Dumb Barter: A Seminal Form of Negotiation

Negotiation requires communication, but not necessarily verbal exchanges. Parties may make adjustments in their offers and counteroffers incrementally by other means. Historical and anthropological analysis reveals numerous examples of such negotiations.

How do parties manage to strike a deal in situations characterized by such obstacles as complete distrust — and even hostility — major cultural differences, extreme difficulties in reaching the trading place, and several security risks? The practice presented here, the “dumb barter,” has been observed in various parts of the world (and especially West Africa) for millennia. A close examination of this unlikely — but real and effective — process yields useful insights into the fundamental nature of negotiation. Furthermore, it sheds light on how a trading system can be kept stable for a long time, even when the parties to the exchange are more foes than friends.

West Africa was not, as was previously believed, isolated from the main currents of international trade. Long before Asian sailors on the eastern coasts and Portuguese on the western side began trading with local people, a number of tribes traveled vast distances for business purposes.

For many centuries, sub-Saharan Africa, particularly the Western Sahel, was rich in gold and poor in salt. Salt from the desert was needed by the people of Sahel to flavor and preserve their food. It was also prized as a substance that healed wounds. Gold had obvious value, especially for people around the Mediterranean Sea. Merchants from what is now the southern part of Morocco, the Tafilelt and Sigilmasa regions, would bring down to Sahel and Sudan blocks of salt taken from Taghaza, in the heart of the Sahara, and trade it for gold dust. Because of this trade, cities grew and flourished as they became commercial centers (Péres 1937). Western Sudan, which the Arabs called Bilad al-Sudan, “the land of the black man,” produced large amounts of gold until about 1500 AD. The British even called the areas southwest of Timbuktu “Guinea,” from which the British gold coin takes its name. Later, what is nowadays known as “Ghana” was named “The Gold Coast.”

The Taghaza–Timbuktu route was an extremely difficult way to travel — there were few wells along the route, little food, and travelers were vulnerable to attacks by armed bands raiding caravans, pillaging and killing. It was a two-month journey for a caravan of heavily loaded camels. Nevertheless, for more than two thousand years, traders followed this route, until the end of the eighteenth century, when salt’s value declined and the mining of Sudanese gold more or less stopped, replaced by gold from other parts of the world. Up until that time, exchanges werecarried out by means of a very specific negotiation method known as “dumb barter.”

Dumb barter involves no direct contact between negotiators. Members of one group leave goods at a neutral location, usually on the edge of their territory, and then leave. Some time later, members of the other group leave something next to the goods as a proposal for an exchange. The first group then returns and either picks up the goods that were left by the others or leaves them there until such time as additions or subtractions make the deal acceptable.

The process provides no opportunity for either party to talk to the other, which is why this form of negotiation has been called a “dumb barter” or “silent trade.” In French, it was sometimes termed “commerce par dépôt” (“deposit trade”). This process was essentially characterized by the series of deposits and the expected convergence of assessments regarding their value.

Trade need not always involve verbal agreement, and dumb barter has been observed in many preindustrial groups. In fact, the absence of oral communication was not the reason for this type of interaction, but rather a mere consequence. Avoiding a face-to-face encounter was essential for the proper implementation of this form of negotiation. Thus, the term “deposit trade” better reflects the true reality of this bargaining process. Although there is no verbal communication, a number of signals are exchanged, which turned this indirect encounter into genuine interaction. The environment in which the dumb barter took place was characterized by global distrust, the likelihood that one or both parties could cheat, and the risk of violent conflict. But nonetheless, the need to trade opened up a window of opportunity.

Dumb barter has been explained as a way to do business with “savages,”“rude tribesmen,” and highly aggressive people. In the case of the Saharan trade routes, presumably no member of either party wanted to run the risk of open conflict, attack, or kidnapping, which is why it avoided direct encounter with the other party. (They could have been kidnapped in order to obtain information about the source of the goods they were offering, i.e., gold.) In a worst-case scenario, the Africans could have risked being enslaved or even slaughtered.

Shyness, fear, xenophobia, and language difficulties have also been cited as possible motives for initiating a dumb barter negotiation. It was still the most effective method for trading with hostile tribes. Thus, economic exchanges could develop, for it maintained peace in the trading place. It may well have been the first form of market neutrality (Grierson 1903). This type of bargaining also had a real advantage in that it avoided one party having to abide by the onerous constraints imposed by hospitality laws (Grierson 1903).

According to historical records, dumb barter dates back at least twenty-five hundred years. Accounts of this form of trade are found as far back as the writings of the Greek historian Herodotus.

El Yaqut, a thirteenth-century Arab writer, reported this practice in the following way: “Having reached the shores of the river of gold, the traders from Sigilmasa beat their big drums, whose sound goes very far, and lay down on the riverside, by small heaps marked up with the name of each owner, salt, animals, copper rings, and blue pearls; then they move half a day away. After the Negroes cross over the river, inspect the goods, lay down the equivalent in gold, and then vanish. The merchants from Sigilmasa come back and take the gold if they consider that the quantity is enough, and if not they leave the heaps as they are and leave. The Negroes then come back and so on until they reach an agreement. Then the whites leave, beating their drums the way they came” (Delafosse 1911: 46). Ibn El-Ouardi, writing at the same time, added that if the bargaining lasted too long, the Africans sometimes became furious, destroyed the goods, and took back their gold.

Other versions such as those provided by El Idrisi and El Masud (twelfth century) mention that after having loaded salt when passing through Taghaza, the traders went to Kumbi, where they meet their agents, local people familiar with this type of negotiation. They traveled south together for twenty more days before reaching the “gold river.” They wrote that the Africans were naked, lived in holes, and refused to emerge in the presence of foreign merchants (Bovill 1958).

Alvise Da Cadamosto (1455), a Venetian sailing for the King of Portugal, Henry the Navigator, heard from Arab and Sanhadja (a Berber tribe from Maghreb) merchants that this silent trade was still being conducted. He wrote that “salt caravans from Taghaza traveled south through Timbuktu to the city of Mali where the salt had to be transferred from the camels’ backs to the men’s heads, as the terrain was considered highly unsuitable for camels because of swamps, improper food and the risk of catching fevers or illnesses endemic to this part of Africa” (Bovill 1958: 195). The men formed a great army, walking until they reached a river (it is not known exactly which river) a long way away. Then, they piled the salt in rows, each marking his own, before retiring half a day’s journey, and the silent barter continued as described above. Al Mansur, wrote in 1600 of “deformed Negroes who will never be seen in the act of trading with the Moors or any stranger,” but who were nevertheless praised as being the most reliable traders in the world (Bovill 1958: 188).

At the end of the seventeenth century, Sidi U Ali — a marabout, a religious, and political leader of Tazeroualt, a kingdom in the south of Morocco — was still sending a salt caravan every other year to “Guinea.” The caravan would bring back a fine powdered gold, loading up four, five, and sometimes even six camels to return to Morocco (Jacques-Meunié 1982).

Mouëtte, a French traveler held prisoner in Morocco, recounted another variant of the dumb barter as conducted in Timbuktu between traders from Tafilalt and the Sudanese (Barbot 1682/2008, 1992). The exchange site was a gunshot away from the residence of the commander of the frontier. There, the Moors met an Arab, acting on behalf of the commander, who instructed traders on how to negotiate without meeting the Sudanese. Fiber mats were spread on the ground and salt was displayed. The Arab noted the names of the merchants and the amount of salt they brought. Then, the exchange took place. The Sudanese selected a suitable heap, deposited their gold dust next to it, and left. The process continued until an agreement was reached. The agreement was confirmed when the Arab go-between placed a handful of salt next to the gold and took one-twelfth of the salt and one ounce out of each pound of gold as payment to the commander. No words were exchanged. If any disorder erupted, the commander himself punished the culprits by beating them with a rod.

Trading goods, far away, with unknown people with whom there will be no direct contact and usually no middleman sounds as if it would pose an impossible challenge, but this form of negotiation has occurred in several places around the world and has lasted for more than two millennia.

These accounts illustrate some of the issues associated with the dumb barter process, in which intercultural negotiation occurs with neither a broker nor spoken communication and without even the possibility of a face-to-face encounter. First, parties must have some extremely strong incentives to embark on such an adventure. On the Arab side, the investment was made in time and camels, as well as the considerable amount of effort needed to journey to and return from the trading place. The substance of the trade, salt, was a cheap commodity for these traders. They simply picked it up along the way, but carrying it posed a problem.

For their African trading partners, gold was sometimes as easily obtained as salt was for the Arabs, especially after the river flooded when it sometimes appeared by itself as the waters receded. Other times, it was obtained via mining, but not without effort and risk. Gold was only of limited use in the place where it was mined, but it became a precious resource for exchange.

These circumstances alone — the ability of the Arabs to harvest salt and the Africans to mine gold — were insufficient for them to catalyze this bargaining arrangement. Several additional conditions needed to be met. One was that the adventure would need to have a sufficient probability of success to make it worth the effort. There was no way to eradicate all risk, but parties could reduce these risks by beginning with only small quantities of goods for trade in what we might call a step-by-step or contingent approach. Another concern was the parties’ personal safety — that they would be able to avoid attack, kidnapping, and enslavement. This risk was minimized by the parties’ physical separation throughout the process. Because language was of no help, another issue was the parties’ ability to communicate sufficiently to understand the process. Otherwise, if the parties were unable to reach agreement on the rules of the game, the process would never begin. How was such a risky and fragile bargaining system able to survive for nearly two millennia?

Dumb barter is usually described as an encounter between groups who have no other means of contact and who feel great mutual distrust (Sundström 1965). In nonmarket economies, this highly impractical procedure has applied only to reciprocally hostile societies. As soon as the hostilities — and attendant risks — were reduced, groups have tended to shift to face-to-face negotiations. In such situations, rules of conduct have often applied only when dealing with people from one’s own tribe. When meeting aliens, hostile action was often allowed, such as robbery, kidnap, or murder. Thus, the prerequisite for trading was not silence, but rather avoidance of face-to-face confrontation. (The silence was merely a by-product of the process because the parties spoke different languages and thus had no means of communication anyway.)

Nothing in the historical record explains how dumb barter began, although plausible hypotheses have emerged to explain how the process could have started. Initially, the dumb barter process could have been brokered by a trusted third party who acted as a guarantor. It could also have begun by accident. For instance, one of the camels in a caravan may have died en route and because the other camels were already overloaded, the salt was abandoned by the side of the road. On their return, the Arabs noticed that the salt was still there but that some gold dust had been carefully arranged next to it. They interpreted this as an offer, took the gold, and remembered the place. Next time, they again left salt, then waited to see what would happen. Later, they found some gold next to the salt. Thus, they understood that they could leave with the gold and come again maybe next year. In this way, a pattern could have been established.

Trust is not a condition for beginning this kind of negotiation process; on the contrary, the extreme level of distrust prevailing between the two parties is what necessitates dumb barter. Its regularity does not reflect the moral qualities of the participants but rather their understanding of their own self-interest. An implicit code of behavior must be strictly followed; otherwise, consequences could be disastrous. The parties anticipated each other’s behaviors based on experience and developed their negotiation practices accordingly. For instance, they knew that the first offer should not be accepted, but should not also be contested forever. The process thus became ritualized. Finally, endless haggling would have entailed a loss of credibility, and consumed too much time, increasing the risk for the Arabs of catching a local disease and traveling back to Morocco at an improper time when travel conditions would be the hardest.

Once the Arabs laid down the salt, they had little opportunity to cheat the other party and make sizable gains. If the Arabs took the gold and left, the locals would lose nothing because they would still get the salt — the Arab traders would have been unlikely to take the gold and keep the salt. Selling it somewhere else in the vicinity would have been difficult because information travels more quickly than a caravan carrying salt and gold, and news of their perfidy would most likely have spread to other potential parties.

If the African tribesmen ran away with only the salt, their benefit would be limited. The Arabs would interpret such a move as a signal that there would be no deal and would conclude that there was no point in organizing future expeditions. If the Africans ran away with both the salt and the gold, they would have been easily caught and slowed down by their heavy load. Either action could have destroyed the entire trading system and cut off the tribesmen’s supplies of a highly valued commodity. Finally, although European stereotypes of the era depicted the Arabs and Moors as ferocious, merciless slave traders — and their African counterparts as shy, weak, fearful, and unattractive — the risk of destroying the trade would have been high enough to prevent bad behavior, whether the stereotypes were true or not.¹

A possible mutual demonization process, which could prevent parties from negotiating (Mnookin 2010; Faure 2011), would not have proven an insurmountable obstacle for trading between the Arabs and Africans. In the dumb barter context, the point is to trade despite obstacles, whereas in current demonization situations, especially political, the representation (or the social/political construction) of the other as a devil is commonly used as an excuse not to negotiate (Faure 2008).

Dumb barter is said to have required scrupulous honesty from both sides (Sundström 1965), but establishing and maintaining a dumb barter procedure has little to do with moral values. Instead, it suggests that what keeps the system functioning is not the honesty of a group of people but rather their capacity to clearly understand both their short- and long-term interests — each side must believe that its counterpart will behave according to rational self-interest. Honest behavior is thus the rational consequence of fulfilling one’s long-term interests.

Either party would have been unlikely to find an alternative salt merchant or gold trader in the same part of the world. Thus, behaving honestly functions in this process as smart conformity to the implicit rules of the game. The aim of both parties is to keep the process going for as long as it yields joint benefits.

Dumb barter entailed both considerable transaction costs in time and energy and potential physical risk. In at least one case, Arab traders were tempted to make an end run around the process by finding the source of the gold. One account reports that “the merchants once tried to discover the source of the gold by treacherously capturing one of the timid Negroes. He pined to death without saying a word and it was three years before the Negroes would resume the trade” (Bovill 1958: 83).

On another occasion, an emperor of Melli (nowadays Mali), curious to see these people who were described as “Negroes with huge lips hanging on their chest,” ordered that some of them be brought to him. A plan was devised and four were captured. Of these, only one was successfully detained. He never spoke, refused to eat or drink, and, according to the account, died in four days. None of the subsequent emperors ever tried again because, as an unexpected consequence of this ruse, they endured three years with no trade, carrying the salt all the way to the riverside, but ultimately receiving no gold in return (Cadamosto 1789: 58).

Although both the Arab salt merchants and the African gold traders shared the same need to trade, their relationship was characterized by power asymmetry. Apparently the Arabs, who carried weapons and had substantial experience in battle, could have captured or killed the local people if they met in person, and it seems that the gold traders had little capacity to respond, although few details are available. More generally, it seems that dumb barter reduced the risk of power asymmetry when a smaller, weaker, more isolated, and vulnerable group sought to trade with a much more powerful group of outsiders.

In fact, the foreigners were not even that strong — the whole process was built from one party’s perception that the other was stronger. And this reputation was easy for the reputedly more powerful party to maintain because the two sides never met face to face and thus had no way of discovering that their perceptions may have been inaccurate.

Asymmetric power typically increases the difficulty of finding an equilibrium solution. Viewed as an example of “dire necessity driving a timid people to trade with dreaded and more powerful neighbors” (Hakluyt Society 1937: 16), the system was balanced enough to be sustainable. Furthermore, some research indicates that the weaker party often achieves a better negotiation outcome in asymmetric bargaining situations than one would expect (Faure and Klaousen 2000; Zartman and Rubin 2000).

The stronger party’s challenge was to convince the other party that the process is reliable and safe, even if it was not necessarily able to convince the other party of its goodwill. Thus, the dumb barter process can equalize outcomes because it does not allow the stronger party to take advantage of its strength and to gain more concessions through such tactics as intimidation.

The dumb barter process’s most defining characteristic is the lack of direct interaction between parties. Although no words were exchanged and the two parties did not see each other, a number of signals were still exchanged between the Arab merchants and their African trading partners. The Arab merchants announced their arrivals with gongs, drums, and smoke (Herodotus 430 BC; Bovill 1929; De Moraes Farias 1974). They also used signals to indicate that they had moved to a sufficient distance. Thus, in this dumb barter, no one was actually “dumb”— but was always careful not to disrupt the system by keeping out of sight.

In most exchanges, communication involves verbal language, that is, the use of words, usually spoken or written, or both. Words obviously facilitate more efficient communication and ease the relationship-building process and the process of making and receiving offers and counteroffers.

With all the inefficiency and delay associated with a nonverbal bargaining process, an obvious question arises as to why a more direct communication entailing the use of some kind of basic verbal language did not eventually develop from such a trade. One would assume that after repeated negotiations with the same partners, both sides would have an incentive to simplify the process. The issue of power asymmetry is relevant here, I believe. Because they perceived themselves as vulnerable, the African tribesmen avoided exposing themselves any more than was absolutely necessary, which made speaking to the other party and learning any words from each other’s languages impossible. In addition, the duration of the process was probably a less important consideration at that point in history. The whole trip could take almost half a year. A week more would not have made any difference. It was not the length of the trip, per se, that made the most difference; it was the necessity to avoid certain seasonal risks, such as the rainy season in the Sahel or the seasonal flooding south of it.

Among several different bargaining activities, three in particular are relevant to this discussion: conceding, compensating, and constructing. Conceding is an interactive, integrative process aimed at sharing a resource, while compensating is the process of directly exchanging something for something else, and constructing (or creating value) consists of changing the entire structure of the negotiation itself. Dumb barter is an archetypal expression of the compensation paradigm, for the basic mechanism is an exchange of goods of different values. However, adding more gold to the pile (or taking some away if the other party has withdrawn some of its salt) and vice versa typically represents the implementation of a concession–convergence model whose purpose is to find an equilibrium. We have a two-stage process: first, agreeing on a formula of exchange, then dealing with the fine-tuning of the agreement.

Neither side in the salt-for-gold dumb barter seemed to have strong alternatives to negotiation. The Arab traders’ only means of improving their best alternative to a negotiated agreement (BATNA) would have been to reject the other party’s final offer and to find another group of Africans ready to operate the same way, but it seems unlikely that they would have been able to find such a partner. If a second negotiation did not enable them to get a better deal, they risked losing the opportunity to make a deal with the first group. For the Africans, improving their BATNA would have entailed finding other Arab salt traders whose arrival could be predicted — without the use of language, and with little knowledge of these foreigners and theirs plans, it seems improbable that the African tribesmen would have even attempted this.

Fruitful markets usually attract competitors but not in the salt-for-gold dumb barter process. One possible reason was that social control exercised on both sides would have prevented intragroup competition. External competition could have been prevented by the secrecy of the process — the bartering place and the means of initiating the barter were, historians assume, both kept secret. The challenges of their journey made it unlikely that the Arab salt traders were followed. Furthermore, mutual signs of recognition may be exchanged before opening the barter to make sure that everything goes as usual and is done according to the customary ritual.²

Another option for both sides would have been for the Arabsto send a strong message to their counterpart by leaving the gold untouched and destroying the salt or for the Africans to take back the gold and leave the salt. But we have no way of knowing how the other party would have reacted. One possibility is that exercising such an option would have discouraged the other party from returning the next time to participate in the process Thus, I believe the weakness of each party’s BATNA actually contributed to the stability and longevity of the dumb barter process in this region.

It is difficult to ascertain the potential bargaining range for this type of deal. Chronicles and reports provide few clues. A text from the twelfth century (Hamid 1928: 90) mentions that “the amount of gold that was to be obtained from the salt could vary from a certain quantity to its double according to the size of the offer and demand.” Such a margin for maneuver presumably required a considerable number of iterative moves in order to reach a satisfactory agreement.

Using this process would have become far more complex if, for example, the Africans needed something in particular, such as beads and textiles, and the only party that had textiles to trade did not need gold but was willing to trade the textiles for beads. At this point, if the three parties used the dumb barter system, several conditions would have to have been met to have made this multiparty deal possible. Suppose a scenario in which party A was the African gold traders, party B was a group of bead sellers, and party C had textiles to trade. First, party A must have been aware that party C needed beads. Second, parties B and C must also have not known each other. Third, A would have had to conduct a sequential negotiation process, starting with B and ending with C. At the end of the process, if agreement had been reached in both negotiations, A would have obtained beads and textiles. If B and C had known each other, each of them (or both of them together) could have controlled the whole process and weakened A’s bargaining position. In that case, we would have the following negotiation sequence: B–C then B–A, or C–A then C–B, or (B + C) − A.

Fairness is a cultural value that can play an essential role in resource allocation. In a dumb barter, fairness is achieved within the context of a set of subjective evaluations by each party. For the Arab traders, the salt was freely available and cheap, but its value was enhanced by the trouble they took to carry it across deserts and through swamps. For the African traders, gold was a relatively useless metal, but its exchange value made its extraction worthwhile to extract it from the ground despite the dangers of mining.

In both cases, what made the good offered valuable was a combination of the work required to obtain it and to bring it to the bartering place as well as the extent of the other party’s need for it. For the Africans, the salt was a vital necessity. For the Arabs, the question was how much money they could earn in selling the gold back home and whether this made the journey worthwhile.

There was clearly no market price to serve as an objective, independent point of reference. Thus, negotiating became a highly subjective process in which neither side had much information about the other party and its interests, making it more difficult to achieve a price that both parties would agree was fair. Past negotiations could have established points of reference, and each party might have based its assessment on what was agreed upon the previous year. Still, if conditions had changed — for instance, if the Arabs lost half their cargo in an attack or if the Africans had found fewer gold deposits — values would have needed to be recalibrated. This indeterminacy, however, did not automatically lead to failure.

A number of travelers’ accounts report that dumb barter in its basic form remained in use until the beginning of the nineteenth century in this part of the world. It lasted for well over two millennia, in spite of conditions that usually were not conducive to trading. In world trade, markets developed because they were accepted as neutral places. In the case of the dumb barter trade, the parties involved were able to make the place of exchange (here the riverside) a neutral place subjected to its own specific rules, away from the law of the jungle, law of the savannah, or the risks entailed by crossing deserts.

It was only when sources of gold outside Africa increased that the trade went into decline. (The goldmines were probably close to being exhausted by that stage.) An English traveler (Jackson 1814) quoted an invoice sent from Timbuktu to a merchant in Fez in 1790 that included gold dust and gold bars. As late as the beginning of the nineteenth century, another English explorer indicated that beyond Timbuktu lay a country where there was gold. He explained that the inhabitants never showed themselves, but during the night, traders laid down goods, and the next morning, these would be taken and replaced by gold (Lyon in Roncière 1925). But when Mungo Park explored the upper Niger in 1797 and René Caillé reached Timbuktu in 1830, neither of them mentioned anything linked to the dumb barter process.

Several other more modern bargaining scenarios have superficial similarities to this ancient dumb barter process. For example, kidnapping for ransom in twenty-first century in Latin America and the Middle East (especially Mexico and wartime Iraq) has developed as a quasi market in which both parties keep a strict separation, although the obvious differences outweigh the similarities. But in the kidnapping case, parties communicate using written or spoken language, and they usually speak the same language and share the same culture. The substance of the exchange — the hostages and the ransom — is never within physical reach of both parties. The kidnappers may hold both at once, and the power asymmetry is severe. Finally, the dumb barter process leaves both parties better off than before the interaction occurred. Even when the hostage is returned safely, kidnapping for ransom leaves the victims both poorer and traumatized — it only provides actual benefits for one party, the kidnappers (Faure and Zartman 2010).

Dumb barter is typical of premarket economies and has been reported in an impressive number of places around the world. It was carried out in other parts of Africa, such as Ethiopia and the Congo basin, where the Pygmies live. Indeed, the Pygmies may still use this practice. Dumb barter was reported by Indicopleustes (1897), a Greek merchant from Alexandria who made several voyages in the early sixth century. He wrote about a place in Azania on the border between Egypt and Ethiopia, where officials from Axum bartered for gold with beef, salt, and iron (Roncière 1925; Bovill 1958). The practice also seems to have become well established in trading between tribes from East Africa and merchants from India (Sundström 1965).

Dumb barter was used on islands in the Indian Ocean, in Sumatra, Buru, Ceram, Timor, and New Guinea. It was also practiced in Hainan in southern China and in Ceylon, as mentioned by Hsien Fa (414 AD), the fifth-century Chinese traveler. The Semang, often labeled as the Pygmies of Malaysia, used to trade with the Malays by means of dumb barter, trading jungle products for rice, salt, beads, and tobacco (Page 1938). They still practice dumb barter today, depositing their goods in a specific spot, and then departing and returning in a few days to take the articles left by the Malays (Fage 1961).

It was also used in the Americas in places such as Guatemala, the Andes, and the Amazon Basin. Much further north, foodstuffs were exchanged between Canadian aborigines and Inuits, and between Beothuk aborigines and western fishermen in Newfoundland (Pastore 1987). Tribes close to the Arctic Ocean, such as the Aleuts, were familiar with this procedure. The Lapps in northern Scandinavia also used dumb barter to trade furs (Tegengren 1943). Dumb barter is still practiced with small groups such as the Rauté, a nomadic tribe of hunter-gatherers of Nepal who use dumb barter to trade with surrounding Hindu farmers. They trade handmade wooden bowls and boxes to obtain grain, iron, cloth, and jewelry (Fortier 2001).

Dumb barter has been not just a strange, exotic, and exceptional way of trading goods but a rather universal way of insuring a better living for a number of small, isolated societies.

This look at one of the oldest forms of negotiation gives us an opportunity to reflect on the essential elements of bargaining itself. What are the properties that make bargaining unique and distinct from any other activity? Dumb barter is a situation in which metaphor and reality meet. Communication is reduced to a bare minimum. Self-interest outweighs distrust. Ritualization offers a structuring dimension and some level of predictability (Gulliver 1979). We may be observing here the very beginning of diplomatic immunity. The role of the context has been neutralized, setting the stage for a quasi-experimental situation. We have mixed-motive strategies for which each party faces a dilemma and an associated risk. The core characteristic is a process of adjustment, some kind of reciprocity or converging moves. Finally, the hidden variable we have operating here between the parties is what Arthur Denzau and Douglass North (1994) call “reasonably convergent mental models,” which simply make the whole process possible.