The  Subject Cataloging Manual: Subject Headings gives guidance  in  this  matter  in  the instruction sheet  H2190.
It indicates  that  "--Texts"  may  be  assigned  under  "individual lesser-known  (rare   or   exotic)  languages  and  dialects"
to collections of  texts  or  to  individual  works  for  which  the cataloger cannot figure  out  the subject because of inability to
read the language or get a translation of it.
 
This instruction sheet further states  that "--Texts" should  not  be used under languages that are no longer spoken but for which a large  body  of  written  material  exists,  such  as  the  Latin language. "However the subdivision '--Dialects--Texts' may  be used under  any language for collections of  texts  in  such dialects, e.g., English language--Dialects--Texts."
 
Another relevant instruction in that same sheet directs that  one assign a heading (Ethnic group)--Languages--Texts or
(Place)--Languages--Texts only  for "texts in languages spoken by the group for which no separate language heading exists or can be established, e.g. Jews--Languages--Texts."
 
As far as our practice is concerned, a view of  our MARC database shows  a  very  limited usage  of "--Texts"  under uncommon Jewish languages.   "--Texts"  has  never  been  used  under "Judeo-Arabic language" and  has  been used only  one  time each under the  two headings "Ladino language"  and "Judeo-Tajik language." In  fact, apart from these two instances, I  have not found for works which clearly consist of  a text or collection of texts in  an uncommon Jewish language any records which use "--Texts" after the name of the language.

One reason for this lack of such headings may  be that there have been specific headings that  were available and suitable for  the works cataloged, such as: "Jewish religious poetry, Judeo-Arabic," "Judeo-Persian literature," "Judeo-Arabic  poetry"
or "Judeo-Catalan poetry."
 
It seems also worth noting that  some  of  the tables in  the  PJ classification schedule make provision for texts  in lesser known
Jewish languages by assigning a special cutter or number to them.
 
In summary, our usage is too limited to show a trend, but it does appear   that   our   rules   allow   the   use   of   "(Uncommon
language)--Texts" as  a  kind  of  last resort,  when  one  can't determine the content of a text or when the work is  a collection
of  texts (which would presumably be  too difficult to examine in sufficient detail to determine more precise topical headings).
 
Peter Kearney
LC Hebraica Team