A 677 Year Tree-ring Chronology for the Middle Bronze Age

“A 677 YEAR TREE-RING CHRONOLOGY FOR THE MIDDLE BRONZE AGE”

PETER IAN KUNIHOLM - MARYANNE W. NEWTON

INTRODUCTION

In 1973 when I proposed to Prof. Rodney S. Young to undertake the dendrochronology of An- atolia as a dissertation topic, one of his immediate suggestions was: “Go to Tahsin and get some Kiiltepe wood!” Over the years we have collected wood at Kiiltepe thanks to Prof. Ozgiiq’s generos- ity, and, since it is difficult to study the dendrochronology of one site alone, we have also collected comparative wood from Acemhiiyiik thanks to Prof. Nimet Ozgiig. It is therefore our pleasant duty to publish this first report on Middle Bronze Age dendrochronology in honor of both the Man from Kanesh and the Lady from Purushkanda who after Dr. Young’s death became the god-parents of this project. We also thank all the assistant field directors and site supervisors, especially Prof. Kutlu Emre, Dr. Aykut Cmaroj$, and Dr. Aliye Ozten who helped to welcome us and make work so much easier by discussing the excavations which were still in progress and by sharing their ideas at every step.

Kiiltepe and Acemhiiyiik will always be focal points of any study of the Middle Bronze Age, as the type sites for the Assyrian Colony Period, as vast resources of textual and art-stylistic information, and as centers to which many famous Near Eastern rulers of the 18th century BC sent rich gifts, thereby establishing for us the cultural and temporal cross-links so necessary for an understand- ing of the period. Thus, it is only appropriate that any dendrochronological study of the MBA begin with the wood of Kiiltepe and Acemhiiw.

Our Middle Bronze Age work occurred at a time when we were also trying to develop long tree-ring chronologies which would span the approximately 3700 years from the prcsent to the MBA. The most recent summary of this effort is P.I.Kuniholm and C.L.Striker, “Dendrochronologi- cal Investigations in the Aegean and Neighboring Regions, 1983-1986,” Journal of Field Archaeology 14 (1987) 385-398.

GOALS:

Several questions immediately came to mind as we started to measure the Middle Bronze Age wood (actually charcoal). They are listed below with the results as of October 1988:

QUESTION ANSWER

58 years before the Acemhiiyiik bulildings.

58 years after Wargama Saray and the same year as Hatipler Tepesi.

58 years after Wargama Saray and the same year as Sankaya.

1.

2.

3.

Relative construction date for the Wargama Saray at Kiiltepe?

Relative construction date for the Sankaya Palace at Acemhiiyiik?

Relative construction date for the Hatipler Tepesi at Acemhiiyuk?

280

4.

5.

6 .

7. 8.

9.

10.


Relative construction date(s) for the Kultepe I1 buildings and therefore the difference in time between Kultepe I1 and Kultepe Ib? (Present system based principally on lirnu-names and to a lesser extent on pottery and sealings.) No wood yet from Kultepe 11.

Relative dates between the Palace and Karum at Kultepe? No suitable wood yet from the Karum.

Destruction dates for any of these buildings? No. or at least an indication from repairs to the buildings that would give us an idea of approximately how many years they were in use before they burned to the ground?

One indication from Kultepe. Relative construction dates for the lesser buildings at Acemhiiyiik? One. Could we construct a long Middle Bronze Age tree-ring chronology by means of which other Anatolian MBA sites could be dated? Yes. Could we “wiggle-match” this long chronology to find where it fits on the radiocarbon time- scale? Results only 3/7 complete. Absolute dates for the Middle Bronze Age? Not yet.

COMMENT ON HOW TO READ THE FOLLOWING CATALOGUES OF SAMPLES:

SAMPLES

Sampling numbers (KUL-1, KUL-2, etc.) refer to the order in which the specimens were col- !ected.

PROVENIENCES

Grid square references and room numbers, in the absence of a published plan of the WarSama Saray and of Hatipler Tepesi, were kindly provided by the excavators, Prof. Tahsin Ozguq, Prof. Kutlu Emre, and Prof. Nimet Ozgus. The name of the scientist who provided a species identification is listed in parentheses. Other information includes the function of the timber, if known.

WINGS

A,B,C,D, etc., refer to the fragment that was measured. Sometimes more than one fragment could be combined into a longer sequence (CDEF COMB). The numbers refer to the measurable rings on each sample.

RELATIVE DATES

The counting system is a purely arbitrary one based on KUL-13, one of the longest and clearest ring-sequences from this building. Its innermost ring, the pith (Turkish rnerkez), was arbitrarily as- signed a date of Relative Year 1001. All other dates at Kultepe and Acemhuyiik are given on the basis of this counting system. Thus KUL-17 starts at Relative Year 998 or three years earlier than KUL-13, and so-forth. Both the innermost ring and the outermost ring for each sample are given. If no numbers are given in the column marked RELATIVE DATES, this means that we have not been able to find an acceptable fit, and the sample remains undated. The measurements for each piece, however, are on our computer file and the pieces may some day be datable as we continue to collect Middle Bronze Age wood for comparison.

After the last preserved ring is one of the following notations:

B = Bark (Turkish kabuk). This means that the last preserved ring is from the year in which the tree was cut. No rings have been lost either to the carpenter’s tools or to the action of the fire which destroyed the building. (This is the most desirable kind of sample.)

Very few rings, if any, are missing. This is a subjective judgment on the part of the person measuring the piece and is based on the condition of the exterior of the wood (for example, with the bark peeled off but with no rings removed).

v =

A 677 YEAR TREE-RING CHRONOLOGY 28 1

w = This means that we have no way of proving how many rings are missing. (This is the least desirable kind of sample.)

The last rings can be counted but not measured, either because the sample is in bad condition or because the rings are impossibly small. The last measured ring is listed in the text under PROVENIENCES.

++ =

KULTEPE, WARSAMA SARAYI, CATALOGUE

SAMPLE PROVENIENCE RINGS RELATIVE DATES First Last Year Year

CITADEL, BURNED LATE PALACE (WARSAMA SARAYI) (KULTEPE KARUM Ib) CROSSDATED SAMPLES:

KUL - 3 No Grid No. Charred board, adzed on four sides. found on floor. Submitted 12 Au- gust 1975 by Prof. Kutlu Emre.

A = 68 1095 - 1 1 6 2 ~

KUL - 5 Grid No. Y/22. Part of transverse beam in N-S wall, just above socle. Charcoal fragments. Submitted 3 October 1975.

B = 135 1038 -1172~

5A, 7AD COMB. = 73

5C, 7BE COMB. = 84

KUL - 6 Grid No. V/23. W side of palace. From SW comer of room S of one from which KUL-5 was collected Five kerpig pylons to E, dividing the room in two. Charcoal fragments. Pinus sp. (silvestris group) (Dr. B. F. Kukachka).

A = 73

B = 53

-COMB.= 73

C = 70

D = 81

E = 66

F - 91 CDEF COMB. := 134 1039 -1172~

The following charcoal samples were collected in 1987. All are transverse beams at floor level from the E W wall in Grid Squares 2/23 and aa/23.

KUL - 11 Measured to RD 1166. Last six rings can be counted but not measured.

A = 72+6 1095 -1172++~

KUL - 12 A = 76 1080 - 1 1 5 5 ~

KUL - 13 A = 173 1001 -1173~

KUL - 14 A = 153 1020 -1172~

KUL - 15 A = 82 1092 -1173~

KUL- 16 A = 153 969 - 1 1 2 1 ~

282 PETER IAN KUNIHOLM - MARYANNE W. NEWTON



CITADEL, BURNED LATE PALACE (WARSAMA SARAYI) (KULTEPE KARUM Ib) CROSS- DATED SAMPLES (continued):

KUL - 17 A = 74 1100 -1173~

KUL - 18 Juniperus sp. (Dr. W. H. Schoch).

A = 174

KUL - 19 Later repair. See discussion below.

A = 90

998 -1171B

1145 - 1 2 3 4 ~

KUL - 20 A = 88

B = 91

c = 77

ABCCOMB.= 92 1082 -1173~

KUL - 21 ABCCOMB.= 93 1077 -1169~

KUL - 23 Extremely difficult piece to measure. The first 328 years are secure, but there appears to be a missing ring at Relative Year 1080. This piece and the problems with the last rings are discussed below. Juniperus sp. (Dr. W. H. Schoch). Also a repair or part of a later building program.

763 - 1 1 9 1 ~ ABCDEFG COMB. = 428 f

CITADEL, BURNED LATE PALACE (WARSAMA SARAYI) (KULTEPE KARUM Ib) SAM- PLES WHICH D O NOT CROSSDATE:

KUL- 1

KUL- 2

KUL- 4

KUL- 7

Grid No. 2/19-20. Heavily eroded lumps of wood, slightly charred on outside. Excav- ated in 1973, submitted in Feb. 1974 by Ugur Silistreli. KUL - 1A identified as Ce- drus libani Loud., KUL - 1B identified as Pinus silvestris L. (Dr. B. Aytug).

A = 34

B = 28

Grid No. Y/20. Pinus silvestris L. (Dr. B. Aytug). Excavated and submitted as above.

A = 5 1

East retaining wall of palace terrace. 26 m.f from NE comer. Transverse beam, in fragments. Centered .74 m. above bottom of foundation course. Submitted 3 October 1975 by Prof. Kutlu Emre.

A = unmeasurable

S side of palace, room with some stone paving on sides, 5th room from E side of palace. Grid square AA/23. Charcoal lumps (full cross-sections), found piled on N wall. 7A identified as Pinus sp. (silvestris group), 7BC identified as Juniperus sp. (Dr. B. F. Kukachka).

7AD, 5A COMB. = 73

7BE, 5C COMB. = 84

A 677 YEAR TREE-RING CHRONOLOGY 283



CITADEL, BURNED PALACE (WARSAMA SARA YI) (KULTEPE KARUM Ib) SAMPLES WHICH DO NOT CROSSDA TE (continued):

KUL- 8 Grid No. AA/31 and bb/29. Bag of oak fragments, eleven of which have been combined to form a small master chronology 84 years long which does not crossdate well with the Kiiltepe conifer master.

MASTER = 48

KUL - 9 Collected 1982. Provenience unknown.

A = 84

1987 sampling. This specimen is a transverse beam at floor level from the E-W wall in Grid Squares 2/23 and aa/23.

KUL - 22 (Erratic growth.) A = 91

DISCUSSION:

First, note in the catalogue and in the Kultepe bar graph (fig. 1) how many pieces have last preserved rings in Relative Year 1173 or have last preserved rings at the most one or two years earlier. In one instance. KUL - 18, the bark (kabuk) is present in 1172. We propose, therefore, that Relative Years 1172/3 are the construction years for the Wargama Saray. Of course, other pieces lost a number of rings due to the action of the wood-cutters and to the fire which destroyed the palace.

Second, we always look to see if there are some pieces which are later than the master chronology, thereby indicating that they are from later repairs. This would be especially likely if the Waraa- ma Saray were long-lived. Here the excavator’s opinion regarding the life-span of the building is im- portant. We understand from the excavators of both Kiiltepe and Acemhuyiik that their palaces may have had long lives (100 years or more). Thus it was with great interest that we recently discovered that one hitherto uncrossdated piece, KUL - 19 from Grid Squares 2 /23 and aa/3, fits at Relative Date 1234 or 61 years after most of the Kultepe timbers were cut. This means that the building was at least 61 years old at the time KUL - 19 was cut, possibly older if we make an allowance for missing exterior rings on KUL - 19. KUL - 23 from the same Grid Squares is also later than the majority. So far these are the only indications of later building activity. It is essential for determining the longevity of these buildings for additional wood of this kind to be recovered. We note that other pieces of wood from the same wall have terminal rings in 1173, but we do not have a ready explanation why there should be this difference unless KUL - 19 were a fallen timber from a repair to the roof or some kind of intervention in the wall, traces of which have disappeared since it was excavated several years ago.

Third, undated pieces are sometimes as interesting as the dated pieces. Note in the catalogue that there are no undated pieces with 100+ rings, seven with 50+ rings (which may some day be crossdated), and only two with 50- rings (which have almost no chance of ever being crossdated). Work on newly-collected samples from the WarSama Sarayl continues.

The heartwood-sapwood boundary, so often useful in determining the nearness of the terminal growth-ring in oaks, is less useful in conifers and completely indistinguishable in the charcoal which makes up almost 100 of our Bronze Age dendrochronological material.

KULTEPE. WARSAMA S A R A Y I

1145

800 1234

900

7 63

1000

1190

1092 1173

I - - = sapwood

-- - h e a r t wood

b a r k or- t e r m i n a l +I= growth r i n g

A= l a s t p r e s e r v e d r i n g

1082 1173

1020 1172

1038 1172

N m P

998

CFI m

1172

2 P

1077

E

1169

r

969 1121


A C E M H W K , SARIKAYA PALACE, CATALOGUE


SARIKA YA PALACE, Second level, CROSSDATED SAMPLES (for room locations see notations on plan from Belleten 1979). AU samples are carbonized: Collected 10 October 1973.

ACM- 5

ACM 6

ACM- 7

ACM-- 8

Grid No. PA/46, Room 29. Cedrus libani, sp. (Dr. W. H. Schoch). Two beam fragments.

A = 171 * 969 - 1 1 3 9 ~

B = 222 *lo08 - 1 2 2 9 ~

Grid No. PA/46, Room 29. Cedrus libani sp. (Dr. W. H. Schoch). Beam fragment, (last ring present on all that could be seen of the beam.)

A = 275 * 964 -1238~

Grid No. PA/46, Room 29. Cedrus libani Loud. (Dr. B. Aytug). Charcoal beam fragment.

A = 182 *lo35 -1239~

Grid No. PA/46, Room 29. Stored in room 17, Grid No. NA/45. Cedrus libani Loud. (Dr. B. Aytug). Charcoal lump. Charred outer face.

A = 100 *lo42 - 1 1 4 1 ~

Collected from the depot of the Ankara Museum 19 February 1974:

ACM - 12 Cross-section of charcoal log. Cedrus libani, sp. (Dr. W. H. Schoch). Unknown provenience.

A = 254 * 910 - 1 1 6 3 ~

*These six cedar pieces form a tightly fitting master chronology called ACM - 666, but because of the idiosyncratic nature of cedar they have been averaged separately from the pieces in the ACM - 999 Master composed of other conifers from Acemhuyiik. The problems which ACM - 6 and 7 cause with the interpretation of the Acemhuyiik wood are discussed below. NOTE: All Acemhuyiik dates are based on the Kultepe Relative Dating system.

Collected 10 August 1974:

ACM - 13

ACM - 14

ACM - 15

N side of palace, 3rd room from E, doorway to E-W hall between Rooms 1 and 4. E reveal of door. Half-section, juniperus sp. L. (Dr. B. Aytug).

AB = 156 1076 - 1 2 3 1 ~

See ACM - 13. Also from E reveal of door. Juniperus sp. (Dr. W. H. Schoch), Three-quarter section.

AB = 126 1104 -1229~

See ACM - 13. Also E reveal of door. Juniperus sp. L. (Dr. B. Aytug). Quarter-section. Squared on two sides.

AB = 142 1090 - 1 2 3 1 ~

286 PETER IAN KUNIHOLM - MARYANNE W. NEWTON

A C E M H W K , HATIPLER TEPESI, CATALOGUE


HA TIPLER TEPESI, L A R G E BURNED BUILDING, CROSSDA TED SAMPLES:


ACM - 17 Grid No. FA/29. Threshold between rooms 1 and 2. Door jamb, N reveal. (Tenon in side = ACM - 17B). Pinus nigru / silvestris (Dr. W. H. Schoch).

A = 95 1118 - 1 2 1 2 ~

ACM - 18 Grid No. FA/30. Room 3. SE comer of grid near wall (party wall with Room 2) Beam beds run N-S. “DoSeme t ah ta l am izlerinin oldugu oda.” Pinus silvestris L. (Dr. B. Aytug). Section.

A = 67 1165 -1231~

ACM - 20 Grid No. HA/GA-28. Room 23. Half-section. Charcoal. Round side down, = ?threshold. Ring 64 complete all around arc. Pinus nigrulsilvestris (Dr. W. H. Schoch).

A = 64 1168 -1231~


ACM - 21 Grid No. HA/28. Room 23. West side of room. Westemmost transverse beam (of 15) projecting N-S from S wall. Footing for S wall of room. Ju- niperus sp. L. (Dr. B. Aytug and Dr. W. H. Schoch).

A = 64

E = 156

1166 -1229~

1076 -1231~

ACM - 22 Random from HTP. Room 24. Not in situ when recovered. Probably a continuation of N-S wall in GA/28 and GA/29. Radial section. Pinus nigru / silvestris (Dr. W. H. Schoch).

A = 65 1166 -1230~

ACM - 23 Grid. No. GA/30. NE comer of Room 3. N wall <,Jundation, SW comer of wall. Cross-section. Buxus sp. (Dr. W. H. Schoch), ?ark preserved.

A = 207 1024 -1230B

ACM - 24 SE corner of HTP palace. Entryway from S. Beams laid parallel for entryway. Ring 158 on one radius only. Cedrus libani Loud. (Dr. B. Aytug).

A = 158 1074 -1231~


DISCUSSION:

Sankaya and Hatipler Tepesi have a large number of timbers which were cut in Relative Year (Kultepe Dating System) 1231 (see fig. 2). In spite of the disastrous fire which destroyed these structures, at least ten pieces have a last preserved ring ranging from 1229-1231. Apparently the practice of the Acemhiiyiik carpenters was to peel their timbers but not to shape them in any radical fashion. Our first conclusion, then, is that both structures at Acemhiiyiik were built in the same year. Prof. Nimet Ozgiic now informs us that she thinks that they may be wings of the same palatial building with Sankaya serving as the “palace” and Hatipler Tepesi serving as a storage or support building. Two cedar pieces, ACM - 6 and 7 from Room 29, have last rings in 1238 and 1239 re- spectively. We do not have a ready explanation for the 7-8 year discrepancy other than that Room 29 may have undergone a modification 7-8 years after the basic construction.

NON-CHRONOLOGICAL COMMENT ABOUT WOOD SPECIES:

Of the 35 samples which have received specialist identification, 12 are pine, 12 are juniper, 9 are cedar, 1 is boxwood, and 1 is oak. Pines and junipers thus represent two-thirds of the wood used in construction at Kiiltepe and Acemhiiyiik. Note the remarkable absence of oak, so common in medieval structures in Anatolia.

Cedars are always of interest for the Bronze Age because of the vast quantity of cedar wood presumably exported from the Levant to Egypt and preserved in Egyptological collections around the world. The 330 year Cedar of Lebanon chronology from Sankaya immediately becomes some- thing with which we can try to match Egyptian (= Lebanese) cedar wood. We are already at work on a comparative study of the growth characteristics of living cedars from Anatolia and Lebanon.

288

1058


1239

9 0 0

9 6 4 1238

9 4 2

- - = sapwood

= heart wood - bark or terminal growth ring

-I = last preserved ring

1227

1 0 2 4

947

Hatip 2 4 1074

1227

10

10

1118

Hatip 21E 6

Sarikaya 13 6

1212

Sarikaya 15 1090

1168 * 1165 w

910

Hatip 23 1 1

1163

n

1042

1166 I Hatip 2 1 A i

Sarikaya 1 4 1104

1141

Sarikaya 5 8 I

9 6 9

1008 1

1139

1231

23 1

23 1

23 1

1231

1 2 3 1

1230

1230

1229

1229

1229

Figure 2, Acemhiiyiik Bar Graph


A C E M H W K , SARIKAYA PALACE, CATALOGUE


SARIKAYA PALACE, second level, SAMPLES WHICH DO NOT CROSSDATE:

Collected 10 October 1973:

ACM - 3 Grid No. PA/42. Collected from Room 5 , original provenience lost. From fallen roof beams? Last ring is complete around circumference except for adzed area. Wrapped in hasir or matting. Pinus silvestris (Dr. B. Aytug).

A = 56

ACM - 4 Same room and context as ACM -3. Pinus silvestris (Dr. B. Aytug). Two cross-sections.

A = 39

B = 47

Collected from the Ankara Museum Depot 19 February 1974:

ACM - 11 End of charcoal beam. Full section. Chiselled holes for tenons (which are still in place) on opposite sides. Pinus nigra / silvestris (Dr. W. H. Schoch). Unknown provenience (but from Sankaya).

A = 91

Collected 10 August 1974.

ACM - 16 E side of Sankaya Palace, Room 7. Narrow ? hallway S of stone wall footing. Bark? on outside. Juniperus sp. (Dr. W. H. Schoch). Erratic growth.

A = 141

Collected 3 July 1981:

ACM = 27 Same provenience as ACM-28. Half of a whole fallen log. Erratic growth.

A = 91

ACM - 28 Recovered from Room 5. Original provenience lost, but from Sankaya. Fallen roof? beam. A = Section of carbonized whole log. Evidence of peel- ing/adzing on one side. B = small additional piece which may preserve terminal ring. Pinus nigru / silvestris. (Dr. W. H. Schoch). Erratic growth.

A = 103

B = Not measured

290

SAMPLE

H A TIPL E R D A TE: Collected 10

ACM - 19

ACM - 25


A C E M H W K , HATIPLER TEPESI, CATALOGUE PROVENIENCE RING RELATIVE DATES

First Last Year Year

TEPESI, L A R G E BURNED BUILDING, SAMPLES WHICH DO N O T CROSS-

August, 1974:

Grid No. FA/29. Beam end found lying loose in NE corner of Room 11. Beetle galleries on two opposite sides. Flat on one side. Bevelled at 45 O on other. Very crudely chopped. Juniperus sp. (Dr. W. H. Schoch).

No Grid No. NE room, near N wall (next to door to E-W gallery). Wood cross-section. Perhaps threshold between Rooms 12 and 23. Juniperus sp. (Dr. W. H. Schoch).

A = 85

A = 152

MISCELLANEOUS PIECES FROM A C E M H U Y U K FROM CONTEXTS OTHER THAN SARI- K A Y A A N D HATIPLER TEPESI

PRIVA TE HOUSE O N THE HUYUK: thought to be approximately contemporary with Kultepe Karum Ib) Collected 10 October 1973. From NW trench. ACM - 1 Charcoal beam section collected in the depot of the Nigde Museum. Cedrus

libani (Dr. W. H. Schoch).

Charcoal beam section collected in the depot of the Nigde Museum. From same log as ACM - 1.

A = 281 947 -1227~ ACM - 2

A = not measured

ACM - 9 Charcoal beam collected 19 February 1974 from the depot of the Ankara Museum. Remainder left in Ankara. From same log as ACM - 1.

A = 286 942 -1227~

DISCUSSION:

The last rings of ACM - 1 and 9 make the timber approximately contemporary with the San- kaya Palace. However, this is one timber only, and any data from such a small sampling should be treated with extreme caution.

Collected 3 July 1981:

ACM - 26 Grid No. U/55, West trench, Layer I11 (Excavator’s comment: same as Sankaya, 3rd Palace.) Possibly fallen from W side of building (as yet unex- cavated). Lies above floor.

A = 35 ACM - 29 West trench, N side, post, burned on top. Max. D .=. 285m. Max. pre-

served Ht .=. 55m. Juniperus sp. (Dr. W. H. Schoch). A = 190

Collected 12 July 1982 from M.T.A. Ankara: ACM - 31 Central trench, Level 11. Juniperus foetidissima. 2 FH cores from badly rot-

ted log. Not measured.


PORSUK/ULUKI$LA:

In the spring and summer of 1988, just as this report was being prepared, Prof. Olivier Pelon kindly provided us with a number of samples from his excavations at Porsuk/Ulukqla. We also thank Dr. Sylvestre DuprC for his help and commentary. Although the Porsuk work is far from complete, and a final Porsuk dendrochronological report can be written only after at least another season’s excavations and collection, it is clear that we have already a 321 year long master chronology from Porsuk that overlaps the 469 year long Kultepe/Acemhuyiik chronology by 113 years and extends it downward by 208 years. We therefore have added the measurements from Porsuk/Ulu- kqla to the Middle Bronze Age Master Chronology, making it 677 years long. See Table 1, below.

METHODS BY WHICH THE CHRONOLOGIES WERE MATCHED:

Study of the living conifer forests of Anatolia has shown us that it is possible to achieve syn- chronisms over long distances, as far as from Kars to the Aegean and even to central Greece. Thus, it is not surprising that we find excellent visual and statistical fits between our three sites. The.visual fit can be seen in the composite graph in Fig. 3. The statistics used to support our fits are as follows:

......................................................................................................................................................................................... Site A Site B Distance t-score Overlap Trend D-Score

Kultepe Acemhuyiik 160 km. 7.37 100 yrs. 65.7 O h 117.7

KUL/ACM Porsuk 120-180 6.65 113 yrs. 66.1 ‘/o 106.9

Acemhuyiik Porsuk 120 km. 6.60 113 yrs. 63.4 ‘/o 88.4

Kultepe Porsuk 180 km. 2.68 55 yrs. 64.8 O h n/a

.........................................................................................................................................................................................

NOTE: These are standard statistical tests. The trend coefficient or percentage of agreement is described by B. Huber in J. Fletcher, ed., Dendrochronology in Europe, BAR 51 (1978) 23. The t-test is described by M.G.L. Baillie and J.R.Pilcher, “A Simple Crossdating Program for Tree-Ring Research,” Tree-Ring Bulletin 33 (1973) 7-14. The D-Score (significant only if the overlap or n- is over 100 years) is described by B. Schmidt, Dendrochronologie und Ur- und Fruhgeschichte, Koln, Habilitation thesis, 1987, 13.

NOTE ON HOW READ THE TABLE OF TREE-RING INDICES:

Growth percentages for each year are shown in the left-hand table, and the numbers of samples from which each percentage was derived are shown in the right-hand table. For example, in the Middle Bronze Age Master Table which follows, Relative Year 1335’s ring-growth was 45% of normal, an average that was derived from 39 specimens from 39 trees. Note also that the last half of this table is more significant than the first half because it is based on a greater number of samples. For example, in Relative Years 792 and 793 ring growth was only 44% of normal, but this observa- tion is based on only one sample from one tree. When more wood is collected from the earlier cen- turies of this period and if the new samples show the same fluctuation in the same two years, then we will be able to say that this is indeed a significant deviation from normal growth. “Problem” pieces have been omitted from the indices, as were the cedars which have their own growth curve.

292

Kiiltepe, Acemhuyiik and Porsuk/Uluki$la

SUMMARY ID = 133951


Table 1

TREE RING INDICES FOR MIDDLE BRONZE AGE

DATE 0 1

763 0 0 770 113 108 780 102 172 790 85 66 800 93 110 810 98 104 820 103 120 830 81 80 840 66 86 850 97 149 860 149 113 870 123 154 880 34 49 890 26 60 900 65 140 910 27 45 920 160 126 930 132 129 940 110 35 950 183 164 960 75 45 970 174 127 980 93 94 990 125 103

1000 118 66 1010 99 90 1020 48 35 1030 42 72 1040 78 87 1050 112 101 1060 110 84 1070 143 66 1080 95 137 1090 126 130 1100 120 113 1110 79 115 1120 92 108 1130 91 89 1140 85 88 1150 123 88 1160 114 122 1170 92 99 1180 88 87 1190 79 68 1200 78 71 1210 127 93 1220 114 79 1230 101 68 1240 127 90 1250 92 108 1260 122 112 1270 111 119 1280 108 117 1290 140 119 1300 86 83 1310 96 87 1320 132 128 1330 100 109 1340 94 41 1350 84 103 1360 158 118 1370 63 78 1380 117 124 1390 55 79 1400 122 80 1410 138 85 1420 118 126 1430 89 72 1440 0 0

2 3

0 93 125 89 165 148 44 44

126 119 87 86 83 93 95 84 91 102

132 120 157 132 133 146 26 69 37 57

150 92 55 23

142 218 131 69 38 28

159 120 68 104 97 74

140 112 46 156

150 86 44 96 52 35 82 51 75 116 91 83

104 66 85 123 86 134 99 127

113 119 96 108

111 88 81 68 85 99 97 126

110 101 86 I8 97 95

101 109 106 96 125 87 109 97 74 61

110 75 99 115

125 109 68 87

130 136 113 89 52 78 96 88 67 114

116 142 85 82 85 55 92 133 84 69

122 96 117 49 76 78 92 86 54 74 91 127 0 0

4 5

74 81 67 107

114 66 51 76 65 97 94 28 98 115 62 68

147 100 121 136 134 90 70 97 92 85 60 112

100 104 55 26

212 153 65 77 24 43 98 100 67 80 87 84 91 111

163 75 54 84 43 109 44 81 54 39 99 78 52 62

100 134 174 136 104 86 128 104 117 122 76 110 93 103 88 103

110 102 68 83 65 65 94 87 61 93 96 128

121 128 90 129

126 131 62 87

128 125 155 146 112 101 84 104 81 116

101 93 51 70 86 106

119 115 126 45 60 87 50 110

108 114 71 52

123 114 102 65 85 108 91 98

101 111 146 145

0 0

6 7

120 104 105 105 73 81 93 108

130 89 46 79

107 94 111 92 112 149 92 105

102 99 107 110 51 37

123 123 107 43 39 92 40 58 62 9 45 68

101 123 118 159 35 83

114 111 72 146 86 70 62 72 58 44 97 101 81 88 54 71

117 77 101 97 72 107 99 67 90 118 87 120 94 81 82 77

124 123 89 121 77 98

110 153 89 88

111 79 86 38

105 103 146 137 51 86

132 112 108 104 97 97

103 75 138 56 90 75 81 108

135 152 98 55 47 95

101 105 144 183 143 128 87 80

114 123 60 95 86 82

156 75 76 117

148 160 0 0

8

107 98 82

143 137 102 98 97

149 158 122 144 30

150 85

152 83 77 96

130 162 142 69 85 88 58 76 77 60 81 80

138 130 136 112 108 85

101 102 118 98

115 100 69 73 76

142 97 88

140 90 74 91 99 59

135 73 95

116 175 80 91 77

120 116 125 109 179

0

9

89 92 83

126 104 153 100 91 22

157 114 107 41

119 50

25 1 118 112 159 130 156 105 78 62 74 69 40 57

107 73

114 167 125 115 107 104 97 83

125 107 72 91 82

102 99 77

131 101 156 110 71 78

111 87 62

125 92 73

121 172 96

103 100 102 114 95 79

217 0

NUMBER OF SAMPLES FOR MIDDLE BRONZE AGE

DATE 0 1 2 3 4 5 6 7 8 9

763 0 0 0 1 1 1 1 1 1 1 770 1 1 1 1 1 1 1 1 1 1 780 1 1 1 1 1 1 1 1 1 1 790 1 1 1 1 1 1 1 1 1 1 8 0 0 2 2 2 2 2 2 2 1 1 1 810 1 2 2 2 2 2 3 3 3 3 820 3 3 3 2 2 2 3 3 3 3 830 3 3 3 3 3 3 3 3 3 3 840 3 3 3 3 3 3 3 3 3 3 850 3 3 3 3 3 3 3 3 3 3 860 3 3 3 3 3 3 3 3 3 3 870 3 3 3 3 3 3 3 3 3 3 880 3 3 3 3 3 3 3 3 3 3 890 3 3 3 3 3 3 3 3 3 3 900 3 3 3 3 3 3 3 3 3 3 910 3 3 3 3 3 3 3 3 3 3 920 3 3 3 3 4 4 4 4 4 4 930 4 4 3 3 3 3 3 2 2 2 940 2 2 2 2 2 2 3 3 3 3 950 3 3 3 3 3 3 3 3 3 3 960 3 3 3 3 3 3 3 3 3 4 970 4 4 4 4 4 2 2 2 2 2 980 2 3 3 3 3 3 3 3 3 3 990 3 3 3 3 3 3 3 3 4 4

1000 4 5 5 5 5 5 5 5 5 5 1010 5 5 5 5 5 5 5 5 5 5 1020 6 6 6 6 6 6 6 6 6 6 1030 6 6 6 6 6 6 6 6 7 8 1040 8 8 8 8 8 8 9 9 9 9 1050 9 9 9 9 9 10 10 10 10 10 1060 10 10 10 10 10 10 10 10 10 10 1070 10 10 10 10 11 12 14 15 15 15 1080 16 17 19 19 19 19 19 19 19 19 1090 20 21 23 23 23 24 24 25 25 25 1100 26 26 26 27 28 29 29 29 29 28 1110 28 28 28 28 28 28 28 28 29 30 1120 30 29 28 28 28 28 28 28 28 28 1130 28 28 28 28 27 27 26 26 26 26 1140 26 25 25 25 26 26 25 25 25 25 1150 25 25 25 25 25 25 25 25 25 25 1160 26 26 25 25 25 25 27 27 28 29 1170 29 29 29 23 19 19 20 21 21 21 1180 21 21 21 21 21 21 21 21 21 21 1190 21 21 21 21 21 21 21 22 23 23 1200 23 23 23 23 23 23 23 23 23 23 1210 23 23 23 22 22 22 23 23 23 24 1220 24 26 25 26 27 27 27 28 29 27 1230 24 20 15 15 15 15 15 15 15 16 1240 21 20 19 19 19 19 19 19 18 17 1250 17 17 17 18 19 18 20 19 18 19 1260 19 19 21 22 21 21 21 22 22 23 1270 23 24 26 26 27 27 27 27 28 29 1280 28 30 29 28 30 30 28 28 28 28 1290 28 28 29 29 29 30 30 30 31 30 1300 30 29 30 30 30 30 30 30 31 32 1310 33 33 34 34 34 37 38 38 40 41 1320 39 40 40 39 41 42 42 42 41 41 1330 40 39 39 40 40 39 39 38 38 39 1340 38 38 39 39 39 39 38 38 39 39 1350 39 39 40 40 41 41 41 39 38 37 1360 35 33 32 30 29 29 28 27 27 27 1370 27 27 26 25 24 23 21 20 21 22 1380 21 21 21 21 22 21 21 21 21 21 1390 21 21 21 20 20 20 20 20 20 20 1400 20 19 19 18 18 18 18 17 17 17 1410 17 17 17 16 16 16 15 15 15 15 1420 15 14 14 14 12 12 12 12 9 8 1430 8 5 4 4 4 4 3 3 3 2 1440 0 0 0 0 0 0 0 0 0 0


RELATIVE OR COMPARATIVE CHRONOLOGIES AND DATES: We now have a 677 year long skeleton to which we hope to attach the flesh of Middle Bronze

Age history. As new information becomes available, our challenge will be to see where specific Bronze Age events fit in. The approximate time-scale of this combined chronology from the three sites is from the Early Bronze Age or the 23rd century BC (the birthdate of the oldest sample at Kultepe) to the 16th century BC or the time of the Old Kingdom Hittites (the latest rings at Por- suk/Ululugla). This is a non-dendrochronological estimate, based on the finding of bullae, in the destruction of the Sarkaya Palace, from the early part of the reign of Samgi-Adad I. Four of these have been attributed to his tenth year (Prof. Nimet Ozgii~, pers. comm.) This year of Samgi-Adad’s becomes the terminus ante quem for our Relative Year 1231. If the Acemhiiyiik buildings had the century or century-and-a-half life-span that Prof. Nimet Ozgii~ has suggested, the effect would be to move the entire dendrochronology upward (earlier) by that amount of time so that it would begin in approximately the 24th century and end in the 17th century. If more wood from repairs at the continuing excavations at Acemhuyiik can be found, we will have an idea of how much earlier the absolute chronology should be adjusted. At any rate, the parts of the whole dendrochronology will move upward or downward together.

Moreover, a glance at the composite graph (Fig. 3) and the 68 year difference between the building years at Kultepe and Acemhuyiik also raises the question as to whether the Wargama Saray could have been built as early as Kultepe I1 and destroyed during Kultepe Ib.

We are informed by Messrs. Pelon and DuprC that an Old Kingdom Hittite date for the Porsuk destruction from which our charcoal comes is probable. We have not yet found either bark or terminal rings there. The continuing work at Porsuk wiU no doubt add useful information to what we already have in hand, especially if the data include bullae and tablets attributable to specific kings.

The gap between the end of the 677 year Middle Bronze Age tree-ring chronology and the be- ginning of the 806 year Gordion chronology is now approximately one century or less. We had hoped that wood from Magat might fiU part of the gap, but unfortunately it was not long-lived. The best hope appears to be the latest Hittite levels at Porsuk. When the chronologies are bridged we will have a continuous sequence that is over a millennium and a half long, covering the period from Early Bronze into the Iron Age.

ABSOLUTE RADIOCARBON DATES FOR NUMBERED RINGS FROM KULTEPE AND

In the absence of a continuous tree-ring chronology which extends back from the present to the Middle Bronze Age, we thought it would be worth trying to “wiggle-match” the Kultepe/ Acemhuyiik ring-sequence with the radiocarbon time-scale. In theory, it should be possible to pin the sequence down to within +20 years. Seven samples of numbered rings spread over the length of the master chronology were sent last spring for radiocarbon analysis to Dr. Bernd Kromer at the University of Heidelberg and more are going to him now that the chronology has been extended, but as of this writing only three of the tests are complete. A joint report wiU be published as soon as the data become available.

A C E M H ~ K

ACKNOWLEDGMENTS: The authors thank the General Directorate of Antiquities and Museums for general research

permission, and the excavators for invitations to their sites.

For assistance in the laboratory in sample preparation, measurement, and analysis we thank Ca- rol B. Griggs, Hope E. Kuniholm, Shana L. Tarter, Robert S . Pohl, Jane E. Reagen, and Roger P. Boissonnas.

For financial support we thank the National Endowment for the Humanities, the Institute for Aegean Prehistory, the National Geographic Society, the Samuel H. Kress Foundation, the David and Lucile Packard Foundation, the National Science Foundation, and a group of private contribu- tors.

Documents

A 677 Year Tree-ring Chronology for the Middle Bronze Age