Thsäv (updated OP)

Museum for the best conlanging and conworldery threads. Ask mods to move threads here.
Post Reply
User avatar
Z500
Lebom
Lebom
Posts: 85
Joined: Sun Jan 30, 2005 3:01 pm
Contact:

Thsäv (updated OP)

Post by Z500 »

some of you may remember Thsjēzb. as the name might suggest, Thsäv is a descendant. grammatically, it's pretty similar to Thsjēzb, but it has lots of consonant elision, diphthongs, vowel mutation, and semi-irregular inflections. i posted this on another board already and got some feedback, and i'd like to see what you guys here think.

Code: Select all

Thsäv [t͜sʰaːβ] is very similar to Shess in its grammar. The distinct character of Thsäv phonology comes from extensive elision.

typology

Thsäv is a moderately inflecting and somewhat isolating language.
adjectives precede the words they modify, but numbers and possessive pronouns come after.

word order

word order in Thsäv is mostly free. words may appear anywhere in the sentence. some limitations on word ordering:

 - modifiers must precede the word they modify (with the exception of numbers and, optionally, possessive pronouns)
 - the verb must be at the end of a subordinate phrase

words may be ordered to show emphasis, by moving them closer to the end of the sentence or nearer to other relevant words. aesthetic concerns, too, may govern the ordering of words.

the "default" word order is OVS, with VOS as a common dialectal variation.


phonology

stops <k p t kh ph th g b d> /k p t̪ kʰ pʰ tʰ g b d̪/
fricatives <h s z v> /h s z β/
nasals <m n mh nh> /m n̪ m̥ n̪̊/
taps <r gr br/rb> /ɾ ɾˠ ɾʷ/
approximants <w j ğ y̆> /w j jˠ ɥ/

when <s> occurs after <kh ph th> it becomes an aspirated affricate [ksʰ psʰ tsʰ]
<w j ğ y̆>, instead of being independent phonemes, impart labialization, palatalization, velar-palatalization, and labio-palatalization to the preceding consonant, respectively (e.g. ny̆ād /nᶣɒːd/)
<m n mh nh> assimilate before consonants with a different POA (e.g. pamgat [pɐŋ.gɐʔ])
<g b d z v m n> + <h> across word boundaries causes deletion of <h> and breathiness of following vowel (/pʉb.hɒɾ/ [pʉɒ̤ɾ])
<r> + <b> across word boundaries is pronounced the same as <br>
<t d> are dental, <th> is alveolar
<t> is realized as [ʔ] at the end of a word and before <n nh>
    in some dialects, /p/ and /k/ also become [ʔ] at the end of a word, rounding and backing the preceding vowels respectively
<th> is realized as [ʔ] before <kh ph th>
<k p t> + <h> sequences across word boundaries may merge to <kh ph th>
aspirated stops cannot coexist in adjacent syllables or the same syllable. if one stop is stressed, it will remain aspirated; otherwise, it is the second stop that will remain aspirated
<r> is realized as [ɾˠ] when next to <k>, and as [ɾʷ] when next to <b> or <w>
<rg> is realized as [ɾˠ]
<rd> is realized as [ɾ]; it is not modified by adjacent velar or rounded consonants like <r>


        front   mid     back
close   iː      ʉ       uː
           ɪ    (ɪ̈)
mid     ɛ ɛː            o
open    aː      ɐ (ɐː)  ɒː

short <a e i o y> /ɐ ɛ ɪ~ɪ̈ o ʉ/
long <ā ä ē ī ō ū> /ɒː ɛː aː~ɐː iː oː uː/

<o> is lowered to [ɒ] before the taps <r gr br>

vowels can be made breathy-voiced with an underdot or following <h>; e.g. <āh> /ɒ̤ː/, <oh ọ> /o̤/


syllable structure

syllables are in the form (C)(A/T)V(C)


stress

most words are stressed at the first syllable. words beginning with <yr> and <ar> are stressed on the second syllable. word stress is pronounced slightly louder than the following syllables and with a higher pitch, the pitch falling with each subsequent syllable. the next stressed syllable will rise in pitch again, though slightly lower than in the preceding word.

some words merge due to phonetic mutation (next section), being pronounced as one word with two stressed syllables.

voiced stop elision

historically, <g b d> lenited between vowels and before consonants. in AD the lenited consonants have disappeared, leaving diphthongs in place of sequences of <vowel> + <voiced stop> + <vowel>, and compensatorily lengthened vowels before <voiced stop> + <consonant> clusters. if the first vowel was followed by <b>, then it is rounded; it is backed if followed by <g>. <d> has no effect on the preceding vowel.

  g     b
a [ɑ]   [ɶ]
e [ɤ]   [œ]
i [ʊ]   [ʏ]
o [o]   [o]
y [u]   [ʉ]
ā [ɒː]  [ɒː]
ä [ɑː]  [ɶː]
ē [ʌ͓ː]  [œː]
ī [ɤː]  [yː]
ō [oː]  [oː]
ū [uː]  [uː]

[ʊ] and [ʌ͓ː] are unrounded
[ɤ] and [ɤː] are mid-back

despite the name, <z v> are also included in the process of voiced stop elision. <z> is treated identically to <d>, and <v> to <b>.

short vowels are reduced to a non-syllabic glide when adjacent to a long vowel (e.g. /ɶaː/ [ɶ̯aː]). /ɪ o ʉ/ reduce to [j w ɥ].

thagān  /tʰɑ̯ɒːn/
nhedā   /n̥ɛ̯ɒː/
ybä     /ɥaː/
ygān    /wɒːn/

short vowels also reduce before a diphthong

ky ygit
/kʉ uɪʔ/
[kᶣuɪʔ]

gyd     [gʉd]  "fish"
ygyd    [uʉd]  "fish (acc)"
thagyd  [tʰɑʉd] "of a fish"
ogyd    [oʉd]   "to a fish"
pagyd   [pɑʉd]  "at a fish"

there is much variation between the pronunciation of these diphthongs in Thsäv dialects.

/tʰɑʉd/ [tʰɑud] [tʰɑɥd] [tˠʰʉd] [tʰʊːt] [tʰoːt]


grammar


cases

Thsäv has five cases which it uses to indicate a word's function in a sentence.

 - the nominative case expresses the subject of a sentence
 - the accusative case expresses the direct object of a sentence
 - the dative-genitive case expresses the indirect object and possession
 - the lative case expresses motion and change
 - the locative case expresses locations in space and time

pronouns

       nom     acc    d-g     lat     loc
1sg    siph    meph   thsip   dihp    phip
2sg    pakh    amk    phak    okh     phak
3sg    phä     amph   thä     oph     paph
1pl    syt     ast    thsyt   dyht    phyt
2pl    pāt     ant    tāt     oth     phāt
3pl    phew    ampho  thew    pho     pho

the genitive pronouns follow the modified noun when used as possessive pronouns.

demonstratives

this    thä
that    that
yon     thēg
what,   thēd
which

question words

what     red
who      rebim
where,   parj
when     
why      orj
how many 

nouns

nouns have five cases. the nominative is unmarked. the other four have prefixes.

nom     -
acc     y- before stops, m- before vowels and approximants
dat-gen th(a)-
lat     o- (before consonant), w- (before vowel)
loc     p(a)-

many nouns begin with a voiced stop or with <v z>. these nouns undergo phonetic mutation when prefixed with one of the case prefixes. the same type of mutations occur when a bare noun exists in the middle of a sentence. here nouns are given declined in each of the five cases with some dialectal pronunciations.

fish
gyd     /gʉd/
ygyd    /uʉd/   [wʉd]   [uɥd]
thagyd  /tʰɑʉd/ [tʰˠʉd] [tʰɑɥd]
ogyd    /oʉd/   [wʉd]   [oɥd]
pagyd   /pɑʉd/

limb
bä      /baː/
ybä     /ɥaː/
thabä   /tʰɶ̯aː/ [tʰʷaː]
obä     /waː/
pabä    /pɶ̯aː/ [pʷaː]

east
gān     /gɒːn/
ygān    /wɒːn/
thagān  /tʰɑ̯ɒːn/
ogān    /wɒːn/
pagān   /pɑ̯ɒːn/

smoke
zit     /zɪʔ/
yzit    /ʉɪʔ/
thazit  /tʰɐɪʔ/
ozit    /oɪʔ/
pazit   /pɐɪʔ/

alcohol
git     /gɪʔ/
ygit    /uɪʔ/       [wɪʔ]
thagit  /tʰɑɪʔ/     [tʰˠɪʔ] [tʰɑjʔ]
ogit    /oɪʔ/       [wɪʔ]
pagit   /pɑɪʔ/      [pˠɪʔ]

son
bäh     /ba̤ː/
ybäh    /ɥa̤ː/
thabäh  /tʰɶ̯a̤ː/     [tʰʷa̤ː]
obäh    /wa̤ː/
pabäh   /pɶ̯a̤ː/


prepositions

dative-genitive
ohm     away from (idiomatic)
ma      for
be      out of, from (idiomatic)

lative
gro     through
ohm     off of, away, from
mat     to
mor     into

locative
pak     after, behind
gahd    before, ahead
thet    beside
mor     in
ag, ga  on
am, ma  over
dat     under
mat     with


numbers
                                   power   derivation
one      hor            [hɒɾ]
two      pam            [pɐm]
three    deht           [dɛ̤ʔ]
four     näm            [naːm]
five     zet            [zɛʔ]
six      prit           [pɾɪʔ]
seven    mjät           [mʲaːʔ]
eight    mūd            [muːd]
nine     thwek          [tʰwɛk]
ten      pyb            [pʉb]        1
hundred  pūv            [puːβ]       2
thousand pāh            [pɒ̤ː]        3
million  āhgat (āht)    [ɒ̤ːɐ̯ʔ]       6   many-ORD; "large number"
billion  pāhgat (pāht)  [pɒ̤ːɐ̯ʔ]      9   thousand million
trillion papāht         [pɐ.pɒ̤ːʔ]    12  thousand billion
quadr.   āhtat          [ɒ̤ː.tɐʔ]     15  million billion
quint.   pāhtat         [pɒ̤ː.tɐʔ]    18  thousand quadrillion
sext.    tāhtat         [tɒ̤ː.tɐʔ]    21  billion trillion
sept.    āhphat         [ɒ̤ː.pʰɐʔ]    24  million quintillion
oct.     papāhtat       [pɐ.pɒ̤ː.tɐʔ] 27  trillion quadrillion
non.     paphāt         [pɐ.pʰɒːʔ]   30  trillion quintillion
dec.     tapāht         [tɐ.pɒ̤ːʔ]    33  quadrillion quintillion
cent.    āhpaht         [ɒ̤ː.pɐ̤ʔ]     36  many decillion

[dɛ̤ʔ.pʉɒ̤ɾ]
dehtpybhor
three-ten-one
thirty-one

[puː.muː.pʉː.mʲaːʔ]
pūvmūdpybmjät
hundred-eight-ten-seven
187

[naːm.puː.pɐm.pʉ]
nämpūvpampyb
four-hundred-two-ten
420

four billion; eight hundred fifteen million; one hundred sixty-two thousand; three hundred forty-two
nämpāhgat mūdpūvpybzetāhgat pūvpritpybpampāh dehtpūvnämpybpam
4,815,162,342

ordinals are expressed with the -gat suffix, with the sole exception of gāh "first."

pamgat
two-ORD
second

pampybzetgat
twenty-fifth

fractions are expressed by giving the denominator in the genitive case. one half is brigt [ɾʷʊːʔ]

[tʰɐɛt ɒɾ]*
thazet hor
one fifth

* chain of two rules: aspiration (/t/ + /h/ -> [tʰ]), deaspiration is unstressed position after aspirated stop ([tʰ] -> [t])

[tʰɐ.pʉː dɛ̤ʔ]
thapyb deht
three tenths

[tʰɒ̤ːɐ̯t ɒɾ]
thāhgat hor
one millionth


phrases

In Thsäv, as in the related Memeyk languages, phrases are the building blocks of sentences. Phrases have two types of word order. Independent phrases have free word order, tending towards VOS:

[ɥaːʔ tsʰɪp bɛ.kʉɾ.to piː.βɒ̤ːd]
Ybjäth thsip bekyrto Pīvāhd?
ACC-book my read.3-PAST Pivahd?
Did Pivahd read my book?

In subordinate phrases, the verb is required to be at the end; the word order overall tends toward SOV. In this order, the verb and head of the phrase are adjacent to the modified word.

[uɪʔ tsʰɪp kɛɾ.mo kɐ.kɒː kᶣaː]
[Ygit thsip kermo] kakā ky ä!
[ACC-alcohol my steal.3-PAST] cloud is that
That's the cloud [that stole my alcohol!]

a subordinate phrase can also function as a noun:

[mɒɾ pɐˈɾoˌpʰʏ̯ɑː tʰɐ.pɐm mɒɾ pɐ̯aːp pʰuː ɥuː sɛɾʔ]
Mor parophib ag thapam mor pazäp phūg yyg sert.
in LOC-bush as GEN-two in LOC-hand bird ACC-same say.3
they say that a bird in the hand is worth [the same as] two in the bush

verb:

[ɒ̤ː ɥaː.tʰo ɛ.kʉɾ.to ʉtʰ aːm ba̤ː]
Āhd ybjätho bekyrto yth äm bäh
be.3.IRR PAST many ACC-book-PL read.3-PAST boy
there was a boy that read a lot of books

adjective:

[tʰaː ɐʔ pɐ.pʲɛʔ waːɾʔ kᶣɒːʔ]
thä dat papjet wärt ky āt
GEN-it under LOC-sun shine.3 was.3 color
it had a color that shone in the sun

adverb:

[mɐʔ o.wɪn.to tʰɐʔ pɐ.nᶣɒː ɛʉ̤.tʰo pɾɐ.tʰo]
mat owinto that pany̆ād ezyhtho pratho
to LAT-house that LOC-night drink.1-PAST walk.1-PAST
I walked home drunk that night


adjectives

Thsäv adjectives are derived from verbs. instead of conveying an action of some sort, as regular verbs do, they are verbs that are stative. when used as a predicate, they behave exactly like verbs:

[mɐɾ wɪn.to tʰaː]
mard winto thä
red.3 house 3SG.GEN
his/her house is red

when used attributively, adjectives appear before the modified noun, comprising a one-word subordinate phrase. attributive adjectives do not take verb endings:

[mɐː wɪn.to tsʰɪp kʉ taː]
mad winto thsip ky thä
red house my is this
this is my red house

when used attributively, adjectives are negated with the adverb met "not":

[paː.ɾɐ ɥɒ̤ː mɛʔ nɶː pʰaː / naː.ɾʷɐ ʉ ɐ̤ʔ ɾˠo wɒːw mʲaːk naːb]
Pädra dy āhg met näb phä, näbra dy hat gro ogādo mjäk näb
lie-3.IRR POT forever not dead it, die-3.IRR POT strange through LAT-time also death
that which is not dead may eternal lie, and with strange aeons even death may die.


adverbs

adverbs in Thsäv don't take any endings. they simply appear next to the verb they modify. they may come before or after the verb in an independent phrase, though in a subordinate phrase they are locked at the end of the phrase.

[ɾˠo.kɛ.ɾo pʰjˠɒː sʲɛo̤ mɐ tʰɑ̯ʉː ɾˠɛ.tʰɪp]
Grokero phjğād sjeoh ma thagyd grethip
out-go-3-PAST today again for GEN.fish husband
my husband went out to fish again today


verbs

verbs are used to convey action. Thsäv has free word order, but the verb tends toward the beginning of the sentence in an independent phrase, though is locked at the end of a subordinate phrase.

irregular verbs

to be

realis
1p	gā
2p	es
3p	ky

irrealis
1p	ām
2p	im
3p	äm

the realis form is used for the present tense. by itself, the irrealis form is used for the negative.

Ky pä wäkh
be.3 here wolf.PL
the wolves are here

Äm pä wäkh
be.3.IRR here wolf.PL
the wolves are not here

when the adverb äg "later" is used with an irrealis form of "to be," it expresses the future tense. the adverb yth "before" used with the realis form expresses the past tense. to make the irrealis copula negative, add the adverb "met".


regular verbs

regular verbs have two conjugations, to which are added affixes to further specify meaning. the two conjugations are realis, which has to do with states that have or are existing; and irrealis, which deals with states that do not exist. the realis endings vary widely depending on the end of the verb; the irrealis endings are regular.

overview

        realis      irrealis
1p      -h, -s, -ā  -ma
2p      *           -mi
3p      -r, -r-, -y -ra

            conjugation     affix
past tense  realis          -o after consonant (pronounced -w before another vowel), -w after vowel
negative    irrealis        -t
future      irrealis        -
subjunctive irrealis        -s
optative    irrealis        y-
imperative  realis          y-

- the negative and past tense conjugations may be combined by adding -to (-tw before a vowel) to the irrealis verb
- the negative conjugation cannot coexist with conjugations besides the past tense. if another conjugation is used, the adverb "met" can be added to the verb.

irrealis

the irrealis endings for a verb are regular; they do not change depending on the final sound of the verb.

the future tense is formed with a verb with an irrealis ending.

Pany̆ād mat owinto nāgma
LOC-night to LAT-house come-1.IRR
I will come home tonight

this verb form is not limited to expressing the future tense, however. it carries a broader connotation of any event which has not occurred or is untrue; the various irrealis endings exist to further refine the function of the verb. here an irrealis verb is used to express the consequence of a hypothetical action: 

[tʰaː wʉ.tuːk dʉm.ɾɐ wʉː ɒ̤ːm]
Thä ogytūk dymra ygyd hām
DAT.3SG LAT-smart make-3.IRR ACC-fish eat
eating fish makes you smart


realis ending details

the endings for realis verbs are more complicated. the ending used is decided by the type of sound ending the stem. the endings used are detailed below:

stem ending     change to

1p
<k p t>         <kh ph th>
<kh ph th>      <khs phs ths>
vowel + <g b d> vowel + <g b d> + ā (see voiced stop elision section above)
other consonant ending + <ā>
vowel           vowel + <s>

2p
<k t>           <s>
<g d>           <z>
vowel           vowel + <j>
other           consonant + <j> (palatalization)

3p
single          <r> + consonant, deleting <r> in syllable onset
consonant
cluster         cluster + <y>
vowel           vowel + <r>

brod "low, poor"
    broda "I am poor," broz "you are poor," vord "he/she is poor," brōma "I will be poor"
    [ɾʷoɐ], [ɾʷoz], [βoɾ], [ɾʷoː.mɐ]
nhed "to swing"
    nhedā "I swing," nhez "you swing," nherd "he/she/it swings," nhedmi "you will swing"
    [n̥ɛ̯ɒː], [n̥ɛz], [n̥ɛɾ], [n̥ɛː.mɪ]
pot "to scoop"
    poth "I scoop," pos "you scoop," port "he/she/it scoops," potra "it will scoop"
    [potʰ], [pos], [poɾʔ], [pot.ɾɐ]
prat "to walk"
    prath "I walk," pras "you walk", part "he/she/it walks," pratho "I walked"
    [pɾɐtʰ], [pɾɐs], [pɐɾʔ], [pɾɐtʰ.o]
sjem "to cover"
    sjemā "I cover," sjem "you cover," sjerm "he/she/it covers," sjemo "you covered"
    [sʲɛm.ɒː], [sʲɛm], [sʲɛɾm], [sʲɛm.o]
y̆ip "to imitate"
    y̆iph "I imitate," y̆ip "you imitate," y̆irp "he/she/it imitates," y̆irpo "it imitated"
    [ɥɪpʰ], [ɥɪp], [ɥɪɾp], [ɥɪɾp.o]
Last edited by Z500 on Thu Jul 14, 2011 2:55 pm, edited 2 times in total.
scientists have discovered a capsule that makes you not a gullible fucktard!

TomHChappell
Avisaru
Avisaru
Posts: 807
Joined: Wed Dec 28, 2005 2:58 pm

Re: Thsäv

Post by TomHChappell »

Z500 wrote:word order is free, but tends towards VOS in independent phrases and SOV in subordinate phrases.
Odds are, then, that its recent or near-recent ancestor was an SOV language but it has evolved to be a VOS language.

Is such diachronics attested in natlangs?

I know word-order does change, and I know that subordinate clauses tend to be more conservative than main clauses; but the only examples I know about are SVO from SOV.

Is a complete reversal in word-order (OSV from VSO or vice-versa, OVS from SVO or vice-versa, SOV from VOS or vice-versa) actually attested among natlangs?

Is a rotation (one of {OSV, SVO, VOS} from one of the others in that set, or one of {OVS, SOV, VSO} from one of the others in that set) attested among natlangs?

[EDIT]:
But she gives no reference for that one sentence (though she does for each of the others.)
[/EDIT]

And BTW, how does it achieve free-word-order? Is it double-marking (usually both head and dependent are marked)? Or are all dependents marked (e.g. a robust case-system)? Or are all heads marked (e.g. robust agreement among verbs and adpositions and modifiers with the gender and number of the nouns involved)?

Z500 wrote:<r > + <b > across word boundaries is pronounced the same as <br >
(1) Brackets and Slashes
If you're talking about phonemes (speech-sounds as thought of in your conspeakers' metal conception of their language) rather than graphemes (letters or characters in your romanization, or "con-script" characters in your conlang's "native" orthography), use slashes instead of angle-brackets. That is, /r/ instead of <r >, and /b/ instead of <b >, and /br/ instead of <br >.

If you're talking about phones (speech-sounds as distinguishable by a spectrogram or other scientific instruments used by linguists) rather than either phonemes or graphemes, use square brackets instead of slashes or angle-brackets. That is, [r ] instead of /r/ or <r >, and instead of /b/ or <b >, and [br] instead of /br/ or <br >.

If you're talking about graphemes instead of phonemes or phones, use angle brackets instead. That is, <r > instead of /r/ or [r ], and instead of /b/ or , and <br > instead of /br/ or [br ].

(2) Metathesis across a word-boundary? Is such a thing attested among natlangs?

Apparently so; see the links in this search. Singapore English, Istanbul Judeo-Spanish, are the two examples I can easily see on the first page.

Does your 'lang prohibit metathesis across a phrase-boundary? One of the hits in that search seems (upon not actually reading it so much as just looking at it) to hint that (nearly?) universally languages do not allow metathesis across phrase-boundaries.

It would be very interesting if your 'lang does something no natlang does, especially if "naturalism" and "realism" weren't among your design goals. OTOH if some natlang does do that, it would be very interesting to know that your conlang does something that's rare, but not unknown, among natlangs.

Z500 wrote:syllable structure

Thanks for including that!


Z500 wrote:syllables are in the form (C)(A/T)V(C)

What's "(A/T)" stand for, again?


Z500 wrote:stress

Thanks for including that!
And the rest of the phonology, too.


Z500 wrote:nouns

Thanks for including enough examples!


Z500 wrote:fractions are expressed by giving the denominator in the genitive case. one half is brigt [ɾʷʊːʔ]
That's cool! 8)
What a neat idea!
In my conlangs I might use the partitive or the elative of the denominator.
That's if those cases exist and if cardinal numbers have noun forms.

English uses the ordinal form of the denominator; this is a leftover from a phrase where "one-third (1/3) of ..." was said as "the third part of ..." and so on. Halves and quarters have their own words outside of that system, although "the fourth part of ..." was used concurrently with "a quarter of ...".


Z500 wrote:phrases
In Thsäv, as in the related Memeyk languages, phrases are the building blocks of sentences. Phrases have two types of word order. Independent phrases have free word order, tending towards VOS:

I am sure, based on your examples (which I don't quote), that you meant "clauses" instead of "phrases".
I don't know what your native language is, but in case it's French instead of English, linguists writing in French use "proposition" where those writing in English would use "clause", and French-writing linguists use "phrase" where English-writing ones would use "sentence".


Z500 wrote:In subordinate phrases, the verb is required to be at the end; the word order overall tends toward SOV. In this order, the verb and head of the phrase are adjacent to the modified word.

Which verb? Which phrase? What's the head of that phrase?
Your example seems to answer that;


Z500 wrote:[kᶣuɪʔ tsʰɪp kɛɾ.mo kɐ.kɒː aː]
Ky [ygit thsip kermo] kakā ä!
is [ACC-alcohol my steal.3-PAST] cloud that
That's the cloud [that stole my alcohol!]

Here, the RC (relative clause) must precede the noun it modifies. (Though genitives and demonstratives must follow the noun they modify.)
The RC is externally headed, and the RC's nuclear verb immediately precedes the head-noun.

How would you say
"That's the alcohol [that my cloud stole!]"
Ky [kakā thsip kermo] ygit ä!
is [cloud my steal.3-PAST] ACC-alcohol that


Z500 wrote:when used attributively, adjectives are negated with the adverb met "not":

[paː.ɾɐ ɥɒ̤ː mɛʔ nɶː pʰaː / naː.ɾʷɐ ʉ ɐ̤ʔ ɾˠo wɒːw mʲaːk naːb]
Pädra dy āhg met näb phä, näbra dy hat gro ogādo mjäk näb
lie-3.IRR POT forever not dead it, die-3.IRR POT strange through LAT-time also death
that which is not dead may eternal lie, and with strange aeons even death may die.

I didn't quite understand this.
Can't adjectives be used attributively but not be negated?
Also, is your examples translation as near to perfect as you can make it?


Z500 wrote:regular verbs
overview

realis irrealis
1p -h, -s, -ā -ma
2p * -mi
3p -r, -r-, -y -ra

Which participant's person is being agreed with here?
What about verbs that have no participants?


Z500 wrote: conjugation affix
past tense realis -o after consonant (pronounced -w before another vowel), -w after vowel
negative irrealis -t
future irrealis -
subjunctive irrealis -s
optative irrealis y-
imperative realis y-

Why is the imperative considered realis instead of irrealis?


__________________________________________________________________________________________


All told, an excellent start, more completed than many conlangs are when first posted, with many good ideas already evident.

You have maybe one notational difference (the brackets and slashes thing) that you probably should conform to the rest of us; and one terminology difference (phrase instead of clause) that you also should probably conform to the rest of us.

But, just putting up enough to criticise, is a major accomplishment; and you've done better than that IMO.

User avatar
Basilius
Avisaru
Avisaru
Posts: 398
Joined: Thu Nov 16, 2006 5:43 am
Location: Moscow, Russia

Re: Thsäv

Post by Basilius »

Z500: many interesting bits in your description... which I haven't read to the end yet, but I already have a question: how do you pronounce a sound with "velar-palatalization"?
TomHChappell wrote:
Z500 wrote:word order is free, but tends towards VOS in independent phrases and SOV in subordinate phrases.
Odds are, then, that its recent or near-recent ancestor was an SOV language but it has evolved to be a VOS language.

Is such diachronics attested in natlangs?
Odds are that the ancestral language had a much freer WO, and it got frozen differently in main and subordinate clauses.

The reason for that could be, for example, that subordinate clauses used the default placement of out-of-focus constituents, while main clauses typically included at least one in-focus constituent, and required fronting and/or postponing certain constituents depending on discourse factors, and then generalized a statistically common version of *that*.

But nobody cares these days about making the difference between rigid WO's (e. g. VOS where anything else is forbidden) and free WO's (with e. g. VOS statistically prevailing but other WO's also available), so forget what I said.
Basilius

User avatar
Z500
Lebom
Lebom
Posts: 85
Joined: Sun Jan 30, 2005 3:01 pm
Contact:

Re: Thsäv

Post by Z500 »

all excellent questions, tom. thanks for taking the time to critique my work. i'll try to answer them all thoroughly. also, i've updated the OP with some changes i've made since i first read your response.
TomHChappell wrote:
Z500 wrote:word order is free, but tends towards VOS in independent phrases and SOV in subordinate phrases.
Odds are, then, that its recent or near-recent ancestor was an SOV language but it has evolved to be a VOS language.

Is such diachronics attested in natlangs?

I know word-order does change, and I know that subordinate clauses tend to be more conservative than main clauses; but the only examples I know about are SVO from SOV.

Is a complete reversal in word-order (OSV from VSO or vice-versa, OVS from SVO or vice-versa, SOV from VOS or vice-versa) actually attested among natlangs?

Is a rotation (one of {OSV, SVO, VOS} from one of the others in that set, or one of {OVS, SOV, VSO} from one of the others in that set) attested among natlangs?
not sure. this behavior is common with Thsäv's relatives though. i realize it's kind of unrealistic - it's one of the things i plan to address with the coming overhaul to my confamily
And BTW, how does it achieve free-word-order? Is it double-marking (usually both head and dependent are marked)? Or are all dependents marked (e.g. a robust case-system)? Or are all heads marked (e.g. robust agreement among verbs and adpositions and modifiers with the gender and number of the nouns involved)?
ok, so word order isn't entirely free. heads must come after their modifiers. the sole exception is possessive pronouns and numbers, the latter of which is required to come after the noun.

Z500 wrote:<r > + <b > across word boundaries is pronounced the same as <br >
(1) Brackets and Slashes
If you're talking about phonemes (speech-sounds as thought of in your conspeakers' metal conception of their language) rather than graphemes (letters or characters in your romanization, or "con-script" characters in your conlang's "native" orthography), use slashes instead of angle-brackets. That is, /r/ instead of <r >, and /b/ instead of <b >, and /br/ instead of <br >.

If you're talking about phones (speech-sounds as distinguishable by a spectrogram or other scientific instruments used by linguists) rather than either phonemes or graphemes, use square brackets instead of slashes or angle-brackets. That is, [r ] instead of /r/ or <r >, and instead of /b/ or <b >, and [br] instead of /br/ or <br >.

If you're talking about graphemes instead of phonemes or phones, use angle brackets instead. That is, <r > instead of /r/ or [r ], and instead of /b/ or , and <br > instead of /br/ or [br ].

i'm aware of the distinctions. i believe i used the various brackets consistently. what's in the angle brackets, specifically, more closely reflects the origins of Thsäv's phonology. <gr> and <rb> clusters actually had /g/ and /b/ in them (though not necessarily [g] and )

(2) Metathesis across a word-boundary? Is such a thing attested among natlangs?

Apparently so; see the links in this search. Singapore English, Istanbul Judeo-Spanish, are the two examples I can easily see on the first page.

Does your 'lang prohibit metathesis across a phrase-boundary? One of the hits in that search seems (upon not actually reading it so much as just looking at it) to hint that (nearly?) universally languages do not allow metathesis across phrase-boundaries.

well that depends entirely on whether i think it should or not when i'm translating, so it could go either way. it's one of those things that's still kind of up in the air

It would be very interesting if your 'lang does something no natlang does, especially if "naturalism" and "realism" weren't among your design goals. OTOH if some natlang does do that, it would be very interesting to know that your conlang does something that's rare, but not unknown, among natlangs.

you've more or less hit the nail on the head here

Z500 wrote:syllable structure

Thanks for including that!


Z500 wrote:syllables are in the form (C)(A/T)V(C)

What's "(A/T)" stand for, again?

sorry. approximant/tap


Z500 wrote:stress

Thanks for including that!
And the rest of the phonology, too.


Z500 wrote:nouns

Thanks for including enough examples!


Z500 wrote:fractions are expressed by giving the denominator in the genitive case. one half is brigt [ɾʷʊːʔ]
That's cool! 8)
What a neat idea!
In my conlangs I might use the partitive or the elative of the denominator.
That's if those cases exist and if cardinal numbers have noun forms.

that makes sense. Thsäv's ancestors used to have similar cases, but they got whittled down to just the five. dative-genitive seemed the most appropriate

English uses the ordinal form of the denominator; this is a leftover from a phrase where "one-third (1/3) of ..." was said as "the third part of ..." and so on. Halves and quarters have their own words outside of that system, although "the fourth part of ..." was used concurrently with "a quarter of ...".

Z500 wrote:phrases
In Thsäv, as in the related Memeyk languages, phrases are the building blocks of sentences. Phrases have two types of word order. Independent phrases have free word order, tending towards VOS:

I am sure, based on your examples (which I don't quote), that you meant "clauses" instead of "phrases".
I don't know what your native language is, but in case it's French instead of English, linguists writing in French use "proposition" where those writing in English would use "clause", and French-writing linguists use "phrase" where English-writing ones would use "sentence".

there didn't seem to be much difference to me. it was just a word choice i went with when i started.


Z500 wrote:In subordinate phrases, the verb is required to be at the end; the word order overall tends toward SOV. In this order, the verb and head of the phrase are adjacent to the modified word.

Which verb? Which phrase? What's the head of that phrase?
Your example seems to answer that;


Z500 wrote:[kᶣuɪʔ tsʰɪp kɛɾ.mo kɐ.kɒː aː]
Ky [ygit thsip kermo] kakā ä!
is [ACC-alcohol my steal.3-PAST] cloud that
That's the cloud [that stole my alcohol!]

Here, the RC (relative clause) must precede the noun it modifies. (Though genitives and demonstratives must follow the noun they modify.)

i've made a bunch of changes since i posted this so i can't remember if this is in my OP, but this isn't the case. demonstratives precede their noun, as do genitives. possessive pronouns, however, generally follow the noun, but may come before.

The RC is externally headed, and the RC's nuclear verb immediately precedes the head-noun.

How would you say
"That's the alcohol [that my cloud stole!]"
Ky [kakā thsip kermo] ygit ä!
is [cloud my steal.3-PAST] ACC-alcohol that

lose the accusative y in front of git and you've got it.


Z500 wrote:when used attributively, adjectives are negated with the adverb met "not":

[paː.ɾɐ ɥɒ̤ː mɛʔ nɶː pʰaː / naː.ɾʷɐ ʉ ɐ̤ʔ ɾˠo wɒːw mʲaːk naːb]
Pädra dy āhg met näb phä, näbra dy hat gro ogādo mjäk näb
lie-3.IRR POT forever not dead it, die-3.IRR POT strange through LAT-time also death
that which is not dead may eternal lie, and with strange aeons even death may die.

I didn't quite understand this.
Can't adjectives be used attributively but not be negated?[/quote]
they sure can be negated. not sure where you got that from. my apologies if my post was confusing

Also, is your examples translation as near to perfect as you can make it?

well, i'll admit the vocabulary is a bit lacking; 500-odd words, a lot of which are actually grammatical items that aren't even used independently, but incorporated into words in Thsäv and related languages. also, Thsäv's verbal conjugations don't lend it to exact translation of English perfects and past perfects etc.; maybe this is something i can address, but i've generally just gone with what felt "right" when doing translations.


Z500 wrote:regular verbs
overview

realis irrealis
1p -h, -s, -ā -ma
2p * -mi
3p -r, -r-, -y -ra

Which participant's person is being agreed with here?[/quote]
the subject.

What about verbs that have no participants?

so where's the action coming from, and what's it doing? an action with unknown or generic participants can use the 3rd person, and a verb may have no objects if they are assumed from context (Thsäv is pro-drop, can't remember if i made that clear)


Z500 wrote: conjugation affix
past tense realis -o after consonant (pronounced -w before another vowel), -w after vowel
negative irrealis -t
future irrealis -
subjunctive irrealis -s
optative irrealis y-
imperative realis y-

Why is the imperative considered realis instead of irrealis?

part laziness. i didn't really feel like coming up with a different affix (and honestly i can't remember where that y- came from anymore), and i felt like throwing a little irregularity in there just for the hell of it.


All told, an excellent start, more completed than many conlangs are when first posted, with many good ideas already evident.

You have maybe one notational difference (the brackets and slashes thing) that you probably should conform to the rest of us; and one terminology difference (phrase instead of clause) that you also should probably conform to the rest of us.

But, just putting up enough to criticise, is a major accomplishment; and you've done better than that IMO.
scientists have discovered a capsule that makes you not a gullible fucktard!

TomHChappell
Avisaru
Avisaru
Posts: 807
Joined: Wed Dec 28, 2005 2:58 pm

Re: Thsäv

Post by TomHChappell »

Z500 wrote:
me wrote:
Z500 wrote:phrases
In Thsäv, as in the related Memeyk languages, phrases are the building blocks of sentences. Phrases have two types of word order. Independent phrases have free word order, tending towards VOS:
I am sure, based on your examples (which I don't quote), that you meant "clauses" instead of "phrases".
I don't know what your native language is, but in case it's French instead of English, linguists writing in French use "proposition" where those writing in English would use "clause", and French-writing linguists use "phrase" where English-writing ones would use "sentence".
there didn't seem to be much difference to me. it was just a word choice i went with when i started.
"Clause" comes from a Latin word meaning "closed".
A clause expresses a complete thought.
It has a verb and all the verb's participants or arguments or complements, whether nouns or pronouns.

See:
http://www.sil.org/linguistics/Glossary ... Clause.htm
and then
http://en.wikipedia.org/wiki/Clause.

A phrase, on the other hand, is a group of words all related to each other; it typically consists of some kind of head-word and all its dependent words or sub-phrases. A phrase needn't be an entire clause; it would be ungrammatical to just say it because it would not be a complete sentence, it would just be a fragment of one.

Most modern theories of syntax analyze most languages' clauses as being built out of phrases, which in turn are built out of smaller phrases, and so on, until you get down to the individual words. Those theories are called "phrase-structure grammars".

See
http://www.sil.org/linguistics/Glossary ... Phrase.htm
and then
http://en.wikipedia.org/wiki/Phrase.

We''ll all understand you better, and vice-versa, if you use those two terms as explained above.

User avatar
Z500
Lebom
Lebom
Posts: 85
Joined: Sun Jan 30, 2005 3:01 pm
Contact:

Re: Thsäv (updated OP)

Post by Z500 »

ok, i see the difference now. properly speaking then, most instances of "phrase" should probably be replaced with "clause" in my grammar
scientists have discovered a capsule that makes you not a gullible fucktard!

Post Reply