How to use Russian morphology in Oracle Text (in DBMS 11g R2)?

Question

Oracle

Hello colleagues.
The official documentation for Oracle Text 11 g Release 2 (11.2) ( link ) says that Russian morphology is supported at the word stem level (stemmer). Tried this in practice:
Table:

select * from docs;
 
        ID TEXT
---------- --------------------
         1 читать
         2 читаю
         3 читал
         4 чтение
         5 sing
         6 sang
         7 singing
         8 sung

Create a lexer:

exec ctx_ddl.create_preference('MYLEXER', 'world_lexer');

Create an index:

create index i_docs on docs (text) indextype is ctxsys.context
   parameters ('LEXER MYLEXER stoplist CTXSYS.EMPTY_STOPLIST');

The following query for English text produces the correct result (4 entries):

SELECT SCORE ( 1 ), text
  FROM docs
  WHERE CONTAINS (text, '$sing', 1 ) > 0
  ORDER BY SCORE ( 1 ) DESC;

And a request for Russian text produces only one entry:

SELECT SCORE ( 1 ), text
  FROM docs
  WHERE CONTAINS (text, '$читать', 1 ) > 0
  ORDER BY SCORE ( 1 ) DESC;

At the sql.ru forum I read that the original encoding of the instance itself is probably to blame. My parameters:

NLS_CHARACTERSET     CL8ISO8859P5
NLS_NCHAR_CHARACTERSET    AL16UTF16

Tried on the 12th version of the DBMS (on UTF8 in 12.1.0.2), the result is the same.
Has anyone faced a similar issue? Did you manage to connect Russian morphology?