Absolute age range of the Late Cypriot IIC Period on Cyprus.

MANNING, STURT W. ; WENINGER, BERNHARD ; SOUTH, ALISON K. 等

Introduction

Late Cypriot (LC) IIC marks the high point of 2nd-millennium BC civilization on Cyprus. A number of monumental buildings occur at various sites around the island, usually associated with large settlements of 10+ ha (Negbi 1986; Astrom & Herscher 1996), and there is extensive evidence of involvement in Mediterranean trade systems (Knapp & Cherry 1994: 123-67). Until now, absolute dates for the period have been estimated solely on the basis of material culture associations between Cyprus and the historically dated civilizations of Egypt and the Near East (often indirectly via imports of Aegean ceramics both to Cyprus, and Egypt and the Near East). No direct and independent chronometric data existed. We report sets of radiocarbon determinations relevant to the beginning and end of LCIIC. These data offer a firm chronology for this key period, and, moreover, demonstrate the general validity of the conventional Late Bronze Age (LBA) chronologies for the region.

LCIIC period and traditional dating

The formal definition of the LCIIC period was codified by Astrom (1972a; 1972b; see also Kling 1989: 6-55), and then articulated in subsequent excavations (Kling 1989: 55-82). It represents the climax of LC civilization, and correlates with the final century or so of the great palatial civilizations of the Aegean and Near East. The close of the period coincides with the region-wide collapse of the LBA palatial trading systems of the east Mediterranean (Drews 1993)--a process that began c. 1200 BC. This approximate date is determined from material culture linkages with the historically derived chronology of Egypt (Kitchen 1996).

On Cyprus the evidence for the close of the LCIIC period varies. Some sites were abandoned or destroyed around the close of LCIIC, but others were not, or were rebuilt. Further, whereas past scholarship held that there was a clear distinction in the material culture between LCIIC and the subsequent LCIIIA period, today it is argued that there is no clear distinction between the end of LCIIC and the start of LCIIIA. Kling (1989) therefore introduced the concept of an LCIIC/IIIA transition phase. This detailed issue is not the subject of the present paper focused on the absolute chronology of the main traditional LCIIC period; thus the LCIIC/IIIA transition phase will be considered as approximately synonymous with the early LCIIIA period in the text below.

On the basis of extensive ceramic associations and exchanges, LCIIC is broadly correlated with the Late Helladic (LH) IIIB period in the Aegean, notwithstanding certain problems of detail (Kling 1989: 170-71). Some scholars have speculated that the close of the period lies slightly later in the early LHIIIC period, but the evidence is not clear-cut. In view of the Aegean-Cypriot-Egyptian linkages, and their interpretation, a variety of fairly similar dates have in turn been proposed for LCIIC: TABLE 1.

TABLE 1. Dates proposed for the LCIIC period on the basis of
Aegean-Cypriot-Egyptian material culture linkages.

scholars dates proposed for LCIIC period

Astrom 1972b: 760-62 1320-1190 BC
Merrillees 1977:46 1325/1300-1225/1200 BC
Warren & Hankey 1989: 118, 169 late 14th century BC to unspecified
 point after 1185/1180 BC
Merrillees 1992: 51 1300-1200 BC

However, in view of the indirect and sometimes circular method of art-historical dating employed in all these estimates, the need for direct and independent dating evidence is self-evident. And, in reverse, if independent absolute dates are available for LCIIC, then these are relevant to the chronology of the other contemporary east Mediterranean civilizations.

This project

We sought suitable samples for radiocarbon dating relevant to either the beginning or end of the LCIIC period from recent excavations or study at several of the main sites on Cyprus: FIGURE 1. The aim was to date LCIIC in absolute, calendar, terms. The 58 samples and their contexts are summarized in TABLE 2. The radiocarbon measurements were converted (calibrated) into calendar year dating probability ranges with the OxCal computer programme (Ramsey 1995 and later versions; OxCal 3.5 manual) and the current recommended INTCAL98 radiocarbon calibration dataset (Stuiver et al. 1998). Combined calibrated age ranges (i.e. combining the individual calibrated probability distributions via Bayesian rules for the combination of probabilities), and weighted averages of radiocarbon ages before calibration (where satisfactory given a chi-square test), are as defined and calculated by the Combine and R_Combine functions of OxCal. The individual calibrated calendar dating probability distributions for the samples are shown in FIGURES 2-5. We turn now to the attempt to date the beginning and the end of LCIIC from these data.

[ILLUSTRATION OMITTED]

[Graphs omitted]

TABLE 2. Radiocarbon samples in this study. Where not stated otherwise,
samples are wood charcoal and so set termini post quos for their find/
use contexts. Short-lived samples, e.g. grain, olive seeds, grape seed,
prunus seed, offer radiocarbon ages approximately contemporary with
find context.

Period Alassa Palaeotavema Apliki Karamallos House A
 Building II room 3 (Taylor 1952; Kling
 (Hadjisawas 1996) 1989: 85-87)

LCIIB
 (i) AA33445, 33446, 33447,
 33448, 33449, roof-timber forming
 (i) KN-4623, 4624, 56-year dendrochronological
 from construction sequence. Probably relates to
LCIIC of building (main earier LCIIC cutting and use
 ashlar building at (possible range early-mature
 site), and relating LCIIC).
 to original
 construction or
 use in LCIIC.

 (ii) AA33440, 33441, 33442,
LCIIC/IIIA 33443, 33444, small branches
Trans. and stems, [is less than or equal
Phase to] 5 yrs. growth,comprising
 basket. Dates LCIIC/IIIA trans,
 phase or early LCIIIA.
 (iii) AA33450, 33451, 33452,
 33452A, 33453, 33454, from
 cereal sample 3. Dates early
 LCIIIA or LCIIC/IIIA trans, phase
LCIIIA (ii) KN-4625, 4846, burning and abandonment.
 from final use of
 building in LCIIIA
 period.

Period Kalavasos Ayios Maroni
 Dhimitrios Building X Aspres
 (South 1997) (Manning
 1998: 48-53)

LCIIB
 14 samples on elements
 of substantial roof-beams
 (KN-4615 to 4622, 4653,
 4654, 4726 to 4728,
 4829) and 2 samples
 perhaps from doorway OxA-8373, a
 elements (KN4829 4782). grape seed
 Building constructed associated
LCIIC between later LCIIB and with a pithos-
 early LCIIC, and then base
 significantly reconstructed installation,
 during LCIIC. Wood likely Dates earlier-
 cut and used somewhere mature LCIIC.
 between later LCIIB and
 during (earlier) LCIIC.

LCIIC/IIIA
Trans.
Phase

LCIIIA

Period Maroni Tsaroukkas
 Building 1
 (Manning 1998)

LCIIB

 (i) OXA-8351, olive seed
 from pre-final Building 1
 context. Sets precise
 terminus post quem for
 final later LCIIC structure
LCIIC (ii) OXA-8265, 8266, 8267,
 olive seeds from final use
 horizon in building in later
 LCIIC. Date later LCIIC.
 (iii) OXA-8372 from late
 LCIIC context with building
 collapse and dumping.
 Date either late LCIIC or
LCIIC/IIIA possibly post-LCIIC,
Trans.
Phase

LCIIIA

Period Maroni Tsaroukkas Maroni
 Building 2 Voumes
 (Manning 1998) Ashlar
 Building room
 4 (Cadogan
 1996)

 (i) OXA-8353, 8355, 8237, 8323, from
 pre-Building 2 contexts of LCI-IINB
 date.

LCIIB

 (ii) OXA-8348, 8349, from primary
 construction of Building 2. OXA-8352,
 from LCIIC industrial use context.
 Samples offer termini post quos for the
 above contexts.
 (iii) OXA-8324, olive seed from final
 use of building. Dates later LCIIC.
 (iv) OXA-8321, 8326, 8354, collapse
 and abandonment phase,
LCIIC late LCIIC. Samples derive from
 LCIIC construction or use KN-4547, olive
 phase. Offer termini post quos for seeds from
 these, near olive
 (v) OXA-8350, 8264, 8325 press. Dates
 8322, from post-use debris/dumping final use of
 of original LCIIC material. As this room and
 (vi) for charcoal samples (OXA-8264, building late
 8325). Consistent with mixed material in LCIIC.
LCIICIIIIA in a dump, OXA-8350, a prunus seed,
Trans. and OXA-8322, an olive seed, offer
Phase significantly differing ages, and
 appear torelate to different original
 use contexts: plausibly earlier and
 late LCIIC respectively.

LCIIIA

The absolute date of the beginning of the LCIIC period

The radiocarbon determinations on parts of roof-beams or doorways from Kalavasos Ayios Dhimitrios, and the two determinations on long-lived architectural samples from LCIIC contexts at Alassa Palaeotaverna, provide date ranges for tree-rings employed in, respectively, later LCIIB/early LCIIC, or LCIIC, architecture. They offer termini post quos for the LCIIC cultural phase. Many of the tree-rings dated are significantly older than the date the trees were felled for use in construction (none are preserved to bark). This tendency may have been exaggerated if original outer rings burnt and peeled away in the major fire destruction conspicuous especially at Kalavasos Ayios Dhimitrios Building X. As evident in FIGURE 4, several of the Kalavasos samples reflect wood growing in the 16th-15th centuries BC (also the case for KN-4624 from Alassa). However, among the distribution of ages in the overall set, the pattern of latest ages represented should give a likely indication of the ages of outer rings from the roof-beam/architecture samples, and so an estimate towards the likely cutting/use date. Several samples include calibrated age ranges down to, or around, 1400 BC, and two determinations in particular provide calibrated age ranges predominantly in the 14th century BC: KN-4654, 4726 -- respective calibrated age ranges in detail at 68.2%, 1 [Sigma], confidence 14301370 BC (41.7%) or 1360-1310 BC (26.5%), and 1400-1290 BC (60.3%) or 1280-1260 BC (7.9%) (Note: the 1 [Sigma] range is the most likely 68.2% of the total 100% dating probability; individual parts of the 1 [Sigma] range may also be noted with their probability in % terms -- of the overall total 100% -- together they add up to the 1 [Sigma], 68.2%, of the total 100% dating probability). Overall, a real calendar date range in the 14th century BC therefore appears plausible for the late LCIIB and/or early LCIIC construction.

A further, more specific, terminus post quem is available from the Pin us nigra charcoal from the LCIIC-constructed roof at House A room 3, Apliki Karamallos. The tree-rings were examined and one group found to comprise a 56-year tree-ring chronology (APK-2A). It is not possible to date such a short sequence by dendrochronology. The sample was thus cut into five decadal samples (rings 1001-1010, 1011-1020, 1021-1030, 1031-1040, 1041-1049) with a view to obtaining a radiocarbon wiggle-matched chronology for the fixed sequence: FIGURE 6. The calculated calendar age range for the most recent decade centres on the interval 1340-1331 BC, and, with error allowances, may be defined as 1365-1295 BC at 1 [Sigma], 68.2%, confidence, or 1380-1250 BC at 95.4%, 2 [Sigma], confidence. There were then 12 further rings present, and an unknown number of missing rings: so we have a terminus post quem for the roof c. 1322.5 [+ or -] 35 BC at 1 [Sigma] confidence. The wood was cut and used plausibly in the earlier LCIIC period (and before the final LCIIC/IIIA transition phase occupation at the site). Even allowing for missing rings to bark beyond the last preserved ring, it is clear that this wood was probably felled somewhere late in the 14th century BC, and no later than a couple of decades into the 13th century BC.

[ILLUSTRATION OMITTED]

Several determinations on charcoal samples from Building 2 at Maroni Tsaroukkas are also relevant. Two charcoal samples related to the construction (roofing matter?) of the building (OxA-8349, 8348) offer similar age ranges: combined calibrated age range of 1400-1310 BC at 1 [Sigma] confidence. A terminus post quem and/or start date for points within LCIIC somewhere in the 14th century BC is suggested. The charcoal samples from the abandonment phase (OxA-8321, 8326, 8354) or dumping (OxA-8264, 8325) also reflect wood either originally used in architecture during LCIIC (compare OxA-8349, 8348), or wood employed in artefacts, or as firewood, and so on, in LCIIC. The calibrated age ranges, for the respective termini post quos, favour the 14th-13th centuries BC (see FIGURE 5).

Finally, a few data on short-lived samples relevant to the date of the earlier LCIIC period are presented (TABLE 3). OxA-8351 provides an age range contemporary with shortly before the construction of LCIIC Building 1 at Maroni Tsaroukkas and indicates that a point in earlier LCIIC probably lay in the 14th century Be. OxA-8373, from an earlier-mature LCIIC context, offers a very similar age range, as does OxA-8350, which is also likely to derive originally from the same period.

TABLE 3. Short-lived radiocarbon samples relevant to the date of the
earlier LCIIC period. Note: the total 1[Theta] range is the most likely
68.2% of the overall 100% dating probability for each sample when
calibrated. The most likely 1[Theta], 68.2%, range may be in more than
one part, as here.

sample 1[Theta], 68.2%, calibrated
 age range BC

OxA-8351 olive seed from 1410-1310 1280-1260
Maroni Tsaroukkas, pre-Final Building 1 (65.3%) (2.9%)

OxA-8373 grape seed from 1430-1290 1280-1260
Maroni Aspres (61.4%) (6.8%)

OxA-8350 prunus seed from 1410-1300 1280-1260
Maroni Tsaroukkas, Building 2 (64.3%) (3.9%)

Overall, the common calibrated ranges for all these radiocarbon data relevant to early to earlier LCIIC are either c. 1395-1375 BC or 1340-1315 Be. The shape of the radiocarbon calibration curve -- that is the history of variations in natural radiocarbon levels in the atmosphere -- explains the dichotomy: see FIGURE 6. How do we choose which offers a better start range for the period? First, we may note that the radiocarbon wiggle-matched likely earlier LCIIC tree-ring sequence from Apliki favours a date no earlier than the second half of the 14th century BC and probably one towards its end (or later). Second, we may note the mature LCIIB ceramics found at Amarna in Egypt (Merrillees 1968: 78-88; Astrom 1972b: 761). These imports arrived during the short-lived existence of this capital city between Akhenaten year 5 and Tutankhamun year 3 (c. 1349-1334 BC: Kitchen 1996). They demonstrate that c. 1395-1375 BC is far too early a date for the subsequent LCIIC period, which instead must begin no earlier than the late Amarna period (i.e. after 1340/1330 BC). Thus the dating range c. 1340-1315 BC best describes an approximate starting date for the LCIIC period.

The absolute date of the close of the LCIIC period

This is defined by the short-lived samples from final use contexts late in the period, with the long-lived samples from the succeeding LCIIIA period offering consonant evidence (termini post quos for LCIIIA). The short-lived samples associated with the late LCIIC period in this study (all from the Maroni sites) are: OxA-8265, 8266, 8267, 8324 and KN-4647. OxA-8322 from Building 2 at Maroni Tsaroukkas may also probably be associated with this group. These samples ought to reflect similar calendar ages. The combined calibrated age range for the set at 1 [Sigma] confidence is 1260-1240 BC (10.6%), or 1220-1130 BC (57.6%).

The relatively wide and imprecise calibrated age range for this set is explained by the shape of the radiocarbon calibration curve in the period from the 13th to 12th centuries BC (see FIGURE 6). In the period from 1265 BC to 1135 BC there are several quite distinct `wiggles' in the curve: for example, in the 40-year period from 1225 to 1185 BC, the radiocarbon timescale ranges from 3010 [+ or -] 17 BP to 2921 [+ or -] 18 BP. Within measurement errors, all the later LCIIC samples on short-lived samples could date either 1265-1245 BC, or 1225-1185 BC, or in the next several decades down to the significant upwards wiggle at 1135 BC -- which offers a final solid intercept for all the determinations. There is thus a built-in ambiguity at this period. However, there are ways we can try to resolve this. First, we can examine in more detail the relation of these dates to the calibration curve, and, second, we can consider whether the short-lived samples from a subsequent early LCIIIA (or LCIIC/IIIA transition phase) context at Apliki provide clarification.

All the samples above came from similar late LCIIC contexts. They are not LCIIIA (or LCIIC/ IIIA transition phase). Thus, although two of the samples in particular offer calibrated age ranges which, while including the decades around 1200 BC, extend and even favour ages later in the 12th-11th centuries BC (OxA-8324, 8322), the real calendar age interpretation for these samples from the possible ranges should be so as to enable them to be at least fairly similar in real calendar age to the other samples (assuming data are `good'). The common temporal span for the entire set of samples at 1 [Sigma] confidence consists in detail of 1254-1244 BC (10.7%), or 1212-1198 BC (15.5%), or 1180-1150 BC (35-1%), or 1144-1138 BC (6.9%). Common-sense would indicate that somewhere c. 1212-1150 BC (50.6% in total) is the reasonable best estimate for the common age range.

To go further, we can consider (i) the basket matter and (ii) the cereal samples from the final occupation at Apliki. The context is early in the LCIIIA period (or the LCIIC/IIIA transition phase).

i The brushwood samples forming the basket consisted of short-lived material ([is less than or equal to] 5 years growth). Of course, the manufacture of the basket may have occurred a little while before its final use, and then destruction in the burning of the room. Nonetheless, since the basket was in use at the end of occupation, and the working lifetime of such a basket was probably not long, the short-lived branches forming the basket probably were growing no earlier than the very late LCIIC period and in fact more likely in the early LCIIIA period (LCIIC/IIIA transition phase). The radiocarbon determinations on these samples therefore offer approximate dates for the end of LCIIC or a little later. The calibrated age ranges of the five individual radiocarbon measurements centre around 1200 BC (FIGURE 3), and, combined, offer a calibrated age range in detail at 1 [Sigma] confidence of 1290-1279 BC (8.2%), or 1262-1210 BC (42-5%), or 1199-1191 BC (6.1%), or 1175-1168 BC (4.3%), or 1140-1131 BC (7.0%). While not conclusive, much of the 13thcentury BC dating possibility may be ruled out given the evidence from the prior late LCIIC samples, and, if one also rules out the late possibility on the 1135 BC wiggle as not compatible with one date (AA33443) and barely compatible with another (AA33442), these samples favour a calendar date somewhere from late in the 13th century BC to before 1167 BC. They would thus appear to support selection of an end date for the main LCIIC period before the second half of the 12th century BC, and, indeed, favour a range at the earlier end of the possible span determined above (i.e. more c. 1212-1198 BC or 1180-1168 BC than towards c. 1150 BC).

ii The cereal samples offer radiocarbon ages approximately contemporary with the destruction of Apliki, and so offer a terminus ante quem for the end of the main LCIIC period. The six determinations from the cereal cache ought to be more or less exactly contemporary in calendar age, and the radiocarbon ages form a tight set. The combined calibrated age range at 1 [Sigma] confidence is 1260-1230 BC (15.2%) or 1220-1120 BC (53.0%). The statistically consistent weighted average radiocarbon age of the set is 2959 [+ or -] 20 BP, which yields an almost identical calibrated age range at 1 [Sigma] confidence of 1260-1230 BC (14.8%), or 1220-1120 BC (53.4%). The notable observation is that the calibrated age range for these slightly later -- in real calendar time -- samples is almost identical to the calibrated age ranges for the later LCIIC samples and end of LCIIC/early LCIIIA samples. It would therefore seem that late LCIIC must lie in the earlier part of this common range, the end of LCIIC around the middle, and early LCIIIA towards the end. The next step is to try to quantify this common-sense observation.

Trying to resolve the c. 1200 BC ambiguity

It is apparent that close dating of the end of LCIIC requires resolution and quantification of several data sets, which are known to be discrete and sequential (in stratigraphic and calendar terms), but which yield largely overlapping calibrated radiocarbon age ranges.

We have a sequence consisting of:

i A clear terminus post quem for later LCIIC in the form of the five Apliki roof-timber samples -- weighted average age for the set is 3069 [+ or -] 24 [sup.14]C years BP.

ii The later LCIIC data from short-lived samples from Maroni. The Maroni sites were all abandoned late in LCIIC, but before the end of the overall Cyprus-wide LCIIC period (e.g. slightly before the destruction of Building X in late LCIIC at Kalavasos: Cadogan 1996: 17). The weighted average age for the set is 2971 [+ or -] 14 [sup.14]C years BP.

iii The short-lived branches from the Apliki basket, which date no earlier than very late LCIIC and more likely early LCIIIA (or LCIIC/IIIA transition phase) weighted average age for the set is 2995 [+ or -] 24 [sup.14]C years BP.

iv The cereal samples dating the destruction of Apliki -- weighted average age for the set is 2959 [+ or -] 20 [sup.14]C years BP.

We may seek to employ this calendar/stratigraphic sequence, in combination with the calibration of the radiocarbon data, to obtain more precise calibrated age ranges for each of the elements of the sequence via the Bayesian statistical methods available in the Sequence function of the OxCal software (see Ramsey 1995 and OxCal 3.5 manual). The Sequence function incorporates the stratigraphic information and uses a form of Markov Chain Monte Carlo sampling to estimate the ordered (i.e. constrained) sets of probability distributions. The results (see FIGURE 7) at 1 [Sigma] confidence of a typical analysis are:

i Apliki roof: 1391-1310 BC (66.4%), or 1270-1267 BC (1.8%),

ii Maroni later LCIIC: 1262-1229 BC (62.3%), or 1216-1209 BC (5.9%),

iii Apliki basket: 1242-1210 BC (41.6%), or 1201-1190 BC (13.3%), or 1177-1164 BC (13.4%)

iv Apliki cereal: 1201-1188 BC (6.0%), or 1181-1125 BC (62.2%).

[Graph omitted]

From this we see that a point in later LCIIC most likely lies somewhere 1262-1229 BC. The actual end of the LCIIC period was at least a few years later, and probably one to a few decades later than this Maroni datum: e.g. perhaps somewhere between c. 1250-1200 BC. The Apliki basket samples remain rather ambiguous in terms of a clear date. One final consideration may therefore be noted. Although all samples are short-lived branch matter and in total seem to represent a calendar window of no more than about 5 years, there is a significant range -- some 95 [sup.14]C years from 3050 [+ or -] 55 BP to 2955 [+ or -] 55 BP -- among the five radiocarbon ages in the set (contrast tight consistent sets for Apliki cereal samples or roof-timber sequence). This implies that the samples derive from a period when there was a very sharp change in atmospheric radiocarbon levels over only a few years -- enabling the large range in the radiocarbon measurements to be present. While by no means definitive, the most plausible places such a sub-decade variation from 3050 [+ or -] 55 BP to 2955 [+ or -] 55 BP could derive from, within the possible date ranges determined above, are on the steep slope centred c. 1215 BC, or either side of the sharp wiggle in the radiocarbon calibration curve c. 1195 BC (either 1205-1195 BC or 1195-1185 BC): see FIGURE 6. In turn, this points to two scenarios: a `high' one where LCIIC ends by, or shortly before, c. 1215 BC, or a `lower' one where LCIIC ends by, or shortly before, c. 1200-1190 BC. This leaves the subsequent early LCIIIA (or LCIIC/IIIA transition phase) destruction of Apliki. On the `high' scenario above, this could be either 1201-1188 BC or somewhere 1181-1125 BC; on the `lower' scenario above, only the 1181-1125 BC range appears plausible. But, if it is thought likely that the Apliki basket was not in use for a very long period of time (i.e. less than 30 years), then a date after c. 1160 BC seems unlikely for the Apliki destruction.

The final evidence from this project consists of the two LCIIIA samples from Alassa. Their combined calibrated age range at 1 [Sigma] confidence is 1260-1240 BC (6.5%), or 1220-1110 BC (52.7%), or 1100-1080 BC (5.8%), or 1060-1050 BC (3.2%). The main probability centres on the 12th century BC. KN-4625 derives from wood found near a hearth in use during the LCIIIA period, and KN-4846 is charcoal from a LCIIIA fill outside the north wall of Building II. These determinations suggest a 12th-century BC range for the post-LCIIC phase at Alassa, consistent with the data from the Apliki cereal samples for a point in early LCIIIA.

Conclusions

We propose an absolute age range for the main LCIIC period from c. 1340-1315 BC to c. 1200 BC +20/-10 (and no later than c. 1168 BC at the extreme). This estimate for the LCIIC period is the first direct, and independent, temporal definition of this important cultural phase. It is thus very welcome to find that the dates determined are almost identical to those estimated previously from the study of material culture linkages within the eastern Mediterranean (TABLE 1). Therefore, the large set of radiocarbon data from this project, including extensive data from short-lived samples contemporary with use-contexts (cf. arguments of James et al. 1998: 36-8 based on a claimed lack of such) strongly support the general range of the conventional later LBA chronologies of Cyprus and the Aegean (see previously e.g. Manning & Weninger 1992). In turn, this means that the conventional chronology of Egypt -- the source of the dates for the conventional LBA chronologies of the Aegean, Cyprus and so on via material culture linkages -- is approximately correct. It cannot be incorrect by a large amount, such as a century (Hagens 1996; 1999), let alone the c. 250 years argued for by James et al. (1991; 1998) and Rohl (1995). This project therefore marks, again, an end to suggestions of radically lower LBA chronologies in the eastern Mediterranean and Near East.

Acknowledgements. Permission for obtaining samples was provided by the Department of Antiquities, Cyprus. The radiocarbon analyses were funded by the University of Koln (Koln dates), NERC (Oxford dates), and the NSF (Arizona dates). We thank the three laboratories for their work. The dendrochronological analyses were carried out in the Malcolm and Carolyn Wiener Laboratory for Aegean and Near Eastern Dendrochronology, Cornell University. We thank Vasiliki Kassianidou for her assistance with the Apliki samples. We thank the ANTIQUITY referees for their comments, also Maryanne Newton and Mary Jaye Bruce.

References

ASTROM, P. 1972a. The Late Cypriote Bronze Age architecture and pottery, as P. Astrom, The Swedish Cyprus Expedition, vol. IV, part 1C. Lund: The Swedish Cyprus Expedition.

1972b. The Late Cypriote Bronze Age. Relative and absolute chronology, foreign relations, summary and historical conclusions, in L. Astr6m & P. Astrom (ed.), The Swedish Cyprus Expedition, vol. IV, part 1D: 675-781. Lund: The Swedish Cyprus Expedition.

ASTROM, P. & E. HERSCHER. (ed.). 1996. Late Bronze Age settlement in Cyprus: function and relationship. Jonsered: Paul Astroms Forlag. Studies in Mediterranean Archaeology Pocket-book 126.

CADOGAN, G. 1996. Maroni: change in Late Bronze Age Cyprus, in Astrom & Herscher (ed.): 15-22.

DREWS, R. 1993. The end of the Bronze Age: changes in warfare and the catastrophe ca. 1200 Be. Princeton (NJ): Princeton University Press.

HADJISAVVAS, S. 1996. Alassa: a regional centre of Alasia? in Astrom & Herscher (ed.): 23-38.

HAGENS, G. 1996. A critical review of dead-reckoning from the 21st Dynasty, Journal of the American Research Centre in Egypt 33: 153-63.

1999. An ultra-low chronology of Iron Age Palestine, Antiquity 73: 431-9.

JAMES, P., N. KOKKINOS & I.J. THORPE. 1998. Mediterranean chronology in crisis, in M.S. Balmuth & R.H. Tykot (ed.), Sardinian and Aegean chronology: towards the resolution of relative and absolute dating in the Mediterranean: 29-43. Oxford: Oxbow.

JAMES, P., I.J. THORPE, N. KOKKINOS, R. MORKOT & J. FRANKISH. 1991. Centuries of darkness. London: Jonathan Cape.

KITCHEN, K.A. 1996. The historical chronology of ancient Egypt, a current assessment, Acta Archaeologica 67: 1-13.

KLING, B. 1989. Mycenaean IIIC:1b and related pottery in Cyprus. Goteborg: Paul Astroms Forlag. Studies in Mediterranean Archaeology 87.

KNAPP, A.B. & J.F. CHERRY. 1994. Provenience studies and Bronze Age Cyprus: production, exchange and politico-economic change. Madison (WI): Prehistory Press.

MANNING, S.W. 1998. Tsaroukkas, Mycenaeans and Trade Project: preliminary report on the 1996-1997 seasons, Report of the Department of Antiquities, Cyprus: 39-54.

MANNING, S.W. & B. WENINGER. 1992. A light in the dark: archaeological wiggle matching and the absolute chronology of the close of the Aegean Late Bronze Age, Antiquity 66: 636-63.

MERRILLEES, R.S. 1968. The Cypriote Bronze Age pottery found in Egypt. Goteborg: Paul Astroms Forlag. Studies in Mediterranean Archaeology 18.

1977. The absolute chronology of the Bronze Age in Cyprus, Report of the Department of Antiquities, Cyprus: 33-50.

1992. The absolute chronology of the Bronze Age in Cyprus: a revision, Bulletin of the American Schools of Oriental Research 288: 47-52.

NEGBI, O. 1986. The climax of urban development in Bronze Age Cyprus, Report of the Department of Antiquities, Cyprus: 97-121.

RAMSEY, C.B. 1995. Radiocarbon calibration and analysis of stratigraphy: the OxCal program, Radiocarbon 37: 425-30.

ROHL, D.M. 1995. A test of time 1: the Bible -- from myth to history. London: Century.

SOUTH, A.K. 1997. Kalavasos-Ayios Dhimitrios 1992-1996, Report of the Department of Antiquities, Cyprus: 151-75.

STUIVER, M., P.J. REIMER, E. BARD, J.W. BECK, G.S. BURR, K.A. HUGHEN, B. KROMER, G. MCCORMAC, J. VAN DER PLICHT & M. SPURK. 1998. INTCAL98 radiocarbon age calibration, 24,000-0 cal BP, Radiocarbon 40: 1041-83.

TAYLOR, J. DU PLAT. 1952. A Late Bronze Age settlement at Apliki, Cyprus, The Antiquaries Journal 32: 133-67.

WARREN, P. & V. HANKEY. 1989. Aegean Bronze Age chronology. Bristol: Bristol Classical Press.

Received 9 November 1999, revised 10 January 2000, accepted 13 April 2000, revised 16 March 2001

STURT W. MANNING, BERNHARD WENINGER, ALISON K. SOUTH, BARBARA KLING, PETER IAN KUNIHOLM, JAMES D. MUHLY, SOPHOCLES HADJISAVVAS, DAVID A. SEWELL & GERALD CADOGAN(*)

(*) Manning, Department of Archaeology, University of Reading, PO Box 218 Whiteknights, Reading RG6 6AA, England. S.W. Manning@reading.ac.uk Weninger, Institut der Ur- und Fruhgeschichte, Radiocarbonlabor Weyertal 125, Koln 5000, Germany. South, Vasilikos Valley Project, 7733 Kalavasos, Larnaca, Cyprus. Kling, 176 Lost Nation Road, Essex Junction VT 05452, USA. Kuniholm, The Malcolm & Carolyn Wiener Laboratory for Aegean & Near Eastern Dendrochronology, Department of the History of Art & Archaeology, G-35 Goldwin Smith Hall, Cornell University, Ithaca NY 14853-3201, USA. Muhly, American School of Classical Studies, Odos Souedias 54, 10676 Athens, Greece. Hadjisavvas, Department of Antiquities, Ministry of Communications & Works, PO Box 2024, Nicosia, Cyprus. Sewell, Environmental Systems Science Centre, University of Reading, PO Box 238 Whiteknights, Reading RG6 6AL, England. Cadogan, Culworth, Banbury OX17 2AT, England.