Warlpiri language

Warlpiri has a standard three-vowel system, similar to that of Classical Arabic, with a phonemic length distinction creating a total of six possible vowels.

As shown in the chart, Warlpiri distinguishes five positions of articulation and has oral and nasal stops at each position. The oral stops have no phonemic voice distinction, but they display voiced and unvoiced allophones. Stops are usually unvoiced at the beginning of a word and voiced elsewhere. In both positions, they are usually unaspirated.

Warlpiri, like most other Australian languages, has no fricative consonants.

Warlpiri syllables are quite constrained in structure. All syllables begin with a single consonant, there are no syllable-initial consonant clusters and no syllable begins with a vowel. After the consonant is a single long or short vowel, which is sometimes followed by a single closing consonant. Open syllables are much more common than closed ones. No syllable ends with a stop or with the retroflex flap /ɽ/.

The most common kind of consonant cluster occurs when a syllable ends with a nasal consonant and the next syllable begins with the corresponding stop, but other clusters like /rk/ and /lp/ also occur.

Stress is not generally distinctive but is assigned by rule. Polysyllabic words receive primary stress on the first syllable, with secondary stresses tending to occur on alternate syllables thereafter; this rhythm may be broken by the structure of the word and so some three-syllable stress groups occur.

If two adjacent syllables in a Warlpiri morpheme have high vowels, those high vowels are almost always alike: both /u/ or both /i/. The number of Warlpiri roots with adjacent syllables having /u/ and /i/ is very small. Both progressive and regressive vowel harmony occur. In progressive vowel harmony, the second vowel changes to match the first; in regressive harmony, the first changes to match the second.

The tendency to prefer adjacent high vowels to be identical also spreads across morpheme boundaries within a word. Adding a suffix to a word can place a /u/ and an /i/ in contact. When that happens, one of the vowels tends to assimilate by changing to match the other vowel. That kind of assimilation is called vowel harmony and is common in the world's languages. It is found, for example, in Finnish, Hungarian, Mongolian and Turkish.

Regressive harmony occurs only when a tense suffix is attached to a verb (see below). For example, when the verb panti- (class 2) is placed in the past tense with the suffix -rnu, the result is panturnu, not *pantirnu. Progressive harmony occurs with most other kinds of suffixes. For example, when the ergative case suffix -ngku is attached to the noun karli 'boomerang', the result is karlingki, not *karlingku.

On occasion, long chains of high vowels can assimilate, each forcing the next. For example, when the class 2 verb kiji- is attached to the past tense suffix -rnu, the resulting word is kujurnu.

No Warlpiri word begins with an alveolar consonant; the first consonant of a word must be bilabial, palatal, retroflex or velar. Exceptions include borrowings such as tala 'dollar', from English dollar.

All Warlpiri words end in vowels. A word whose final meaningful component ends in a consonant is usually "corrected" by appending a meaningless suffix, usually -pa.

Warlpiri verbs are built from a few hundred verb roots, distributed among five conjugation classes. Two of the classes contain the vast majority of verb roots; the other three classes have only a few roots each.

A large class of modifying prefixes, or preverbs, are used to create verbs with specific meanings. For example, the verb root parnka- means 'run' by itself, and wurulyparnka- means 'scurry into hiding'. The preverb wuruly- is used with a few other verb roots to form other verbs of hiding or seclusion. Preverbs are sometimes reduplicated for emphasis or to change the meaning.

Most preverb-verb combinations are fixed in the lexicon, and new combinations cannot be created freely. However, a few preverbs are very productive and can be combined with many different roots, and some roots accept almost any preverb.

The verb root is followed by a tense suffix. There are five tense suffixes for each conjugation class, as shown in the following table. (Some optional variations have been omitted.)

Warlpiri nouns are assembled from thousands of roots, with a rich array of derivational techniques such as compounding and derivational suffixes. Plurals are formed by reduplication of the root.

Each full Warlpiri clause may contain an auxiliary word, which, together with the verb suffix, serves to identify tense and to clarify the relationship between main and dependent clauses. Common auxiliaries include ka (present tense), kapi (future tense), kaji (conditional). The auxiliary word is almost always the second word of a clause.

The auxiliary word also functions as the home for an elaborate family of suffixes that specify the person and number of the subject and object of the clause. They are similar to the familiar conjugational suffixes that agree with the subject in Indo-European languages, but in Warlpiri, they are placed on the auxiliary instead of on the verb and agree with the object as well as the subject.

An example of a suffixed auxiliary word can be seen in the farewell, kapirnangku nyanyi 'I will see you.' Here, kapi indicates future tense, -rna indicates first-person singular subject 'I', -ngku indicates second-person singular object 'you' and nyanyi is the nonpast form of the class 3 verb 'see'.

In the past tense, the auxiliary word often drops out completely. In that case, the agreement suffixes attach instead to the first or second word of the clause, as in nyangurnangku 'I saw you'.

The junction at which the agreement suffixes are attached can trigger progressive vowel harmony. Thus, nyanyi kapingki '(S)he will see you' shows the vowel of the suffix -ngku (second-person singular object) assimilating to the final vowel of kapi.