Survey of Prehistoric Sites in Mahra, Eastern Yemen

ISSUE NO. 6 JULY. 2002 A Semi-Annual Archaeological Refereed Journal on the Arab World Editorial Board Editor-in-Chief Prof. Abdul Rahman T. AL-Ansary Editors Dr. Khaleel I. Al-Muaikel Dr. Abdullah M. Alsharekh Publisher Abdul Rahman Al-Sudairy Foundation Opinions presented in Adumatu do not necessarily reflect those of the Editorial Board or the Publisher ©All Rights Reserved For The Publisher. Contents EDITORIAL PAPERS • Survey of Prehistoric Sites in Mahra, Eastern Yemen. • The Middle and Late Bronze Age Climate of Ya'amoun in Northern Jordan Using Oxygen Isotope Analysis from Human Tooth Enamel. • Hima: Center of Prehistoric Art and Culture in Southern Arabia. • Archaeology and Development: Dams and other Dangers. BOOK REVIEWS • Al-Ula & Mada'in Salih: The Civilization of Two Cities. A. R. Al-Ansary and Hussein Abu Al-Hassan. • Islam: Art and Architecture. Markus Hattstein, Peter Delius (eds). ARABIC SECTION EDITORIAL PAPERS • A New Inscription among the Inscriptions of D Smwy. • Approaches and Styles of Arab Date Determination in Light of Early Arabic Inscriptions. • The Kods Coinage During the Umayyads and the Abbassids Dynasties. • Rare Fatimid Dinar in the Name of Caliph Al-Mustansir Billah, Minted in Sicily in 442 A.H. • An Islamic Inscription of a Lady from Al-Qatif, Dammam Museum—Eastern Province. SYMPOSIA AND CONFERENCES • Third International Congress on the Archaeology of the Ancient Near East. • "The Calligraphy Center of the Library of Alexandaria: Intent and Objectives" - A Symposium. BOOK REVIEWS • La Triade Astrale Divine un des aspects des relations du Maghreb antique avec l'Afrique de l'Est et le Sud de l'Arabie. Dr. Afraa Ali Al-Khatib. 4 Jeffrey I. Rose 7 Dr. Abdulla A. Al-Shorman 21 Dr. Majeed Khan 27 Dr. Harriet Crawford 35 Prof. David F. Graf. 39 Dr. Michael Decker & J. Eric Cooper 42 4 Dr. Amida M. Sholan 7 Dr. M. K. Al-Morekhi 15 Prof.Raafat M.El-Nabrawy 27 Dr. Khalaf F. Al-Tarawneh 49 Prof. Ahmed O. Zailaie 57 Prof. Zeidan A. Kafafi 67 Dr. Khalid M. Azab 70 Dr. Faraj Allah A. Yousef 75 Issue No. 6 July 2002 Survey of Prehistoric Sites in Mahra, Eastern Yemen Jeffrey i. Rose Abstract. A recent survey in the governorate of Mahra, Yemen produced surface collections of lithic artifacts rich in bifacial tools and blades. Technological analysis of the Mahra collection suggests a simple blade industry unrelated to the Upper Palaeolithic blade traditions found in the deserts of the northern Arabian Peninsula. A large percentage of the Mahra assemblage consists of specialized bifa- cial tools characterized as piano-keeled naviforms. Correlates to these tool types have been discovered throughout Oman. It is suggested the artifacts are associated with an Early/Middle Holocene wet- phase lasting from 10,000 to 5,000 BP. Introduction: In the summer of 2000, an archaeological re- connaissance was conducted in the governorate of Mahra, eastern Yemen. The survey was car- ried out by a small group made up of both Americans and Yemeni from the General Or- ganization for Ancient Monuments and Manu- scripts (GOAMM). The primary goal of this project was to find evidence of prehistoric hab- itation and to locate contexts that may contain stratified Pleistocene deposits. The brief ten- day survey was focused in the Wadis Thabut and Faydami on the sandy plain surrounding Al-Ghaydah, and the narrow strip of coastline extending eastward toward Hawf on the Omani border (fig. 1). A total of 14 sites, all appear- ing to date to the Holocene, were identified and are briefly described in the following report. Geography and Climate: The governorate of Mahra is situated in the easternmost province of Yemen. With only 110,000 inhabitants (half of whom are Bedouin and the other half fishers), occupying approxi- mately 57,000 square kilometers, Mahra is one of the least populated and least developed prov- inces in all of Yemen. As of the time our sur- vey was carried out, there were no paved roads connecting the Mahra Governorate with the rest of the country. Because of the logistical difficulties inhibiting travel to this distant gov- ernorate, Mahra has received very little prior archaeological attention. The only previous survey was conducted by Arnirkhanov (1994) in the mid-1980s. Presently, the Mahra Ar- chaeological Project (MAP), led by Dr. Juris Zarins, is actively engaged in archaeological research in this region. While investigating Iron Age trade routes for frankincense and myrrh in 1997, Zarins' team discovered poten- tial Palaeolithic artifacts ranging from Lower Palaeolithic handaxes to Upper Palaeolithic blade industries (Zarins et al., n.d.). The Mah- ra Prehistoric Survey was implemented in 2000 to further investigate these finds. The interior of Mahra consists of low mountains, typically less than 1000 meters in elevation. The basement rock is composed of Paleocene and Eocene limestones, marls, shales, and sandstones (Birse et al. 1997). Obtaining raw material would not have been a problem for human groups occupying this region—flint sources are ubiquitous and of excellent quality for knapping. Tabular blocks were observed eroding from exposed limestone terraces, and an abundance of rounded nodules are available within the wadi channels and scattered throughout the Issue No. 6 July 2002 ISSN: 1319-8947 (pp. 7-20) 7 Jeffrey I. Rose 8 Issue No. 6 July 2002 Survey of Prehistoric Sites in Mahra, Eastern Yemen peneplain. Seasonal drainage channels incise the rocky landscape. Along the coast a narrow plain serves as a buffer between the low mountains and the Arabian Sea. Jutting across the Omani border into eastern Mahra is the westernmost extent of the Dhofar mountain range. Upon the flanks of this low mountain chain there is a 20,000-hectare forest made up of Tamarindus, Commiphora, and Anogeissus. The forest is sustained by the Southwest Monsoon System that cycles seasonally through the Indian Ocean and is responsible for depositing ap- proximately 200mm of rainfall over eastern Mahra per annum (Schyfsma 1978). During certain periods of the Pleistocene (IOS 5e, 5a, 3) and early Holocene, the Southwest Monsoon System appears to have increased in intensity (Sanlaville 1992; Cleu- ziou et al.1992). These periodic fluctuations are attributed to 1) a rise in continental albedo in the northern hemisphere, 2) retreating ice sheets in Europe and Asia, and 3) an increase in water surface temperature throughout the western Indian Ocean (Clemens et al. 1991; Zonneveld et al. 1997). As a result of the strengthened monsoons, seasonal storms were driven further north into the hinterlands of the Arabian Peninsula and deposited significantly greater amounts of rainfall over the land (Sarnfheim 1972; Kutzbach 1981). Sub-tropical environments, now limited to a handful of forests located at high elevations, characterized much of southern Arabia during the Pleistocene and Holocene mesic phases. The vast seas of sand currently blanketing South Arabia's lowland basins and plains were once covered in grasslands, open savanna, and stabilized dunes. Increased rainfall produced playa lakes and seasonal drainages throughout these landscapes, providing enough freshwater to facilitate hominid habitation (McClure 1974; McClure 1978; Garrard and Harvey Adtimatu: 1981; Schulz and Whitney 1986; Schulz and Whitney 1987; Lezine et al. 1998). Survey Methods: The goal of this survey was to identify set- tings with a potential for in situ Pleistocene deposits. Much of the work was conducted by traveling via automobile to a variety of geomorphic regions, in order to assess the presence/absence of Pleistocene accumula- tions. The survey covered approximately 100 km along the coastal plain (fig. 1), 10% of which was intensively examined. Among the areas specifically targeted were relic ter- races near ancient watercourses, caves, and rockshelters. Upon identification of these loci, survey was carried out on foot to collect archaeological materials. If artifacts were found, a more extensive reconnaissance strategy was employed to de- termine the horizontal extent of the site and to collect a sample of material for analysis. Tools, cores, and debitage encompassing all size ranges were collected. Each site was numbered consecutively, in the order of its discovery, beginning with the prefix MPS (Mahra Prehistoric Survey). It must be em- phasized that collection techniques focused on obtaining diagnostic tools and debitage. The assemblages are not unbiased, represen- tative samples. For this reason a detailed technological analysis of the lithics was not carried out, as it would not adequately de- scribe the reduction sequence. Only a small sample of the collection could be exported from Yemen, further hindering classification of the artifacts. Sites and Artifacts: Fourteen sites were discovered, of which ten are located on relic alluvial terraces overlook- ing the low plain around Al-Ghaydah (fig. 1). This flattened lowland extends from the town Issue No. 6 July 2002 9 Jeffrey I. Rose 1 »■«« On I X 00 1 <N*> 1 1 X =tfc © o On — © NO 70.3 25.3 CN S ro CN <N On NO oo 11.6 00 00 c—/ o WO o o (S <N P-vd o NO r-NO* o 00 (N ro 00 On NO ON On r~ m *—* </"> 77.8 r-' 14.9 00 on 49.3 16.9 15.6 NO CN m 00 co <N 65.4 r» 17.3 On to On * 75.4 t-~ 00 ro ON ! ON NO ro ro e ro V) a. H H U fa o u o o o o o x> & f I • O 3 w o W ^ ^ w « si » -2 2^ o 6 <« c -r P _v« y „s, *~. '5 53 S 3 -2 P -2 % afl C i H o. c > P\3 S3 <u "3 J3 w Q < S H 10 Issue No. 6 July 2002 Survey of Prehistoric Sites in Mahra, Eastern Yemen of Al-Faydami in the north to Nishtun in the south, and gradually ascends westward approx- imately 100 kilometers from the coast toward the Hadramaut Plateau. The plain is bounded by low mountains to the north and south. An- cient fluvial systems draining from these mountains carved a network of relic terraces into the landscape. The sites of MPS 1, 2, 3, 8, 9, 10, 11, 12, 13, and 14 (fig. 1) yielded identical chipped stone industries. Table 1 presents the frequencies of various lithic types present at these terrace sites. MPS 2 is omitted due to the extremely low piece count. Because of the biased nature of the MPS assemblages, the lithic categories used for this analysis are loosely defined, and are intended only to give a general sense of the technology and typology. The "flake" category encompasses flakes, retouched flakes, and pri- mary flakes. The few bladelets discovered by the survey are lumped together with the blades. The type classified as "piano-keeled navi- forms" include all pieces belonging to that technological continuum, from crude rough- outs to completed tools. The MPS prehistoric terrace sites are lithic scatters of indeterminate expanse sprawling across the relic alluvial terraces. Within these terraces are interstratified outcrops of high quality, fine-grained flint. Tabular blocks of raw material are found eroding from the inselbergs, and rounded nodules blanket the surface of the deflated terraces. The ubiquitous flint debitage, cores, and tools found throughout the surface of the landscape suggest the sites may have func- tioned as primary reduction workshops. Nearly every one of the terraces surveyed contained some degree of reworked lithic material lying on the surface. In some cases refits could be made, indicating the artifacts were still in their primary context. The surveyed terraces range in elevation from 42 to 153 masl; their surfaces are charac- terized as deflated alluvium. In general, the ar- tifacts bear little evidence of water wear, while patination ranges from medium (orange) to heavy (burgundy). The artifacts encompass a wide range of sizes, from small chips (>2 cm) to large tools and cores (<15 cm). This pattern demonstrates minimal post-depositional aeo- lian and/or alluvial sorting. The richest concentrations of material were found on terraces along the Wadis Faydami and Thabut (fig. 1). Large, thick blades with unfaceted striking platforms make up a high percentage of debitage from these localities (fig. 2). Three types of cores are recognized: the first are single and multiple-platform amor- phous flake cores derived from rounded nod- ules, the second type consists of unidirectional blade cores manufactured from tabular flint (fig. 3), and the third type is characterized as parametal, recurrent exploitation of the flint nodule. The latter two cores types tend to be pyramidal in shape with unmodified and una- braded platforms. Platform angles on the blade cores are typically quite steep, averaging around 90-degrees. Based on the deep bulbar scars, it appears that blanks were produced us- ing hard hammer percussion. A number of tools were found that do not fit into any current typological category. They are naviform in plan view with a piano-keeled cross-section (fig. 4). These artifacts are bifa- cially and unifacially worked. Typically, there is little to no retouch on the ventral face, and on the dorsal face steep flakes are removed to produce the diagnostic keeled cross-section. A large percentage of artifacts of this type were found, representing various stages of reduc- tion. Other tool forms discovered on the relic terraces include: backed bifacial chopping tools (fig. 5), ad hoc retouched pieces, and var- ious scraping tools. Analysis of tool frequen- cies is purposely omitted from this report, as it Issue No. 6 July 2002 11 Jeffrey I. Rose 12 Issue No. 6 July 2002 Survey of Prehistoric Sites in Mahra, Eastern Yemen Issue No. 6 July 2002 13 Jeffrey I. Rose Li 5 cm Fig. 4*PIano-kceled. naviform tools 14 Issue No. 6 July 2002 Survey of Prehistoric Sites in Mahra, Eastern Yemen mm would only be representative of our sampling biases. Among the ten relic terrace sites, two yielded evidence of architectural remains. Two stone features were discovered at MPS 1 and MPS 10. In the case of MPS 1, the in- stallation consists of a circle of rocks pre- served up to 1/2 meter high. The outer wall of the structure is approximately .64 m thick; the total diameter of the feature ranges be- tween 4.2 and 4.9 m. The circle of rocks was constructed on the eastern edge of a terrace overlooking the coast. No artifacts were in- side the feature, while a handful of chipped stone pieces, including one piano-keeled nav- iform, were discovered immediately adjacent to the circle. The site itself is situated atop an inselberg in proximity to the present coastline. The inselberg is bounded on its northern side by an active wadi channel that drains into the Gulf of Qamar (fig. 1). The second stone circle was discovered at MPS 10. This feature is considerably more prominent than that recognized at MPS 1, the piled rocks reach nearly 11/4 meters in height (fig. 6). This installation is described as a solid cairn of stone piled in a circle. The cairn aver- ages approximately 4.3 meters in diameter. In the center there is a circular depression about .8 m in diameter and .6 m deep, giving the entire structure a volcano-like appearance. In addition, two parallel rows of stones extend about three meters outward from the feature. The installation was constructed on one of the highest hilltops in the immediate region, 153 masl. Similar to the feature at MPS 1, several artifacts were found immediately adjacent to the stone circle, including two piano-keeled naviforms. The purpose of these stone features is unknown, though similar structures were identified in the province of Hadramaut and are thought to have served as post- Aaumalu Palaeolithic burials. An alternate interpreta- tion posits these features were used in some capacity for hunting. The location of MPS 1, overlooking the confluence of a wadi chan- nel and the coastline, would have provided favorable conditions for hunting game at- tracted to the freshwater. Until one of the circles is excavated, however, these ideas are purely conjectural. The four other sites discovered during the Mahra Palaeolithic Survey were historic in age, and thus not examined in any great detail. In two cases (MPS 4 and MPS 5) the age of the site was determined based on the presence of artifacts associated with triliths—small piles of stones specially placed in a line. These fea- tures, dating to the Iron Age, are commonly found throughout southern Arabia and are thought to mark the ancient frankincense route to and from the Dhofar region in Oman (Zarins 1998). The final two historic sites, MPS 6 and 7, yielded only light concentrations of lithic arti- facts. While both sites were found on scree slopes below caves, it is likely that the artifacts originated from sediments within these caves. Flint material was found in conjunction with non-diagnostic ceramics. The chipped stone recovered from these sites was significantly less patinated than the material discovered on the alluvial terraces around the Al-Ghaydah plain. Discussion: Archaeologists working in Oman have re- ported assemblages similar in nature to the material collected from the terraces around Wadis Faydami and Thabut. Villiers-Petocz (1989) published a collection of lithic materi- al held by the Omani Department of Antiqui- ties. He describes bifacially worked pieces, discovered in multiple localities throughout Issue No. 6 July 2002 15 Jeffrey I. Rose Survey of Prehistoric Sites in Mahra, Eastern Yemen Jeffrey I. Rose Oman, which are similar in shape and tech- nology to the piano-keeled naviforms found in Mahra. These pieces were found in con- junction with a blade industry, similar to the Mahra findings. Smith (1977) reports bifa- cially worked piano-keeled pieces from southeast Oman. Whalen (personal commu- nication) describes artifacts fitting this description found in a survey along the Wadi Ghadun in southwestern Oman. It should be noted that this tool type has not yet been found west of Mahra. It may be that the dis- tribution of this industry is limited to Oman and eastern Yemen. Based on the consistent presence of piano- keeled naviform tools, it is clear the ten relic terrace sites discovered on this survey are homogenous. It is yet to be determined, how- ever, to which phase of time these collections can be attributed. An analysis of the blade technology excludes certain periods. The blades were produced by hard hammer percus- sion, forming 90-degree platform angles on unabraded and unmodified striking platforms. In addition, there were no core tablets associat- ed with these surface collections. This pattern suggests that the Mahra technology does not have affinities with the Upper Palaeolithic blade technology of northern Arabia (Coinman 1997), ruling out a later Pleistocene age for these artifacts. The aforementioned Omani assemblages are posited to fall somewhere between the late Pa- laeolithic and post-Palaeolithic periods. In considering current palaeoenvironmental data, it is unlikely early human populations would have inhabited marginal environments in south- ern Arabia between 30,000 and 17,000 B.P. This phase was one of the driest the Arabian Peninsula has ever been subjected to, more arid than the present day. The bracket for the Mah- ra lithic industry, thus, can be narrowed to sometime after 17,000 B.P. It is suggested these occurrences roughly coincide with a pro- nounced wet-phase during the early and middle Holocene, lasting from ca. 10,000 to 5,000 B.P. (Wilkinson 1997). Jeffrey I. Rose: Department of Anthropology - Southern Methodist University - Dallas, TX 75275 (214) 768-3924. Jeff-rose@email.com (SJjJLlI (_5j_a_>JI j <i-» ttl JuLlJI J isy'tW JL-JI n"n i U f\jj i^Uj . n_i JL-ojI «Lcl i i.-> £_jrJ( 3j_$_a 4jCj_»_>J, (j-wj-jifcUl Jjlf "ill . ^ ■ > au > I idiljjl {y& t—aJGj Lg-La cjuu£ 4 it hi J^>-JP <Lcj o -»JU o i^-i^ . 4_ajjjl!I 5^j^_>JI 4 j ui JLc-Ii ^jLit-uo ^>-J (^jJI <(jr^^' g ? Lajjuj -tjl <ir *L>-jI 3-)Li (_-_) iCiljJ^I {y> £>"J' I-*-* ' niVi^l JL3J .it-sjUJI 4jj__u_a 4jj9jjj JISLuiLj C-oa--ulj i4_>jjl «Lul uj 0_>jJI J_l_3 .4__tu 0 , • • • (jJJ ^ • , • • • j_« Co»lj ^-iJI i3Uj injJI j? i^-IjVl 4jbj_>JI 3_JojJI 4_j_i_jJb Jaxjp »v-jljjVl tiliJt ij\ .(^WO- 4 ;, ii i (^da-LJI ^jj 4-lLc i-SjLiUI) ja-sL>JI 18 Issue No. 6 July 2002 Survey of Prehistoric Sites in Mahra, Eastern Yemen ■ ———■^»<MWIW....................I»..............MW^IIIimiliaMaWWWMMMWMMMBIMBMMMWiWMIWWMMBMMM^aMM Notes: Members of MPS thank GOAMM and their very capable representatives who provided vital support in this reconnais- sance effort. Specifically, we would like to acknowledge the invaluable assistance of His Excellency Dr. Yusef Abdul- lah, Abdulbaset Noman, and Ahmed Shamsan for their aid. In addition, I would like to express my sincere gratitude to Dr. Christopher Edens for his assistance in both the pre- paratory and concluding phases of this project. I thank Dr. Anthony E. Marks for his assistance in analyzing the stone artifacts. I would also like to thank Dr. Farouk El-Baz and Mutlu Ozdogan of the Center for Remote Sensing at Boston University for producing the topographic maps used by the survey. Most importantly, I wish to express my gratitude to my family for their unwavering support. The Mahra Prehistoric Survey was made possible by a grant from the Ameri- can Institute for Yemeni Studies. Rererences: Amirkhanov, H. A. 1994. "Research on the Palaeo- lithic and Neolithic of Hadramaut and Mahra". Arabi- an Archaeology and Epigraphy 5: 217-28. Birse, A. C. R, W. F. Bott, J. Morrison, and M.A. Sam- uel 1997. "The Mesozoic and Early Tertiary tectonic evolution of the Socotra area, eastern Gulf of Aden, Yemen. Marine and Petroleum Geology 14 (6): 675. Clemens, S., W. Prell, D. Murray, G. Shimmield, and G. Weedon 1991. "Forcing Mechanisms of the Indian Ocean Monsoon". Nature 353 : 720-5. Cleuziou, S., M.-L . Inizan, and B. Marcolongo 1992. "Le Peuplement Pre-et Protohistorique du Systeme Flu- viatile Fossile du Jawf-Hadramawt au Yemen". Paleor- ient 18:5-29. Coinman, N. R. 1997. "Upper Paleolithic Technolo- gies: Core Reduction Strategies". In: H.G.K. Gebel, Z. Kafafi, and GO. Rollefson (ed.). The Prehistory of Jordan, II. Perspectives from 1997, Studies in Early Near Eastern Production, Subsistence, and Environ- ment 4. Berlin: Ex Oriente. Garrard, A.N., and C.P.D. Harvey 1981. "Environment and Settlement During the Upper Pleistocene and Holo- cene at Jubbah in the Great Nafud, northern Arabia". Atlal 5 : 137-48. Kutzbach, J. E. 1981. "Monsoon Climate of the Early Holocene: Climate Experience with the Earth's Orbital Parameters for 9,000 Years Ago". Science 214 : 59-61. Lezine, A.-M., J.-P. Saliege, C.R. Wertz, F. Wertz, and M.-L. Inizan 1998. "Holocene Lakes from Ramlat as- Sab'atayn (Yemen) Illustrate the Impact of Monsoon Activity in Southern Arabia". Quaternary Research 50 : 290-9. McClure, H. A. 1974. The Arabian Peninsula and Prehistoric Populations. Miami: Field Research Pro- jects. McClure, H. A. 1978. "Ar Rub' al Khali." In: Saad S. Al-Sayari and Josef G. Zotl (ed). Quaternary Period in Saudi Arabia, Vol. 1: Sedimentological, Hydrogeo- logical, Hydrochemical, Geomorphological, and Cli- matological Investigations in Central and Eastern Saudi Arabia, Wien: Springer-Verlag. Sanlaville, P. 1992. "Changements Climatiques dans la Peninsule Arabique Durant le Pleistocene Superieur et l'Holocene". Paleorient 18 : 5-25. Sarntheim, M. 1972. "Sediments and History of the Postglacial Transgression in the Persian Gulf and North- western Gulf of Oman". Marine Geology 12 : 245-66. Schulz, E. and J. Whitney 1986. "Upper Pleistocene and Holocene Lakes in the An Nafud, Saudi Arabia". Hydrobiologia 143 : 175-90. Schulz, E. and J. Whitney 1987. "Upper Pleistocene and Holocene Paleoenvironments in the An Nafud, Sau- di Arabia. In: G. Matheis and H. Schandelmeier (ed). Current Research in African Earth Sciences. Rotter- dam: A. A. Balkema. Schyfsma, E. 1978. "Climate." In: Saad S. Al-Sayari and Josef G. Zotl (ed), Quaternary Period in Saudi Arabia, Vol. 1: Sedimentological, Hydrogeological, Issue No. 6 July 2002 19 Jeffrey I. Rose Adumatti Hydrochemical, Geomorphological, and Climatologi- cal Investigations in Central and Eastern Saudi Ara- bia. Wien: Springer-Verlag. Smith, G. H. 1977. "New Prehistoric Sites in Oman". Journal of Oman Studies 3 : 71-81. Villiers-Petocz, L. E. 1989. "Some Notes on the Lithic Collections of the Oman Department of Antiquities". Journal of Oman Studies 10 : 51-9. Wilkinson, T. J. 1997. "Holocene Environments of the High Plateau, Yemen: Recent Geoarchaeological Inves- tigations". Geoarchaeology 12 (8): 833-64. Zarins, J. 1998. "View from the South: The Greater Ara- bian Peninsula." In: D. Henry (ed). Prehistoric Archae- ology of Jordan. Oxford: BAR International Series 705. Zarins, J., G. R. Hedges, and R. G. Blom n.d., "Proposal for the Second Reconnaissance of the Mahra Archaeological Project." Unpublished report. Zonneveld, K., G. Ganssen, S. Troelstra, G. Versteegh, and H. Visscher 1997. "Mechanisms Forcing Abrupt Fluctuations of the Indian Ocean Summer Monsoon During the Last Deglaciation". Quaternary Science Review 16: 187-198. 20 Issue No. 6 July 2002