path: root/doc/ParadigmsRus.tex

This is an API for the user of the resource grammar for adding
lexical items. It gives functions for forming expressions of open
categories: nouns, adjectives, verbs.

Closed categories (determiners, pronouns, conjunctions) are
accessed through the resource syntax API, {\tt Structural.gf}.

The main difference with {\tt MorphoRus.gf} is that the types
referred to are compiled resource grammar types. We have moreover
had the design principle of always having existing forms, rather
than stems, as string arguments of the paradigms.

The structure of functions for each word class {\tt C} is the following:
first we give a handful of patterns that aim to cover all
regular cases. Then we give a worst-case function {\tt mkC}, which serves as an
escape to construct the most irregular words of type {\tt C}.
%However, this function should only seldom be needed: we have a
%separate module {\tt IrregularEng}, which covers all irregularly inflected
words.

The following modules are presupposed:
\begin{verbatim}
  resource ParadigmsRus = open
    (Predef=Predef),
    Prelude,
    MorphoRus,
    CatRus,
    NounRus
    in {

  flags  coding=utf8 ;
\end{verbatim}

\textbf{Parameters}

To abstract over gender names, we define the following identifiers.
\begin{verbatim}
  oper
    Gender : Type ;

    masculine : Gender ;
    feminine  : Gender ;
    neuter    : Gender ;
\end{verbatim}

To abstract over case names, we define the following.
\begin{verbatim}
    Case : Type ;

    nominative    : Case ;
    genitive      : Case ;
    dative        : Case ;
    accusative    : Case ;
    instructive   : Case ;
    prepositional : Case ;
\end{verbatim}

In some (written in English) textbooks accusative case
is put on the second place. However, we follow the case order
standard for Russian textbooks.
To abstract over number names, we define the following.
\begin{verbatim}
    Number : Type ;

    singular : Number ;
    plural   : Number ;

    Animacy: Type ;

    animate: Animacy;
    inanimate: Animacy;
\end{verbatim}

\textbf{Nouns}
Best case: indeclinabe nouns: {\cyr kofe}, {\cyr
pal\cyrsftsn{}to}, {\cyr VUZ}.
\begin{verbatim}
     mkIndeclinableNoun: Str -> Gender -> Animacy -> N ;
\end{verbatim}

Worst case - give six singular forms:
Nominative, Genetive, Dative, Accusative, Instructive and Prepositional;
corresponding six plural forms and the gender.
May be the number of forms needed can be reduced,
but this requires a separate investigation.
Animacy parameter (determining whether the Accusative form is equal
to the Nominative or the Genetive one) is actually of no help,
since there are a lot of exceptions and the gain is just one form less.
\begin{verbatim}
     mkN  : (nomSg, genSg, datSg, accSg, instSg, preposSg,
             nomPl, genPl, datPl, accPl, instPl, preposPl: Str)
             -> Gender -> Animacy -> N ;
\end{verbatim}
({\cyr \em muzhchina, muzhchinu, muzhchine, muzhchinu,
muzhchino\cyrishrt{}, muzhchine}

\noindent {\cyr \em muzhchin\cyrery{}, muzhchin, muzhchinam,
muzhchin, muzhchinami, muzhchinah})

\vspace{5mm}

The regular function captures the variants for some popular nouns
endings from the list below:
\begin{verbatim}
     regN             : Str -> N ;
\end{verbatim}

Here are some common patterns. The list is far from complete.

\subsubsection{Feminine patterns}

\noindent feminine, inanimate, ending with "-a", Inst -"{\cyr
mashin-o\cyrishrt{}}":
\begin{verbatim}
     nMashina         : Str -> N ;
\end{verbatim}
feminine, inanimate, ending with "-a", Inst -"{\cyr
edinic-e\cyrishrt{}}":
\begin{verbatim}
     nEdinica         : Str -> N ;
\end{verbatim}
feminine, animate, ending with "-a":
\begin{verbatim}
     nZhenchina       : Str -> N ;
\end{verbatim}
feminine, inanimate, ending with "{\cyr g\_k\_h-a}":
\begin{verbatim}
     nNoga            : Str -> N ;
\end{verbatim}
feminine, inanimate, ending with "-{\cyr -iya}":
\begin{verbatim}
     nMalyariya       : Str -> N ;
\end{verbatim}
feminine, animate, ending with "{\cyr -ya}":
\begin{verbatim}
     nTetya           : Str -> N ;
\end{verbatim}
feminine, inanimate, ending with "-{\cyr \cyrsftsn{}}"(soft sign):
\begin{verbatim}
     nBol             : Str -> N ;
\end{verbatim}

\subsubsection{Neuter patterns}

\noindent neutral, inanimate, ending with "-ee":
\begin{verbatim}
     nObezbolivauchee : Str -> N ;
\end{verbatim}
neutral, inanimate, ending with "-e":
\begin{verbatim}
     nProizvedenie    : Str -> N ;
\end{verbatim}
neutral, inanimate, ending with "-o":
\begin{verbatim}
     nChislo          : Str -> N ;
\end{verbatim}
neutral, inanimate, ending with "-{\cyr oe}":
\begin{verbatim}
     nZhivotnoe       : Str -> N ;
\end{verbatim}

\subsubsection{Masculine patterns}

\noindent Ending with consonant:

\noindent masculine, inanimate, ending with "-{\cyr el}"- "{\cyr
pep-la}":
\begin{verbatim}
     nPepel           : Str -> N ;
\end{verbatim}
animate, "{\cyr brat-\cyrsftsn{}ya}":
\begin{verbatim}
     nBrat            : Str -> N ;
\end{verbatim}
same as above, but inanimate:
\begin{verbatim}
     nStul            : Str -> N ;
\end{verbatim}
"{\cyr malyshe\cyrishrt{}}":
\begin{verbatim}
     nMalush          : Str -> N ;
\end{verbatim}
"{\cyr potol-ok, potol-ka}"
\begin{verbatim}
     nPotolok         : Str -> N ;
\end{verbatim}
the next four differ in plural nominative and/or accusative
form(s):

\noindent {\cyr bank-i}(Nom=Acc):
\begin{verbatim}
     nBank            : Str -> N ;
\end{verbatim}
same as above, but animate:
\begin{verbatim}
     nStomatolog      : Str -> N ;
\end{verbatim}
"{\cyr adres-a}" (Nom=Acc):
\begin{verbatim}
     nAdres           : Str -> N ;
\end{verbatim}
"{\cyr telefony}" (Nom=Acc):
\begin{verbatim}
     nTelefon         : Str -> N ;
\end{verbatim}
masculine, inanimate, ending with "{\cyr \cyrsftsn{}}" (soft
sign):
\begin{verbatim}
     nNol             : Str -> N ;
\end{verbatim}
masculine, inanimate, ending with "{\cyr -en\cyrsftsn{}}":
\begin{verbatim}
     nUroven          : Str -> N ;
\end{verbatim}

Nouns used as functions need a preposition. The most common is with Genitive.
\begin{verbatim}
     mkFun            : N -> Prep -> N2 ;
     mkN2             : N -> N2 ;
     mkN3             : N -> Prep -> Prep -> N3 ;
\end{verbatim}

\subsubsection{Proper names}

{\cyr Ivan, Masha}:
\begin{verbatim}
     mkPN             : Str -> Gender -> Animacy -> PN ;
\end{verbatim}
\begin{verbatim}
     nounPN           : N -> PN ;
\end{verbatim}

On the top level, it is maybe {\tt CN} that is used rather than {\tt N}, and
{\tt NP} rather than {\tt PN}.
\begin{verbatim}
     mkCN             : N -> CN ;
     mkNP             : Str -> Gender -> Animacy -> NP ;
\end{verbatim}

\textbf{Adjectives}
Non-comparison (only positive degree) one-place adjectives need 28
(4 by 7) forms in the worst case: (Masculine  | Feminine | Neutral
| Plural) * (Nominative | Genitive | Dative | Accusative Inanimate
| Accusative Animate | Instructive | Prepositional). Notice that 4
short forms, which exist for some adjectives are not included in
the current description, otherwise there would be 32 forms for
positive degree.

The regular function captures the variants for some popular
adjective endings below. The first string agrument is the
masculine singular form, the second is comparative:
\begin{verbatim}
     regA             : Str -> Str -> A ;
\end{verbatim}

\noindent Invariable adjective is a special case: {\cyr haki,
mini, hindi, netto}:
\begin{verbatim}
     adjInvar         : Str -> A ;
\end{verbatim}

Some regular patterns depending on the ending.

\noindent ending with "{\cyr y\cyrishrt{}}":
\begin{verbatim}
     AStaruyj         : Str -> Str -> A ;
\end{verbatim}
ending with "{\cyr i\cyrishrt{}}", Gen - "{\cyr
malen\cyrsftsn{}k-ogo}":
\begin{verbatim}
     AMalenkij        : Str -> Str -> A ;
\end{verbatim}
ending with "{\cyr i\cyrishrt{}}", Gen - "{\cyr horosh-ego}":
\begin{verbatim}
     AKhoroshij       : Str -> Str -> A ;
\end{verbatim}
ending with "{\cyr o\cyrishrt{}}", plural - "{\cyr molod-ye}":
\begin{verbatim}
     AMolodoj         : Str -> Str -> A ;
\end{verbatim}
ending with "{\cyr o\cyrishrt{}}", plural - "{\cyr kak-ie}":
\begin{verbatim}
     AKakoj_Nibud     : Str -> Str -> Str -> A ;
\end{verbatim}

Two-place adjectives need a preposition and a case as extra arguments.

"{\cyr delim na}":
\begin{verbatim}
     mkA2             : A -> Str -> Case -> A2 ;
\end{verbatim}

Comparison adjectives need a positive adjective (28 forms without
short forms). Taking only one comparative form (non-syntactic) and
only one superlative form (syntactic) we can produce the
comparison adjective with only one extra argument - non-syntactic
comparative form. Syntactic forms are based on the positive forms.

\begin{verbatim}
     mkADeg           : A -> Str -> ADeg ;
\end{verbatim}
On top level, there are adjectival phrases. The most common case
is just to use a one-place adjective.
\begin{verbatim}
     ap               : A  -> IsPostfixAdj -> AP ;
\end{verbatim}

\textbf{Adverbs}
Adverbs are not inflected. %Most lexical ones have position after the verb. Some can be preverbal (e.g. {\it always}).
\begin{verbatim}
     mkAdv            : Str -> Adv ;
\end{verbatim}

\textbf{Verbs}

In our lexicon description ({\it Verbum}) there are 62 forms: 2
(Voice) by { 1 (infinitive) + [2(number) by 3
(person)](imperative) + [ [2(Number) by 3(Person)](present) +
[2(Number) by 3(Person)](future) + 4(GenNum)(past) ](indicative)+
4 (GenNum) (subjunctive) } Participles (Present and Past) and
Gerund forms are not included, since they fuction more like
Adjectives and Adverbs correspondingly rather than verbs. Aspect
is regarded as an inherent parameter of a verb. Notice, that some
forms are never used for some verbs. %Actually, the majority of verbs do not have many of the forms.
\begin{verbatim}
  Voice: Type;
  Aspect: Type;
  Tense : Type;
  Bool: Type;
  Conjugation: Type ;
\end{verbatim}
"{\cyr gulya-Esh\cyrsftsn{}, gulya-Em}":
\begin{verbatim}
  first: Conjugation;
\end{verbatim}

\noindent Verbs with vowel "{\cyr \cyryo}": "{\cyr
da\cyryo{}sh\cyrsftsn{}}" (give), "{\cyr
p\cyrsftsn{}\cyryo{}sh\cyrsftsn{}}" (drink):
\begin{verbatim}
  firstE: Conjugation;
\end{verbatim}

\noindent "{\cyr vid-Ish\cyrsftsn{}, vid-Im}":
\begin{verbatim}
  second: Conjugation;
\end{verbatim}
"{\cyr hoch-Esh\cyrsftsn{}, hot-Im}":
\begin{verbatim}
  mixed: Conjugation;
\end{verbatim}
irregular:
\begin{verbatim}
  dolzhen: Conjugation;

  true: Bool;
  false: Bool;

  active: Voice ;
  passive: Voice ;
  imperfective: Aspect;
  perfective: Aspect ;
\end{verbatim}

The worst case need 6 forms of the present tense in indicative
mood ({\cyr ya begu}, {\cyr ty bezhish\cyrsftsn{}}, {\cyr on
bezhit}, {\cyr my bezhim}, {\cyr vy bezhite}, {\cyr oni begut}), a
past form (singular, masculine: {\cyr ya bezhal}), an imperative
form (singular, second person: {\cyr begi}), an infinitive ({\cyr
bezhat\cyrsftsn{}}). Inherent aspect should also be specified.
\begin{verbatim}
     mkVerbum : Aspect -> (presentSgP1,presentSgP2,presentSgP3,
                           presentPlP1,presentPlP2,presentPlP3,
           pastSgMasculine,imperative,infinitive: Str) -> V ;
\end{verbatim}

Common conjugation patterns are two conjugations: first - verbs
ending with {\cyr -at\cyrsftsn{}/-yat\cyrsftsn{}} and second -
{\cyr -it\cyrsftsn{}/-et\cyrsftsn{}}. Instead of 6 present forms
of the worst case, we only need a present stem and one ending
(singular, first person): {\cyr ya l\cyryu{}bl\cyryu{}}, {\cyr ya
zhdu}, etc. To determine where the border between stem and ending
lies it is sufficient to compare first person from with second
person form: {\cyr ya l\cyryu{}bl\cyryu{}}, {\cyr ty
l\cyryu{}bish\cyrsftsn{}}. Stems should be the same. So the
definition for verb {\cyr l\cyryu{}bit\cyrsftsn{}} looks like:
\texttt{regV Imperfective Second }"{\cyr l\cyryu{}b}" "{\cyr
l\cyryu{}}" "{\cyr l\cyryu{}bil}" "{\cyr l\cyryu{}bi}" "{\cyr
l\cyryu{}bit\cyrsftsn{}}";
\begin{verbatim}
     regV : Aspect -> Conjugation -> (stemPresentSgP1,
       endingPresentSgP1,pastSgP1,imperative,infinitive: Str) -> V ;
\end{verbatim}

For writing an application grammar one usually doesn't need the
whole inflection table, since each verb is used in a particular
context that determines some of the parameters (Tense and Voice
while Aspect is fixed from the beginning) for certain usage. The
{\it V} type, that have these parameters fixed. We can extract the
{\it V} from the lexicon.
\begin{verbatim}
     mkV              : Verbum -> Voice ->  V ;
     mkPresentV       : Verbum -> Voice -> V ;
\end{verbatim}

Two-place verbs, and the special case with direct object. Notice
that a particle can be included in a {\tt V}.

\noindent "{\cyr vo\cyrishrt{}ti v dom}", "{\cyr v}", accusative:
\begin{verbatim}
     mkV2             : V   -> Str -> Case -> V2 ;
\end{verbatim}
{\cyr slozhit\cyrsftsn{} pic\cyrsftsn{}mo v konvert}:
\begin{verbatim}
     mkV3             : V -> Str -> Str -> Case -> Case -> V3 ;
\end{verbatim}
"{\cyr videt\cyrsftsn{}}", "{\cyr l\cyryu{}bit\cyrsftsn{}}":
\begin{verbatim}
     dirV2            : V -> V2 ;
     tvDirDir         : V -> V3 ;
\end{verbatim}

The definitions should not bother the user of the API. So they are
hidden from the document.