The Contribution of Vernacular Names for Pearl Millet to Its Early History in Africa and Asia

THE CONTRIBUTION OF VERNACULAR NAMES FOR PEARL MILLET TO ITS EARLY HISTORY IN AFRICA AND ASIA

Paper prepared for the RIHN Symposium ‘Small millets in Africa and Asia’ Tokyo September 19-20 th , 2010

Preprint Draft only; Please don’t quote without permission

Roger Blench Kay Williamson Educational Foundation 8, Guest Road Cambridge CB1 2AL United Kingdom Voice/ Fax. 0044-(0)1223-560687 Mobile worldwide (00-44)-(0)7967-696804 E-mail [email protected] http://www.rogerblench.info/RBOP.htm

This printout: February 10, 2012

TABLE OF CONTENTS

ACRONYMS ...... i 1. Introduction...... 1 2. Taxonomy and distribution...... 1 3. Pearl millet in Africa...... 2 3.1 Archaeobotany...... 2 3.2 Vernacular terms in West Africa ...... 3 3.3 Berber...... 5 3.4 Pearl millet names in the Bantu world...... 6 4. Pearl millet in Asia...... 6 4.1 India and China...... 6 4.2 Europe and the Near East...... 7 5. Synthesis and conclusions...... 7 References...... 8

TABLES

Table 1. Archaeological records of domesticated pearl millet 3 Table 2. #mar(d)a , a West African root for pearl millet 3 Table 3. Berber terms for pearl millet 5 Table 4. Pearl millet names in India 7

MAPS

Map 1. Language subgroups and ethnic groups in Table 2 4 Map 2. Pearl millet in India 6 Map 3. Diffusion of pearl millet cultivation in prehistory 8

ACRONYMS

* regular reconstruction # quasi-reconstruction based on rapid inspection of forms AD Anno Domini BC Before Christ BP Before present

i Linguistic evidence for pearl millet: Roger Blench Circulation draft

1. Introduction Pearl or bulrush millet is a characteristic staple of Sub-Saharan Africa, where it was almost certainly domesticated. However, it seems to have spread to India as early as 4000 bp, and thence across the Himalayan region to China, where it is first recorded around 0 AD. However, solid archaeobotanical evidence for much of this is lacking. Linguistic evidence can thus contribute to determining the pattern of the spread of pearl millet. The first review to explore African and Asian terminology for pearl millet is Portères (1958), who remains the only author to offer a global synthesis and indeed to discuss the problem of semantic shift between cereal species. This paper 1 reviews the linguistic, historical and archaeological evidence for this remarkable spread across the Old World.

Methodological considerations in the use of linguistic data are established in the companion paper on Asian small millets (cf. also Blench 2006, 2008). Conventions for tabulating linguistic data are also given in the same paper. Agronomic and linguistic data on millets are often compromised by aggregation. For example, the FAO database groups together all millets (without even listing species) into a single statistical set. 2. Taxonomy and distribution Linnaeus originally placed pearl millet cultigens into two separate species ( P. glaucum and P. americanum ) of the genus Panicum . Brunken et al. (1977) further reduced the number of species in section Penicillaria to two, on morphological and cytological grounds; P. purpureum was maintained as a separate species because of tetraploid chromosome number and perennial life cycle. All the diploid cultivated, weedy and wild taxa that frequently hybridise without genetic barriers, are classified under a single species, P. americanum. Clayton and Renvoize (1986) demonstrated that the taxonomically correct name for cultivated pearl millet is Pennisetum glaucum (L.) R. Br. They recognized the weedy forms (colloquially called shibras) as P. sieberanum and their wild progenitor as P. violaceum. P.violaceum differs from pearl millet in having involucres that are sessile, deciduous at maturity and always contain a single spikelet. Shibras differ from pearl millet in having deciduous involucres, and from P. violaceum in having deciduous and shortly pedicelled involucres. Pennisetum americanum and Pennisetum typhoides are commonly found as synonyms. The millets of Northern Ghana are so distinctive that they were originally classified as a separate species, Pennisetum gambiense , Stapf and Hubbard. This was later considered unnecessary and they are now classified as race globosum (Brunken et al. 1977; Appa Rao et al. 1982, 1985).

Pearl millet is usually characterised as a crop of the semi-arid zone of Africa and the majority of cultivation is found there today (Busson et al. 1965). Apart from its use as a staple cereal, it is used as animal fodder, for construction, beer and medicine (Andrews & Kumar 1992; Hamon 1993; Burkill 1994: 313–317). Millet is the most important cereal in Saharan oases, although it is gradually yielding to wheat, although it remains important as a fodder plant (Gast and Adrian 1965; Gast 1968). However, prior to the introduction of maize, millet played an important role in farming systems even in the humid coastal regions of West Africa. Brunken et al (1977) point to the restricted distribution of their race leonis , a high humidity type confined to coastal areas of Sierra Leone and Liberia. Perrot (1990) shows that unpublished journals of the seventeenth and eighteenth centuries clearly describe millet cultivation in the lagoon areas of Côte d’Ivoire. Archaeological evidence has shown that it was apparently cultivated far south in Cameroun, in a region which is now rainforest, during the period 2400-2200 BP (Neumann et al. 2011). Millet was first grown in Europe perhaps as early as the eighth century and these forms, under the name Pennisetum spicatum , are still grown in Spain and North Africa. Pearl millet is also a significant crop in semi-arid India, namely Rajasthan, Maharashtra, Gujarat, Uttar Pradesh and Haryana. It is particularly important in the drier, hotter parts of western Rajasthan, the only part of the country where it remains the staple food.

1 This is an expansion of a section of a paper on African millets given in October 2010, to the RIHN Roundtable, Tokyo. Other African millets are dealt with in a separate paper and indigenous Asian millets in yet a third paper. Thanks to Koen Bostoen for keeping me up to date on African publications and to Maarten Kossmann for sending me an advance copy of his material on Berber. 1 Linguistic evidence for pearl millet: Roger Blench Circulation draft 3. Pearl millet in Africa

3.1 Archaeobotany Archaeobotanical evidence for pearl millet in Africa is sparse, but gradually improving (Nixon et al. 2011). The first authors to claim an African origin for pearl millet were Koernicke & Werner (1885) and this as been generally accepted. Wild relatives are found on the southern edge of the Sahara (Chevalier 1932:888- 890; Brunken et al. 1977) and it is usually considered that this was its locale of domestication (De Wet 1995a). Portères (1962) argued that the diversity of races in the agronomic record implied multiple domestications, and there is no direct archaeobotanical evidence associated with many of the races identified in Brunken et al. (1977). However, the collections of cultivated races analysed in Oumar et al. (2008: 493) show very limited microsatellite diversity and their accessions do include a wide range of locales, including both high-humidity zones and Saharan oases.

Archaeobotany has placed pearl millet domestication on the margins of the Sahel more than 3000 years ago (Tostain 1998; D’Andrea et al. 2001). Recent proposals situate this event in the third millennium BC in the far western Sahel, perhaps in Mauritania and/or northeast Mali (Fuller et al. 2007). One of the emblematic sites of Central Nigeria, the Nok region, has produced a large amount of millet dated to 800-450 cal BC (Kahlheber et al. 2009). Perhaps even more surprising is the fact that millet was also cultivated in areas much further in regions that are now rainforest (Höhn et al. 2007). Millet is still grown as a ritual crop in the sandy, semi-arid areas of the eastern parts of coastal Ghana. The polysemy of ‘millet’ and ‘food’ in a coastal language like Ga in Ghana is a striking indication of its former importance in a region which has now almost entirely switched to maize.

The levels of Dhar Tichitt in Mauretania provide the most detailed sequence of wild and cultivated Pennisetum s. Re-examination of the ceramics from Dhar Tichitt and Oualata revealed that domesticated Pennisetum was grown at least as early as 3500 BP (Amblard 1996). A find of comparable age (3460±200 BP (TO-8172, 1980-1520 cal BC)) has been reported from the site of Birimi in NE Ghana (D’Andrea, Klee & Casey 2001; D’Andrea & Casey 2002). Slightly later is the dune site Ti-n-Akof in Burkina Faso where grains of domesticated Pennisetum have been found, directly dated to 2840±49 BP. In Gajiganna in Nigeria, domesticated Pennisetum occurs used as pottery temper in phase IIb dated to 2930±60 BP (Neumann 1999). Currently the earliest Pennisetum glaucum is that recorded the Malian Neolithic sites of Karkarichinkat from 2500-2000 BC (Manning et al. 2011). Nixon et al. (2011) provide a useful table of recent archaeobotanical records of pearl millet.

Beyond Africa, the evidence is extremely weak, plagued by misidentifications. Nonetheless, at least some of the material from India is indeed pearl millet (Fuller et al. 2004). Table 1 summarises the present evidence for archaeological millet in Africa and India;

2 Linguistic evidence for pearl millet: Roger Blench Circulation draft Table 1. Archaeological records of domesticated pearl millet Region Country Site Date(s) Reference Africa Mauretania Dhar Tichitt 1936-1683 BC cal, 1881-1527 BC Amblard & Pernes cal 1989; Amblard (1996) Mali Karkarichinkat 2500-2000 BC Manning et al. (2011) Burkina Faso Ti-n-Akof 2840 ± 49 BP (cal. 1035-916 BC, Neumann (1999:77) UtC-4906) Nigeria Kursakata Directly dated to 2430±70 BP Neumann et al. (1996), (UtC 5452) and 2290±70 BP Klee et al. (2000) Nigeria Nok 800-450 cal BC Kalhheber et al. (2009) Ghana Birimi 3460±200 BP (TO-8172, 1980- D’Andrea, C., Klee, M. 1520 cal BC) & J. Casey (2001) Cameroun Abang 400-200 cal BCE Hohn et al. 2007; Minko’o, Neumann et al. 2011 Bwambé- Sommet Asia India Rangpur ca. 1800-1200 BC Herman (1997) Hallur ? 2200-1800 BC Fuller (2003) Source: expanded from Blench (2006)

3.2 Vernacular terms in West Africa The most comprehensive compilation of pearl millet names for West Africa is in Burkill (1994), although this is not a linguistic analysis. Philippson & Bahuchet (1996) and Bostoen (2007a, b) have analysed vernacular terms in the Bantu languages of Eastern and Southern Africa, and these are considered in §3.4.

If millet did indeed originate in the western edge of West Africa this is not easily discerned from the vernacular names. Compared with sorghum, these are surprisingly diverse and do not seem to form a significant pattern over a wide area. Within Bantu there are large zones with relatively homogeneous terms, but otherwise there is a single widespread root, something #mar(d)a , attested in both Chadic and Plateau languages of Nigeria, as well as some Kwa languages significantly west of this region. This root is first identified and discussed in Longtau (2008). Evidence for this root is given in Table 2;

Table 2. #mar(d)a , a West African root for pearl millet

Ph Family Subgroup Language Attestation Gloss Source AA Chadic West Hausa maiwa millet Ab49 AA Chadic West Bole mòr ɗo pearl millet RS AA Chadic West Ngamo mòr ɗò millet RS AA Chadic West Geji mar ɗa millet RMB AA Chadic West Ngizim mar ɗ̃ û millet RS AA Chadic West Karekare màr ɗo millet RS AA Chadic West Kushi moo ɗo millet RMB AA Chadic West Miya màywá millet RS AA Chadic West Mwaghavul mààr millet RMB AA Chadic West Fyer mar millet RMB AA Chadic West Sirzakwai mar ɗay millet RMB AA Chadic West Zaar màrwá millet BCa NC EBC Plateau Ninzo amar millet RMB NC EBC Plateau Ningye mwan millet RMB NC EBC Plateau Anib àm ɛ̂ n millet RMB NC EBC Plateau Nyankpa imala millet RMB NC EBC Plateau Ashe i-ma millet RMB NC EBC Plateau Id ũ imara millet RMB NC EBC Plateau Shang mara millet RMB

3 Linguistic evidence for pearl millet: Roger Blench Circulation draft Ph Family Subgroup Language Attestation Gloss Source NC EBC Plateau Jili amo millet RMB NC EBC Plateau Sambe tìk àmâr millet RMB NC EBC Plateau Kwa ŋ mɛr millet RMB NC EBC Plateau Ya ŋkam marak millet RMB NC EBC Plateau Tarok imar millet Lo08 NC EBC Plateau Sur mər millet RMB NC EBC Plateau Pe ime millet RMB NC Adamawa Yoti múri millet K&Y NC Kwa Ga ŋmãã̀ ̀ millet, food Kr99 NC Kwa Adyukru máy ˋ mil ALKwCI NC Volta-Niger Nupoid Nupe mã̀yì millet Ba14 NC Bantu Jarawan Mbula mara millet MR NC Bantu Jarawan Mbat máár millet MR

Map 1 shows the locations of the language subgroups and individual ethnic groups listed in Table 2;

Map 1. Language subgroups and ethnic groups in Table 2

The presence of -ɗ- in the Chadic languages such as Karekare màr ɗo, argues that this is closer to the original form, which was simplified when borrowed into Plateau languages and Hausa (which is also likely from an ecological point of view). Pearl millet must have subsequently been carried along the coast, reaching Ga and Adyukru; hence the examples of the beach cultivation discussed above. However, the Hausa names have probably been borrowed into other languages, such as Miya, which looks suspiciously similar to Hausa. The widespread distribution of this term in Central Nigeria must reflect a major expansion of pearl millet at some later era, and this could be linked to the apparent significance of millet at the Nok sites mentioned in Kahlheber et al. (2009). There is no apparent link with the Bantu terms (see §3.4) which suggests that pearl millet may have spread southwards to the Bantu area from the Lake Chad region rather than from Central Nigeria.

Another, unrelated root occurs in the Central Chadic languages, found in the Mandara Mountains, principally in Northern Cameroun. Table 3 shows this root, reconstructed by Richard Gravina (p.c.) as something like # həya ; 4 Linguistic evidence for pearl millet: Roger Blench Circulation draft

Table 3. A Central Chadic root for millet Language Attestation Gloss Podoko hiyá ,-ə mil Hdi hiya le sorgho, le mil Mandara hiá mil (m) Lamang xíyá millet Sharwa hay ən graine Tsuvan he le mil Ouldeme hā y mil millet Moloko hay mil millet Vame āhá y mil millet Gidar haya mil Psikye xá millet, corn Bana xà mil (nom géné rique) Source : Compiled by Richard Gravina

It is possible that is was from this region that the millets of the Cameroun sites derive, but we would need more evidence from the poorly documented Adamawa languages that today exist in the region between the two language areas. 3.3 Berber Pearl millet was cultivated across most of North Africa at one time and a word for it is attested in most Berber languages. Meillassoux (1972: 390) discusses the reference in Ibn Battuta to the trade in anli , pearl millet, in the 14 th century in the Iwalata area, using the Berber term. Kossmann (p.c.) observes that it is probably proto-Berber, and he proffers the following attestations, shown in Table 4;

Table 4. Berber terms for pearl millet Lect Attestation Tashelhiyt anili, aynli Ntifa illan Kabyle ilni Ouargla in ǝlli Ghadames alele El Foqaha an ä́li, elli Awdjilah ílli, élli Tuareg en ăle Zenaga i’llän

There is something of a puzzle here, since if pearl millet was domesticated in the southwestern Sahara it is hard to see how it could have been part of proto-Berber, as the direction of expansion of Berber is usually considered to be from the northeast. Moreover, the archaeobotany of crop repertoires in the Saharan oases usually show wheat and barley in the initial phases with pearl millet appearing somewhat later. It is suggested here that the early Berbers encountered the crop as a staple in the hands of those they traded with in the Central Sahara. Only later did they begin to cultivate it for themselves in the oases. The origin of the word itself is problematic, as it is not obviously similar to any sub-Saharan language. Tukulor nutil and Songhay hèèni are possible resemblances but hardly convincing.

Documentation of millet names elsewhere in Africa remains very weak. Ethiopian sources lack significant names apart from Tigrinya mashela and Oromo bultuk (cf. Westphal 1975). The Arab traveller, Al Idrisi in Ethiopia. A more important lacuna is the region between ( ُد ) mentions the cultivation of dukhn ,(1154) Cameroun and Uganda, where very few names have been recorded, although pearl millet must have travelled along this corridor to reach East Africa.

5 Linguistic evidence for pearl millet: Roger Blench Circulation draft 3.4 Pearl millet names in the Bantu world Almost the whole of Central, Eastern and Southern Africa is the domain of the Bantu languages, except for Nilotic outliers in the east and Khoisan languages in the south. Much of this region initially appears unsuitable for the cultivation of pearl millet, especially the equatorial rain forest. However, the archaeological evidence from Cameroun shows that it can be inappropriate to read climatic conditions back into the past, especially where the introduction of crops with greater tolerance of humidity (maize and cassava) have obscured subsistence patterns that may have obtained until the sixteenth century. Bernard Clist (p.c.) reports that when the site of Rivière Denis on the Gabonese coast was occupied earlier than 3000 BP the region was savanna.

The linguistic evidence for Bantu pearl millet names has been studied by Phillipson & Bahuchet (1996) 2 and later Koen Bostoen (2007b). Bostoen concluded that there are two widespread roots for pearl millet *-cáng ʊ́ and *-bèdé , both of which are phonologically regular, suggesting that they represent inherited vocabulary rather than loanwords. The common Bantu term in the west is *-cáng ʊ́ , attested in Guthrie’s zones B, C, D, H, K, L, M and R. Nearly all present-day *-cáng ʊ́ reflexes are applied to bulrush millet, rather than sorghum, maize or other cereals. Languages with a *-cáng ʊ́ reflex are spoken in transition areas between the rain forest and the savannah, such as the Bandundu and Lower Kasai areas in the DRC. This suggests strongly that bulrush millet was known to early Bantu speakers, as does the archaeobotanical evidence.

The geographical distribution of the term *-bèdé is confined to the eastern and southern part of the Bantu domain, with reflexes attested in Guthrie’s zones D, E, F, G, J, K, L, M, N, P and S3. The referent of these is much more diverse, referring to pearl millet, finger millet, sorghum or maize. The term itself appears to have been borrowed from Nilotic languages, spoken in Uganda, Southern Sudan and Northern Kenya. Bostoen (2007b) provides a table of Nilotic terms for ‘sorghum’ which are typically bé(e)l or similar and which appear to be the source of the Bantu root. This implies that the Bantu emerged from the rainforest having entirely lost cereal cultivation. When they encountered Nilotes growing sorghum, they borrowed the term but applied it to a range of cereal crops they adopted from encounters with different resident cultivators, eventually carrying the term down towards Southern Africa. 4. Pearl millet in Asia

4.1 India and China Map 2. Pearl millet in India Portères (1976:433-441) provides a detailed overview of the subtypes of pearl millet in Africa and argues that the morphological evidence strongly supports early transmission to India. The absence of finds in the Nile Valley and the Near East support a sea-route. Although the cultivation of Pennisetum millet is well-established in many parts of India (Watt 1889-93: P.384), the archaeological record is frankly weak, with grains dated by context rather than directly. However, recent reviews of the evidence have confirmed that at least some records can be considered well-founded (Fuller 2003; Blench 2003). Possehl (1986) and Weber (1998) have both considered the impact of African millets on the subsistence systems of the subcontinent.

Map 2 shows the synchronic distribution of pearl millet in India. In the Indian literature, pearl millet is sometimes known as ‘black Italian millet’ and earlier sources give Holcus spicatus and Panicum

2 Thanks to Gerard Philippson for sending the original datasheets for this study. 6 Linguistic evidence for pearl millet: Roger Blench Circulation draft spicatum as scientific names. Table 5 shows a compilation of Indian names for pearl millet.

Table 5. Pearl millet names in India Phylum Language Transliteration Script #""""! ا Indo-European Urdu, Sindhi, Punjabi ba'ajra Indo-European Rajasthani, Gujarati, Marathi bajri बाजर Indo-European Hindi, Urdu and Punjabi bajra बाजरा Indo-European Nepali bājuro बाजरोु Dravidian Kannada sajje ಸೆ Dravidian Telugu sajjalu Dravidian Telugu ga ṇṭ e Dravidian Telugu kambu Dravidian Malayalam kampam Dravidian Tamil kambu க

Southworth (2005:204) attributes *b ājjara to Proto-North Indo-Aryan, but the suspicious similarity of names is may argue for a relatively recent secondary dispersal. Dravidian has a much wider diversity of names, and on the basis of what remains a rather small sample of terms, it is likely that pearl millet ws first adopted in a Dravidian area and only later spread further north. Zide & Zide (1976:8) give a proto-Munda term, * gan(-)gay , but this is generic for ‘large cereal’ and does not provide and convincing evidence pearl millet was known to the proto-Munda. None of the Indian names resemble any African terms.

The eFlora of China gives pearl millet as yu gu , 御谷, and notes that it is cultivated in north and east China. Bray (1984) gives a literary reference ca. 0 AD to a tall cereal crop with ‘seeds like mung beans’ which might be pearl millet, or simply a distinct cultivar of sorghum. There seems to be a significant lacuna in the information concerning its history and distribution. It is possible that it was brought to China at the same time as sorghum and finger-millet but did not spread and the modern cultivation represents a recent and unrelated introduction, perhaps for livestock feed. Bradley (2011) has no reference to it in reviewing the linguistic materials on millets. 4.2 Europe and the Near East Brunken et al. (1977) observe that pearl millet is absent from the historical and archaeological record in Ancient Europe and the Near East (cf. also Zohary & Hopf 2000). They quote the Spanish botanist Colmeiro who claimed that pearl millet was introduced into southern Spain following the Moorish invasions in the eighth century. The well attested forms for millet in the Berber languages (Table 4) do indeed make this a credible transfer. Neither of these terms have any obvious etymology, and are quite unlike Berber. There is no certain evidence for pearl millet in Ancient Egypt and earlier identifications of millet in intestines were probably sorghum (Darby et al. 1977:494; Germer 1985:224). However, it is found in early medieval Nubia; Fuller & Edwards (2001) report seeds of Pennisetum glaucum at Nauri in the sixth century. A survey of Oman by Hammer et al. (2009) recorded its presence, but whether this is an ancient crop, a medieval introduction from India or a modern fodder crop is unclear. Portères (1976) refers to the cultivation of pearl millet in Europe in the sixteenth century. 5. Synthesis and conclusions Pearl millet has historically been an important crop, carried across most of Africa and Eurasia over the preceding five millennia. Its ability to survive in drought-prone regions where soil fertility is low, as well as the importance of the stalks for livestock fodder, argue that it should be accorded greater significance in the international agronomic research system. Similarly, while the archaeobotanical picture in Africa is beginning to become clearer, there are striking gaps elsewhere in the world. China remains almost unknown, despite the historical and ethnographic records. Similarly, the movement into Europe, which may well have happened twice, once from the Maghreb and once from the east, is barely documented.

Bearing these caveats in mind, Map 3 presents a speculative map of the early diffusion of pearl millet

7 Linguistic evidence for pearl millet: Roger Blench Circulation draft cultivation.

Map 3. Diffusion of pearl millet cultivation in prehistory

Areas of secondary diversification, as suggested by linguistic evidence, are marked as blue ellipses. The proposed region of domestication follows recent literature, but it would not be difficult to imagine new archaeobotanical materials resituating this zone.

References Al-Idrisi 1154. Description de l'Afrique et de l'Espagne . [Trans. R.P.A. Dozy and M.J. de Goeje. 1969]. Amsterdam: Oriental Press. Amblard, S. & J. Pernes 1989. The identification of cultivated pearl millet ( Pennisetum ) amongst plant impressions on pottery from Oued Chebbi (Dhar Oualata, Mauretania). African Archaeological Review , 7:117-126. Amblard, Sylvie 1996. Agricultural evidence and its interpretation on the Dhars Tichitt and Oualata. In G. Pwiti and R. Soper. Aspects of African Archaeology, Papers from the 10 th Congress of the PanAfrican Association of Prehistory and Related Studies. 421-427. Harare: University of Zimbabwe Publications. Andrews, D.J. and K.A. Kumar 1992. Pearl millet for food, feed and forage. Advances in Agronomy , 48:90- 139. Appa Rao, M.H. Mengesha and V. Subramanian. 1982. Collection and preliminary evaluation of sweet-stalk pearl millet ( Pennisetum) . Economic Botany , 36:286-290. Appa Rao, S. Melak M. Mengesha and D. Sharma. 1985. Collection and Evaluation of Pearl millet (Pennisetum americanum ) germplasm from Ghana. Economic Botany , 39:25-38. Blench, Roger M. 2003. The movement of cultivated plants between Africa and India in prehistory. In: K. Neumann, A. Butler & S. Kahlhaber (eds.) Food, fuel and fields: progress in African Archaeobotany . 273-292. Köln: Heinrich-Barth-Institut. Blench, Roger M. 2006. Archaeology, Language and the African Past . Lanham: Altamira Press. Blench, Roger M. 2008. Using linguistics to reconstruct African subsistence systems: Comparing crop names to trees and livestock. In Timothy P. Denham, José Iriarte and Luc Vrydaghs (eds), Rethinking Agriculture: Archaeological and Ethnoarchaeological Perspectives, 598-644. California: Left Coast Press. Bostoen, Koen. 2007a. Bantu plant names as indicators of linguistic stratigraphy in the Western Province of 8 Linguistic evidence for pearl millet: Roger Blench Circulation draft Zambia. In Doris Payne (ed.), Selected Proceedings of the 37th Annual Conference on African Linguistics , 16-29. Somerville: Cascadilla Press. Bostoen, Koen 2007b. Pearl millet in early Bantu speech communities in Central Africa: A reconsideration of the lexical evidence. Afrika und Übersee 89: 183-213. Bradley, David 2011. Proto-Tibeto-Burman Grain Crops . Circulated Version of talk given at Cornell University, 2011. Brunken, J., J.M.J. de Wet and J.R. Harlan 1977. The morphology and domestication of pearl millet. Economic Botany , 31:163-174. Burkill, H.M. 1994. The Useful Plants of West Tropical Africa, Families E-I. Royal Botanic Gardens, Kew. Busson, F. et al. 1965. Plantes alimentaires de L'Ouest Africain . Ministère de la Coopération, Paris. Chevalier, A. 1932. Les productions végétales du Sahara. Revue de Botanique Appliquée et d’Agriculture Tropicale, 12:669-924. Clayton, W.D. & Renvoize, S. 1986. Genera Graminum . HMSO. D’Andrea, A.C., Casey, J. 2002. Pearl millet and Kintampo subsistence. African Archaeological Review 19:147–173 D’Andrea, A.C., Klee, M., Casey, J. 2001. Archaeological evidence for pearl millet ( Pennisetum glaucum ) in sub-saharan West Africa. Antiquity 75:341–348. Darby William J., Ghalioungui, Paul & Louis Grivetti 1977. The gift of Osiris . [2 vols] London/New York/San Francisco: Academic Press. Fuller, D.Q. 2003. Indus and non-Indus agricultural traditions: local developments and crop adoptions on the Indian Peninsula. Indus Ethnobiology: New Perspectives from the Field. S. A. Weber and W. R. Belcher eds. 343-396. Lanham; Boulder; New York; Oxford, Lexington Books. Fuller, D.Q. & D.N. Edwards 2001. Medieval plant economy in Middle Nubia: preliminary archaeobotanical evidence from Nauri. Sudan & Nubia, 5: 97-103. Fuller, D.Q. & R. Korisettar, Venkatasubbaiah, P.C., Jones, M.K. 2004. Early plant domestications in southern India: some preliminary archaeobotanical results. Vegetation History and Archaeobotany , 13: 115-129. Fuller, D.Q., Macdonald K., Vernet, R. 2007. Early domesticated pearl millet in Dhar Nema (Mauritania): evidence of crop processing waste as ceramic temper. In: Cappers R (ed) Fields of change. Proceedings of the 4th internat workshop for African archaeobotany . Barkhuis & Groningen University Library, Groningen, pp 71–76. Gast, M. 1968. Alimentation des populations de l'Ahaggar: étude ethnographique . Mem. du Centre de Recherches anthropologiques, préhistoriques et ethnographiques. Mem. 8. Arts et Metiers Graphiques. Paris : Flammarion. Gast, M. & Adrian, J. 1965. Mils et Sorgho en Ahaggar . Mem. du Centre de Recherches anthropologiques, préhistoriques et ethnographiques. Paris: Arts et Métiers Graphiques. Germer, R. 1985. Flora des Pharaonischen Agypten . Mainz am Rhein: Von Zabern. Hammer, K., J. Gebauer, S., Al Khanjari & A. Buerkert 2009. Oman at the cross-roads of inter-regional exchange of cultivated plants. Genetic Resources & Crop Evolution , 56:547–560. Hamon, S. ed. 1993. Le mil en Afrique. Montpellier: ORSTOM. Höhn, A., S. Kahlheber, K. Neumann, and A. Schweizer 2007. Settling the rain forest – the environment of farming communities in southern Cameroon during the first millennium BC. In J. Runge, editor, Dynamics of forest ecosystems in Central Africa during the Holocene: Past – Present – Future . London: Taylor and Francis, pp 29-41. Kahlheber, Stefanie, Koen Bostoen & Katharina Neumann 2009. Early Plant Cultivation in the Central African Rain Forest: First Millennium BC: Pearl Millet from South Cameroon. Journal of African Archaeology, 7 (2): 253-272. Kahlheber, Stefanie, A. Höhn, and N. Rupp 2009. Archaeobotanical Studies at Nok sites: an Interim Report. Nyame Akuma , 71: 2-15. Klee, M., Zach, B.& K. Neumann 2000. Four thousand years of plant exploitation in the Chad Basin of northeast Nigeria I: the archaeobotany of Kursakata. Vegetation History and Archaeobotany 9:223-237. Koernicke, F. & H. Werner 1885. Handbuch des Getreidesbaues . Berlin: Verlag von Paul Paery. Longtau, Selbut 2008. Linguistics and Archaeology: Historical Inferences from the Analysis of Cognates for Bulrush Millet in Plateau and a Review of Research Methodology . Paper presented at the Conference of the Linguistic Association of Nigeria, Uyo, 2007. 9 Linguistic evidence for pearl millet: Roger Blench Circulation draft Manning, Katie, Ruth Pelling, Tom Higham, Jean-Luc Schwenniger and Dorian Q Fuller 2010. 4500-year old domesticated pearl millet ( Pennisetum glaucum ) from the Tilemsi Valley, Mali: new insights into an alternative cereal domestication pathway. Journal of Archaeological Science 38 (2): 312-322. Meillassoux, Claude 1972. L'Itineraire D'Ibn Battuta de Walata a Malli. Journal of African History , 13, (3): 389-395. Needham, Joseph & Francesca Bray 1984. Science & Civilization in China, Vol. 6, Part 2 . Cambridge: Cambridge University Press. Neumann, Katharina 1999. Early plant food production in the West African Sahel. In: M. Van Der Veen (ed.) The exploitation of plant resources in Ancient Africa. 73-80. New York: Kluwer/Plenum. Neumann, Katharina, Aziz Ballouche & Marlies Klee 1996. The emergence of plant food production in the West African Sahel: new evidence from northeast Nigeria and northern Burkina Faso. In G. Pwiti and R. Soper. Aspects of African Archaeology, Papers from the 10 th Congress of the PanAfrican Association of Prehistory and Related Studies. 441-448. Harare: University of Zimbabwe Publications. Neumann, Katharina, Koen Bostoen, Alexa Höhn, Stefanie Kahlheber, Alfred Ngomanda & Barthelémy Tchiengué, 2011. First farmers in the Central African rainforest: A view from southern Cameroon. Quaternary International , 249: 53-62. Nixon, Sam, Mary Anne Murray & Dorian Q. Fuller 2011. Plant use at an early Islamic merchant town in the West African Sahel: the archaeobotany of Essouk-Tadmakka (Mali). Vegetation History and Archaeobotany , 20:223–239. Oumar, Ibrahim, Cédric Mariac, Jean-Louis Pham & Yves Vigouroux 2008. Phylogeny and origin of pearl millet ( Pennisetum glaucum ) [L.] R. Br) as revealed by microsatellite loci. Theoretical and Applied Genetics , 117:489–497. Perrot, Claude-Hélène 1990. Semailles et moissons dans la région d’Assinie vers 1700. Journal des Africanistes , 60(1):9-25. Philippson, Gérard & Serge Bahuchet 1996. Cultivated Crops and Bantu Migrations in Central and Eastern Africa: a linguistic Approach. Azania , 29/30: 103-120. Portères, R. 1958. Les appellations des céréales en Afrique: V. Le Millet, Teff de l’Abyssinie. Journal de l’Agriculture Tropicale et Botanique Appliqué , 5:454-463. Porteres, R. 1962. Berceaux agricoles primaires sur le continent Africain. J. African Hist. 3:195-210. Portères, R. 1976. African cereals: eleusine, fonio, black fonio, teff, Brachiaria, Paspalum, Pennisetum and African rice. In: Harlan J.R., De Wet J.M.J. & Stemler A.B.L. (eds.) Origins of African plant domestication . (Mouton) The Hague 1976:409-452. Possehl, Gregory L. 1986. African millets in South Asian protohistory. In: J. Jacobson. (ed.) Studies in the archaeology of India and Pakistan . 237-256. New Delhi: Oxford & IBH. Southworth, Franklin C. 2005. Linguistic archaeology of South Asia . London/New York: RoutledgeCurzon. Tostain, Serge. 1998. Le mil, une longue histoire: hypothèses sur sa domestication et ses migrations. In Monique Chastanet (ed.), Plantes et Paysages d Afrique. Une histoire à explorer , 461-490. Paris : Karthala. Watt, G. 1889-93. (ed.) A Dictionary of the Economic Products of India . 6 vols. Calcutta: Government of India. Weber, S.A. 1998. Out of Africa: the initial impact of millets in South Asia. Current Anthropology, 39(2), 1998:267-274. Westphal, E. 1975. Agricultural Systems in Ethiopia . Wageningen, Centre for Agricultural Publishing and Documentation. Zide, A.R.K. and Zide, N.H. 1976. Proto-Munda cultural vocabulary: evidence for early agriculture. In: International Conference on Austroasiatic Linguistics, lst, Honolulu, 1973. Austroasiatic studies. Edited by Philip N. Jenner, Laurence C. Thompson, Stanley Starosta, pp. 1295-1334. Hawaii: University Press of Hawaii. Zohary, D. & M. Hopf 2000. Domestication of plants in the old world: the origin and spread of cultivated plants in West Asia, Europe and the Nile Valley . New York: Oxford University Press.