Bidhaawyeet, Tubdhaawi, تُبڈاوِ
Native toSudan, Eritrea, Egypt
RegionRed Sea
Native speakers
(2.5 million cited 1993–2020)[1]
Latin alphabet (limited use)
Arabic alphabet (limited use)
Official status
Recognised minority
language in
Language codes
ISO 639-2bej
ISO 639-3bej

Beja (Bidhaawyeet or Tubdhaawi) is an Afroasiatic language of the Cushitic branch spoken on the western coast of the Red Sea by the Beja people. Its speakers inhabit parts of Egypt, Sudan and Eritrea. In 2017 there were 2,220,000 Beja speakers in Sudan according to Ethnologue. In 1993 there were 77,000 speakers of Beja in Egypt, current number of speakers today in Egypt is unknown. As of 2020 there are an estimated 201,000 Beja speakers in Eritrea. The total number of speakers in all three countries is 2,498,000.[2]


The name Beja, derived from Arabic: بجا, romanizedbijā, is most common in English-language literature. Native speakers use the term Bidhaawyeet (indefinite) or Tubdhaawi (definite).[3]


Beja is held by most linguists to be part of the Cushitic branch of the Afroasiatic family, constituting the only member of the Northern Cushitic subgroup. As such, Beja contains a number of linguistic innovations that are unique to it, as is also the situation with the other subgroups of Cushitic (e.g. idiosyncratic features in Agaw or Central Cushitic). The characteristics of Beja that differ from those of other Cushitic languages are likewise generally acknowledged as normal branch variation.[4]

The relation of the Northern Cushitic branch of Cushitic to the other branches is unknown. Christopher Ehret proposes, based on the devoicing of Proto-Cushitic voiced velar fricatives, that Northern Cushitic is possibly more closely related to South Cushitic than to the other branches.[5]

The identification of Beja as an independent branch of Cushitic dates to the work of Enrico Cerulli between 1925 and 1951.[6] Due to Beja's linguistic innovations, Robert Hetzron argued that it constituted an independent branch of Afroasiatic.[7] Hetzron's proposal was generally rejected by other linguists, and Cerulli's identification of Beja as the sole member of a North Cushitic branch remains standard today across otherwise divergent proposals for the internal relations of the Cushitic language family.[8]


The ostracon which Francis Llewellyn Griffith believed bore writing in the Blemmye language. Gerald M. Browne and Klaus Wedekind believe this to be the ancestor of the modern Beja language.

Christopher Ehret proposes the following sequence of sound changes between Proto-Cushitic and Beja:[9]

  1. PC *ts' → *c' (alveolar ejective affricate becomes palatal ejective stop)
  2. PC *t' → *ts' (dental ejective stop becomes alveolar ejective affricate)
  3. *C' → C (ejectives become their non-ejective voiceless counterparts)
  4. [+lateral/+obstruent] → [+retroflex/+obstruent] (that is, /dl/ and /tl'/ become /ɖ/ and /ʈ/, respectively)
  5. PC *dz → *ts (voiced alveolar affricate becomes voiceless)
  6. *tss; *cʃ (voiceless alveolar affricate becomes a fricative, voiceless palatal plosive becomes a postalveolar fricative
  7. *pb
  8. PC *ɣʷ → * (labialized voiced velar fricative becomes voiceless)
  9. *x⁽ʷ⁾ → *k⁽ʷ⁾ (velar fricatives become plosives)
  10. PC *ɬr /V_V (lateral fricative becomes alveolar tap between vowels)
  11. PC *ɬl /#_ (lateral fricative becomes lateral approximant word-initially)
  12. PC *z → j /V_ (a consonant of unknown value becomes palatal approximant after vowels)
  13. PC *z → d /#_ (the same consonant of unknown value becomes voiced alveolar stop word-initially)
  14. PC *ŋ, *ɳn (all nasals but /m/ collapse into alveolar nasal)

Ehret's reconstructed Proto-Cushitic /z/ is not a voiced alveolar fricative, but a consonant of unknown value. Ehret proposes that it might be a voiced palatal plosive /ɟ/.[10]

Some linguists and paleographers believe that they have uncovered evidence of an earlier stage of Beja, referred to in different publications as "Old Bedauye" or "Old Beja." Helmut Satzinger has identified the names found on several third century CE ostraca (potsherds) from the Eastern Desert as likely Blemmye, representing a form of Old Beja. He also identifies several epigraphic texts from the fifth and sixth centuries as representing a later form of the same language.[11] Nubiologist Gerald Browne, Egyptologist Helmut Satzinger, and Cushiticist Klaus Wedekind believed that an ostracon discovered in a monastery in Saqqarah also represents the Old Beja language. Browne and Wedekind identified the text as a translation of Psalm 30.[12]


Labial Dental/
Retroflex Palatal Velar Glottal
plain labialized
Nasal m n (ɳ) (ɲ) (ŋ)
Plosive voiceless t ʈ k ʔ
voiced b d ɖ ʤ ɡ ɡʷ
Fricative voiceless f s ʃ (x) h
voiced (z) (ɣ)
Trill r
Approximant w l j

Nasals other than /m/ and /n/ are positional variants of /n/. The consonants /χ/ and /ɣ/ only appear in Arabic loanwords in some speakers' speech; in others', they are replaced by /k/ or /h/ and /g/. Some speakers replace /z/ in Arabic loanwords with /d/.[14]

Beja has the five vowels /a/, /e/, /i/, /o/, and /u/. /e/ and /o/ only appear long, while /a/, /i/, and /u/ have long and short variants.[15]

Front Central Back
Close i u
Open a

Beja has pitch accent.[16]


Both Roman and Arabic script have been used to write Beja.[17] The Roman orthography below is that used by the Eritrean government and was used in a literacy program at Red Sea University in Port Sudan from 2010 to 2013. Three Arabic orthographies have seen limited use: The first below was that used by the now defunct Website Sakanab; the second was devised by Muhammad Adaroob Muhammad and used in his translation of E.M. Roper's Beja lexicon; the third was devised by Mahmud Ahmad Abu Bikr Ooriib, and was employed briefly at Red Sea University in 2019. No system of writing has gained wide support. The only system to have been employed in publications by more than one writer is the Latin script.[18]

Beja Orthographies
IPA Roman[19] Arabic (Usakana)[19] Arabic (MHM)[20] Arabic (RSU)[21]
ʔ ' ء
b b ب
d d د
ɖ dh ذ ݚ
f f ف
g g ق گ
ɣ gh غ
gw قْو گْو
h h ه
ʤ j ج
k k ك
kw كْو
l l ل
m m م
n n ن
r r ر
s s س
ʃ sh ش
t t ت
ʈ th ث
w w و
x kh خ
j y ي

In the Roman orthography, the vowels are written with the letters corresponding to the IPA symbols (i.e., ⟨a, e, i, o, u⟩). Long vowels are written with doubled signs. As /e/ and /o/ cannot be short vowels, they only appear as ⟨ee⟩ and ⟨oo⟩, respectively.

The single ⟨e⟩ sign, however, does have a use: To distinguish between /ɖ/ and /dh/, ⟨dh⟩ is used for the former and ⟨deh⟩ for the latter. Similarly, ⟨keh⟩ is /kh/, ⟨teh⟩ is /th/, ⟨seh⟩ is /sh/. Single ⟨o⟩ is not used.

In all Arabic orthographies, short vowels are written with the same diacritics used in Arabic: fatḥah for /a/ (ﹶ), kasrah for /i/ (ﹺ), ḍammah for /u/ ( ُ). 'Alif (ا) is used as the seat for these diacritics at the beginning of a word. Long /aː/ is written with 'alif (ا) preceded by fatḥah, or alif maddah (آ) when word-initial. Long /iː/ is written with yā' ي preceded by kasrah. Long /uː/ is written with wāw و preceded by ḍammah. The systems vary on the representation of long /eː/ and long /oː/. In the Usakana system, /eː/ is written with a modified Kurdish yā' ێ; in the system devised by Muhammad Adaroob Muhammad it is represented by yā' with a shaddah يّ; in the Red Sea University system, it is not distinguished from the yā' for /j/ or /iː/. In the Usakana system, /oː/ is written with a modified Kurdish wāw ۆ; in the system devised by Muhammad Adaroob Muhammad it is represented by wāw with a shaddah وّ; in the Red Sea University system, it is not distinguished from the wāw for /w/ or /uː/.

Pitch accent is not marked in any orthography. In Wedekind, Wedekind, and Musa (2006 and 2007), stressed syllables are indicated in boldface.[22]

In addition to these two systems and several academic systems of transcribing Beja texts, it is possible that Beja was at least occasionally written in the Greek alphabet-based Coptic script during the Middle Ages.[23]


Nouns, articles, and adjectives

Beja nouns and adjectives have two genders: masculine and feminine, two numbers: singular and plural, two cases: nominative and oblique, and may be definite, indefinite, or in construct state.[24] Gender, case, and definiteness are not marked on the noun itself, but on clitics and affixes. Singular-plural pairs in Beja are unpredictable.

Plural forms

Plurals may be formed by:

A small number of nouns do not distinguish between singular and plural forms. Some nouns are always plural. A few nouns have suppletive plurals.[26]

Case and definiteness

A noun may be prefixed by a clitic definite article, or have an indefinite suffix. Definite articles indicate gender, number, and case. The indefinite suffix marks gender only, and does not appear in the nominative case.[27] For feminine common nouns, the indefinite suffix is -t; for masculine nouns and feminine proper nouns, -b. The indefinite suffixes only appear after vowels. The definite article is proclitic. It has the following forms with masculine monosyllabic nouns that do not begin with /h/ or /ʔ/ (note that an initial glottal stop is usually omitted in writing, and that all words that appear to be vowel-initial actually begin with a glottal stop):

  Nominative Oblique
Singular uu- oo-
Plural aa- ee-

The feminine definite articles begin with t but are otherwise identical (tuu-, too-, taa-, tee-). With nouns longer than one syllable and with nouns that begin with /h/ or /ʔ/, reduced forms of the definite article are used which do not distinguish between cases, but maintain gender distinctions. In some dialects (e.g. that described by Wedekind, Wedekind, and Musa for Port Sudan) the reduced forms maintain number distinctions; in others (e.g. that described by Vanhove and Roper for Sinkat) they do not.


Possessive relationships are shown through a genitive suffix -ii (singular possessed) or -ee (plural possessed) which attaches to the possessing noun. If the possessing noun is feminine, the genitive marker will begin with t; if the possessed is feminine, the suffix will end with t. When the suffix does not end with the feminine marker t, it reduces to -(t)i, whether singular or plural (that is, the singular/plural distinction is only marked for feminine possessa). Because this suffix adds a syllable to the noun, full forms of articles cannot be used; thus, the article on the noun itself does not indicate case. However, agreeing adjectives will be marked for oblique case. No article or indefinite suffix may be applied to the possessed noun. The possessed noun follows the possessor. Examples:

(The noun tak 'man' has the suppletive plural (n)da 'men'; raaw 'friend' has the shortened plural raw 'friends'.)

Postpositions follow nouns in the genitive. Examples:[28]


Adjectives follow the nominal heads of noun phrases. They agree in gender, number, case, and definiteness, and carry case and definiteness markers of the same form as nouns.


Clauses may be composed of two noun phrases or a noun phrase and a predicative adjective followed by a copular clitic. The copula agrees in person, gender, and number with the copula complement (the second term), but the first- and third-person forms are identical. The copular subject will be in the nominative case, the copular complement in the oblique. Oblique -b becomes -w before -wa. Copular complements that end in a vowel will employ an epenthetic y between the final vowel and any vowel-initial copular clitic.

Masculine Feminine
singular -u -tu
plural -a -ta
singular -wa -tuwi
plural -aana -taana
singular -u -tu
plural -a -ta



Beja verbs have two different types, first noted by Almkvist: "strong verbs," which conjugate with both prefixes and suffixes and have several principal parts; and "weak verbs," which conjugate with suffixes only and which have a fixed root. Verbs conjugate for a number of tense, aspect, modality, and polarity variations, which have been given different names by different linguists:

Roper Wedekind, Wedekind, and Musa Vanhove
Conditional Past Continuous Aorist
Past Indicative Past Perfective
Present Indicative Present Imperfective
Future Indicative Future Future
Intentional Desiderative
Imperative Imperative Imperative
Optative Jussive Optative

(Roper analyzes additional subjunctive forms where Wedekind, Wedekind, and Musa, and Vanhove see a conditional particle.)

Each of the above forms has a corresponding negative. (Vanhove refers to the imperative negative as the "prohibitive".) The past continuous and past share a past negative. Negative forms are not derived from corresponding positive forms, but are independent conjugations.

Every verb has a corresponding deverbal noun, which Wedekind, Wedekind, and Musa refer to as a "noun of action", Vanhove calls an "action noun", and Roper a "nomen actionis". Numerous serial verb constructions exist which connote different aspectual and potential meanings.[31]


The third person masculine singular positive imperative is the citation form of the verb. Weak verbs have a long final suffix -aa while strong verbs have a short final suffix -a. For both weak and strong verbs, the negative imperative is formed by an identical set of prefixes baa- (for masculine singular and common plural) and bii- (for feminine singular). Strong verbs use a negative imperative root which has a lengthened vowel.[32]

Weak Verb: giigaa "go away"
  Positive Negative
2sg.m giig-aa baa-giig-aa
2sg.f giig-ii bii-giig-ii
2pl giig-aana baa-giig-aana
Strong Verb: fidiga "open"
  Positive Negative
2sg.m fidig-a baa-fidiig-a
2sg.f fidig-i bii-fidiig-i
2pl fidig-na baa-fidiig-na

Deverbal and dependent forms

Deverbal noun

Every Beja verb has a corresponding deverbal noun (Wedekind, Wedekind, and Musa: "noun of action"; Vanhove: "action noun"; Roper: "nomen actionis"). For weak verbs, the deverbal noun is formed by a suffix -ti attached to the imperative root (see above). For strong verbs, deverbal nouns are not entirely predictable.


There are patterns in strong verb deverbal nouns related to the structure of the citation form of the verb. However, these are not consistent.[33]

Deverbal adjective

A further derived form is a suffix -aa attached to the citation root, and then followed by -b for masculine nouns and -t for feminine. Examples:

Citation Form Masculine Feminine Meaning
diwaa diw-aab diw-aat asleep
afooyaa afooy-aab afooy-aat forgiven
adhidha adhidh-aab adhidh-aat hobbled
nikwiyi nikwiy-aat pregnant

This form may be used as an adjective, but it is also employed in the construction of multiple conjugated negative forms. Wedekind, Wedekind, and Musa analyse this form as a participle.[34] Martine Vanhove analyses it as a manner converb -a.[35]

Tense-conjugated forms

Past continuous/aorist

The past continuous stem for strong verbs is not derivable from any other verb stem. The negative of the past continuous is identical to that of the past: There is only one past tense negative form. For both weak and strong verbs, the past negative is formed through a deverbal participial or converbal form (see above) followed by the present negative of the irregular verb aka "to be".[36]

Wedekind, Wedekind, and Musa describe the past continuous as being used for "habitual, repeated actions of the (more distant) past."[37] It is the verb conjugation used for counterfactual conditionals, which leads to Roper's identifying this tense as the "conditional". It is also frequently used in narratives.[38]

Weak Verb: giigaa "go away"
  Positive Negative
1sg giig-i giig-aab kaaki
2sg.m giig-tiya giig-aab kittaa
2sg.f giig-tiyi giig-aat kittaayi
3sg.m giig-i giig-aab kiiki
3sg.f giig-ti giig-aat kitti
1pl giig-ni giig-aab kinki
2pl giig-tiina giig-aab kitteena
3pl giig-iin giig-aab kiikeen
Strong Verb: fidiga "open"
  Positive Negative
1sg iifdig fidg-aab kaaki
2sg.m t-iifdig-a fidg-aab kittaa
2sg.f t-iifdig-i fidg-aat kittaayi
3sg.m iifdig fidg-aab kiiki
3sg.f t-iifdig fidg-aab kitti
1pl n-iifdig fidg-aab kinki
2pl t-iifdig-na fidg-aab kitteena
3pl iifdig-na fidg-aab kiikeen

The past or perfective stem for strong verbs is identical to the citation form (imperative) stem, with predictable phonetic modifications.[39] The negative is identical to that of the past continuous/aorist (above).

Weak verb: giigaa "go away"
  Positive Negative
1sg giig-an giig-aab kaaki
2sg.m giig-taa giig-aab kittaa
2sg.f giig-taayi giig-aat kittaayi
3sg.m giig-iya giig-aab kiiki
3sg.f giig-ta giig-aat kitti
1pl giig-na giig-aab kinki
2pl giig-taana giig-aab kitteena
3pl giig-iyaan giig-aab kiikeen
Strong verb: fidiga "open"
  Positive Negative
1sg a-fdig fidg-aab kaaki
2sg.m ti-fdig-a fidg-aab kittaa
2sg.f ti-fdig-i fidg-aat kittaayi
3sg.m i-fdig fidg-aab kiiki
3sg.f ti-fdig fidg-aab kitti
1pl ni-fdig fidg-aab kinki
2pl ti-fdig-na fidg-aab kitteena
3pl i-fdig-na fidg-aab kiikeen

The present or imperfective has two stems for positive strong verbs, while the negative strong stem is identical to that used for the imperative (and thus also for past/perfective verbs).[39] Weak negative verbs add the prefix ka- to positive past/perfective forms.

Weak verb: giigaa "go away"
  Positive Negative
1sg giig-ani ka-giig-an
2sg.m giig-tiniya ka-giig-taa
2sg.f giig-tinii ka-giig-taayi
3sg.m giig-inii ka-giig-iya
3sg.f giig-tini ka-giig-ta
1pl giig-nay ka-giig-na
2pl giig-teena ka-giig-taana
3pl giig-een ka-giig-iyaan
Strong verb: fidiga "open"
  Positive Negative
1sg a-fandiig kaa-fdig
2sg.m fandiig-a kit-fidig-a
2sg.f fandiig-i kit-fidig-i
3sg.m fandiig kii-fdig
3sg.f fandiig kit-fidig
1pl ni-fadig kin-fdig
2pl ti-fadig-na kit-fidig-na
3pl i-fadig-na kii-fdig-na

The strong future stem is described differently by Wedekind, Wedekind, and Musa and by Vanhove. Both agree that it is a fixed stem followed by a present/imperfective conjugated form of the verb diya "to say." Wedekind, Wedekind, and Musa's strong stem is similar to the past continuous/aorist stem (next section), and identical for all numbers, genders, and persons, except the first person plural, which has a prefixed n-. For Vanhove, there are distinct singular and plural stems which are identical to the past continuous/aorist first person singular and plural, respectively.[40] Similarly, for weak verbs, Wedekind, Wedekind, and Musa have a future stem ending in -i with a first person plural -ni, followed by a present tense/imperfective conjugation of diya. Vanhove sees the -i as a singular future, and the -ni as a general plural. For negative verbs, the negative present/imperfective of diya is used as the conjugated auxiliary.

Weak Verb: giigaa "go away"
  Positive Negative
1sg giig-i andi giig-i kaadi
2sg.m giig-i tindiya giig-i kiddiya
2sg.f giig-i tindii giig-i kiddiyi
3sg.m giig-i indi giig-i kiidi
3sg.f giig-i tindi giig-i kiddi
1pl giig-ni neeyad giig-ni kindi
2pl giig-i teeyadna (Vanhove: giig-ni teeyadna) giig-i kiddiina (Vanhove: giig-ni kiddiina)
3pl giig-i eeyadna (Vanhove: giig-ni eeyadna) giig-i kiidiin (Vanhove: giig-ni kiidiin)
Strong Verb: fidiga "open"
  Positive Negative
1sg iifdig andi iifdig kaadi
2sg.m iifdig tindiya iifdig kiddiya
2sg.f iifdig tindii iifdig kiddiyi
3sg.m iifdig indi iifdig kiidi
3sg.f iifdig tindi iifdig kiddi
1pl niifdig neeyad niifdig kindi
2pl iifdig teeyadna (Vanhove: niifdig teeyadna) iifdig kiddiina (Vanhove: niifdig kiddiina)
3pl iifdig eeyadna (Vanhove: niifdig eeyadna) iifdig kiidiin (Vanhove: niifdig kiidiin)

(NB: Wedekind, Wedekind, and Musa see verbs of the form CiCiC as having identical past continuous [aorist] and future stems. Some verbs of other forms have different stems, which would lead to a greater divergence between the forms described by them and those described by Vanhove.) E.M. Roper, describing the same dialect as Vanhove, identifies the stem employed as being identical to the past continuous/aorist (for him, "conditional"—see above), just as Vanhove does. However, he understands the form with n- as being used only with the first person plural, as Wedekind, Wedekind, and Musa do.[41]


In addition to the future, Bidhaawyeet has a similar form expressing desire to undertake an act or intention to do so. The citation root takes a suffix -a for all persons, genders, and numbers, and is followed by a present tense/imperfective conjugated form of the verb diya "to say", as the future is.[42]

Jussive, optative, potential

There is distinct disagreement between the major grammars of the past century on the modal conjugation or conjugations referred to as "jussive," "optative," and "potential."

Wedekind, Wedekind, and Musa describe a "jussive" with the following paradigm. For strong verbs, the first person is based on the past/perfective stem, and the persons are based on the future stem; no negative jussive is given:

Weak verb: giigaa "go away"
1sg giig-atay
2sg.m giig-ata
2sg.f giig-ati
3sg.m ba-giig-iiyay
3sg.f ba-giig-tiyay
1pl giig-niiyay
2pl giig-aana
3pl ba-giig-iinay
Strong verb: fidiga "open"
1sg fidg-atay
2sg.m fidg-ata
2sg.f fidg-ati
3sg.m ba'-iifidig-ay
3sg.f bat-iifidig-ay
1pl niifdig-ay
3pl ba-'iifdig-naay

They give various examples of the jussive with translations into English, in order to give a sense of the meaning:

Vanhove identifies a complex "potential" form composed of a nominalizing suffix -at followed by a present/imperfective reduced conjugation of the verb m'a 'come' (eeya in the non-reduced present/imperfective).

Weak verb: giigaa "go away"
1sg giig-at
2sg.m giig-at-a
2sg.f giig-at-i
3sg.m giig-at eeyini
3sg.f giig-at eetnii
1pl giig-at eenay
2pl giig-at-na
3pl giig-at een
Strong verb: fidiga "open"
1sg fidg-at
2sg.m fidg-at-a
2sg.f fidg-at-i
3sg.m fidg-at eeyini
3sg.f fidg-at eetnii
1pl fidg-at eenay
2pl fidg-at-na
3pl fidg-at een

Vanhove describes the potential as expressing "epistemic modalities of inference or near-certainty."[44] Examples below, with the potential verbs in bold:

Additionally, she recognizes an optative with positive and negative polarity. The positive optative is formed from a prefix baa- to the past continuous/aorist. The negative construction is more complex. In some dialects, the final -aay of most forms of the weak negative is a short -ay:

Weak verb: giigaa "go away"
  Positive Negative
1sg baa-giig-i baa-giig-aay
2sg.m baa-giig-tiya bit-giig-aay
2sg.f baa-giig-tiyi bit-giig-aay
3sg.m baa-giig-i bii-giig-aay
3sg.f baa-giig-ti bit-giig-aay
1pl baa-giig-ni bin-giig-aay
2pl baa-giig-tiina bit-giig-eena
3pl baa-giig-iin bii-giig-eena
Strong verb: fidiga "open"
  Positive Negative
1sg baa-'iifdig baa-fdig
2sg.m baa-t-iifdig-a bit-fdig-a
2sg.f baa-t-iifdig-i bit-fdig-i
3sg.m baa-'iifdig bii-fdig
3sg.f baa-t-iifdig bit-fdig
1pl baa-n-iifdig bin-fdig
2pl baa-t-iifdig-na bit-fdig-na
3pl baa-'iifdig-na bii-fdig-na

Vanhove gives no explanation for the use of the optative positive. The optative negative is used in conditional clauses with meanings of incapacity and necessity:


Through lexicostatistical analysis, David Cohen (1988) observed that Beja shared a basic vocabulary of around 20% with the East Cushitic Afar and Somali languages and the Central Cushitic Agaw languages, which are among its most geographically near Afroasiatic languages. This was analogous to the percentage of common lexical terms that was calculated for certain other Cushitic languages, such as Afar and Oromo. Václav Blažek (1997) conducted a more comprehensive glottochronological examination of languages and data. He identified a markedly close ratio of 40% cognates between Beja and Proto-East Cushitic as well as a cognate percentage of approximately 20% between Beja and Central Cushitic, similar to that found by Cohen.[46]

A fairly large portion of Beja vocabulary is borrowed from Arabic. In Eritrea and Sudan, some terms are instead Tigre loanwords.[47] Andrzej Zaborski has noted close parallels between Beja and Egyptian vocabulary.[48]

The only independent Beja dictionary yet printed is Leo Reinisch's 1895 Wörterbuch der Beḍauye-Sprache. An extensive vocabulary forms an appendix to E.M. Roper's 1928 Tu Beḍawiɛ: An Elementary Handbook for the use of Sudan Government Officials, and this has formed the basis for much recent comparative Cushitic work. Klaus and Charlotte Wedekind and Abuzeinab Musa's 2007 A Learner's Grammar of Beja (East Sudan) comes with a CD which contains a roughly 7,000-word lexicon, composed mostly of one-word glosses. Klaus Wedekind, Abuzeinab Muhammed, Feki Mahamed, and Mohamed Talib were working on a Beja-Arabic-English dictionary, but publication appears to have been stalled by Wedekind's death.[49] Martine Vanhove announced a forthcoming Beja-Arabic-English-French dictionary in 2006. It has not yet been published.[50] The Beja scholar Muhammed Adarob Ohaj produced a Beja-Arabic dictionary as his masters thesis in 1972. It has not yet been published.[51]

Swadesh List

The following list is drawn from Wedekind, Wedekind, and Musa's 2007 grammar and Roper's 1928 handbook. Nouns are given in indefinite accusative forms (the citation form); unless marked otherwise, forms that end in ⟨t⟩ are feminine and all others are masculine. Verbs are given in the singular masculine imperative.

No. English Beja
1 I aneeb (acc), ani (nom)
2 you (singular) barook (m.acc), batook (f.acc), baruuk (m.nom), batuuk (f.nom)
3 we hinin
4 this oon- (m.acc), toon- (f.acc), uun- (m.nom), tuun- (f.nom)
5 that been
6 who aab (acc), aaw (nom)
7 what naa
8 not [1]
9 all kass-
10 many gwidaab
11 one gaal (m), gaat (f)
12 two maloob
13 big win
14 long saraaraab
15 small dibiloob
16 woman takat
17 man (adult male) tak
18 man (human being) m'aadamiib
19 bird kilaay (s), kilay (p)
20 dog yaas (s), yas (p)
21 louse taat (s), tat (p)
22 tree hindiib (s/p)
23 seed teeraab (s/p)
24 leaf raat (s), rat (p)
25 root gadam (s), gadamaab (p)
26 bark (of a tree) adhift (s), adhifaat (p); hindi shadhiidh
27 skin adeeb
28 meat shaat
29 blood booy
30 bone miitaatt (s), miitatt (p)
31 fat (noun) l'aab; dhhaab
32 fire n'eet
33 egg kwhiib
34 horn d'aab
35 tail niiwaat
36 feather tambaat/timbaat
37 fish ashoob
38 hair hamoot
39 head girmaab
40 ear angwiil (s), angwil (p)
41 eye liiliit[2]; gw'aj
42 fingernail n'af
43 nose ginuuf (s), ginif (p)
44 mouth yaf (s), yafaab (p)
45 tooth kwiriib
46 tongue (organ) miid(al)aab (s), miid(al)ab (p)
47 foot ragad (s), ragadaab (p)[3]
48 knee gindhif/gindif (s), gindhifaab/gindifaab (p)
49 hand ayeeb
50 belly f'iib
51 neck alaat
52 breast nigw
53 heart (organ) gin'aab
54 liver seet
55 to drink gw'a
56 to eat tamaa
57 to bite finika
58 to see rhaa
59 to hear maasiwa
60 to know kana
61 to sleep diwaa
62 to die yaya
63 to kill dira
64 to swim nidabaa
65 to fly firaa
66 to walk hiireeraa
67 to come m'aa[4]
68 to lie (as in a bed) b'iya
69 to sit sa'a; thathaa
70 to stand gada
71 to give hiya
72 to say diya
73 sun yint
74 moon tirigt
75 star hayiikw
76 water yam
77 rain bireeb
78 stone aweeb
79 sand isseet
80 earth burt; haash
81 cloud baal (f.s), baalaat (p); sahaab
82 smoke eegaab
83 ashes n'eetehaash
84 to burn liwa (tr), lawa (intr)
85 road darab
86 mountain ribaab
87 red adaroob
88 green sootaay[5]
89 yellow
90 white eeraab
91 black hadal
92 night hawaad (s), hawad (p)
93 warm naba'
94 cold m'akwaraab; l'aab
95 full ataab
96 new gayiib
97 good daayiib
98 round kwadhaadh; kwalaal
99 dry balamaab
100 name sim
  1. ^ Beja handles negation through distinct negative polarity conjugation. There is no lexical "not."
  2. ^ In some dialects liiliit means "pupil."
  3. ^ Ragad refers to the foot and leg.
  4. ^ This is a rare suppletive imperative. Other forms of the verb have no /m/ and are constructed around a consonantal root /j/.
  5. ^ Sootaay covers the blue-green range.


1 2 3 4 5 6 7 8 9 10
m gaal maloob mhay fadhig ay asagwir asaramaab asumhay ashshadhig tamin
f gaat maloot mhayt fadhigt ayt asagwitt asaramaat asumhayt ashshadhigt tamint

"Ten" has combining forms for the production of teens and products of ten. Numbers from 11 to 19 are formed by tamna- followed by the units. E.g., "fourteen" is tamna fadhig. Combining ones use the form -gwir; e.g., "eleven" is tamnagwir. "Twenty" is tagwuugw. "Twenty-one" is tagwgwagwir. "Thirty" is mhay tamun; "forty" is fadhig tamun; "fifty" is ay tamun; etc. "One hundred" is sheeb. For higher numbers, Beja-speakers use Arabic terms.

Ordinal numbers are formed by the addition of a suffix -a. "First" is awwal, borrowed from Arabic.

"Half" is tarab. Other fractions are borrowed from Arabic.[52]


Beja has an extensive oral tradition, including multiple poetic genres. A well-known epic is the story of the hero Mhamuud Oofaash, portions of which have appeared in various publications by Klaus Wedekind.[53] An edition appears in Mahmud Mohammed Ahmed's Oomraay, published in Asmara.[54] In the 1960s and '70s, the Beja intellectual Muhammed Adarob Ohaj collected oral recordings of poetic and narrative material which are in the University of Khartoum Institute of African and Asian Studies Sound Archives. Didier Morin and Mohamed-Tahir Hamid Ahmed have used these, in addition to their own collections, for multiple academic publications in French on Beja poetics[citation needed].[51] Red Sea University and the NGO Uhaashoon worked with oral story-tellers to produce a collection of 41 short readers and a longer collection of three short stories in Beja between 2010 and 2013.[55]


  1. ^ Beja at Ethnologue (25th ed., 2022) Closed access icon
  2. ^ "Beja". Ethnologue. Retrieved 11 September 2023.
  3. ^ Wedekind, Wedekind & Musa 2004–2005, p. 10; Vanhove 2017, p. 6.
  4. ^ Zaborski 1988, p. 491.
  5. ^ Ehret 2008, p. 159.
  6. ^ Lamberti 1991, p. 553.
  7. ^ Hetzron 1980.
  8. ^ Vanhove 2020a, p. 301; Tosco 2000.
  9. ^ Ehret 2008, pp. 153–155.
  10. ^ Ehret 2008, p. 155.
  11. ^ Satzinger 2014.
  12. ^ Browne 2003; Wedekind 2010, Wedekind identified this psalm as number 29, but made it clear that he was referring to the Septuagint. Different traditions number the psalms differently and the psalm in question is more commonly numbered 30 today, as here.
  13. ^ Wedekind, Wedekind & Musa 2004–2005, §26; Vanhove 2014, §2.1.
  14. ^ Wedekind, Wedekind & Musa 2004–2005, §§31–32.
  15. ^ Wedekind, Wedekind & Musa 2004–2005, §§26, 30.
  16. ^ Wedekind, Wedekind & Musa 2004–2005, §27.
  17. ^ Wedekind, Wedekind & Musa 2004–2005, §25.
  18. ^ Wedekind, Wedekind & Musa 2007, p. 11.
  19. ^ a b Wedekind, Wedekind & Musa 2007.
  20. ^ محمد أدروب محمد.
  21. ^ مهمود واحمد ابوبكر أوريب 2019.
  22. ^ Wedekind, Wedekind & Musa 2004–2005; Wedekind, Wedekind & Musa 2007.
  23. ^ Browne 2003.
  24. ^ Appleyard 2007.
  25. ^ Vanhove 2014, §3.1.5.
  26. ^ Wedekind, Wedekind & Musa 2007, p. 43.
  27. ^ Roper 1928, §43.
  28. ^ Roper 1928, §§56–57.
  29. ^ Wedekind, Wedekind & Musa 2007, p. 45; Vanhove 2014, §3.1.1.
  30. ^ Roper 1928, §63.
  31. ^ Wedekind, Wedekind & Musa 2007, pp. 150–158; Vanhove 2014, §3.2.6ff.
  32. ^ Wedekind, Wedekind & Musa 2007, pp. 97–100; Vanhove 2014, §
  33. ^ Wedekind, Wedekind & Musa 2007, pp. 159–161; Vanhove 2014, §§3.3.1–3.3.2.
  34. ^ Wedekind, Wedekind & Musa 2007, pp. 56–57.
  35. ^ Vanhove 2014, §
  36. ^ Wedekind, Wedekind & Musa 2007, p. 101.
  37. ^ Wedekind, Wedekind & Musa 2007, p. 155.
  38. ^ Wedekind, Wedekind & Musa 2007, p. 183; Vanhove 2014, §
  39. ^ a b Wedekind, Wedekind & Musa 2007, p. 100.
  40. ^ Wedekind, Wedekind & Musa 2007, p. 101; Vanhove 2014, §
  41. ^ Roper 1928, §183.
  42. ^ Wedekind, Wedekind & Musa 2007, pp. 154–155; Vanhove 2014, §
  43. ^ Wedekind, Wedekind & Musa 2007, pp. 162–166.
  44. ^ Vanhove 2014, §
  45. ^ Vanhove 2014, §; Vanhove 2011.
  46. ^ Vanhove 2020a, p. 302.
  47. ^ Wedekind, Wedekind & Musa 2007, p. 17.
  48. ^ Zaborski 1988.
  49. ^ Wedekind et al. 2017.
  50. ^ Vanhove 2006, p. 6.
  51. ^ a b Hamid Ahmed 2010.
  52. ^ Wedekind, Wedekind & Musa 2004–2005, §§250–257.
  53. ^ Wedekind 2006; Wedekind & Mohammed 2008.
  54. ^ Mohammed Ahmed 2004.
  55. ^ Swackhamer 2013.


  • Almkvist, Herman (1881). Die Bischari-Sprache Tū-Beḍāwie in Nordost-Afrika. Upsala: Akademisch Buchdruckerei.
  • Appleyard, David (2007). "Beja Morphology". In Kaye, Alan S. (ed.). Morphologies of Asia and Africa. Winona Lake, Ind.: Eisenbrauns. pp. 447–479. ISBN 978-1-57506-109-2.
  • Browne, Gerald (2003). Textus blemmyicus in aetatis christianae. Champaign, Illinois: Stipes Publish L.L.C. ISBN 978-1-58874-275-9.
  • Ehret, Christopher (2008). "The primary branches of Cushitic: Seriating the diagnostic sound change rules". In Bengston, John (ed.). In Hot Pursuit of Language in Prehistory: Essays in the four fields of anthroplogy. Amsterdam: John Benjamins Publishing Company. pp. 149–160. ISBN 9789027232526.
  • Hamid Ahmed, Mohamed-Tahir (2010). "L'Espace et la poésie bedja (Soudan)". Journal des Africanistes. 79 (2): 171–187. Retrieved 11 November 2017.
  • Hetzron, Robert (1980). "The Limits of Cushitic". SUGIA. 2: 7–125.
  • Lamberti, Marcello (1991). "Cushitic and Its Classifications". Anthropos. 86 (4/6): 552–561.
  • أوريب, مهمود واحمد ابوبكر (2019). تِگْرايَاوَ إِسَوادِيتْ كَنَانْ إِكُوَرانِي هَرْفيِ. Port Sudan: Red Sea University.
  • محمد, محمد أدروب. قاموس بِذاوِيّت/عربي. Port Sudan.
  • Mohammed Ahmed, Mahmud (2004). Oomraay. Asmara: Semhar Press.
  • Reinisch, Leo (1893). Die Beḍauye-Sprache in Nordost-Afrika. Wien: Kaiserlich Akademie der Wissenschaften.
  • Roper, E.M. (1927). "The Poetry of the Haḍenḍiwa". Sudan Notes and Records. 10: 147–158.
  • Roper, E.M. (1928). Tu Beḍawiɛ: An Elementary Handbook for the Use of Sudan Government Officials. Hertford: Stephen Austin and Sons, Ltd.
  • Satzinger, Helmut (2014). "The Barbarian Names on the Ostraca from the Eastern Desert (3rd Century CE)". Inside and Out. Interactions between Rome and the Peoples on the Arabian and Egyptian Frontiers in Late Antiquity. LAHR. pp. 199–212. ISBN 978-90-429-3124-4.
  • Swackhamer, Jeanette. "Dabaab". Bidhaawyeetiit Kitaabaati Kitab. Retrieved 11 November 2017.
  • Swackhamer, Jeanette (2013). "Ooktaab Oowin". Bidhaawyeetiit Kitaabaati Kitab. Retrieved 11 November 2017.
  • Tosco, Mauro (2000). "Cushitic Overview". Journal of Ethiopian Studies. 33 (2): 87–121.
  • Vanhove, Martine (2011). "Towards a semantic map of the Optative in Beja (North-Cushitic)". In Busetto, Luca; Sottile, Roberto; Tonelli, Livia; Tosco, Mauro (eds.). He bitaney lagge. Dedicato a / Dedicated to Marcello Lamberti. Saggi di Linguistica e Africanistica. Essays in Linguistics and African Studies. pp. 231–246.
  • Vanhove, Martine (2006). "The Beja Language Today in Sudan: The State of the Art in Linguistics". Proceedings of the 7th International Sudan Studies Conference April 6 – 8 2006. University of Bergen.
  • Vanhove, Martine (2014). "Beja Grammatical Sketch". doi:10.1075/ ((cite journal)): Cite journal requires |journal= (help)
  • Vanhove, Martine (2017). Le Bedja. Leuven: Peeters. ISBN 978-9042935167.
  • Vanhove, Martine (2020a). "North-Cushitic". In Rainer Vossen; Gerrit J. Dimmendaal (eds.). The Oxford Handbook of African Languages (PDF). Oxford University Press. pp. 300–307. doi:10.1093/oxfordhb/9780199609895.013.63. ISBN 9780199609895.
  • Vanhove, Martine (2020b). "Beja". In Christopher Lucas; Stefano Manfredi (eds.). Arabic and contact-induced change. Berlin: Language Science Press. pp. 419–439. ISBN 978-3-96110-251-8.
  • Wedekind, Klaus; Mohammed, Mahmud (2008). "A Beja saga in four dialects: Lexical and other differences". In Takács, Gábor (ed.). Semito-Hamitic Festschrift for A. B. Dogopolsky and H. Jungraithmayr. Berlin: Reimer. pp. 266–378.
  • Wedekind, Klaus; Wedekind, Charlotte; Musa, Abuzeinab (2004–2005). Beja Pedagogical Grammar (PDF). Aswan and Asmara.((cite book)): CS1 maint: location missing publisher (link)
  • Wedekind, Klaus; Wedekind, Charlotte; Musa, Abuzeinab (2007). A Learner's Grammar of Beja (East Sudan). Cologne: Rüdiger Köppe Verlag.
  • Wedekind, Klaus (2006). "Beja Narratives: Pursuit of Participants and Analysis of Aspects". Ethnorema. 2: 101–128. ISSN 1826-8803.
  • Wedekind, Klaus (2010). "More on the Ostracon of Browne's Textus Blemmyicus". Annali, Sezione Orientale. 70: 73–81. ISSN 0393-3180.
  • Wedekind, Klaus; Musa, Musa; Mahamed, Feki; Talib, Mohamed (2017). "Forthcoming - Klaus Wedekind, with Abuzeinab Musa, Feki Mahamed, Mohamed Talib, et. al., "Trilingual dictionary of Beja" [arranged by the provenience of entries: S. Egypt, E. Sudan, N. Eritrea]". Retrieved 11 November 2017.
  • Zaborski, Andrzej (1988). "Remarks on the Verb in Beja". In Arbeitman, Yoël (ed.). Fucus: A Semitic/Afrasian Gathering in Remembrance of Albert Ehrman. John Benjamins Publishing. pp. 491–498. ISBN 978-9027235527.
  • Zaborski, Andrzej (1997). "Problems of the Beja Present Seven Years Ago". Lingua Posnaniensis. 39: 145–153.