Yasur-Lau Assaf. Household Archaeology in Ancient Israel and Beyond
Подождите немного. Документ загружается.
22 james w. hardin
1984; Meyers 2002a, 2003b). Further contributions to the understand-
ing of these same issues have been made by those engaged in feminist
and other gender studies, many of which are cited above.
In spite of a number of cautionary tales provided by and from
ethnographic and ethnoarchaeological data concerning inferences
made about the archaeological record (e.g., Cranstone 1971; David
1971; Bonnichsen 1973; Stanislawski 1973; Horne 1982; Kramer 1982a;
Newell, 1987), I believe that there is ample evidence demonstrating
their usefulness of these types of data and this evidence more than war-
rants their consideration for use in understanding ancient households.
is is particularly true in the Levant, where the ethnographic and eth-
noarchaeological data most oen cited derive from the same greater
geographical regions as the archaeological remains being investigated.
When approached cautiously, these data prove useful for understand-
ing many aspects of archaeological materials, including the function
of certain artifacts, the execution and organization of certain activi-
ties, and social organization at the household level. is is especially
the case when ethnographic and ethnoarchaeological data accompany
other sources of data such as historical texts. An example of the judi-
cious use of textual information and ethnography, ethnohistory, and
ethnoarchaeology is demonstrated in the work of D. Schloen (2001).
A Case Study
Schloen’s ambitious work, e House of the Father As Fact and Symbol:
Patrimonialism in Ugarit and the Ancient Near East
, presents foremost
an explanatory model for understanding ancient Near Eastern society
holistically. In this work, he addresses the existence in the wider Mid-
dle East and eastern Mediterranean region of a strong, markedly per-
sistent organization at the household level that included architectural
compounds occupied by patrilineal, patrilocal, and extended family
households. e theoretical/philosophical bases for his understand-
ing of the extended family household are the hermeneutics of Paul
Ricoeur and the patrimonialism and agency theories of Max Weber
(2001: 7–9). e primary unit of examination for this model is the
“house of the father.”
e “house of the father” identies a typically recurring and his-
torically common patrilocal, extended “Mediterranean”-type house-
hold (what Schloen calls the “joint family”). Indeed, this household
type was pervasive and appears in many dierent social and political
understanding houses 23
contexts and is manifested in various ways, even in the face of dif-
fering levels of political organization and control imposed from the
outside. For Schloen, the household’s “diverse manifestations were
developments of a coherent pattern of meaningful social action that
was universally understood and so endured for thousands of years in
the ancient Levant and its environs” (2001: 1). He understands the
household as a “demographic and economic fact” and a “powerful
political and religious symbol.” is understanding of the household
as both fact and symbol allows Schloen to move beyond functional
approaches that oen overemphasize or exclusively emphasize mate-
rial and ecological factors.
4
e ancient evidence Schloen uses to apply and illustrate his model
comes from Bronze Age Ugarit and from Israel in the subsequent Iron
Age (the latter is to be addressed by him more thoroughly in a future
second volume). He refers to the Iron Age as the “Axial Age”—an age
of empires in the Near East—when he believes societies underwent
fundamental and dramatic shis in their social and economic orga-
nization (2001: 1–2). In spite of this fundamental change, he demon-
strates through archaeological and textual evidence from both Ugaritic
and Israelite societies that the joint-family Mediterranean-type patri-
local household was recognized as the cultural norm as “outer fact and
inner symbol” (2001: 1).
Schloen further demonstrates the persistence and historical com-
monness of the patrilocal, extended Mediterranean household through
a study of ethnographic data and historical records, and he documents
this household type from a variety of periods and places along the Med-
iterranean littoral.
5
Some of his most convincing data are provided by
4
Schloen is justiably suspicious of functional approaches (e.g., Wilk and Rathje
1982) that implicate strong correlations between resource availability (land and labor)
and ecology on the one hand with household structure and composition on the other
(Schloen 2001: 117–119). Functionalist approaches oen are particularly insensitive
to human agency or culture-specic traditions that may supersede many other con-
cerns (e.g., ecology, availability of resources, etc.). ey also have been criticized for
treating household and kinship structure and organization as static (Schloen 2001:
131–133 citing and addressing Rosenfeld 1972, 1983 and Eickelman 1989). In spite of
Schloen’s and others’ caveats, much ethnographic data gathered and presented under
functionalist, theoretical frameworks continues to be useful in a number of ways. In
fact, Schloen himself uses such descriptive, functional works in his reconstructions of
Near Eastern society, as described in the main text here (2001: 133).
5
Laslett (1983) already noted the statistically preponderant existence of a typical
patrilineal complex household in the Mediterranean region.
24 james w. hardin
three spatially and temporally separate, and well-documented, prein-
dustrial examples. ese include Roman Egypt, Renaissance Tuscany,
and Ottoman Syria (Schloen 2001: 127–133).
6
He includes additional
historical data from feudal Europe (especially France and Italy), the
poleis of classical Greece, cities of the Roman Empire, Islamic cities
from Morocco to Syria dating up to the modern period, and informa-
tion provided by Muslim village ethnographies and ethnoarchaeolo-
gies. In these examples, as well as in others, he sees a similar pattern
unfolding of household organization, size, and composition.
is study by Schloen is quite comprehensive in its examination of
social organization at the household level for ancient Ugarit and Israel.
In addition to illuminating the organization and commonness of the
“house of the father,” Schloen’s work demonstrates the usefulness not
only of historical data, but also of ethnographic/ethnoarchaeological
data to archaeological investigation and interpretation at the house-
hold level. It highlights the quantity and quality of the multivariate
data available for household research in the Levant and the broader
Mediterranean littoral.
Archaeological work like Schloen’s, which focuses on a better
understanding of household organization, is becoming more typi-
cal of research in the Levant. In addition, household archaeology is
beginning to impact Levantine archaeology in many of the same ways
that it has over the last three or four decades impacted archaeology
in the New World. With regard to the rich and varied data available
for household study, it is lamentable that those of us working in the
Levant began to focus our archaeological strategies only during the last
decade or so on better understandings of household organization and
what this organization may mean for understanding past societies at
many levels of organization. e advances in understanding already
made in this rather brief period attests to the importance of household
archaeology. Already we are beginning to better understand the physi-
cal context of households, the activities in which they are involved, the
way these activities as well as the households are organized, and what
6
For Roman Egypt, Schloen (2001) uses Bagnall’s and Frier’s (1994) analysis of
data from a number of cities and towns, but especially written data from Hermopolis,
Oxyrhynchos, and Panopolis. For Renaissance Tuscany, he cites the analysis of Her-
lihy and Klapisch-Zuber (1985) that uses data from the Florentine Census of 1427 to
address household composition. A number of sources for Ottoman Syria are cited,
including Establet and Pascual (1994) and Marcus (1989) for Aleppo and Damascus.
understanding houses 25
this organization may inform regarding other aspects of past societ-
ies, like gender. e use of multivariate sources of data, which we are
fortunate to possess, can only increase the strength of our inferen-
tial strategies as we approach a varied and rich archaeological record.
e power of these inferences also increases with the adoption of new
methodological strategies for the collection of archaeological materi-
als from domestic contexts and from recent developments borrowed
from the hard sciences. ese developments make possible the better
understanding of activity areas and domestic space and notably include
studies of oors through the analysis of microdebris, phytoliths, and
soil chemistry that now accompany the similar but more traditional
studies of larger artifacts. Household archaeology in the Levant has
much to oer—not only for better understanding households locally,
but also for testing our archaeological inferences generally.
HOUSEHOLD ARCHAEOLOGY IN ISRAEL: LOOKING INTO
THE MICROSCOPIC RECORD
1
Ruth Shahack-Gross
Introduction
Household archaeology is an area of research that is strongly inu-
enced by anthropology, and it is quite well developed in North Amer-
ican scholarship, where it relies heavily on ethnographic analogies
as a basis for archaeological interpretation. To date, archaeologists
view households as systems of membership (Allison 1999a), thus the
goals of household archaeology research as set forward by American
anthropologists include the reconstruction of such things as archi-
tecture, activities, economy, human interrelationships, attitudes and
traditions, household composition (e.g., number of occupants, ages,
genders), status, changes in socioeconomic organization, relationship
of the household to the environment (both ecological and social), pro-
duction vs. consumption, and ritual/symbolism. Allison (1999a and
references therein) demonstrates that, with the advent of ethnographic
and ethnoarchaeological studies, came the realization of the complex-
ity of households and their identication. us, Allison (1999a) rightly
suggested that two steps of analysis should be conducted:
1. Households should be identied based on architectural parameters
and the activities within them identied using material culture
remains and pictorial and textual data.
2. All other goals of research (i.e., socioeconomy, symbolism, organi-
zation, etc.) might be inferred from studies conducted in step 1.
1
I thank the organizers of the household archaeology session for inviting me to
participate in it. e case studies presented here are the result of teamwork between
the Kimmel Center for Archaeological Science and the Tel Dor and Megiddo excava-
tions. I am grateful to all team members.
28 ruth shahack-gross
Clearly, these two steps point to a gap in our ability to reliably and
condently reconstruct households in antiquity, with the results of
step 1 being more reliable as they are based on actual nds, and the
results of step 2 being less reliable because they include more assump-
tions. However, even step 1 involves a number of assumptions and
some degree of uncertainty. ese assumptions start in the eld with
the excavator’s interpretation of features, their temporal associations,
and the location of oors.
In order to study the material remains on a house oor, the oor
itself rst needs to be identied. In most Levantine sites, where build-
ing with mudbricks was common, oors are rarely paved or plastered,
and unequivocal identication of dirt oors is an incredibly dicult
task. Oen only artifacts that are assumed to be associated with what
is presumed to be a oor will be studied; this introduces one type
of uncertainty into the process of stage 1 analysis. A second type of
uncertainty introduced into the analysis is due to the fragmentary or
selective nature of archaeological assemblages that is caused by natu-
ral dierential preservation as well as by methods of excavation (e.g.,
hand-picked vs. sieved materials, pore size of sieves used in the eld,
and selection of indicative sherds over nonindicative ones). ese two
types of uncertainty, the identication of oors and the fragmentary/
selective nature of assemblages, might therefore be regarded as leading
to a cumulative analytical error.
As in every scientic method, analytical errors may be reduced
through the use of new approaches. In order to help reduce them in
household archaeology studies, the oors in the studied household
must be identied unequivocally, and complete assemblages of artifacts
and/or ecofacts should be studied. is can be carried out through the
random and nonselective sampling of sediments and artifacts. Study-
ing complete assemblages on the macroscopic and microscopic levels
is not always practical due to the large numbers of nds from historic
contexts. In this article I argue that the microscopic study of tell sedi-
ments helps to reduce the uncertainty of step 1 analysis in household
archaeology research because whole sediment samples are studied
without selection of materials based on size (i.e., no sieving) or shape
(i.e., how indicative a material is). e sampling method is random,
and bias is thus negligible. e number of samples should be relatively
large with some overlap in sampling certain contexts, so the repro-
ducibility of the method can be quantitatively evaluated. For example,
Albert et al. (2008) conducted phytolith analyses at Tel Dor studying
household archaeology in israel 29
several pairs of the same sediment, and they were able to determine
the uncertainty inherent in their analysis, which was found to be in
the order of only a few percentage points.
e identication of activity areas is one of the stated goals of
household archaeology. In the Near East, spatial analyses have been
conducted using macroscopic artifacts and installations (e.g., Gadot
and Yasur-Landau 2006), sieved microartifacts (e.g., Rosen 1989,
1993), micromorphology (Matthews et al. 1997; Matthews 2005), and
phytolith analysis (Albert et al. 2008). Elemental analyses (e.g., phos-
phate analysis) are routinely carried out in Central America (Terry
et al. 2004) and have recently been conducted in Israel at the Tell es-
Ṣa/Gath excavations. To my knowledge, at this point in time there
has not been a study in the eld of household archaeology in which
all approaches—macroscopic, microscopic, paleobotanical, molecular
and elemental—have been integrated in order to understand activity
areas.
Many of the approaches mentioned above for the identication of
activity areas are related to the subdiscipline known as “geoarchaeol-
ogy,” in which there is strong reliance on the microscopic record. Geo-
archaeological methods, especially micromorphology, are well suited
to identifying the exact location of oors, which is a necessary rst
step in locating activity areas. While lime-plastered oors might be
identied with a relatively high degree of certainty (but, see below for
an example of an erroneous identication), dirt oors, as mentioned
above, are very dicult to identify. Oen decisions about the loca-
tion of a dirt oor are based on eld observations of the relationship
between surfaces and walls and the foundations of built installations
such as basins and tabuns or large storage vessels. ese identications
are, however, not straightforward, as certain installations may have
been placed below or above the actual oor level. Because the weight
load of people and/or animals on activity surfaces results in compaction
and the development of characteristic elongated, subhorizontal, micro-
scopic voids (Goldberg and Whitbread 1993; Gé et al. 1993; see also
Shahack-Gross et al. 2003, for compaction under livestock enclosures),
dirt oors are best identied using micromorphological methods.
Further complicating the identication of activity areas is, once
oors are identied, the assumption that remains of activities on
these oors represent the use-life of the household. However, ethno-
archaeological and experimental studies caution that artifacts repre-
senting primary activities on oors are rare. For example, Hayden and
30 ruth shahack-gross
Cannon (1983) showed that macroscopic artifacts on oors may rep-
resent curated items kept as provisional discard. Ziadeh-Seely (1999)
showed that macroscopic artifacts in abandoned structures represent
post-abandonment activities, mostly storage and livestock keeping.
ese studies were recently supported from the point of view of the
microscopic record by Tsartsidou et al. (2008), who demonstrated
that phytoliths from abandoned structures primarily represented post-
abandonment storage and livestock keeping activities. erefore, it
is very important to be aware that most artifacts on oors, whether
macroscopic or microscopic, probably derive from secondary post-
abandonment activities unless other evidence is provided to show that
the oor remains are the result of primary deposition.
In this paper, I present two examples of studies that demonstrate
the usefulness of geoarchaeological studies for understanding archaeo-
logical households. e rst, originally published by Shahack-Gross
et al. in 2005, shows how micromorphology established the true loca-
tion of oors in a monumental Iron Age building at Tel Dor. e sec-
ond, originally published by Shahack-Gross et al. in 2009, shows how
activity areas outside houses shed light on activities within houses.
Example 1: Identifying Floors: An Example from Iron Age Tel Dor
Field archaeologists identify oors by “feel”; that is to say, “oors” are
oen surfaces that peel o as they are excavated. Shahack-Gross et al.
(2005), however, showed that, in certain cases, such surfaces may in
fact be only the upper level of the residue of activity remains. Such was
the case in a room excavated in the 1990s inside the so-called Monu-
mental Building at Iron Age Tel Dor. In this building, white surfaces,
which formed undulating layers between the room walls, were identi-
ed as plaster oors as they peeled o during excavation. e eld
archaeologists, however, were puzzled, as these “oors” were empty
of nds and no in situ macroscopic pottery was discovered on these
surfaces.
In the early 2000s, geoarchaeological research was conducted at Tel
Dor by the Kimmel Center for Archaeological Science. During the
course of this research, it became apparent that the white “plaster”
surfaces in the Monumental Building were composed primarily of
plant phytoliths (Shahack-Gross et al. 2005). ese are microscopic
bodies that form in plants while they are alive and remain as fossils
household archaeology in israel 31
aer the plant organic matter degrades. A detailed study of the white
surfaces in the building revealed that they were composed of over 90%
grass phytoliths. In addition to phytoliths, the white layers included
other mineral remains, such as sea sand from the nearby Mediterra-
nean beach, calcium-phosphate nodules that indicate an in situ decay
of organic matter in antiquity, and calcitic dung spherulites, which are
microscopic spheres that form in the guts of animals (mostly in rumi-
nants) and are excreted with their dung. It was therefore concluded
that the white layers are not lime plaster oors, but the remains of
livestock dung (Shahack-Gross et al. 2005).
Two possible interpretations have been suggested to explain the
presence of livestock dung in the building: the dung was used as oor-
ing material (i.e., dung plaster) or the room was used as a livestock
pen. e Tel Dor sediments included high quantities of sand that is
not easily compacted under load, thus, typical elongated voids did not
develop to their fullest and were dicult to identify. is initially sug-
gested that the white material might be the remains of dung plaster
that could have been used as a oor. To examine this possibility, dung
plasters from two ethnographic contexts were examined micromor-
phologically (see data in Shahack-Gross et al. 2005) and it became
apparent that such plasters are a mixture of soil and dung dominated
by the soil fraction. is was not the case at Tel Dor, however, and
the possibility that the white layers in the building represent a pur-
posely built oor is highly unlikely. On the other hand, supporting
evidence for the second possibility, that the white layers represent
activity remains, was found in the form of compression elongated
voids in the dung material. ese, along with microlamination of the
phytoliths throughout the layer, indicate that the dung accumulated
slowly during activities (based on Shahack-Gross et al. 2003; Macphail
et al. 2004). us, the inevitable conclusion was that livestock were
housed in the room under investigation in the Iron Age Monumental
Building at Tel Dor and that the white layers represent residues of
activity. e possibility that the dung remains accumulated while the
building was abandoned was rejected on the grounds of continuity
in the ceramic assemblage associated with the oors and lls of the
studied room, and on the repeated nature of the oor-ll sequences
(see details in Shahack-Gross et al. 2005). e recognition that the
white surfaces do not actually represent oors therefore implied that
the true oor is the surface on which the livestock was kept, i.e., an
earthen oor located below the white layer. e results of the study by