[image ALT: Much of my site will be useless to you if you've got the images turned off!]
Bill Thayer

[image ALT: Cliccare qui per una pagina di aiuto in Italiano.]

[Link to a series of help pages]
[Link to the next level up]
[Link to my homepage]

This webpage reproduces an article in the
Journal of Near Eastern Studies
Vol. 11, No. 4 (Oct. 1952), pp291‑295

The text is in the public domain.

This page has been carefully proofread
and I believe it to be free of errors.
If you find a mistake though,
please let me know!

 p291  Did Roman Commercial Competition Ruin South Arabia?

George F. Hourani

It is not easy to know precisely to what extent the prosperity of ancient South Arabia was dependent on international commerce. The valleys and terraces of Yemen provide plenty of rich agricultural land, and the Periplus of the Erythraean Sea in the middle of first century A.D. says that this country (the hinterland of the port of Muza) "produces grain in moderate amount, and a great deal of wine," and that for this reason little grain or wine was imported.1 The Ḥaḍramawt has much cultivated land in its long valleys, and its ancient port of Cane is also said by the Periplus to import little grain or wine.2 In the plains east of Yaman and north of the Ḥaḍramawt Mountains, cultivation once extended far into what is now the gravel desert of Sabatayn, as is evident from the many ruins of ancient cities and irrigation works in that area. On the other hand, cultivated area has to be considered in relation to total population, and this is a wholly unknown factor. But at least it can be said with confidence that the wealth of those people above the subsistence level was derived from their well-attested commercial activity.3 The Periplus lists the various articles imported: mainly clothing and spices through Muza and clothing and metals through Cane. Even the upkeep of irrigation systems in the vicinity of trade routes probably depended on the country's commercial prosperity, for "the value of the trade made it worth while to attend to the general security; this in turn made it possible to maintain irrigation and agriculture."4

Thus when the South Arabian economy fell into serious ruin, at some time not earlier than the second century B.C. and not later than the sixth A.D., the change may fairly be attributed in the first place to a decline in the country's foreign trade. But we do not know just when or why this decline began. In the standard works on Arabian history it is usually said to be due to Greek and Roman mercantile competition, beginning in the first century B.C. and working powerful effects in the first century A.D.5 A representative statement of this view is given in Hitti's History of the Arabs, from which I quote a few sentences:

In the course of this first Ḥimyarite period [115 B.C.A.D. 300] the zenith of the South Arabian power was passed. So long as the Yamanites monopolized the maritime trade of the Red Sea they prospered; but now the control was slipping out of their hands. The Periplus of the Erythraean Sea (A.D. 50‑60), the first record of organized trading with the East in vessels built and commanded by subjects of a Western power, marks the turning-point of the tide of commerce. . . .

 p292  When Egypt under the Ptolemies became once more a world power the first attempt was made to contest the supremacy of the sea with the South Arabians. . . . Rome, which captured Egypt from the Ptolemies about the middle of the first century B.C., followed the Ptolemies in the policy of maritime competition against the Arabians and in the desire to free Egypt from commercial dependence upon al‑Yaman. . . .

The entry of Roman shipping into the Indian Ocean sounded the knell of South Arabian prosperity.6

My aim is to show that such a view is unwarranted: that the evidence does not point to any decline in South Arabian prosperity as a whole in the first century B.C. or the first A.D. and that the decline, when it did occur, cannot be considered an effect of the Greco-Roman commercial activity down the Red Sea in those two centuries.

I shall begin with a brief survey of Greco-Roman activity in trading with India via the Red Sea. Voyages from Ptolemaic Egypt to India began after Eudoxus of Cyzicus led the way in 120 B.C. and became more frequent after Hippalus had discovered direct ocean sailing in the Indian Ocean with the southwest monsoon. Hippalus is now generally dated between 120 and 90 B.C. The sailings from Egypt fell off in the middle of the first century B.C., owing to the disturbed conditions in the Mediterranean world, but revived again under Augustus. Strabo saysa that under the latest Ptolomies "not even twenty ships [annually] dared to traverse the Arabian gulf so as to come out beyond the straits, whereas now great fleets are sent to India and the Ethiopic headlands"; in another place he gives the figure of 120 ships as the size of the annual expedition which was setting out from Myus Hormus for India in 25‑24 B.C.7 For the first century A.D. the Periplus and Pliny show Greco-Roman shipping still very active in the Red Sea and beyond; and quantities of Roman coins of this period have been found in southern India.b In the second century Ptolemy Claudius shows increased acquaintance with the geography of the countries bordering the Indian Ocean.

Is there any evidence for a South Arabian economic decline coincident with this period of Greco-Roman activity? If there is, it has not been stated by our historians. But let us gather for ourselves some indications which could be suggested.

1. The Periplus indicates that before direct sailings from Egypt to India started, Aden used to be the leading port of international commerce between West and East; but the city was subdued by "Caesar," "not long before our time,"8 and it is now no more than "a village by the sea." Whoever this "Caesar" was who occupied Aden for a short or long period, the fact is clear that the port suffered a decline when it lost its international role as a market, owing to the changed conditions brought about by the direct sailings. This is substantial evidence as far as it goes. But it only refers to the fate of one city. The Periplus describes two other ports as flourishing, Muza on the coast of Yaman and Cane in Ḥaḍramawt; and these may well have absorbed the former trade of Aden or replaced it with commerce of a different kind in the new conditions.

2. The important excavations of the American expedition in the Wādi Bayḥān in progress since 1949, have revealed the end of the state of Qatabān in the first century B.C. The capital, Tamnaʿ, was destroyed by fire, ca. 50 B.C., and thereafter Qatabān was incorporated in the  p293 kingdom of Ḥaḍramawt.9 But this again is of no more than local significance. It could perhaps be explained entirely by the fortunes of war; but, if the event has an economic background, it may well have been due to the replacement of the overland caravan routes, passing from the Indian Ocean through the Western Ḥaḍramawt (Qatabān) and on to Najrān and the north, by the cheaper sea route which the Arabs were forced to use in face of Greco-Roman competition.

So far as I know these are the only two facts which may indicate economic decline in any part of South Arabia in late Hellenistic and early Roman times. Over against them there are other facts which reveal a prosperous state of affairs in the first century A.D. At the time of the Periplus Muza is a busy market town full of Arab shipowners and sailors; it has a varied trade with Egypt, Barygaza in India, Somaliland, and Rhapta in East Africa. Rhapta is tributary to Muza, and the Muzans sail there in their own ships. Through Muza the Sabaean king receives gifts of horses and asses, valuable clothing, and manufactured work in gold, silver, and bronze; these are sent by the Roman emperor, presumably to insure commercial privileges and protection for Roman merchants.10 Cane is the port for the export of frankincense and aloes, and she trades with Egypt, India, and the Persian Gulf.11 Pliny writes (A.D. 77) that the demand for frankincense has increased so much recently in the Roman Empire that the gum is now gathered twice a year and the lumps are not given a chance to grow to full size; that the amount consumed at Poppaea's funeral alone (A.D. 66) was more than the whole annual produce of Arabia; and that India, China, and Arabia are receiving Roman currency at the rate of one hundred million sesterces a year.12 Indirect evidence for the fortunes of South Arabian commerce is provided by the history of the caravan cities of Nabataea. The prosperity of Petra in particular was almost entirely dependent on caravan traffic between Egypt and Syria, on one side, and South Arabia, on the other.13 Now the ruins of Petra show definitely that the city was flourishing in the first century A.D. After the Roman occupation in the time of Trajan there was probably a slow decline; but then we find Petra's place taken by Jerash, Bosra, and other cities of the Decapolis which received the caravan trade.14 We may also make mention of the political condition of South Arabia for what it is worth as an indication of economic conditions. In the course of the first century B.C. practically the whole country came under the rule of the two kingdoms of Saba and Dhu‑Raydān in the west and Ḥaḍramawt in the south, and from our still limited information we may gather that this political stability endured in the first century A.D. There was even some overseas expansion, for the kingdom of Axum was founded in this century by colonists from South Arabia, while ports of East Africa were under Arab political control.

Now, if the evidence has been correctly interpreted so far, we can say briefly that Greco-Roman trade began late in the second century B.C. and became substantial  p294 by 24 B.C. at latest, while South Arabian economic decline did not start until the second century A.D. at the earliest. Is it possible to find a causal connection between the two events? Allowing fully for delayed effects in matters of economic history, I do not see how a trade situation which was in full play by 24 B.C. can be considered the cause of changes which did not begin to appear for over a century after that date.

I suspect that the theory being criticized was never based primarily on historical evidence, which we have seen to be generally unfavorable,15 but on mistaken conceptions about the normal course of events in commercial history. It may have been assumed, by Schoff or some predecessor, that an increase in Greco-Roman trade must have meant a decrease in South Arabian trade in the same regions. The idea of trade as a fixed quantity tends to persist in popular economic thinking. One answer to it, in our particular case, is that the trade between the Roman Empire and India, into which the merchants of Egypt entered so vigorously, expanded enormously in the period with which we are concerned, so that there may still have been room for South Arabians to take part in it with their own shipping. But apart from such direct participation — for which there is actually little positive evidence — the ports of Muza, Cane, and Ocelis must have drawn wealth from the expenditures of the northern seamen in them, and the Sabaean king is known to have received payment from Rome for trading facilities in these ports.16 In modern times ports like Aden, Malta, and Port Said have flourished entirely in foreign shipping. A port is doomed by being in the backwaters, not by being on the main lines of international trade. But, most important of all, the direct trade of the Mediterranean world with South Arabia was very greatly increased by the rise in the demand for spices, which followed on the general rise in the security and prosperity of the Roman Empire in the first century A.D. And, finally, South Arabian trade with East Africa and with the Persian Gulf was little touched by the new competitor.

Another conception which has been applied indiscriminately is that of economic imperialism, the use of armed forces by states in concerted effort to secure trade monopolies. Such a policy has of course been put into practice by many nations — a pertinent example is the use of naval force by the Portuguese against the Arabs to secure for themselves the Indian trade in the sixteenth century. But there is no evidence that the Romans resorted to such an active policy in the Red Sea and beyond. If they sent an expedition under Aelius Gallus, if they captured and garrisoned Aden or Socotra, it was not for the purpose of excluding other nations from trade but solely to obtain security for their own traders.

To find the conditions which would naturally bring ruin to South Arabian commerce, we have to look in the third and following centuries. The contraction of the entire Mediterranean economy in the third century led to a shrinkage in the demand for oriental products. The demand for myrrh and frankincense dried up as Christianity replaced paganism, for the early Christians were averse to pagan  p295 odors. What through-trade with the Indian Ocean remained to the merchants of Egypt was now transacted largely through the Abyssinian port of Adulis, and thus the Arabs were largely deprived even of the profits of their ports of call. What then took place in South Arabia is still unknown; definite evidence of its economic history in these obscure centuries now waits upon the further progress of excavation.

University of Michigan

The Author's Notes:

1 Periplus, ed. H. Frisk (Göteborg, 1927), Chap. 24.

2 Ibid., Chap. 28.

3 Particularly noteworthy is the assertion of Agatharchides (late second century B.C.), "On the Erythraean Sea," Chap. 102 in K. Müller, Geographi Graeci Minores (Paris, 1882): "For no nation seems to be wealthier than the Sabaeans and Gerrhaeans, who are the agents for everything that falls under the name of transport from Asia and Europe."

4 F. Stark, The Southern Gates of Arabia (London, 1936), p205.

5 W. H. Schoff, The Periplus of the Erythraean Sea (New York, 1912), pp5‑6; D. G. Hogarth, Arabia (London, 1922), p5; D. E. O'Leary, Arabia before Muhammad (London, 1927), p80; P. K. Hitti, History of the Arabs (4th ed.; London, 1949), pp58‑60; C. Brockelmann, History of the Islamic Peoples (New York, 1947), p3; H. St. J. Philby, The Background of Islam (Alexandria, 1947), p102.

6 Op. cit., pp59‑60.

7 Geography XVII.1.13; II.5.12.

8 Νῦν δὲ οὐ πρὸ πολλοῦ τῶν ἡμετέρων χρόνων Καίσαρ αὐτὴν κατεστρέψατο, chap. 26.

9 W. F. Albright, "The Chronology of Ancient South Arabia," in BASOR, No. 119 (October 1950).

10 Periplus, Chaps. 16, 21, 23, 24.

11 Ibid., Chaps. 27‑28.

12 Nat. Hist. XII, chap. 32, secs. 5862; chap. 41, sec. 83

13 Traffic with Mesopotamia naturally passed through more northerly cities such as Palmyra. To the south, Petra had sea routes via Aelana and Leuce Come, but these could never have handled much traffic owing to the inhospitable character of the sea and shore in that region.

14 See M. Rostovtseff, Caravan Cities (Oxford, 1932).

15 The destruction of Tamnaʿ was unknown when the theory was formed. On the other hand, it used to be thought that Aden was "destroyed," not "subdued" — a mistaken translation of κατεστρέψατο in Periplus, Chap. 26; and Hippalus and the subsequent rise of substantial Graeco-Roman trade with India used to be dated much later, in the middle decades of the first century A.D. Thus on balance the theory once seemed more plausible than it does today.

16 Periplus, Chaps. 23‑24.

Thayer's Notes:

a XVII p798.

b Many, many thousands. An exhaustive discussion and listing of Roman coins reported thruout India is provided by Robert Sewell, Journal of the Royal Asiatic Society, Oct. 1904; the listing for Southern India starts at p623. Many more have since been found, of course.

[image ALT: Valid HTML 4.01.]

Page updated: 18 Sep 12