Low Riparian: an analytic alien language
Posted: Sat Jan 12, 2019 6:55 pm
I created this language for a LARP about an alien society that's much like a whitewashed Victorian England, with one key difference: there are four genders. Mostly I was trying to test-run different pronoun systems, and the one we ended up using in the LARP was not the one I created for this language, but I did end up with a language with about 300 lexical items and fairly complete grammar. It’s a monosyllabic analytic language with complex phonotactics and no tones; its syntax is SVO and mostly head-final, with coverbs, serial verbs, and relational nouns rather than adpositions. It is more isolating than English but less than Chinese.
I bring up the gender issue just to explain why there are no kinship terms in the language - they’d be completely different from anything human. But everything else is intended to be for the most part naturalistic. I could shoehorn kinship terms, in either the gender system of the aliens (called "Trave," /tɹei̯v/, which is also the Low Riparian word for a person) or the human one, but instead of inventing new vocab I am going to describe the language as I created it in the summer of 2015.
Phonology
Note: while /t d n̥ n/ are phonetically alveolar, they are phonologically in the same series as the dental fricatives, not /s z/.
There are five vowels: /a e i o u/, <a e i o u>. Any sequence of the form /i u/ + /a e o/ or /a e o/ + /i u/ is an admissible diphthong; the sequences /iu/ and /ui/, pronounced [ju] and [wi], round up the inventory of diphthongs. The triphthongs are of the form /i/ + /a e o/ + /u/ or /u/ + /a e o/ + /i/. If due to borrowing, a triphthong such as <iei> is formed, the first element is dropped.
A syllable consists of an optional onset, a vowel (which may be a diphthong or triphthong), and an optional coda. The onsets are of the form /(s)(C)(L)/ or /z(L)/. Here, (L) indicates an optional <l r>, and (C) indicates an optional consonant other than <s z l r>. If the onset is of the form /sC(L)/, then the consonant C must be voiceless, but we will Romanize such clusters as <sm sn sñ sng> to avoid writing a superfluous <h>. The codas are of the form /(L)(C)(s)/ or /(L)z/, with the same requirement that if an /s/ is present, the consonant C must be voiceless.
Phonetically, voiceless nasals are glottalized in syllable codas, and aspirated in syllable onsets. In addition, although phonologically nasals in consonant clusters with /s/ are considered voiceless, they are neither aspirated nor glottalized, and are usually partially voiced.
The postalveolar obstruents are conventionally treated as /tʃ dʒ ʃ ʒ/, but their realization can be retroflex, alveolo-palatal, or even fully palatal. Alveolo-palatal realizations are common in onsets before /i/ or after /s/, and retroflex ones before /ɹ/ or in codas after /ɹ/. The onset /sʃ/, as well as /stʃ/ in less careful speech, is usually pronounced [ʃ:], which again may be realized as [ɕ:] or [ʂ:]. This assimilation is occasionally encountered in codas, but is uncommon and stereotyped.
Finally, observe that any combination of onset, vowel, and coda is an admissible syllable, even complex ones such as <smlieurchs>. Most syllables are far less complex than the maximum permitted, just as in English, most syllable codas are not as complex as in <sixths>.
In practice, it is not common for codas to be of the form /LC(s)/ after a triphthong or closing diphthong. Historically, closing diphthongs would not form before <l r>, and to the extent they exist in Low Riparian, it’s due to borrowing. It’s considered an urban uneducated working-class feature to drop the second element of the diphthong in this circumstance (thus, pronouncing <deirn>, “skeleton,” as <dern>). It’s also not common for coda to be of the form /Cs/ after a closing diphthong, but it’s more common than /LC/ or /LCs/, and no native speaker simplifies the vowel or the coda in this case.
In addition, the bare consonants <e o> never end a syllable; there’s always a closing diphthong element or a coda, and in borrowings of open syllables they’re replaced with <ei ou>. In contrast, <a i u> can end a syllable, although it’s rare for all three.
Orthography
We'll only ever deal with Romanized orthography, but for completeness, Low Riparian orthography is described. It is a featural system, similar to Tengwar; Tengwar itself would be very good for Low Riparian, since the presence of voiceless nasals makes its base 4*3*2 system especially easy (and again, orthographically /t d n̥ n/ are treated as dental). Like Hangeul, each syllable is written in a block, making it hard to shoehorn words that do not fit into native phonology; unlike Hangeul, the native phonotactics is so permissive that this is never a problem.
While the consonants are perfectly phonemic, the vowels are merely graphemic. The surrounding languages have seven vowels, including /ø y/, Romanized <ö ü>; like /i/, /y/ can be used as a semivowel as well as a vowel nucleus. All but the most pretentious Low Riparians pronounce them /e i/, with one exception: if an <ü> forms a diphthong or triphthong with an <i>, such as <üei> or <iü>, then it is treated as an /u/. Some less educated Low Riparians even spell common borrowings that should be spelled <ö ü> as <e i>.
I bring up the gender issue just to explain why there are no kinship terms in the language - they’d be completely different from anything human. But everything else is intended to be for the most part naturalistic. I could shoehorn kinship terms, in either the gender system of the aliens (called "Trave," /tɹei̯v/, which is also the Low Riparian word for a person) or the human one, but instead of inventing new vocab I am going to describe the language as I created it in the summer of 2015.
Phonology
MOA\POA | Labial | Dental | Alveolar | Postalveolar | Velar |
Stop/affricate | /p b/ <p b> | /t d/ <t d> | /tʃ dʒ/ <ch j> | /k g/ <k g> | |
Fricative | /f v/ <f v> | /θ ð/ <th dh> | /s z/ <s z> | /ʃ ʒ/ <sh zh> | /x ɣ/ <kh gh> |
Nasal | /m̥ m/ <mh m> | /n̥ n/ <nh n> | /ɲ̊ ɲ/ <ñh ñ> | /ŋ̊ ŋ/ <ngh ng> | |
Approximant | /ɹ/ <r> | ||||
Lateral approximant | /l/ <l> |
There are five vowels: /a e i o u/, <a e i o u>. Any sequence of the form /i u/ + /a e o/ or /a e o/ + /i u/ is an admissible diphthong; the sequences /iu/ and /ui/, pronounced [ju] and [wi], round up the inventory of diphthongs. The triphthongs are of the form /i/ + /a e o/ + /u/ or /u/ + /a e o/ + /i/. If due to borrowing, a triphthong such as <iei> is formed, the first element is dropped.
A syllable consists of an optional onset, a vowel (which may be a diphthong or triphthong), and an optional coda. The onsets are of the form /(s)(C)(L)/ or /z(L)/. Here, (L) indicates an optional <l r>, and (C) indicates an optional consonant other than <s z l r>. If the onset is of the form /sC(L)/, then the consonant C must be voiceless, but we will Romanize such clusters as <sm sn sñ sng> to avoid writing a superfluous <h>. The codas are of the form /(L)(C)(s)/ or /(L)z/, with the same requirement that if an /s/ is present, the consonant C must be voiceless.
Phonetically, voiceless nasals are glottalized in syllable codas, and aspirated in syllable onsets. In addition, although phonologically nasals in consonant clusters with /s/ are considered voiceless, they are neither aspirated nor glottalized, and are usually partially voiced.
The postalveolar obstruents are conventionally treated as /tʃ dʒ ʃ ʒ/, but their realization can be retroflex, alveolo-palatal, or even fully palatal. Alveolo-palatal realizations are common in onsets before /i/ or after /s/, and retroflex ones before /ɹ/ or in codas after /ɹ/. The onset /sʃ/, as well as /stʃ/ in less careful speech, is usually pronounced [ʃ:], which again may be realized as [ɕ:] or [ʂ:]. This assimilation is occasionally encountered in codas, but is uncommon and stereotyped.
Finally, observe that any combination of onset, vowel, and coda is an admissible syllable, even complex ones such as <smlieurchs>. Most syllables are far less complex than the maximum permitted, just as in English, most syllable codas are not as complex as in <sixths>.
In practice, it is not common for codas to be of the form /LC(s)/ after a triphthong or closing diphthong. Historically, closing diphthongs would not form before <l r>, and to the extent they exist in Low Riparian, it’s due to borrowing. It’s considered an urban uneducated working-class feature to drop the second element of the diphthong in this circumstance (thus, pronouncing <deirn>, “skeleton,” as <dern>). It’s also not common for coda to be of the form /Cs/ after a closing diphthong, but it’s more common than /LC/ or /LCs/, and no native speaker simplifies the vowel or the coda in this case.
In addition, the bare consonants <e o> never end a syllable; there’s always a closing diphthong element or a coda, and in borrowings of open syllables they’re replaced with <ei ou>. In contrast, <a i u> can end a syllable, although it’s rare for all three.
Orthography
We'll only ever deal with Romanized orthography, but for completeness, Low Riparian orthography is described. It is a featural system, similar to Tengwar; Tengwar itself would be very good for Low Riparian, since the presence of voiceless nasals makes its base 4*3*2 system especially easy (and again, orthographically /t d n̥ n/ are treated as dental). Like Hangeul, each syllable is written in a block, making it hard to shoehorn words that do not fit into native phonology; unlike Hangeul, the native phonotactics is so permissive that this is never a problem.
While the consonants are perfectly phonemic, the vowels are merely graphemic. The surrounding languages have seven vowels, including /ø y/, Romanized <ö ü>; like /i/, /y/ can be used as a semivowel as well as a vowel nucleus. All but the most pretentious Low Riparians pronounce them /e i/, with one exception: if an <ü> forms a diphthong or triphthong with an <i>, such as <üei> or <iü>, then it is treated as an /u/. Some less educated Low Riparians even spell common borrowings that should be spelled <ö ü> as <e i>.