New to conlanging, help appreciated

[Arzemju]

Hi, i'm Arzemju /arzemjy/ and I'm intrested in conlanging since almost a year, but everytime I start a phoneme inventory I change my mind and mess it up, I never had a lexicon with over 16 words, I never suceeded in making a real conlang.

My goal now is to make an Altaic language, in which I want to put great efforts and alot of my spare time. I read the Language Construction Kit like 10 times and even with that help I can't make a conlang. I've been on the ZBB for some weeks and it seems that conlanging is harder than I thought reading the LCK.

Any help will be appreciated, thanks!

[Ossicone]

Just stick with.
Pick a phonology and don't deviate from it.
I had to force myself quite a bit to stick with the phoneme inventory of my latest conlang.

It's not your native language so the words won't come intuitively to you. And you won't know what works until after you have a good number of words. I'd say after ~100 words, if you are having issues you can make some minor changes.

Also, do you mind if I ask how old you are? You seem young, and if you are, I suggest patience. Conlanging is a lot of work, a lot of tweeking, and a learning process. It won't be perfect from the start.

[MrKrov]

Altaic is fairly vague, in the language groups included and thus the kinds of language you may be looking for. More input please.

[Arzemju]

@Ossicone I am 15 years old, and I know lot of patience is required to achieve a good conlang. I don't want a perfect one from the start, I want to just prove myself I can make a conlang; after that goal achieved i'll probably work more seriously on conlanging, who knows?

@MrKrov I'd like to make it part of the Turkic branch of Altaic languages.

Btw, a little thing: how do I pronounce 'turkic' ? (IPA transcription prefered)

[Ilaeriu]

Yay! Finally I meet a conlanger my own age :) I'm 15 too.

I agree, you really need to just find a phonology and stick with it. You're not limited to only one conlang for your entire life - you're entitled to make as many as you wish, with different phonology. You say you don't want a perfect one, so the phonology doesn't really need to be perfect, now does it? :)

If you really need to change it, I suggest starting out with a very small phonology, and then adding some sounds later rather than subtracting.

All in all, don't think too much about it. Conlanging is very much a creative process as well. Good luck :D

[MrKrov]

Some generalizations:
Phonology: front rounded vowels and vowel harmony
Nouns: plural suffix, possessive suffixes, case suffixes. No noun classess/gender.
Verbs: negation/tense/aspect/mood
Particples for relative clauses
Exclusively suffixing
Word order: SOV, left-branching

Needn't worry too much. Even Zamenhof could make one.

[Bristel]

Btw, a little thing: how do I pronounce 'turkic' ? (IPA transcription prefered)

/tɜrkɪk/ is how I pronounce it.

[Arzemju]

Some generalizations:
Phonology: front rounded vowels and vowel harmony
Nouns: plural suffix, possessive suffixes, case suffixes. No noun classess/gender.
Verbs: negation/tense/aspect/mood
Particples for relative clauses
Exclusively suffixing
Word order: SOV, left-branching

Needn't worry too much. Even Zamenhof could make one.

That's what I have actually (there are both roman and cyrillic letters, not useful actually though, it'll serve later, with the conculture)

/p b t ʦ d k g q/ <p b t c d k g q> <п б т ц к г қ>
/m n ŋ/ <m n ñ> <м н ң>
/r/ <r> <р>
/f v s z ʃ ʒ χ ʁ h/ <f v s z ş-ç j x ǵ h> <ф в с з ш-щ ж х ғ һ>
/l j/ <l y> <л ј>

/a i e y ɯ ɤ o u ə/ <a i e ü ı ö o u ə> <а і е ұ и ө о ү ә>

For aesthetic purposes, I didn't take the basic OU, but instead took the vertical one.

[Aevas]

Just stick with.
Pick a phonology and don't deviate from it.
I had to force myself quite a bit to stick with the phoneme inventory of my latest conlang.

I quite disagree with this. Personally I think it's better to make a rough sketch of the phonology, just as for everything else. The fine tuning can always be made later.

I think this is a common perception, among new conlangers at least, that because you can't make words without sounds, the phonology is seen as the foundation of the entire language and needs to be rock solid before going further - something that leads to overemphasis on this relatively small part of a language. Phonology is really just like any other part of the language, and it's very well possible to just sketch it out roughly at first and then go back and improve and adjust as ideas come. Besides, a lot of new 'phonologies' only include half of the things a real phonology should include, at best, which makes me wonder why it's seen as so important to complete something that often isn't complete anyway.

[Ossicone]

Just stick with.
Pick a phonology and don't deviate from it.
I had to force myself quite a bit to stick with the phoneme inventory of my latest conlang.

I quite disagree with this. Personally I think it's better to make a rough sketch of the phonology, just as for everything else. The fine tuning can always be made later.

I think this is a common perception, among new conlangers at least, that because you can't make words without sounds, the phonology is seen as the foundation of the entire language and needs to be rock solid before going further - something that leads to overemphasis on this relatively small part of a language. Phonology is really just like any other part of the language, and it's very well possible to just sketch it out roughly at first and then go back and improve and adjust as ideas come. Besides, a lot of new 'phonologies' only include half of the things a real phonology should include, at best, which makes me wonder why it's seen as so important to complete something that often isn't complete anyway.

I guess it's a personal thing. I like to have it set, or I try and fit too many ideas in at once. I guess it's also bad phrasing on my part -- I don't mean the entire phonology should be layed down from the beginning -- more just the phoneme selection.

For me, I just find that if I don't lay down a road map, I will deviate towards English/Spanish. The sounds help me get a feel for the rest.

[kadani]

Just stick with.
Pick a phonology and don't deviate from it.
I had to force myself quite a bit to stick with the phoneme inventory of my latest conlang.

I quite disagree with this. Personally I think it's better to make a rough sketch of the phonology, just as for everything else. The fine tuning can always be made later.

I think this is a common perception, among new conlangers at least, that because you can't make words without sounds, the phonology is seen as the foundation of the entire language and needs to be rock solid before going further - something that leads to overemphasis on this relatively small part of a language. Phonology is really just like any other part of the language, and it's very well possible to just sketch it out roughly at first and then go back and improve and adjust as ideas come. Besides, a lot of new 'phonologies' only include half of the things a real phonology should include, at best, which makes me wonder why it's seen as so important to complete something that often isn't complete anyway.

Maybe it could be expressed as: Don't invalidate constructions by later changes to the phonology. Rejistanian sounded quite different when I started it but I can still sing the KaMaRi song because nothing was actually invalidated and no permissible cluster has become unpronouncable.

[Sankon]

/p b t ʦ d k g q/ <p b t c d k g q> <п б т ц к г қ>
/m n ŋ/ <m n ñ> <м н ң>
/r/ <r> <р>
/f v s z ʃ ʒ χ ʁ h/ <f v s z ş-ç j x ǵ h> <ф в с з ш-щ ж х ғ Һ>
/l j/ <l y> <л ј>

/a i e y ɯ ɤ o u ə/ <a i e ü ı ö o u ə> <а і е ұ и ө о ү ә>

For aesthetic purposes, I didn't take the basic OU, but instead took the vertical one.

I would probably prefer <й> for /j/ than <j>. <J>, in my opinion, looks fugly with Cyrillic.

Why do you use <ұ> for /y/? It's only used in Kazakh, and then for /ʊ/, not /y/. <ӱ> would probably be a better choice.

Why don't you use <и> for /i/ and <ы> for /ɯ/? Though <ы> is used for /ɨ/, /ɨ/ is close enough to /ɯ/ that it would be better.

You forgot a romanization for /d/.

[sangi39]

/p b t ʦ d k g q/ <p b t c d k g q> <п б т ц к г қ>
/m n ŋ/ <m n ñ> <м н ң>
/r/ <r> <р>
/f v s z ʃ ʒ χ ʁ h/ <f v s z ş-ç j x ǵ h> <ф в с з ш-щ ж х ғ Һ>
/l j/ <l y> <л ј>

/a i e y ɯ ɤ o u ə/ <a i e ü ı ö o u ə> <а і е ұ и ө о ү ә>

For aesthetic purposes, I didn't take the basic OU, but instead took the vertical one.

I would probably prefer <й> for /j/ than <j>. <J>, in my opinion, looks fugly with Cyrillic.

Why do you use <ұ> for /y/? It's only used in Kazakh, and then for /ʊ/, not /y/. <ӱ> would probably be a better choice.

Why don't you use <и> for /i/ and <ы> for /ɯ/? Though <ы> is used for /ɨ/, /ɨ/ is close enough to /ɯ/ that it would be better.

You forgot a romanization for /d/.

I'm going to have to agree with Sankon here. On which graphemes are being used but I'm also a little uneasy about the vowel system. If the intent was a Turkic language then that /ɤ/ seems more likely to be /ø/ which seems to have been the intent given the orthography. If we follow Tuvan and Kyrgyz, which have that vowel system bar /ə/ then you could do:

/i y ɯ u/ <и ү ы у>
/e ø o/ <е ө o>
/a/ <a>

... and /a/ could then pattern as the back equivalent of /e/ as it does, IIRC, in Tuvan.

Alternatively, if we follow the pattern in Altai you could use <ӧ> and <ӱ> to represent /ø/ and /y/ respectively but in general it seems where these two sounds appear together they're either <ӧ> and <ӱ> or <ө> and <ү> but I could be mistaken.

I'd also have to say I prefer <й> as well. Use of <j> in Cyrillic seems quite rare, limited in general to Serbian and Macedonian when used to write /j/.

Nice use of the Uzbek alphabet for the uvular fricatives as well. I think you could take this pattern further and write /h/ using <ҳ> and given how you write /q/ you've already got a use for the descender as "indicates the sound is further back in the mouth than the sound represented by the unmodified grapheme".

Other than that I'd also point out the missing grapheme for /d/ but I guess that's just a missing letter :)

[Arzemju]

/p b t ʦ d k g q/ <p b t c d k g q> <п б т ц к г қ>
/m n ŋ/ <m n ñ> <м н ң>
/r/ <r> <р>
/f v s z ʃ ʒ χ ʁ h/ <f v s z ş-ç j x ǵ h> <ф в с з ш-щ ж х ғ Һ>
/l j/ <l y> <л ј>

/a i e y ɯ ɤ o u ə/ <a i e ü ı ö o u ə> <а і е ұ и ө о ү ә>

For aesthetic purposes, I didn't take the basic OU, but instead took the vertical one.

I would probably prefer <й> for /j/ than <j>. <J>, in my opinion, looks fugly with Cyrillic.

Why do you use <ұ> for /y/? It's only used in Kazakh, and then for /ʊ/, not /y/. <ӱ> would probably be a better choice.

Why don't you use <и> for /i/ and <ы> for /ɯ/? Though <ы> is used for /ɨ/, /ɨ/ is close enough to /ɯ/ that it would be better.

You forgot a romanization for /d/.

I'm going to have to agree with Sankon here. On which graphemes are being used but I'm also a little uneasy about the vowel system. If the intent was a Turkic language then that /ɤ/ seems more likely to be /ø/ which seems to have been the intent given the orthography. If we follow Tuvan and Kyrgyz, which have that vowel system bar /ə/ then you could do:

/i y ɯ u/ <и ү ы у>
/e ø o/ <е ө o>
/a/ <a>

... and /a/ could then pattern as the back equivalent of /e/ as it does, IIRC, in Tuvan.

Alternatively, if we follow the pattern in Altai you could use <ӧ> and <ӱ> to represent /ø/ and /y/ respectively but in general it seems where these two sounds appear together they're either <ӧ> and <ӱ> or <ө> and <ү> but I could be mistaken.

I'd also have to say I prefer <й> as well. Use of <j> in Cyrillic seems quite rare, limited in general to Serbian and Macedonian when used to write /j/.

Nice use of the Uzbek alphabet for the uvular fricatives as well. I think you could take this pattern further and write /h/ using <ҳ> and given how you write /q/ you've already got a use for the descender as "indicates the sound is further back in the mouth than the sound represented by the unmodified grapheme".

Other than that I'd also point out the missing grapheme for /d/ but I guess that's just a missing letter :)

I did forgot <d>, just an error ^^
As for the choices of vowels, I didn't want to take letters too much like russian, but I guess it's easier with them. So here is the revised script (with /w/ added):

/p b t ʦ d k g q/ <p b t c d k g q> <п б т ц д к г қ>
/m n ŋ/ <m n ñ> <м н ң>
/r/ <r> <р>
/f v s z ʃ ʒ χ ʁ h/ <f v s z ş-ç j x ǵ h> <ф в с з ш-щ ж х ғ ҳ>
/l j w/ <l ğ w> <л й ў>

/i y ɯ u/ <i ü ı u> <и ү ы у>
/e ø o/ <e ö o> <е ө o>
/a/ <a> <a>

And nice idea with the descenders, but I will keep the Uzbek barred <ғ> which looks better.
Should I also add other letters for phonemes present in Kazakh,Uzbek,Tajik,Mongolian,Turkish when i'll borrow words? Or I could also adapt them with the present script?

[kadani]

I did forgot <d>, just an error ^^
As for the choices of vowels, I didn't want to take letters too much like russian, but I guess it's easier with them. So here is the revised script (with /w/ added):

/p b t ʦ d k g q/ <p b t c d k g q> <п б т ц д к г қ>
/m n ŋ/ <m n ñ> <м н ң>
/r/ <r> <р>
/f v s z ʃ ʒ χ ʁ h/ <f v s z ş-ç j x ǵ h> <ф в с з ш-щ ж х ғ ҳ>
/l j w/ <l ğ w> <л й ў>

/i y ɯ u/ <i ü ı u> <и ү ы у>
/e ø o/ <e ö o> <е ө o>
/a/ <a> <a>

And nice idea with the descenders, but I will keep the Uzbek barred <ғ> which looks better.
Should I also add other letters for phonemes present in Kazakh,Uzbek,Tajik,Mongolian,Turkish when i'll borrow words? Or I could also adapt them with the present script?

I think you will most likely adapt phonemes for you conlang anyways... so if you keep a phonemic spelling you want to write the letters as your conculture speaks them. There is the idea of spelling words according to their origin like the English do. but this seems strange here. Diacritics, I can understand tp a certain point, but new letters seem much effort for your conpeople.

[sangi39]

And nice idea with the descenders, but I will keep the Uzbek barred <ғ> which looks better.

Oh, I didn't mean for you to do otherwise :)

Should I also add other letters for phonemes present in Kazakh,Uzbek,Tajik,Mongolian,Turkish when i'll borrow words? Or I could also adapt them with the present script?

This is a matter of personal choice methinks. For example, you could go as far as Japanese goes and simply represent foreign words within the native constraints of the language, e.g. "sport">"supootsu". Alternatively you could use native letters to represent foreign sounds but without making the distinction clear, e.g. native <г> could be used to write the native voiced velar plosive and then also be used to write a voiced uvular plosive as well at which point it becomes an orthographic irregularity. Further, you could use certain digraphs, e.g. the voiced uvular fricative could be represented as <гқ> or some similar combination. This may still have the ambiguity mentioned above but it spreads the possible readings out from one grapheme to a set of digraphs and trigraphs, limiting the possibilities. For example, Welsh <si> can be /si/ or /S/ but imagine if <s> could be /s/ or /S/. The possible reading /S/ would be placed on every occurrence of <s> rather than <si> increasing ambiguity.

And, like you said, you could always modify certain letters within the script using conventions already there. For example, the descender already has the vague implication I mentioned above. Using this, a descender could be added to <г> to indicate the voiced uvular plosive and likewise a descender could be added to <ғ> to indicate a voiced glottal fricative. OTOH you could simply borrow the native way the sound is written, which could bring the soft and hard signs into the orthography.

You could even mix the various methods up a bit. Say, for example, that <ь> ends up being borrowed when a word from some language is spelt exactly how they spell it but it marks the schwa, a sound already part of this language. From here you could use <ь> to indicate foreign sounds which are more centralised. For instance <иь> could end up representing /ɨ/ and <уь> could likewise represent /ʉ/.

[masako]

Seriously...don't obsess over one aspect of your conlang for too long...it kills everything else.

[Arzemju]

Seriously...don't obsess over one aspect of your conlang for too long...it kills everything else.

I know, that's what always kills my projects of conlang: the phoneme inventory...

Okay now I finished it, what should I go on now?
Any links on what I need for the conlang?

[sangi39]

Seriously...don't obsess over one aspect of your conlang for too long...it kills everything else.

I know, that's what always kills my projects of conlang: the phoneme inventory...

Okay now I finished it, what should I go on now?
Any links on what I need for the conlang?

Well, one starting point is the Wikipedia linguistics article, specifically this bit: http://en.wikipedia.org/wiki/Linguistic ... _divisions

I find Wikipedia quite good at the beginning just because if you have a starting point, simply following links to subsections means you can get really quite in-depth quite quickly. For example, going to "morphology" one of the of first major things you come across is "inflection" and "word-formation". In "inflection" one of the first major things is "declension" and "conjugation" and in "declension" you find "number", "case" and "gender". So going through the articles like this you can get at first a brief overview of a particular subject, e.g. morphology, syntax, semantics, etc. and then each subsequent article provides increasing amounts of information without going to fast.

If you're looking at doing something typically Turkic then there's always the Turkic article on Wikipedia. From here you can go to the pages of specific languages and through some of these you can find info on morphology, syntax, etc.

Note, though, that dsepite the "Wikipedia isn't reliable" idea, predominantly espoused by teachers and lecturers since it indicates a student's laziness and it can be constantly edited, it does actually have references for most statements in the articles meaning you can go out into the wider collection of resources from a single location.

[Arzemju]

I have redone the vowel harmony, having three kinds of vowels: front and back. Front and back vowels can't appear in the same word.

/i e y a/ <i e ü a>
/ʊ ɯ u ɔ/ <y ı u o>

Is it good?