Name mangling (Was: Re: First Sound Recording of Asha'ille!)

On Mon, Mar 07, 2005 at 07:59:44PM +0100, Henrik Theiling wrote:
> Hi!
>
> Benct Philip Jonsson <bpj@...> writes:
[...]
> > I figured that, but any word adopted into Sohlob would still have to
> > be adapted to its phonology, restrictive phonotactics and vowel
> > harmony, and to the rule that no word may end in a vowel (tho they
> > may end in /j/, /w/ or /h/, which latter has a [?] allophone
> > phrase-finally.  Arsaey Aengosaey [,&rs&j ,&NQ's&j] is actually the
> > closest match.
>
> There are languages that obfuscate names badly by this, of course.  I
> just love Chinese for it's system...
Yes. It's both phonological and conventional mangling. (Convention in
the sense that proper names prefer to be trisyllabic, so random
syllables are added/dropped to/from the original.) The result can be
completely unrecognizable. :-)


> One thing that's typical for my langs is that labials are missing and
> that /j/, as a palatal, is not always there.  Clusters may be very
> restricted, too.
>
> Just to play the game, examples from Tyl Sjok:
>    - 'Henrik': [hEn=Xik=],
>    - 'Björn':  [djEl3n=] oder [gjEl3n=],
>    - 'Arthaey Angosii': [?al=sE ?ENgVsi] -- quite close, I think. :-)
Ebisédian has rather pathological rules about names. All proper names
are always prefixed with a proper name marker, inflected for gender. A
strict CVCV...CV(C) structure is enforced. Masculine names prefer
earlier stress and feminine names prefer ultima stress. All bets are
off on vowels, esp. in any case other than the locative, due to
ablaut-like case inflection. So:

Henrik --> ehe'n3rik [?E'hEn@\r`ik] (locative)
           ehi'n3r0k [?E'hin@\r`Ak] (originative)
           eha'n3r3k [?E'han@\r`@\k] (conveyant)

Björn --> ej0'rin [?E'dZAr`in] (locative)
          ejo'r0n [?E'dZor`An] (originative)
          eja'r3n [?E'dZar`@\n] (conveyant)

Arthaey -> `yarathei' [Hy?ar`aTE'?i] (locative)
           `y3r3th30' [Hy?@\r`@\T@\'?A] (originative)
           `yaratha3' [Hy?ar`aTa'?@\] (conveyant)

(I won't even try to transliterate "Angosii".)


[...]
> Qthyn|gai is worse, as words would get the normal word prefix to them
> including class etc.  Further, they'd get the stem 'name' to indicate
> a name.  Apart from that, only three vowels are available [a i M],
> only few clusters exist (strange ones), so the basic structure is CV,
> labials are missing, too, and no palatals, no l/r (only [tK] and [qK]).
> We get:
>
>    - 'Henrik':          [hAnMXikM]       Well...
>    - 'Björn':           [gIhMn]          This is close to impossible.
>    - 'Arthaey Angosii': [hasai haNgMsi]  Well...
[...]

Nice. This sounds almost as bad as Ebisédian. :-P

Tatari Faran is much better. Although it also has rather limited
syllabic structure available, its phonological rules are much more
conducive to transliteration:

Henrik -> henri' [hEn.4i?] (isn't that much better? :-P)
Bjorn  -> joron ['dzO.4On]
Arthaey Angosii -> aratei ankosi [?a4a'tej an.'kOsi]

However, one ought not to be fooled by this lucky sample set... the
phonemic equivalence of [4] and [d] as well as the lack of [l] can do
some rather infelicitous things (to quote Sally) to such names as:

Rachel -> datsere [da.'tsE.4E]
Robert -> doberet [dO.bE.'4Et]
Roland -> doran ['dO.4an]
Rudy -> duri ['du.4i] (you can see how [4] and [d] swap here)
Lisa -> disa ['di.sa]

(Besides these, if your name happens to be Karen, try your *best* to
get another, native, name instead when visiting Fara. In TF, _karen_
[ka'4En] means "shoe", which is probably not what you want for a
name!)


T

--
MAS = Mana Ada Sistem?

Replies