Wiktionary:About Hittite
1=Language considerations (Hittite)Please see Module:checkparams for help with this warning.
|
This is a Wiktionary policy, guideline or common practices page. Specifically it is a policy think tank, working to develop a formal policy. |
| Policies – Entries: CFI - EL - NORM - NPOV - QUOTE - REDIR - DELETE. Languages: LT - AXX. Others: BLOCK - BOTS - VOTES. |
Introduction[edit]
The aim of this page is to serve as a guide for creating Hittite entries on English Wiktionary.
Lemmatization[edit]
Hittite lemmas should always be listed in cuneiform script. For nouns, the nominative singular should always be used as lemma. For adjectives we use the nominative singular of the common gender, and for verbs we use the infinitive. If the lemma form is not attested, the page must be treated as a reconstruction (e.g. *antaraš). The page should say on which attested forms the word is reconstructed. For words where Hittite scribes used voiced and voiceless signs interchangeably, voiceless orthographic variants have preference over voiced ones (e.g. attaš). Morphemes, however, should be lemmatized at their broad transcription. Suffixes should not include the desinence unless it's a heteroclite suffix (e.g. -nu-, never -numanzi).
Romanizations[edit]
There are three romanizations systems for Hittite. The first is the one-to-one transliteration. Here each sign is written between hyphens with its corresponding romanized form. Its purpose is to give an accurate representation of the original cuneiform script.
The second one is the broad transcription; it aims to present the language in a legible and practical way for its study. Generally, to transcribe words, the hyphens from the transliteration are removed, and adjacent repetitions of identical vowels are simplified (e.g., a-ša-an-zi > ašanzi, na-at > nat, but ši-uš > šiuš). Adjacent identical consonants are not simplified but remain geminate (ap-pa-an-zi > appanzi). Redundant vowels are expressed with a macron (e.g. e-eš-ḫar > ēšḫar), and silent vowels are written between brackets (e.g. at-ta-az, at-ta-za > attaz(a)). Signs of ambiguous voicing should always take the voiceless romanization unless it is geminate to a voiced sign, in that case it should conform to the voicing of the unambiguous sign. This romanization does not pretend to accurately reflect Hittite's pronunciation; that function is taken by the phonemic transcription.
Cuneiform[edit]
To write in cuneiform, we use a tool module called typing-aids. In order to type a word we must use the template {{chars}}. In the first parameter we set hit, and in the second we give the transliteration of the word we want written. For example, to write attaš in cuneiform script, the following code must be written: {{subst:chars|hit|at-ta-aš}}, which will automatically display "𒀜𒋫𒀸".
Phonemic transcription[edit]
As stated above, the phonemic transcription aims to represent Hittite's pronunciation in an accurate way. This will serve as guidance for Hittite phonemic reconstruction.
Consonants[edit]
| Stops | fortis | /pː/ | /tː/ | /kː/ | /kːʷ/ |
|---|---|---|---|---|---|
| lenis | /p/ | /t/ | /k/ | /kʷ/ | |
| Resonants | fortis | /rː/ | /lː/ | /nː/ | /mː/ |
| lenis | /r/ | /l/ | /n/ | /m/ | |
| Fricatives | fortis | /sː/ | /χː/ | /χːʷ/ | |
| lenis | /s/ | /χ/ | /χʷ/ | ||
| Affricate | /t͡s/ | ||||
| Glides | /w/ | /j/ | |||
Fortis and Lenis[edit]
All consonants in Hittite except for glides could be geminate. Those consonants that descend from the PIE voiceless series are often called the fortis series, while the simple ones that descend from the voiced and voiced aspirates are called the lenis. The exact phonetic difference between the fortis and lenis series is debated. Here we treat the distinction as always being one of length. Geminate "z", however, represents the sequence /t͡stː/ and /t͡st/, since it comes from the PIE double dental sequence /DD/.
Due to the imperfect script in which Hittite is written, there is no way to know whether an initial stop was fortis or lenis. (However, one may make an assumption about the pronunciation of a term if its etymology is known.) For simplicity, we assume all fortis consonants were lenited in initial position. Therefore, paḫḫur is transcribed as /paχːʷr̩/, never as /pːaχːʷr̩/.
Consonant clusters[edit]
Because each sign in cuneiform had at least one vowel, initial consonant clusters cannot be written without one silent vocalic element (e.g. ka-ra-a-wa-ar ~ karāwar ~ /ˈkraːwr̩/, *an-ta-ra-an ~ antaran ~ /n̩tran/). Sometimes these consonant clusters are not difficult to spot when the value of the vowel is inconsistent. Generally, whenever there is doubt concerning possible silent vowels, one should abstain from giving the pronunciation unless the etymology of the word is known.
Gemination has the same issue. So, whenever there is a consonant cluster involving a stop, there might be ambiguity of whether the stop is fortis or lenis (e.g. aš-a-an-ta-an ~ asāntan ~/ʔsaːntːan/).
Vowels[edit]
| Front | Central | Back | |
|---|---|---|---|
| Close | /i/, /iː/ | /ɨ/ | /u/, /uː/ |
| Mid | (/ə/) | ||
| /ɛ/, (/ɛː/) | /ɔ/, (/ɔː/) | ||
| Open | /a/, /aː/ | ||
Plene-spelling[edit]
Plene-spelling was mostly used to represent long vowels. In word initial position, plene-writing was used to denote a glottal stop (e.g. e-eš-zi /ˈʔɛst͡si/, a-a-an-si /ˈʔaːnsi/). In uncertain vowel qualities, such as e & i, it can be used to disambiguate (e.g. har-ke-e-er /ˈχarkɛɾ/).
U and Ú[edit]
Hittite had two signs for "u", written each 𒌋 (U) and 𒌑 (Ú). Even though they might appear interchangeably in some words, they mostly happen in complementary distribution. This means that the signs must have denoted different phonemic realizations; therefore, it's generally assumed that in certain environments "U" was pronounced /ɔ/ just like in Hurrian.
Whenever "u" was plene-spelled with a "U", this meant that the sequence should be analyzed as /ɔ/ (e.g. u-uC= /ɔC/, Cu-u-CV= /Cɔ/), unless it's followed by a vowel (e.g. Cu-u-a= /Cua/). On the other hand, when it's plene-spelled with a "Ú", then it represents the phoneme /u/ or /uː/ (the plene spelling would sometimes be used to disambiguate). All "u" adjacent to an "ḫ" or preceding a "p" or an "r" should always be transcribed as /ɔ/. In final positions, the sequence Ca-u and Ci-u should be transcribed as /Caɔ/ and /Ciɔ/, respectively. Before an "n" it sould always be transcribed as /ɔ/ unless it's word initial; however, the genitive word for dog is 𒆪𒌑𒈾𒀸 (ku-ú-na-aš), which was remodeled by influence of the nominative. The sequences Ca-ú-un, Ca-ú-uḫ, Ca-ú-up and Ca-ú-ur should be transcribed as /Cawɔn/, /Cawɔχː/, /Cawɔp/ and /Cawɔr/, respectively. The signs "Cu" "Ú" should always be transcribed as /u/, /uː/ or /w/ elsewhere.
Alternating i and e[edit]
All alternating "e"s and "i"s that precede an "š" and are found word initially or follow a consonant should be transcribed as /ɨ/.
/a/ and /ə/[edit]
The phonemes /a/ and /ə/ were indistinguishable from one another in cuneiform script. Even though we know they were distinct because the have two completely different etymological origins, for the sake of simplicity we ignore their distinction and treat them both under the same phoneme /a/. In addition, cuneiform "*a" from laryngeal origin should also be transcribed as /a/, as in "arunaš".
Syllabic resonants[edit]
Resonants, just like in PIE, could be the nucleus of a syllable as well (e.g. ker). They are written as m̩ n̩ r̩ l̩. In cuneiform they were always written with the "a" sign of the shape "aR", making them indistinguishable from the sequence /CaRC/.
Vowel length[edit]
The phonemes /aː/, /uː/, and /iː/ should only be transcribed when they're plene-written. The sounds /a/, /u/, and /i/ should go elsewhere. The phoneme /aː/ should also be transcribed for the thematic vowel of a-stem oxytones instead of /a/, and whenever a tonic "a" is found in an initial or final syllable. The vowels /ɛ/ & /ɛː/ appear in complementary distribution. The phoneme /ɛː/ appears in monosyllabic words and tonic open syllables, while /ɛ/ appears elsewhere. For the sake of simplicity we treat them under the same phoneme /ɛ/. It is likely that the phoneme /ɔ/ behaved symmetrically to /ɛ/ & /ɛː/, having a long allophone in the same phonetic environments.