Printer Friendly

Mancala players at Palmyra.



Palmyra is known as a former outpost of the Roman Empire, whose riches were accumulated when it prospered as a caravan city in the Syrian desert on the route from the Mediterranean to the Gulf (Rostovtzeff 1932). Its role as a trading centre was dependent on the Romans, who needed to maintain security in the desert to control their empire (Gawlikowski 1994: 32). When, at the end of the third century AD, Palmyra's queen Zenobia famously defied the Roman emperor in a quest for an independent state, he destroyed the city. Arabs occupied part of the city from the seventh century onwards, but the city never returned to its former glory.

While political influence on Palmyra was Roman, cultural influence was mostly Parthian (Seyrig 1950: 1; Browning 1979: 36). Culture found expression in art, architecture, textiles, jewellery, perfume and pottery and all have contributed to our understanding of daily life in Palmyra. This study concerns another aspect of culture--board games deduced from marked and modified stone (Figure 1). These show a preference for a mancala-type game also found in neighbouring regions.


On the track of mancala

Board games are known to follow trade routes both in antiquity and in modern times (de Voogt 1999). Gaming pieces and dice are the most common finds in the archaeological record since boards made of wood are rarely preserved. But although wood was often used for playing boards, there are also numerous examples of boards carved on stone, so-called 'graffiti games' (Schadler 1994, 1998; Mulvin & Sidebotham 2004; Bell & Roueche 2007). Their date of origin is difficult to determine and only the archaeological context can shed light on their history.

Roman and Greek board games, as opposed to Egyptian and Babylonian ones, are described in ancient written sources (Austin 1934, 1935). These include the Greek game of five lines, morris and backgammon types. But several games, including mancala, have no reference in the literature and their rules and playing practices are unknown. Research by Schadler (1998), Mulvin and Sidebotham (2004) and Bell and Roueche (2007) has uncovered mancala-type games in Roman territory. In Asia Minor, Schadler found the Greek game of five lines that underwent a transformation into something with a strong similarity to so-called mancala games.

Mancala games use rows of cup-shaped holes and a proportionate number of identical playing counters--commonly seeds, shells or stones--that are distributed one-by-one in consecutive holes. The board consists of two or more rows of holes. Each of the two players usually owns one row and the object of the game is to capture the majority of the counters.

The method by which this is achieved depends on a great variety of playing rules that are found all round the world. Modern mancala games are found in the Caribbean, parts of South America, all over Africa and most parts of Asia including the Middle East (Murray 1952; de Voogt 1997). Examples of Syrian mancala are found both in museum collections (de Voogt 1997: 19) and sold in stores from Damascus to the Palmyran tourist shops. As in most areas of the world, these games are made of wood and are sold without playing rules.

Mancala at ancient Palmyra

In Palmyra, two types of games resembling mancala survive as stone carvings. One, referred to here as Roman mancala, has a configuration of two rows of five holes. The other has two rows of seven holes or sometimes even more and is called Syrian mancala. The location of these games on the site as it stands uncovered by archaeological excavation, suggests two separate introductions. This expands the ideas of Schadler concerning the presence of mancala in Roman Asia Minor and adds the complication of different variations of mancala that have appeared in separate time periods.

In September 2009, a brief survey was conducted to map the presence of graffiti (board) games on the marble and stone surfaces of the excavated city. The Palmyran ruins range from large temples and a colonnade, to living quarters and army camps, as well as funerary buildings. With the exception of the funerary sites and the later addition of an Ottoman fortress on top of a hill, the present survey's search for graffiti games included temples, Diocletian's army camp, a theatre and the entire colonnade with its adjoining official buildings and residential quarters. At least 14 Roman mancala games were identified on the theatre steps, on the floors of two houses and on the side of the Tetrapylon, as well as on a loose stone in the Temple of Nebo and one along the colonnade (Figure 2). Half of the total was located in the Temple of Baal, which was also the site of all 18 games of Syrian type (Figure 3).

Even so, it cannot be claimed that the search was exhaustive. Large parts of the city are not excavated and, since the presence of these games is dependent on the condition of the stone and the exposure of floors, there may be many more games hiding under the sand or even erased by erosion. This collection of examples is only a preliminary pass, but it produced a convincing set of games that illustrate playing practices in Palmyra.

No excavation took place to uncover these games. All reported game boards are in plain sight or partially hidden by nearby rocks. No pieces or playing rules have been reported but if the games were indeed mancala games then it is unlikely that the pieces, large numbers of pebbles, shells or seeds, would be recognised during an excavation.

Roman mancala

The term 'Roman mancala' is, at first glance, misleading since there is no proof that the game originated with the Romans or that this game is actually a mancala game. There are no descriptions of such a game in the Roman written record (Schadler 1998) and, although many examples of graffiti boards that look like mancala boards have been found in Roman Asia Minor, it is not possible to prove that the rules of the game were similar to those of today's mancala games.


Much of the definition of mancala games today is based on the appearance of the board: rows of cup-shaped holes. Even if the playing of Roman mancala is much different from today's games, it may still warrant the term 'mancala' on the basis of its appearance. Whether it should also be called 'Roman' remains a question that can only be answered after the complete distribution of this game becomes known.

There is evidence that Roman mancala was a game and not, for example, a divinatory or calculating device. In Palmyra and other sites, the boards are sometimes found in a game context. The Palmyran theatre shows the game in the same location as other Roman graffiti games on the theatre steps (Figure 4). Also, the contextual features of the boards points towards a game rather than anything else, such as outer ledges of monuments and theatre seats as well as areas that provide shade in the late afternoon. In contrast to other Roman sites, Palmyra shows almost exclusively mancala-type games consisting of two rows of holes and hardly any other configuration of holes or other type of Roman game.

The predominance of mancala as opposed to other games in Palmyra may be explained by the erosion of the stones that allow cup-shaped incisions to be recognised somewhat more easily than incised lines. Also, it is possible there is some researcher bias, to the effect that the researcher may have developed an eye for this rather than for other games. Still, only one of the mancala games found on the entire site was in the company of other games, suggesting that these mancala games may have outshone other games in popularity.

The Roman mancala boards in the Temple of Baal were found mostly near remains of the buildings that have been identified as a banqueting house as well as a sacred altar and basin (Browning 1979:112-14), all in front of the entrance to the temple proper. Of the games found in the Temple of Baal, as is true also for many of the Syrian mancala boards that were found, the number of holes per row is not always clear. For the purposes of this study, a game is considered Roman if it consists of two rows of not more than five holes.


Surprisingly, no games were discovered in the Diocletian camp despite the limited erosion of the floors. Many Roman games are found in army camps on the fringes of the Roman Empire (Schadler 1994) but in Palmyra it might be inferred that the players were Palmyran citizens rather than Roman soldiers. On the other hand, the theatre is also a common site of graffiti games, and it is no surprise that at least two were discovered in the theatre at Palmyra. Substantial erosion of the theatre step surfaces may have destroyed many others.


The example on the Tetrapylon is remarkable in that it has end-holes, one additional hole at each end of the two rows: in traditional mancala games these are used for storage of counters (Figure 5). There are other examples with at least one storage hole near the rows of holes, indicating that storage holes were an optional addition. An additional hole that is further removed from the two rows can also be found, making the purpose of these holes as storage holes (for captured counters, for instance) all the more likely (Figure 6).

The position of the boards on the carving surface indicates that the players probably faced each other with the board placed lengthwise between them. Particularly when a game is found on a ledge, the game cannot sit sideways between two opponents if the ledge is not wide enough. In today's mancala it is highly uncommon to place a board lengthwise but it is not impossible.

Syrian mancala

In modern-day Syria two sets of mancala rules are documented: one is called 'Mangaley' and the other is known as 'crazy Mangaley' (Culin 1896), the latter played mostly by women. Both games are exclusively played on boards of two rows of seven holes. Murray (1952: 165-8) suggests that an earlier version may have used two rows of six. Indeed, the rules of the game may allow different configurations of boards, a flexibility known for a number of other mancala games in Asia and Africa (Murray 1952; de Voogt 1997). In this study, a game is termed Syrian if it has two rows of seven or more. There are many examples of games with exactly two rows of seven (Figure 7). The examples with more than seven holes per row are likely to be a variation, as is common for many Asian games, where similar rules can be applied to different boards. Alternatively, holes may have been eroded or replaced to facilitate play in later years.


Since both Roman and Syrian configurations can still be found in the Temple of Baal, it is not unlikely that Roman mancala was transformed to Syrian. Some boards suggest as much since the rows are not complete or the added holes are somewhat different in shape (Figure 8).

The majority of mancala games in Palmyra are found inside or around the Temple of Baal, which was transformed into a fortress after Roman times (Figure 3). Since many peoples occupied this site, it is not certain which group should be connected to the Syrian version of the game, but they were most probably Muslims--Arabs or Ottomans. The classic Arabic records have little to say about mancala. The Arabs mention the game once and from this reference the word 'mancala' was coined in the anthropological literature (Murray 1952: 158). The rules or the configuration of the board are not recorded. Neither do the Arabs include mancala in their religious discussion of games in which the games of chess and nard (Arabic backgammon) are prominently featured (Wieber 1972). Evidence of game rules that resemble today's mancala games does not appear until the sixteenth century or later (Murray 1952; de Voogt 1997), which largely coincides with the arrival of the Ottomans in Palmyra.




At least 18 Syrian-type boards could be identified. Since this was an army base and fortress, it is not surprising to find graffiti games. The high number may be partly explained by the surfaces, which are relatively free of erosion. Yet, it is tempting to speculate that the soldiers were either very bored or quite addicted to this type of game.

The games were found mostly on a ledge or floor on the north and north-eastern parts of the outer wall or the inner temple. These locations provide shade during the day or at least in the afternoon. The smaller buildings, next to which most of the Roman games were found, had probably been destroyed by the time the Arabs occupied the Temple of Baal. No games were found inside the inner temple area. The stones located in today's north-eastern part of the fortress were sorted and arranged by archaeologists so that the original location of the game board is not always clear. As a consequence, one game was found in a vertical position (Figure 1). Others were found on fragments of stone, indicating that columns had collapsed before the game was chiselled into the rock. Although this may have happened in Roman times, it is more likely to have occurred after the sack of the city and the destructive earthquakes that hit the city afterwards. Only one Roman mancala game was found on a piece of rock that was originally part of a tall structure.


There is no early written record of (board) games in Palmyra. There is a cursory mention of mancala in the Arab literature but Roman texts do not mention any game resembling the mancala-type games found in Palmyra. The extensive presence of such games in both the old town and the Temple of Baal suggest at least two introductions of mancala-type games in Palmyra.

The first, Roman mancala, was played prior to the third century AD and has been noted in other cities of Roman Asia Minor. Although its origin remains unknown, its distribution is, so far, associated only with the Roman Empire. Palmyra is unusual in displaying this game without any apparent reference to the Greek game of five lines, as is attested in other Roman sites and with few examples of other Roman games. Also, the game appears to have been popular with the Palmyrans themselves rather than with Roman soldiers.

The Syrian game of two rows of seven holes represents a second introduction of mancala and its distribution is limited to the former Temple of Baal, adapted as a fort in the Arab period in or after the seventh century. The presence of a garrison and the common presence of games in military outposts partly explain this distribution. The distribution of this game outside Palmyra and outside of Syria is largely unknown.

Graffiti games in present-day excavation sites illustrate distribution patterns of playing practices and cultural influences whose elucidation would benefit by more research in neighbouring regions. A systematic description of graffiti board games in Roman Asia Minor and beyond is only in its infancy. But we may predict that both the origin of Roman mancala and the Arab or Ottoman relationship to Syrian mancala can be ascertained if the full extent of their distribution becomes known. Meanwhile, Palmyra features as a prominent mancala-playing haven in the Middle East.


Particular thanks go out to Bassam Show for his assistance with transport and translations, as well as Arnold de Voogt who was instrumental in the efficient search for games in Palmyra. I wish to thank the American Museum of Natural History in New York for their support and for making this research possible. I owe thanks to Jennifer Steffey for her detailed art work, as well as to David Hurst Thomas, Diederik Burgersdijk, Peter Whiteley, and Connie Dickmeyer for their suggestions, support and advice.

Received: 16 November 2009; Accepted: 31 December 2009; Revised: 8 March 2010


AUSTIN, R.G. 1934. Roman board games. I. Greece & Rome 4(10): 24-34.

--1935. Roman board games. II. Greece &Rome 4(11): 76-82.

BELL, R.C. & C.M. ROUECHE. 2007. Graeco-Roman pavement signs and game boards, in I.L. Finkel (ed.) Ancient board games in perspective: 106-109. London: British Museum Press.

BROWNING, I. 1979. Palmyra. Park Ridge (NJ): Noyes Press.

COLIN, S. 1896. Mancala, the national game of Africa. Washington: [s.n.].

DE VOOGT, A.J. 1997. Mancala board games. London: British Museum Press.

--1999. Distribution of mancala board games: a methodological inquiry. Journal of Board Games Studies 2: 10-14.

GAWLIKOWSKI, M. 1994. Palmyra as a trading center. Iraq 56: 22-33.

MULVIN, L. & S.E. SIDEBOTHAM. 2004. Roman game boards from Abu Sha'ar (Red Sea Coast, Egypt). Antiquity 78: 602-17.

MURRAY, H.J.R. 1952. A history of board games other than chess. Oxford: Clarendon Press.

ROSTOVTZEFF, M.J. 1932. Caravan cities. Oxford: Clarendon Press.

SCHADLER, U. 1994. Latrunculi--ein verlorenes strategisches Brettspiel der Romer. Homo ludens: der spielende Mensch 4: 47-67.

--1998. Mancala in Roman Asia Minor? Journal of Board Game Studies 1:10-25.

SEYRIG, H. 1950. Palmyra and the East. The Journal of Roman Studies 40: 1-7.

WIEBER, R. 1972. Das Schachspiel in der arabischen Literatur von den Anfangen bis zur zweiten Halfte des 16. Jahrhunderts. Walldorf-Hessen: Verlag fur Orientkunde.

Alex de Voogt, American Museum of Natural History, Central Park West at 79th St.,New York, NY 10024-5192, USA
COPYRIGHT 2010 Antiquity Publications, Ltd.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2010 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Research
Author:de Voogt, Alex
Article Type:Report
Geographic Code:4EUIT
Date:Dec 1, 2010
Previous Article:Revisiting Indian Rouletted Ware and the impact of Indian Ocean trade in Early Historic south Asia.
Next Article:An early medieval symbol carved on a tree trunk: pathfinder or territorial marker?

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters