Petrography of Ceramics from Bhairana

Received : 24-08-2012Revised : 26-10-2012Accepted : 05-11-2012

Petrography of Ceramics from Bhirrana: A Preliminary Study

K. Krishnan, L.S. Rao1 (late), V. Vinod2, Smitha S. Kumar, Prabhin Sukumaran3 and Dilip Kumar Kushwaha4

Department of Archaeology and Ancient History, Faculty of Arts, the Maharaja Sayajirao University of Baroda, Vadodara 390002Email: [email protected]. Archaeological Survey of India, Old High Court Building, Nagpur 4400012. Department of Archaeology, University of Kerala, Kariavattom Campus, Thiruvananthapuram 6955813. Department of Geology, Faculty of Science, the Maharaja Sayajirao University of Baroda, Vadodara 3900024. Department of Ancient Indian History, Culture and Archaeology, Gurukul Kangri University, Haridwar,

Uttarakhand 249404

K. Krishnan et al., Man and Environment XXXVII(2): 18-27 [2012]. © Indian Society for Prehistoric and Quaternary Studies

AbstractRepresentative samples recovered from excavations at Bhirrana, Haryana were subjected to the hardness test, porosity study, and thin-section analysis. From the study it appears that at least three sources of raw materials were initially exploited and after the Hakra phase one source was completely discarded. The study further revealed that there were a few ceramic workshops/workshop complexes at Bhirrana. Textural analyses indicate that different paste recipes were used for making different types of vessels.

IntroductionArchaeological investigations at Bhirrana (29° 33’ N; 75° 33’ E), Fatehabad District, Haryana brought to light evidence of evolution of material culture from the beginning of the Hakra ware phase to a fully fl edged Harappan Settlement (Rao et al. 2004-2005, Rao et al. 2005-2006, Rao 2005-2006). The earliest occupation of the site is designated as Period IA - Hakra ware culture followed by Period IB - Early Harappan, Period IIA - Early Mature Harappan (transitional) and Period IIB - Mature Harappan. A variety of ceramics were recovered from this site. These were subjected to a series of scientifi c tests to understand the evolution of ceramic materials at the site as well as the provenance of the raw materials used and the different manufacturing techniques employed in the manufacture of the ceramics. This paper presents the results of the macroscopic and microscopic studies carried out and also attempts to interpret its cultural implications.

Materials and MethodsThe ceramics from Bhirrana were classifi ed by the excavators into different wares based on its

external features such as colour, slip pattern, slip-body relation, painting, texture and inferable fi ring conditions. The wares present at the site include Mud appliqué ware, Grey ware, Red ware, Red ware with black slip, Red ware (incised), painted Red ware, Painted Red ware (slipped), Red ware with buff slip, Deep Red ware, Red ware/buff painting (combed pattern), Red ware (combed pattern), Buff-slipped Red ware, Chocolate-slipped ware, Bichrome ware, deep incised ware and Tan slipped ware. One hundred and twenty representative samples from dwelling pits 14, 5, 7 and 8 were analysed.

The ceramic colours were identifi ed by comparing its colour with their equivalents in the Munsell Chart for Ceramic Colour Designations (Peabody Museum Papers Vol.: 38). The hardness of the sherds was determined by performing the scratch test using Mohs scale. The porosity of each sherd was measured using the water absorption technique (Shepard 1965). Thin-sections of all ceramics were studied under the polarizing microscope (Leitz Laborlux 12 Pol D.). The mineralogy of the non-plastic inclusions was identifi ed with the help of their optical properties (Phillips and Griffen 1981). These samples were then clubbed under different fabric groups. The term fabric here has been defi ned based purely on the composition. It includes the mineralogy and texture. The texture was identifi ed

19


using the parameters, such as, grain-shape, grain size and relative abundance of specifi c grain categories (Bullock et al. 1985). Representative thin-sections from each fabric group and sub-group were subjected to point counting using a Swift Automatic Point Counter connected with a stepping stage. A total number of 300 points were counted in each representative sample by setting the stepping stage to single jump mode.

ResultsMacroscopic StudiesThe dominant colour of ceramics was Red and Grey with its varieties. It was observed that the colour of majority of the Red wares were comparable with their equivalents in the Munsell Chart designated as 10 R 6/8, that is, 55.9% of the Red wares had the same colour equivalent. The second largest colour category was 10 R 6/6 (16.1%). The other colours represented in this category were 10 R 5/8 (5.4%) and 10 R 3/2 (5.4%); 10R 4/1 (4.3%), 10 R 5/3 (2.2%) and the remaining colours showed 1.1% presence. In case of Grey ware, it was observed that majority of them had 10 R 5/1 (45.5%) followed by 10 R 4/1 (27.3%). The other colours available in Grey ware category were 10 R 5/2 (9.1%), 10 R 2.5/1 (9.1%) and 10 R 3/1 (9.1%). Hardness is the resistance offered by the surface when scratched with a mineral. It was recorded for all varieties of Red ware and Grey ware. It was found that among the Red wares, the hardness index 6 was the most common hardness (90.9%), followed by 7 (5.5%), 5 (1.8%), and 4 and 8 (0.9%). In case of Grey wares the dominant hardness was 6 (66.6%) and 8 (33.3%). It was interesting to note that the common hardness index was 6 in both the wares. The porosity study revealed that the fl uctuation in the value of apparent porosity was too high in most of the ceramic types except in the case of Red wares and Chocolate slipped wares (Fig. 1).

Microscopic StudiesBased on the microstructural characters, the ceramics from Bhirrana were grouped into ten main fabric groups, namely, A, B, C, D, E, F, G, H, I and J (Fig. 2 and 3). Of these main groups C, D and H have subgroups, namely C1, D1 and H1.

Fabric Group AFabric Group A (Fig. 4.1, please see inside front cover) consisted of Chocolate slipped ware, varieties

of Red wares, Mud appliqué wares, Tan slipped ware and deep incised wares. It was essentially a quartz-feldspar fabric (Fig. 2). The frequency of grains within this group varied from 50-60% except in two cases where the frequency was 30-40%. The grains were moderately sorted and did not show any preferred orientation. The grain size distribution character ranged from unimodal to bimodal. Voids were less. The dominant non-plastic inclusions were quartz and feldspar. The feldspars included both fresh and altered plagioclase with isolated occurrence of microcline. Also present in this fabric group were micas (both muscovite and biotite), calcites sometimes of cryptocrystalline nature occupying pores, olivine, augite, and on a few occasions, grog was also present.

Fabric Group BFabric Group B (Fig. 4.2) consisted of varieties of Red wares, Grey wares, Chocolate slipped wares and Mud appliqué wares. It was a quartz feldspar fabric (Fig. 2). The frequency of grains varied from 30-50% except in two cases, that is, in one case the frequency was as low as 20% and in the other as high as 60-70%. The grains were moderately to well-sorted and some of them showed a parallel orientation. The grain size distribution character ranged from unimodal to bimodal. Voids were present, but mostly due to grain fallouts. The dominant non-plastic inclusions were quartz and feldspar. The feldspars included both fresh and altered plagioclase. In addition to these, mica (both muscovite and biotite), augite and on a few occasions grog was also present.

Fabric Group CFabric Group C (Fig. 4.3) includes Deep incised ware, Red ware and Painted Red ware. It was a quartz-feldspar fabric (Fig. 2). The frequency of the grains was 20% except in one case where it varied from 10-20%. The grains were moderately sorted, except one case of poor sorting and a great majority of them showed a parallel orientation. The grain size distribution character ranged from semi-bimodal to bimodal. Voids were rare. The dominant non-plastic inclusions were quartz and feldspar. The feldspars included both fresh and altered plagioclase. In addition to these, mica-schist, grog, mica (both muscovite and biotite), augite and cryptocrystalline calcite were also present.

Man and Environment XXXVII(2) – 2012

20

Fig. 1: Apparent porosity of ceramic fabrics of different archaeological categories

Fig. 2: Mineralogy of ceramic fabric groups

Fig. 3: Texture of ceramic fabric groups

Fabric Group C1Fabric Group C (Fig. 4.4) had one sub-group C1 comprising of Mud appliqué ware and varieties of Red wares. It was a quartz-feldspar fabric (Fig.

2). The grains displayed parallel orientation. The frequency of grains was 30-40% except in one case where it was 50%. The grains were moderately to well sorted and showed parallel orientation. The grain

21


size distribution character ranged from unimodal to semi-bimodal. Voids were rare. The dominant non-plastic inclusions were quartz and feldspar. The feldspars included both fresh and altered plagioclase. In addition to these, mica-schist, grog, mica (both muscovite and biotite), augite, cryptocrystalline calcite were also present.

Fabric Group DFabric Group D (Fig. 4.5) consisted of a variety of Red wares. It was a quartz feldspar fabric (Fig. 2). The orientation of longitudinal particles was more or less parallel except in one case where no particular feature was observed. The frequency of grains was very low (as low as 2 to 3%) except in one case where it was 20%. The inclusions were moderately to well-sorted and were more or less showed parallel orientation. The grain size distribution character ranged from unimodal to semi-bimodal. Voids were rare. Non-plastic inclusions present were quartz, feldspar, mica, along with calcite and argillaceous inclusions. In one case grog was also present.

Fabric Group D1Fabric Group D1 (Fig. 4.6) included Red and Grey wares. It was a quartz feldspar fabric (Fig. 2). The longitudinal particles displayed parallel orientation. The frequency of grains was 40-50%. These were well-sorted. The grain size distribution character was unimodal. The dominant non-plastic inclusions were quartz and feldspar (fresh and altered). The other inclusions were mica and grog.

Fabric Group EFabric Group E (Fig. 4.7) consisted of varieties of Red ware, Bichrome ware, Chocolate slipped ware and Grey ware. It was a quartz feldspar fabric (Fig. 2). The longitudinal particles had a parallel orientation. The frequency of grains varied from 20-60%. The grains were moderately to well-sorted. The grain size distribution character ranged from unimodal to bimodal. Voids were few. The dominant non-plastic inclusions were quartz and feldspar. The feldspars included both fresh and altered plagioclase with isolated occurrence of microcline. In addition to these, mica (both muscovite and biotite), calcites sometimes of cryptocrystalline nature within the pores, augite, microcline feldspar and on a few occasions grog, were also present. Sandstone fragments and mica schists were also encountered in one section.

Fabric Group FFabric Group F (Fig. 4.8) had Bichrome wares, varieties of Red ware and Mud appliqué wares. It was a quartz-feldspar-calcite fabric (Fig. 2). The longitudinal grains in the fabric showed parallel orientation. The frequency of grains varied from 40-60%. The degree of sorting was poor to moderate. The grain size distribution character was bimodal. Voids were rare. The dominant non-plastic inclusions were quartz and feldspar; however, calcite was also fairly well represented. The feldspars included both fresh and altered plagioclase with isolated occurrence of microcline. In addition to these, mica (both muscovite and biotite) and grog were also present.

Fabric Group GFabric Group G (Fig. 4.9) consisted of varieties of Mud appliqué ware, Red ware, Bichrome ware, Chocolate slipped ware and Grey ware. It was a quartz fabric (Fig. 2). Longitudinal grains showed parallel orientation in a few sections. The frequency of grains varied from 10-40%. The degree of sorting ranged from poor to well-sorted. The grain size distribution character ranged from unimodal to bimodal. Voids were present, but mostly due to grain fallouts. The non-plastic inclusions present were quartz, feldspar (both fresh and altered), hornblende, augite, grog, mica schist, granite, mica (biotite and muscovite) and mica laths.

Fabric Group HFabric Group H (Fig. 4.10) included Painted Red ware, Chocolate-slipped ware and Bichrome ware. It was a quart-feldspar-mica fabric (Fig. 2). A few of the sections showed that longitudinal grains had a parallel orientation. The frequency of grains varied from 10-30%. The degree of sorting ranged from poor to well-sorted. The grain size distribution character ranged from unimodal to bimodal. Voids were rare. The non-plastic inclusions present were quartz, feldspar, grog, mica (biotite and muscovite) and calcite.

Fabric Group H1Fabric Group H1 (Fig. 4.11) had fi ne and coarse Bichrome ware, Mud appliqué ware, Chocolate-slipped ware, Tan slipped ware, a variety of Red ware and Buff (slip) ware. It was a quartz-feldspar fabric (Fig. 2). The grains had more or less parallel orientation. The frequency of grains varied from 10 to 40% except one where it was between 50-


22

60%. The degree of sorting was poor to well-sorted. The grain size distribution character ranged from unimodal to bimodal. Voids were present, but mostly due to grain fallouts. The non-plastic inclusions were quartz, plagioclase feldspar (fresh and altered), mica, mica schist, calcite, argillaceous inclusions, microcline feldspar, grog, bioclasts, re-crystallized cryptocrystalline calcite, granite and augite.

Fabric Group IFabric Group I (Fig. 4.12) consisted of Red wares, Chocolate slipped wares, Grey wares and Buff (slip) red ware. It was a quartz fabric (Fig. 2). Longitudinal grains showed parallel orientation. The frequency of grains varied from 10-30% except in two cases where the frequency was 5-10% and 60-70%. The degree of sorting was moderate to well-sorted. The grain size distribution character ranged from unimodal to bimodal. Voids were present, but mostly due to grain fallouts. The non-plastic inclusions were quartz, feldspar [Plagioclase (fresh and altered)], microcline, grog, mica (biotite and muscovite), mica-schist, calcite encrustation in pores, calcite, augite and cryptocrystalline silica.

Fabric Group JFabric Group J (Fig. 4.13) consisted of Bichrome ware, Mud Appliqué wares and Grey ware. It was a quartz-feldspar fabric (Fig. 2). It had longitudinal grains that displayed parallel orientation. The frequency of grains varied from 20-40% except in one case where the frequency was 50 to 60%. The degree of sorting was poor to well-sorted. The grain size distribution character ranged from unimodal to bimodal. Voids were rare. The non-plastic inclusions were quartz, feldspar, mica, grog, calcite and augite. Of these grog could not be quantifi ed.

DiscussionInitially, although we noted several shades of Red and Grey wares among the ceramics from Bhirrana, however, in reality these differences are marginal. A ceramic gets its colour due to a combination of several factors. These can be broadly stated as dependent on (a) the amount of colouring agent that is originally present within the raw material used for preparing the surface coat and the body, (b) the amount of natural colouring agent deliberately added by the potter, and (c) the chemical alterations that take place on the colouring agent during fi ring due

to fi ring temperature, fi ring conditions, duration of fi ring and position of vessel within the kiln. The range of difference in colour may also be due to the heterogeneity in the distribution of the colouring agents. The position of vessel within the kiln also invariably affects the temperature that is received by each pot. Ethnographic studies by various investigators (Krishnan and Rao 1994, Krishnan and Adikari 2008) have shown that there is a temperature fl uctuation within the kilns. Similar studies in closed kilns and open kilns have revealed temperature fl uctuation within the kiln (Rao 1991). For instance, in a closed kiln the temperature fl uctuation was approximately 50° C, where as in an open-fi re baking kiln (with 2.5 m diameter) the fl uctuation was approximately 180° C. This fl uctuation can vary depending upon several conditions. For instance, in the case of open-fi re baking kilns wind activity can cause even more temperature fl uctuation. Hence, variation in the colour of the vessels does not always relate to a deliberate attempt from the potters’ side to create different tones of the same colour. This does not discount the use of the Munsell Chart to describe the colour of a ceramic fabric as its use for identifying the ceramic colour designations can act as a standard tool to compare the presence and absence of different shades in different archaeological sites of the same cultural period.

While studying the surface features of the archaeological ceramics, it was observed that several of the sherds have fi re clouding on it. This feature appears due to the inadequate reach of oxygen to those areas. Such vessels/sherds with colour defects are usually found in vessels from kilns where the draught of air is not under proper control. In most of the cases, fi re clouding marks are found on vessels baked in open-fi re baking kilns. Therefore, it may not be inappropriate to state that majority of the sherds studied from Bhirrana belonged to vessels baked in open-fi re baking kilns.

In brief, the vessels of Bhirrana were made via multiple techniques. Paddle and dabber marks are sometimes visible on several sherds. These kinds of marks occur due to the enlargement of the vessels that are already thrown on the wheel. Thus, from the analysed samples, we have noted an example of the slabbing techniques datable to the Hakra phase, and also one of the coiling techniques datable to the Early Mature Harappan period. With regard to the shaping technique, one of the ceramics, a Painted Red

23


ware lid, appeared to have been made employing the slabbing technique. This falls into the archaeological category of Painted Red ware and was recovered from the dwelling pit No.5, belonging to the Hakra phase. An example of the coiling technique used for the construction of a vessel was also noted on a Red ware (slip) recovered from layer 6 of trench YC2/II, that dated back to the Early Mature Harappan period. However, there was not a single case of a vessel that employed the hand-made or moulded technique found within the samples analysed.

The archaeological categories, such as, Mud Appliqué ware, Bichrome ware, Tan-slipped ware and Chocolate-slipped ware all fall in the category of Red ware, both macroscopically as well as microscopically. All these categories have been identifi ed by looking at the surface features of the ceramics. Of these, Mud appliquéing could have been employed for functional aspects, whereas the Bichrome, Tan slipped and Chocolate-slipped ware may have been used for aesthetic purposes, or an attempt from the potter’s side to get rid of the monotony of colours.

The hardness of majority of the sherds was found to be 6 (90.9% of the analysed samples). This physical property is related to the composition and technology of the product. A common hardness may indicate that there was a consistency in the preparation of clay paste and uniformity in maintaining the fi ring conditions.

In general, the porosity of almost all wares was found to be fl uctuating. But, when viewed independently with regard to specifi c archaeological categories and cultural periods, a slightly interesting pattern developed. The porosity fl uctuation was noticeably higher in the ceramics of the Hakra period. For instance, 15.77 for Bichrome wares, 8.7 for Chocolate-slipped wares, 15.22 for Mud appliqué wares, 10.84 for Red wares and 10.48 for Grey wares. This fl uctuation underwent slight reduction

in the Early Harappan period, that is, 15.69 for Mud appliqué wares, 4.24 for Red wares and 2.03 for Grey wares. The fl uctuation was notably much lesser in the Early Mature Harappan period, that is, 1.96 for Bichrome wares, 3.02 for Chocolate-slipped wares, 5.23 for Mud appliqué wares and 6.66 for Red wares. The Chocolate-slipped ware of the Mature Harappan showed considerable fl uctuation, that is, 9.11, while in case of other ceramics it was 1.71 for Mud appliqué wares, 6.24 for Red wares and 2.84 for Grey wares (Table 1). Thus, it can be observed that the porosity fl uctuation reduces considerably as one move towards the Mature Harappan phase. It may also be noted that the apparent porosity of ceramics increases towards the Mature Harappan phase. Several factors can lead to porosity, one of which is due to the burning away of organic temper, such as, vegetable inclusions either present in the raw clay or deliberately added to the clay or the sieved fi brous animal excretions. The voids thus formed contribute to the amount of porosity, however, nothing could be gathered about the genesis of pores as the nature of voids was less clear in the thin-sections, except in case of grain fall-outs. Higher porosity adds to the longevity and quality of vessels that are either frequently heated or used as water containers and this can be consciously achieved by making deliberate efforts to improve the quality of clay paste while preparing it. A marginal difference in the porosity that is observable in case of Red wares of the Early Mature Harappan and the Mature Harappan, may be considered as an indicator to state that by then the paste preparation techniques followed in the workshop/s that produced Red wares became standardized. It may be further noted that the tendency to standardize the clay paste began in the Early Harappan phase itself. Further it appears that the quantity of organic ingredients during the Hakra phase was much higher than in the later phases.

Table 1: Fluctuations in apparent porosity in different periods

Ware Category Hakra Period Early Harappan Early Mature Harappan Mature HarappanBichrome ware 15.77 - 1.96 9.11Chocolate-slipped Ware 8.7 - 3.02 -Mud appliqué ware 15.22 15.69 5.23 1.71Red ware 10.84 4.24 6.66 6.24Grey ware 10.48 2.03 - 2.84


24

The ceramics from Bhirrana were grouped into ten main fabric groups. This grouping was done based on the mineralogical composition of the non-plastic inclusions. The non-mineral phases, such as, grog, though recorded, have been kept out of this main discussion, as its quantity is very low and also does not appear to have been deliberately added by the potters. Further, based on the dominant minerals and texture the fabric groups have been assigned names. These are (i) quartz-feldspar fabric tempered with silt and very fi ne sand, (ii) quartz-feldspar fabric tempered with very fi ne and fi ne sand, (iii) quartz-feldspar fabric tempered with different grades of sand (very fi ne, fi ne, medium and coarse sand), (iv) quartz-feldspar-calcite fabric tempered with different grades of sand (very fi ne, fi ne, medium and coarse sand), (v) quartz fabric tempered with different grades of sand (very fi ne, fi ne, medium and coarse sand), (vi) quartz-feldspar-mica fabric tempered with silt, feldspar fabric tempered with different grades of sand (very fi ne, fi ne, medium and coarse sand) and (vii) quartz fabric tempered with silt and sand. In short it can be seen that the various fabric groups from the analysed ceramics at Bhirrana are quartz–feldspar, quartz-feldspar-calcite, quartz-feldspar-mica, feldspar, and quartz fabric tempered with different grades of sand. Of the minerals mentioned above, quartz and feldspar are invariably present in all the fabric groups with their quantity varying, however, relative abundance of calcite and mica along the other minerals, such as, augite, olivine, hornblende and very few fragments of rocks may throw some light into the provenance of the clay. Texturally fabric groups B, G and I are the fi nest ones (Fig. 3). Fabric groups E, H1 and J are of medium texture (Fig. 3). Fabric groups A, C, C1, D, D1, F and H are the coarse ones (Fig. 3). Although a common textural similarity is visible (Fig. 3) in case of fabric groups C, C1, D and D1, they had to be grouped separately based on the variation in the relative abundance of minerals (Fig. 2).

The archaeological categories falling in group A are Chocolate slipped ware (3), various types of Red wares (4), Mud appliqué ware (5), Tan slipped ware (1) and Deep incised ware (2); B are varieties of Red ware (5), Grey ware (1), Chocolate slipped ware (2), Mud Appliqué ware (3); C are Deep Incised ware (1) and varieties of Red ware (3); C1 are varieties of Red ware (3) and Mud Appliqué ware (1); D are varieties of Red ware (4); D1 are varieties of Red

ware (2) and Grey ware (1); E are varieties of Red ware (4), Bichrome ware (2), Chocolate slipped ware (1) and Grey ware (1); F are Bichrome ware (1), Red ware (1) and Mud appliqué ware (1); G are Mud appliqué ware (4), Bichrome ware (4), varieties of Red ware (5), Chocolate slipped ware (2), Mud appliqué ware (3) and Grey ware (3); H are varieties of Red ware (2), Chocolate slipped ware (1) and Bichrome ware (1); H1 are varieties of Bichrome ware (4), Mud appliqué ware (2), Chocolate slipped ware (3), Tan slip ware (1), varieties of Red ware (5) and Buff (slip) ware (1); I are varieties of Red ware (4), Chocolate slipped ware (3), Grey ware (3) and Buff slip Red ware (1); and J are Bichrome ware (1), Mud appliqué ware (2) and Grey ware (1). The mineralogical composition of the non-plastic inclusions within the thin-sections of the analysed samples does show a commonality and also minor variations. Such features are likely to occur within alluvial regions. The site of Bhirrana is located in an area that occupies a unique morphological position in the Indo-Gangetic alluvial plain with the Himalayan ranges to the north, the Yamuna River to the east, the Thar Desert to the west, and the Aravalli Ranges to the south. The major part of the area is covered by Indo-Gangetic alluvium deposited during the Quaternary over the rocks of Vindhyan super group (Precambrian) and is represented by geomorphic features associated with a mature river system. The major geomorphological features associated with Haryana are the Older Alluvium, Newer Alluvium and Aeolian deposits. The Older Alluvium constitutes 90 % of the alluvial cover of Haryana and the Newer Alluvium covers the rest. The Aeolian deposits occur mainly over the Older Alluvial deposit and these are mainly deposited in the west and south of Haryana. Aeolian deposits in the Newer Alluvium occur as lensoidal bodies within the fl uvial deposits, whereas, aeolian sediments in the Older Alluvium occur as dunes and sand sheets (Fig. 5). Surface aeolian sediments in the Bhirrana area are generally fi ne-grained and moderately well sorted (Saini 2003; Ansari et al. 2012).

Therefore, precise provenance analysis of the ceramics based on the composition of the non-plastic inclusions becomes diffi cult in this area, but, by considering the susceptibility of weathering of different non-plastic inclusions, it is possible to conclude that a minimum of three raw material deposits were exploited by the settlers of Bhirrana

25


Fig. 5: Geolological and geomorphological map of the study area (After Saini 2003)


26

during the aforesaid archaeological phases. It also appears that one of the raw material sources was not used after the Hakra phase.

Another interesting feature noticed during the analyses was that the same ceramic wares fell into different fabric groups or different ceramic wares fell into the same fabric group. This indicates that the same type of raw material was used for making different ceramic wares. Thus, the macroscopic analysis of the wares produced different groups than the microscopic analysis. The methods adopted for macroscopic categorisation of ceramics are dependent on the pronounced physical property of the ceramics, such as, colour, externally visible texture, surface features, specifi c kinds of decoration etc. Although these properties have close affi liation with several cultural traits, they fall within the same fabric group as their raw material compositions don’t differ much. This also suggests that potters used multiple raw material sources to make a single ceramic ware. In short, it may not be wrong to state that different wares were made using the same kind of raw materials and also different types of raw materials were used to make the same type of ceramics. There is no direct relationship between ware type and a specifi c raw material source. This indicates that several potters or several workshops produced the pottery that has been studied from the site. In other words, a single workshop or a workshop complex was probably engaged in the production of a range of vessels.

The stratigraphic context of the samples that fell in each Fabric Group category was further examined. It was found that Fabric A included samples from dwelling pits 14, 5, 7, 8, belonging to the Early Harappan, Early Mature Harappan and Mature Harappan periods. The samples from dwelling pit 14 are from a pit sealed by layer 8. The contexts of samples from dwelling pits 5, 7 and 8 have not been supplied in the sample details. The Early Harappan samples come from layer 7, Early Mature Harappan from layer 6, and Mature Harappan from layers 3 (YB2/III) and 4 (YB2/II). These samples include a variety of ceramic types. Fabric Group B samples come from dwelling pits 14, 5, 7, 8, Early Harappan, Early Mature Harappan and Mature Harappan. Sample from dwelling pit 14 is from a pit sealed by layer 8. The Early Harappan samples come from layer 8 of the same trench, while the Early Mature Harappan samples come from layer 5 (YF2/I) and Mature Harappan samples come from layers 2 (YB2/

III) and 4 (YB2/II). Fabric Group C samples come from dwelling pits 7 and 8. Fabric sub-group C1 samples come from dwelling pits 5 and 8. Fabric Group D samples are from dwelling pit 8, Early Mature Harappan layer 5 (both from YF2/1) and Mature Harappan layer 3 (YB2/III). Fabric sub-group D1 samples are from dwelling pit 14, pit sealed by layer 8, Mature Harappan pit sealed by layer 2 (YB2/III) and layer 4 (YB2/III). Fabric Group E samples are from dwelling pit 14, 5, 7 and 8 and also from Early Mature Harappan. The sample from dwelling pit 14 is from a pit sealed by layer 8 and the one from Early Mature Harappan is from layer 5 of YF2/I. Fabric Group F materials are from dwelling pits 7 and 8. Fabric Group G materials come from dwelling pit 14, 5, 7, 8, Early Harappan, Early Mature Harappan and from Mature Harappan levels. The samples from dwelling pit 14 are all from a pit sealed by layer 8. The Early Harappan pottery samples are from layer 7 (YF2/IV). The Early Mature Harappan pottery comes from layers 5 (YF2/I) and 6 (YC2/II). The Mature Harappan pottery are from layers 2 (YB2/III) and 4 (YC2/II), and one comes from a pit sealed by layer 2 (YB2/III). The pottery that falls in Fabric Group H belongs to the dwelling pit 8, Early Mature Harappan and Mature Harappan phases. The Early Mature Harappan pottery comes from layer 6 (YC2/II). The Mature Harappan ceramic belongs to layer 4 (YB2/III). Fabric sub-group H1 has ceramics from dwelling pit 14, 5, 8, Early Harappan and Early Mature Harappan phases. The samples from dwelling pit 14 come from a pit sealed by layer 8. The Early Harappan sample comes from YF2/II and III. The Early Mature Harappan samples come from layers 5 (YF2/I) and 6 (YC2/II) (two samples). Fabric Group I comes from dwelling pit 14, 5, 7, 8, Early Harappan and Mature Harappan phase. The three samples from dwelling pit 14 are from a pit sealed by layer 8. The Early Harappan pottery comes from YF2/IV. The four samples from Mature Harappan are from layer 2, 3, 4 (YB2/III) and 4 (YC2/II). The four samples fall in Fabric Group J, which come from the dwelling pits 5, 7 and Early Harappan phase. The sample from Early Harappan phase comes from layer 7 (YF2/IV). Given the small sample size, it is diffi cult to present an intra-site distribution analysis of the various Fabric Groups. Nevertheless, , we still attempted to make some observations based on our sample size. The samples falling in Fabric Group C and sub-group C1 are exclusively from the dwelling pits 5, 7

27


and 8. The pottery types are varieties of Red wares, Deep Incised ware and Mud appliqué ware. The minerals present as non-plastic inclusions in these potteries are quartz, feldspars (fresh and altered), mica (biotite and muscovite), mica-schist, augite and cryptocrystalline calcite. A few pieces of grog were noticed in Fabric Group C pottery samples, however, it did not appear to be added deliberately. The minor difference between C and C1 is in its texture. Both are essentially quartz feldspar fabrics, with a minor difference in their clay paste preparation. It is possible to assume that each fabric group represents a particular technological tradition (related with raw material collection, processing etc) or in another way represents tradition within one or two workshops. The apparent porosity of C and C1 do not show much fl uctuation except in one case, that is, in C.

SummaryWhile summing up the above discussion certain propositions can be forwarded. The ceramics recovered from Bhirrana were made from the clay from the Indo-Gangetic alluvium as evident from the mineralogy of the non-plastic inclusions. Different kinds of ceramics were made in the multiple workshop complexes as ceramics of various archaeological category fall within the same Fabric Group or vice versa. The vessels at Bhirrana were shaped using multiple techniques. A great majority of the vessels were fi red in the open-fi re baking kilns as evident from the fi re clouding observed. The raw material source for ceramic production in different cultural phases remained the same, except in a single case; a different source, which appeared to have been used only during the Hakra phase. From the microstructural analysis, it appears that the clay paste preparation method seems to have improved from the Hakra phase to the Mature Harappan phase. Red ware and Chocolate slipped ware are the archaeological categories which refl ect a high degree of technological skills of the manufacturers.

AcknowledgementThe authors would like to acknowledge the members of the Excavation Branch I, Archaeological Survey of India for allowing them to study the ceramic samples.

ReferencesAnsari, Z.R., L.A.K. Rao and A.Yusuf 2012; GIS

based Morphometric Analysis of Yamuna Drainage Network in Parts of Fatehabad Area of Agra District, Uttar Pradesh, Journal of Geological Society of India 79(5): 505-514.

Bullock, P., N. Fedoroff, A. Jogerius, G. Stoops, T. Tursina and U. Babel 1985. Handbook for Soil Thin-section Description. Wover Hampton: Waine Research Publication.

Krishnan, K. and G. Adikari 2008. Ceramic Analysis of the Anuradhapura Hinterland, Archaeologia 4: 42-59.

Krishnan, K. and V. Rao 1994. A Study of Clay Paste Preparation through Textural Analysis. South Asian Studies 10:113-117.

Munsell Chart for Identifi cation of Ceramic Colour Designations. Peabody Museum Papers Vol: 38, Baltimore.

Phillips, R.W. and D.T. Griffen 1981. Optical Mineralogy: The Non-Opaque Minerals. San Francisco: W.H. Freeman & Co.

Rao, L.S. 2005-2006. The Harappan Spoked Wheels Rattled Down the Streets of Bhirrana, Dist. Fatehabad, Haryana, Puratattva 36: 59-67.

Rao, L.S., N. Sahu, U.A. Shastry, P. Sahu and S. Diwan 2005-2006. Bhirrana Excavation – 2005-06, Puratattva 36: 45-49.

Rao, L.S., N.B. Sahu, P. Sahu, S. Diwan and U.A. Shastry 2004-2005. New Light on the Excavations of Harappan Settlement at Bhirrana, Puratattva 35: 60-75.

Rao, V. 1991. Harappan Pottery Manufacture: A Preliminary Ethnographic Investigation. M.A. Dissertation (Unpublished), Vadodara: M.S. University of Baroda.

Saini, H.S. 2003. Sedimentological Characters of the Late Quaternary Aeolian Sediments of Haryana, Proceedings of the Indian National Science Academy 69(2): 201-205.

Shepard, A.O. 1965. Ceramics for Archaeologists. Washington DC: Carnegie Institute of Washington.

K. Krishnan et al., Man and Environment XXXVII(2): 18-27 [2012]. © Indian Society for Prehistoric and Quaternary Studies

Fig. 4: Microstructural characters of the ceramics from Bhirrana

Petrography of Ceramics from Bhairana

Documents