Yasur-Lau Assaf. Household Archaeology in Ancient Israel and Beyond
Подождите немного. Документ загружается.
372 michael d. press
the hypothetico-deductive method that Hodder believes to be largely
illusory: whereas Hodder states that most archaeological analysis is
hermeneutic, I would suggest that Hodder’s model is merely a slightly
modied hypothetico-deductive method. Beyond recognizing that
archaeologists begin their analyses with some prior knowledge and
assumptions about the material in question, and recognizing the pos-
sibility for shis in interpretation and questions during the process of
analysis, Hodder’s model simply moves from questions to excavation
to data analysis to interpretation to consideration of other sites (1992:
Fig. 22)—in the general manner of a deductive approach.
Perhaps, then, we might acknowledge that no suggested method for
dealing with the interpretive problem of identifying cultic contexts is
purely deductive or purely inductive. Elements of both approaches are
used, consciously or not, in each archaeologist’s approach. A demon-
stration of this hybridization is provided by Holladay’s (1987) analysis
of the religion of Israel and Judah in Iron II. Holladay proposes a
model (1987: 266–269) of the types of religious groups and the types
of cultic contexts associated with each, which we might expect in the
Iron II Levant. e model is only presented, however, aer the evi-
dence from the presumed cultic contexts is investigated (252–265);
moreover, his analysis of the data and classication of these contexts
(270–274) are not based on objective criteria derived from the model
itself, but from an inductive comparison of the dierent contexts to
each other. In fact, Holladay himself states at the outset that his overall
approach in the essay (not exactly that of the analysis of cultic contexts
just described) moves between inductive and hypothetico-deductive
reasoning (1987: 251).
I believe that, to deal with a complex interpretive problem such as
identication of cultic contexts, it is best (if not necessary) to take
advantage of both deductive and inductive methods. On the one hand,
the archaeologist should use generalizing observations, based on eth-
nographic and ethnohistorical information, to set the boundaries and
general principles (to “calibrate” the evidence, as Kyriakidis [2007c:
20] has put it). On the other, the archaeologist should investigate cultic
contexts in the specic culture in question, or—if there is no sucient
evidence—those of neighboring contemporary cultures, or immedi-
ately antecedent or subsequent cultures that display some relation-
ship—as long as this latter material is used cautiously. I recognize that
a proper application of this approach would require a comprehensive
study of cultic contexts cross-culturally—something that is not feasible
a problem of definition 373
for researchers investigating questions on specic cultures, and some-
thing (to my knowledge) that has not been accomplished in anthro-
pology, although such a general study would be a valuable grounding
for particularist analyses. ese suggestions, then, are meant as general
principles to be applied only to the extent possible.
Summary of Methodological Suggestions
To summarize the methodological discussion above, there are a few
suggestions I would draw for approaching the problem of identifying
cultic contexts:
1. Explicit denitions: I do not necessarily recommend the ones sug-
gested here, but simply the principle of making clear statements of
denition.
2. Explicit analogies: It is essential to state clearly the comparisons that
serve as the basis of our interpretations, whatever those compari-
sons are.
3. A suggested approach to analogies: Beyond the suggestion of explicit
comparisons, I would further suggest that both generalizing and
particularist comparisons be used: rst cross-cultural ones, based on
ethnographic and ethnohistoric data, in order to “calibrate the evi-
dence”; then, those from the specic culture, or closely related ones.
e Problem of Domestic Cult
Having addressed some basic issues of identifying cultic contexts, we
can now move to look specically at domestic cult. e term domestic
might seem, at least supercially, to be more straightforward. Never-
theless, its precise place in the dichotomy of “ocial” vs. “popular”
religion—or of “public” vs. “private”, or of “elite” vs. “non-elite”—is in
fact not a simple one. Quite simply, these dichotomies, though oen
used interchangeably (see, for instance, Dever 2005: 5–6), are not iden-
tical; for instance, the houses of the “elite” might be loci of cult, which
would be just as “domestic” as that of “non-elite” houses. is is not
the place to review in detail the development of scholarship on ocial
and popular religion (for detailed surveys, see especially the work of
Berlinerblau [1993, 1996: 17–43; 2001]) For now, it is sucient to note
374 michael d. press
that recent literature on this topic has suggested a move away from
a simple dichotomy toward a more diverse categorization of cult.
4
It
would be best, then, to understand “domestic” cult as acts of worship
situated in the house (of whatever type of house), i.e., at the lowest
level of social organization.
While there has been a growth in interest in domestic and “popu-
lar” religion in theoretical discussions over the last few decades, few
attempts have been made to develop methods for identifying domestic
cult archaeologically. Renfrew’s extensive essay on cult identication
includes only a brief section on domestic cult practice (1985: 22–23),
presumably because his main concern was with the Phylakopi sanctu-
ary. He suggests that, while the four main aspects of religious ritual he
has identied should still be found in a domestic setting, there may be
lower degrees of “separateness” and “specialization” than those found
in communal cult, and a more modest scale. Gilmour (1995: 12, 16)
echoes these concerns. He formulates a set of ve characteristics that
set “household” cult apart from “public” cult: the context is not exclu-
sively cultic; the “cult-specic artifacts” are fewer and simpler; the
scale is smaller; continuity of cult is unlikely; and “folk” deities and
“unconventional forms” might be expected (Gilmour 1995: 16). Nei-
ther Renfrew nor Gilmour, however, provides any examples or sum-
mary of data to support these generalizations. As a result, while these
generalizations may appear sensible, they can be taken as little more
than speculative.
e lack of clear methodological frameworks is reected in the pro-
posals of specic domestic cult contexts in the southern Levant. A good
example can be seen in the spatial analysis of activity in an Iron Age
house at Tell Halif (Hardin 2004). While the presentation as a whole
4
To note a few examples briey: Albertz (most recently 2008: 94; also 1978, 1994:
esp. 19) has proposed a model with three types or levels of religious activity: state
religion, family religion, and local religion; he suggests (2008: 94) that these essen-
tially correspond to Holladay’s (1987: 268–269) model of “established worship” on
national/town and neighborhood levels, alongside “tolerated nonconformist worship.”
Wallace’s general anthropological model of four levels of cult institutions: individual,
shamanic, communal, and ecclesiastical (1966: 84–101, esp. 86–88), has been adapted
by Wright (1994: 72–73) to a hierarchy of Mycenaean sites. For Israelite religion, Zevit
(2003: 230–233) has suggested that cultic activities were associated with each of the
nested kinship groups in society—the individual, the “father’s house” (beit ʾav), the
clan, the tribe, and the people. us, recent scholarship has recast the dichotomy of
“ocial” and “popular” religion as a hierarchy of multiple levels, with persons fully
capable of participating in more than one.
a problem of definition 375
is fairly successful, the reconstruction of part of the “living room” as a
cultic area (Hardin 2004: 76–77, 79; see also Borowski 1995: 148, 151)
is questionable. e entire assemblage of “cultic” nds from this area
of the room is two large worked stones, a fenestrated stand fragment,
and a pillar gurine head. Besides the small quantity and fragmentary
nature of the nds, suggesting that they are not in primary use context,
the artifacts themselves are not unquestionably “cultic” (or otherwise
religious) in nature.
5
Similarly, the cult corner suggested by Herr in a
house at Tall al-ʿUmayri (2006) is based on meager evidence of ambig-
uous nature (such as standing stones). Gadot and Yasur-Landau, on
the other hand, suggest that domestic cult in the Megiddo Level K-4
courtyard building was spread throughout the house (2006: 591), but
they make use of the same basic arguments: in this case, supposed
“ritual” objects (gurines, stone stelae, a ceramic stand, and special
ceramic forms) were not concentrated in any specic area of the house
but found throughout.
e nds from a series of Iron Age houses at Tall Jawa (Daviau
2001) are more substantial. Schmitt (2008: 164) and Albertz (2008:
95–96) have both pointed to the impressiveness of the “domestic cult
assemblages” found here, and Schmitt states that they have caused him
to reinterpret the cultic nature of Ashdod Area D (2008: 164). Despite
the larger scale of these nds, however, the interpretation of the Jawa
material, again, ultimately rests on questionable arguments. While
Daviau tries to dene a typical Ammonite domestic cultic assem-
blage on the basis of the Jawa material, the reports of the excavations
(Daviau 2003; also 2002) do not report that the objects were found
in any clear association, but that one or two such objects (gurines,
ceramic stands, etc.) might occur in a room mixed with hundreds of
sherds of pottery, several grindstones, and other apparently utilitarian
objects. In the end, the domestic cult assemblage of Jawa is essentially
dened on the basis of the unique or unusual objects found in each
house, objects that are labeled “cultic” with little further discussion.
e above discussion points to the need for clear methods of domes-
tic cult identication. As with the identication of cultic contexts
generally, if we are to employ generalizations in any form to identify
5
For the gurines and ceramic stands sometimes having noncultic functions, see
discussion and references above. e worked stones are suggested to be masseboth, but
the shape of the stones is unusual; in addition, standing stones appear to have had a
variety of functions, including noncultic ones (Graesser 1972).
376 michael d. press
domestic cult (as advocated by Renfrew and Gilmour) we must start
with some sort of objective criteria. Ethnographic and ethnohistorical
data need to be investigated and compared in order to identify basic
problems, trends, and possibilities in domestic cult practice. As a start,
I will attempt to demonstrate how such data might be used below:
I will use one modern example and two from the ancient world. In
particular, I will pay attention to the types of issues highlighted by
Renfrew and Gilmour. For the purposes of this demonstration, their
suggestions can be stated as two basic questions:
1. Are there dedicated objects in domestic cult?
2. Are there dedicated cult places in the house?
In fact, Renfrew (1985: 22) suggests that both conditions must be met
in order to identify cult practice within the house.
1) In a modern Jewish home, a series of religious objects might
regularly be stored or on display, for example in a hutch or on shelves
in the living room: a kiddush cup, shabbat candle holders, a meno-
rah (hanukkiyah), a seder plate, etc. At the time of the appropriate
calendrical rituals, these objects are taken out and brought to other
parts of the house—most likely the dining room, but perhaps also
to the kitchen, to another part of the living room, etc.—to be used
there. Meanwhile, other rituals (especially those performed by more
observant Jews) might be performed elsewhere in the house, such as
recitation of prayers (e.g., Modeh Ani, in bed), or putting on tallit and
tellin before going to services.
From this example, we see that many rituals may have special ob-
jects associated with them, but others may not—or may not have objects
associated with them at all (such as prayer). As for a dedicated place,
this may be more problematic. e area of concentration of ritual
objects is not one of performance but of display or storage. Mean-
while, although the performance of religious rituals may tend to clus-
ter in a specic room of the house (such as the dining room), many
or even most rooms may see some religious ritual. On top of this, we
may note that even the area with the greatest concentration of ritual
performance is not dedicated only—or even primarily—to religious
ritual.
6
6
For a similar modern example from a very dierent context, namely Greece, see
Kyriakidis 2007c: 18.
a problem of definition 377
is modern example is of course merely a single case, and can hardly
be generalized. A much wider range of cultic and other religious ritual
examples would need to be studied to provide a proper range of pos-
sibilities and trends in ritual performance. Nevertheless, this example
raises several issues that can be kept in mind when considering ancient
cult and religious ritual. If this diusion of ritual throughout the house
can be true for modern societies, then what of ancient ones, where it is
commonly acknowledged that religion was not compartmentalized, but
rather a much more essential part of everyday life? On the one hand,
we might expect more abundant examples or ritual behavior; at the
same time, we might expect religious ritual activities to be even more
scattered through the house, and not concentrated in a specic area.
7
is then leads to the ancient examples:
2) One of the clearest examples of a small-scale, low-level cultic
practice in the Hebrew Bible is the baking of cakes for the Queen of
Heaven (Jer. 7: 17–18; 44: 15–19, 25). e precise setting of this rit-
ual, or set of rituals, is unclear, although it appears to center on the
house as the members of the family are invoked: the children gather
the wood, the fathers build the re, and the women knead dough for
7
Related to this problem is the identication of “cult corners” in the archaeologi-
cal record. e term cult corner or cultic corner was rst introduced into Palestinian
archaeology by Y. Shiloh in a paper on early Iron II cultic contexts (1979). Shiloh
identied a series of tenth-century cult corners in Megiddo, notable among them
Locus 2081 (1979: 149). He continued (1979: 150–151, 156) by noting the existence of
groups of artifacts (altars, ceramic stands, etc.) similar to those from Megiddo at other
sites, such as Lachish and Taʿanach (Lapp’s “Cultic Structure”), in order to identify
cult corners at the latter sites as well. e cultic nature of these contexts, however, is
by no means a consensus view. For example, Yeivin (1973) and Fowler (1984a) both
disputed the identication of the “Cultic Structure” as religious in nature, suggesting
instead that it was part of a house, with perhaps a kitchen and/or a storeroom. Even
Rast (1994), though highly critical of Fowler and Yeivin, provides an interpretation
that is not necessarily contradictory: he in fact argues that only the artifacts from the
structure are necessarily “cultic,” with the “Cultic Structure” itself (and other similar
contexts, such as Megiddo Locus 2081) being the house of a priestly family where
cultic items from a temple were stored. is interpretation is extremely signicant.
First, although Rast defends Lapp’s label of the building as a “Cultic Structure,” this
usage would stretch the meaning of the term cultic beyond real analytic value (i.e., the
context itself would not at all be a locus of “cultic” activity). In addition, even Rast
has concluded that the Taʿanach “Cultic Structure” (and Megiddo Locus 2081) is not a
“cult corner” but simply a storage area. us, some of the very denitional examples of
the concept provided by Shiloh are undermined. is is not to say that “cult corners”
do not exist in the southern Levant; rather, I only suggest that they need to be clearly
proven and not assumed to be a common feature. Gilmour, at least, has gathered some
evidence to suggest that cult corners are a fairly universal feature (1995: 15), but more
investigation of both generalizing and culture-specic data is needed.
378 michael d. press
the cakes (Jer. 7: 18). is interpretation, as centering on small fam-
ily units and the house, has in fact been put forth by several scholars
(Ackerman 1999: 27; 2003: 461–464; Zevit 2001: 554–555; Dever 2005:
190–191). It is hard to glean further information on the ritual due to
the abbreviated nature of the references. Nevertheless, the items used
in this ritual do not, for the most part, seem to be of a unique nature:
it involves, as described in Jeremiah, rewood and dough. An excep-
tion might be, if we understand the verb lĕhaʿăṣibāh in Jeremiah 44:19
to mean “to make in her image,” as is commonly done (Ackerman
1999: 26; Zevit 2001: 554; Stager 2000: 7 for further discussion and
references), the use of special molds for the dough (Ackerman 1999:
26; Stager 2000). Meanwhile, the evidence for the associated rituals
listed, pouring libations and burning incense (Jer. 7: 18; 44: 18–19,
25), is unclear: these cultic acts may or may not have been performed
with special items. As for the possible locus of performance within the
house, this again is not stated; there is no suggestion of a special loca-
tion, however, and the preparation and baking of dough would suggest
an ordinary oven and kitchen.
3) J. C. Wright (1994: 56–60) suggested modifying Kilian’s concep-
tion (1988a) of the “wanax” ideology to a “hearth-wanax” ideology:
this term encapsulates the idea that the megaron of the Mycenaean
palace, and specically its central hearth, were a major focus of cult.
He further proposed that this cult was in fact a monumental extension
of cult rituals conducted in ordinary houses at the hearth. ere is, in
fact, little direct evidence for cultic activity at these royal hearths (see
Darcque 2005: 175). Although he does not explicitly state it (perhaps
because the evidence would be well known to many of his readers),
Wright is presumably basing this reconstruction of Mycenaean rituals
on evidence from historical Greece, when there was a central public
hearth and associated state cult of Hestia (at least at Athens), which
was an extension of household cult (see, e.g., Nilsson 1972: 72–77;
Mikalson 2005: 160). If such rituals can be reconstructed for Myce-
naean Greece (a plausible, but by no means certain, suggestion; cf.
Darcque 2005: 175–177), then these rituals would not have been the
only or even primary use of these rooms. e primary function of the
hearth appears to have been for heat and light, with a secondary usage
for cooking, particularly in the main room of the house (Shear 1968,
1987: 146; Shaw 1990; Darcque 2005: 175).
ese examples, then, suggest that the identication of objects and
areas of the house connected with cult may be much less straightforward
a problem of definition 379
than oen assumed. I would use them instead to suggest an alternative
method of identication of domestic cult: while not ignoring the pos-
sibility of dedicated objects or areas of the house, this method would
focus on identifying specic cultic behaviors in the culture in ques-
tion, and then, on determining the traces (the material correlates) that
these behaviors would leave in the archaeological record. As suggested
above, this method would privilege analogies from the culture itself, or
(if necessary) from closely related cultures. e two ancient examples
above, then, would function well with the proposed application of
these methodological suggestions below, to Iron Age Philistia. ere
is little evidence, in terms of either clear archaeological data or textual
data, for Philistine domestic cult practices; as a result, we must turn to
neighboring or antecedent/subsequent cultures. For Iron I, this would
involve looking especially to Mycenaean Greece (and Cyprus), while
for Iron II, the neighboring cultures of Judah, Israel, and Phoenicia
would be most appropriate.
Application
We can now turn to the application of the methodological suggestions
above, in Iron Age Philistia. ere are, in my view, two contexts (or
sets of contexts) that can be identied, with fair certainty, as cultic: the
seventh-century temple at Ekron (Gitin 2003; Gitin et al. 1997) and
the Iron I Qasile temples (Mazar 1980). e Tel Miqne/Ekron struc-
ture is quite clearly identied as a temple complex on the basis of an
inscription (Gitin et al. 1997). e Qasile structures can be identied
as cultic above all through the concentrations of nds on and around
a raised platform (Locus 133) in Stratum X (Mazar 1980: 38–39, Fig.
11, pls. 11:3–12:3) and in Pit 125 in Stratum XI (Mazar 1980: 24–25,
Table 17).
8
ese two concentrations would seem to correspond to
behaviors such as leaving votives or other oerings around a platform/
altar, and collecting old oerings in a favissa (or bothros); the latter
is not necessarily a cultic ritual itself, but an associated behavior. To
these two contexts we can add a third, the favissa from Tel Yavneh
8
Note also that the nds of the pit include the large anthropomorphic vessel
(Mazar 1980: 78–81), which, as Mazar notes, might be connected with the terracotta
female gures known from some Mycenaean cult sites (see below).
380 michael d. press
(Zier and Kletter 2007). While simply a pit without any identied
structures associated (Zier and Kletter 2007: 7), the wealth of nds
from this context suggests it is the favissa from a presumed temple.
Beyond these examples, a number of other contexts have been sug-
gested to be cultic. At Batash, Mazar identied Locus 910a in Area E as
a cult corner (1997b: 218, 222, 259). A number of contexts at Ashdod
and Miqne have been identied as cultic, most forcefully by T. Dothan
(for recent summaries see T. Dothan 2003; 2002; Schmitt 2008; Zevit
2001: 132–142). ese range in interpretation from temples (Miqne
Buildings 350 and 351) to shrines of various sizes to public build-
ings or houses with some cultic functions (e.g., the Ashdod Area H
complex, including the “apsidal building”, from Strata XIII–XII; see
T. Dothan 2003: 200–202; M. Dothan and Ben-Shlomo 2005: 23–25).
Many of these structures, however, have been more recently reinter-
preted as residences, workshops, or secular public buildings (see esp.
Yasur-Landau 2002, Mazow 2005).
9
eir status, then, is unclear, and
calls for further analysis.
It is impossible for me to review all of these contexts in detail here.
Instead, I will choose two examples with which to test some of the
methodological suggestions: Room 16 in Miqne Field INE East Slope,
Stratum VIIA Phase 9B (Fig. 1), dating to roughly the middle of the
twelh century (Meehl, Dothan and Gitin 2006: 34–42); and Room
5032 in Ashdod Area H, Stratum XIb (Fig. 2), dating to the early elev-
enth century (M. Dothan 1971: 159–161; Dothan and Ben-Shlomo
2005: 31–34). I have chosen these contexts for two reasons. First, the
stratigraphy and architecture of both rooms and their surrounding
areas have recently been published in full. In addition, T. Dothan has
made a series of cumulative arguments for the cultic nature of several
contexts at Miqne and Ashdod, of which Room 16 and its features are
part (see, e.g., T. Dothan 2003; 2002; 1998: 155–158). Meanwhile, the
nds of gurines in contexts such as Ashdod Room 5032 have been
one of the main types of evidence adduced for Philistine domestic cult
(e.g., Schmitt 2008, esp. 166–167; Yasur-Landau 2001: 335).
1. Tel Miqne Room 16 (Fig. 1)
Room 16 is one of 10 rooms or areas (nos. 9–18) excavated in Field
INE East Slope and Sondage, Stratum VIIA Phase 9B (Meehl, Dothan,
9
See also L. Mazow in this volume.
a problem of definition 381
and Gitin 2006: 34, plans Stratum VIIA 9B4–9B1). e other areas
include ve rooms (9, 12–14, and 18) interpreted as domestic areas,
two large open activity areas (11 and 17) interpreted as industrial
areas, two rooms (10 and 16) considered to be cultic, and a passageway
(15). Rooms 12–14 and 16–18, and Passageway 15, are adjacent and
appear to be part of a single structure. Of these rooms, Room 16 is
by far the largest—it is described as “almost double the size of Rooms
13 and 14 together” (Meehl, Dothan, and Gitin 2006: 34). ere are
three main architectural features of this room, not necessarily all in
use at the same time but built and modied over the four subphases of
the room: Bench 100021, built against the west wall; Platform/Hearth
100023, built of mudbricks and paved with potsherds, and with an
associated ash layer on top; and stone Bathtub 68047. In addition,
there were several re pits in use in this room during its dierent sub-
phases, including three over the course of Phase 9B1 (Meehl, Dothan,
and Gitin 2006: 40–41). ere were few small nds on the surfaces
associated with these features; the only ones noted are a carnelian lotus
seed pendant along with some charred olive seeds in Phase 9B2.
Now we will briey turn to an evaluation of Dothan’s arguments
on the basis of the methodological suggestions. Nowhere does Dothan
dene the term cultic, leaving the interpretation of the contexts some-
what open-ended. Nevertheless, some of the specic analogies she
uses (see below) as well as her reference to “cultic praxis” in connec-
tion with Renfrew’s framework (Dothan 2003: 191), suggest that she
is primarily viewing cult as specically involved with worship. e
basis of her interpretation of Room 16 and other contexts as cultic
is the varying combinations of features (hearths, bathtubs, benches,
columns, and platforms) and artifacts (gurines, zoomorphic vessels,
bronze and iron objects, ivory objects, incised scapulae, and miniature
vessels) found in them (see Dothan 2003: esp. 210). us Room 16
described above is interpreted as cultic explicitly on the basis of its
hearth, bench, and bathtub (Meehl, Dothan, and Gitin 2006: 34–35,
41). In addition, the features of this room form part of a cumulative
inductive argument based on comparison with other contexts with
such features and material culture nds. e argument is not based
on the inherent features of these objects, however; it is based on a set
of analogies, largely explicit, with rooms containing similar features
and artifacts in the contemporary Aegean and Cyprus. ese analo-
gies involve comparisons with a series of buildings, including the pal-
aces at Mycenae, Tiryns, and Pylos (each with central hearth in the
megaron); Asine House G (the plan of large central hall anked by