Hello,
To create a proto-lang, I am starting with a (mostly) regular sketched out language, in that there is not much in the way of morphophonolgical interactions.
I was wondering what are some ways to create more complex systems? I realise some languages are quite regular (Turkish, Japanese), but even then it is not simply 'add suffix X' for everything.
Any ideas/examples/comments would be useful! I am especially interested to hear about how natlangs have become more/less regular (or complex in their morphology) . Thanks!
Introducing irregularity
Re: Introducing irregularity
Not an easy question.
ti -> si change in Finnish generated quite much irregularity.
*läkt- 'leave'
*läkt-e-n 'I leave'
*läkt-e-e 'He leaves'
*läkt-i-n 'I left'
*läkt-i 'He left'
1. ti -> si affects the past forms.
Läksin 'I left'
Läksi 'He left'
Clusters of two plosives are lenited kt -> ht
2. The present forms are affected
*lähten 'I leave'
lähtee 'he leaves'
Plosives are lenited when they start a closed syllable. So called consonant gradation.
3. Consonant gradation affects sg1 present, but not the ks cluster of past forms.
lähden 'I leave'
So now we have three stems:
Lähd-e-n
Läht-e-e
Läks-i-n
Läks-i
New analogical past forms are created and are in free variation with the old ones.
lähd-i-n
Läht-i
That actually creates one stem more.
So, basically irregularity is gained by accumulating sound changes.
ti -> si change in Finnish generated quite much irregularity.
*läkt- 'leave'
*läkt-e-n 'I leave'
*läkt-e-e 'He leaves'
*läkt-i-n 'I left'
*läkt-i 'He left'
1. ti -> si affects the past forms.
Läksin 'I left'
Läksi 'He left'
Clusters of two plosives are lenited kt -> ht
2. The present forms are affected
*lähten 'I leave'
lähtee 'he leaves'
Plosives are lenited when they start a closed syllable. So called consonant gradation.
3. Consonant gradation affects sg1 present, but not the ks cluster of past forms.
lähden 'I leave'
So now we have three stems:
Lähd-e-n
Läht-e-e
Läks-i-n
Läks-i
New analogical past forms are created and are in free variation with the old ones.
lähd-i-n
Läht-i
That actually creates one stem more.
So, basically irregularity is gained by accumulating sound changes.
My meta-thread: viewtopic.php?f=6&t=5760
Re: Introducing irregularity
I'd generally agree with Omzinesý on that. One of the easiest ways to introduce morphological irregularity certainly is sound change, and even analogy (as discussed over in the L&N Q&A thread, some Spanish verbs irregularly insert a /g/ in various areas of their paradigms through analogy with certain verbs which now, through anology, no longer have /g/ in that position).
This is especially true when you get one or both of a merger or a split occurring. Taking a Finnish-like example, to build on Omzinesý's points, where there are the following "proto"-nouns *lakti, *kakte, *maksi and *vakse which take the suffix -n in the accusative singular and null suffix in the nominative.
To begin with, final *e shifts to /i/, yielding:
lakti : laktin
kakti : kakten
maksi : maksin
vaksi : vaksen
Then t becomes s before i:
laksi : laksin
kaksi : kakten
maksi : maksin
vaksi : vaksen
Next k becomes h before a plosive:
laksi : laksin
kaksi : kahten
maksi : maksin
vaksi : vaksen
So you have four verbs which all have the form -ksi in the nominative singular, but which have three different forms in the accusative without any way of predicting which form a given noun ending in -ksi might take.
This is especially true when you get one or both of a merger or a split occurring. Taking a Finnish-like example, to build on Omzinesý's points, where there are the following "proto"-nouns *lakti, *kakte, *maksi and *vakse which take the suffix -n in the accusative singular and null suffix in the nominative.
To begin with, final *e shifts to /i/, yielding:
lakti : laktin
kakti : kakten
maksi : maksin
vaksi : vaksen
Then t becomes s before i:
laksi : laksin
kaksi : kakten
maksi : maksin
vaksi : vaksen
Next k becomes h before a plosive:
laksi : laksin
kaksi : kahten
maksi : maksin
vaksi : vaksen
So you have four verbs which all have the form -ksi in the nominative singular, but which have three different forms in the accusative without any way of predicting which form a given noun ending in -ksi might take.
You can tell the same lie a thousand times,
But it never gets any more true,
So close your eyes once more and once more believe
That they all still believe in you.
Just one time.
But it never gets any more true,
So close your eyes once more and once more believe
That they all still believe in you.
Just one time.
Re: Introducing irregularity
Speaking of Japanese, one of the major sources of irregularity in Japanese is contractions that are non-universal sound changes. /k/ was deleted before /i/ in some verb and adjective forms and nowhere else, creating irregularity:
Classical Japanese: omoki, omoku / kaku, kakite
Modern Japanese: omoi, omoku / kaku, kaite
Another thing that happened was there was some sort of suffix in Old Japanese that was present on independent nouns but not when they were compounded, and this remains in the modern language in a number of compounds, for example:
火 hi - 炎 honō
木 ki - 木の実 konomi
目 me - 瞼 mabuta
上 ue - 上着 uwagi
白 shiro - 白髪 shiraga
This can be even more obscured by other contractions/changes e.g.:
神 kami + 戸 he → 神戸 Kamube → Kaube → Kōbe (the city).
Classical Japanese: omoki, omoku / kaku, kakite
Modern Japanese: omoi, omoku / kaku, kaite
Another thing that happened was there was some sort of suffix in Old Japanese that was present on independent nouns but not when they were compounded, and this remains in the modern language in a number of compounds, for example:
火 hi - 炎 honō
木 ki - 木の実 konomi
目 me - 瞼 mabuta
上 ue - 上着 uwagi
白 shiro - 白髪 shiraga
This can be even more obscured by other contractions/changes e.g.:
神 kami + 戸 he → 神戸 Kamube → Kaube → Kōbe (the city).
Re: Introducing irregularity
On the topic of analogy—when does analogy start to be used? When sound change destroys existing paradigms too much? How much irregularity can a language handle before it starts to invoke analogy? Or is it just one of those 'up to you' choices?sangi39 wrote:I'd generally agree with Omzinesý on that. One of the easiest ways to introduce morphological irregularity certainly is sound change, and even analogy (as discussed over in the L&N Q&A thread, some Spanish verbs irregularly insert a /g/ in various areas of their paradigms through analogy with certain verbs which now, through anology, no longer have /g/ in that position).
This is especially true when you get one or both of a merger or a split occurring. Taking a Finnish-like example, to build on Omzinesý's points, where there are the following "proto"-nouns *lakti, *kakte, *maksi and *vakse which take the suffix -n in the accusative singular and null suffix in the nominative.
Re: Introducing irregularity
It seems to be sort of "up to you", I think. Some languages maintain previous levels of irregularity while others level it by means of analogy. Analogy basically seems to take hold when speakers spot a pattern and apply it to a word that previously did not fit that pattern. What exactly that pattern is can be fairly hard to predict.Davush wrote:On the topic of analogy—when does analogy start to be used? When sound change destroys existing paradigms too much? How much irregularity can a language handle before it starts to invoke analogy? Or is it just one of those 'up to you' choices?sangi39 wrote:I'd generally agree with Omzinesý on that. One of the easiest ways to introduce morphological irregularity certainly is sound change, and even analogy (as discussed over in the L&N Q&A thread, some Spanish verbs irregularly insert a /g/ in various areas of their paradigms through analogy with certain verbs which now, through anology, no longer have /g/ in that position).
This is especially true when you get one or both of a merger or a split occurring. Taking a Finnish-like example, to build on Omzinesý's points, where there are the following "proto"-nouns *lakti, *kakte, *maksi and *vakse which take the suffix -n in the accusative singular and null suffix in the nominative.
It's also worth pointing out that analogy can sometimes increase the number of words which don't follow more widespread patterns, e.g. dove in some dialects of English instead of dived by means of analogy with drove (dive has been a weak verb as far back as Old English). So sometimes analogy can work to increase the number of exceptions to a rule rather than heading solely towards a single pattern (in this case, rather than shifted drove to drived to fall in line with the majority of other verbs, the similar sounding root dive took a past tense form which is much less common).
You can tell the same lie a thousand times,
But it never gets any more true,
So close your eyes once more and once more believe
That they all still believe in you.
Just one time.
But it never gets any more true,
So close your eyes once more and once more believe
That they all still believe in you.
Just one time.