Spelling and Pronunciation
The Dha-Patu alphabet consists of twenty-six letters, of which all but two have English equivalents:
The seven digraphs dh gh kh ng sh th zh are transliterations of single letters: ð ʒ x ŋ š þ ž.
The names of vowels are the same as their sounds; the names of consonants are formed by adding ë, representing the mid-central vowel in 'what, love, gun': hence bë, dë, dhë, etc. In acronyms, this vowel is elided when the consonant is juxtaposed with a vowel: e.g., the Roman numerals IX, XI are pronounced ikh, khi.
Phonetic variations are indicated by diacritics. Weak vowels are marked with a grave accent (as the a in 'sofa', e in 'bet', i in 'pin', o in 'moll' (for dialects that distinguish it from 'mall'), u in 'butcher'), long vowels with the circumflex. In some cases (involving foreign loan-words), these distinctions are contrastive: cf. te 'this', tê 'tea'; lèm ‘eleven’, lêm “lame” (a general-purpose pejorative). Also, å is like the interjection 'aw' (as opposed to 'ah'); ä represents a in 'as'; ö ü are pronounced as in German.
The vowels i u, in conjunction with other vowels, are pronounced like English y and w, except after l, r, or consonant-clusters: e.g., ai like 'eye', oi as in 'oil'. This occurs both within and between words (but only when they are closely linked syntactically, i.e. within a noun- or verb-phrase): e.g. laim, pronounced much like the English word 'lime'; dha itun, pronounced much like 'thy tune'. (Note, however, that Dha-Patu vowels are always pronounced more briefly and tautly than their English equivalents.) The consonants m n ng [ŋ] l r can be pronounced without accompanying vowels, like the final syllables in bottom, button, bottle, butter. These semivowels and vocalic consonants may be marked with breves and macrons, respectively.
A weak vowel between two consonants is represented with an apostrophe: e.g. p'tor 'two', shob'd 'person'. Depending on the dialect, this vowel may be realized as either ë or an echo of the stressed vowel: pëtor ~ pòtor, shobëd ~ shobòd. (The latter case is similar to English pronunciations of 'robot' as either 'row-but' or 'row-bought'.) The weak vowel is lost in liaison with other words that end or begin with vowels: e.g. dhe ptor 'pair'.
When the same vowel is duplicated (which only occurs between words), or when a vowel and a diphthong fall together, a medial consonant is inserted (but not written). Depending on the dialect, it might be an aspirate h or a semivowel: e.g. dhi (h)aire, 'thee-high-ray'; dhi (j)itun, pronounced 'thee-ye-tune'.
Normally, primary stress falls on the first syllable of the root, except that it never falls on weak vowels, as in p'tor. No native word has more than three syllables, and most have only one or two, so secondary stress occurs only in longer, borrowed words, and in compounds; in these cases, it falls on the final syllable. Primary stress in a compound word is marked with an acute accent.
Some extra letters are used in foreign loan-words: c [ts], č (ch in 'church'), ď (j ~ g as in 'judge'), h j (as in 'hallelujah'), q (a back velar), w, y (as in Ancient Greek). The letters æ œ are equivalent to ä ö.
Foreign borrowings are often inconsistent, and variant forms are the rule rather than the exception. Generally speaking, proper nouns are taken both morphologically and orthographically in the nominative singular form; otherwise the term tends to be reduced and assimilated as much as possible: cf. Kosmos the Cosmos, kosm a cosmos; Homo genus Homo, om human (being).
Sometimes different derivations are taken as doublets: cf. filosofia (system of) philosophy, filosofos philosopher, (profession of) philosophy. The latter has the variants filosof ~ filosofê for the common and feminine genders, respectively.