Dholavira weights
by
Chandrakant Doshi
Key words: Harappa, Indus, Dholavira, Gujarat, ancient weights
Weights have been recovered from most major sites of the Harappan civilization, including Dholavira, a Harappan site in Kutch, Gujarat. It was excavated from 1989 by R. S. Bisht and in the excavation report Bisht lists 996 weight1specimens. Dholavira flourished in third millennium BCE and has revealed a few unique features not identified elsewhere in the Harappan Civilization, including the heaviest weight recovered from any site. Bisht has identified 3 different weight systems. One of these is the standard Harappan system with the unit weight of 13.625 grams. This note examines the weights that conform to the standard Harappan system.
Weight collection at Dholavira consists of 20 different shapes and 44 different materials. Many of the shapes and materials have not been seen among the recovered weights from earlier excavated sites at Harappa2, Mohenjo Daro2and Chanhu Daro2. These shapes and materials are listed in tables below.
One useful way of studying the relations between the weights and highlight the systematic nature of the system is by an appropriate chart. A weight system would show the characteristic staircase where the flats or rungs are produced by specimens with values in close proximity to each other. A log chart is an appropriate way of displaying this systematic nature as it can accommodate the wide range of values on a single chart.
A chart of all the 996 specimens is shown in Image 1. The weights are all sorted in ascending order and converted to log2, as this provideshigh resolution and highlights the binary steps of the weight system. The converted values are plotted on the chart. The specimens within 5% of the Harappan weight system are over plotted in red on the chart.
Table 2 lists all the shapes and their frequencies. The largest number of weights are listed as cuboids (387), followed by spheroids (124).
The chart below, Image 2, is a plot of all the 996 weights. For this chart, the weights are all arranged in alphabetical order and then, for each shape, the weights are sorted in ascending order. This way all the shapes can be identified with the plot of each shape starting at the bottom of the chart and reaching towards the top. The plot of the cuboid specimens, which distinctly shows the characteristic staircase of a weight system, is over plotted in red for easy identification.
Weights are made of 44 different materials. These are listed in Table 3.
A few of the weights with sufficient number of specimens are plotted in the chart below, Image 3. First, the weights are separated by material and, for each material, the weight list is sorted in ascending order and log of each weight to base 2 is plotted on the chart below.The weights of 6 different materials are displayed on the chart.
The vertical scale of the chart is adjusted to show the binary nature of the system at lower weight values. 3.77 on the vertical axis corresponds to 13.625 grams. Weights of chert, agate and sandstone show distinct staircase, indicating many specimens at the nominal values of the weight system. That is not the case with the other three materials.
Another way of characterising a weight system is by plotting its frequency distribution. This is shown in Image 4. As most of the weight specimens follow a binary scale, the range of the distribution is from 0.75 to 1.50 normalised values. Height of the columns in the chart represent the frequency. The columns in blue comprise all the specimens within a given range, irrespective of their condition of preservation. Red columns represent only perfect, meaning undamaged, specimens.
The peak frequency is at the nominal values, i.e. 1.00 on the chart. R. S. Bisht has identified 3 different weight systems, which all contribute to the distribution. Frequency distribution of only the chert specimens gives a better representation of the Harappan weight system. This is shown in Image 5.
A. S. Hemmy analysed the Harappan weight system and established the unit of the system at 13.6253grams. Weights below the unit formed a binary sequence of successive halves. Heavier weights followed the sequence of 1, 2, 4, 10 for each decade starting from the unit weight. At the top end the sequence was 100, 200, 400, 800 as no weights at ratio 1000 were recovered at that time. Only two damaged specimens support Hemmy’s top ratio of 800.
M. S. Vats carried out excavation at Harappa in the 1930s, although the excavation report was not published until 1940. In the report Vats mentions a heavy weight of 26535.6 grams4, though this is not included in the official weight5list. No reason is given for this omission.
Among the Dholavira weights is one perfect specimen of limestone, weighing 13720 grams, which is close to the ratio 1000. (Acc No. 54404). The heaviest weight is also a limestone specimen weighing 35700 grams, (Acc No. 54550). This is 31% higher than the nearest attainable nominal value of 27250 grams. The condition of preservation of this last piece is not recorded.
Two additional weight systems Bisht reports have not been analysed in this note. This will be undertaken at a later date. Any weight specimens conforming to these additional systems recovered from other sites will have to be examined at the same time.
The analysis of the weights that conform to the standard Harappan weight system confirms the remarkable accuracy of these weights as first observed by A. S. Hemmy6and Dholavira weights demonstrate that this standard was prevalent throughout the Harappans civilization.
Acknowledgements
I would like to thank Harappa.com for putting the link to the Dholavira Excavation Report on Tweeter (@Ancient_Indus). This note would not have been possible without access to the material found at Dholavira.
The link to the report: fb.me/8z9Nkaucl
References:
1/. Weights and Measures, page 334 onwards, Dholavira Excavation Report, available Online: http://asi.nic.in/pdf_d…/dholavira_excavation_report_new.pdf
The link was provided by Harappa.com@Ancient_Indus
2/. Weights from these sites were analysed in a previous paper, which can be viewed here: https://jignashi.blogspot.com/2010/01/
3/. Chanhu Daro Excavations, 1943 pp 236
4/. Excavations at Harappa, M. S. Vats, 1940, p. 57, item 8
5/. Excavations at Harappa, op cit, pages 363-365
6/. Further Excavations at Mohenjo Daro, E. J. H. Mackay,
1938, p. 603
1938, p. 603
© 2019 Chandrakant Doshi
* * *
