1Abstract. In order to re-construct the Harappan stone-bead techno-system, the model based on Cambay data (see Chapter 6) was first modified according to the specificities of the archaeological data.

2In particular, the ancient polishing procedures were different from the present-day ones. We had therefore to consider techniques practiced in Cambay prior to the introduction of electricity as well as those traditional operations still practiced in Yemen (see Chapter 3). Quantitative data have been modified according to each procedure.

3Secondly, a specific count of archaeological materials retrieved during excavation had to be made. These specifics, seen in the model as a form of constraint, will allow us to estimate the number of workers required to fabricate a plausible number of beads, and following from that, to envision the socioeconomic situation in which this took place (regular vs. occasional work; domestic vs. workshop vs. industrial; independent vs. attached specialists, etc.).

4For this purpose, we aggregated together beads recovered from all Harappan sites. We had to be satisfied with an overall estimate based on the number of “classical Harappan” beads, which because they are exceptional have always attracted the attention of excavators both in the Indus culture area and in Mesopotamia, where they are also found. In sum, it appears that the number of large classical Harappan beads does not exceed a few hundred, and are probably fewer than 500. Medium-and small-sized beads are always more numerous in proportion to large beads recovered at Harappan sites and are, grosso modo, 10 times and 100 times more numerous, respectively. The following counts result: 500 qsl, 5,000 qsm, 50,000 qsp.

5The question was then to assess the representativeness of beads recovered with respect to that in the parent population: 1 to 100 or 1 to 1,000? We hypothesize 1 to 100for thefollowingfour reasons:

1. If the number of excavated beads represents 1 to 1,000 in the parent population, that would imply that the original population was 500,000 qsl, 5 million qsm, and 50 million qsp. How then do we justify the loss of this many million beads? In effect it is a matter of material: on the one hand, beads are neither destructible nor recyclable; on the other, as pieces of jewelry, we assume beads always attract the attention of excavators and pot-hunters, alike.
2. If the number of excavated sites compares to the number of known sites surveyed on the order of 1,000, then the number of large sites excavated is quite small, on the order of10. But these large sites, excavated rather extensively, have never produced many beads (Mohenjodaro was excavated on the order of 106,000 m²!) (Jansen 1994:267).
3. Large classical Harappan carnelian beads have been imitated in baked clay (e.g., at Nausharo; Jarrige 1994:289), allowing us to suppose that their number was limited for several reasons (the problem of symbolic or physical access to large beads).
4. The intensive production of beads results in large amount of wastes (chalcedony and camelian flakes) which have never been found on any site (including Lothal).

6In adopting the hypothesis 1 to 100, we thus suppose that overall Harappan bead production produced 50,000 qsl (superior quality large), 500,000 qsm (superior quality medium), and 5 million qsp (superior quality small) beads.

7After processing this number of beads according to activity analysis, the number of workers and the techno-system were re-constructed. The following interpretative propositions obtain:

• The number of specialists implied in the production of stone beads was limited.
• Specialists worked in an integrated manner, executing all tasks of the technical sequence.
• In estimating the limited annual production of beads, specialists were apparently involved in other work as well, and not only in bead manufacture. These other tasks would have been, in part, craft specialties.
• Harappan stone-bead production appears to be so low that one cannot hypothesize a commercial enterprise in the proper sense of the term. One must therefore envision some kind of ad hoc productive system responding to a limited demand (religious, elites, etc.). Production for elites allows us to suppose that all the stone-bead specialists were “attached” sensu Brumfiel and Earle (1987).

8These interpretations contribute to the more general debate concerning the role of craftsmanship in the formation of the state. Contrary to what was intuitively supposed, the evidence here suggests that craftspeople involved in lapidary work were too few to represent a real “class of craftsmen”. The development of stone-bead production appears to have been a response to the demand of an elite. It cannot be taken as an argument in favor of theories according to which the origin of the state is to be found in a problem of the redistribution of goods.

9– The knapping of “classical Harappan” beads may have been carried out by only a very few persons (2 or 3 at any one time), those who were able to practice on a sufficient scale of production, and who could therefore learn the craft. They could not have been made by the craftspeople, present on numerous sites, whose major production was of small stone beads. What type of demand would have been great enough, while remaining nonetheless limited, to incite a few craftspeople (2 or 3) to develop the skills necessary for the manufacture of these exceptional beads? Archaeological data suggest that “classical Harappan” beads were intended for deities, and not for kings as in Mesopotamia.

10This interpretation contributes to the more general debate concerning the factors that played a role in the process of urbanization.