Natural Language Grammars for an Information System

NATURAL LANGUAGE GRAMMARS FOR AN I N FOR M A TI ON SYSTEM L u i s de S o p e ~ a CENTRO DE I N F O R M A T I C A Universidad del Pais Vasco BILBAO-SPAIN t h e same t e c h n o l o g y : t h e f i r s t version of USL was d e v e l o p e d f o r German by H. Lehmann N. O t t and M Z o e p p r i t z ( 4 , 5 , 6 ) a t t h e IBM Heidelberg Scientific Center. Subsequently and using it as a basis a Spanish version was also written (8), as well as an English one. ABSTRACT The User Specialty Languages (USL) System is an applications independent natural language interface to a Relational Database System. It provides non DP-trained people with a tool to introduce, query, manipulate and analyse the data stored in a Relational Database via natural language. USL interfaces with different languages; in the present paper the gran~nar developed for Spanish is presented, and compared with the German granTnar which was previously implemented and upon which it is based. Their main differences are pointed out, and the generality of the system to deal with other natural languages shown. The c o m p o n e n t s of USL a r e : a parser, a grammar w i t h a structural vocabulary, a set o f i n t e r p r e t a t i o n routines associat e d t o t h e grammar r u l e s , and t h e Data Base Management System, System R. t h e system has been d e s i g n e d i n such a way t h a t o n l y t h e grammar and t h e s t r u c t u r a l vocabulary need t o be changed i n o r d e r t o s h i f t f r o m one l a n g u a g e t o a n o t h e r ; t h e r e s t o f t h e i t e m s a r e common and can be s h a r e d by a l l language i n t e r f a c e s . 1. INTRODUCTION In a d d i t i o n t o t h e system components a s u b j e c t - d e p e n d e n t v o c a b u l a r y has t o be d e f i n e d by t h e u s e r f o r each a p p l i c a t i o n t o g e t h e r with h i s / h e r database r e l a t i o n s and v i e w s . The U s e r S p e c i a l t y Languages (USL) Sysan applications independent natutem is ral language i n t e r f a c e to a Relational D a t a b a s e . I t was d e s i g n e d t o p r o v i d e non DP-trained users with a t o o l to i n t r o d u ce, query, m a n i p u l a t e and a n a l y s e t h e d a t a s t o r e d i n a Data Base. USL i s p l a c e d on t o p o f a R e l a t i o n a l Data Base Management System, System R ( 1 ) , and i t s f u n c tion is to translate natural language input sentences typed in" by t h e u s e r s into queries written in the f o r m a l query l a n g u a g e o f System R. The p r e s e n t p a p e r c e n t e r s a r o u n d t h e German and t h e Spanish Grammars d e v e l o p e d for USL. Section 2 describes briefly the grammar s t r u c t u r e and s c o p e . S e c t i o n 3 d e a l s w i t h a comparison o f b o t h grammars and p o i n t s o u t t h e i r main d i f f e r e n c e s . The work r e p o r t e d h e r e was d e v e l o p e d during the a u t h o r ' s stay at the Heidelberg Scientific C e n t e r o f IBM Germany i n 1981. A main o b j e c t i v e o f USL was t o a c h i e v e a p p l i c a t i o n s independence, i . e . to d e s c r i be t h a t s u b s e t o f n a t u r a l l a n g u a g e ( m o r phology, structural words, syntactic s t r u c t u r e s and t h e i r i n t e r p r e t a t i o n ) that can be used i n t h e c o n t e x t o f d a t a b a s e interrogation, and that constitutes a c o r e t o be used i n d i f f e r e n t domains. At t h e s a m e t i m e USL aimed a t o b t a i n i n g i n terfaces for diferrent languages using 2. T h i s s e c t i o n r e p o r t s t h e main f e a t u r e s o f t h e grammar a c c e p t e d by t h e USL p a r s e r . Detailed descriptions of t h e German and Spanish v e r s i o n s can be f o u n d i n r e f e r e n ces ( g ) and ( 8 ) , r e s p e c t i v e l y . Permission to copy without fee all or part of this material is granted provided that the copies are not made or distributed for direct commercial advantage, the ACM copyright notice and the title of the publication and its date appear, and notice is given that copying is by permission of the Association for Computing Machinery. To copy otherwise, or to republish, requires a fee and/or specific permission. 1983ACM0-89791-107-5/83/006/0075 GRAMMAR OF THE USL SYSTEM $00.75 75 2.1. GRAMMAR STRUCTURE define vocabulary words, constructs from strings: is written in a modiT h e USL g r a m m a r fied Backus-Naur F o r m . H e r e we s h a l l brief ly describe its elementsand the structures it uses. The - basic elements '0123456789' - creating NOMEN:+MAS,+SG:FPE-NOMEN('PAIS') (= 'PAIS'; T h e f e a t u r e s for M a s c u l i n e a n d S i n g u l a r are a s s i g n e d to the c o n s t r u c t defined as N O M E N , w h i c h s p a n s the s t r i n g 'PAIS' ('country'). FPE-NOMEN refers to the n a m e of the r e l a t i o n to be a c c e s s e d f o r interpretation of the noun, here the s a m e w o r d 'paLs' is used. are: PRIMITIVES: Letters, digits symbols; they have to be the grammar. Example: by and special declared in PRIMITIVE<DIGIT>; <- G r a m m a r and strings hand side, on the left. taching an be: CONSTRUCTS:Grammatical categories def~ ned for verbs (VERB), nouns (NOMEN), adjectives (ADJ), prepositions (PREP), noun phrases (NP), l o c a l (ABL) and temp o r a l (ABT) complements, v e r b and one o r more complements (Verb Complex, VC) v e r b and a11 i t s complements (Sentence K e r n e l , SK), e t c . Example o f a Construct declaration: rules proper: Constructs may appear on the rightone or several constructs For example, the rule atadjective to a noun could <NOMEN:I:FPE-ADJ(2,1)><<NOMEN:(MAS=MAS(2))!(FEM=FEM(2)), (SG=SG(2))!(PL=PL(2))><ADJ>; <VERB> CONSTRUCT; - where for simplicity only the tests for gender and number agreement have been included on the right-hand s i d e NOMEN construct. The resulting construct retains the features of the original NOMEN via the parameter 1, and the semantic routine FPE-ABJ is called for interpretation of adjectives modifying nouns. FEATURES: They q u a l i f y C o n s t r u c t s and belong t o one o f t h r e e t y p e s : i n t e g e r ( v a l u e s are i n t e g e r numbers), l o g i c a l ( v a l u e s are + and - , o r 1 and O), and case ( v a l u e s belong t o a s e t o f s o called c a s e v a l u e s , which must a l s o be d e f i n e d . Examples: <SG> FEATURE LOGICAL; singular etc. of <LAB> FEATURE qualifies (construct <TYP> for nouns, number. adjectives, 2.2. the different PREP). GRAMMAR SCOPE T h e USL g r a m m a r describes a subset of natural language to be u s e d in database interrogation; therefore many structures not important in this context h a v e n o t been included. However, both the German and the Spanish Grammars are quite comprehensive; they provide a similar coverage of the language, which includes: - Wh-questions - Yes/no questions - Commands Statements Negation - Adjectives - Genitive Attributes - Appositions Noun Complements Relative Clauses - Quantifiers - Comparatives - Coordination - Possessive Pronouns - Locative Adverbials - Temporal Adverbials Functions of sum, maximum, minimum, average, number Use of variables and functions INTEGER; prepositions FEATURE CASE; quaZifies plements have to <NI> feature singular verbs according to the comthey require. Values of TYP be defined as, for example: CASEVALUE; - for intransitive only). verbs (Nominative - {NAD>CASEVALUE; - f o r verbs w i t h Nominative, and D a t i v e complements. Accusative - The r u l e s o f the grammar make use o f the elements t h a t have p r e v i o u s l y been d e c l a r e d . They have a s p e c i a l symbol t o separate left and right-hand sides: members on the r i g h t are i n p u t e l e m e n t s ; after application o f the r u l e t h e y are s u b s t i t u t e d by the elements s p e c i f i e d on the l e f t . - There are f o u r t y p e s o f r u l e s c b a r a c t e r i s e d by the symbol Used t o s e p a r a t e both s i d e s ; here we s h a l l d e s c r i b e o n l y the two most i m p o r t a n t t y p e s : <= F i x e d Token r u l e s : t h e y are 3. THE S P A N I S H A BRIEF used t o The 76 firts GRAMMAR: PARALLEL USL WITH grammar GERMAN was written for German; t h e Spanish grammar i s based on it, t h e r e f o r e t h e r e are many r u l e s s i m i larly formulated, sometimes only with slight variants. In this section the main differences between both grammars w111 be p o i n t e d out, as it c a n be f o u n d after comparing the descriptions given in references (g) and (8). 3.1. feature is set on all these Noun Phrases); 'a' may a l s o introduce an Indirect Object: e.g. 'amar A Juan', 'vender algo A alguien'. The preposition 'para' i n t r o d u c e s Indirect Objects and the D a t i v e feature is set on t h e s e N o u n P h r a s e s . As to the Genitive case, it d o e s not e x i s t as s u c h in S p a n i s h . c) Rules for nouns with determiners are very simple in Spanish, only gender and number checks are needed. German has to restrict the case features too, and distinguish between nominal inflection, adjective inflection, and the determiners 'wieviel', 'lauter/nur', and 'der/ein' Many more rules much longer and complicated are therefore necessary in German, while only three rules are needed in Spanish. MORPHOLOGY USL Grammars make no d i s t i n c t i o n between m o r p h o l o g y and s y n t a x ; t h e same k i n d o f r u l e s s e r v e s f o r both p u r p o s e s . Rules defining morphology are of course very d i f f e r e n t i n b o t h grammars: noun and a d j e c t i v e affixes, inflection o f nouns and a d j e c t i v e s a r e m u c h more s i m p l e i n t h e Spanish grammar, as o n l y p l u r a l s need be d e s c r i b e d f o r nouns, and f e m i n i n e s and p l u r a l s f o r a d j e c t i v e s , and no d e c l e n s i o n s e x i s t . But a t t h e same time these c o n s t r u c t s p r o v i d e l e s s i n f o r mation, as o n l y g e n d e r and number a r e marked, and n o t h i n g h e l p s d e t e r m i n e t h e case (unless a preposition is later attached), as occurs in German. d) Spanish and German use the same three gender features for masculine, feminine, and neuter. However, neuter nouns de n o t properly exist in Spanish as they do in German: the neutral article '1o' determines adjectives used as nouns, Iike '1o bello', '1o util', '1o me]or' (the beautiful', 'the usefui', 'the best'). They always have a collective and abstract meaning which is not expected to occur in database interrogation. Therefore, apart from a rule attaching the article '1o' to a noun defined as neutral (which the vocabulary definition programs allow ), no further use is made of the feature in the grammar. Spanish Other possible uses of '1o' preceding relative clauses of prepositional phrases, e.g. '1o que paso', lo de sLempre', have not been considered. German must also account for special cases, like nouns with adjective inflection, and nouns and names without article, which are not necessary in Spanish. As a result, several features used in German are not defined in Spanish. 3.2. SYNTAX I n many cases r u l e s a r e v e r y s i m i l a r l y formulated, only small changes can be detected by looking at the features checked on t h e r i g h - h a n d s i d e s , t h e ones s e t on t h e r e s u l t i n g c o n s t r u c t s , and t h e semantic r o u t i n e s invoked. e) In the rules for the formation o f t h e SENT construct (the one spanning the whole input string), one is provided to transform a declarative sentence into a yes/no question. This transformation in the interpretation of a statement as an interrogation is done if a '?' is found at the end of the declarative sentence. In Spanish, the freedom in word order makes it possible to ask a question using a linear order just by a change in the intonation; when typing, a question mark is required to indicate this change. For example, the sentence 'Maria vive en B i l bao' is a statement ('Maria lives in Bilbao'), but 'Maria viva en B i l b a o ? ' is an interrogation ('Does Maria live in Bilbao?'). a) Adjectives in Spanish can be placed either before or after a noun. The difference is mainly stylistic, but sometimes also the meaning changes depending on the adjective position (e.g. 'hombre pobre' vs. 'pobre hombre', caballo grande' vs. 'gran caballo'). For the purpose of USL, the most usual, noun-adjective sequence i s t o be e x p e c t e d , even though rules ape also supplied for the reversed sequence. In German the adjective-noun sequence is only allowed. b) Even if nouns are morphologically unmarked for case in Spanish, the grammar makes use of case features. Any noun phrase not preceded by a preposition is in principle candidate for subject (Nominative) and direct object (Accusative). preposition 'a' may introduce a The Direct Object (for persons or any kind of 'personified' thing - but this has not been considered, and the Accusative f) The interpretation of of a Verb Complex or special in Spanish. 'ningun' ('no') is in negation should actually already implied by the way the sentences 77 negation in front Sentence Kernel is If the quantifier the Verb Group, no apply but the one quantifier. In this 'No v i v e 'Ningun ('Does ningun empleado no empleado vive employee en live en 'Compran ('Do all Bilbao?' Bilbao?' 'Tienen todos los ('Do all countries in B i l b a o ? ' ) w i l l get the s a m e i n t e r p r e t a t i o n , the f i r s t one has an a d d i t i o n a l in f r o n t of the s e n t e n c e . todos los employees even if negation are correctly 'Compran empleados coche?' buy a car?') paises capital?' have a capital?') interpreted, coche todos capital todos los los but empZeados?' g) 'Tienen paises?' The rules for Verb Complex (VC) and S e n tence Kernel (SK) formation are quite similar. Only in S p a n i s h when a Subject N o u n P h r a s e is a t t a c h e d to a VC, a f e a t u r e c a l l e d S U B J is c h e c k e d in both c o n s t r u c t s , a n d the N o u n P h r a s e c h e c k e d for p r e s e n c e of the q u a n t i f i e r 'todos' ('all'). T h e f i r s t c h e c k is t e n t a t i v e a n d t r i e s to syntactic ambiguities between avoid Subject and Accusative Object arising in some sentences. For e x a m p l e the s e n t e n c e s 'Exporta Espa~a vino?' and ' E x p o r t a v i n o EspaBa' (which both mean 'Does Spain export wine?') have two parses, 'vino' being subject in one parse and Direct Object in the other. This occurs because the only knowledge USL u s e s f o r disambiguation is the one contained in the strutture of database relations, a n d no r e a l knowledge representation device guides the system. The database c o n t e n t s are not d u p l i c a t e d in a d i c t i o n a r y , and t h e r e f o r e every unidentified string found in the i n p u t s e n t e n c e i s i n t e r p r e t e d as a d a t a base v a l u e . I n o u r example n e i t h e r ' E s p a ~a' nor ' v i n o ' a r e d e f i n e d i n t h e v o c a b u l a r y I b o t h o f them are v a l u e s s t o r e d i n t h e d a t a b a s e , and t h e system w i t h o u t s e m a n t i c knowledge i s u n a b l e t o d i s c a r d ' v i n o ' as subject of 'exportar'. which have exactly the same m e a n i n g as before, would be interpreted as if all employees bought t h e same c a r , all countries had the same c a p i t a l , due to the order of the quantifiers in the input sentence. Therefore the order of complements must be reversed for a correct interpretation of the s e n t e n c e s m e a n i n g . h) When the main verb is 'ser' or 'estar' ('be') s p e c i a i c h e c k s a r e s o m e t i m e s n e c e s If an adjective is attached as sary. complement to the verb a gender check is needed between adjective and the other Nominative, e.g.: 'Son casadas 'Que secretarias 1as secretarias?' son casadas?' where the adjective 'casada' 'secretaria' . On t h e other agreement is avoided for Nominatives: 'Quien es/quienes taria de Juan?' son el agrees with hand, number coordinated jefe y la secre- i) No p r e c e d i n g Genitive Atributes exist in Spanish, there are only genitives following their head nouns: 'salario de Juan' vs. two possibilities in German: ' Gehalt von Meier', Melers Gehalt' The f e a t u r e SUBJ t r i e s t o c h a r a c t e r i z e nouns and names and t h e matching v e r b s o f which they only can be subject: here 'exportar' and ' E s p a f i a ' . However t h e f e a t u r e i s n o t y e t c o m p l e t e l y used because t h i s would i m p l y , among o t h e r t h i n g s , t h e definition o f the p r o p e r names s t o r e d in t h e d a t a b a s e , which was up t o now a v o i d e d by USL . I f r e a l use i s t o be made o f t h e feature it will be n e c e s s a r y t o d e f i n e i t as integer, and assign its different values to different groups of verbs and subject candidates, or to define different categories of subjects (e.g. persons, animals, countries, etc.) and specify for each verb the classes of subjects it requires. This will allow to perform a semantic typing inside the syntax. j) When defining Prepositions the German grammar must duplicate many of them to account for their use with Accusative of Dative noun phrases. In Spanish this is not necessary as no case needs to be associated with prepositions, only one definition is needed. Rules h a v e t o be p r o v i d e d in Spanish for any sequence of complements after a verb requiring prepositional complements, even if actually some o f the structures thus obtained would sound odd and will probably not be u s e d , but they are grammatically correct. T h e s e c o n d c h e c k f o r a Noun P h r a s e w i t h 'todos' is important when t h e Noun P h r a s e is attached as Subject to the right of a VC. Due t o t h e USL l e f t to right interpretation for the scope of quantifiers, if the NP i s the first complement attached to the verb the interpretation will be correct. Otherwise the order of complem~ts should be r e v e r s e d so that t h e NP i s m o v e d to the first position. This is done by application of a special semantic routine. For example, the sentences: k) To d e s c r i b e Local Adverbials less rules are needed in Spanish, as less particular c a s e s n e e d t o be a c c o u n t e d for. They allow for the formation of interrogatives using 'donde' ('de donde', 'hasta donde'), and for adverbs with more than one preposition, like 'desde encima de', 'pot debajo de' i) For 78 Temporal Adverbials there are many differently f o r m u l a t e d r u l e s . For example, the f o l l o w i n g formats are allowed f o r a d a t e i n t h e Spanish grammar: 419/1956 4-IX-1956 04.09.1956 c u a t r o de s e p t i e m b r e de 1 9 5 6 4 de s e p t i e m b r e de 1956 26 de e n e r o ' S e c r e t a r i a cuyos j e f e s ' ( ' S e c r e t a r y whose m a n a g e r s ' ) ' N a c i o n e s cuyas s u p e r f i c i e s ' ( ' N a t i o n s whose s u r f a c e s ' ) e n e r o de 1 . g 7 9 a~o 1979 aSo 1 9 7 g mes de e n e r o mes de e n e r o de 1 9 7 9 dia/viernes 26 d e e n e r o In the r u l e s f o r R e l a t i v e Clause f o r m a t i o n t h e maximum i n f o r m a t i o n a b o u t t h e r e f e r e n t f e a t u r e s i s p i c k e d up f r o m the whole clause and transmitted to the resulting Noun Phrase, in o r d e r t o c o r r e c t l y i d e n t i f y the referent. When d e a l i n g w i t h D e c l a r a t i v e Sentences r u l e s are w r i t t e n i n t h e Spanish Grammar t o a l l o w f o r more t h a n one complement t o be p l a c e d t o t h e l e f t of a verb; this is intended f o r R e l a t i v e Clauses l i k e : A date can also be a t t a c h e d to a t i m e expression, e.g.: 'el 5 de a g o s t o a las 16:20'. Care must be t a k e n to a v o i d ambiguities between two possible uses of the word 'horas' as a m e a s u r e of a t i m e i n t e r v a l (meaning 'hours'), and as a p o i n t in t i m e (mean~g 'o'clock'): ' P a i s e s a l o s qua I t a l i a ' P r o d u c t o s que I t a l i a Actually, because o f Spanish l a r g e l y f r e e word o r d e r , i t i s a l s o p o s s i b l e f o r more t h a n one complement t o a p p e a r on t h e l e f t o f t h e main v e r b . However, most o f t h e s e c o n s t r u c t i o n s sound odd o r u n n a t u r a l and are not expected. ' d e n t r o de 3 horas' ' d u r a n t e 2 h o r a s y 10 m i n u t o s ' 'antes de las 5 horas' 'alas 3 h o r a s 10 m i n u t o s ' More t h a n one p r e p o s i t i o n is introduce a temporal adverbial: 'desde a n t e s del m a t t e s ' , ' h a s t a d e s p u e s de las 8'. allowed to n) Verbs o f t y p e NAA ( r e q u i r i n g N o m i n a t i v e and two A c c u s a t i v e s ) and NAG ( r e q u i r i n g N o m i n a t i v e , A c c u s a t i v e and G e n i t i v e ) do not e x i s t in Spanish, t h e r e f o r e the r u l e s written i n t h e German Grammar f o r these a d d i t i o n a l o b j e c t s need n o t be p r o v i d e d . m) The syntax of R e l a t i v e Clauses is q u i t e different in S p a n i s h , d u e to the p a r t i c u lar p r o p e r t i e s and uses of S p a n i s h Relative Pronouns. Also, as R e l a t i v e Clauses are subordinate, w o r d o r d e r is d i f f e r e n t in G e r m a n and the g r a m m a r m u s t be p r o v i d e d with a special section to d e s c r i b e these sentences with verb final word order. This is not n e c e s s a r y in S p a n i s h because all clauses, e i t h e r m a i n of s u b o r d i n a t e , have a l w a y s a l a r g e l y f r e e w o r d order. o) The r u l e s f o r C o o r d i n a t i o n in Spanish f o l l o w q u i t e c l o s e l y t h e German model. Only t h e c o o r d i n a t i o n o f N o u n Phrases has been k e p t l e s s c o m p l i c a t e d p r o b a b l y due n o t t o the g r e a t e r s i m p l i c i t y o f c o o r d i n a t i o n in Spanish but t o t h e s m a l l e r s o p h i s t i c a t i o n o f t h e b a l a n c e checks i n t h e c o o r d i n a t e d structures. These rules will certainly have t o be r e v i s e d . Most of the Spanish relative pronouns c o n v e y l i t t l e i n f o r m a t i o n a b o u t the c h a r a c teristics of their referent noun: 'que T can have a referent of a n y gender and number, 'cual' and 'quien' are o n l y m a r k e d for singular, 'cuales' and 'quienes' for plural, unless an article is p l a c e d in f r o n t of the r e l a t i v e : On a c o o r d i n a t e d N o u n Phrase t h e p l u r a l f e a t u r e i s always s e t , the s i n g u l a r feature o n l y i n case t h e c o n j u n c t i o n i s ' o ' ( ' o r ' ) and a t l e a s t one o f t h e c o n s t i t u e n t s i s singular, because i n t h i s case t h e v e r b can a g r e e w i t h i t s n e a r e s t N o u n Phrase, o r w i t h t h e c o o r d i n a t e d N o u n P h r a s e . The m a s c u l i n e f e a t u r e i s s e t i f one o f t h e Noun Phrases a r e f e m i n i n e . ' P a i s e s a los que e x p o r t a I t a l i a ' ( ' C o u n t r i e s to w h i c h I t a l y e x p o r t s ' ) ' £ m p l e a d o s de los c u a l e s ( ' E m p l o y e e s of w h o m J o h n exporta vino' e x p o r t a a Alemania' J u a n es jefe' is m a n a g e r ' ) S p e c i a l r u l e s a r e needed f o r c o o r d i n a t i o n with the conjunction 'sing' ('but'), because t h e f i r s t element o f the c o o r d i n a t i o n must be n e g a t e d , e . g . : ' n o j e f e s s i n g e m p l e a d o s ' , ' n o 5000 s i n g 6 0 0 0 ' . T h e last r e l a t i v e p r o n o u n 'cuyo' ( ' c u y a ', ' cuyos', 'cuyas': 'whose') admits any referent but must agree in gender and n u m b e r w i t h the n o u n f o l l o w i n g it: ( I n some cases o f s i n g u l a r nouns c o n j u n c t e d by ' y ' , e s p e c i a l l y i f t h e r e i s no a r t i c l e p r e c e d i n g t h e second noun, t h e r e s u l t i n g c o n j u n c t e d N o u n Phrase can a l s o be s i n g u lar: 'la e n t r a d a y s a l i d a de a v i o n e s se s u s p e n d i o ' . As t h e s e case a r e v e r y p a r t i c u l a r t h e y have n o t been c o n s i d e r e d ) . ' £ m p l e a d o s c u y o jefe' ('Employees whose manager') 'Pals c u y a p o b l a c i o n ' ('Country whose population') 79 relational database c o u l d be o b t a i n e d . The system is ready for use in real applicat~n environments, and user experiments and studies have even been performed with the original German version (2,3). The Spanish version is less developed and tested, and needs further revisions and improvements, but we h o p e i t will soon reach t h e same degree of applicability of its German counterpart. P) As to P e r s o n a l P r o n o u n s some of t h e m can o n l y o c c u p y s p e c i a l p l a c e s in s e n t e n c e s , and this must be accounted for: the personal pronouns 'el', 'ella, 'ellos', and 'ellas' ('he', 'she' and masculine and feminine 'they% respectively) function a s n o r m a l Noun P h r a s e s ; t h e y c a n be s u b j e c t of sentences, and become accusative or But d a t i v e if p r e c e d e d by a p r e p o s i t i o n . the P e r s o n a l Pronouns 'io' , 'la', 'los', 'las' ( a c c u s a t i v e ) , and 'le', 'se' ( a c c u s a t i v e and d a t i v e ) can o n l y be p l a c e d to the left of the verb, and immediatly preceding it. However, all the rules dealing with Personal Pronouns only describe thelr syntax, no i n t e r p r e t a t i o n routines are provided by the system . REFERENCES 1) A s t r a h a n M . M . , et. al. ' S y s t e m R: R e l a t i o n a l A p p r o a c h to D a t a base M a n a g e m e n t ' ACM T r a n s . on D a t a b a s e Systems, vol.1, n£ 2 , J u n e 1 9 7 6 . Exemples: 'Quien le necesita?' ('Who needs him?') 2) Kettler W., Schmidt A., Zoeppritz M. 'Erfahrungen mit zwei natuerlich-sprach lichen abfragesystemen', TR 8 1 . 0 1 . 0 0 1 , IBM Germany, Heidelberg Scientific Center, 1.981. 3) Lehmann H., Ott N., Zoeppritz M. 'User experiments with Natural Language for Data Base access' Proceedings 7 th. International ConfeLinguisitics, rence on Computational Bergen, 1.978. 4) L e h m a n n H. 'Interpretation of Natural an Information System', Research and Development, 5, september 1.978. 'Que p a i s io e x p o r t a (a A l e m a n i a ) ? ' ( ' W h i c h c o u n t r y e x p o r t s it (to G e r m a n y ) ? ' ) 'Quien le vende ('Who sells him 'Quien ('Whow (un auto)?' (a car)?') se io vende?' s e l l s it to him?') And s o m e t i m e s a Personal Pronoun is used as a redudant dative to emphasize or further explain the indirect object of the sentence: ~Quien (le) ('Who s e l l s vende autos a Italia?' c a r s to I t a l y ? ' ) 4. provided to describe this O t t N . , M. Z o e p p r i t z ' USL An Experimental Information System Based on Natural Language' in L Bolc (ed.): Natural Communication with Computers, vo] . 2, Carl Hanser Verlag, Muenchen-Wien, 1.979. 6) L e h m a n n H. 'A System for Answering German' Proc. 6th. International sium, Cambridge - England, parti- CONCLUSIONS overview of the grammar accepted by the User Specialty Languages parser has been given. The main lines of the Spanish Grammar have been described, together with its particularities in relation with the German one upon which it is based. An 7) It has been shown that there are many small differences and minor details changing from one grammar to the other, but the main lines have been kept, and what is more important, the interpretation routines needed for the semantic part of t h e USL S y s t e m h a v e b e e n a l s o used in the Spanish version almost unchanged. This indicates that this same a p p r o a c h can apply to other languages (as has already been done for example with English), and just by writing a g r a m m a r i n t h e USL f o r m a t a language interface to a whole natural in of n£ 5) '(Le) suministra Italia vino a Alemania?' ('Does Italy supply Germany with wine?) Rules are cularity. Language IBM J. vol. 22, Questions ALLC 1.980. Real Academia Espa~ola 'Esbozo de una nueva gramatica Lengua espa~ola' Espasa-Calpe, Madrid, 1.979. in Sympo- de la 8) S o p e R a L. ' G r a m m a r of S p a n i s h for U s e r S p e c i a l t y Languages' TR 8 2 . 0 5 . 0 0 4 , IBM G e r m a n y , Heidelberg Scientific Center, 1.982. 9) 80 Z o e p p r i t z M. ' Syntax for ty Languages forthcoming. German in System' the User Special