Читать книгу Botanical Aspects of Environment and Economy at Gordion, Turkey - Naomi F. Miller - Страница 10

1

Archaeological Background

The archaeological site of Gordion is most famous as the home of the Phrygian king Midas and as the place where Alexander the Great cut the Gordian knot on his way to conquer Asia. Located in central Anatolia near the confluence of the Porsuk and Sakarya rivers, Gordion also lies on historic trade routes between east and west, as well as north to the Black Sea. Very favorably situated for long-distance trade, Gordion’s setting is marginal for cultivation, but well-suited to pastoral production. It is therefore not surprising that with the exception of a single Chalcolithic site (Kealhofer 2005), the earliest settlements in the region are fairly late—they date to the Early Bronze Age (late 3rd millennium BC). The earliest known levels of Gordion, too, date to the Early Bronze Age, and occupation of at least some part of the site was nearly continuous through at least Roman times (second half of the 1st century BC); a Medieval settlement is also attested (Voigt 2005). Pre-Chalcolithic occupation in this part of the Sakarya valley is evidenced by abraded Late Paleolithic flint tools that erode out of Pleistocene conglomerates and occasionally turn up in flotation samples and other excavated sediments.

The most prominent sites in the archaeological region in which Gordion lies are Gordion itself and at least 100 Phrygian-period burial mounds. Archaeological surveys have recorded sites mostly dating between the Early Bronze Age and the modern era (Kealhofer 2005). Gordion is comprised of the 13-ha Yassihöyük (literally, “flat mound”), also referred to as the Citadel Mound (and in some earlier project publications as the City Mound), which is surrounded by an inner town and fortification system (Küçük Höyük and Kuş Tepe) that encloses an area of 51 ha. By the mid-1st millennium BC, settlement had expanded to an extensive outer town with an estimated total settlement area of about 1.5 square km (Ben Marsh, e-mail 9/16/08). The plant remains discussed in this report all come from excavations in the eastern part of the Citadel Mound.

Gordion is known through both history and archaeology. The best-known ancient references to Phrygian Gordion and its king Midas are found in Herodotus’s Histories. Other ancient references, mostly Greek, occur in the works of Xenophon, Arrian, Plutarch, and Livy. Modern archaeological interest in Gordion came through Classicists’ knowledge of ancient Greek contact with the Phrygian world. The ancient settlement mound was identified as Gordion and excavated in 1901 by Gustav and Alfred Körte (Körte and Körte 1904; Sams 2005:10). A University of Pennsylvania team led by Rodney S. Young, a professor of Classical Archaeology, began excavations in 1950.

Young’s excavations (1950–1974) focused on the Early Phrygian levels at Gordion and Middle Phrygian burial mounds. This work established a rough chronological framework for the region. Analysis and conservation continued under the direction of Keith DeVries after Young’s death in 1974. Fieldwork, however, was suspended until 1987, when a small team from the University of Pennsylvania Museum assessed the possibilities for a new project. In cooperation with the University of North Carolina, Chapel Hill, the Penn Museum renewed excavation under the direction of Mary M. Voigt in 1988. (For the history of the excavations, see DeVries et al. n.d.; Sams 2005; Voigt 2005.) Voigt established a stratigraphic sequence for the site based on the excavations of 1988 and 1989. Paleoethnobotanical research is an integral part of the renewed program of excavation and surface survey at Gordion. Since the 1990s, extensive excavation of Phrygian and later deposits has been carried out. Analysis of those archaeobotanical remains has begun (Marston 2003, 2010; Miller 2007).

Charred plant remains from Gordion provide the best evidence for tracing long-term changes in vegetation and plant use that in turn reflect many aspects of ancient economy and society in the Sakarya basin over several millennia. That the remains originate from a single site is an important limitation for a study that seeks to understand regional trends. Nevertheless, many specific questions can be addressed with these data concerning the nature of the original vegetation, the relationship between agriculture and pastoral production, irrigation, and ethnic markers. This report deals with archaeobotanical remains dating from the Late Bronze Age to the Medieval period that were excavated during the 1988 and 1989 seasons at Gordion. The assemblage consists of charcoal hand-picked during excavation and charred seed and wood remains obtained by the flotation of systematically collected soil samples. In subsequent years, I conducted informal botanical surveys in the region and collected voucher specimens and comparative material that are currently housed in my laboratory at the University of Pennsylvania Museum, Philadelphia. This work has informed both the identifications and interpretations presented here.

Stratigraphy and Chronology

The excavations of 1988/89 were limited to the Yassihöyük Citadel Mound. Young’s work had exposed the royal precinct—or at least elite quarter—of the Early Phrygian period, about 5 m below the modern surface; the excavated area covers about 2.5 ha (Voigt and Henrickson 2000a:39). To minimize the excavation area needed to obtain a stratigraphic sequence, Voigt set the upper excavation units (Operations 1, 2, and 7) at the edge of Young’s main excavation area. The uppermost level included one of Young’s sherd dumps and then descended from Medieval deposits down to the Early Phrygian royal precinct. Physically but not stratigraphically discontinuous, the lower units (Operations 3–6; 8–11; 14 [below 3–6]) were placed in an Early Phrygian courtyard area and the architectural remains from that level are preserved for touristic purposes (Voigt 2005; Fig. 1.1). The excavation of the lower trenches extended to a small area of Middle Bronze Age date. The project used a lot and locus system for excavation, recording, and analysis. In particular, a lot represents a contiguous unit of excavated earth, ideally from a single depositional stratum; it is the basic unit of excavation. A locus is comprised of one or more contiguous lots that ideally represent a “significant stratigraphic unit.” Lots and loci may also be arbitrarily defined (for example, in exploratory trenches).

Voigt has discussed the stratigraphy and the cultural and historical associations in detail (1994). She developed the YHSS numbering system—a shorthand representation of the stratigraphic analysis—to aid in the recording and sorting of the various data classes generated by the project. Re-analysis of artifacts and stratigraphy, along with new radiocarbon dates of short-lived seeds, led to a major revision of the chronology of the first millennium BC levels (DeVries et al. 2003; Voigt 2005) that is used in this volume.

The Yassihöyük Stratigraphic Sequence and Characteristics of Deposits Sampled for Botanical Remains

The Yassihöyük Stratigraphic Sequence (YHSS) assigns the excavated deposits to broad chronostratigraphic units that roughly correspond to more traditional archaeological periods (Voigt 1996). Numbered one to ten from top to bottom (Table 1.1), each of these large units is divided into a series of stratigraphic contexts defined with a minimum three-digit code (thus, deposits within YHSS 7 are assigned a number between 700 and 799). Decimal places are added as the complexity and understanding of the deposits warrant (thus, 725 is a floor deposit of a burned building in YHSS 7, and 725.04 is an oven within that building). The Early Phrygian Destruction level (YHSS 6A) is at the base of the upper trenches and the top of the lower ones. The discussion here emphasizes the time periods for which there is substantial archaeobotanical data.

MIDDLE BRONZE AGE (YHSS 10). These deposits pre-date 1500 BC. A single deposit of less than 1 cubic m volume was sampled; two samples from an erosion surface were analyzed.

LATE BRONZE AGE (YHSS 8/9), ca. 1500–12th century BC. Initially YHSS 9 was assigned to the Early Hittite Empire period; the excavated area (and flotation samples taken) consisted primarily of lensed trash and some exterior surfaces; there were no structures. YHSS 8 was assigned to the Late Hittite Empire. The only structure was single-room CBH, a stone-lined cellar with no internal features. Samples analyzed from this phase are mainly from pits, a hearth, and floor deposits. According to Voigt (1996), samples from YHSS 8 and 9 can be grouped for comparisons with the Early Iron Age and later deposits, since there is no break in the cultural sequence at this time.

Fig. 1.1 1988/1989 excavation units. Top: Early Phrygian Destruction Level (plan) (source: Gordion archive). Bottom: Three burnt buildings discussed in this volume.

Table 1.1. Yassıhöyük stratigraphic sequence, approximate dates (source: Voigt 2005:27)

YHSS 1	Medieval	13–14th century AD
YHSS 2	Roman [not in these samples]	early 1st–5th century AD
YHSS 3	Hellenistic	330–mid-2nd century BC
YHSS 4	Late Phrygian	540–330 BC
YHSS 5	Middle Phrygian	800–540 BC
YHSS 6A	Early Phrygian (“Destruction Level”)	900–800 BC
YHSS 6B	Early Phrygian (courtyards)	950–900 BC
YHSS 7	Early Iron Age	12th century–950 BC
YHSS 8/9	Late Bronze Age	1500–12th century BC
YHSS10	Middle Bronze Age	2000–1500 BC
—	Early Bronze Age [not in these samples]	2500–2000 BC

EARLY IRON AGE (YHSS 7), ca. 12th century–950 BC. The Early Iron Age (EIA) deposits are treated as one chronological unit, though they can be put in three stratigraphic groups. Samples from the earliest EIA, YHSS stratum numbers 730 and higher, come from various features (ovens, pits) associated with domestic structures and activities. A burnt reed structure (BRH, stratum number 725) is roughly in the middle of the deposits assigned to YHSS 7. Due to the in situ charring, the floated material is not comparable to ordinary occupation debris and so is listed and treated separately in this report. The most recent samples from YHSS 7 are mostly from wash and later Early Iron Age pits (705).

EARLY PHRYGIAN PERIOD COURTYARDS (YHSS 6B), 950–900 BC. The distinct stratigraphic break between YHSS 7 and 6 signals a change in function from ordinary domestic to elite quarters. YHSS 6B, which consists primarily of clay fills and construction debris, yielded very few botanical remains.

EARLY PHRYGIAN DESTRUCTION LEVEL (YHSS 6A), ca. 900–800 BC. On a grander scale, the buildings of the Destruction Level suffered the fate of the burnt reed structure 725. Similarly, the charred construction debris and in situ room contents are not comparable to ordinary occupation debris and are treated separately in this analysis. The deposits analyzed here come from the anteroom of Terrace Building 2. Broadly, there is clear stratigraphic continuity between YHSS 6B and 6A, but the YHSS 6B deposits excavated in 1988/1989 are in the center of the old excavation, and the YHSS 6A deposits are at its edge.

MIDDLE PHRYGIAN (YHSS 5), ca. 800–540 BC. Soon after the fire in YHSS 6A, the center of the Citadel Mound was leveled and covered with a thick (4 to 6 m) layer of clay (Voigt 2007). This phase is poorly represented in the stratigraphic sounding, so only a few samples were taken, mostly from post-occupation deposits within the cellar of Middle Phrygian building I:2 and a few later pits. This makes generalizations difficult.

LATE PHRYGIAN (YHSS 4), ca. 540–330 BC. Thanks to a large number of trash-filled pits in the excavated area, many flotation samples yielding quite a bit of material were taken. There are also a few samples from hearths. However, remains of structures were fragmentary due to Hellenistic stone-robbing. Despite the fairly small exposure the neighborhood can be characterized as an “industrial” area (Voigt 1996).

HELLENISTIC (YHSS 3), ca. 330–mid-to-late 2nd century BC. Two phases have been distinguished, YHSS 3B, ca. 330–mid-3rd century BC, and YHSS 3A, mid-3rd–mid-2nd century BC. The industrial nature of the excavated area continues in the lower part of this stratum (YHSS 3B), and most of the samples come from a series of hearths. During YHSS 3A, Galations (European Celts) arrived at Gordion. A burned structure, part of the Galatian “Abandoned Village” of YHSS 3A, lies above 3B. Functionally, these flotation samples are most usefully compared to those of the YHSS 7 BRH structure and Terrace Building 2A of the YHSS 6 Early Phrygian Destruction Level. A few Late Hellenistic (YHSS 3A) pits and wall fragments lie above.

MEDIEVAL (YHSS 1), 13th–14th century AD. Voigt (1994) reserved YHSS 2 for Roman period deposits; in the 1988/1989 excavation area, however, there is a stratigraphic gap. Roman material has been excavated in the northwestern zone of the Citadel Mound elsewhere on the site recently (see Goldman 2005; Marston 2010; Miller 2007a, 2007b). The few Medieval samples in the YHSS sounding come primarily from a few pits and an oven.

Yassihöyük Stratigraphic Sequence in Cultural Context

All archaeological periods were important for the people living in them, but some stand out thanks to the breadth and depth of present-day knowledge of the time. Texts—those that were never lost as well as those known only from excavation—are an independent source of information against which one can compare the archaeological materials. Also, members of the Gordion team, working with excavation, survey, archival, and other data, continue to refine our understanding of the sequence.

BRONZE AGE SETTLEMENT. In addition to Gordion, there are a few Early Bronze Age sites within a 10-km radius of the site. During the Middle and especially the Late Bronze Age in the region, Gordion was in the orbit of the Hittite empire (Voigt 1994:276). Despite the uncertain environment for farming, there were many settlements within 15 km of the site (Kealhofer 2005). Perhaps integration into the Hittite economy allowed people simply to move away or to trade for foodstuffs in bad years, or the local adaptation would have encouraged putting more effort into herding to see people through hard times.

THE PHRYGIAN QUESTION. In line with Herodotus’s and Strabo’s writings, the Phrygians are thought to have originated in southeastern Europe (Sams 1988; Voigt and Henrickson 2000a, 2000b). Keith Devries (2000:18) has mapped the plausible extent of Phrygia (at least 7th to 4th centuries BC) in west central Anatolia through rock inscriptions in the Phrygian language and other epigraphic finds. Sometime after Hittite (YHSS 9–8) ceramic evidence of connections with the Sakarya valley ceases and before the establishment of the royal precinct (YHSS 6), Phrygians had settled at Gordion. Voigt (1994:277) sees a stratigraphic break between YHSS 8/9 and 7, along with a suite of cultural changes. Voigt and Henrickson (2000:46) have argued the changes reflect the arrival of a new group of people, the Phrygians. For example, a possible ceramic marker is the Early Iron Age handmade pottery characteristic of early YHSS 7, which replaced the wheel-made Hittite ceramics; among other possibilities, at the very least this would indicate a change in ceramic production and distribution (Henrickson 1993). Despite the apparent continuity in settlement between the earlier and later YHSS 7 Early Iron Age deposits, the pottery at the end of the phase is again wheel-made, and indeed, is indistinguishable from that of Early Phrygian YHSS 6B.

THE EARLY PHRYGIAN DESTRUCTION LEVEL (YHSS 6B). Rodney Young’s major investigation of the Yassihöyük mound stopped at the Early Phrygian royal precinct. The area exposed by his excavation included presumed royal residences in the center and at the edge a series of attached megarons (Terrace Buildings 1–8), the back walls of which presented a single face to the central area. These buildings appear to have functioned as service buildings for the elite quarter. The buildings had been destroyed in a catastrophic fire, now dated to about 800 BC. Though no skeletons were found, the fire was so intense it melted pottery and vitrified the silicates in some of the wood and seeds. Young and others associated the fire with the Kimmerian invasion mentioned by ancient authors, but even before the current re-dating to 800 BC, that view was not tenable (Voigt 2007). Rather, the Early Phrygian rulers had begun a major revamping of the fortification system at the time of the catastrophic, but accidental conflagration. A clear stratigraphic break, the Destruction Level is culturally continuous with what lies above.

MIDDLE PHRYGIAN REBUILDING (YHSS 5). One of the most mysterious aspects of Gordion is the clay layer that seals the Destruction Level. Over much of the excavated area, the buildings built into the clay layer follow the general lines of the earlier, now buried, structures. It is therefore not surprising that “The YHSS 5 (Middle Phrygian) ceramic assemblage is clearly derived from that of YHSS 6 (Early Phrygian) both typologically and technologically” (Henrickson 1993:132). Henrickson remarks that this assemblage is restricted to local types. It is during the Middle Phrygian period that the settlement expanded considerably. Excavation and surface survey suggest relatively dense occupation over an area of approximately 160 ha (Voigt and Henrickson 2000a); the Citadel Mound lay in the middle of a settlement whose maximum extent was about 1 km north-south and 2 km east-west. This archaeological evidence for expanded settlement, the massive earth-moving and reconstruction of the palace quarter, continued tumulus building, and a plethora of imported wares suggest it was a very prosperous time (DeVries 2005; Henrickson 1993:140; Voigt 2005). Regional survey, too, suggests the Middle Phrygian was a time of prosperity and agricultural expansion (Kealhofer 2005). This conclusion is fully consistent with what one might expect, given the textual evidence for the power and expansion of Gordion during the 8th century.

The outstanding feature of the Middle Phrygian and subsequent landscapes was the burial tumuli that dot the countryside, especially Tumulus MM (“Midas Mound”) and the cluster nearby. About a hundred tumuli have been mapped. They are distributed within a 10-km radius of the Citadel. Tumulus building in the region ended by the 2nd century BC.

LATE PHRYGIAN ECONOMIC EXPANSION (YHSS 4). The Late Phrygian phase at Gordion is the time of the Persian/Achaemenid conquest. Gordion’s political importance probably had waned, but it appears to have been a prosperous economic center; most of the Greek pottery comes from these deposits, demonstrating contact with the west, as well (Henrickson 1993; Voigt 1994).

HELLENIZATION AND THE MEETING OF PEOPLES. In 333 BC, Alexander the Great arrived at Gordion and incorporated the area into his emerging empire. In the Early Hellenistic (YHSS 3B) ceramic assemblage, “the adoption of Greek forms becomes even more pervasive, affecting even basic types like cooking pots” (Henrickson 1993:155). Around 250 BC, finds, both spectacular and quotidian, confirm the Celtic (Galatian) occupation at Gordion attested by ancient texts (Dandoy et al. 2002; Voigt 2003).

MEDIEVAL (YHSS 1). During the Medieval period new cultural interactions might have had some affect on land use. In the case of Gordion, there is enough pig bone to suggest the presence of a resident non-Muslim population. We might expect that the influx of Central Asian Turkic tribes and political unification of new regions under Islam to have influenced trade networks and the material, including plants, that traveled along the routes.

Archaeobotanical Questions

The previous sections give some general archaeological and cultural background. Samples from the stratigraphic excavation contain a record of close to 2000 years. Data from plant macroremains, charred wood, seeds, and other plant parts can address a number of issues concerning ancient plant use, land use, and landscape. The long sequence allows us to trace vegetation history in the region and evaluate the extent and nature of human impact. Charred wood indirectly provides evidence of forest composition, and the remains themselves come from fuel and construction. From the seeds of cultigens and wild plants we can infer the relative importance of agriculture and pastoralism over time. Somewhat more directly, the charred remains leave evidence of crop choice. The intensity of land use for agricultural and pastoral pursuits would have varied, too. In conjunction with the other archaeological interpretations, the botanical data can enrich our understanding of agriculture and economy in the Sakarya valley.

Original Vegetation, Climate, and Changes in Land Use Intensity

Even today, and certainly in antiquity, climate is one of the major determinants of vegetation. For central Anatolia, Aytuğ (1970) proposes a landscape of anthropogenic steppe, certainly around Ankara but even around Gordion. As Walter points out, “Die Grenze zwischen Wald und Steppe wird in Zentralanatolien noch dadurch kompliziert, daß dieses Land keine Hochebene im eigentlichen Sinne darstellt. Vielmehr wechseln weite Beckenlandschaften (als ‘ova’ bezeichnet) mit Gebirgsrücken ab. Auf den höheren Erhebungen findet man noch Waldreste, während die tiefer liegenden Teile baumlos sind” (1956:97). [The boundary between forest and steppe in central Anatolia is complex, as this land is not a plateau in the proper sense. Basin landscapes, called ‘ova,’ alternate with mountain ridges. On the higher slopes one finds relict woodland, while the low-lying parts are treeless.] He uses an analogy between Ankara and Salt Lake City to conclude that the natural vegetation would be grassy steppe. At least in the United States, comparable Artemisia steppe occurs in Nevada, with less than 300 mm (winter) rainfall. Around Ankara, in a fenced area, Walter saw perennial grasses, including various Stipa, Bromus tomentellus, B. erectus, Festuca sulcata, Phleum sp., Melica sp., and other plants. He therefore suggests, at least for Ankara, an original Stipa–Bromus tomentellus steppe, and similar vegetation along the route to Eskişehir. Artemisia fragrans grows at the same elevation range.

Two types of natural vegetation characterize the central Anatolian steppe: perennial grassland and Artemisia steppe. Botanists have argued about whether the Artemisia steppe is disturbed grassland or original vegetation cover (Walter 1956:98). I think it likely that around Gordion, whose elevation is so close to the steppe-forest boundary, relatively favorable conditions prevailed, allowing a dense grass cover that could have supported grazing animals, presumably wild in the distant past, but herds of domestic sheep and goat by the Middle Bronze Age. Note that Marsh (2005:168) found “typical grassland soils” in the Sakarya valley below later erosion deposits.

Catchment

Regional surveys and excavation at the site of Gordion give some evidence of population and land use in the Sakarya valley over the archaeological sequence. In different periods, the area from which food supplies were drawn would have expanded or contracted according to population levels and exchange relations (political, social, or economic) with people beyond the valley. With few exceptions, most of the food plants could easily have been produced locally; what cannot be determined from the remains is whether they actually were.

Irrigation

A variety of evidence can potentially bear on the question of whether or not crops were irrigated. The first thing to consider is whether it would have been desirable and possible to irrigate. Given the erratic nature of the climate, any technique that would even out harvests from year to year would be a good thing, especially in those time periods, such as the Middle Phrygian, when there was a relatively high population density. Since the late 1950s when the Sakarya was straightened, the river has been down-cutting the plain, and irrigation requires the use of pumps. Aerial photographs from the 1950s show a very different meandering river regime, but the annual flooding of the first half of the 20th century may itself be a relatively recent phenomenon, post-dating the archaeological deposits (Marsh 2005).

Several types of botanical evidence can address the question, but not all are relevant to the data currently available from Gordion.

1. Weed seeds of irrigated and unirrigated fields. Due to the unfortunate (for the archaeobotanist) practice of suppressing weed growth in the fields, I am unable to evaluate the ancient seed assemblage through comparison with the modern field weed composition. The evidence of the sedges, however, supports the view that habitats available for grazing were not constant, and that irrigation was most significant during the Middle Phrygian and Medieval periods.

2. Crop choice. Some crops would have been irrigated because they are summer-grown (millets, and in the Medieval samples, cotton and rice). The samples from the 1988/1989 excavation have few millets and show a suggestive association with the sedge seeds in the Medieval period, but not earlier. If wheat and barley were irrigated, one might expect some association with seeds of wet areas. Namely, in a situation (including the present) where both are cultivated, wheat is more likely than barley to be irrigated because it is less drought resistant and, favored as food, is the more valuable crop. Similarly, six-row barley is more likely to be irrigated than the two-row type. The notable stability in the proportion of wheat to barley reveals no identifiable change in irrigation practices of the two major cereals (wheat or barley).

3. Measurements of cereal grains. As discussed above, there is a similar lack of positive evidence for changes in irrigation practices based on the plumpness of the wheat and barley grains.

Ben Marsh (in Voigt and Young [1999:n. 6]) has suggested that clean clay used to cap the Early Phrygian level on the Citadel Mound may have consisted of sediments originating from “hydraulic work at the time of the reconstruction (e.g., digging irrigation canals or drainage ditches)”; both activities, especially the former, support an interpretation that land use for agriculture intensified. It may be no accident, then, that two indications of a relative shift toward the agricultural side of the agropastoral continuum date to this period: a dip in the proportion of sheep and goat and an increase in the wild seed to cereal ratio (see discussion in concluding chapter).

Population Movement

Several questions specific to the culture history of Gordion will also be addressed. For example, do changes in the agropastoral economy reflect changing ties to the world beyond the Sakarya valley? Anatolia has long been a crossroads between east and west, and north and south. Based on both ancient texts and modern archaeology, Gordion has attracted scholarly attention concerning several ancient episodes of migrations, or at least of population movement. One group of questions for Young, Sams, Voigt, and others is: When and under what circumstances did Phrygians arrive in Anatolia, and can they be identified by non-linguistic material remains? The same questions can be asked of the Celtic (Galatian) arrival and presence. Voigt and Henrickson’s stratigraphy-based analyses of changes in material culture have generated several hypotheses in this regard. Social, political, and ethnic environment all may affect the agropastoral economy; assigning changes in the archaeobotanical record exclusively to these specific factors would be unwise.

By phrasing these questions somewhat less specifically, however, the archaeobotanical remains could provide some illumination as well. We all know that pots do not equal people, and archaeological cultures (typically recognized by pottery) do not equal ethnic groups. It is hard to think that plant remains could unequivocally distinguish Gordion’s place in the orbit of the Hittites (YHSS 8/9) from its independence during the heyday of the Phrygians (YHSS 7, 6, 5), or mark the Persian conquest (YHSS 4), the arrival of the Celts (Galatians) (YHSS 3), or contacts with the wider Islamic world (YHSS 1). Voigt (1994:276) suggests that certain kinds of domestic, relatively private, habits can help identify cultural markers. Examples include hearth and fireplace form, which could relate to food preparation customs; one archaeobotanical contribution to the discussion is food remains.

Cultural Affiliation

One of the results of the Gordion archaeobotanical study is that much of the evidence for environment and land use in the Sakarya valley shows incremental change that is not correlated in any obvious way with the apparent changes in the population or its cultural affiliation. As will be demonstrated in this study, the most significant change in agricultural strategy occurred during Early and Middle Phrygian times, when cultural continuity prevailed. Despite the dramatic history of population movement and replacement in the Sakarya valley, agricultural strategies appear to have been remarkably stable. I suggest that at a given level of technology within the Near Eastern agricultural tradition, the harsh environment of the Sakarya valley strongly constrains the agricultural possibilities, and that when any newcomers arrived, it behooved them to learn how to be successful farmers from the local population, if they did not already know. This is not to deny any agricultural innovation at all, but that of necessity it was cautiously applied. In conjunction with data and interpretations generated by other researchers, however, two possible expressions of Phrygian identity may be suggested (see discussion in concluding chapter): the consumption of einkorn and a possible “heirloom” artifact made of alder.