Jeff Gaillard, M.A., (


The southern African Lemba have long maintain that they are a Jewish people directly descended from Jewish Babylonian exiles via Yemen. Cultural (linguistic) and Y-chromosome genetic analysis of their male lineages confirms a link to other Semitic-speaking peoples. The Lemba have also clearly have been influenced by neighboring their Bantu-speaking peoples but at their core, their historical Jewishness appears validated genetically. Disaggregating cultural and biological data allows us to better understand the specific history of the Lemba.  

The Lemba, an endogamous ethnic group located in southern Africa, have long held the belief of their Jewish ancestry. They purport to be a lost tribe of Israel.  According to their oral history, their ancestors came to Africa from Sena.  Once in Africa, they rebuilt their ancient city and later built Great Zimbabwe.  They were forced to leave Sena, because they had broken the holy laws by eating unclean food, and were dispersed among the Gentiles like other Jews.

While traveling in Africa, they followed the ngoma, which has been compared to the Ark of the Covenant, mentioned in the Hebrew Bible (Parfitt 2000).

An ancient ngoma found in a cave near the Limpopo was of similar size to the Ark: 1.4 feet high and 2.5 feet in circumference; it also had four handles at each corner…As the Ark had the priestly caste to tend it – the Levites – so the ngoma had the priestly caste of the Lemba.  The ngoma, like the Ark, was carried into battle and led the people on their wanderings throughout Africa; it was too holy to be placed on the ground, so at the end of a day’s march, it was hung from a tree or a special platform was constructed for it (Parfitt 2000:187-188).

 Today, the Lemba are commonly called the “Black Jews,” however, they were once referred to as “Kruger’s Jews,” as it was believed that Paul Kruger, President of the Transvaal, was responsible for their discovery (Thomas et al. 2000).  The Lemba do not eat pork and practice circumcision.  Lemba means the people who refuse, i.e., refuse to eat pork and to marry Gentiles. Biblical and Semitic names are common among the Lemba (Parfitt 2000). They speak a language known as Hiberu (Jones 1996).  “Physically, many Lemba have a distinctive appearance: angular features with prominent hooked nose” (Van Warmelo 1974:81).  Parfitt (2000) has also noted that some of the Lemba are similar to Middle Eastern Jewry or Peninsula Arabs.  It has been suggested that the Lemba might be of Muslim origin, as Islam has similar dietary laws and social customs as Judaism (Thomas et al. 2000; Mandivenga 1983). Jones posits that “Lemba Jewishness stems in part from a diluted and almost forgotten Islam” (Jones 1996:158).

Historically, the Lemba have always been a group that has produced goods, such as smelted metals and fine pottery.  This set them apart from their Bantu speaking neighbors, who have primarily been occupied with stock-keeping and grain-growing (Van Warmelo 1974).  Often the Lemba were employed in the courts of Bantu kings, as confidants and advisors.  The kings were privy to the Lemba’s savvy in trading, which resulted in a natural shrewdness.  In light of this data and further research, Van Warmelo  was convinced that the Lemba were of Semitic origin.  He offers the following:

All these facts point to the conclusion that the Lemba are the descendants of Semitic traders from the East Coast, in other words, of Arabs.  This is supported by the texts of prayers which have more recently come to light.  They are unintelligible and look unlike Bantu, but suggest some mangled suras from the Koran.  The words Saidi sangu or Zaidi zangu are constantly repeated and are said to mean “my master” (cf. Ar. seyyid “lord”).  In the family name Hadzhi (Haji) and the oath Seremane (Ar. Sulaiman) we see a link with what is mentioned from the Arab side further down (Van Warmelo 1974:82).

 Mourant (1978) suggested that the Lemba represented a thoroughly African blood-group picture, even though they displayed Jewish cultural features.  Acknowledging that the Lemba had done skilled metalwork, particularly for the Arabs.  He opines that:

This suggests possible connections with the modern Yemenite and with African Jewish silversmiths, and possibly also with the nomadic Jewish blacksmiths of Khaibar in Arabia, and other Jewish blacksmiths on the northern boarders of the Sahara (Mourant 1978:37).

 The Lemba unequivocally believe that they are direct descendants of Jewish exiles who returned from Babylon.  There is a record of “the sons of Sena’ah, three thousand nine hundred and thirty” in the Old Testament (Ezra 2:35, Nehemiah 7:38).  Parfitt (2000) has located a city in modern Yemen, called Sena, which the Lemba believe is their ancestral home.

According to Jones (1996), many people have an unclear association with Judaism, yet feel that they have a special connection to ancient Israel.  In many cases, upon close examination, their belief in Jewish ancestry is not valid.  The Lemba have been accused of wishful thinking (Jones 1996; Parfitt 2000).  The mythic Sena may not be the city of Sena in Yemen, but possibly was located in Africa.  In David Livingstone’s Travels, Parfitt (2000) uncovered the following passage:

The village of Senna stands on the right bank of the Zambesi…the soil is fertile; but the village, being in a state of ruin, and having several pools of stagnant water, is very unhealthy (Parfitt 2000:196).

 Could Livingstone have been describing the original ancestral home of the Lemba?  Parfitt (2000) was not convinced that Livingstone was describing the place of origin of the Lemba, however, the passage does deserve exploration.  Hiberu, which sounds much like Hebrew, is not a Semitic language (Jones 1996), but related to Shona, an African language. Jones (1996) avers that the Lemba have simply incorporated a European myth of ancient Israelite builders of Great Zimbabwe as part of their history. He posits that the incorporation of myth as history has allowed the Lemba to become a “Lost Tribe” with a noble past.  Jones is adamant that “the Lemba have no direct intellectual connection with Judaism, but picked up its elements from a related source” (Jones 1996:159), such as Arab traders who took local wives and passed on Moorish customs.  Jones believes this would account for circumcision and the strict Lemba dietary laws.  Within the scope of population genetics, Jones’ argument seems plausible, as genes are definitely passed on by mating.

 Diversity in systems of genetic variation (Jones 1996) are now being used to accurately map the paths of populations.  In order to ascertain the best possible scientific answer, with respect to the origin of the Lemba, a review of the current Y chromosome research concerning them is necessary.  The information obtained via genetic studies allows us to better weigh the claims of oral histories and any written records.

Movement of people between faiths confuses the biological frontiers as, whatever their religion, their descendants will preserve the genes of their unconverted ancestors.  Sometimes, a conversion long ago has genetic and political resonances today (Jones 1996:156).


The Y chromosome has been very useful in molecular biological studies.  It is a single haploid DNA chain (Hammer 1994; Spurdle et al. 1994), which does not recombine, except for the psuedoautosomal region, and is passed on paternally.  The Y chromosome allows for the construction of patrilineal genealogical cladograms, similar to those used in mitochondrial DNA (Skorecki 1997).  Polymorphisms of the Y chromosome which mutate often include microsatellites and the single-nucleotide substitutions, along with biallelic YAP Alu insert (Thomas et al. 2000).

Biallelic polymorphisms are called “unique event polymorphisms” (UEP), since they are not likely to have occurred more than one time in the evolution of humans (Thomas et al. 1998; 2000).  The UEP’s categorize the Y chromosomes into dissimilar genealogical groupings. Y chromosomes are important in observing geographically distant populations, which have experienced protracted admixture with outside groups (Thomas et al. 1998; 2000). 

Spurdle and Jenkins (1996) examined the allele frequencies of RFLP polymorphisms in the Lemba and several other populations (including Semitic groups) to secure genetic affinities and the possible origin of the Lemba.  They compared frequency observation of four different RFLP polymorphisims, specifically: YAP loci, p12F2, p49a and pD31, which yielded the following:


The results suggest that ³ 50% of the Lemba Y chromosomes are Semitic in origin, ~ 40% are African, and the ancestry of the remainder cannot be resolved.  These Y-specific genetic findings are consistent with Lemba oral tradition, and analysis of the Jewish people and their association with Africa indicates that the historical facts are not incompatible with theories concerning the origin of the Lemba (Spurdle and Jenkins 1996:1126).


 Spurdle and Jenkins (1996) noted that a frequency of 2.6 for the p12F2/TagI 8-Kb allele was present in the Lemba.  This finding is of import, since this allele is not present in African populations, nor was it found to be present in a sample of 60 Polynesians.  The researchers were certain that the p12F2 data accurately reflects input into the Lemba gene pool from “Caucasoid” males.

This data is in agreement with the result from the three other Y-linked RFLP’s, namely: p YAP, p49a/TagI and pDp31 (Spurdle et al. 1994; Sprudle and Jenkins 1996).  Nine dissimilar haplotypes have been detected in Lemba samples. Six of the haplotypes occur at frequencies > .05 (Spurdle and Jenkins 1996).

There is indication of the Ht4 haplotype (which is typically African) found at a frequency of .20, which is a clear marker of African male gene flow. This African gene flow into the Lemba gene pool is joined by high frequencies of Ht7, Ht8 and Ht11, indicating “Caucasoid” male gene flow.  These three haplotypes accounted for 53% of the Lemba Y chromosomes that the researchers studied (Spurdle and Jenkins 1996).  They concluded that the Ht7, Ht8 and Ht11 haplotypes are typical of Jewish populations, since another study yielded similar frequencies in a Lebanese control group (Santachiara Benerecetti et al. 1993), which could be representative of all Semitic groups (Spurdle and Jenkins 1996).

These same haplotypes are also present in the South African Asiatic Indian population.  It is not possible to determine the exact source of “Caucasoid” gene flow with just these three haplotypes. The Ht13 haplotype, found at a frequency of .13 in South African Asiatic Indians (Spurdle and Jenkins 1996), is completely absent from the Lemba.  Thus, the “Caucasoid” male genes may be safely assigned as Semitic.  Gene flow from European “Caucasoids” is not a possibility, as their signature haplotype, Ht15, is not present in the Lemba gene pool (Torroni et al. 1990; Spurdle and Jenkins 1996).

It is interesting to note that the allele frequencies of the Lemba are significantly different from those of both the Bantu-speaking African population (P << .005) and the South African European population (P < .005) but not from those of the South African Jewish group (.05 < P < .10) [(Spurdle and Jenkins 1996:1129)].

 Spurdle and Jenkins (1996) Y-specific genetic data does not allow them to accurately establish a Jewish or Arabic Semitic ancestry for the Lemba.  However, they offer the following explanation:

This is not surprising, since a common ancient history of Jews and Arabs is reflected both in their languages and as similarities in the stories of Judaism and Islam recorded in the Jewish Bible and the Koran, respectively (Spurdle and Jenkins 1996:1131).

 The researchers suggest that certain aspects of the Lemba culture would favor Jewish ancestry, as opposed to an Arabic one (Spurdle and Jenkins 1996).  Many of the food laws among the Lemba are Jewish and not Islamic, e.g., the practice of separating meat and milk.  Muslim law forbids the consumption of alcohol (Mathivha 1992), yet several Lemba ritual sacrifices involve the use of alcohol.  The method of male circumcision among the Lemba is quite different from that practiced amongst Muslims (Spurdle and Jenkins 1996), but is similar to that which was performed in biblical times (Gutmann 1987). Thus, “historical data concur with many of the suggestions based on Lemba oral history” (Spurdle and Jenkins 1996:1131).


Before the destruction that took place in Jerusalem in the sixth century BC, many Jews freely entered into Yemen (Nyrop 1985).  They lived and worked as skilled artisans

 in Sena as specialist in metalwork and pottery (Goitein 1971).  These facts coincide with Lemba oral history, which states that their ancestors who came from Sena (only males made the voyage to Africa) were artisans and craftsman (Van Warmelo 1974).  Spurdle and Jenkins opine that “it is entirely possible that the ancestors of the Lemba were Jewish craftsmen and traders from Sena in Yemen” (Spurdle and Jenkins 1996:131).


The Lord said to Moses:
Then you shall bring Aaron and his sons to the door of the tent of meeting, and shall wash them with water, and put upon Aaron the holy garments, and you shall anoint him and consecrate him, that he may serve me as priest.  You shall bring his sons also and put coats on them, and anoint them, as you anointed their father, that they may serve me as priests; and their anointing shall admit them to a perpetual priesthood throughout their generations (Exodus 40:12-15).

According to the Hebrew Bible, Aaron and his male descendants were selected to be perpetual priests.  Thus, Aaron was the first Cohen, and the priestly line is known as Cohanim (Travis 1998; Thomas et al. 1998; Parfitt 2000; Thomas et al. 2000).  Jewish males are separated into three distinct castes, specifically “Cohanim (the paternally inherited priesthood), Leviim (non-Cohen members of the paternally defined priestly tribe of Levi), and Israelites (all non-Cohen and non-Levite Jews)” [Thomas et al. 2000:675]. Thomas et al. (1998) conducted Y chromosome research on present day Cohanim and Levites. The researchers were able to “trace the origin of Cohen chromosomes to about 3,000 years before present, early during the Temple period” (Thomas et al. 1998:138).  They found that Levite Y chromosomes are very diverse, but the “Cohen chromosomes are homogeneous” (Thomas et al. 1998:138).  Their research would become pivotal concerning the Lemba and their disputed origins.

For highly polymorphic, single-locus systems, the identification of haplotypes with restricted distributions may provide “signatures” of ancient connections that have been partially obscured by subsequent mixing with other populations.  Gene flow from the Cohenim could account for the presence of the Cohen modal haplotype in both Ashkenazic and Sephardic Israelites, or it could be a signature of the ancient Hebrew population.  The Cohen modal haplotype may therefore be useful for testing hypotheses regarding the relationship between specific contemporary communities and the ancient Hebrew population (Thomas et al. 1998:139).

 Thomas et al. (2000) examined six distinct populations, namely: Lemba, Ashkenazic Jews, Bantu, Yemeni-Hadramant, Yemni-Sena, and Sephardic Jews, in order to obtain a clearer picture of Lemba paternal genetic heritage. Thus, 399 Y chromosomes were analyzed for six biallelic markers and six microsatellites within the six populations (Thomas et al. 2000).   They discovered that one of the Lemba clans, the Buba, exhibited a high frequency of the Y chromosome Cohen modal haplotype (CMH).  Interestingly, according to Lemba oral history, it was the Buba who led the Lemba out of Judea to Yemen, where they built the city of Sena (Thomas et al. 2000).

Parfitt (2000) has established that the Lemba are currently composed of > 12 clans, and some of these correlate with place names in Yemen, specifically the eastern Hadramaut.  Ironically, Parfitt (2000) postulates that he has located the original city of Sena, of Lemba oral history, in this same region.  Amongst the Lemba, the Buba clan is the most important in some rituals, is believed to be the oldest, and is respected as senior clan (Parfitt 2000).

Thomas et al. (2000) noted that 12 polymorphisms were prevalent in the multiple Jewish populations that they sampled: “six microsatellites: DYS19, DYS388, DYS390, DYS391, DYS392, and DYS393) and six UEP (YAP, SRY4064, sY81, SRYt465, 92R7, and Tat” (Thomas et al. 2000:675). They found the CMH frequency in lay Jews to be moderate at (~.12) and at a low frequency or absent amongst Palestinian Arabs, Armenians, Greeks, Cypriots, Yakut, Mongolians and Nepalese.  As the CMH is of high frequency (~. 50) in the Jewish priesthood (Thomas et al. 2000), the researchers believed that an absence of this haplotype among Bantu populations, would buttress the Lemba assertion of paternal Jewish ancestry, “especially if its frequency is relatively low in other Semitic groups” (Thomas et al. 2000:675).

It was found that the YAP+ chromosome frequency in the Lemba samples were intermediate between that of the Bantu and Semitic groups (Thomas et al. 2000).  The CMH or Haplotype 34 was found to be absent among the Bantu and Sena sampled populations (Thomas et al. 2000). A high proportion of Bantu Y chromosomes were found in the Lemba, along with Haplotype 34 (CMH) at .088, which is representative of all Lemba Y chromosomes (Thomas et al. 2000).

Microsatellite variability revealed that a high repeat number of DYS388 alleles is a unique feature of the Near East (Thomas et al. 2000). Among European samples (Kayser et al. 1997), no variation of DYS388 was found, and the samples had a low repeat number allele.  Thomas et al. (2000) found a low repeat number of DYS388 alleles among their Bantu samples, but found their Semitic samples had Y chromosomes with a high incidence of high repeat number of DYS388 alleles.

Lemba Y chromosomes, on the other hand, were frequently closely associated with either Bantu or Semitic Y chromosomes, but only very rarely with both of them simultaneously.  Given that differences can be observed between Semitic and Bantu chromosomes, it is possible to suggest a Semitic or Bantu origin for Lemba Y chromosomes, on the basis of their genealogical proximity to Bantu and Semitic types (Thomas et al. 2000:681).

 Thus, 32.4% of Lemba Y chromosomes can be labeled of Bantu origin, and the majority (67.6%) of Semitic origin (Thomas et al. 2000).  With respect to the earlier mentioned Buba clan, the researchers noted the following:

Of particular interest is the Buba clan, since membership of this clan and possession of the CMH are significantly associated (P < .001).  Seven of the 11 clan-designated Lemba CMH Y chromosomes came from members of this clan, whereas 7 (Northern Province, 4/4; and Sekhukuneland, 3/9) of the 13 Buba have the CMH (Thomas et al. 2000:682, 684).

 Y chromosome haplotypes that are present in more than one population, with their absence from other population groups, would indicate a common origin or male mediated gene flow (Thomas et al. 2000). Although support of Jewish genetic contribution to the Lemba is valid, with the presence of the CMH, the researchers caution that “it is possible that the Lemba CMH Y chromosomes are a consequence of a relatively recent event that, in Lemba oral tradition, has acquired a patina of antiquity” (Thomas et al. 2000:685).  “There are, nevertheless, some subtle patterns in the genes that uncover the hidden history of those who carry them” (Jones 1996:153).  It appears that there is a need for caution in relating oral history and genetic data to verify historical facts. However, the preponderance of the evidence (Thomas et al. 2000) currently supports the Lemba claim of Judaic ancestry.

Thomas et al. (2000) conclusions concerning the Lemba are delimited as follows:

The genetic evidence revealed in this study is consistent with both a Lemba history involving an origin in a Jewish population outside Africa and male-mediated gene flow from other Semitic immigrants (both of these populations could have formed founding groups for at least some of the Lemba clans) and with admixture with Bantu neighbors; all three groups are likely to have contributors to the Lemba gene pool, and there is no need to present an Arab versus a Judaic contribution to that gene pool, since contributions from both are likely to have occurred.  The CMH present in the Lemba could, however, have an exclusively Judaic origin (Thomas et al. 2000:685).


Demic diffusion (the genetic absorption of one population by another) did not take place with the Lemba and their Bantu neighbors (Jones 1996). Van Warmelo explicates the reason for the Lemba’s endogamy:

The Lemba did not give their daughters in marriage to any but their own people, and avoided marriage with Bantu women.  This strict endogamy was the secret of their survival as a distinct people. It was based on the dogma of the “uncleanness” of non-Lemba, who ate what the Lemba had been taught was forbidden, viz. pork, certain other animals and the flesh of cattle not Kosher–killed according to their law (Van Marmelo 1974: 81-82).

 The Lemba’s endogamy has allowed them to remain biologically distinct from the Bantu, even though they do share some similar genetic traits.  The Lemba’s genes link them together into a diffuse family more clearly related to Semitic peoples. On 31 October 1999, several of the Lemba elders were invited to Pretoria, to the residence of President Thabo Mbeki. After the visit, the Lemba have become more recognized as an African Jewish tribe in southern Africa (Parfitt 2000).

 An American Jewish educator, Yaacov Levi, with degrees in animal and fisheries science came to South Africa on 1 December 1999 (Parfitt 2000). He arrived ostensibly to teach the Lemba normative Judaism. Subsequently, many American Jewish communities have sent aid and thousands of books to the Lemba (Parfitt 2000).  These events will likely help the Lemba achieve their goal, i.e., of being recognized as a lost tribe of Israel, and validation of their oral history.  Jones may be correct, “sometimes genes confirm the details of history” (Jones 1996:154).***


I would like to thank Dr. Fatimah Jackson, Dr. M.F. Berry, A. Awolana, A. Santos, A. Webb-McNeil, Professor Ken Kidd, Dr. Andrew Pakstis, and Dr. S.O.Y. Keita 


  1. Goitein SD (1971) Jewish society and institutions under Islam.  In: Ben–Sasson HH, Ettinger S (eds) Jewish society through the ages.  Shocken, New York:170-184.

  2. Gutmann J (1987) The Jewish life cycle: iconography of religions.  Sec 23: Judaism. EJ Brill, Leiden.
  3. Hammer MF (1994) A recent insertion of an Alu family member on the Y chromosome is a useful marker for human population studies. Mol. Biol. Evol. 11: 749-761.
  4. Jones S (1996) In the Blood – God, Genes and Destiny. HarperCollins, London.
  5. Kayser M, Caglia A, Corach D, Fretwell N, Gehrig C, Graziosi G, Heidorn F, et al. (1997) Evaluation of Y-chromosomal STRs: a multicenter study. Int. J. Legal Med. 110: 125-133.
  6. Mandivenga EC (1983) Islam in Zimbabwe. Mambo Press, Gweru, Zimbabwe.
  7. Mathiva MER (1992) The Basena/Vamwenya/Balemba.  Merester, Toyander, South Africa.
  8. Mourant AE, Kopec AC, Domaniewska – Sobczak K (1978) The genetics of the Jews.  Clarendon, Oxford.
  9. Nyrop RF (1985) The Yemens – country studies.  American University Foreign Area Studies.  US Government Printing Office, Washington, DC.
  10. Parfitt T (2000) Journey to the vanished city – the search for a lost tribe of Israel.  Vintage, New York.
  11. Santachiara Benerecetti As, Semino O, Passarino G, Torroni A, Brdicka R, Fellous M, Modiano G (1993) The common, Near-Eastern origin of Ashkenazi and Sephardi Jews supported by Y-chromosome similarity.  Ann. Hum. Genet. 57: 55-64.
  12. Skorecki K, Selig S, Blazer S, Bradman R, Bradman N, Warburton PJ, Ismajlowicz M, et al. (1997) Y chromosomes of Jewish priests.  Nature 385:32.
  13. Spurdle AB, Hammer MF, Jenkins T (1994) The Alu polymorphism in southern African populations and its relationship to other Y-specific polymorphisms.  Am. J. Hum. Genet. 54: 319-330 (Erratum: Am. J. Hum. Genet. 54: 929).
  14. Spurdle AB, Jenkins T (1996) The origins of the Lemba “Black Jews” of southern Africa: evidence from p12F2 and other Y-chromosome markers.  Am. J. Hum. Genet. 59: 1126-1133.
  15. Thomas MG, Skorecki K, Ben-Ami H, Parfitt T, Bradman N, Goldstein DB (1998)  Origins of Old Testament priests. Nature 394: 138-140.
  16. Thomas MG, Parfitt T, Weiss DA, Skorecki K, Wilson JF, le Roux M, Bradman N, Goldstein DB (2000) Y chromosomes traveling south: the Cohen modal haplotype and the origins of the Lemba – the “Black Jews of southern Africa.” Am. J. Hum. Genet. 66: 674-686.
  17. Terroni A, Semino O, Scozzari R, Sirugo G, Spedini G, Abbas N, Fellous M, et al. (1990) Y chromosome DNA polymorphisms in human populations: differences between Caucasoids and Africans detected by 49a and 49f probes.  Ann. Hum. Genet. 54: 287-296.
  18. Travis J (1998) The priests’ chromosome? DNA analysis supports biblical story of the Jewish priesthood. Science News 154: 218-219.
  19. Van Warmelo NJ (1974) The classification of cultural groups.  In: Hammond-Tooke WD (ed) The Bantu-speaking peoples of southern Africa. Routledge & Keagan Paul, London: 56-84.

© W. Montague Cobb Research Laboratory

Download the PDF of this Article! >