Construcciones, usos y variación desde el latín hasta la actualidad
Edited By Carlota de Benito Moreno and Álvaro S. Octavio de Toledo y Huerta
Este volumen constituye la primera monografía que aborda el haz entero de construcciones en las que, a lo largo de la historia, ha participado el auxiliar más conocido del español, HABERE > haber. Catorce especialistas de universidades europeas y americanas trazan, desde muy diversas ópticas teóricas (lingüística de área, gramática formal, gramática de construcciones, sociolingüística, dialectología o gramaticalización), una perspectiva de conjunto, exhaustiva en los detalles y novedosa en sus hipótesis, sobre los distintos caminos que históricamente ha emprendido el verbo ‘haber’ y hoy caracterizan su comportamiento en español (y otras lenguas románicas aquí abordadas: catalán, francés, portugués...), adentrándose igualmente en aspectos generales de la variación y el cambio morfosintácticos.
Existentials in relative clauses: a contrastive corpus study of Spanish haber and French y avoir
Eberhard Karls Universität Tübingen
One important use of haber is an impersonal existential construction (1). French y avoir (1b) and English there be (1c) are also examples of typical existential constructions and function quite parallel to Spanish haber (for a recent cross-linguistic definition of existential constructions and a description of their typical properties see McNally (2011), with special reference to Spanish see Fernández Soriano / Táboas 1999: 1754–1759).
(1) a. Hay un libro sobre la mesa
b. Il y a un livre sur la table
c. There is a book on the table
The noun phrase following the existential (in the above examples un libro/ un livre/ a book) is usually called ‘pivot’, everything following it is called ‘coda’ (sobre la mesa / sur la table/ on the table) (cf. McNally 2011: 1831). As the main function of existential sentences consists in introducing new discourse referents, the pivot is generally indefinite. This phenomenon holds across many languages and is known as the Definiteness Effect or the Definiteness Restriction (for an overview see McNally 2011; with special reference to Spanish and French see Suñer 1982, Leonetti 2008, Dobrovie-Sorin / Beyssade 2004, 2012). Despite an on-going debate about the language-specific details and the exact linguistic nature of the Definiteness Restriction, there is general agreement that simple existential clauses with definite pivots, in Spanish as well as in French and English, require special contextualization to be acceptable (2a-c).
(2) a. #Hay el libro sobre la mesa
b. #Il y a le livre sur la table
c. #There is the book on the table
(3) a. El libro está sobre la mesa
b. Le livre est sur la table
c. The book is on the table
This observation has led to an approach in terms of a “locative paradigm” (Freeze 1992: 554), that considers existential sentences (1a-c) and locative sentences (3a-c) as being closely connected (among others Lyons 1967, Clark 1978, Freeze 1992, Partee / Borschev 2007, Koch 2012; an early study of the links between haber, estar and definiteness in Spanish is Bull 1943). In relation to the Definiteness Effect, Carlson (1977) observed for English that relativization of the existential pivot is possible only in a very special class of relative clauses, that he called Amount Relatives (4). In amount relatives, the antecedent is not denoting an individual (e.g. the man) but rather an amount of something (that many men), and thus conforms to the Definiteness Restriction on the pivot of existentials (cf. Carlson 1977, McNally 2008, Dobrovie-Sorin / Beyssade 2004, 2012).
(4) a. Every man there was on the life-raft died (= 6b in Carlson 1977: 521)
b. The men (that) there were in Austria liked Bob (= 14a in Carlson 1977: 526)
Dobrovie-Sorin / Beyssade (2004, 2012) adapt Carlson’s observation to French, and claim that in French the class of existentials in relative clauses is larger than in English, but is still limited to head nouns of the semantic type ‘property’ (excluding individuals) (Dobrovie-Sorin / Beyssade 2004: 91, cf. also McNally 2008, who presents similar conclusions on English). Just like Carlson (1977), they explain this phenomenon with the Definiteness Effect. According to all these studies, relative clauses with existentials headed by singular count nouns (i.e. head nouns that denote individuals) are neither possible in French nor in English (5a, b).
(5) a. *Le livre qu’il y avait dans la bibliothèque de ma grand-mère a brûlé (= 92 in Dobrovie-Sorin / Beyssade 2012:124)
b. *The man that there was in Austria likes Bob (= 14c in Carlson 1977: 526)
As the Definiteness Effect is at work in Spanish, too, it will be interesting to compare the findings for English and French with Spanish data. Indeed, the Definiteness Effect in Spanish declarative clauses with haber (ex. 2a) is even stricter than in French (Leonetti 2008, Meulleman 2012). We thus should expect Spanish haber to allow only amount relatives as well. However, this is not the case. Relative clauses containing haber and being headed by singular count nouns (6) are, on the contrary, frequently used in Spanish.
Escandell / Leonetti (1998) already point out that the Definiteness Effect seems to disappear in Spanish relative clauses headed by the pivot of haber and explain (1998: 262–263):
De acuerdo con el análisis más plausible de las restrictivas, éstas son modificadores de una proyección nominal, y no del SN entero; en otras palabras, el antecedente es sólo el N (o el N’) que precede a la relativa, y no incluye, por tanto, al artículo. […] De esta forma el antecedente no posee ninguno de los rasgos referenciales aportados por el determinante (en el caso que nos ocupa, lo decisivo es que no posee el rasgo de definitud), y, en consecuencia, no se produce incompatibilidad alguna con la construcción existencial.
At this point, however, the question arises why this explanation should not also work for French or English, i.e. why haber may occur freely inside restrictive relative clauses, whereas y avoir and there be may not, although the Definiteness Effect holds in all the three languages. Obviously, there must be reasons beyond the Definiteness Effect. In order to shed further light on the issue, I conducted an extensive corpus study on French and Spanish relative clauses with y avoir and haber. It soon turned out indispensable to also consider their counterparts with être and estar. As Spanish and French belong to the same language family and existential haber and y avoir are very similar to each other, (they even originate from the same source construction, namely (hic/ibi) habet that was already used as an existential in Vulgar Latin, see Fernández Soriano / Táboas 1999: 1757), I focused on these two languages, and included English only occasionally into the comparison. The study is based on corpus data from Frantext catégorisé1 and CORPES XXI2 for French and Spanish literary language and on the C-Oral Rom corpus (Cresti / Moneglia 2005) for spoken language. In order to directly compare Spanish with French, and also with English, I consulted two parallel corpora from the OPUS collection (Tiedemann 2012), namely Open Books and Open Subtitles.3 After a summary of the situation in French in chapter 2, I present and discuss the Spanish corpus data in chapter 3. In chapter 4, I suggest a new explanation, which is not based on the Definiteness Effect. Chapter 5 gives a conclusion. ← 193 | 194 →
2. The situation in French
Dobrovie-Sorin / Beyssade (2004: 85–92; 2012: 117–125) adapt Carlson’s (1977) theory of amount relatives to French. They observe that French y avoir is not as restricted to amount relatives as English there be, but is still restrained to special types of head nouns, namely property denoting nouns (Dobrovie-Sorin / Beyssade 2004: 91). They also note that y avoir inside French relative clauses is very rare, and report that they only found around 30 examples in Frantext, a database of French literary language (Dobrovie-Sorin / Beyssade 2004: 90; 2012: 122). Their observation is partly confirmed by my study. I used a slightly different corpus, namely Frantext catégorisé. When this study was carried out, it included 1.940 texts, mostly novels, and contained 127.515.618 words. Probably thanks to the more refined search possibilities offered by Frantext catégorisé, and also to the successive enlargements of the corpus since 2004, I found around 300 examples of y avoir inside relative clauses, which is much more than Dobrovie-Sorin/Beyssade (2004) did. The examples, however, are mostly consistent with the observations made by Dobrovie-Sorin/Beyssade. Most of them are amount relatives, with a nonspecific, non individual head noun, that can easily be interpreted as denoting an amount, like in (7a). Sometimes, there is even an explicit quantitative term, like la quantité in (7b) or plusieurs in (7c). Example (7c) also clearly emphasizes that the head noun plusieurs des hommes is not specific but rather meant as describing types or kinds of different men, i.e. this example is related to the quantification over kinds, that is possible for pivots of existential constructions (cf. McNally 2011).4
(7) a. Je m’imagine sans cesse que tu vas voir les belles jeunes femmes qu’il y a dans la ville
b. La quantité de chefs - d’oeuvre qu’il y a dans cette ville [= Rome], c’est étourdissant!
c. Un homme qui lit une feuille d’annonces matrimoniales peut délivrer, tour à tour, plusieurs des hommes qu’il y a en lui : l’homme qui rit, l’homme qui convoite, l’homme qui réfléchit
So far, all examples are consistent with the theory of the extended class of amount relatives, defended by Dobrovie-Sorin / Beyssade (2004: 92): “Il semble […] qu’on puisse fournir une explication qui prolonge celle de Carlson (1977b) [= Carlson 1977] tout en rendant compte d’un plus grand nombre de données : la quantité est un cas particulier d’expression dénotant un non particulier.5 Donc, parmi les ← 194 | 195 → relatives construites avec il y a, on trouvera aussi, en plus de relatives de quantité, des relatives de propriétés, de genre, des espèces,…”
Yet there are also counterexamples. Head nouns that denote specific individuals should be excluded from relativization with y avoir. Both Carlson (1977: 526) and Dobrovie-Sorin / Beyssade (2012: 121, 124) explicitly state that relativization with there be or y avoir of singular count nouns is impossible (cf. ex. 5 in the introduction). In my corpus study, however, I found evidence against this claim:
(8) a. Je n’ai pas touché à la haute chaise de bébé qu’il y avait dans son cabinet de travail
b. J’ai pensé à une pompe, dit Wolf […] à une pompe qu’il y avait dans le jardin des voisins, avec un battant, et peinte en vert
c. C’est une maladie qu’il y a dans la famille de mon père… Mon grand-père a été pris et avant lui d’autres tantes et cousins: une sorte de paralysie
All these examples clearly involve a singular, specific count noun, whether with a definite (8a) or an indefinite (8b, c) article.6 Both concrete objects (8a, b) and abstract entities (8c) can be referred to. These counterexamples are however very rare, and make up for less than 5 % of the (already few) 300 occurrences of y avoir inside relative clauses. A search of the same corpus for relative clauses with the copula être returned more than 2000 results, i.e. more than seven times more than with y avoir. Relative clauses with être are used with singular (9a) and plural count nouns (9b), abstract nouns (9c), mass nouns (9d), and mass nouns with an explicit specification of amount (9e). Example (9f) illustrates a case where the noun is given a kind interpretation, quite parallel to example (7c). Être inside relative clauses is thus much less restricted than y avoir: it can be used in all contexts where y avoir is used, and, additionally, in contexts where y avoir is hardly ever used.
(9) a. J’ai sorti la grande valise de famille qui était sous le lit
b. Mon père, par exemple, accuse le voisin de lui avoir volé des planches qui étaient dans le fond du jardin
c. Vallespir : le plus joli nom de vallée qui soit dans notre langue
d. Il dit que le poisson et le riz qui étaient sur la table s’étaient refroidis et qu’il en avait apporté d’autres
e. Il lui faut dresser les quinze kilos de charbon qui sont sur la pelle, les envoyer au fond
The same is true for spoken language. A search of the French corpus of C Oral Rom (Cresti / Moneglia 2005) for relative clauses with y avoir returned only 10 hits, most of them being amount relatives. Among the 10 examples, there is however also one example with a singular count noun as head of the relative clause (10).7
(10) / j’allais travailler là / je passais sous le pont / qu’il y a sous les escaliers de la gare /
In summary, the analysis of Dobrovie-Sorin / Beyssade (2004, 2012) accounts for the majority of the facts, but not for all the data. Existential y avoir inside relative clauses is seldom used, and mostly with nonspecific, non individual denoting nouns. But there are also some examples of modification of singular count nouns by relative clauses with y avoir. The alternative predicative structure with a relative clause formed with the copula être is much more common, and even used in contexts with existential meaning (9c). Bearing these French facts in mind, let us now turn to the Spanish data.
3. The Spanish data
Spanish offers a completely different picture. Although the Definiteness Effect is at work in Spanish, too, and is even stricter than in French (Leonetti 2008; Meulleman 2012), existential haber is easily found inside relative clauses. I used CORPES XXI for literary language and C Oral Rom for spoken language. As CORPES XXI covers all geographical varieties of Spanish and also contains data from other than literary texts, I restricted my search to a subcorpus, containing only written texts from Spain. This subcorpus had less than 67 million words, and was thus only half as big as Frantext catégorisé. However, searching for relative clauses with haber and its pivot as antecedent returned more than 1500 hits, i.e. the number of Spanish hits exceeded the French result by far. Searching for the alternative copula structure likewise gave more than 1500 results. The same phenomenon can be observed for spoken language. Whereas the French corpus in C Oral Rom only contained 10 relevant sentences, the Spanish corpus included 50 sentences, i.e. five times as much.8 Relevant relative clauses with estar were a ← 196 | 197 → bit less numerous than with haber in C Oral Rom, and amounted to 39. In summary, existentials inside relative clauses are much more frequent in Spanish than in French. Furthermore, the approximately equal amount of structures with haber and with estar indicates that these two structures are real alternatives to each other.
3.1 Relative clauses with haber
Among the occurrences of relative clauses with haber, there are examples of amount relatives parallel to the French examples cited above, either with explicit reference to an amount (11a), with a definite plural article (11b), or with a mass noun (11c):9
(11) a. Parecía un edificio de viviendas, pero muchas de ellas debían de estar destinadas a oficinas, como revelaba la gran cantidad de placas que había a un lado del portal
b. Gómez rebuscó entre los papeles que había encima de su mesa hasta que encontró un par de folios grapados
c. Con esta cocción tan lenta la grasa que hay debajo de la piel se funde
But there are also numerous examples of relative clauses that are headed by singular count nouns, both with indefinite (12a, b) and definite (12c, d) article:
(12) a. Volvió la cabeza y se detuvo un momento a curiosear en un corcho que había en la pared y en el que estaban clavadas un montón de fotos
b. La directora se quitó las gafas y las metió en un estuche de metal que había encima de la mesa
c. Toñín mira con ansiedad el pequeño aparato que hay sobre la mesa
d. Cogí la grapadora que había sobre su mesa y me puse a inspeccionarla
In all these examples, haber is used with a noun that refers to a specific, concrete object. An amount, property or kind reading seems to be excluded in these examples. This shows that Spanish haber, unlike English there be or French y avoir, is perfectly capable of appearing inside relative clauses headed by specific singular count nouns, as long as the relative clause is restrictive (cf. on this latter point Escandell / Leonetti 1998). In all these examples the relative clause serves to establish the reference of the antecedent (i.e. the pivot). Whereas the “normal” function of a presentational existential clause is the first mention of a new discourse referent, (such as a book in There’s a book on the table or, slightly modifying (12a), un corcho in Había un corcho en la pared), this first mention has already happened when the ← 197 | 198 → relative clause with haber occurs (e.g. by a verb of vision or perception like curiosear (12a) or mirar (12d)). The relative clause then specifies the location of the new discourse referent and thus serves to complete its discourse introduction.10 In order to get a more complete picture of the data, let us now consider parallel structures with estar instead of haber.
3.2 Relative clauses with estar
Estar can be found inside relative clauses with a large variety of antecedents: singular count nouns headed by a definite article (13a), an indefinite article (13b), or plural count nouns (13c, d).
(13) a. El cabo le indica amablemente el teléfono que está encima de su mesa
b. […] dijo apuntando a su casco, un Champion Universal de factura británica que estaba sobre la mesa
c. Luego clavó los ojos en unos papeles que estaban sobre la mesa, a su derecha
d. Barrió con el brazo las tazas y platos y tarros de mermelada que estaban sobre la encimera hasta hacer que saltaran al suelo y estallaran en mil pedazos
In declarative clauses, haber and estar are in complimentary distribution, depending on the definiteness of the noun (see among many others Bull 1943, Fernández Soriano / Táboas 1999: 1756, cf. also examples (1a, 3a) above). As we have seen in (11)–(13), the definiteness of the pivot noun phrase will certainly not be decisive for choosing which predicate appears inside the relative clauses. But are there other factors that could determine whether haber or estar is used? In some contexts, haber and estar appear to be in free variation inside relative clauses:
(14) a. Para aliviar el estrés, masajea la línea horizontal que hay debajo de toda la almohadilla. Corresponde a la zona del diafragma y ayuda a liberar tensiones
b. Masajea la línea que está debajo de los dedos antes de llegar al gordo. Ésta es la zona de los hombros.
Both examples are produced by the same author, in the same book, and appear in the same chapter, a passage on reflexology. They are only separated by few paragraphs, and their context is exactly the same. The sentences are thus almost identical, except that in one of them the author uses haber, in the other estar. This may happen for stylistic reasons (not to repeat the same linguistic structure over ← 198 | 199 → and over again), or just by chance because the author does not feel any difference between estar and haber in this context and in consequence does not pay much attention to the question. However, there are also subtle differences in the use of haber and estar inside relative clauses. With a human antecedent, for example, haber is not completely excluded but rare, and is only used with unspecific reference (15a). In the overwhelming majority of examples estar is used, both with specific or unspecific reference (15b, c).
(15) a. Sentir que todas las personas que hay alrededor te miran
b. Coged una copita de champán – dijo a los dos hombres que estaban junto a él
c. A veces, las personas que estaban dentro de la tienda llevaban ya un rato observándolos
Unlike haber, estar has also the capacity to appear inside appositive relative clauses (16) (cf. also Escandell / Leonetti 1998). Furthermore, estar rather than haber is used when the emphasis is on the exact location of the entity (17).
(16) La patrona alargó la mano, medio dormida, para descolgar el teléfono, que estaba sobre la mesilla
(17) Decidir puntos que están dentro o fuera de zonas poligonales dadas
On the other hand, haber is used for less important details. These are useful for fixing the reference, but do not present important or unexpected information. Reconsider (12a, b) from above (repeated here for convenience):
(12) a. Volvió la cabeza y se detuvo un momento a curiosear en un corcho que había en la pared y en el que estaban clavadas un montón de fotos
b. La directora se quitó las gafas y las metió en un estuche de metal que había encima de la mesa
Where else, if not on the wall, should a pin board with fotos be attached? Similar remarks hold for (12b). Un estuche de metal is rather an unimportant detail of the story, without further consequence for the action, and its location encima de la mesa is everything else than unexpected. Furthermore, there were no estar examples parallel to (11a), i.e. examples with explicit reference to an amount or with a quantitative term such as cantidad heading the relative clause. This use seems to be restricted to haber. Likewise, when the indefinite pronoun lo is the antecedent of the relative clause (this may be considered as the ultimate case of a non-profiled substance), mostly haber is used in the relative clause (18):
(18) De noche y desde fuera, no se distingue lo que hay en el salpicadero de un coche
The lines separating haber from estar in relative clauses are thus everything else than clear cut, although the communicative weight of the introduced entity partly ← 199 | 200 → explains the variation. Important, salient, or well profiled entities, e.g. human beings, are rather introduced using estar in the relative clause, whereas less important details and less profiled entities, such as mass nouns or the indefinite pronoun lo, tend to be introduced via a haber relative. As the numerous examples of almost identical contexts for one or the other variant show, this seems yet to be rather a question of stylistic effects than of grammatical constraints.11 In addition, emphasis on the exact location requires estar, whereas explicit reference to an amount demands a haber relative clause.
Summing up, haber and estar are almost completely interchangeable inside relative clauses. This result is unexpected, given that the Definiteness Effect is very strong in Spanish and the restriction to amount relatives observed in the literature on English and French should hold for Spanish as well. On the other hand, Escandell / Leonetti’s (1998) convincing explanation of the Spanish facts (see introduction) should easily be transferable to English and French, so that the question arises why y avoir or there be should not also be able to modify nouns in the same way as Spanish haber.
Moreover, the few French corpus occurrences with y avoir inside a relative clause modifying a singular count noun (see above) indicate already that it is not completely impossible to have y avoir inside relative clauses that are not of the extended amount relative type, but only rare or unusual.12 We thus have to ask why the structure is so unusual in French, or, put differently, what makes Spanish different. ← 200 | 201 →
4. An alternative explanation
The starting point of our explanation goes back to Keenan / Comrie (1977). They develop a cross-linguistic “Accessibility Hierarchy” of relativized positions and state that the subject position is most easily relativized cross linguistically. Further research on the issue has added robust empirical support to their thesis. Subject relatives appear to be not only cross-linguistically less marked, but also more frequent than object or oblique relatives. Both Reali / Christiansen (2007) and Duffield / Michaelis (2011) report high percentages (≥ 65 %) of subject relative clauses in English corpora. This frequency effect is valid for French and Spanish as well. We analysed for each language approximately 300 relative clauses, chosen randomly from the corpora Frantext and Corpes XXI, and counted 61 % subject relatives in French, and 67 % subject relatives in Spanish. A second case in point comes from acquisition data. Keenan / Comrie (1977: 89–90) already cite evidence from language production and comprehension studies that show an advantage of subject relative clauses, both in English and French. In a recent study, Ezeizabarrena (2012) confirms also for Spanish an earlier acquisition date of subject relatives. We thus assume both for French and Spanish a preference for subject relative clauses, with higher frequency and greater ease of comprehension and production.
In principle, this should be of no further concern to our study, because, according to traditional grammars, the pivots of both y avoir and haber are considered being the objects of the construction. Both the French and Spanish relative clauses under examination should thus be object relative clauses. In practice, however, Spanish relative clauses with haber share much more properties with subject relative clauses than French y avoir relative clauses do. We can see this clearly by comparing the relative clauses with the existential y avoir / haber (that should be object relative clauses) to the ones containing the copula être/ estar (that are subject relative clauses). French has different relative pronouns, according to the function inside the relative clause: subject relative clauses are introduced by the pronoun qui, object relative clauses by que. Furthermore, the existential y avoir requires in its conjugated form the expletive subject il. The French structure qu’il y a PP as compared to the subject relative clause qui est(sont) PP, is thus doubly marked for its object function: first by the form of the object relative pronoun, second by the expletive pronoun il occupying the subject position inside the relative clause.13 In contrast, Spanish que hay PP, in theory also an object relative ← 201 | 202 → clause, resembles much more the subject relative clause que está(n) PP.14 In both cases, the uniform relative pronoun que is used, and there is no expletive subject with haber. With singular pivots, there is thus no formal difference between the two structures. Only with plural pivots, there is agreement with estar, whereas haber as an impersonal construction should not display agreement (at least following normative grammars).15 However, the pivot of existential/presentative constructions in general (Koch 2003), and the pivot of haber in particular, are known to vacillate between object and subject status (for an overview see García Yebra 1983). Pluralization of haber in different varieties of Spanish is a recurrent subject of linguistic studies, and is ultimately interpreted as indicating a process of language change, where haber is seen no longer as impersonal, but as a personally construed verb with the pivot being the subject (among others Hernández Díaz 2006: 1146–1153, Waltereit / Detges 2008, Brown / Rivas 2012, Claes 2014, and the contributions to this volume).
In summary, speakers of Spanish use relative clauses with haber as a means of fixing the reference of new discourse referents, whereas relative clauses with y avoir are rather avoided in French. Instead, relative clauses with être or another verb (e.g. se trouver) are used, or there is even no relative clause at all. Comparing French and English versions of Spanish relative clauses with haber in the parallel corpora Open Books and Open Subtitles gives further insight in the issue. In most cases, the Spanish relative clause corresponds to a simple prepositional phrase (22b, c). In other cases, it corresponds to a relative clause with être/to be or a ← 202 | 203 → different verb (23b, c). Remarkably, this is nearly always a verb yielding a subject relative clause.16
(22) a. Lo metí en un vaso que había al lado del sofá y me di la vuelta
b. [Je] l’ai mis dans un gobelet près du canapé, et je me suis retourné
c. So I […] stuff it in a paper cup next to the sofa, and turn around
(23) a. -¡No os cortarán la cabeza! -dijo Alicia, y los metió en una gran maceta que había allí cerca
b. «Vous ne serez pas décapités,» dit Alice ; et elle les mit dans un grand pot à fleurs qui se trouvait près de là
c. ‘You shan’t be beheaded!’ said Alice, and she put them into a large flower-pot that stood near
Even in cases with a clear specification of amount, speakers of English and French do not use an amount relative with y avoir or there be but prefer a subject relative (qui existe) or an infinitive (to be found) (24):
(24) a. Y persuadiré al sensible Aquiles… con ciertos hechos fascinantes… sobre la gran cantidad de oro que hay en Troya
b. J’ai des arguments… tels que la quantité d’or qui existe à Troie
c. And I will persuade the sensitive Achilles…… with certain fascinating facts…… about the great store of gold to be found in Troy
French y avoir and English there be have been claimed to allow relativization of their pivots only in so called amount relative clauses (Carlson 1977, Dobrovie-Sorin / Beyssade 2012), or, extending the original analysis, with property or kind denoting head nouns (Dobrovie-Sorin / Beyssade 2004, McNally 2008). In both cases, the head nouns comply with the Definiteness Restriction. Furthermore, relativization out of existentials seems to be rare, at least in French (Dobrovie-Sorin / Beyssade 2004: 90).17 When we transfer these analyses to Spanish haber, we would expect similar results, because Spanish haber also obeys the Definiteness Restriction. However, this is not the case. First of all, Spanish relative clauses with haber are much more frequent than their French counterparts. Second, they are ← 203 | 204 → in no way restricted to property denoting antecedents but combine easily with all sorts of head nouns, especially singular count nouns. Escandell / Leonetti (1998) develop a convincing explanation, why this is possible (despite the strong Definiteness Effect in Spanish). A priori, there is no reason why this explanation should not also work for French or English, i.e. why French y avoir or English there be should exclude singular count nouns as antecedents in such a categorical way as is assumed in the literature. And indeed, our corpus study in Frantext catégorisé and C-Oral Rom revealed some examples of singular count nouns figuring as heads of relative clauses with y avoir. Rather than being completely ungrammatical, the structure thus seems to be only rare.
We relate the differences between French and Spanish in using y avoir and haber in restrictive relatives to a blocking effect on y avoir relatives by the cross linguistic preference for subject relatives. Whereas the pivots of both y avoir and haber should be objects according to normative grammars, haber is much more open for reanalysis than y avoir. Especially in relative clauses, the object marking of relative clauses with haber is less explicit than with y avoir. Taking the strong cross linguistic preference for subject relative clauses into account, y avoir is thus handicapped by its clearer status as object relative and gives a rather clumsy expression in relative clauses. Similar remarks may apply to English there be. While the verb in English there be agrees with the pivot,18 and the pivot could thus be interpreted as subject, there is still the expletive pronoun there, filling the canonical subject position, and complicating the relative clause “unnecessarily” (as compared to a simple copula relative clause). Haber relative clauses, in contrast, can quite easily be interpreted as subject relative clauses, and thus compete freely with estar relative clauses. Stating the location of a newly introduced entity is achieved by different means in the three languages. French and English use a prepositional phrase, or a subject relative clause with être/to be or another verb. Relative clauses with y avoir / there be are rare. Spanish easily uses relative clauses with haber to complete the introduction of new discourse referents by indicating their location. Thanks to their similar structure, haber and estar compete with each other and can specialize: haber to the unmarked case, where neither the newly introduced entity nor its location is important or unexpected, estar to the marked case, with special emphasis to the referent or its location. ← 204 | 205 →
Brown, Esther L. / Rivas, Javier (2012): “Grammatical relation probability: how usage patterns shape analogy”, Language Variation and Change 24, 317–341.
Bull, William (1943): “Related Functions of Haber and Estar”, The Modern Language Journal 27:2, 119–123.
Carlson, Gregory (1977): “Amount relatives”, Language 53, 520–542.
Claes, Jeroen (2014): “A Cognitive Construction Grammar approach to the pluralization of presentational haber in Puerto Rican Spanish”, Language Variation and Change 26, 219–246.
Clark, Eve (1978): “Locationals: existential, locative and possessive constructions”, in Joseph Greenberg (ed.), Universals of Human Language, Stanford: Stanford University Press, 85–126.
CORPES XXI = Real Academia Española: Corpus del Español del Siglo XXI (CORPES) [online], <http://www.rae.es>.
Cresti, Emanuela/ Moneglia, Massimo (2005): C-ORAL-ROM. Integrated Reference Corpora for Spoken Romance Languages, Amsterdam / Philadelphia: John Benjamins.
DeMello, George. (1991): “Pluralización del verbo haber en el español hablado culto de once ciudades”, Thesaurus 46, 445–471.
Dobrovie-Sorin, Carmen / Beyssade, Claire (2004): Définir les indéfinis, Paris: CNRS.
Dobrovie-Sorin, Carmen / Beyssade, Claire (2012): Redefining Indefinites, Dordrecht: Springer.
Duffield, Cecily / Michaelis, Laura (2011): “Why subject relatives prevail: constraints versus constructional licensing”, Language and Cognition 3, 171–208.
Escandell Vidal, María Victoria / Leonetti, Manuel (1998): “Construcciones existenciales y oraciones de relativo”, in Giovanni Rufino (ed.), Atti del XXI Congresso Internazionale di Linguistica e Filologia Romanza, vol. II, Tübingen: Niemeyer, 261–272.
Ezeizabarrena, Maria José (2012): “Children do not substitute object relatives with subject relatives in every romance language: the case of Spanish”, Revue Roumaine de Linguistique 57, 161–181.
Fernández Soriano, Olga / Táboas Baylín, Susana (1999): “Construcciones impersonales no reflejas”, in Ignacio Bosque and Violeta Demonte (eds.), Gramática descriptiva de la lengua Española, vol. II, Madrid: Espasa, 1723–1778.
Frantext = ATILF-CNRS / Université de Lorraine: Base textuelle FRANTEXT [online], <http://www.frantext.fr>. ← 205 | 206 →
Freeze, Ray (1992): “Existentials and other locatives”, Language 68, 553–595.
García Yebra, Valentín (1983): “¿Complemento directo o sujeto con las formas unipersonales de haber?”, Revista de Filología Española 63, 33–71.
Hernández Díaz, Axel (2006): “Posesión y existencia. La competencia de haber y tener y haber existencial”, in Concepción Company (ed.), Sintaxis histórica de la lengua española, vol. I:2, México: UNAM / FCE, 1055–1160.
Keenan, Edward L. / Comrie, Bernard (1977): “Noun phrase accessibility and universal grammar”, Linguistic Inquiry 8, 63–69.
Koch, Peter (2003): “From subject to object and from object to subject: (de)personalization, floating and reanalysis in presentative verbs”, in Giuliana Fiorentino (ed.), Romance languages. Transitivity in Romance languages, Berlin / New York: De Gruyter, 153–185.
Koch, Peter (2012): “Location, existence, and possession: a constructional-typological exploration”, Linguistics 50, 533–603.
Leonetti, Manuel (2008): “Definiteness effects and the role of the coda in existential constructions”, in Henrik Høeg-Muller (ed.), Essays on nominal determination: from morphology to discourse management, Amsterdam / Philadelphia: John Benjamins, 131–162.
Lyons, John (1967): “A note on possessive, existential and locative sentences”, Foundations of Language 3, 390–396.
McNally, Louise (2008): “DP-internal only, amount relatives, and relatives out of existential sentences”, Linguistic Inquiry 39, 161–169.
McNally, Louise (2011): “Existential Sentences”, in Klaus von Heusinger et al. (eds.), Semantics, Berlin / New York: De Gruyter, 1829–1848.
Meulleman, Machteld (2012): “Degrees of grammaticalization in three Romance languages: a comparative analysis of existential constructions”, Folia Linguistica 46, 417–451.
Milsark, Gary (1977): “Toward an explanation of certain peculiarities in the existential construction in English”, Linguistic Analysis 3, 1–30.
Partee, Barbara / Borschev, Vladimir (2007): “Existential sentences, BE, and the genitive of negation in Russian”, in Ileana Comorovski and Klaus von Heusinger (eds.), Existence: Semantics and Syntax, Dordrecht: Springer, 147–190.
Reali, Florencia / Christiansen, Morten (2007): “Processing of relative clauses is made easier by frequency of occurrence”, Journal of Memory and Language 53, 1–23.
Tiedemann, Jörg (2012): “Parallel Data, Tools and Interfaces in OPUS”, in Proceedings of the 8th International Conference on Language Resources and Evaluation (LREC’2012) [online], <http://opus.lingfil.uu.se>.
Waltereit, Richard / Detges, Ulrich (2008): “Syntactic change from within and from without syntax: a usage-based analysis”, in Ulrich Detges and Richard Waltereit (eds.), The paradox of grammatical change: perspectives from romance, Amsterdam / Philadelphia: John Benjamins, 13–30.
1 ATILF-CNRS / Université de Lorraine: Base textuelle FRANTEXT [online], <http://www.frantext.fr>. Frantext catégorisé (a subcorpus of Frantext) is tagged for part of speech information and thus offers better search options than Frantext.
4 Cf. for a parallel case also Dobrovie-Sorin / Beyssade (2004: 91).
5 This is not a typo: “non particulier” is correct here (the text should not read “nom particulier”), because the authors refer to non individual variables, not to proper names.
6 This is all the more remarkable, because according to Carlson (1977: 525) and McNally (2008: 162; although McNally nuances the issue a bit), NPs with indefinite article should not be possible at all as heads of existential relative clauses.
7 According to the C Oral Transcription conventions, / signifies “non conclusive prosodic break” (Cresti / Moneglia 2005: 26). There occurs a prosodic break directly before the relative pronoun in example (10). As this sentence is the only example of this construction in the whole corpus, I do not want to draw any conclusions on that.
8 One of the main concerns of the creation of the C Oral Rom corpus was to provide comparable corpora (Cresti / Moneglia 2005: 1), such that we can directly compare the numbers for French and Spanish.
9 The examples are all taken from CORPES XXI, but the results hold for spoken language as well.
10 Often, the relative clause is necessary to justify the definite article. A modified version of (12d) without the restrictive relative clause but with the definite article Cogí la grapadora y me puse a inspeccionarla sounds strange because la grapadora has not been mentioned in the preceding context, and is thus not identifiable for the hearer.
11 Suñer (1982: 324–339) discusses another subtle difference of haber and estar, namely a depictive ‘on-stage’ effect when estar is used in declarative clauses with indefinite subjects (e.g. En la esquina hay un vigilante nuevo vs. En la esquina está un vigilante nuevo) and comments: “If the notion of markedness were duly exploited in syntax-semantics, the hay sentences would represent the unmarked construction and the estar sentences would be the marked member of the pair.” In our corpus examples, there were both relative clauses with depictive uses of estar and haber, as well as relative clauses with non-depictive uses of both verbs, such that Suñer’s distinction seems to be relevant only for declarative clauses. The idea of haber being the unmarked member of the haber/estar pair, however, fits nicely our observations concerning the communicative weight of the antecedent and the preference for haber in unimportant (~unmarked) contexts.
12 Similarly, McNally (2008) argues that there be should not be restricted to the sole class of amount relatives neither.
13 Similar remarks may apply to English with the expletive there.
14 Cf. also Ezeizabarrena (2012: 162–167) for a nice overview of ambiguous subject/object relative clauses in Spanish.
15 There are some few occurrences of pluralized haber in relative clauses headed by plural nouns in CORPES XXI (i, ii) and one example in C Oral Rom (iii). The examples in CORPES XXI were nearly all from American Spanish, and most of them were press texts, not literature. The spoken example was produced by a speaker of South American origin (Chile). The forms of haber were either imperfect or future, complying with observations from the literature that haber-pluralization occurs most easily in these forms and mainly in American Spanish (DeMello 1991).
i) Si deseas tener más información sobre todas las presentaciones que habrán en Lima en los próximos meses, visita nuestro blog En vivo y en concierto
ii) Las pocas plantas que habían en la Unidad ya se habían marchitado
iii) Porque / se cortaban / unos eucaliptus grandes que habían en [/] en la / cuestión así de la casa /
16 The original versions of most of the texts are English, which means that mostly both French and Spanish sentences are translations. This shows that Spanish translators, instead of just copying the original syntactic structure, deliberately chose to use the relative clause with haber, a fact that emphasizes even more the Spanish preference for relative clauses with haber.
17 Neither Carlson (1977) nor McNally (2008) address this issue for English.
18 At least in standard varieties.