Hadza Language – Classification, Phonology, Grammar, and Numerals
Hadza language is an isolate language spoken by close to 1,000 Hadza tribe members along the coasts of Lake Eyasi. The Hadza people are the last full-time hunter-gatherers in Africa. Hadza language is one of only three East African languages with clicks. Despite the limited number of speakers, the Hadza language is used extensively, with most kids learning it. However, the language is categorized as a vulnerable language by UNESCO.
Name
The Hadza are known by many names in the literature. The word ‘Hadza’ itself translates to ‘human being.’ ‘Hadzabe’ or ‘hazabee’ is the plural form, and ‘Hadzapi’ or ‘hazaphii’ translates to ‘they are men.’ ‘Hatsa’ and ‘Hatza’ are older German spellings. Hadza language is occasionally distinguished as ‘Hadzane or ‘hazane,’ which means ‘of the Hadza.’
‘Tindiga’ is from Kiswahili (watindiga) which means ‘people of the marsh grass,’ a reference to the big spring in Mangola, and kitindiga, which means the language of the same. Apparently, ‘kindiga’ is of the same form derived from one of the native Bantu languages, probably Isanzu. ‘Kangeju’ (pronounced Kangeyu) is an antiquated German name of unclear origin. ‘Wahi’ (pronounced Vahi)is the German spelling of the Sukuma name for either a Sukuma clan that traces its origin to the Hadza or the Hadza that live west of the lake.
Hadza Language Classification
According to Starostin 2013 and Sands 1998, Hadza is a language isolate. Along with its neighbor Sandawe, many linguists once categorized the Hadza language as a Khoisan language, mainly because they both have clicks. However, Hadza has very few proposed lexical cognates with either Sandawe or other reputed Khoisan languages. Even many of the proposed cognates appear doubtful. For instance, the links with Sandawe are Cushitic loanwords. In contrast, the links with southern Africa are so scanty (often solitary consonant-vowel syllables) that they are most probably coincidental. A few terms link Hadza language with Oropom, which may be spurious. The numerals piye /pie/ (two) and itchâme /it͡ʃʰaame/ ‘one’ imply a connection with Kw’adza, an extinct language spoken by hunter-gatherers who may have recently shifted to Cushitic. (Higher numerals were typically borrowed in both languages).
Hadza language has no dialect, although there are some regional vocabularies, particularly Bantu loan words. These loan words are more numerous in the western and southern areas of high bilingualism. Ethnologue marked the Hadza language as “threatened.”
Learn Hadza Language Phonology
The syllable structure of Hadzane is limited to CV, sometimes CVN if nasal vowels are analyzed as coda nasals. Vowel-initial syllables don’t occur initially. Medially, they may be equal to /hV/ – at least; there are no known minimal pairs of /h/ vs. zero.
Hadzane is noted for its medial clicks, i.e., clicks within morphemes hence why we always call it Hadza click language. The Nguni Bantu languages and Sandawe are also known to have this distribution. However, it is not found in the Khoisan languages of southern Africa. Historically, some of these words are derived from clicks in initial positions (many seem to manifest lexicalized reduplication, while some are as a result of prefixes). However, others are opaque. As in Sandawe, most medial clicks in Hadzane are glottalised, but not all: naxhi ‘to be crowded,’ laqo ‘to trip someone,’ puche ’a spleen,’ khaxxe ‘to jump,’ shenqe ‘to peer over,’ binxo ‘to carry kills under one’s belt,’ haqqa-ko ’a stone,’ tacce ’a belt,’ tanche ‘to target,’ minca ‘to lick one’s lips,’ penqhenqhe or peqeqhe ‘to hurry,’ keqhe-na ‘slow,’ exekeke ‘to listen.’
The Hadza Language Tone
Neither pitch accent nor lexical tone has been demonstrated for Hadzane. Also, there is no known grammatical use of tone/stress or lexical minimal pairs.
Hadza Language Vowels
There are five vowels in the Hadza language [a e i o u]. When the intervocalic [ɦ] is omitted, long vowels may occur. For instance, [kʰaː] or [kʰaɦa] ‘to climb,’ but some words are not indicated with [ɦ], as [boko] ‘to be sick’ vs. [boːko] ‘she.’ All vowels that appear before voiced nasal clicks and glottalized nasals are nasalized. Speakers differ on whether they hear them as VN sequences or as nasal vowels. Though uncommon, invariable nasals do occur. However, they do not occur before consonants that have a point of articulation to assimilate to. In such positions, [CVNCV] and [CṼCV] are allophones. However, since VN cannot occur at the terminal position in a word or before a glottal consonant, where there are only nasal vowels, nasal vowels may be allophonic with VN in all positions.
Hadza Language Consonants
Dental-alveolar | Velar | Post alveolar ~palatal | Labial | Glottal | |||||||||
Lateral | Central | Labialised | Plain | Lateral | Central | ||||||||
Stop | Aspirated | tʰ ⟨th⟩ | kʷʰ ⟨khw⟩ | kʰ ⟨kh⟩ | pʰ ⟨ph⟩ | ||||||||
Ejective | pʼ ⟨bb⟩2 | ||||||||||||
Tenius | t ⟨t⟩ | kʷ ⟨kw⟩ | k ⟨k⟩ | p ⟨p⟩ | ʔ ⟨–⟩ | ||||||||
Voiced | d ⟨d⟩ | ɡʷ ⟨gw⟩ | ɡ ⟨g⟩ | b ⟨b⟩ | |||||||||
Nasal | n ⟨n⟩ | ŋʷ ⟨ngʼw⟩ | ŋ ⟨ngʼ⟩ | ɲ ⟨ny⟩ | m ⟨m⟩ | ||||||||
Tenius prenasalised | ƞt ⟨nd⟩ | ƞkʷ ⟨ngw⟩ | ƞk ⟨ng⟩ | ƞp ⟨mb⟩ | |||||||||
Aspirated prenasalised | ƞtʰ ⟨nt⟩ | ƞkʰ ⟨nk⟩ | ƞpʰ ⟨mp⟩ | ||||||||||
Click | Glottalised nasal | ᵑǀ͡ʔ ⟨cc⟩ | ᵑǁ͡ʔ ⟨xx⟩ | ᵑǃ͡ʔ ⟨qq⟩ | (ᵑʘʷ
~ ᵑʘ͡ʔ ⟨mcw⟩)1 |
||||||||
Nasal | ᵑǀ ⟨nc⟩ | ᵑǁ ⟨nx⟩ | ᵑǃ ⟨nq⟩ | ||||||||||
Aspirated | ᵏǀʰ ⟨ch⟩ | ᵏǁʰ ⟨xh⟩ | ᵏǃʰ ⟨qh⟩ | ||||||||||
Tenius | ᵏǀ ⟨c⟩ | ᵏǁ ⟨x⟩ | ᵏǃ ⟨q⟩ | ||||||||||
Affricate | Aspirated | t͜sʰ ⟨tsh⟩ | c͜ʎ̥˔ʰ ⟨tlh⟩3 | t͜ʃʰ ⟨tch⟩ | |||||||||
Ejective | t͜sʼ ⟨zz⟩ | k͜xʷʼ ⟨ggw⟩ | k͜xʼ ⟨gg⟩4 | cʎ̥˔ʼ ⟨dl⟩3 | t͜ʃʼ ⟨jj⟩ | ||||||||
Tenius | t͜s ⟨ts⟩ | cʎ̥˔ ⟨tl⟩3 | t͜ʃ ⟨tc⟩ | ||||||||||
Voiced | d͜z ⟨z⟩ | d͜ʒ ⟨j⟩ | |||||||||||
Tenius prenasalised | ƞt͜s ⟨nz⟩ | ƞt͜ʃ ⟨nj⟩ | |||||||||||
Aspirated prenasalised | ƞt͜sʰ ⟨nts/ns⟩ | ƞt͜ʃʰ ⟨ntc⟩ | |||||||||||
Approximant | ɾ ~ l ⟨l, r⟩5 | (w ⟨w⟩)7 | (j ⟨y⟩)7 | ɦ ⟨h⟩7 | |||||||||
Fricative | ɬ ⟨sl⟩ | s ⟨s⟩ | (x ⟨hh⟩)6 | ʃ ⟨sh⟩ | fʷ ⟨f⟩ | ||||||||
- The nasalization of the glottalized nasal click sounds is noticeable on preceding vowels. It is, however, not noticeable during the hold of the click itself, which is silent because of concurrent glottal closure. The labial [ᵑʘʷ] (or [ᵑʘ͡ʔ]) is found in a mimetic word where it alternates with [ᵑǀ].
- The palatal affricates could be articulated with an alveolar onset (/t͜/), although this is not required.
- The labial ejective /pʼ/ is found in only a few words.
- The velar ejective /k͜xʼ/ varies between a central affricate [k͜xʼ], a plosive [kʼ], a fricative [xʼ], and a lateral affricate [k͜ʼ]. The remaining central ejective fricatives can appear as ejective fricatives, too (i.e. [xʷʼ], [ʃʼ], [sʼ]).
- The lateral approximant /l/ sounds like a flap [ɾ]between vowels and sometimes elsewhere, particularly in rapid speech. [l] is the most common post-pausa and in repeated syllables, for example, in lola, sp. rabbit. A lateral flap realization [ɺ] is also possible.
- The unvoiced velar fricative [x] is known from only one word, where it alternates with /kʰ/.
- The NC sequences can only be found in word-initial positions in loan words. The nasal consonants /ɡ dʒ ŋ ɲ ɡʷ d ŋʷ/, the voiced obstruents and perhaps /dz/ (on a darker background) also appear to have been borrowed (Elderkin 1978).
- Zero onset and [ɦ] seem to be allophones. [j,w] may be allophones of [i,u] and what are usually transcribed as [w] next to a back vowel or [j] when positioned before a front vowel. For example, the message copula transcribed as –ha, -a, -ya, -wa are nothing more than transitions between vowels.
Hadza Language Orthography
Anyawire and Miller have devised a practical orthography for the Hadza language (Miller et al., 2013). This orthography is not being used by any Hadza language speaker as of 2015. It, therefore, has limited value for communicating in Hadza. It bears broad similarities with the orthographies of neighbouring languages like Iraqw, Isanzu, Sandawe, and Swahili. The apostrophe, which is pervasive anthropological literature transcription but causes difficulties in literacy, is not used. Vowel sequences indicate glottal stops, i.e., ⟨bee⟩ is transcribed as /beʔe/, as in ⟨Hazabee⟩ /ɦadzabeʔe/ ‘the Hadza’). True vowel sequences are separated by w or y ( /pie/ ‘two’ is written ⟨piye⟩. In some cases, an h may be justified, and ejectives and glottalised clicks by consonant lengthening (apart from reduced ⟨dl⟩ instead of *ddl for /cʼ/). The ejectives are based on voiced consonants, ⟨zz bb ggw gg dl jj⟩, because these are otherwise mostly found in borrowings and therefore not common. Tch / tʃʰ/ and tc /tʃ/ are as in Sandawe, sl /ɬ/ as in Iraqw. Ultimately, this is a French convention). Nasalised vowels VN rimes are <in un en an>. Long vowels are <aha> or <â> where they result from an elided /ɦ/. A tonic syllable is generally not marked but may be written with an acute accent ⟨á⟩.
Hadza Language Grammar
Hadza language is a head-marking language in both noun phrases and clauses. Word order is flexible; the default word order is VSO, although both VOS and fronting SVO are common. The order of noun, attributive, and determiner also varies, though with morphological implications. There is gender and number agreement on both verbs (for subjects) and attributives (for noun heads).
Usually, the first syllable of a word with a long vowel and tonic accent is reduplicated to indicate ‘just’ (meaning either ‘solely’ or ‘merely’). This reduplication is quite common. It occurs on both verbs and nouns. Also, reduplication can be used to emphasize other things like the pluractional infix ⟨kV⟩ and the habitual suffix –he.
Nouns and Pronouns in Hadza Language
In Hadza tribe language, nouns have grammatical number (plural and singular) and gender (feminine and masculine). They are indicated by suffixes as follows:
Singular | Plural | |
Masculine | -bii | |
Feminine | -ko | -bee |
The feminine plural denotes mixed natural gender, e.g., Hazabee’ the Hadza. The grammatical singular is transnumeric for many animals, as in the English language: dongoko ‘zebra’ (either single or a group). Vowel harmony may be triggered by the masculine plural, e.g., dongobee ‘zebras’ (denoting an individuated number), dungubli ‘zebra bucks.’ The diminutive suffix –nakwe and some kin terms take –te in the singular masculine (m.sg), which is otherwise unmarked.
In Hadza language, gender is used in a metaphoric sense. Ordinarily feminine words are made masculine if they are noticeably thin, while ordinarily masculine words are made feminine if they are noticeably round. Also, gender distinguishes things like tubers (feminine) and their vines (masculine) or berry trees (feminine) and their berries (masculine). Non-count nouns tend to be grammatically plural, for example, atibii ‘water’ (cf. ati ‘rain,’ atiko ‘a spring’).
However, the names of dead animals don’t follow this pattern. For instance, calling attention to a dead zebra uses the term hantayii (hantayee, plural hantayatee (rare) and hantayitchii). This is because these forms are imperative verbs and not nouns. The morphology is more unambiguous in the imperative plural when talking to more than one person: hantayetate, hantâte, hantayitchate, hantatate (replace the final –te with –si when talking to only men; check below for the verbal object suffixes -a-, -ta-, -itcha-, -eta-).
The Copula
The –pi and –pe forms of nouns usually seen in the anthropological literature (actually –phii and –phee) are copular, e.g., dongophee ‘they are zebras.’ The copular suffixes differentiate gender in all persons as well as clusivity in the first person. They are:
Masculine (singular) | Feminine (singular) | Feminine (plural) | Masculine (plural) | |
1.ex | -nee | -neko | -‘ophee | -‘uphii |
1.in | -bebee | -bibii | ||
2 | -tee | -teko | -tetee | -titii |
3 | -a | -ako | -phee | -phii |
Words with high vowels (u, i) tend to raise preceding mid vowels to be high, just as –bii does. The 3. sq copula tends to sound like a –wa (ko) or –ya(ko) after high and often mid vowels: /ea, oa/ ≈ [eja, owa], and transcriptions with y and w are common in the literature.
Hadza Language Pronouns
Demonstrative and personal pronouns in Hadza language are
Masculine (singular) | Feminine
(singular) |
Masculine (plural) | Feminine
(plural) |
|
1.inclusive | Ono | Onoko | unibii | onebee |
1.exclusive | ûbii | ôbee | ||
2 | the | theko | ithibii | ethebee |
3.given | bami | bôko | bii | bee |
3.proximal | hama | hako | habii | habee |
3.invisible | himiggê | himiggîko | himiggîbii | himiggêbee |
3.distal | naha | nâko | nâbii | nâbee |
There are other 3rd-person pronouns, which include some compound forms. The formation of adverbs entails the addition of the locative –na to the third person pronouns, e.g., beena ‘there,’ hamana ‘here’ himiggêna ‘behind/in there,’ and naná ‘over there.’
Adjectives and Verbs in Hadza Language
To indicate pluractionality, an infix (kV), where V is an echo vowel, comes after the initial syllable of verbs.
The copula was discussed above. Hadza language has many auxiliary verbs: ka- and iya- ~ ya- ‘and then,’ subjunctive i-, and negative akhwa- ‘not.’ Their inflection may not follow a defined pattern or have various inflectional endings from those of lexical verbs, which are as follows:
Tense-aspect-mood Inflections in Hadza
Anterior/non-past | Posterior/past | Veridical condition | Potential condition | Hortative/imperative | Purposive (subjunctive) | |
1sg | -ˆta | -naa | -nikwi | -nee | -na | |
1.ex | -‘ota | -‘aa | -‘ukwi | -‘ee | -ya | |
1.in | -bita | -baa | -bikwi | -bee | (use 2pl) | -ba |
2sg | -tita ~ -ita | -taa | -tikwi | -tee | -‘V | -ta |
2f.pl | -(e)têta | -(e)tea | -ˆtîkwi | -etee | ||
2m.pl | -(i)tîta | -(i)tia | -itii | -(ˆ)si | ||
3m.sg | -eya | -amo | -kwiso | -heso | -ka | -so |
3f.sg | -ako | -akwa | -kwiko | -heko | -kota | -ko |
3f.pl | -ephee | -ame | -kwise | -hese | -keta | -se |
3m.pl | -iphii | -ami | -kwisi | -hisi | -kitcha | -si |
The functions of the posterior and anterior differ between auxiliaries; with lexical verbs, they are past and non-past. The veridical and potential conditionals reflect the extent of certainty that an event would have occurred. First-person singular non-past ˆta and a few other forms lengthen the preceding vowel. The 1.ex forms apart from –ya start with a glottal stop. An echo vowel follows the imp.sg, which is a glottal stop.
Habitual forms take –he, which tends to shorten to a long vowel, before these ending. In some Hadza language verbs, the habitual has become lexicalised (marking the 3.posterior forms with a glottal stop). Therefore an actual habit takes another –he. Many compound tense-aspect-moods occur by doubling the terminal inflections. Many additional inflections have not been worked out.
The inflectional endings take the form of clitics and may occur with an adverb before the verb, leaving the verb stem bare (verb root and object suffixes).
Attributives
As is common in the area, only a few bare-root adjectives exist in Hadza language, like pakapaa ‘big’. Most attributive forms have a suffix with cross-gender number marking: -i (f.sg. and m.pl.) or -e (m.sg. and f.pl.). These suffixes agree with the noun they qualify. The–i suffix tends to trigger vowel harmony. For instance, the adjective one- ‘sweet’ has these forms
One sweet
Sg. | pl | |
f | unîko | onêbee |
m | onê(onehe) | unîbii |
The copula may replace the –bee/-ko/-bii. However, the i/e cross-number marking remains.
Adjectives, demonstratives, and other attributives may come before or after a noun; however, nouns only take their gender number endings when they come first in the nominal group: Manako unîko’ tasty meat’,ondoshibii unîbii’ sweet cordia berries,’ but unîbii ondoshi and unîko mana. In the same vein, dongoko bôko but bôko dongo ‘those zebra.’
Hadza language verbs can also be made attributive, e.g., akwiti ‘the woman (akwitiko) who is talking,’ diozo ‘to say.’ The attributive form is combined with the copula when forming the progressive aspect: dluzîneko ‘I am talking’ (female speaker), dlozênee ‘I am speaking’ (male speaker).
Object Marking
Verbs can take up to two object suffixes: an indirect object (IO) and a direct object (DO). These only differ in the 3sg and 1ex. The indirect object suffixes are also used on nouns when indicating possession, e.g. (mako-kwa ‘my pot,’ mako-a-kwa ‘it is my pot’).
Object/possessive suffixes
Singular | Plural | ||||
DO | IO | DO | IO | ||
1.ex | -kwa | -oba | -oya | ||
1.in | -ona ~ -yona | ||||
2m | -ena | -ina | |||
2f | -na | ||||
3m | -a ~ -ya ~ -na | -ma | -itcha | ||
3f | -ta | -sa | -eta | ||
Verbs only have two object suffixes when the first (the direct object) is third-person. In such a case, the direct object reduces to the attributive suffix’s form: -i (f.sg. / m.pl.) or -e (m.sg. / f.pl.); only context tells the intended combination of gender and number. Also, third-singular direct objects reduce to this form in the imperative singular; third-plural change their vowels but don’t merge with the singular: check ‘dead zebra’ in the nouns above for an example.
Hadza language Word Order
The factors that govern the word order within nominal groups are unknown. The constituent order is usually SXVO (X is an auxiliary) for an emphasized or new subject, with the subject moving further back (XVOS, XVSO, and XSVO) or simply omitted. Where semantics, context and the verbal suffixes do not disambiguate, verb-noun-noun is understood as VSO.
Numerals in Hadza Language
Prior to the introduction of Kiswahili, the Hadza did not count. Piye ‘two’ and itchâme ‘one’ are native numerals. Sámaka ‘three’ is a loan word from Datooga, while ikhumi ‘ten,’ bothano ‘five’ and bone ‘four’ are from Sukuma language. Aso ‘many’ is frequently used instead of bothano for five. There is no systematic method of expressing other numbers without using Kiswahili.
According to Dorothea Bleek, piye might have a Bantu origin; the nearest locally is Nyaturu –βĩĩ. An l/r comes between the vowels in other Bantu languages. The similitude of two and one to Kw’adza noted above was first recognised by Sands.
Names of Dead Animals
Animal |
General name |
Celebratory name |
lion | séseme | hubuwee |
giraffe | zzókwanako | háwayii |
zebra | dóngoko | hantáyii |
ostrich | khenangu | hushúwee |
buffalo | naggomako | tíslii |
eland | khomatiko | hubuwii} |
rhinoceros | tlhákate | hukhúwee |
baboon | neeko | nqokhówii |
leopard | Tcanjai, nqe | henqêe |
impala | p(h)óphoko | dlunkúwii |
hartebeest | qqeleko
bisoko |
zzonowii |
wildebeest | ||
hippoppotamus | wezzáyiko |
kapuláyii |
elephant | beggáuko | |
small antelope | hingcíyee | |
other large antelope | hephêe | |
boar | kwa’i |
hatcháyee |
warthog | dláha |
Hadza language has gotten some attention for a dozen ‘triumphal’ (Blench) or ‘celebratory’ (Woodburn) names for dead animals. These names are used to announce a kill. In the imperative singular, they are:
The names are somewhat generic. Any spotted cat may be referred to as henqêe, the same way any running ground bird may be called hushúwee. ‘Eland’ and ‘lion’ use the same root. According to Blench (2008), this may have something to do with the fact that eland is considered a magical creature in the region.
An indirect object suffix may be used to refer to the individual who killed the game. Compare hanta – ‘zebra’ with the more ordinary verbs, kw– ‘to give’ and ghasha ‘to carry’ in the imperative plural and singular (Miller 2009).
hanta-i-i hanta-i-ko-o
hanta-ta-te hanta-i-kwa-te
zebra-do.3fs-imp zebra-do.3fs-io.3fs-imp
“A zebra!” “I got a zebra!
qhasha-i-i kw-i-ko-o
qhasha-ta-te kw-i-kwa-te
carry-do.3fs-imp give-do.3fs-io.3fs-imp
“Carry it! “Give it to me!”
Theories about Early Human Language
In 2003, the press widely reported suggestions made by Joanna Mountain and Alec Knight of Stanford University that the first human language may have had click sounds. The supposed proof for this suggestion is genetic: speakers of Hadza and Jul’hoan have the most divergent known mitochondrial DNA of any human populations, implying that they were the first, or at least among the first, surviving humans to have split off the family tree. In plain terms, humanity’s three fundamental genetic divisions are the Jul’hoan, the Hadza and relatives, and everybody else. Due to the fact that two of the three divisions speak languages with click sounds, perhaps their shared ancestral language, by implication the ancestral language of all mankind, had click sounds as well.
However, apart from the genetic interpretation, this conclusion is based on many unsupported assumptions:
- Unlike the Bantu Nguni languages (Xhosa, Zulu etc.)and Yeyi, none of the two groups borrowed click sounds as part of a sprachbund;
- Sound change, a very common linguistic phenomenon, did not affect any of the languages to the point that its basic phonology became unrecognisable;
- Both groups have maintained their languages without language shift since they diverged from the rest of mankind; and
- Neither the ancestors of the Hadza nor those of the Jul’hoan developed click sounds on their own as the creators of Damin did.
No evidence exists that these assumptions are correct or even probable. The linguistic opinion is that click consonants may actually be a relatively new development in human language. They are not more unyielding to change or more likely to be linguistic relics than other sounds. Also, click sounds are easily borrowed. For instance, ǁXegwi, a Khoisan language, is believed to have ‘reborrowed’ click sounds from Bantu languages, which had initially borrowed clicks from Khoisan languages. The article by Mountain and Knight is the most recent in a long line of speculation on the ancient origin of click consonants. The speculations have largely been motivated by the obsolete idea that primitive people use primitive languages – an idea that lacks empirical backing.
Hadza Language Audio
Hadza Language In Popular Culture
In the sci-fi novel by Peter Watts Blindsight, the Hadza language is presented as the human language most related to the ancient language of vampires, citing the disproved theory that click sounds are good for hunting.
For more articles related to Tanzania languages click here!