1

SOME CORRECTIONS BY STURT W. MANNING TO:

Manning, S.W. and Kromer, B. 2011. Radiocarbon Dating Iron Age Gordion and the Early Phrygian

Destruction in Particular. In C.B. Rose and G. Darbyshire (eds.), The New Chronology of Iron Age Gordion:

123-153. Gordion Special Studies VI. Philadelphia: University of Pennsylvania Museum of Archaeology

and Anthropology.

This document lists some corrections (typos, or mistakes, or things missed between original paper and

publication) to the above chapter (hereafter ‘the printed text’). I acknowledge that these issues were spotted and

raised by Douglas J. Keenan in an email of 1 February 2012. This present document is dated 2 February 2012.

The items noted at this time are as follows:

1. Table 6.1. Unfortunately the last line (row) of the original Table was not printed. Thus one 14

C date (Hd-23655)

is not listed there. There are thus in all 15 14

C dates on short-lived samples (barley, lentils or flax seeds) from the

YHSS 6A Destruction Level at Gordion as shown in Figures 6.3a and 6.3b of the printed text. Another

inconsistency/typo is that the quoted calibrated age ranges were meant to be rounded to the nearest 5 years, but

this was not done for Hd-23638 in the printed table (it should be 895-800). Finally, it should be made clear that

the calibrated age ranges are dates BC. A corrected version of Table 6.1 is provided below. Corrections are

highlighted in yellow.

Hd Lab No. Sample Context Sample Code Material Dated

Gordion YHSS Phase

14C Age BP SD δ13C

Calibrated Age BC at 2σ

21191 CMaG-YH 30416 Barley 7A 2699 18 -23.24 900-810

21229 CMaG-YH 59024 Charcoal 7A 2872 35 -23.91 1200-920

21231 CMaG-YH 59025 Charcoal 7A 2808 26 -24.48 1040-890

21236 CMaG-YH 33521 Lentils 6A 2641 25 -22.28 840-785

21304 CMaG-YH 33555 Flax seeds 6A 2655 19 -25.47 840-795

21310 CMaG-YH 59026 Pine charcoal 7A 2823 33 -23.65 1090-890

21311 CMaG-YH 59029 Pit 93 Frags Oak & Pine Late 5A 2462 33 -24.39 760-410

21318 CMaG-YH 33335 Bitter Vetch 7A 2728 29 -23.53 930-810

21327 CMaG-YH 59023 Pine charcoal 7A 2789 24 -23.57 1010-850

21332 CMaG-YH 32525 Barley 6A 2678 18 -22.96 895-800

21352 CMaG-YH 32385 Barley 6A 2674 20 -23.24 895-795

21354 CMaG-YH 33379 Wheat 7A 2726 24 -22.38 920-815

21358 CMaG-YH 33243 Lentils 6A 2650 34 -22.95 900-780

21361 CMaG-YH 59028 Roof thatch/ reeds= Phragmites? 6A 2692 26 -24.21 900-800

21364 CMaG-YH 59050 Pine charcoal 7A 2742 16 -24.12 920-830

21866 YH 33394 Op 10, 37 Lot 90 33394 Wheat 7A 2733 23 -20.27 925-820

21867 YH 33402 Op 10, 25,Lot 87 33402 Wheat 7A 2760 22 -20.75 980-830

21880 YH 59027 GOR 1999 59027 1A Roof thatch/ reeds= Phragmites? 6A 2686 21 -24.14 895-800

21885 YH 32138 32138 Lot Roof thatch/ reeds= Phragmites? 6A 2703 29 -24.95 910-800

23609 YH 33554 B 33554 Barley 6A 2669 12 -22.68 835-800

23610 YH 33575 B 33575 Lentils 6A 2683 12 -22.14 895-800

23611 YH 33243 B 33243 Lentils 6A 2688 14 -22.80 895-805

23612 YH 33568 B 33568 Barley 6A 2680 16 -22.57 895-800

23624 YH 33568 33568 Barley 6A 2686 15 -22.56 895-800

23638 YH 33554 33554 Barley 6A 2683 12 -22.80 895-800

23639 YH 33555 33555 Flax seeds 6A 2714 11 -24.98 900-820

23644 YH 33243 33243 Lentils 6A 2673 12 -22.66 840-800

23645 YH 33575 33575 Lentils 6A 2694 13 -22.17 895-805

23655 YH 33555 B 33555 Flax seeds 6A 2707 11 -25.26 895-800

Table 1. Printed text Table 6.1 (p.133) REVISED. Revisions highlighted in yellow.

2

2. p.135. Left Column. Line 29. Typo. The number of barley samples is 6 (and not 7 as in the printed text). On

behalf of the two authors I apologize for failing to spot this typo. The weighted averages quoted and calendar age

ranges remain as printed (except on some runs the 2σ (95.4% probability) calibrated range for the barley might

change by 1 year, starting 834 BC, as in Figure 2 below, versus 835 BC as on p.135 of the printed text). Figures 1

and 2 below show the respective weighted averages and calibrated ranges in detail. The OxCal 3.10 runfile and

data used are:

Plot { Phase "Gordion TB2A Destruction YHSS6A" { R_Combine "All Lentils" { R_Date "Hd-21358" 2647 32; R_Date "Hd-23644" 2673 12; R_Date "Hd-23611" 2688 14; R_Date "Hd-23645" 2694 13; R_Date "Hd-23610" 2683 12; R_Date "Hd-21236" 2641 25; }; R_Combine "All Barley" { R_Date "Hd-23638" 2683 12; R_Date "Hd-23609" 2669 12; R_Date "Hd-23624" 2686 15; R_Date "Hd-23612" 2680 16; R_Date "Hd-21352" 2674 20; R_Date "Hd-21332" 2678 18; }; }; };

Figure 1. The weighted average of the 6 radiocarbon measurements on lentil samples from YHSS 6A (Destruction

Level) contexts: calibrated calendar probability distribution and calibrated age ranges. The IntCal04

radiocarbon calibration curve is used (as in the Manning and Kromer printed text) with curve resolution set at 1.

Atmospheric data from Reimer et al (2004);OxCal v3.10 Bronk Ramsey (2005); cub r:1 sd:12 prob usp[chron]

950CalBC 900CalBC 850CalBC 800CalBC

Calibrated date

2550BP

2600BP

2650BP

2700BP

2750BP

2800BP

Radio

carb

on d

ete

rmin

ation R_Combine All Lentils : 2680.08±6.03192BP

68.2% probability 827.4BC (68.2%) 812.4BC 95.4% probability 835.6BC (95.4%) 807BC X2-Test: df=5 T=5.4(5% 11.1)

3

Figure 2. The weighted average of the 6 radiocarbon measurements on barley samples from YHSS 6A

(Destruction Level) contexts: calibrated calendar probability distribution and calibrated age ranges. The

IntCal04 radiocarbon calibration curve is used (as in the Manning and Kromer printed text) with curve

resolution set at 1.

The above Figures report results from OxCal 3.10 which was the current version of this software when the

Manning and Kromer chapter was written. They also use IntCal04 which was the then standard radiocarbon

calibration curve (for references see the Manning and Kromer printed text pp.150-153). Today (2 February 2012)

OxCal 4.17 is the current version of the software, and the current international radiocarbon calibration dataset is

IntCal09 (which for the relevant time period uses the same data as IntCal04). Figures 3 and 4, for interest, show

the data in Figures 1 and 2 repeated with OxCal 4.17 and IntCal09 and with curve resolution 1. The outcomes are

almost identical, with variations in calculations of no more than 1 year in total (either radiocarbon years, or

calendar age ranges).

Figure 3. As Figure 1 but with OxCal 4.17 and IntCal09.

Atmospheric data from Reimer et al (2004);OxCal v3.10 Bronk Ramsey (2005); cub r:1 sd:12 prob usp[chron]

950CalBC 900CalBC 850CalBC 800CalBC

Calibrated date

2550BP

2600BP

2650BP

2700BP

2750BP

2800BP

Radio

carb

on d

ete

rmin

ation R_Combine All Barley : 2678.21±5.9948BP

68.2% probability 826.5BC (68.2%) 811.7BC 95.4% probability 834.3BC (95.4%) 806.8BC X2-Test: df=5 T=1.1(5% 11.1)

4

Figure 4. As Figure 2 but with OxCal 4.17 and IntCal09.

3. To confirm and clarify (in view of the correction to printed text Table 6.1 reported in 1. above): the total

number of radiocarbon dates on short-lived samples (barley, flax seeds or lentils) from the YHSS 6A Destruction

Level reported in the Manning and Kromer printed text is 15: comprising 6 dates on barley samples, 6 dates on

lentil samples, and 3 dates on flax seed samples, as shown in the printed text in Figures 6.3a and 6.3b. And as

listed in the revised version of Table 6.1 provided in 1. above as Table 1. The barley and lentil sets both provide

consistent ages and combine to offer the weighted averages and calibrated age ranges shown in Figures 1-4 above.

The three dates on flax seeds vary a little more. Two are a little older than the other flax seed date, and all 12 of

the dates on barley and lentils from YHSS 6A. A plot showing all 15 of the radiocarbon dates on the short-lived

samples from YHSS 6A is shown in Figure 5 using OxCal 4.17 and IntCal09. The calibrated age ranges for each

of the individual dates and the two weighted averages shown are listed below in the Appendix. The weighted

average and calibrated calendar age ranges from (i) all the barley and lentil samples (n=12); and (ii) all 15 dates

are shown in detail in Figures 6 and 7. Figure 8 shows the data in Figure 6 allowing for the regional radiocarbon

offset factor, ΔR, discussed in the printed text p.146 (of 15.4±11.3 14

C years).

Conclusions:

The 15 radiocarbon dates on short-lived sample material, which should provide dates close to, or very close to, or

contemporary with, the time of the YHSS 6A Destruction Level all favour a date range in the 9th century BC.

Only 4 of the 15 dates even include any date range at 95.4% probability after 800 BC, and even then these date

ranges end (rounded to whole years) 799 BC, 797 BC, 789 BC and 786 BC. Two dates on flax seeds (Hd-23639,

Hd-23655) seem to indicate slightly older material (versus the other 13 data) but even then their age ranges

include calibrated calendar dates down to 824 BC or 818 BC – and hence these data are potentially compatible

with the other 13 data. The similarity of the dates in this overall YHSS 6A Destruction Level set is consistent with

their all coming from the same destruction horizon and providing a close dating for this. If we use the weighted

average of all 15 data, the likely (94.5% probability) calendar age range is 839-808 BC (see Figure 7). If we

exclude the flax seeds on the basis that two (of three) are seemingly perhaps a little older – and so we try to get as

late a date as possible – then the weighted average of the 12 data on barley and lentils gives a 95.4% probability

range of 834-808 BC (Figure 6). If we further allow for a possible regional radiocarbon offset (on the basis of data

available from a detailed radiocarbon study of tree-ring series from Gordion) as perhaps affecting these samples –

5

see discussion in the printed text p.146 – then the adjusted weighted average for the 12 dates on barley and lentil

samples then yields a 95.4% probability range of 832-802 BC. And so, as argued in the printed text, the evidence

is consistent with a date for the YHSS 6A Destruction Level around or very shortly after the last few decades of

the 9th century BC – and so the new revised Gordion Iron Age chronology presented in Rose and Darbyshire

(2011): The New Chronology of Iron Age Gordion.

Finally, I consider an up-dated revision to the analysis shown in the printed text Figures 6.10 and 6.11. I revise

this sequence analysis in two ways: (i) I include the possible regional radiocarbon offset as in Figure 8 of this text;

and (ii) I treat the YHSS 6A Destruction Level set with a Tau_Boundary model in OxCal (see Bronk Ramsey C.

2009. Bayesian analysis of radiocarbon dates. Radiocarbon 51(2):337-60), with the subsequent boundary

(labelled as “E” in Figure 9) offering an estimate of the date of the actual destruction (shortly or very shortly after

the date of the plant material found in the destruction). For this analysis, see Figure 9, and for a detail showing the

“E” Boundary which models the date range of the YHSS 6A Destruction Level, see Figure 10. The analysis offers

a very good overall agreement (e.g. OxCal Amodel value of 147.5), with just one date, Hd-21236, failing to offer

a satisfactory individual OxCal agreement value (54.9 < 60) – and this is a very minor disagreement for a

complex analysis. The modelled calendar age range for the YHSS 6A Destruction Level from this analysis (using

the Tau_Boundary model for YHSS 6A and using the possible regional radiocarbon offset of 15.4±11.3 14

C years)

is 825-803 BC at 95.4% probability (see Figure 10). This finding again clearly supports and requires the New

Chronology of Iron Age Gordion.

(Figures 5-10 and the Appendix follow)

6

Figure 5. Plot showing the calibrated calendar age probabilities, and the calibrated age ranges at 68.2% and

95.4% probability (upper and lower lines under each histogram), for all 15 radiocarbon data on short-lived

samples (barley, flax seeds and lentils) from the YHSS 6A Destruction Level. All calendar Ages are dates BC (or

CalBC) The calibrated calendar age probabilities and ranges for the weighted average of the 12 data on barley

and lentils, and for all 15 data, are shown at the bottom of the figure – for more details on these see Figures 6

and 7.

7

Figure 6. The weighted average of the 12 radiocarbon measurements on barley and lentil samples from YHSS 6A

(Destruction Level) contexts: calibrated calendar probability distribution and calibrated age ranges. Data from

OxCal 4.17 and IntCal09 with curve resolution set at 1.

Figure 7. The weighted average of the 15 radiocarbon measurements on short-lived sample matter (barley, flax

seeds and lentils) from YHSS 6A (Destruction Level) contexts: calibrated calendar probability distribution and

calibrated age ranges. Data from OxCal 4.17 and IntCal09 with curve resolution set at 1. (Compare with Figure

6.16 in the printed text – the difference is the curve resolution in Figure 6 here is 1, whereas the curve resolution

is 5 in Figure 6.16 A in the printed text.)

8

Figure 8. As Figure 6 but with a ΔR adjustment of 15.4±11.3

14C years – see the printed text p. 146 for discussion

and explanation. (Compare with Figure 6.16 B in the printed text - the difference is the curve resolution in Figure

8 here is 1, whereas the curve resolution is 5 in Figure 6.16 A in the printed text.)

9

Figure 9. Iron Age sequence analysis for Gordion. This is slight revision (see text above for explanation) of the

analyses shown in the printed text Figures 6.10 and 6.11. The dendro dates, approximate ±5 years, come from

Manning et al. (2001; 2010 – for references, see the printed text pp.150-153). For the “E” Boundary (the date of

the modelled date range for the YHSS 6A Destruction), see Figure 10.

10

Figure 10. Detail of Boundary “E” from Figure 9 offering an approximate date for the actual YHSS 6A

Destruction.

Appendix: Calibrated calendar age ranges BC for the 15 radiocarbon dates on short-lived samples from

the YHSS 6A Destruction Level (see Table 1 above – revised version of Table 6.1 in the printed text), and

for the weighted average of the barley and lentil samples (n=12), and for the weighted average for all 15

samples. Data from OxCal 4.17 and IntCal09 with curve resolution set at 1.

OxCal v4.1.7 Bronk Ramsey (2010); r:1 Atmospheric data from Reimer et al (2009); Gordion YHSS6A, Destruction Level, TB2A, Short-Lived Samples, n=15 All Barley, n=6 Hd-23638 R_Date(2683,12) 68.2% probability 831.4BC (68.2%) 811.5BC 95.4% probability 892.5BC ( 6.3%) 875.9BC 842.8BC (89.1%) 804.5BC Hd-23609 R_Date(2669,12) 68.2% probability 823.5BC (68.2%) 807.5BC 95.4% probability 833.7BC (95.4%) 802.5BC Hd-23624 R_Date(2686,15) 68.2% probability 835.4BC (68.2%) 810.3BC 95.4% probability 894.6BC (12.8%) 868.6BC 861.7BC ( 1.5%) 855.5BC 847BC (81.2%) 804.6BC Hd-23612 R_Date(2680,16) 68.2% probability

831.5BC (68.2%) 809.5BC 95.4% probability 894.2BC ( 8.8%) 870.1BC 845.8BC (86.6%) 801.8BC Hd-21352 R_Date(2674,20) 68.2% probability 831.1BC (68.2%) 806.4BC 95.4% probability 894.5BC ( 8.8%) 869.1BC 859.2BC ( 0.1%) 858.5BC 847BC (86.4%) 798.9BC Hd-21332 R_Date(2678,18) 68.2% probability 832.1BC (68.2%) 808BC 95.4% probability 894.5BC ( 9.5%) 869.1BC 860BC ( 0.3%) 858.2BC 846.7BC (85.6%) 800.7BC All Flax Seeds, n=3 Hd-21304 R_Date(2655,19) 68.2% probability 820BC (68.2%) 801.5BC 95.4% probability 834BC (95.4%) 796.5BC

11

Hd-23639 R_Date(2714,11) 68.2% probability 894.4BC (34.9%) 869.4BC 849.5BC (33.3%) 829BC 95.4% probability 899.3BC (95.4%) 823.5BC Hd-23655 R_Date(2707,11) 68.2% probability 893.4BC (26.2%) 873.7BC 846.3BC (42.0%) 822BC 95.4% probability 896BC (95.4%) 817.6BC All Lentils, n=6 Hd-21358 R_Date(2647,32) 68.2% probability 826.4BC (68.2%) 796.5BC 95.4% probability 894.6BC ( 6.4%) 868.4BC 859.6BC ( 0.6%) 855.3BC 848.5BC (88.4%) 785.5BC Hd-23644 R_Date(2673,12) 68.2% probability 825.9BC (68.2%) 808.8BC 95.4% probability 837.2BC (95.4%) 802.4BC Hd-23611 R_Date(2688,14) 68.2% probability 836.5BC (68.2%) 810.8BC 95.4% probability 894.6BC (13.7%) 868.5BC 861.9BC ( 1.6%) 855.4BC 846.9BC (80.1%) 805.7BC Hd-23645 R_Date(2694,13) 68.2% probability 886.2BC ( 3.1%) 883.4BC

840.6BC (65.1%) 812.1BC 95.4% probability 894.9BC (19.6%) 867BC 863.3BC ( 3.6%) 853.3BC 848.5BC (72.1%) 808.4BC Hd-23610 R_Date(2683,12) 68.2% probability 831.4BC (68.2%) 811.5BC 95.4% probability 892.5BC ( 6.3%) 875.9BC 842.8BC (89.1%) 804.5BC Hd-21236 R_Date(2641,25) 68.2% probability 816.4BC (68.2%) 796.5BC 95.4% probability 835.8BC (95.4%) 789BC All Barley and Lentil Data, n=12, Weighted Average All Barley & Lentil data R_Combine(2679.14,4.25203) 68.2% probability 826.5BC (68.2%) 812.5BC 95.4% probability 833.9BC (95.4%) 807.6BC X2-Test: df=11 T=6.5(5% 19.7) All 15 Short-Lived Sample Data, Weighted Average All 15 data R_Combine(2685.23,3.66103) 68.2% probability 829.6BC (68.2%) 814.6BC 95.4% probability 885.4BC ( 0.9%) 882.4BC 838.3BC (94.5%) 807.8BC X2-Test: df=14 T=21.9(5% 23.7)

Sturt W. Manning, 2 February, 2012.