18 Afro-Asiatic Overview

Afro-Asiatic (or Afroasiatic; also known as “Hamito-Semitic”) is an entity of genetically related languages which is often labeled a macro-family or language phylum due to the number and typological diversity of its member languages and the chronological depth of this entity. In any case the choice of the label is arbitrary and depends on taxonomic preferences of particular schools and individual authors within the framework of genealogical classification, since in this sphere there is no generally accepted and scientifically based taxonomy.

The name of this language phylum has a long history. The term “Semitic” was proposed by Leibniz in 1710 (French translation: Leibniz 2000). He was followed by German orientalist Schlözer¹ (1781: 161) who wrote:

From the Mediterranean Sea to the Euphrates and from Mesopotamia down to Arabia, as is known, only one language reigned. The Syrians, Babylonians, Hebrews and Arabs were one people. Even the Phoenicians who were Hamites spoke this language, which I might call Semitic.

The term “Semitic” was derived from the name of the elder son of Noah, Shem (Sem), who was the eponymous ancestor of certain peoples speaking the languages in question according to the biblical genealogy of peoples (Gen. 10). Following Schlözer, Renan (1855) introduced the term “Hamitic” (chamitique in French) for languages that exhibit common features with Semitic languages but are too different from them to be included in the Semitic entity, namely Ancient Egyptian, Coptic, and Berber. Renan also used the term “Cushitic” for the non-Semitic languages spoken in Ethiopia and the Horn of Africa. At first, these languages were labeled Ethiopian, but later this name was applied only to Semitic languages of Ethiopia. According to the biblical genealogy of peoples, Ham (Kam, Kham) is the second son of Noah, and Cush and Mizraim (Egypt in biblical Hebrew) are the sons of Ham. Cush (Kush) in Ancient Near East roughly corresponded to Nubia. Thus, Ancient Egyptian and Cushitic fit well into the biblical genealogy as members of the Hamitic entity. It must be noted that the biblical genealogy of peoples may in no way be considered linguistically based; rather, it reflects the geopolitical realities of the epoch, i.e. Canaan, according to this genealogy, is the son of Ham, while Elam is the son of Sem.²

Quite logically, from the 1860s the term “Hamito-Semitic” (henceforth H-S; also “Semito-Hamitic”) became widely used; among the pioneers in the field, R. Lepsius, L. Reinisch, T. Benfey, and F. Müller must be mentioned. Further development of H-S studies mainly consisted in the rapid growth of relevant linguistic data and the number of languages considered H-S. All languages showing certain parallels with but not sufficient parallels to classify them as Semitic, were treated as “Hamitic” within the H-S entity. In this context the problem of the affiliation of Hausa and Chadic languages in general became a most important issue. Already in the mid-nineteenth century the idea of linking Chadic with H-S was expressed by Newman (1844), and this hypothesis was taken up by other pioneers of H-S studies. It is necessary to emphasize here that Chadic languages form a unique group as to its composition, including more than 150 languages and dialect clusters with rather limited numbers of speakers, plus Hausa—the second biggest language of Africa south of Sahara after Swahili—and possessing a very high functional status in comparison with all other Chadic languages. Hausa shows various evident morphological parallels with classical Semitic languages and Ancient Egyptian, first of all in personal pronouns and preverbal markers. It may be surmised that if the sociolinguistic role of Hausa had been played by one of the Central Chadic languages, where isoglosses of this kind are mostly absent, the study of Chadic in the H-S context would probably have been delayed for decades.

The increase in linguistic data made it clear that the languages traditionally classified as “Hamitic” do not form a coherent linguistic unit within H-S, the only common feature being their not belonging to the Semitic family. Following scholars supporting this idea (Cohen 1952 and others), Greenberg (1955, 1963) in his studies of African linguistic classification argued that, since no “Hamitic” entity exists, all language groups included in it should be considered as independent branches within H-S on a par with Semitic languages. As a next step he proposed eliminating the term “Hamitic” from genealogical linguistic nomenclature and replacing H-S with “Afro-Asiatic” (AA), as this is the only important language unit embracing both Asia and Africa. However, this term is not very convenient because it may mean all the languages of Asia and Africa. Therefore, Diakonoff (1988) proposed changing it to “Afrasian”, which is widely used in Russian studies and also by some Anglophone authors. Yet other variants proposed by certain scholars—“Erythraic”, “Lisramic”, “Noahitic”—were not accepted by the linguistic community. At present, both variants are in wide use: AA and H-S or “Semito-Hamitic” (following European academic traditions), often on a par with AA. It may be noted that today the reason for eliminating the term “Hamitic” is not as valid as it was in the 1950s, when it implied a dichotomy into Semitic and Hamitic branches. The label H-S is just a conventional one, analogous to “Indo-European” or “Indogermanisch”; nobody would consider the latter a binary unit consisting of Indian and European (or even Indian and Germanic) language groups.³

The AA language phylum is one of the most intriguing and complex issues within the framework of comparative linguistics. It includes languages with ancient written traditions (Ancient Egyptian, Akkadian, Eblaite, Amorite, Phoenician, Ugaritic, etc.) as well as dozens of unwritten languages in sub-Saharan Africa. The chronological gap between classical dead and modern unwritten languages is uncontested in historical linguistics. Moreover, and in contrast to Indo-European, the ancient written monuments are attested only in two groups of AA: Semitic and Egyptian.

The paradigm of comparative linguistics (in the sense of Kuhn 1962) was drawn up in the 1870s within Indo-European studies by a group of scholars known as neogrammarians. However, non-paradigmatic conceptions in the field of linguistic reconstruction and genetic classification played an important part in AA studies, especially the attempts to explain lexical, phonetic, and grammatical parallels within AA by language mixing as a result of areal contacts and ancient processes of convergence. One must not forget that, in the mentality of the nineteenth and the first decades of the twentieth century it was very hard to accept the idea of a common origin of the languages of the great ancient civilizations and world religions on the one hand and “minor” unwritten languages, as spoken in western and central Sudan, on the other. This approach reached its apogee in the work of the distinguished German Africanist Meinhof (1912). He built his Hamitic hypothesis on the Hamitic theory as well known from the field of physical and social anthropology and history, according to which pastoralists of Hamitic origin brought to Black Africa a highly developed civilization and then mixed with local people. Meinhof surmised that “Hamitic” languages in West Africa emerged as a result of the mixture of the languages of incoming pastoral Hamites with local languages. Following Lepsius (1863), he emphasized the role of grammatical gender as a key linguistic criterion for classifying a language as “Hamitic”. According to the neogrammarian paradigm, however, typological criteria are not appropriate for postulating linguistic affinity, even though in Bantu comparative linguistics nominal classification is traditionally considered obligatory for classifying a language as Bantu (cf. Guthrie 1967–71). Most probably Meinhof, as a pioneer of Bantu comparative studies (Meinhof 1910), had chosen grammatical gender as a primary criterion for his classification of “Hamitic” because of its typological parallelism to the nominal classification of Bantu. Hence, he classified as “Hamitic” some Chadic languages possessing grammatical gender, and labeled this group as “Chado-Hamitic”, while other languages obviously related to them were classified as Chadic and non-H-S. On the basis of the nominal classification as the only linguistic criterion, he also included in his “Hamitic” family Fula (Peul), part of the Nilotic languages (referred to as “Nilo-Hamitic”), and some other languages of East and southern Africa. Meinhof’s nomenclature partially survived into the mid-twentieth century, when Greenberg published his classification of H-S and later changed its name to AA. Greenberg also excluded from this unit all languages classified by Meinhof as “Hamitic” on the mere basis of the typological criterion of grammatical gender or nominal classification.

Other scholars, albeit non-supporters of Meinhof’s approach, were reluctant to consider Chadic languages to be H-S. Such an attitude was somehow preserved until the 1980s, when Chadic languages were included in a volume on sub-Saharan languages (see Manessy 1981) but not in the H-S volume of the same series (see Cohen 1988).

The evolution of the status of Chadic within AA may be described as centripetal; the position of these languages as a self-standing unit within AA is no longer contested. In contrast to this the evolution of the genealogical status of Cushitic languages was centrifugal. The idea of excluding Bedauye (Beja) from Cushitic, while keeping it in AA, got no support. But the hypothesis of considering West Cushitic as an independent branch (i.e. Omotic) of AA (Fleming 1969; Bender 1975, 1976) is now supported by the majority of experts. Omotic is even considered by some scholars to be the most divergent branch of AA, while some others still support the idea of its Cushitic affiliation. This controversy reflects a general theoretical problem of genetic classification within the neogrammarian paradigm. The genealogical tree remains a sound and well-established model for presenting diachronic splitting within a genetic entity of languages. Postulating such an entity (absolute classification) as well as most closely interrelated language groups on the highest levels of the genealogical tree does not normally provoke much doubt. However, splitting on the medium level often remains obscure, since it may be very difficult or even impossible to unambiguously distinguish between genetically relevant isoglosses and areal ones induced by contacts between related languages.

Thus, it may be concluded that in the present state of research, the AA language phylum consists of five or six branches: Semitic, Egyptian, Berber, Cushitic, Omotic (or a single Cush-Omotic), and Chadic (see Map 18.1).

As to the external links of AA, the idea of relating Semitic to Indo-European was first proposed at the beginning of the twentieth century and was later developed for the whole AA phylum within the framework of Nostratic theory, bringing together several language families and phyla: Indo-European, AA, Uralic, Altaic, Kartvelian, and Dravidian. The further development of Nostratic studies may help to resolve the problem of the AA Urheimat (original homeland). At present the AA-internal arguments indicate eastern Sahara and adjacent territories as the most probable AA Urheimat, while Nostratic theory makes the Fertile Crescent hypothesis more probable. According to different hypotheses, the possible time span for the differentiation of Proto-AA varies from 1,000 to 7,500 bc; a more realistic estimate was provided by Diakonoff (1988: 23): “the speakers of Egyptian were the first to break away from the basic Proto-Afrasian nucleus not later than the 8th millennium B.C.”

The present state of the art in comparative AA linguistics may be characterized as follows: regular sound correspondences and important and diverse morphological isoglosses, as well as a rich common lexicon, create a solid base for future research. Despite its long history, however, AA comparative linguistics is still in its infancy. The few existing comparative vocabularies of AA are far from standard norms.

A comprehensive monographic overview of AA has been published by Diakonoff (1988). The comparison of this study with its earlier version (Diakonoff 1965) makes it possible to appreciate the changes in methodological approaches and main trends in AA comparative studies. Further progress is to be expected from comparative studies and linguistic reconstructions on lower levels of all AA groups and subgroups. In view of the enormous chronological gap between various branches and the time-depth of Proto-AA, only step-by-step reconstructions on intermediate levels may lead to conclusive results. A large number of languages are not documented and analyzed at all; for many others only short wordlists and very brief grammatical notes are available; thus, field research and documentation of non-described languages have absolute priority in AA linguistics.