Files
gf-core/lib/resource/doc/gfdoc/ParadigmsFre.html
2007-12-12 20:30:11 +00:00

415 lines
11 KiB
HTML

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML>
<HEAD>
<META NAME="generator" CONTENT="http://txt2tags.sf.net">
<TITLE> French Lexical Paradigms</TITLE>
</HEAD><BODY BGCOLOR="white" TEXT="black">
<P ALIGN="center"><CENTER><H1> French Lexical Paradigms</H1>
<FONT SIZE="4">
<I>Last update: 2007-07-06 09:17:50 CEST</I><BR>
</FONT></CENTER>
<P></P>
<HR NOSHADE SIZE=1>
<P></P>
<UL>
<LI><A HREF="#toc1">Parameters</A>
<LI><A HREF="#toc2">Nouns</A>
<UL>
<LI><A HREF="#toc3">Compound nouns</A>
<LI><A HREF="#toc4">Relational nouns</A>
<LI><A HREF="#toc5">Relational common noun phrases</A>
<LI><A HREF="#toc6">Proper names and noun phrases</A>
</UL>
<LI><A HREF="#toc7">Adjectives</A>
<UL>
<LI><A HREF="#toc8">Two-place adjectives</A>
</UL>
<LI><A HREF="#toc9">Adverbs</A>
<LI><A HREF="#toc10">Verbs</A>
<UL>
<LI><A HREF="#toc11">Two-place verbs</A>
<LI><A HREF="#toc12">Three-place verbs</A>
<LI><A HREF="#toc13">Other complement patterns</A>
</UL>
</UL>
<P></P>
<HR NOSHADE SIZE=1>
<P></P>
<P>
Produced by
gfdoc - a rudimentary GF document generator.
(c) Aarne Ranta (<A HREF="mailto:aarne@cs.chalmers.se">aarne@cs.chalmers.se</A>) 2002 under GNU GPL.
</P>
<P>
Aarne Ranta 2001 - 2006
</P>
<P>
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.
</P>
<P>
Closed categories (determiners, pronouns, conjunctions) are
accessed through the resource syntax API, <CODE>Structural.gf</CODE>.
</P>
<P>
The main difference with <CODE>MorphoFre.gf</CODE> 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.
</P>
<P>
The structure of functions for each word class <CODE>C</CODE> is the following:
first we give a handful of patterns that aim to cover all
regular cases. Then we give a worst-case function <CODE>mkC</CODE>, which serves as an
escape to construct the most irregular words of type <CODE>C</CODE>.
However, this function should only seldom be needed: we have a
separate module <A HREF="../../french/IrregFre.gf"><CODE>IrregFre</CODE></A>,
which covers all irregularly inflected verbs.
</P>
<PRE>
resource ParadigmsFre =
open
(Predef=Predef),
Prelude,
CommonRomance,
ResFre,
MorphoFre,
CatFre in {
flags optimize=all ;
</PRE>
<P></P>
<A NAME="toc1"></A>
<H2>Parameters</H2>
<P>
To abstract over gender names, we define the following identifiers.
</P>
<PRE>
oper
Gender : Type ;
masculine : Gender ;
feminine : Gender ;
</PRE>
<P></P>
<P>
To abstract over number names, we define the following.
</P>
<PRE>
Number : Type ;
singular : Number ;
plural : Number ;
</PRE>
<P></P>
<P>
Prepositions used in many-argument functions are either strings
(including the 'accusative' empty string) or strings that
amalgamate with the following word (the 'genitive' <I>de</I> and the
'dative' <I>à</I>).
</P>
<PRE>
accusative : Prep ;
genitive : Prep ;
dative : Prep ;
mkPrep : Str -&gt; Prep ;
</PRE>
<P></P>
<A NAME="toc2"></A>
<H2>Nouns</H2>
<PRE>
mkN : overload {
</PRE>
<P></P>
<P>
The regular function uses heuristics to compute the
plural and the gender from the singular. The plural
heuristic currently
covers the cases <I>pas-pas</I>, <I>prix-prix</I>, <I>nez-nez</I>,
<I>bijou-bijoux</I>, <I>cheveu-cheveux</I>, <I>plateau-plateaux</I>, <I>cheval-chevaux</I>.
The gender heuristic is less reliable: it treats as feminine all
nouns ending with <I>e</I> and <I>ion</I>, all others as masculine.
</P>
<PRE>
mkN : (cheval : Str) -&gt; N ;
</PRE>
<P></P>
<P>
Adding gender information widens the scope of the regular pattern.
</P>
<PRE>
mkN : (foie : Str) -&gt; Gender -&gt; N ;
</PRE>
<P></P>
<P>
In the worst case, both singular and plural forms and the gender are needed.
</P>
<PRE>
mkN : (oeil,yeux : Str) -&gt; Gender -&gt; N ;
</PRE>
<P></P>
<A NAME="toc3"></A>
<H3>Compound nouns</H3>
<P>
Some nouns are ones where the first part is inflected as a noun but
the second part is not inflected. e.g. <I>numéro de téléphone</I>.
They could be formed in syntax, but we give a shortcut here since
they are frequent in lexica.
</P>
<PRE>
mkN : N -&gt; Str -&gt; N
} ;
</PRE>
<P></P>
<A NAME="toc4"></A>
<H3>Relational nouns</H3>
<P>
Relational nouns (<I>fille de x</I>) need a case and a preposition.
</P>
<PRE>
mkN2 : N -&gt; Prep -&gt; N2 ;
</PRE>
<P></P>
<P>
The most common cases are the genitive <I>de</I> and the dative <I>à</I>,
with the empty preposition.
</P>
<PRE>
deN2 : N -&gt; N2 ;
aN2 : N -&gt; N2 ;
</PRE>
<P></P>
<P>
Three-place relational nouns (<I>la connection de x à y</I>) need two prepositions.
</P>
<PRE>
mkN3 : N -&gt; Prep -&gt; Prep -&gt; N3 ;
</PRE>
<P></P>
<A NAME="toc5"></A>
<H3>Relational common noun phrases</H3>
<P>
In some cases, you may want to make a complex <CODE>CN</CODE> into a
relational noun (e.g. <I>la vieille église de</I>). However, <CODE>N2</CODE> and
<CODE>N3</CODE> are purely lexical categories. But you can use the <CODE>AdvCN</CODE>
and <CODE>PrepNP</CODE> constructions to build phrases like this.
</P>
<A NAME="toc6"></A>
<H3>Proper names and noun phrases</H3>
<P>
Proper names need a string and a gender. If no gender is given, the
feminine is used for strings ending with <I>e</I>, the masculine for other strings.
</P>
<PRE>
mkPN : overload {
mkPN : Str -&gt; PN ;
mkPN : Str -&gt; Gender -&gt; PN
} ;
</PRE>
<P></P>
<A NAME="toc7"></A>
<H2>Adjectives</H2>
<PRE>
mkA : overload {
</PRE>
<P></P>
<P>
For regular adjectives, all forms are derived from the
masculine singular. The heuristic takes into account certain
deviant endings: <I>banal-banale-banaux</I>, <I>chinois-chinoise-chinois</I>,
<I>heureux-heureuse-heureux</I>, <I>italien-italienne</I>, <I>jeune-jeune</I>,
<I>amer-amère</I>, <I>carré- - -carrément</I>, <I>joli- - -joliment</I>.
</P>
<PRE>
mkA : (cher : Str) -&gt; A ;
</PRE>
<P></P>
<P>
Often just the feminine singular is deviant.
</P>
<PRE>
mkA : (sec,seche : Str) -&gt; A ;
</PRE>
<P></P>
<P>
This is the worst-case paradigm for the positive forms.
</P>
<PRE>
mkA : (banal,banale,banaux,banalement : Str) -&gt; A ;
</PRE>
<P></P>
<P>
If comparison forms are irregular (i.e. not formed by <I>plus</I>, e.g.
<I>bon-meilleur</I>), the positive and comparative can be given as separate
adjectives.
</P>
<PRE>
mkA : A -&gt; A -&gt; A
} ;
</PRE>
<P></P>
<P>
The functions create by default postfix adjectives. To switch
them to prefix ones (i.e. ones placed before the noun in
modification, as in <I>petite maison</I>), the following function is
provided.
</P>
<PRE>
prefixA : A -&gt; A ;
</PRE>
<P></P>
<A NAME="toc8"></A>
<H3>Two-place adjectives</H3>
<P>
Two-place adjectives need a preposition for their second argument.
</P>
<PRE>
mkA2 : A -&gt; Prep -&gt; A2 ;
</PRE>
<P></P>
<A NAME="toc9"></A>
<H2>Adverbs</H2>
<P>
Adverbs are not inflected. Most lexical ones have position
after the verb.
</P>
<PRE>
mkAdv : Str -&gt; Adv ;
</PRE>
<P></P>
<P>
Some appear next to the verb (e.g. <I>toujours</I>).
</P>
<PRE>
mkAdV : Str -&gt; AdV ;
</PRE>
<P></P>
<P>
Adverbs modifying adjectives and sentences can also be formed.
</P>
<PRE>
mkAdA : Str -&gt; AdA ;
</PRE>
<P></P>
<A NAME="toc10"></A>
<H2>Verbs</H2>
<P>
Irregular verbs are given in the module <CODE>IrregFre</CODE>.
If a verb should be missing in that list, the module
<CODE>BeschFre</CODE> gives all the patterns of the <I>Bescherelle</I> book.
</P>
<P>
Regular verbs are ones with the infinitive <I>er</I> or <I>ir</I>, the
latter with plural present indicative forms as <I>finissons</I>.
The regular verb function in the first conjugation recognizes
these endings, as well as the variations among
<I>aimer, céder, placer, peser, jeter, placer, manger, assiéger, payer</I>.
</P>
<P>
Sometimes, however, it is not predictable which variant of the <I>er</I>
conjugation is to be selected. Then it is better to use the function
that gives the third person singular present indicative and future
((<I>il</I>) <I>jette</I>, <I>jettera</I>) as second argument.
</P>
<PRE>
mkV : overload {
mkV : (finir : Str) -&gt; V ;
mkV : (jeter,jette,jettera : Str) -&gt; V ;
</PRE>
<P></P>
<P>
The <CODE>IrregFre</CODE> list gives some verbs as two-place. These verbs can be
reused as one-place verbs.
</P>
<PRE>
mkV : V2 -&gt; V
} ;
</PRE>
<P></P>
<P>
The function <CODE>mkV</CODE> gives the default compound auxiliary <I>avoir</I>.
To change it to <I>être</I>, use the following function.
</P>
<PRE>
etreV : V -&gt; V ;
</PRE>
<P></P>
<P>
This function turns a verb into reflexive, which implies the auxiliary <I>être</I>.
</P>
<PRE>
reflV : V -&gt; V ;
</PRE>
<P></P>
<A NAME="toc11"></A>
<H3>Two-place verbs</H3>
<P>
Two-place verbs need a preposition, except the special case with direct object.
(transitive verbs).
</P>
<PRE>
mkV2 = overload {
mkV2 : V -&gt; V2 = dirV2 ;
mkV2 : V -&gt; Prep -&gt; V2 = mmkV2
} ;
</PRE>
<P></P>
<A NAME="toc12"></A>
<H3>Three-place verbs</H3>
<P>
Three-place (ditransitive) verbs need two prepositions, of which
the first one or both can be absent.
</P>
<PRE>
mkV3 : overload {
mkV3 : V -&gt; V3 ; -- donner,_,_
mkV3 : V -&gt; Prep -&gt; V3 ; -- placer,_,dans
mkV3 : V -&gt; Prep -&gt; Prep -&gt; V3 -- parler, à, de
} ;
</PRE>
<P></P>
<A NAME="toc13"></A>
<H3>Other complement patterns</H3>
<P>
Verbs and adjectives can take complements such as sentences,
questions, verb phrases, and adjectives.
</P>
<PRE>
mkV0 : V -&gt; V0 ;
mkVS : V -&gt; VS ;
mkV2S : V -&gt; Prep -&gt; V2S ;
mkVV : V -&gt; VV ; -- plain infinitive: "je veux parler"
deVV : V -&gt; VV ; -- "j'essaie de parler"
aVV : V -&gt; VV ; -- "j'arrive à parler"
mkV2V : V -&gt; Prep -&gt; Prep -&gt; V2V ;
mkVA : V -&gt; VA ;
mkV2A : V -&gt; Prep -&gt; Prep -&gt; V2A ;
mkVQ : V -&gt; VQ ;
mkV2Q : V -&gt; Prep -&gt; V2Q ;
mkAS : A -&gt; AS ;
mkA2S : A -&gt; Prep -&gt; A2S ;
mkAV : A -&gt; Prep -&gt; AV ;
mkA2V : A -&gt; Prep -&gt; Prep -&gt; A2V ;
</PRE>
<P></P>
<P>
Notice: categories <CODE>V2S, V2V, V2Q</CODE> are in v 1.0 treated
just as synonyms of <CODE>V2</CODE>, and the second argument is given
as an adverb. Likewise <CODE>AS, A2S, AV, A2V</CODE> are just <CODE>A</CODE>.
<CODE>V0</CODE> is just <CODE>V</CODE>.
</P>
<PRE>
V0, V2S, V2V, V2Q : Type ;
AS, A2S, AV, A2V : Type ;
</PRE>
<P></P>
<!-- html code generated by txt2tags 2.3 (http://txt2tags.sf.net) -->
<!-- cmdline: txt2tags -thtml -\-toc french/ParadigmsFre.txt -->
</BODY></HTML>