Norse Platters from Bornais, by Clara Freer & Ian Dennis, Cardiff University

Norse Platters from BornaisClara Freer & Ian Dennis

Bornais, South Uist • Large settlement• Spans 5th cent AD to 15th

cent AD• Numerous separate area

excavations

BORNAIS

Bornais, South Uist

• Mounds 2 and 2A main excavation areas

Bornais, South Uist

• Mound 2 focus of the settlement– Houses excavated

Objectives

• Understand the nature of platter production

• Consider manufacturing constraints (including the amount of clay in weight used, the temper, size, styles and different forms found on the Bornais excavations)

• Investigate their use as baking plates, how effective they are and to examine their durability

• Investigate gendered production through an examination of the size of the finger print impressions found on platters from Bornais

• Examine and compare our platters to the ones from Bornais

Various Platter Sherds from Bornais

Organic impressions on the underneathSmooth sand underneath, with finger impressionsvisible

Making the Platters

Weighing the clay and temper, platter 1 had a 50% ratio. 40 to 50% temper was noted in all of the Bornais platters examined.

Shaping the platter. It was noted that the Bornais platters had an average thickness of 8mm

Sand & chaff from barley was used to stop the platter sticking to the surface, again this was evident on the Bornais platters with organic impressions on the underneath.

Perforating the platter with antler pin.

The Making Process, Observations & Temper

Inclusion frequency chart for temper

The main outcrops of the Lewisian complex are on the islands of the Outer Hebrides, including Lewis, from which the complex takes its name. These rocks are largely igneous in origin, mixed with metamorphosed marble, quartzite and mica schist with later intrusions of basaltic dykes and granite magma.

Making Process, Observations & Temper

• Fine temper with gneiss, 40% temper

• Coarse Temper with large inclusions of gneiss, 50% temper

Temper from the Bornais platters was examined.

Temper Test Plates

Platter # Clay weight

Temper weight

Temper %

Total wet weight

Diameter X (wet)

Diameter Y (wet)

Total dry weight

Diameter X (dry)

Diameter Y (dry) % Shrink (av) Wght loss

Test platter 01 100g 10g 10% 110g 118mm 111mm 87.9g 111mm 103mm 7.0% 20.0%

Test platter 02 100g 20g 20% 120g 113mm 114mm 96.6g 108mm 109mm 4.5% 19.5%

Test platter 03 100g 30g 30% 130g 124mm 116mm 106.9 118mm 111mm 4.7% 17.8%

Test platter 04 100g 40g 40% 140g 126mm 125mm 116.9g 117mm 121mm 5.2% 16.5%

Test platter 05 100g 50g 50% 150g 132mm 130mm 126.8g 127mm 124mm 4.3% 13.4%

Test platter 06 100g 60g 60% 160g 133mm 132mm 138.8g 128mm 127mm 3.8% 13.3%

Lewisian Gneiss stone used as temper , we replicated this by using ground gneiss from Bornais.

Mini platters made to test the temper and shrinkage when dry. 10% to 60% temper.

10% 20% 30%

40% 50% 60%

Average shrinkage of the mini platters was 5% and the average weight loss was 17%

Finger Print Testing

Platter # Clay weight

Temper weight

Temper %

Total wet weight

Diameter X (wet)

Diameter Y (wet)

Total dry weight

Diameter X (dry)

Diameter Y (dry) % Shrink (av)

Fingerprints 01 100g 50g 50% 141mm 131mm 136mm 127mm 3.3%

Ian FP 01 14.6mm 13.5mm 13.9mm 12.8mm 5.0%Ian FP 02 17.3mm 13.4mm 16.1mm 12.6mm 6.5%Ian FP 03 16.8mm 14.2mm 15.2mm 12.2mm 7.1%Ian FP 04 15.2mm 11.7mm 14.7mm 10.5mm 6.8%

Total Ave Shrink 6.4%

Kirsty FP 01 13mm 12.1mm 11.5mm 10.7mm 11.5%Kirsty FP 02 12.3m 11.5mm 11.3mm 10.1mm 10.2%Kirsty FP 03 10.5mm 10.4mm 8.8mm 9.4mm 12.9%Kirsty FP 04 9.4mm 9.4mm 8.1mm 8.6mm 11.3%

Total Ave Shrink 11.5%

Clara FP 01 13.7mm 12.5mm 11.5mm 10.2mm 17.6%Clara FP 02 14.9mm 13.4mm 13.4mm 11.7mm 11.4%Clara FP 03 14.4mm 11.8mm 12.9mm 9.7mm 14.2%Clara FP 04 12.5mm 14.2mm 11.1mm 11.8mm 14.1%

Total Ave Shrink 14.3%

Illustration of a Bornais platter showing finger impressions from Mound 2 SF(3846)

3 hands used in the test, one male and two female to see if there was a size variation

Shrinkage was also noticed in the finger imprints with the average being 14.3%

The results show that it was difficult to discern male and female finger imprints on the test platters.

Platter 1 Platter 2

Platter 3 Platter 4

Finished platters. All the platters were left for 7 days to dry at room temperature before using.

Platter 1 had an average thickness of 6mm (not used due to cracking).Platter 2 had an average thickness of 12mmPlatter 3 had an average thickness of 7mmPlatter 4 had an average thickness of 7mm

Average weight loss 10.3%Average diameter loss 4.9%

Platter #Clay

weightTemper weight

Temper %

Total wet weight

Diameter X (wet)

Diameter Y (wet)

Total dry weight

Diameter X (dry)

Diameter Y (dry)

% Shrink (av)

Wght loss

Platter 01 1000g 500g 50% 1493.9g 370mm 355mm 1278g 352mm 332mm 5.7% 14.5%

Platter 02 2000g 800g 40% 2831g  410mm  430mm 2725g   385mm  410mm  5.9% 3.8% 

Platter 03 1500g 600g 40% 2123g 430mm 420mm 1855g 411mm 407mm 4% 12.6%

Platter 04 2000g 1000g 50% 3044g 530mm  510mm  2726g  510mm   490mm 4%  10.4% 

Platter shrinkage and weight reductionduring the drying process

Platter 1, showing cracks on the edge due to drying to quickly

The Test Platters

Peat was used as the fuel for the fire, as found and identified at Bornais hearths. Peat temperature reached 950 degrees before levelling out to an average of 580 degrees when embers.

Test platters were placed on to and above the peat embers

Platter Fire temperature

Time added Position Usage Time removed Temperature upon

removal Survival

Temper test 10% 560 12:05 on embers none 12:07 N/A Broken

Temper test 20% 560 12:10 on embers none 12:40 280 Fired

Temper test 30% 520 12:20 above embers none 13:05 280 Fired

Temper test 40% 580 12:30 on embers none 13:00 285 Fired

Temper test 50% 540 12:25 on embers none 12:55 280 Fired

Temper test 60% 580 12:45 on embers none 13:05 290 Fired

Test platter 1 , popped and fractured

Upon removal the surface temperature was immediately taken from each platter. Averaging 285 degrees.

Cooking on the Large Platters

The temperature of the fire was taken before using the platters. Averaging 631 degrees.

Platter placed just above the fire supported by the stones on either side.

Flat bread placed on the platter when water splashed on evaporated off (10mins)

Flat bread cooking nicely and was ready to eat after 10 minutes.

Meat cooking on the platter. Ready after 15minutes

Platter Fire temperature Time added Position Usage Time removed Temperature upon removal Survival

Platter 2 547 15:15 above embers bread and meat 17:00 290 Fired

platter 3 697 13:10 above embers bread and meat 14:10 250 Fired

platter 4 650 15:30 high above embers bread 17:00 200 Fired

The fire was re-stoked and extended for two large platters. Platter 2 positioned just above the fire.

The bread upon platter 4 cooked more evenly than that of platter 2 just above embers, which tended burn if not turned regularly.

Platter 2 (with the bread) & platter 4, which was positioned 100mm above the fire to compare heat range and cooking effectiveness.

The bread upon platter 4 while on platter 2 fish and vegetables are cooking.

The surface temperature of platter 2 during the cooking process reached 290 degrees, while platter 4 reached a surface temperature of 200 degrees towards the centre. This was lower than all the other platters we used (as expected, it is higher up) but gave more flexibility and control in the baking/cooking of the bread.

Platter 2, showing the burnt residues on the surface after the cooking of meats. None of these residues are visible on the Bornais platters. (No fats were found on the Bornais platter sent for lipid analysis).

Platter 4, showing the surface after the cooking of the flat bread.

Post examination of the Platters

Platter 4, showing the surface after the cooking of the flat bread. The darkened area is very similar to the platter(SF 3846) from Bornais. Suggesting the use and cooking method we used.

Bornais Platter, Small find # 3846, context 1053 Mound 2.

Bornais Platter without organic impressions on the underneath, with no sooting visible.

Platters 3 & 4, placed above the embers, showed no sooting and a red tinge on the underneath after usage. This replicates closely the platter sherds found at Bornais.

Platter 3, the underneath after usage.

Platter 2, placed partly on the embers, showing the sooting on the underneath after usage. The area of reddening was where the platter was above the embers. Very similar to the Bornais platter sherds.

Bornais Platter with organic impressions on the underneath, with no sooting visible.

• Organic impressions on the underside successfully replicated with no sooting visible when placed above the fire. Organic and sand substrate found to be useful in platter manufacture.

• Optimal temper ratio found to be 2:1.• Platters would have been placed above (not on) the fire. • Durability of the platters found to be poor, one or two uses

only before the platter begins to crack and break up, likely to be single use items .

• Primarily used for the making of breads• Not possible to discern gender/age from finger impressions on

the platters

Conclusions

I would like to thank:

•A. Lane for his advice and input on the platters from the Hebrides

•Ian Dennis for all his help and advice.

•N. Sharples for allowing access to the Bornais assemblage

•Kirsty Harding for advice on the Bornais ceramics

•Peter Forward & Luc Potts for allowing us to use their back garden for the running of the experiment and making the breads.

•Roeland Paardekooper for his encouragement.

Acknowledgements