<22>

The following section needs to be studied first, if a correct pronunciation of the Beja examples is desired.

<23>

The transcription which is used throughout this book is systematically phonological, and very close to the orthography of Beja which was established in Eritrea. The main difference between the Beja orthography and the transcription used in this book is the use of bold vowel letters. They indicate the pitch-accent.

<24>

There also is an Arabic transliteration of the Beja data. Among the various Arabic transliterations which were in use, this particular one has been promoted in the internet for several years. It distinguishes uu and ii from oo and ee, but its disadvantage is that it uses diacritics to do so. Recently, however, its main website (sakanab) stopped supporting it.

<25>

Since there still are Beja individuals who occasionally use Arabic letters for their language (cf. the “bejaculture” website), the Arabic transliteration - although defunct - has not been deleted from this book.

<26>

The following table is an alphabetical list of the orthographic symbols and their pronunciation (i.e. the IPA values). Some of the sound files are arranged in a different order, such as the files for dh, gh, and kh.

<31>

It will be noticed that some words - including common greetings - are borrowed from Arabic. However, sounds which do not fit the Beja sound system will generally be rejected, yet, depending on the speaker, certain Arabic sounds will be preserved or integrated as table 2 shows.

<32>

As far as loan words are concerned, they will be re-shaped to be acceptable to the Beja language. The shape of loan words is predictable from the typical Beja patterns which will be presented in table 3 further below.

<33>

As far as the places of consonants (C) and vowels (V) in syllables are concerned, Beja words allow for exactly 18 syllable patterns in the word initial syllable, and 12 patterns in all other syllables.

<34>

The most frequent pattern is CV, i.e. a syllable which consists of a consonant followed by a short vowel. But syllables can also be closed with one or two consonants, where the second consonant can only be t. So far this gives 3 patterns, namely: CV as in na 'thing', CVC as in tak 'man', and CVCt as in lamt 'soup'.

<35>

Less frequent are syllables with long vowels [1], which will be written as VV. This gives 3 additional syllables patterns: CVV as in tuu 'the', CVVC as in been 'that', and CVVCt as in diint 'thorn'.

<36>

Only word initial syllables can start with a vowel (or more exactly, with a suppressed hamzah), which gives the following additional patterns: V as in a.ni 'I (1.Sg)', VC as in ar 'boys', VCt as in awt 'honey', VV as in aa.bu 'who (M)', VVC as in aab 'who', and VVCt as in aabt 'who (F)'.

<37>

As a specialty of the Beja phonology there also are syllables in which the initial consonant is followed either by h or by hamzah, abbreviated as H  [2]: CHV as in t'a 'now', CHVV as in m'aa 'come!', CHVCt as in mhay 'three' and as in lhayt 'tomorrow', and CHVVCt as in n'aayt 'goat'. This last set of syllable patterns actually causes one of the systematic differences between northern and southern Beja dialects: What is pronounced as C'V in the North may be pronounced as CV' or as two syllables CV.'V in the South, e.g. s'a, sa', sa.'a 'sit! (M)'.

<40>

The Atman variety of Beja - which serves as the basis for this book - is one of the major Beja dialects. It is spoken by large numbers of people in parts of Port Sudan and in Suakin. Other Beja dialects differ from it in systematic ways, and the differences are limited to certain sounds, certain conditions, and certain groups of words. To the Beja speakers themselves these differences do not present any communication problems - but they may cause some surprises to outsiders. The ways in which the other dialects may differ from the dialect of this book are the following:

<41>

These differences may cause some surprises at first. But since in a particular dialect the same difference will appear again and again, it will be helpful to keep these correspondences in mind. Table 4 below summarizes the main types of correspondences.

<45>

Briefly, the main rules concerning pitch-accent (accent for short) are as follows:

<46>

The word gwhara 'thief', which has the inherent accent on the first a, may illustrate how these rules apply:

<47>

In this section about the addition and deletion of sounds, it is possible to argue in two ways: Either (1) the starting point is a dictionary where certain forms have no vowels, such as t- s- dwl 'you-cause.to-approach', or (2) the starting point is a dictionary with vowels, such as ti- si- dawil. In the first case, there must be rules which add certain vowels in the appropriate places, and in the second case, there must be rules which delete certain vowels in the appropriate places. The result should be the same. The rules here below allow for both perspectives.

<77>

In this presentation of Beja, 11 word classes will be recognized. The terms noun, verb etc. will be used in their traditional sense, and they will be defined more closely in the tables and examples below.

<78>

The 11 word classes are the following:

<79>

The Beja word classes differ from each other by the inflections which they take or which they pass on to other elements in the clause. The inflection takes the form of different kinds of affixes, such as yakan 'I started', yaktaa 'you (M) started', yaktaayi 'you (F) started', etc.

<80>

Verbs take affixes for tense-aspect and mood, as well as person, gender and number. Each verb governs one, two or three noun phrases such as subject, first object or second object, to which it assigns the respective cases - (either 'subject' case or 'object' case).

<81>

Nouns, pronouns, and names are inflected for the cases which they take from the verb. They are also inflected for number (Singular or plural), and they inherently have a certain gender (Masculine or feminine) which they pass on to other words in the phrase. They take demonstrative prefixes and article prefixes, as well as possessive suffixes.

<82>

Adjectives, demonstratives, and numbers are inflected for gender, number, and case, which they take from the head noun of the phrase.

<83>

Adverbs and postpositions only take case and possessive suffixes.

<84>

Conjunctions and interjections are not inflected at all. They do not take any prefixes or suffixes.

<85>

The table (below) summarizes the different word classes and their inflections.

<104>

The affixes of nouns and verbs are different, and each affix has its proper place. Theoretically, verbs or nouns would allow for as many as 7 or 8 affixes following each other - but normally there will just be 1 or 2.

<108>

The language has a number of derivational mechanisms, as the next table shows. Verbs can be derived into Nouns, adjectives into nouns, etc. The mechanisms include both affixation and change of patterns.

<109>

Affixation, for instance, derives tam 'eat' into > tam-ti 'eating'.

<110>

Change of CV patterns, for instance, derives winin (CiCiC) 'be angry' into > wnuun 'anger' (CCuuC).

<111>

The table gives an overview of frequent derivation patterns. The original word class is given in the first column, the derived word class in the second. These derivational affixes, however, cannot be used by non-Beja speakers to create new words at will. Therefore, the table (below) only explains derivations - it is not a pattern for producing them.

<114>

To begin with, here are some Beja vocatives and interjections. There is no need to analyze all words at this point. The analysis of these forms will be introduced step by step.

<22>

The following section needs to be studied first, if a correct pronunciation of the Beja examples is desired.

<23>

The transcription which is used throughout this book is systematically phonological, and very close to the orthography of Beja which was established in Eritrea. The main difference between the Beja orthography and the transcription used in this book is the use of bold vowel letters. They indicate the pitch-accent.

<24>

There also is an Arabic transliteration of the Beja data. Among the various Arabic transliterations which were in use, this particular one has been promoted in the internet for several years. It distinguishes uu and ii from oo and ee, but its disadvantage is that it uses diacritics to do so. Recently, however, its main website (sakanab) stopped supporting it.

<25>

Since there still are Beja individuals who occasionally use Arabic letters for their language (cf. the “bejaculture” website), the Arabic transliteration - although defunct - has not been deleted from this book.

<26>

The following table is an alphabetical list of the orthographic symbols and their pronunciation (i.e. the IPA values). Some of the sound files are arranged in a different order, such as the files for dh, gh, and kh.

<31>

It will be noticed that some words - including common greetings - are borrowed from Arabic. However, sounds which do not fit the Beja sound system will generally be rejected, yet, depending on the speaker, certain Arabic sounds will be preserved or integrated as table 2 shows.

<32>

As far as loan words are concerned, they will be re-shaped to be acceptable to the Beja language. The shape of loan words is predictable from the typical Beja patterns which will be presented in table 3 further below.

<33>

As far as the places of consonants (C) and vowels (V) in syllables are concerned, Beja words allow for exactly 18 syllable patterns in the word initial syllable, and 12 patterns in all other syllables.

<34>

The most frequent pattern is CV, i.e. a syllable which consists of a consonant followed by a short vowel. But syllables can also be closed with one or two consonants, where the second consonant can only be t. So far this gives 3 patterns, namely: CV as in na 'thing', CVC as in tak 'man', and CVCt as in lamt 'soup'.

<35>

Less frequent are syllables with long vowels [1], which will be written as VV. This gives 3 additional syllables patterns: CVV as in tuu 'the', CVVC as in been 'that', and CVVCt as in diint 'thorn'.

<36>

Only word initial syllables can start with a vowel (or more exactly, with a suppressed hamzah), which gives the following additional patterns: V as in a.ni 'I (1.Sg)', VC as in ar 'boys', VCt as in awt 'honey', VV as in aa.bu 'who (M)', VVC as in aab 'who', and VVCt as in aabt 'who (F)'.

<37>

As a specialty of the Beja phonology there also are syllables in which the initial consonant is followed either by h or by hamzah, abbreviated as H  [2]: CHV as in t'a 'now', CHVV as in m'aa 'come!', CHVCt as in mhay 'three' and as in lhayt 'tomorrow', and CHVVCt as in n'aayt 'goat'. This last set of syllable patterns actually causes one of the systematic differences between northern and southern Beja dialects: What is pronounced as C'V in the North may be pronounced as CV' or as two syllables CV.'V in the South, e.g. s'a, sa', sa.'a 'sit! (M)'.