Archaeological Science 40 (2013) 1845e1865 1855

a b

c d

Fig. 4. Distributions of (a) artefacts, (b) stone manuports and jasper akes, (c) bones, and (d) burnt bones on the oor of the Viking Age house. Note that bone distribution data is
excluded from context 844, in the northwest corner of the house, because the bones from this context were accidentally lumped together in a single bag. (For interpretation of the
references to colour in this gure legend, the reader is referred to the web version of this article.)
1856 K.B. Milek, H.M. Roberts / Journal of Archaeological Science 40 (2013) 1845e1865

a b

c d

Fig. 5. Distributions of (a) pH, (b) electrical conductivity, (c) percent loss on ignition at 550  C, and (d) percent loss on ignition at 950  C.
 K.B. Milek, H.M. Roberts / Journal of Archaeological Science 40 (2013) 1845e1865 1857

a b

c d

Fig. 6. Distribution of (a) magnetic susceptibility, (b) barium, (c) calcium, and (d) copper.
1858 K.B. Milek, H.M. Roberts / Journal of Archaeological Science 40 (2013) 1845e1865

a b

c d

Fig. 7. Distributions of (a) magnesium, (b) phosphorus, (c) potassium, and (d) sodium.
 K.B. Milek, H.M. Roberts / Journal of Archaeological Science 40 (2013) 1845e1865
a b
Fig. 8. Distributions of (a) strontium, and (b) zinc.
4.4.2. Central aisle
The layer captured in sample AST01-71, just north of the
hearth, contained only 2e5% charcoal and it did not contain any
of the ash or burnt bone that had characterised the central aisle
contexts in the eld, microrefuse and chemical analyses (Table 4).
Where it was thin sectioned, oor context 864 contained 40e50%
amorphous, decomposed organic matter, an unusually high
concentration of fungal spores (5e10%), and a high proportion of
phytoliths (40e50%). About 20% of the layer consisted of small
(<5 mm) grey aggregates of phytoliths, which were tightly
packed together in randomly oriented, short, articulated
segments identical to reference samples of sheep or goat dung
(Fig. 9c). The deposit did not show any evidence of compaction
by trampling and is therefore interpreted as including the
remains of dung that had been stored next to the hearth for use
as fuel.
4.4.3. Central hearth
The two thin sections taken from the hearth deposits, samples
AST01-74 and AST01-75, contained lenses with either high
concentrations of charcoal or the microcrystallitic CaCO3 granules
and rubied iron nodules characteristic of peat and turf ash, as well
as lenses that contained a mixture of both wood and peat ashes
(Table 4; detailed descriptions and interpretations in
Supplementary data Tables 2.7e2.10). In the alkaline environment
of the hearth, the CaCO3 aggregates normally associated with
charcoal should have survived, and the low frequency of these
aggregates in thin section may be indicative of low-temperature
burning (<400  C) (Simpson et al., 2003). In sample AST01-71,
there was a lens of burnt sh bone that lay at the boundary
between contexts 802 and 831. It is possible that some of the burnt
bones in these ash layers were a product of accidental loss during
1859
cooking, but most of the bone was probably intentionally thrown
into the re once the meal was consumed as a convenient and
sanitary method of waste disposal (Tinsley and McGovern, 2002).
4.4.4. Western side aisle
Sample AST01-67 was taken from the western side aisle,
where there had been no oor layers to record or sample in the
eld, and, on the basis of this negative evidence, where the
excavators had interpreted the location of a oor covering or
wooden platform. In the location where thin section 67 was
taken, however, it was possible to identify a very thin occupation
deposit below the turf collapse: an organic silt loam about 8 mm
thick, containing 5e10% charcoal and <2% bone and burnt bone
(Fig. 9d, Table 4, Supplementary data Tables 2.1 and 2.2). The
deposit had been heavily reworked by soil fauna, but even where
the original fabric had survived bioturbation, neither the orga-
nisation nor the microstructure of the sediment showed any
evidence of compaction by trampling. This provided independent
support for the interpretation derived initially from the eld
evidence, that this part of the oor had not been walked on and
there was probably a raised wooden platform in this area.
Sample AST01-79 captured a small context, 851, which was very
heterogeneous, containing an abundance of charcoal (20e30%) and
amorphous organic matter (30e40%), as well as ash (2e5%), burnt
bone (2e5%), and bone (<2%) (Fig. 9e). The sediment was porous
and uncompacted, and contained relatively large, randomly
oriented charcoal fragments (up to 7 mm), suggesting that the layer
had not been trampled. It could not have been the product of a one-
off dumping event, since there was a lens of very coarse sand and
ne gravel running through the middle of this layer, but it could
have accumulated away from the main pathway of foot trafc,
either underneath or up against a piece of furniture. The horizontal
1860 K.B. Milek, H.M. Roberts / Journal of Archaeological Science 40 (2013) 1845e1865

Fig. 9. Photomicrographs (PPL) of (a) micromorphology sample 80, context 844.3, showing a localised area with horizontally bedded bone, plant mater, and phytoliths surrounded
by sediment disturbed by soil fauna, (b) micromorphology sample 80, context 844.4, showing truncated, articulated phytoliths typical of herbivore dung, (c) micromorphology
sample 71, context 864, showing the tightly packed, randomly oriented segments of articulated phytoliths typical of herbivore dung, (d) micromorphology sample 67, showing
charcoal, bone and decomposing plant matter in an area where there no oor context was identied in the eld, (e) micromorphology sample 79, context 851, showing
uncompacted and relatively large fragments of charcoal, bone and burnt bone, and (f) micromorphology sample 68, context 868, showing very compacted, horizontally bedded
lenses of phytoliths and decomposing plant matter. b: bone, bb: burnt bone, ch: charcoal, o: amorphous organic matter, ph: phytoliths, v: void. (For colour images, the reader is
referred to the web version of this article.)

displacement of artefacts by kicking and scufng, their accumula-
tion on the edges of oors and pathways, and their tendency to
accumulate against physical barriers, have been observed in eth-
noarchaeological and experimental studies (e.g. Nielsen, 1991;
Stockton, 1973; Wilk and Schiffer, 1979).
4.4.5. Eastern side aisle
The oor deposit in the eastern side aisle, context 868, was
described in the eld as a mixture of brown, black, and grey silt and
ash containing rare charcoal and burnt bone fragments. In thin
section AST01-68, however, this layer was composed of a very
highly compacted organic silt loam, with horizontally bedded
lenses of organic matter, articulated phytoliths, and dark greyish
brown organic silts (Fig. 9f). Although it contained the odd charcoal
fragment (<2%), the occupation deposit captured in this particular
sample did not contain any ash. Instead, it consisted almost entirely
of very long, intact strands of herbaceous plant matter (e.g. grass)
that had been heavily compacted and had decomposed,
 K.B. Milek, H.M. Roberts / Journal of Archaeological Science 40 (2013) 1845e1865
undisturbed, in situ. Although there is a remote possibility that this
vegetal matter was derived from cow dung, and was compacted by
animals rather than humans, the layer does not appear to have been
disturbed by hooves and the stabling of cows in this particular area
seems unlikely due to the layers proximity to the central re and
the concentration of bones and artefacts, including a spindle whorl
(Fig. 4). The deposit can best be interpreted as compacted vegetal
bedding material that had been used for a sitting and sleeping area
by the human occupants of the house.
4.4.6. Southwest entrance
Sample AST01-94 was taken from deposits that lled the
southwest entrance. The uppermost layer in the sequence in the
door consisted of medium brown clayey silt with inclusions of the
landnm tephra layer and occasional charcoal ecks (context 763;
Table 4). In the eld, this layer was interpreted as turf collapse, but
in thin section the lowermost part of it was made up of numerous
compact lenses of waterlain silty clay, plant matter, organic soils,
wood ash, and peat ash (Supplementary data Tables 2.17 and 2.18).
Below this occupation deposit the reddish brown soil (context 824)
had also been reworked by water, consisting of multiple, ne lenses
of well-sorted silt in ning-up sequences such as those typically
found in puddles.
5. Discussion
5.1. Interpretation of the use of space in the Viking Age house based
on integrated data
In this 10th-century house, activity areas and associated occu-
pation deposits were divided between three aisles dened by rows
of roof-supporting posts, as well as three main functional areas,
which were interpreted on the basis of integrated eld evidence,
artefact and bone distributions, and geoarchaeological evidence:
a multi-functional central living room that was centred on the
hearth, and two areas that appear to have more specialised func-
tions in the narrow ends of the building (Fig. 10). Linking these
three main functional areas was the central aisle, which contained
very compacted sediments and had clearly been the main corridor
for foot trafc down the length of the building. This central aisle
had received inputs of organic matter and ash, abundant charcoal
and burnt bone inclusions, and chemically it was characterised by
elevations in a suite of elements associated with organic matter and
its ash.
While some of the ash in the oor sediments could be from
accidental spillage from the hearth and the movement of material
by sweeping and trampling, the ubiquity of burnt bones and
charcoal throughout the building, including the entrances and
areas that had been separated from the central living room area by
partition walls, suggests that ash was carried around the house
and deposited deliberately. Although modern analogues must be
used critically, it is worth noting that in 19th- and early 20th-
century Iceland ash was commonly deposited on the oors of turf
houses and animal buildings in order to keep them dry and to ll
in holes (Milek, 2012). The ne particle size of ash residue enables
it to serve as an insecticide, and because it is absorbent it can
protect wooden posts and furnishings from fungal decay (Hakbijl,
2002).
The central hearth was the main source of light and heat in the
building and must have been the focus for all winter/evening
cooking, eating, handcrafts, and social activities. The nd spots of
the jasper strike-a-lights and akes indicate that the re was lit
from its northern end. Several different fuels were used, the most
important being birch wood and peat, but crumbs of sheep/goat
dung found in the thin section taken from the oor just north of the
1861
hearth suggest that dung was also used as a fuel. Bones were
frequently tossed into the re after the meat or sh was consumed,
which would have had the dual effect of feeding the re and
disposing of bone waste hygienically.
On both sides of the hearth the side aisles probably contained
sitting and sleeping spaces. In the western side aisle this was
likely to have been in the form of a raised wooden platform,
against which hearth and food residues accumulated but were
left untrampled, and under which almost no occupation material
accumulated. The cluster of abandoned loomweights on the
south end of this platform suggests that an upright loom had
been located on top of or beside it. In the east side aisle, there
was a bed of compacted hay, which, judging from the straight,
well-dened context boundaries, had probably been contained
by a wooden sill. The artefacts found in the eastern side aisle,
such as the pumice, spindle whorl and knife, suggest that
everyday craftwork took place on this grassy seat. The cluster of
quartz pebbles found in the eastern side aisle also remind us that
social activities such as gaming probably took place in this
central living room.
In the northwest corner of the house, the fragmented vegetal
remains identied in thin section suggest that the rows of posts
observed in the eld dened stalls for housing a few small cows,
sheep and/or goats e at least occasionally e and/or that the area
was used for storing hay or was the location of a sitting/sleeping
area on vegetal bedding material. The presence of omnivore
excrement containing small, weathered/digested bone fragments
suggests that people might sometimes have used this corner as
a lavatory, making the animal stall interpretation the more likely,
but it is entirely possible that this corner of the house had
a combination of these uses, or that the use changed over time. The
charcoal inclusions in this deposit indicate that ash was occasion-
ally sprinkled here to absorb moisture or odours. The turf fragment
found in this deposit hints at the possible use of turf for bedding, as
was common in other parts of the North Atlantic region until the
early modern period (Fenton, 1978).
In the southeast corner of the house there was a small alcove
where the oor deposit contained an exceptionally high concen-
tration of salts. Besides the phosphates and nitrates that could have
been associated with ash residues and decomposing organic matter
e which were actually found in much higher concentrations else-
where in the house e the salts in this area could have been derived
from sea water, seaweed (and its ash), or urine. While urine is
especially rich in N2 it also contains c. 2% Cl, K, SO2 3
4 , PO4 , and
Na . Sea water and its associated products and plants are especially
rich in Na and Cl and also contain Mg2, SO2 2
4 , Ca , K , and

HCO3 . The fact that there is a strong positive correlation between
the high electrical conductivity values and magnesium in this
deposit suggests that it is sea salt, seaweed, or seaweed ash that
were stored and/or used in this corner.
Salt had various practical uses in Viking Age Iceland. Sea salt
obtained by evaporating sea water or burning seaweed may have
been used as a preservative for butter, meat, or sh (Shetelig and
Falk, 1937, 311). However, from the Viking Age to the present, sh
in Iceland has usually been preserved by drying or smoking, and
meat has been preserved by smoking or pickling in sour whey
(Amundsen, 2004; Krivogorskaya et al., 2005). Ethnographic sour-
ces suggest that seaweed ash mixed with water, which produces
alkaline lye, was commonly used in the North Atlantic region for
cleaning and dyeing wool and fulling cloth until the early 20th
century (Buckland and Perry, 1989; Jochens, 1995, 135e140;
Shetelig and Falk, 1937, 332e336; Stead, 1981, 1982). One possible
explanation for the high concentration of salts in the southeast
corner of the house is that seaweed ashes and water were used in
this area to wash or dye wool.
1862 K.B. Milek, H.M. Roberts / Journal of Archaeological Science 40 (2013) 1845e1865

Fig. 10. Interpretive plan of the Viking Age house on Aalstrti, based on integrated eld, geoarchaeological, and artefactual evidence. (For interpretation of the references to colour
in this gure legend, the reader is referred to the web version of this article.)
 K.B. Milek, H.M. Roberts / Journal of Archaeological Science 40 (2013) 1845e1865
5.2. Evaluation of individual geoarchaeological techniques and the
integrated approach
This study of the occupation deposits in the Viking Age house in
Reykjavk illustrates the potential of a multi-method, integrated
approach to the study of activity areas on archaeological sites. The
distributions of artefacts, bones, and ashes, long recognised as
being affected by cleaning and abandonment behaviours and
scavenging by dogs, are of course also very dependent on soil
preservation conditions, which in turn depend on the pH of the soil
solution (in the case of ashes, bones and bone artefacts), and its salt
content (in the case of metals). It is standard practice to take pH
spot tests on a site, but this study has demonstrated how variable
pH and salt content (EC) can be across a site due to the variable
composition of the anthropogenic sediments (e.g. plant matter will
increase acidity, while wood ash will increase alkalinity), as well as
post-depositional conditions (e.g. the quantity and pH of perco-
lating rainwater). These results suggest that on any site where
artefact or bone distributions are used to identify and interpret
activity areas, pH and EC must be tested systematically (e.g. by
sampling on a grid) in order to ne-tune interpretations to micro-
scale preservation conditions.
In this case study, artefact and bone distributions used on their
own provided only limited insight into the use of space in the
building. Burnt bone and charcoal distributions may be used as
proxies for ash dumps on the oors of the house, and provide
information about oor maintenance activities rather than the use
of space. There is no question that with a critical evaluation of the
reliability of artefact nd spots, certain types of artefacts, such as
those associated with craft production (e.g. loomweights, spindle
whorls or pumice in this Viking Age context), re-starting (e.g.
jasper or int akes), or gaming can provide a rich source of
information about the likely locations of these activities. However,
the oors of buildings are not always rich in artefacts, many types
of artefacts cannot provide information about activities (e.g.
beads, nails) and nd spots may frequently be judged to be
unrelated to locations of use. In this study, geoarchaeological
techniques played an essential role in the identication and
interpretation of activity areas, sometimes conrming the eld
evidence, and sometimes providing new information that had not
been detected in the eld.
In this study, magnetic susceptibility provided little new infor-
mation about activity areas, merely conrming that in situ burning
only took place in the central hearth. Likewise, loss on ignition at
950  C conrmed that the hearth contained a concentration of
calcium carbonate-rich ash, and showed how ash had been
distributed around the building e a distribution that overlapped
with that of burnt bones. Loss on ignition at 550  C quantied the
organic matter and charcoal that had been observed in the eld,
and pinpointed the parts of the house that had received the greatest
inputs of this organic material (the central aisle, the eastern side
aisle, and the northwest corner). However, the nature and origin of
the organic matter, and the specic activities that took place in
these areas, could not be determined by this technique. Likewise,
electrical conductivity identied an activity area in the southeast
corner that had resulted in a concentration of soluble salts, which in
this case had not been identied in the eld, but the types of salts
cannot be determined by this technique. While these rapid and
inexpensive methods provide a spatial overview of the locations of
activity areas and spatial variations in preservation conditions,
additional, spatially overlapping data are required to interpret the
original inputs and thereby the activities that created them.
Multi-element analysis and the determination of statistically
signicant correlations were instrumental for the interpretation of
the types of salts found concentrated in the southeast corner of the
1863
building, in this case linking them to sea salts. In addition, the
overlapping distributions of the plant macronutrients P, Ca, K, Mg,
Zn, Ba, the trace element Sr, and the micronutrients Ni, Cu, all of
which are taken up by plants and transferred through the food
chain to animals, lent further support to the understanding of
where there were concentrated inputs of organic matter and ash,
which had initially been based on eld descriptions and loss on
ignition at 550  C. However, because almost all types of decom-
posed, charred, and ashed organic matter will result in the elevated
concentrations of these elements, and different activities do not
have precise chemical ngerprints, multi-element analysis alone
does not enable a precise identication of the inputs or the types of
activities that occurred in the identied activity areas (see also
Wilson et al., 2008). Moreover, because the preservation potential
of different types of elements is highly dependent on the pH of the
soil solution and the presence/absence of xing agents that can
prevent the leaching of some elements (e.g. P xed to clay, Fe or Ca),
and because of preferential loading and possible trapping of
elements such as Ca, P, Sr and Zn in charcoal or bone (Wilson et al.,
2008), the interpretation of multi-element concentrations must
always include a critical assessment of the overlapping data on pH
levels, clay content, and the concentrations of other elements,
charcoal and bone.
In this study, multi-element analysis on its own gave no hint of
the presence of herbivore or carnivore excrement in the north-
west corner of the house, which were identied in thin section.
Phosphorus levels in this area were either at the mean or were
elevated above it by only one standard deviation, and the nutrient
levels (EC) here were low. Because this deposit was dominated by
decomposed organic matter and contained little soil, there was
a lack of clay or iron to which phosphorus could x, and since pH
levels were 6e7, phosphorus in this area could have remained in
mobile ionic form and leached fairly rapidly. In this case study, the
interpretation of the animal stalls, hay storage, and/or vegetal
bedding material in the northwest corner of the building was
achieved only through soil micromorphological analysis, with no
contribution from multi-element analysis or any of the other
geochemical techniques.
Of the geoarchaeological methods employed to detect activity
areas in this case study, micromorphological analysis had the
greatest interpretive power. Most importantly for this study, in
thin section it was often possible to identify and quantify different
types of organic matter (e.g. phytoliths of grass/hay, charred and
uncharred wood and plant remains, sheep/goat dung excrements,
etc.), thus enabling the identication of the area north of the
hearth where sheep/goat dung was stored for fuel, the probable
sitting/sleeping area with grass/hay bedding in the eastern side
aisle, and the probable stabling/hay storage/vegetal bedding area
in the northwest corner of the house. The identication of
uncompacted microstructures in the thin section from the
western side aisle also supported the interpretation initially
proposed by the excavators on the basis of the absence of oor
layers that this was an untrampled area protected by a raised
wooden platform. In thin section it was also possible to identify
the specic types of ashes (e.g. wood ash, peat ash) contained in
the central hearth, and to identify waterlain deposits in the
southwest entrance area, which had not been identied in the
eld. The application of soil micromorphological analysis to the
study of activity areas is limited by the fact that block samples
cannot feasibly be taken systematically across house oors, and by
the fact that sometimes different activities can result in similar
materials and microstructures being visible in thin section.
However, this case study demonstrates that targeted micromor-
phology sampling could be of utmost importance to the inter-
pretation of site activity areas.
1864 K.B. Milek, H.M. Roberts / Journal of Archaeological Science 40 (2013) 1845e1865
6. Conclusion
The results of this integrated geoarchaeological study were
fundamental to the interpretation of the 10th-century building at
Aalstrti 16, in Reykjavik, and have been incorporated into the
museum exhibit built around the preserved foundations of the
house (Reykjavk 871  2: The Settlement Exhibition). However, the
implications of this comparative analysis of geoarchaeological
methods can be applied to the study of activity areas on archaeo-
logical sites anywhere in the world.
All of the geoarchaeological analyses used in this study enhance
eld descriptions by verifying or ne-tuning eld interpretations,
quantifying the components identied in the eld, showing the
degree to which composition and preservation conditions vary
across a single context, and in some cases providing new informa-
tion about sediment characteristics that cannot be observed in the
eld. However, while each sediment analysis provides unique and
valuable information about activity areas, used individually their
interpretive power is limited. In particular, geochemical methods
such as loss on ignition and multi-element analysis are effective at
pinpointing the locations of activity areas that received elevated
inputs of fresh, charred, and/or ashed organic materials, but due to
the wide range of possible sources of organic matter, carbonates, and
elements, it is not possible to make direct links between the spatially
identied activity areas and specic inputs or activities.
The use of multiple, overlapping datasets dramatically increases
the interpretation potential of geochemical methods by allowing
the determination of correlations between them and the critical
assessment of factors that could have affected distribution patterns.
Multi-element distributions in particular can only be interpreted
following an assessment of the degree to which the compositions of
different deposits might have had an effect on element leaching,
binding (e.g. to clay, iron, calcium) or trapping (e.g. by charcoal or
bone). The most effective way to detect and interpret activity areas
on archaeological sites is to integrate as many complementary
methods as possible, using multiple geochemical methods to
identify the possible locations of activity areas, followed by soil
micromorphological analysis to provide detail about the composi-
tion, organisation and compaction of the components in each of the
identied occupation deposits.
Acknowledgements
The excavation was funded by the City of Reykjavk, and the
geoarchaeological research was funded by a SSHRCC Doctoral
Fellowship from the government of Canada, an Overseas Research
Studentship, the Cambridge Commonwealth Trust, Pelham Roberts
and Muriel Onslow Research Studentships from Newnham College,
Cambridge, and Canadian Centennial Scholarships from the Cana-
dian High Commission in London. Garar Gumundsson took the
micromorphology samples, and supervised sampling on site. The
bones were counted by Clayton Tinsley, the thin sections were made
by Julie Boreham, and Steve Boreham and his team in the Depart-
ment of Geography, University of Cambridge, provided technical
support for all of the bulk geochemical analyses that were con-
ducted by K. Milek, except for ICPeAES, which was conducted by ALS
Chemex. Our gratitude is extended to Charles French, Catherine
Hills, Peter Jordan and two anonymous reviewers for their support
and helpful comments on earlier drafts of this paper, and to skar
Gsli Sveinbjarnarson for his assistance with the gures.
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.jas.2012.10.031.
References
Amundsen, C.P., 2004. Farming and maritime resources at Mibr on Flatey in
Breiafjrur, NW Iceland. In: Housley, R.A., Coles, G. (Eds.), Atlantic Connec-
tions and Adaptations: Economies, Environments and Subsistence in the Lands
Bordering the North Atlantic. Oxbow, Oxford, pp. 203e210.
Arnalds, O., Hallmark, C.T., Wilding, L.P., 1995. Andisols from four different regions
of Iceland. Soil Sci. Soc. Am. J. 59, 161e169.
Buckland, P.C., Perry, D.W., 1989. Ectoparasites in sheep from Straborg, Iceland and
their interpretation: piss, parasites and people, a palaeoecological perspective.
Hikuin 15, 37e46.
Bullock, P., Fedoroff, N., Jongerius, A., Stoops, G., Tursina, T., Babel, U., 1985. Handbook
for Thin Section Description. Waine Research Publications, Wolverhampton.
Canti, M.G., 1999. The production and preservation of faecal spherulites: animals,
environment and taphonomy. J. Archaeol. Sci. 26, 251e258.
Canti, M.G., 2003. Aspects of the chemical and microscopic characteristics of plant
ashes found in archaeological soils. Catena 54, 339e361.
Carr, C., 1984. The nature of organization of intrasite archaeological records and
spatial analytic approaches to their investigation. Adv. Archaeol. Method Theory
7, 103e222.
Courty, M.-A., Goldberg, P., Macphail, R., 1989. Soils and Micromorphology in
Archaeology. Cambridge University Press, Cambridge.
Evans, R.T., Tylecote, R.F., 1967. Some vitried products of non-metallurgical
signicance. Bull. Hist. Metall. Group 8, 22e23.
Fenton, A., 1978. The Northern Isles: Orkney and Shetland. John Donald, Edinburgh.
Grnvold, K., skarsson, N., Johnsen, S., Clausen, H.B., Hammer, C.U., Bond, G.,
Bard, E., 1995. Express letter: ash layers from Iceland in the Greenland GRIP ice
core correlated with oceanic and land sediments. Earth Planet. Sci. Lett. 135,
149e155.
Hakbijl, T., 2002. The traditional, historical and prehistoric use of ashes as an
insecticide, with an experimental study on the insecticidal efcacy of washed
ash. Environ. Archaeol. 7, 13e22.
Hreiarsdttir, Eln ., 2005. slenskar Perlur fr Vkingald: Me Viauka um
Perlur fr Sari ldum. Unpublished MA thesis, University of Iceland.
Jochens, J., 1995. Women in Old Norse Society. Cornell University Press, Ithaca.
Johnson, D.L., Hansen, K.L., 1974. The effects of frost-heaving on objects in soils.
Plains Anthropol. 19, 81e98.
Krivogorskaya, Y., Perdikaris, S., McGovern, T.H., 2005. Fish bones and shermen:
the potential of zooarchaeology in the Westfjords. Archaeol. Isl. 4, 31e50.
LaMotta, V.M., Schiffer, M.B., 1999. Formation processes of house oor assemblages.
In: Allison, P.M. (Ed.), The Archaeology of Household Activities. Routledge,
London, pp. 19e29.
Lange, F.W., Rydberg, C.R., 1972. Abandonment and post-abandonment behavior at
a rural Central American house-site. Am. Antiq. 37, 419e439.
Macphail, R., Crowther, J., 2007. Soil micromorphology, chemistry and magnetic
susceptibility studies at Huizui (Yiluo Region, Henan Province, Northern China),
with special focus on a typical Yangshao oor sequence. Indo-Pac. Prehist.
Assoc. Bull. 27, 103e113.
Malinsky-Buller, A., Hovers, E., Marder, O., 2011. Making time: living oors,
palimpsests and site formation processes e a perspective from the open-air
Lower Paleolithic site of Revadim Quarry, Israel. J. Anthropol. Archaeol. 30,
89e101.
Matthews, W., French, C.A.I., Lawrence, T., Cutler, D.F., Jones, M.K., 1997. Micro-
stratigraphic traces of site formation processes and human activities. World
Archaeol. 29, 281e308.
Mehler, N., 2002. Artefacts from the Viking Period. In: Roberts, H.M., Snsdttir, M.,
Vsteinsson, O. (Eds.), Archaeological Investigations at Aalstrti 2001. For-
nleifastofnun slands, Reykjavik, pp. 67e76.
Milek, K.B., French, C.A.I., 2007. Soils and sediments in the settlement and harbour
at Kaupang. In: Skre, D. (Ed.), Kaupang in Skiringssal. Aarhus University Press,
Aarhus, pp. 321e360.
Milek, K., 2012. Floor formation processes and the interpretation of activity areas:
an ethnoarchaeological study of turf buildings at Thver, northeast Iceland.
J. Anthropol. Archaeol. 31, 119e137.
Misarti, N., Finney, B.P., Maschner, H., 2011. Reconstructing site organization in the
eastern Aleutian Islands, Alaska using multi-element chemical analysis of soils.
J. Archaeol. Sci. 38, 1441e1455.
Murphy, C.P., 1986. Thin Section Preparation of Soils and Sediments. AB Academic
Publishers, Berkhampstead.
Nelson, D.W., Sommers, L.E., 1996. Total carbon, organic carbon and organic matter.
In: Sparks, D.L. (Ed.), Methods of Soil Analysis Part 3: Chemical Methods. Soil
Science Society of America, Madison, pp. 961e1010.
Nielsen, A.E., 1991. Trampling the archaeological record: an experimental study. Am.
Antiq. 56, 483e503.
Nordahl, E., 1988. Reykjavk from the Archaeological Point of View. Societas
Archaeologica Upsaliensis, Uppsala.
Pierce, C., Adams, K.R., Stewart, J.D., 1998. Determining the fuel constituents of
ancient hearth ash via ICPeAES analysis. J. Archaeol. Sci. 25, 493e503.
Roberts, H.M., Snsdttir, M., Mehler, N., Vsteinsson, O., 2003. Skli fr vkingald
Reykjavk. In: rbk hins slenzka fornleifaflags 2000e2001, pp. 219e234.
Rolfsen, P., 1980. Disturbance of archaeological layers by processes in the soil. Nor.
Archaeol. Rev. 13, 110e118.
Sampietro, M.M., Vattuone, M.A., 2005. Reconstruction of activity areas at
a formative household in northwest Argentina. Geoarchaeology 20, 337e354.
 K.B. Milek, H.M. Roberts / Journal of Archaeological Science 40 (2013) 1845e1865
Schiffer, M.B., 1996. Formation Processes of the Archaeological Record. University of
Utah Press, Salt Lake City.
Shahack-Gross, R., Albert, R.-M., Gilboa, A., Nagar-Hilman, O., Sharon, I.,
Weiner, S., 2005. Geoarchaeology in an urban context: the uses of space in
a Phoenician monumental building at Tel Dor (Israel). J. Archaeol. Sci. 32,
1417e1431.
Shahack-Gross, R., Simons, A., Ambrose, S.H., 2008. Identication of pastoral sites
using stable nitrogen and carbon isotopes from bulk sediment samples: a case
study in modern and archaeological pastoral settlements in Kenya. J. Archaeol.
Sci. 35, 983e990.
Shetelig, H., Falk, H., 1937. Scandinavian Archaeology. Clarendon Press, Oxford.
Shillito, L.-M., Matthews, W., Almond, M.J., Bull, I.D., 2011. The microstratigraphy of
middens: capturing daily routine in rubbish at Neolithic atalhyk, Turkey.
Antiquity 85, 1024e1038.
Simpson, I.A., Vsteinsson, Orri, Adderley, W.P., McGovern, T.H., 2003. Fuel resource
utilisation in landscapes of settlement. J. Archaeol. Sci. 30, 1401e1420.
Smith, H., Marshall, P., Parker Pearson, M., 2001. Reconstructing house activity
areas. In: Albarella, U. (Ed.), Environmental Archaeology: Meaning and Purpose.
Kluwer Academic Publishers, Dordrecht, pp. 249e270.
Snsdttir, Mjll, 2004. Haust/Autumn 2003. In: Roberts, H.M. (Ed.), Excavations at
Aalstrti, 2003. Fornleifastofnun slands, Reykjavk, pp. 14e27.
Stead, J., 1981. The Uses of Urine, Part I. In: Old West Riding, vol. 1, pp. 12e18.
Stead, J., 1982. The Uses of Urine, Part II. In: Old West Riding, vol. 2, pp. 1e9.
Stein, J.K., 1983. Earthworm activity: a source of potential disturbance of archaeo-
logical sediments. Am. Antiq. 48, 277e289.
Stevenson, M.G., 1982. Toward an understanding of site abandonment behaviour:
evidence from historic mining camps in the southwest Yukon. J. Anthropol.
Archaeol. 1, 237e265.
Stockton, E.D., 1973. Shaws Creek Shelter: human displacement of artefacts and its
signicance. Mankind 9, 112e117.
Stoops, G., 2003. Guidelines for Analysis and Description of Soil and Regolith Thin
Sections. Soil Science Society of America, Madison.
1865
Sullivan, K.A., Kealhofer, L., 2004. Identifying activity areas in archaeological soils
from a colonial Virginia house lot using phytolith analysis and soil chemistry.
J. Archaeol. Sci. 31, 1659e1673.
Sveinbjrnsdttir, Arny E., Heinemeier, J., Gumundsson, Garar, 2004. 14C dating of
the settlement of Iceland. Radiocarbon 46, 387e394.
Tani, M., 1995. Beyond the identication of formation processes: behavioural
inference based on traces left by cultural formation processes. J. Archaeol.
Method Theory 2 (3), 231e252.
Terry, R.E., Fernndez, F.G., Parnell, J.J., Inomata, T., 2004. The story in the oors:
chemical signatures of ancient and modern Maya activities at Aguateca,
Guatemala. J. Archaeol. Sci. 31, 1237e1250.
rarinsson, S., 1987. Geology and physical geography. In: Nordal, J., Kristinsson, V.
(Eds.), Iceland 1986. Central Bank of Iceland, Reykjavik, pp. 1e9.
Tinsley, C., McGovern, T., 2002. Faunal material. In: Roberts, H., Snsdttir, Mjll,
Vsteinsson, Orri (Eds.), Archaeological Investigations in Aalstrti 2001. For-
nleifastofnun slands, Reykjavik, pp. 81e83.
Vizcano, A.S., Caabate, M., 1999. Identication of activity areas by soil phosphorus
and organic matter analysis in two rooms of the Iberian sanctuary Cerro El
Pajarillo. Geoarchaeology 14, 47e62.
Vyncke, K., Degryse, P., Vassilieva, E., Waelkens, M., 2011. Identication of domestic
functional areas. Chemical analysis of oor sediments at the Classical-Hellenistic
settlement at Dzen Tepe (SW Turkey). J. Archaeol. Sci. 38, 2274e2292.
Wandsnider, L., 1996. Describing and comparing archaeological spatial structures. J.
Archaeol. Method Th. 3, 319e384.
Wilk, R., Schiffer, M.B., 1979. The archaeology of vacant lots in Tucson, Arizona. Am.
Antiq. 44, 530e536.
Wilson, C.A., Davidson, D.A., Cresser, M.S., 2008. Multi-element soil analysis: an
assessment of its potential as an aid to archaeological interpretation.
J. Archaeol. Sci. 35, 412e424.
Zielinski, G.A., Mayewski, P.A., Meeker, L.D., Gronvald, K., Germani, M.S., Whitlow, S.,
Twickler, M.S., Taylor, K., 1997. Volcanic aerosol records and tephrochronology
of the Summit, Greenland, ice cores. J. Geophysical. Res. 102, 26,625e26,640.