Phonological assimilation

[mashers]

Am I to assume that phonological assimilation should be avoided when speaking Lojban aloud?

[selpa'i]

Am 15.10.2012 22:31, schrieb mashers:
> Am I to assume that phonological assimilation should be avoided when
> speaking Lojban aloud?

Assimiliation doesn't really happen much in Lojban. You can assimilate
[n] before velar -> [N], but [n] is just an allophone of /n/,
independent of position. In general, Lojban's letters get translated
into phones directly, but each phoneme has a set of allophones which are
each independent on position. As long as you realize a phoneme as any of
those sounds, you're fine.
What kind of assimilation do you have in mind?

mu'o mi'e la selpa'i

--
doị mèlbi mlenì'u
.i do càtlu ki'u
ma fe la xàmpre ŭu
.i do tìnsa càrmi
gi'e sìrji se tàrmi
.i taị bo pu cìtka lo gràna ku


.

[Pierre Abbat]

On Monday, October 15, 2012 13:31:16 mashers wrote:
> Am I to assume that phonological assimilation should be avoided when
> speaking Lojban aloud?

Depends on what you're assimilating. Concerning the nasals, /n/ may be
assimilated to [ŋ] before /g,k,x/, but may not be assimilated to [ɱ] before
/f,v/. Contrariwise /m/ may be assimilated before /f,v/, but not before
/g,k,x/. (I'm not sure, without looking at it and trying to decode some
krokozyabry, whether the refgram mentions [ɱ].)
fonxa
banfi
kumfa
simxu

It's different in Spanish, where "anfibio" is regularly pronounced with [ɱ]; the
"nf", which explains the "nf" in "banfi", is just a spelling rule, and there's
only one distinct nasal preceding a stop or fricative or at the end of a word
(not counting foreign words). And it's different still in English, which
distinguishes three nasals at the end of a word. One must not be too hasty to
be discharged from a hospital, for one would be an impatient inpatient.

There are also several allophones of /r/, which I use in different
environments. This isn't really assimilation, as there's nothing tap-like
about the two vowels in "flira" or roll-like about the "o" or "s" in "morsi".

Assimilation of voicing between stops and fricatives does not occur in lujvo;
instead, one sticks the interfix /ə/ (<y>) between them. If such a pair comes
up in a fu'ivla, you can't stick an interfix, but you can assimilate a
consonant. The consonant is then written as assimilated. For instance, one
can't say "tcadrvisbadne", but one can say "tcadrvispadne" or "tcadrvizbadne".

Pierre
--
When a barnacle settles down, its brain disintegrates.
Já não percebe nada, já não percebe nada.

[mashers]

On Monday, 15 October 2012 21:39:16 UTC+1, selpa'i wrote:

Am 15.10.2012 22:31, schrieb mashers:
> Am I to assume that phonological assimilation should be avoided when
> speaking Lojban aloud?

Assimiliation doesn't really happen much in Lojban. You can assimilate
[n] before velar -> [N], but [n] is just an allophone of /n/,
independent of position. In general, Lojban's letters get translated
into phones directly, but each phoneme has a set of allophones which are
each independent on position. As long as you realize a phoneme as any of
those sounds, you're fine.
What kind of assimilation do you have in mind?

Thanks for clarifying. I was actually thinking of [n] before velar, such as in "gunku". I didn't realise there were any allophones in Lojban. This surprises me.

[mashers]

On Tuesday, 16 October 2012 05:24:08 UTC+1, Pierre Abbat wrote:

On Monday, October 15, 2012 13:31:16 mashers wrote:
> Am I to assume that phonological assimilation should be avoided when
> speaking Lojban aloud?

Depends on what you're assimilating. Concerning the nasals, /n/ may be
assimilated to [ŋ] before /g,k,x/, but may not be assimilated to [ɱ] before
/f,v/. Contrariwise /m/ may be assimilated before /f,v/, but not before
/g,k,x/.

It sounds like allophones are allowed, but one per phoneme. So [n] -> /ŋ/ is allowed, and [m] -> /ɱ/ is allowed, but [n] -> /ɱ/ is not allowed because /ɱ/ and /ŋ/ are effectively contrastive. Is that correct?

 

One must not be too hasty to
be discharged from a hospital, for one would be an impatient inpatient.

Great example :) 


Thanks for all the other points you raised - I'm going to archive your post in my Lojban notes for future reference.