This document requires a Unicode-compatible browser. If you don't see the IPA characters, download the Thryomanes font and set it as the default font for Unicode documents, or as a user-defined font. Some versions of Microsoft Internet Explorer may assume that you have the Lucida Sans Unicode font on your system and display the IPA characters improperly if you don't have that font.
Download the Kjaginic font to see the examples in the Tirelat alphabet.
Tirelat is written in the Kjaginiċ script (pronounced [ˈkjaɡin̪iʦ] or [ˈcjaɟin̪iʦ]), one of the more widespread scripts of Sangari origin. Kjaginic, meaning "bent letters", refers to the curved shape of strokes in the script in contrast to the straight strokes of earlier writing systems such as Tharkania. The basic Kjaginic alphabet as used in Tirelat consists of 25 symbols for consonants and 7 symbols for vowels.
The Kjaginic script contains numerous other letters which are not used in writing Tirelat, but may be used for foreign names. A romanized version of Tirelat exists for use in email, web pages, and other situations where the Kjaginic script is unavailable or inconvenient to use. The introductory Tirelat web pages are written in both Kjaginic and romanized text.
Vowels in Tirelat may be long or short; long vowels are marked by a short vertical line after the vowel letter in the Kjaginic script (e.g., aː iː uː), or the letter "h" in romanized text (ah ih uh). Syllables ending in a short vowel are considered "short", while other kinds of syllables (ending in a long vowel or a consonant) are "long". Stress is predictably placed on the first long syllable of a word, or on the first syllable if no syllable is long. Long words may also have a secondary stress, on the first syllable if neither of the first two syllables have primary stress, or on the first long syllable after the primary stress.
Each letter of the Kjaginic alphabet has a single-character representation in the Tirelat romanization. Digraphs are required for some Kjaginic letters in email and other situations where the proper accented letters are unavailable. For instance, the Tirelat word ʐaŕ (in the Kjaginic alphabet) is transcribed as "žaŕ" in the standard romanization, or "zharh" in the digraph version.
i (i) A high front unrounded vowel, [i], as in "keep". Preceding consonants may be more or less palatalized (for instance, /g/ can be pronounced [ɟ] or [ʝ] when followed by /i/). Examples: short: iŋa (iña) [ˈiŋa] "shoulder", nik (nik) [ˈnik] "mouse", kabi (kabi) [ˈkabi] "overturned"; long: iːdɬa (ihdła) [ˈiːdɮa] "ship", fiːla (fihla) [ˈfiːla] "evening primrose", laliː (lalih) [laˈliː] "maybe".
ɨ (y) A high central unrounded vowel, [ɨ]. Examples: short: ɨzira (yzira) [ˈɨziɾa] "brazil nut", ɖɔpɨk (żopyk) [ˈʣɔpɨk] "hint", ŋakɨ (ñaky) [ˈŋakɨ] "geographical pole"; long: pɨːri (pyhri) [ˈpɨːɾi] "brief", dlɨː (dłyh) [ˈdɮɨː] "ginger".
u (u) A high back rounded vowel, [u], as in "boot". Examples: short: uzga (uzga) [ˈuzɡa] "truck", vujaː (vujah) [vuˈjaː] "goodbye", laːku (lahku) [ˈlaːku] "mask"; long: uːka (uhka) [ˈuːka] "hat", duːna (duhna) [ˈduːna] "sound", piruː (piruh) [piˈɾuː] "when?".
ɛ (e) A mid-low front unrounded vowel, [ɛ], as in "wreck". Examples: short: ɛdri (edri) [ˈɛdɾi] "backward", xɛɬi (xełi) [ˈxɛɬi] "to swim", makɛ (make) [ˈmakɛ] "to give"; long: rɛːv (rehv) [ˈrɛːv] "world", katɛː (kateh) [kaˈtɛː] "arch"
ɜ (ë) A mid-low central unrounded vowel, [ɜ], as in "along". Examples: short: ɜvaʐ (ëvaž) [ˈɜvaʐ] "144", tɜr (tër) [ˈtɜr] "metal", zaːɣɜ (zahġë) [ˈzaːɣɜ] "approximately"; long: lɜːl (lëhl) [ˈlɜːl] "exterior", tɜːma (tëhma) [ˈtɜːma] "mole (animal)".
ɔ (o) A mid-low back rounded vowel, [ɔ], as in "core". Examples: short: ɔmi (omi) [ˈɔmi] "wheel", nɔgli (nogli) [ˈnɔɡli] "big", iːkɔ (ihko) [ˈiːkɔ] "galaxy"; long: ɔːki (ohki) [ˈɔːci] "guinea pig", kɔːta (kohta) [ˈkɔːta] "to carry".
a (a) A low front unrounded vowel, [a], as in French "chat" (intermediate between English "a" in "father" and "cat"). Examples: short: anavɛl (anavel) [ˈanavɛl] "Spanish moss", takki (takki) [ˈtacci] "distant", zanda (zanda) [ˈzanda] "besides"; long: aːki (ahki) [ˈaːci] "easy", taːm (tahm) [ˈtaːm] "morning", jɨraː (jyrah) [jɨˈɾaː] "ginger ale".
ʋ (w) A voiced labial-velar approximant, [w], as in "weird". Examples: ʋɨrra (wyrra) [ˈwɨrːa] "nightjar", miɬʋi (miłwi) [ˈmiɬwi] "soon", naʋ (naw) [ˈnaw] "that (just mentioned)".
m (m) A voiced bilabial nasal, [m], as in "mark". Examples: mɔki (moki) [ˈmɔci] "west", ruːma (ruhma) [ˈruːma] "bark of a tree", ɬɔːm (łohm) [ˈɬɔːm] "flower".
v (v) A voiced labiodental fricative, [v], as in "vat". Examples: vajku (vajku) [ˈvajku] "penguin", mjavi (mjavi) [ˈmjavi] "old", ʂiv (šiv) [ʃiv] "index".
f (f) A voiceless labiodental fricative, [f], as in "fork". Examples: finɖa (finża) [ˈfinʣa] "nostril", tafi (tafi) [ˈtafi] "lip", ʂif (šif) [ˈʃif] "axe".
b (b) A voiced bilabial stop, [b], as in "bell". Between vowels, pronounced as [β] (no English equivalent). Examples: baːvi (bahvi) [ˈbaːvi] "hot", labu (labu) [ˈlaβu] "document", uzuːb (uzuhb) [uˈzuːb] "petroleum".
p (p) An unaspirated voiceless bilabial stop, [p], as in "spark". Examples: pjɛktɜ (pjektë) [ˈpjɛktɜ] "apricot", ɖuːpi (żuhpi) [ˈʣuːpi] "convex", naːsp (nahsp) [ˈnaːsp] "job".
n (n) A voiced alveolar nasal, [n], as in "nice". Examples: nɔx (nox) [ˈnɔx] "street", binda (binda) [ˈbinda] "to own", palaːn (palahn) [paˈlaːn] "cloud".
l (l) A voiced alveolar lateral approximant, [l], as "leaf". Examples: laːʂ (lahš) [ˈlaːʂ] "fire", ɬilli (łilli) [ˈɬilli] "fragile", gzil (gĕzil) [ɡəˈzil] "face"; kɨːla (kyhla) [ˈkɨːla] "cherry".
ɬ (ł / lh) A voiceless alveolar lateral fricative, [ɬ], as in Welsh "lliw"; voiced [ɮ] when adjacent to a voiced stop. Examples: ɬat (łat) [ˈɬat] "language", niːɬi (nihłi) [ˈniːɬi] "east", kiɬ (kił) [ˈciɬ] "ice".
r (r) A voiced alveolar trill, [r], as in Spanish "perro". Between vowels it can be pronounced as an alveolar tap [ɾ], as in Spanish "pero". Examples: [r] rin (rin) [ˈrin] "fish", marga (marga) [ˈmarɡa] "starling", luːr (luhr) [ˈluːr] "clause"; [ɾ] ʐiːri (žihri) [ˈʒiːɾi] "pretty".
ŕ (ŕ / rh) An unaspirated voiceless alveolar trill, [r̥], somewhat like Welsh "rh" but without aspiration; no English equivalent. Examples: ŕaːfi (ŕahfi) [ˈr̥aːfi] "marshmallow", xjaŕi (xjaŕi) [ˈçjar̥i] "rough, uneven", naŕ (naŕ) [ˈnar̥] "science".
z (z) A voiced alveolar sibilant, [z], as in "zone". Examples: zɛvna (zevna) [ˈzɛvna] "desert", kazi (kazi) [ˈkazi] "to live", liz (liz) [ˈliz] "silver".
s (s) A voiceless alveolar sibilant, [s], as in "sack". Examples: saːn (sahn) [ˈsaːn] "foot (part of body)", vaːsa (vahsa) [ˈvaːsa] "lid", tiris (tiris) [ˈtiɾis] "towel".
d (d) A voiced dental stop, [d̪], as in "dental", but pronounced with the tip of the tongue touching the upper front teeth. Between vowels, pronounced as [ð] in "gather". Examples: [d̪] dɨːzi (dyhzi) [ˈdɨːzi] "soft (not hard)", munda (munda) [ˈmunda] "ocean", faraːd (farahd) [faˈɾaːd] "cloth"; [ð] vɛdi (vedi) [ˈvɛði] "narrow", ʐaːdi (žahdi) [ˈʐaːði] "dream"
t (t) An unaspirated voiceless dental stop, [t̪], as in "stop", but pronounced with the tip of the tongue touching the upper front teeth. Examples: tɛza (teza) [ˈtɛza] "bicycle", kittari (kittari) [ˈcittaɾi] "chipmunk", marat (marat) [ˈmaɾat] "window".
j (j) A voiced palatal approximant, [j], as in "yet" or "onion". Preceding consonants may be more or less palatalized. Examples: jari (jari) [ˈjaɾi] "beyond", ʈjagi (ċjagi) [ˈtɕjaɟi] "bent", ŋaj (ñaj) [ˈŋaj] "every".
ʐ (ž / zh) A voiced retroflex fricative, [ʐ] (no English equivalent, but similar to Chinese "r"). When followed by /i/ or /j/, pronounced [ʒ] as in "vision". Examples: [ʐ] ʐar (žar) [ˈʐar] "color", liːʐa (lihža) [ˈliːʐa] "to see", sariːʐ (sarihž) [saˈɾiːʐ] "lava"; [ʒ] ʐim (žim) [ˈʒim] "tree", uːʐi (uhži) [ˈuʒi] "near".
ʂ (š / sh) A voiceless retroflex fricative (as in Klingon "S" or Chinese "sh"). Before /i/ or /j/, like [ʃ] as in "share". Examples: ʂim (šim) [ˈʃim] "eye", sɛʂta (sešta) [ˈsɛʂta] "winter", ɬɔʂ (łoš) [ˈɬɔʂ] "rice".
ɖ (ż / dz) A voiced alveolar affricate, [ʣ], as in "adze". Examples: ɖɛk (żek) [ˈʣɛk] "edge", juːɖa (juhża) [ˈjuːʣa] "self", vɛraːɖ (verahż) [vɛˈɾaːʣ] "puzzle".
ʈ (ċ / ts) A voiceless alveolar affricate, [ʦ], as in "pizza". Examples: ʈɛɬki (ċełki) [ˈʦɛɬci] "excellent", kuːʈi (kuhċi) [ˈkuːʨi] "next", riviʈ (riviċ) [ˈriviʦ] "sign, trace".
ŋ (ñ / ng) A voiced velar nasal, [ŋ], as in "sing". Examples: ŋurmul (ñurmul) [ˈŋurmul] "thunder", dɔːŋ (dohñ) [ˈdɔːŋ] "ring (of a bell)", mɜriŋ (mëri-ñ) [ˈmɜɾiŋ] "circle".
ɣ (ġ / gh) A voiced velar fricative, [ɣ], as in Spanish "lago". Examples: ɣuːl (ġuhl) [ˈɣuːl] "wolf", ʐiɣaːn (žiġahn) [ʒiˈɣaːn] "in exchange for", ʂɛɣ (šeġ) [ˈʂɛɣ] "wood".
x (x) A voiceless velar fricative, [x], as in "loch" or "Bach" (Spanish "baja", German "machen"). Examples: xiri (xiri) [ˈçiɾi] "yellow", xaxt (xaxt) [ˈxaxt] "oak tree" ʂvax (šĕvax) [ʂəˈvax] "cage".
k (k) An unaspirated voiceless velar stop, [k], as in "skip". Examples: kɛʂ (keš) [ˈkɛʂ] "stuff", baːka (bahka) [ˈbaːka] "hill", mik (mik) [ˈmik] "water".
g (g) A voiced velar stop, [ɡ], as in "glass". Examples: gavi (gavi) [ˈɡavi] "to exist", tɛgla (tegla) [ˈtɛɡla] "swallow (bird)", vɔg (vog) [ˈvɔɡ] "computer program".
Tirelat has a few consonant clusters that may be difficult for English speakers. Examples: kifʈ (kifċ) [ˈcifʦ] "lightning bolt", kɔŕʂ (koŕš) [ˈkɔr̥ʂ] "nut", ɬuxk (łuxk) [ˈɬuxk] "smoke", niʂɬa (nišła) [ˈniʂɬa] "potato", tkʋɨr (tkwyr) [ˈtkwɨr] "copper".
Tirelat spelling is, with very few exceptions, phonemic. Each of the contrasting sounds of Tirelat phonology is written with a distinct letter. But the exact sounds represented by these phonemes may vary depending on context. The Tirelat dictionary records a broad phonetic transcription of the sounds of Tirelat words.
Voiceless fricatives are pronounced as voiced when adjacent to voiced stops.
kjɛɬbu (kjełbu) [ˈcjɛɮbu]
Before /i/ and /j/ (and finally after /i/), affricates and velar consonants are somewhat palatalized.
ʈima (ċima) [ˈʨima]
pxaːɖi (pxahżi) [ˈpxaːdʑi]
In a few cases, a schwa [ə] is inserted between two consonants that would otherwise be difficult to pronounce. This sound, written as "ĕ" in romanized text, is left unwritten in the Kjaginic script. In dictionaries, this schwa is indicated with a punctuation mark ('). When prefixes or suffixes are added, this sound is often omitted. Examples: pnaːv (pĕnahv) [pəˈnaːv] "forehead", lpnaːv (lĕ-pnahv) [ləpˈnaːv] "my forehead"; zaːxn (zahxĕn) [ˈzaːxən] "cellar", zaːxnʂuːru (zahxnĕšuhru) [ˌzaːxnəˈʂuːɾu] "cellar door".
Tirelat uses a duodecimal (base-12) system of numbers. The basic digits in the Kjaginic script are the same as the Roman ("Arabic") numerals with the addition of : (10) and ; (11). Digits are divided into groups of three, and a raised dot (') is used as a "duodecimal point": 4A 5B6'39 (4A,5B6.3912 = 101,082.312510).
Phrases are separated by a small raised wedge (,), analogous to the comma, and the end of an ordinary sentence is marked by a lower wedge (.). The wedge is used in combination with a rising diagonal line to mark rising intonation in questions (?), and with a falling diagonal line to mark falling intonation in emphatic statements (!). A horizontal line (-) marks a brief pause. Quotations are set apart by half-brackets ([ ... ]).