NGL: is there anybody still interested in tokcir?
Carlos Th
list is not longer in Yahoogroups.
I wonder if any of the people that has participated in the NGL project
is still there or if there is people still interested in this language.
-- Carlos Th
Javier BF
Hello,
I didn't take part in that list since I've only entered this
groups some months ago. I did once have a look at that list,
though, to see how the language was, but since I only found
messages concerning punctual questions I couldn't manage to
make myself a concrete idea of how the language as a whole
worked. I would appreciate if you could offer us a brief
description of it. Thanks in advance.
Best regards,
Javier
Carlos Th
Well. I have to learn it again my self but here are a few things about
the language:
Alphabet:
' a b c d e f g h i j k l m n o p r s t u v w x y z
All letters have their IPA value (nominally) except for:
{'} is /?/
{c} is /tS/
{j} is /dZ/
{x} is /x/
{y} is /j/
{z} is /Z/
And there are a few allophones like {s} that can be either [s] or [z] or
{h} that is either [h] or [x].
Sylables are (C)V(C). There cannot be more than two vowels in a row nor
more than two consonants in a row. A word cannot start or end with two
consonants or two vowels. There are no geminated consonants or vowels.
All words are stressed in the first syllable unless marked otherwise.
Nouns have three explicit cases and an unmarked form: nominative,
accusative and dative. Unmarked is used with prepositions or in fix
order.
Fix order is SOiVOd where Oi is the indirect object and Od is the direct
Object, any variation of fixed order must be explicitely marked for case
(nominative for subject, accusative for direct object and dative for
indirect object).
There are three explicit numbers and an unmarked form: singular, plural
and paucal. However there is common practice that the unmarked is used
as singular and the paucal is rarerly used... this is probably a
background problem from the current speakers...
There is currently no defined verb system beyond that there are three
tenses (plus unmarked): present, past and future. There are three
competitive verb systems.
What else? No explicit gender, however there are five semantical
genders (or classes of nouns).
Modular design.
A six hundred word lexicon known as Ogden set is the base vocabulary
however there are lots of vocabulary to cover a lot of different and
more specific situations that the ogden set does not cover.
I guess NGL, or tokcir as natively called, would be a fair non-YAEC IAL
(yet another European Conlang) (International Auxiliary Language), but
this is not the point of this language.
-- Carlos Th
[E-mail: if you want to contact me by email, replace my-deja for yahoo.]
Javier BF
> ' a b c d e f g h i j k l m n o p r s t u v w x y z
>
> All letters have their IPA value (nominally) except for:
> {'} is /?/
> {c} is /tS/
> {j} is /dZ/
> {x} is /x/
> {y} is /j/
> {z} is /Z/
>
> And there are a few allophones like {s} that can be either [s] or [z] or
> {h} that is either [h] or [x].
Umm, well, most sincerely I don't find the phoneme chart
wholly consistent. If I've understood correctly, it would be
like:
stops: p-b, t-d, k-g, '
affricates: c-j
fricatives: f-v, s-( ), x/h-( )
nasals: m, n
liquids: l, r
approximants: y, w
vowels: i e a o u
Why is voicing used for the pair f-v but not for s nor x/h?
Why confusing the values of x and h? Why having two letters
for the same phoneme?
Why, if three orders are used for stops and fricatives,
only two are distinguished for nasals?
What about poor letter <q>?
> Sylables are (C)V(C). There cannot be more than two vowels in a row nor
> more than two consonants in a row. A word cannot start or end with two
> consonants or two vowels. There are no geminated consonants or vowels.
O.K. I like that.
> All words are stressed in the first syllable unless marked otherwise.
In which cases would stress not fall in the first syllable?
How would then be it marked, possibly by means of an acute
accent?
> Nouns have three explicit cases and an unmarked form: nominative,
> accusative and dative. Unmarked is used with prepositions or in fix
> order.
If there are prepositions in use, why using suffixed (I guess)
cases as well? Why not using only prepositions for all cases?
> Fix order is SOiVOd where Oi is the indirect object and Od is the direct
> Object, any variation of fixed order must be explicitely marked for case
> (nominative for subject, accusative for direct object and dative for
> indirect object).
>
> There are three explicit numbers and an unmarked form: singular, plural
> and paucal. However there is common practice that the unmarked is used
> as singular and the paucal is rarerly used... this is probably a
> background problem from the current speakers...
Then, why having singular if we can do just with the unmarked?
I would suggest having just unmarked, and marking number
always by means of separate words.
> There is currently no defined verb system beyond that there are three
> tenses (plus unmarked): present, past and future. There are three
> competitive verb systems.
What about aspect? I find aspect to be a much more important
issue than tense.
I don't get the meaning of being "three competitive verb
systems"? Do you mean that there are three different verbal
systems in use, each used by some amount of speakers and each
trying to win over the other two?
> What else? No explicit gender, however there are five semantical
> genders (or classes of nouns).
Which ones? Only five?
> Modular design.
Mmm, I'm not sure what you mean. Do you mean something like
agglutinative structure?
> A six hundred word lexicon known as Ogden set is the base vocabulary
> however there are lots of vocabulary to cover a lot of different and
> more specific situations that the ogden set does not cover.
I think 600 is a too reduced number for a basic vocabulary
but, is that list posted somewhere? I'd like to have a look
at it before saying anything else about it.
> I guess NGL, or tokcir as natively called, would be a fair non-YAEC IAL
> (yet another European Conlang) (International Auxiliary Language), but
> this is not the point of this language.
So, what's its point?
I am currently working on a fair non-YAEC IAL myself, a language
specifically intended as such. I've already asked for help here,
but everywhere I post it the discussion sooner or later deviates
towards other subjects and people forget about my original post.
Maybe I could find some interesting ideas in NGL for it.
Best regards,
Javier
P.S. BTW, what does "NGL" stand for?
Carlos Th
>
> > Alphabet:
> > ' a b c d e f g h i j k l m n o p r s t u v w x y z
> >
> > All letters have their IPA value (nominally) except for:
> > {'} is /?/
> > {c} is /tS/
> > {j} is /dZ/
> > {x} is /x/
Oops! should have been {x} is /S/.
> > {y} is /j/
> > {z} is /Z/
> >
> > And there are a few allophones like {s} that can be either [s] or [z] or
> > {h} that is either [h] or [x].
>
> Umm, well, most sincerely I don't find the phoneme chart
> wholly consistent. If I've understood correctly, it would be
> like:
>
> stops: p-b, t-d, k-g, '
> affricates: c-j
> fricatives: f-v, s-( ), x/h-( )
With the correction:
fricatives: f-v, s-s, x-z, h-( ), h
where [z] and [s] are allophonic variants of {s} /s/.
and [x] and [h] are allophonic variants of {h} /h/.
> nasals: m, n
> liquids: l, r
> approximants: y, w
> vowels: i e a o u
Note that many natural languages have similar irregularities:
> Why is voicing used for the pair f-v but not for s nor x/h?
With the correction: voicing is meaningfull in the pairs f-v and x-z,
while it is meningless in s-s. Note that voicing [h] is not posible and
[x] is an allophone of /h/.
> Why confusing the values of x and h? Why having two letters
> for the same phoneme?
My fault, {x} is /S/.
> Why, if three orders are used for stops and fricatives,
> only two are distinguished for nasals?
As many natural languages, like, say, Spanish.
(well, {n} before a velar should be velarized [N])
> What about poor letter <q>?
{q} has no value as a letter, however is used as a shorthand for the
definite article {ku}.
> > Sylables are (C)V(C). There cannot be more than two vowels in a row nor
> > more than two consonants in a row. A word cannot start or end with two
> > consonants or two vowels. There are no geminated consonants or vowels.
>
> O.K. I like that.
>
> > All words are stressed in the first syllable unless marked otherwise.
>
> In which cases would stress not fall in the first syllable?
> How would then be it marked, possibly by means of an acute
> accent?
Well, actually they are marked with a grave accent. Words with
unstressed prefixes like {in-`} "un-", or loanwords can be stressed in
the second or another syllable.
> > Nouns have three explicit cases and an unmarked form: nominative,
> > accusative and dative. Unmarked is used with prepositions or in fix
> > order.
>
> If there are prepositions in use, why using suffixed (I guess)
> cases as well? Why not using only prepositions for all cases?
Suffixed are used for the core cases, and this violate no universla:
many languages with affixed cases also have apositions (usually at the
opposite end). German, for instance.
> > Fix order is SOiVOd where Oi is the indirect object and Od is the direct
> > Object, any variation of fixed order must be explicitely marked for case
> > (nominative for subject, accusative for direct object and dative for
> > indirect object).
> >
> > There are three explicit numbers and an unmarked form: singular, plural
> > and paucal. However there is common practice that the unmarked is used
> > as singular and the paucal is rarerly used... this is probably a
> > background problem from the current speakers...
>
> Then, why having singular if we can do just with the unmarked?
> I would suggest having just unmarked, and marking number
> always by means of separate words.
If number is not important, use the unmarked. If you want to point that
there was one element you must use the singular.
Note that generic usage: "the lion is a mamal"/"lions are mamals" should
use the unmarked.
> > There is currently no defined verb system beyond that there are three
> > tenses (plus unmarked): present, past and future. There are three
> > competitive verb systems.
>
> What about aspect? I find aspect to be a much more important
> issue than tense.
>
> I don't get the meaning of being "three competitive verb
> systems"? Do you mean that there are three different verbal
> systems in use, each used by some amount of speakers and each
> trying to win over the other two?
It means that there is three different verbal systems in use, all
including diferent ways to mark tense, aspect and mode. Each system is
defined and, unless a vault is made to decide only one of them, a
speaker can choose to use any of them for a given speech.
One of the systems use a series of suffixes, one other use serial verbs
and the third use ... adverbial particles. Using one or the other
system is like using different kinds of languages, like technical
speech, legal speech, colloquial speech, etc.
> > What else? No explicit gender, however there are five semantical
> > genders (or classes of nouns).
>
> Which ones? Only five?
... Things, tools, people, concepts and a fifth I cannot remember right
now. These are classes of nouns that have been identified a posteriory,
not a design goal.
> > Modular design.
>
> Mmm, I'm not sure what you mean. Do you mean something like
> agglutinative structure?
No. This means that each part of the grammar is thought as a module
that should be independent of any other aspect, like the verbal system
and the noun system.
This is a design concept for the development of NGL as a conlang, not
something that will be shown in tokcir as a language.
> > A six hundred word lexicon known as Ogden set is the base vocabulary
> > however there are lots of vocabulary to cover a lot of different and
> > more specific situations that the ogden set does not cover.
>
> I think 600 is a too reduced number for a basic vocabulary
> but, is that list posted somewhere? I'd like to have a look
> at it before saying anything else about it.
I do not remember the exact amount, but this is about the number of
Basic English words.
> > I guess NGL, or tokcir as natively called, would be a fair non-YAEC IAL
> > (yet another European Conlang) (International Auxiliary Language), but
> > this is not the point of this language.
>
> So, what's its point?
The point is to have a collaborative effort to make a language that will
be expresive and will conform to language universals, and that could be
used as L1 by humans, and a language that pretent to be as free as
posible to cultural baggages.
Unlike an IAL, Tokcir do not pretent to be easy to learn / logical /
unambiguous, etc. I guess it will become easier to learn than an
average natural language given the regularity, but it is not what the
language is designed for.
> I am currently working on a fair non-YAEC IAL myself, a language
> specifically intended as such. I've already asked for help here,
> but everywhere I post it the discussion sooner or later deviates
> towards other subjects and people forget about my original post.
> Maybe I could find some interesting ideas in NGL for it.
Probably.
>
> Best regards,
> Javier
>
> P.S. BTW, what does "NGL" stand for?
New Generation Language.
Javier BF
> > affricates: c-j
> > fricatives: f-v, s-( ), x/h-( )
>
> With the correction:
> fricatives: f-v, s-s, x-z, h-( ), h
> where [z] and [s] are allophonic variants of {s} /s/.
> and [x] and [h] are allophonic variants of {h} /h/.
>
> > nasals: m, n
> > liquids: l, r
> > approximants: y, w
> > vowels: i e a o u
>
> Note that many natural languages have similar irregularities:
And? Is then tokcir meant to be a natural language?
Regarding voicing, if it is used for f-v and x-z, I don't
see any good reason for turning [s] and [z] into allophones.
If a natural language has a phoneme chart with such an
inconsistency, it will tend either to extend the voicing
opposition to all fricatives or to anull it.
> > Why is voicing used for the pair f-v but not for s nor x/h?
>
> With the correction: voicing is meaningfull in the pairs f-v and x-z,
> while it is meningless in s-s. Note that voicing [h] is not posible and
> [x] is an allophone of /h/.
Who told you that voicing [h] is not possible!?! Of course
it is possible and there's even an IPA symbol for it. Check
out the chart.
> > What about poor letter <q>?
>
> {q} has no value as a letter, however is used as a shorthand for the
> definite article {ku}.
Why leaving letter q unused while using adding to the alphabet
the apostrophe for the glottal stop? Why not leaving aside
the apostrophe, which is not a proper letter, and simply use
<q> for /?/?
> > If there are prepositions in use, why using suffixed (I guess)
> > cases as well? Why not using only prepositions for all cases?
>
> Suffixed are used for the core cases, and this violate no universla:
> many languages with affixed cases also have apositions (usually at the
> opposite end). German, for instance.
Core cases are not limited to those three. In a sentence
like "I talked about politics", "about politics" is a core
case there.
> > > Fix order is SOiVOd where Oi is the indirect object and Od is the direct
> > > Object, any variation of fixed order must be explicitely marked for case
> > > (nominative for subject, accusative for direct object and dative for
> > > indirect object).
> > >
> > > There are three explicit numbers and an unmarked form: singular, plural
> > > and paucal. However there is common practice that the unmarked is used
> > > as singular and the paucal is rarerly used... this is probably a
> > > background problem from the current speakers...
> >
> > Then, why having singular if we can do just with the unmarked?
> > I would suggest having just unmarked, and marking number
> > always by means of separate words.
>
> If number is not important, use the unmarked. If you want to point that
> there was one element you must use the singular.
And why not saying just "one X" to specify that there was
one element, instead of "X-{singular}"?
> > I don't get the meaning of being "three competitive verb
> > systems"? Do you mean that there are three different verbal
> > systems in use, each used by some amount of speakers and each
> > trying to win over the other two?
>
> It means that there is three different verbal systems in use, all
> including diferent ways to mark tense, aspect and mode. Each system is
> defined and, unless a vault is made to decide only one of them, a
> speaker can choose to use any of them for a given speech.
>
> One of the systems use a series of suffixes, one other use serial verbs
> and the third use ... adverbial particles. Using one or the other
> system is like using different kinds of languages, like technical
> speech, legal speech, colloquial speech, etc.
What a mess!! And do they have their stylistic connotation
defined? OTOH, I don't really see the point of such an
unnecessary complication.
> > > Modular design.
> >
> > Mmm, I'm not sure what you mean. Do you mean something like
> > agglutinative structure?
>
> No. This means that each part of the grammar is thought as a module
> that should be independent of any other aspect, like the verbal system
> and the noun system.
>
> This is a design concept for the development of NGL as a conlang, not
> something that will be shown in tokcir as a language.
I don't really think any part of a conlang can be designed
without taking into account the rest of it. The noun system
must relate to the verbal system because they are both
interrelated in any sentence. Cases are part of the verbal
system as well as of the noun system.
> > > I guess NGL, or tokcir as natively called, would be a fair non-YAEC IAL
> > > (yet another European Conlang) (International Auxiliary Language), but
> > > this is not the point of this language.
> >
> > So, what's its point?
>
> The point is to have a collaborative effort to make a language that will
> be expresive and will conform to language universals, and that could be
> used as L1 by humans, and a language that pretent to be as free as
> posible to cultural baggages.
>
> Unlike an IAL, Tokcir do not pretent to be easy to learn / logical /
> unambiguous, etc. I guess it will become easier to learn than an
> average natural language given the regularity, but it is not what the
> language is designed for.
Which regularity? Do you intend to make the grammar regular
but to keep the sound system so irregular? And why making the
grammar regular at all, if the language is not intended to
be easy to learn nor logical nor unambiguous?
Best regards,
Javier
Carlos Th
> > Note that many natural languages have similar irregularities:
>
> And? Is then tokcir meant to be a natural language?
It is supposed to follow the language universals and behave as a natural
language. It is supposed to be designed to be used as L1 for new
generations (even if nobody will learn it this way).
> Regarding voicing, if it is used for f-v and x-z, I don't
> see any good reason for turning [s] and [z] into allophones.
> If a natural language has a phoneme chart with such an
> inconsistency, it will tend either to extend the voicing
> opposition to all fricatives or to anull it.
I didn't participate in the discutions for orthography and phonetics,
however.
In Spanish I would rather say that [b], [d] and [g] are allophones of
<b> /B/, <d> /D/ and <g> /G/ than viceversa (as usually presented).
This was you have voiceless-voiced opposition in z-d and j-g (Standard
Castillian), while f-b do not oppose (the voiceless is labiodental and
the voiced is bilabial), while you have <s> as /s/ with [z] as a common
allophone, and you have <y> as /j\/ (voiced palatal fricative) without a
voiceless counterpart. In central Colombian dialect, <z> and <s> merge
together breaking the opposition to <d> and <j> is usually [h].
In Swedish there is opposition in the labiodentals: f-v, but all the
other articulations are voiceless: s, sj, tj, h.
In the occlusives, several languages lack a voiceless bilabial
(Arabian), and several lack a voiced velar, so there is no required for
a natural languge to have a complete voiceless/voiced opposition or lack
of.
[...]
> > > What about poor letter <q>?
> >
> > {q} has no value as a letter, however is used as a shorthand for the
> > definite article {ku}.
>
> Why leaving letter q unused while using adding to the alphabet
> the apostrophe for the glottal stop? Why not leaving aside
> the apostrophe, which is not a proper letter, and simply use
> <q> for /?/?
I was not part of the discution and decition, however I do think the
orthography is nicer with {'} as glotal stop than {q}.
> > > If there are prepositions in use, why using suffixed (I guess)
> > > cases as well? Why not using only prepositions for all cases?
> >
> > Suffixed are used for the core cases, and this violate no universla:
> > many languages with affixed cases also have apositions (usually at
the
> > opposite end). German, for instance.
>
> Core cases are not limited to those three. In a sentence
> like "I talked about politics", "about politics" is a core
> case there.
No. "politics" is a core case (the accusative) of the verb "to talk
about". ;-)
[...]
> > One of the systems use a series of suffixes, one other use serial
verbs
> > and the third use ... adverbial particles. Using one or the other
> > system is like using different kinds of languages, like technical
> > speech, legal speech, colloquial speech, etc.
>
> What a mess!! And do they have their stylistic connotation
> defined? OTOH, I don't really see the point of such an
> unnecessary complication.
The designers were testing all this last time I checked. We will see if
this works or if one system should be made official.
> > > > Modular design.
> > >
> > > Mmm, I'm not sure what you mean. Do you mean something like
> > > agglutinative structure?
> >
> > No. This means that each part of the grammar is thought as a module
> > that should be independent of any other aspect, like the verbal
system
> > and the noun system.
> >
> > This is a design concept for the development of NGL as a conlang,
not
> > something that will be shown in tokcir as a language.
>
> I don't really think any part of a conlang can be designed
> without taking into account the rest of it. The noun system
> must relate to the verbal system because they are both
> interrelated in any sentence. Cases are part of the verbal
> system as well as of the noun system.
Of course a few things must be designed before others. You cannot
propose -/yq/ as a suffix for the third person dual irrealis mood when
neither /y/ and /q/ are phonems in the language, there is no dual number
and the mood is supposed to be marked by an auxiliary verb. But this
modular design seem to have worked so far when various people are
colaborating designing the conlang.
[...]
> Which regularity? Do you intend to make the grammar regular
> but to keep the sound system so irregular? And why making the
> grammar regular at all, if the language is not intended to
> be easy to learn nor logical nor unambiguous?
There is no intention of making the grammar regular, the language is
just gravitating that way. NGL is not intented as an IAL.
But, if proposed as an IAL, I don't see the irregularity of the sound
system as an obstacle. Before I became interested in languages and
linguistics I had not realized that {t} and {d} were a voiced/voiceless
opposition: they were just two different sounds no less to be confused
than {t} and {k}.
> Best regards,
> Javier
M.S. Soderquist
Cut, but featuring remarks such as:
>
> What a mess!! And do they have their stylistic connotation
> defined? OTOH, I don't really see the point of such an
> unnecessary complication.
>
(...)
>
> I don't really think any part of a conlang can be designed
> without taking into account the rest of it. The noun system
> must relate to the verbal system because they are both
> interrelated in any sentence. Cases are part of the verbal
> system as well as of the noun system.
>
>
(...)
>
> Which regularity? Do you intend to make the grammar regular
> but to keep the sound system so irregular? And why making the
> grammar regular at all, if the language is not intended to
> be easy to learn nor logical nor unambiguous?
>
This was/is a group project made by democratic process, many issues
decided only after long and healthy debate. That would account for
most of the things that you have criticised. At this point, I don't
think anyone is interested in opening the phonology and orthography
for any further debate. We had to get past that stage at some point. I
do like the sound of the language when I read a passage aloud, and
THAT is the standard by which I would judge the phonological system.
In the end, it is the flow and not the individual sounds that count.
I *personally* like the symmetry of regular grammar, myself, but I
couldn't care less about phonological regularity.
On orthographical questions, many people use {'} for the glottal stop,
and I almost always read it that way when I encounter a conlang I am
not familiar with that uses it. I am rather enamored of {q} for {ku}
anyway.
In the end, the final product includes a lot of compromises, and a lot
of issues remain unresolved. I have perceived your postings as
hostile, and that may be a fault of the medium, but if you can accept
the work that has been done, you are free to jump into the revival of
the project and help shape where it goes from here.
Mia S. Soderquist
Javier BF
> <b> /B/, <d> /D/ and <g> /G/ than viceversa (as usually presented).
Well, it could be discussed whether the main allophones
are occlusives or fricatives. On etymological basis, and
considering that in the past there was a series of voiced
fricatives apart from those fricative allophones, I would
rather place them within the occlusives.
> This was you have voiceless-voiced opposition in z-d and j-g (Standard
> Castillian), while f-b do not oppose (the voiceless is labiodental and
> the voiced is bilabial), while you have <s> as /s/ with [z] as a common
> allophone, and you have <y> as /j\/ (voiced palatal fricative) without a
> voiceless counterpart. In central Colombian dialect, <z> and <s> merge
> together breaking the opposition to <d> and <j> is usually [h].
Mmm, I think you didn't get it right about the structure
of the phoneme chart of Spanish. Labiodentals and bilabials
are at the "same" point of articulation in this system, as
well as velars (g) and uvulars (j), and as well as
interdentals (z) and dentals (d).
The logic of nowaday's Spanish consonant chart is as follows:
.............Labial.....Dental.....Alveolopalatal.....Velar
occl./affr....P-B........T-D..........CH-Y.............K-G
fric..........F..........Z.............S...............J
nas...........M..........N.............Ñ..............._
liq......................L/R/RR.......(LL)............._
As you see, it has a high degree of internal logic, especially
for what regards the occlusive/affricate/fricative part, which
is completely regular: four orders, each with a voiceless
occlusive or affricate, a voiced occlusive-fricative, a
voiceless fricative.
> In Swedish there is opposition in the labiodentals: f-v, but all the
> other articulations are voiceless: s, sj, tj, h.
That's probably because "v" in fact is today's shape of
an ancient "w" which eventually fricativized. So I think you
shouldn't really place "v" together with the other "real"
fricatives if you want to find out the internal logic of its
phoneme chart, but rather with the approximants (i.e., together
with /j/).
> In the occlusives, several languages lack a voiceless bilabial
> (Arabian), and several lack a voiced velar, so there is no required for
> a natural languge to have a complete voiceless/voiced opposition or lack
> of.
Yes, Arabian consonant chart is extremely idiosyncratic in its
nowadays shape; that's to say, quite unusual. The "b" of
today's Arabic is the former "p", as well as today's "j"
is how today appears the former "g". So, you'll only find its
internal logic when cosidering the diacronic facts.
> [...]
> > > > What about poor letter <q>?
> > >
> > > {q} has no value as a letter, however is used as a shorthand for the
> > > definite article {ku}.
> >
> > Why leaving letter q unused while using adding to the alphabet
> > the apostrophe for the glottal stop? Why not leaving aside
> > the apostrophe, which is not a proper letter, and simply use
> > <q> for /?/?
>
> I was not part of the discution and decition, however I do think the
> orthography is nicer with {'} as glotal stop than {q}.
Why? That makes the language difficult to process in databases,
because having a non-letter used as a letter can only cause
trouble. But the worst thing about it is that you are leaving
unused exactly one letter, i.e. exactly the number of letters
you have to add because of leaving that unused, and exactly
the one letter which is most highly appropriate for
representing the glottal stop --I can't think of a better
letter for /?/ than precisely <q>.
> > > > If there are prepositions in use, why using suffixed (I guess)
> > > > cases as well? Why not using only prepositions for all cases?
> > >
> > > Suffixed are used for the core cases, and this violate no universla:
> > > many languages with affixed cases also have apositions (usually at
> the
> > > opposite end). German, for instance.
> >
> > Core cases are not limited to those three. In a sentence
> > like "I talked about politics", "about politics" is a core
> > case there.
>
> No. "politics" is a core case (the accusative) of the verb "to talk
> about". ;-)
No, who told you that? You can only consider "about" as separate
from "politics" if it happened that "to talk about" were a
phrasal verb of the kind vt+adv; but "to talk about" is NOT a
verb of such kind, but of the vi+prep kind. Check it out in
a good dictionary. You can't say: "We talked politics about".
> There is no intention of making the grammar regular, the language is
> just gravitating that way. NGL is not intented as an IAL.
>
> But, if proposed as an IAL, I don't see the irregularity of the sound
> system as an obstacle. Before I became interested in languages and
> linguistics I had not realized that {t} and {d} were a voiced/voiceless
> opposition: they were just two different sounds no less to be confused
> than {t} and {k}.
Of course! And if you ask 99% of Spanish speakers about the
internal logic of its phoneme chart they will have absolutely
no idea about it. But the fact is that a phoneme system, as
well as the grammar, tends to have internal logic and thus
tends to regularization; and that's so even though the native
speakers are not consciously aware of it.
I don't see any good reason to justify the use of an
inconsistent phoneme chart in a DESIGNED language, because
inconsistencies come up in natural languages as a result
of language change through time, and those are
inconsistencies that sooner or later will be resolved by
means of a change (to extend the feature to all the cases
or to anull it), as in the case of Spanish ancient voiced
fricatives which were finally anulled.
Best regards,
Javier
Javier BF
> couldn't care less about phonological regularity.
And why is it that you like to have a regular and symmetric
grammar but don't care a bit to have a regular an symmetric
phoneme chart? The phoneme system is as important a feature
of a language as its grammar.
> On orthographical questions, many people use {'} for the glottal stop,
> and I almost always read it that way when I encounter a conlang I am
> not familiar with that uses it. I am rather enamored of {q} for {ku}
> anyway.
If the language needs exactly 26 letters for its 26 phonemes,
why not using the exactly 26 letters of the standard latin
alphabet, instead of leaving one of them illogically unused
--or even worst: used for a combination of two phonemes which
can prefectly be represented by a combination of the two letters
which represent each one of them-- and which happens to be
precisely the roman letter which most perfectly suits for that
phoneme, and having to introduce then the use of a non-letter
as a letter for the phoneme that could perfectly be represented
by <q> instead?
> In the end, the final product includes a lot of compromises, and a lot
> of issues remain unresolved. I have perceived your postings as
> hostile,
I'm sorry about that. I didn't meant them as hostile. I was
just asking for the reasons to have such inconsistencies in
a designed language.
Best regards,
Javier
Carlos Th
> > In Spanish I would rather say that [b], [d] and [g] are allophones of
> > <b> /B/, <d> /D/ and <g> /G/ than viceversa (as usually presented).
>
> Well, it could be discussed whether the main allophones
> are occlusives or fricatives. On etymological basis, and
> considering that in the past there was a series of voiced
> fricatives apart from those fricative allophones, I would
> rather place them within the occlusives.
Forget etymology. You are making claims on how language evolve to, not
where languages evolve from.
> > This was you have voiceless-voiced opposition in z-d and j-g (Standard
> > Castillian), while f-b do not oppose (the voiceless is labiodental and
> > the voiced is bilabial), while you have <s> as /s/ with [z] as a common
> > allophone, and you have <y> as /j\/ (voiced palatal fricative) without a
> > voiceless counterpart. In central Colombian dialect, <z> and <s> merge
> > together breaking the opposition to <d> and <j> is usually [h].
>
> Mmm, I think you didn't get it right about the structure
> of the phoneme chart of Spanish. Labiodentals and bilabials
> are at the "same" point of articulation in this system, as
> well as velars (g) and uvulars (j), and as well as
> interdentals (z) and dentals (d).
>
> The logic of nowaday's Spanish consonant chart is as follows:
>
> .............Labial.....Dental.....Alveolopalatal.....Velar
>
> occl./affr....P-B........T-D..........CH-Y.............K-G
> fric..........F..........Z.............S...............J
>
> nas...........M..........N.............Ń..............._
> liq......................L/R/RR.......(LL)............._
>
> As you see, it has a high degree of internal logic, especially
> for what regards the occlusive/affricate/fricative part, which
> is completely regular: four orders, each with a voiceless
> occlusive or affricate, a voiced occlusive-fricative, a
> voiceless fricative.
In my dialect, <s> and <z> are articulated the same: alveolar, while
<ch>, <y> and <ń> are definitively palatal and <y> is definitively
fricative.
Well, you can claim that is a variation of that chart, but the fact is
that is not that regular chart.
> > In Swedish there is opposition in the labiodentals: f-v, but all the
> > other articulations are voiceless: s, sj, tj, h.
>
> That's probably because "v" in fact is today's shape of
> an ancient "w" which eventually fricativized. So I think you
> shouldn't really place "v" together with the other "real"
> fricatives if you want to find out the internal logic of its
> phoneme chart, but rather with the approximants (i.e., together
> with /j/).
That eventually fricativized... but that has fricativized. (Actually
there are fricative and approximant allophones.)
> > In the occlusives, several languages lack a voiceless bilabial
> > (Arabian), and several lack a voiced velar, so there is no required for
> > a natural languge to have a complete voiceless/voiced opposition or lack
> > of.
>
> Yes, Arabian consonant chart is extremely idiosyncratic in its
> nowadays shape; that's to say, quite unusual. The "b" of
> today's Arabic is the former "p", as well as today's "j"
> is how today appears the former "g". So, you'll only find its
> internal logic when cosidering the diacronic facts.
In other words, what you are claiming is that if we go to how things
were, then the simetry exists. It does not matter how things are.
The fact is that in Spanish the <f> is labiodental (glotal for some
people) while <b> is bilabial. And <b> is pronunced as [B] far more
times than as [b].
There are many dialectal differences between where <y>, <ll>, <s>, <z>
and <ch> are, but this means that there are at least several dialects
where there are both voiced and voiceless fricatives and there is a
difference in the point of articulation of most of them:
<b>, <d>, <y> and <g> (in my dialect <y> is always fricative: not
aproximant, not affricate, not occlusive), are not articulated the same
than <f>, <s> and <j>.
> > I was not part of the discution and decition, however I do think the
> > orthography is nicer with {'} as glotal stop than {q}.
>
> Why? That makes the language difficult to process in databases,
> because having a non-letter used as a letter can only cause
> trouble. But the worst thing about it is that you are leaving
> unused exactly one letter, i.e. exactly the number of letters
> you have to add because of leaving that unused, and exactly
> the one letter which is most highly appropriate for
> representing the glottal stop --I can't think of a better
> letter for /?/ than precisely <q>.
Why should a language conform to ASCII-US? Not even English conforms to
ASCII-US.
{q} is not a letter in tokcir just as <ţ> is not a letter in Spanish.
{q} is an ideograph that stands for the article {ku}, as {&} stands for
the conjunction {et} and {|} stands for {'ior}, or while <$> stands for
"dollars" in English, and <%> for "percent".
(ASCII-US was designed as a compromise to write US English with 95
characters. If it were designed for tokcir, it would have placed {q}
next to {%} and {&}. Unicode designers had claimed that the code of a
character should not be used for alphabetizing, even if the first 128
codes are shared between Unicode and ASCII-US)
About databases. Any database that can properly alphabetize the <ń> and
the <ch> in Spanish (before the Academia decided that language should
addapt to technology and not otherwise), or the <ä>, <ü> and <w> in
Swidish, will do fine with tokcir.
> > No. "politics" is a core case (the accusative) of the verb "to talk
> > about". ;-)
>
> No, who told you that? You can only consider "about" as separate
> from "politics" if it happened that "to talk about" were a
> phrasal verb of the kind vt+adv; but "to talk about" is NOT a
> verb of such kind, but of the vi+prep kind. Check it out in
> a good dictionary. You can't say: "We talked politics about".
Looks like a duck, walks like a duck, sounds like a duck. "politics" is
the direct object that we talk about.
- What were we talking about?
- Politics.
Etymologically, well, that preposition is attached to the noun, not tho
the verb.
Anyhow, this says nothing about the naturality of some cases marked with
suffixes and some marked with prepositions.
> > There is no intention of making the grammar regular, the language is
> > just gravitating that way. NGL is not intented as an IAL.
> >
> > But, if proposed as an IAL, I don't see the irregularity of the sound
> > system as an obstacle. Before I became interested in languages and
> > linguistics I had not realized that {t} and {d} were a voiced/voiceless
> > opposition: they were just two different sounds no less to be confused
> > than {t} and {k}.
>
> Of course! And if you ask 99% of Spanish speakers about the
> internal logic of its phoneme chart they will have absolutely
> no idea about it. But the fact is that a phoneme system, as
> well as the grammar, tends to have internal logic and thus
> tends to regularization; and that's so even though the native
> speakers are not consciously aware of it.
That is why Arabians, Swedes and Hispanic-Americans have de-regularized
them, keeping an internal logic that is away from the reality.
> I don't see any good reason to justify the use of an
> inconsistent phoneme chart in a DESIGNED language, because
> inconsistencies come up in natural languages as a result
> of language change through time, and those are
> inconsistencies that sooner or later will be resolved by
> means of a change (to extend the feature to all the cases
> or to anull it), as in the case of Spanish ancient voiced
> fricatives which were finally anulled.
Then, we can let the L1 tokcir speakers to adapt the language as they
feel. Anything we decide, even extreme regularity, will be changed
through time.
Anyhow, as a proposed IAL (it means L2), that irregularity will mean
even less.
-- Celyůwi Tompin
[E-mail: if you want to contact me by email, replace my-deja for yahoo.]
.
Carlos Th
> If the language needs exactly 26 letters for its 26 phonemes,
> why not using the exactly 26 letters of the standard latin
> alphabet,
Short answer: Because Tokcir is not Latin. Hmmm. BTW, Latin didn't
use that set of letters... Anyhow, Tokcir is not US English either.
> instead of leaving one of them illogically unused
> --or even worst: used for a combination of two phonemes which
> can prefectly be represented by a combination of the two letters
> which represent each one of them-- and which happens to be
> precisely the roman letter which most perfectly suits for that
> phoneme, and having to introduce then the use of a non-letter
> as a letter for the phoneme that could perfectly be represented
> by <q> instead?
Well. {q} is not a letter in Tokcir, therefor {q} is not an unused
letter. {'} is a letter in Tokcir, therefor Tokcir is using 26 letters
for 26 phonemes:
' a b c d e f g h i j k l m n o p r s t u v w x y z
? a b tS d e f g h i dZ k l m n o p r s t u v w S j Z
Toni Keskitalo
[snip]
>The logic of nowaday's Spanish consonant chart is as follows:
>
>.............Labial.....Dental.....Alveolopalatal.....Velar
>
>occl./affr....P-B........T-D..........CH-Y.............K-G
>fric..........F..........Z.............S...............J
>
>nas...........M..........N.............Ñ..............._
>liq......................L/R/RR.......(LL)............._
>
>As you see, it has a high degree of internal logic, especially
>for what regards the occlusive/affricate/fricative part, which
>is completely regular: four orders, each with a voiceless
>occlusive or affricate, a voiced occlusive-fricative, a
>voiceless fricative.
The Standard Finnish chart is, at least on the first sight, rather
unbalanced:
labial dental velar
stops........p (b)......t (d*).........k (g).....(')
fricatives..(f) v?......s.......(S)....h
nasals.......m..........n..............N
liq./semiv...v.........l/r.......j.....-
The sounds in parentheses don't appear in "native or fully nativized"
words, except for d. In most cases, in "native-looking" words, d is a
so-called "weak grade" of /t/ (like v is of /p/ and - (nothing) is of
/k/). Otherwise it's a loan.
I've included /v/ also on the semivowel row because it often is weaker
than eg. English /v/. Some speakers have a weak glottal stop, whose
earlier existence causes consonants to double in certain word
boundaries, for the great majority of speakers.
My conlang Zegzolt, which I've developed in a rather haphazard way has
this kind of chart:
stops........p b......t d..............k g
"afficates".......... ts......tS.......kh gh
fricatives...f v......s z.....S Z......h
nasals.......m........n................N
liq./semiv...w.......l/r......j........-
In addition, there are palatalized consonants:
Stops: p' t' d' k' g' (no b', though)
fricatives: f' v' s' h' (no z' nor S', Z'. h' may be also "unvoiced
j"...)
nasals: m' n'
liquids: l' r' (these are late borrowings... "original" l' and r'
turned into /j/.)
I wonder if /t'/, /d'/ and /n'/ would actually belong to the same
place of articulation as /tS/ and /S/, /Z/...
Also, in idealized descriptions, there's "unreleased t".
That seems very messy, and I've intended to develop a history of
sounds that would explain some things. I've changed my plans many
times. That's the hazard of "doing the conlang the wrong way"... not
beginning from proto-language.
--
# Replace .invalid with .fi for personal mail only #