«¡La de + N + que…!» The Feminine Definite Article in Spanish Exclamative Clauses

Abstract

The present study explores the exclamative use of the feminine definite article la in structures such as ‘¡La de chicos que besé en la fiesta!’ (How many guys I kissed at the party!). First, a morphosyntactic description of the pattern is offered so as to show that the data under analysis are pseudopartitive constructions which display all the characteristics of primary and partial exclamatives. Building on research on nominal exclamatives, we conclude that these examples are not CPs but indefinite DPs with an exclamative flavor which contain a semi-relative clause introduced by que. Within the framework of Distributed Morphology, we schematize a set of syntactic structures which capture the ‘chimeric’ and hybrid nature of the data, these being halfway between DPs and exclamative clauses. In order to do so, it will be necessary to split the DP into smaller projections (FocP, FinP), since la must move to Spec-FocP to be interpreted as an exclamative operator.

1. Introduction

Among the set of constructions that can be used in General Spanish to express quantitative exclamations (1), it is possible to find the pattern illustrated in (2), which is introduced by the feminine definite article la and the preposition de, followed by an unbounded (plural or uncountable) noun:

(1)	a.	¡Cuántos	chicos	que	besaste	en la fiesta!
		How many.Masc.PL.	boys	that	kissed.2.SG.	in the party
	b.	¡Qué	de	chicos	que	besaste		en la fiesta!
		What	of	boys	that	kissed.2.SG.		in the party
	c.	¡Qué	cantidad	de	chicos	que	besaste	en la fiesta!
		What	quantity	of	boys	that	kissed.2.SG	in the party
	d.	¡Los	chicos	que	besaste	en la fiesta!
		The	boys	that	kissed.2.SG.	in the party
		‘How many guys you kissed at the party!’

(2)	a.	¡La	de	chicos	que	besaste	en la fiesta!
		Art.Fem.Sg	of	boys	that	kissed.2.SG.	in the party
	b.	¡La	de	vino	que	tomaste	en la fiesta!
		Art.Fem.Sg	of	wine	that	drank.2.SG.	in the party
		‘How much wine you drank at the party!’

Interestingly, there is an equivalence in meaning with the examples in (1) in that they license quantitative exclamative interpretations, roughly parallel to ‘the x number/amount of boys/wine,’ the x corresponding to ‘a high degree or quantity.’ Despite this parallelism, the data in (2) are more colloquial and have a less standard status with respect to more canonical exclamatives such as (1a). The crux of the structures in (2) lies in the fact that the feminine article does not precede any noun, nor does it agree with the unbounded noun in adjacency to the preposition de. According to RAE-ASALE (2009, §42.16), the pattern is highly productive and can be found in several varieties of Spanish, although it is relatively recent, considering it was first attested in literary texts at the end of the nineteenth century. Notwithstanding its productivity, with the exception of some occasional references in works on exclamation (Alonso Cortés 1999; Bosque 2017; Gutiérrez-Rexach and Andueza 2017; RAE-ASALE 2009; etc.), the structure has not been explored in depth in the existing current literature.

In the following pages, we will provide a thorough description of this peculiar construction and look into its morphosyntactic nature in an endeavor to contribute to our understanding of less canonical phenomena of linguistic variation in the domain of exclamation. Within the framework of Generative Grammar, more specifically within the Distributed Morphology model (Halle and Marantz 1993, et seq.), the purpose of this paper is to consider the hypothesis that the data under scrutiny are pseudopartitive nominal constructions in which la has become an exclamative, evaluative and existential quantifier equivalent to cuánto/s (‘how many/much’). As opposed to what the standard analyses of this pattern propose (e.g., Bosque 2017; Brucart 1999; RAE-ASALE 2009; etc.), it will be argued that the structures in (2) are not CPs but DPs, whose exclamative flavor arises from the formal requirements imposed by the functional head Focº, the same head which is responsible for the movement of the exclamative quantifier to the edge of the CP in more canonical exclamative clauses such as (1a–c). As will be discussed below, an account in these terms implies drawing a parallelism between the structure of the DP and the CP and, therefore, splitting the DP into smaller projections (Aboh 2004; Roca 2009; Ticio 2003; etc.). In relation to the nominal nature of the data, we will demonstrate that the clause headed by que is a relative clause, akin to those found in cleft sentences.

The paper is organized as follows. Section 2 offers a description of the data focusing on their pseudopartitivity and nominal status (Section 2.1), their exclamative properties (Section 2.2) and the relative clause we propose they contain (Section 2.3). In Section 3, we outline some of the basic theoretical assumptions underlying our research, paying special attention to those pertaining to the internal structure of the DP and the left periphery, which will allow us then to put forward a morphosyntactic analysis of the phenomenon under study in Section 4. Finally, Section 5 includes the conclusions and the implications of our study in our understanding of exclamation in the field.

2. A Chimeric Nominal–Clausal Construction

Ever since Chomsky’s (1970) seminal paper on nominalizations, there have been numerous authors who have observed that there exists a parallelism between the nominal and verbal/clausal domains (Brame 1982; Abney 1986; etc.). More specifically, there are a set of constructions in English (3) and Spanish (4) which have been studied under this logic, by virtue of the fact that, even though their syntactic structure is superficially nominal, they appear to encode meanings typically associated with exclamative clauses (surprise, elation, disbelief, etc.). For this reason, Portner and Zanuttini (2005) dub these structures nominal exclamatives.

(3)	a.	The strange things that he says!
	b.	The things that are in this bag!
	Portner and Zanuttini (2005)

(4)	a.	¡La cerveza que consume la gente en este país!
		‘The beer that people drink in this country!’
	b.	¡La gente de El Salvador que ha muerto en esta guerra!
		‘The people from El Salvador that have died in this war!’
	Torrego (1988)

As will be confirmed after the discussion held in the ensuing paragraphs, the pattern «La de...» turns out to be a fertile territory to study the overlap between these functional categories (DPs and CPs). In this section, we will set out to offer a description of the phenomenon, which will lead us to the conclusion that the data in (2) are halfway between nouns and clauses and can therefore be regarded as nominal exclamatives. To that end, first, in Section 2.1 we will describe the pseudopartitivity of the construction to then demonstrate that the examples exhibit all the characteristics of partial and primary exclamative clauses. The third subsection focuses on the clause introduced by que and develops arguments in favor of the hypothesis that its syntactic structure is that of a relative clause.

2.1. Pseudopartitivity and Nominal Status

Pseudopartitive constructions (5) are made up of two parts linked by the functional preposition de: (a) the head, which includes a quantifier or quantifying expression; (b) the tail, which contains an unbounded bare plural or uncountable noun phrase related to the quantification in (a). Semantically speaking, they introduce one single set, corresponding to the bare nominal, the head denoting a measure or quantity of the N in the tail (for a more detailed description of pseudopartitives see Brucart 1997; Di Tullio and Kornfeld 2012; Mare 2016; RAE-ASALE 2009; among others).

(5)	a.	[Un grupo]HEAD de [estudiantes]TAIL
		‘A group of students’
	b.	[Un poco]HEAD de [cerveza]TAIL
		‘A bit of beer’
	c.	*[Un grupo]HEAD de [los estudiantes]TAIL
		‘A group of the students’
	d.	*[Un poco]HEAD de [persona]TAIL
		‘A bit of person’

As clearly shown in (6) below, «La de…» behaves like a pseudopartitive in that the inclusion of non-bare plural or singular count nouns gives rise to ungrammatical sentences. This means that the noun in the tail must necessarily be unbounded.

(6)	a.	¡[La]HEAD	de	[estudiantes]TAIL	que	vino a la fiesta!
		Art.Fem.Sg	of	students	that	came to the party
		‘How many students came to the party!’
	b.	¡[La]HEAD	de	[cerveza] TAIL	que	tomamos	anoche!
		Art.Fem.Sg	of	beer	that	drank	last night
		‘How much beer we drank last night!
	c.	*¡[La]HEAD	de	[los estudiantes]TAIL	que	vino a la fiesta!
		Art.Fem.Sg	of	the.Pl.Masc students	that	came to the party
		‘*How many the students came to the party!’
	d.	*¡[La]HEAD	de	[persona]TAIL	que	vino a la fiesta!
		Art.Fem.Sg	of	person	that	came to the party
		‘*How many person came to the party!’

Notice, however, that there seem to be two apparent differences between (5) and (6). The first one has to do with the fact that the head of the data in (6) does not appear to contain a quantifier noun, as is the case in (5) with the Ns grupo or poco. This is because the pattern under study emerged from the elision of the feminine noun cantidad (‘quantity’) (Bosque 2017; Brucart 1999; Fernández Ramírez 1951; RAE-ASALE 2009; Roca 2009; Torrego 1988; etc.), which eventually resulted in the grammaticalization of the article and its subsequent reanalysis as a weak evaluative quantifier tantamount to mucho (‘much/many’), given that the reading it yields suggests excess of quantification from the perspective of the speaker (7) (see Section 4). Contrary to other nominal exclamatives (8), the pattern rejects qualitative interpretations; that is, a sentence such as (7b) could never be paraphrased as ¡Qué cosas tan raras por las que se preocupa Mario! (‘How strange the things that Mario worry about!’).

(7)	a.	¡La (cantidad) de cerveza que tomaste en la fiesta!
		=Tomaste mucha cerveza en la fiesta.
		‘You drank too much beer at the party!’
	b.	¡La (cantidad) de cosas por las que se preocupa Mario!
		=Mario se preocupa por muchas cosas.
		‘Mario worries about too many things!’

(8)	¡Las cosas por las que se preocupa Mario!
	Quantitative reading: ‘Mario worries about too many things!’
	Qualitative reading: ‘How strange the things that Mario worry about!’

The second difference between (5) and (6) concerns the definiteness of the structures. A distinguishing feature of pseudopartitives is that they host indefinite nouns or quantifiers (e.g., un centenar de libros, un poco de pan, un grupo de manifestantes, etc.). In spite of the fact that the head of the examples in (6) is the definite article la, the reading we obtain is evidently existential and indefinite, as proven by the paraphrases with quantifiers like mucho (‘too much/many’) in (7) and by the following sentences, in which definiteness effects are not available when combined with the existential verb haber to express a locative meaning:1

(9)	a.	¡La de estudiantes que hay en la nueva universidad!
		Hay muchos estudiantes en la nueva universidad.
		‘There are many students in the new university’
	b.	¡La de libros que hay sobre el escritorio!
		Hay muchos libros sobre el escritorio.
		‘There are many books on the desk’

As is the case with the pseudopartitive constructions in (10), the data analyzed in this paper give rise to alternating agreement patterns, as the verb in the relative clause might agree with the head of the structure or with the unbounded noun in the tail, in which case there would be ad sensum agreement or silepsis.

(10)	a.	Un	grupo de alumnos	trabaja/trabajan	en este bar
		A	group of students	works/work	in this bar
		‘A group of students works/work in this bar’
	b.	Un	montón de personas	visitó/visitaron	las Cataratas del	Iguazú
		A	lot of people	visited.3.Sg/visited.3.Pl	the Falls  of the	Iguazu
		‘A lot of people visited the Iguazu Falls’

(11)	a.	¡La	de	alumnos	que	trabaja/trabajan	en este bar!
		Art.Fem.Sg	of	students	that	works/work	in this bar
	b.	¡La	de	personas	que	visitó/visitaron	las Cataratas del Iguazú!
		Art.Fem.Sg	of	people	that	visited.3.Sg/visited.3.Pl	the falls  of-the Iguazu

Lastly, the examples in (12) and (13) indicate, on the one hand, that «La de...» forms a constituent that can be coordinated and, on the other, that when it comes to agreement, the N in the tail is the element with which the predicative adjectives after a copula agree.

(12)	¡Madre mía, la de gilipollas y la de gilipolleces que tengo que aguantar!
	‘Oh, my god, how many stupid people and things I have to tolerate!’

(13)	a.	¡La	de	alumnos/as	que	están desempleados/as!
		Art.Fem.Sg	of	students.Masc.Pl/Fem.Pl	that	are unemployed.Masc.Pl/Fem.Pl
	b.	*¡La	de	alumnos/alumnas	que	está desempleada!
		Art.Fem.Sg	of	students.Masc.Pl/Fem.Pl	that	are unemployed.Fem.Sg

Crucially, the ungrammaticality of (13b) appears to point to the lack of referentiality of la, which is in accordance with the hypothesis we will further advance in Section 4 that it is a quantifier. In fact, as an anonymous reviewer observes, (13b) contrasts sharply with the grammaticality of the pseudopartitive in (14), where cantidad seems to be a full N, with referential properties, equivalent to cifra ‘figure.’

(14)	¡La cantidad de alumnos que se han inscripto en media hora!
	‘The number of students that have enrolled in half an hour!’

2.2. Exclamation and Clausal Status

One of the key ingredients that characterizes the pattern «La de...» is its exclamative import. Exclamatives are expressive speech acts that manifest an emotional reaction on the part of the speaker, such as surprise, disbelief or elation. In Zanuttini and Portner’s (2003) terms, a sentence as that in (15) implies that the extent or degree to which the mass noun cerveza is predicated exceeds or outranks the range of expected possibilities, which is the reason why, according to the authors, there is a widening effect. The pseudopartitive construction then introduces a variable whose domain includes a set of non-standard propositions, i.e., of all the possible values that the variable X could obtain, the DP belongs to the subset of the least expected ones, to that which can be located in the extremes of implicit pragmatic scales of standardness, expectation or plausibility.

(15)	¡La de cerveza que tomaste en la fiesta!
	‘How much beer you drank at the party!’

The widening of the quantificational domain is closely intertwined with the fact that exclamative sentences generate conventional scalar implicatures in observance with their impossibility of occurring in interrogative or negative sentences (Portner and Zanuttini 2005). This is attributable to the concomitance of exclamation and factivity, as the speaker presupposes the truth of the proposition which the exclamative utterance encodes.

(16)	a.	¡Es increíble la de cerveza que tomaste en la fiesta!
		‘It’s incredible how much beer you drank at the party!’
	b.	*¡No es increíble la de cerveza que tomaste en la fiesta!
		‘*It’s not incredible how much beer you drank at the party!’
	c.	*¿¡Es increíble la de cerveza que tomaste en la fiesta!?
		‘*Is it incredible how much beer you drank at the party!?’

According to Bosque’s (2017) classification of exclamative constructions, the structures under consideration fall within the group of primary exclamatives, given that there is an explicit syntactic or lexical clue that signals its illocutionary force. Said clue is the article la, which, as will be discussed in the next section, behaves like an exclamative operator equivalent to cuántos. Furthermore, our data can be considered definite article degree exclamatives, sometimes also called degree relatives or exclamatives with emphatic articles, such as those illustrated below in (17).2 In both cases, definite articles are used for emphasis.

(17)	a.	¡Los (incontables) sitios que ha visitado este hombre!
		‘The (innumerable) places that this man has visited!’
	b.	¡Lo (muy) inteligente que es María!
		‘How very clever Maria is!’
	c.	¡Lo (increíblemente) rápido que va este coche!
		‘How (incredibly) fast this car runs!’
		Bosque (2017, p. 22)

The application of other typical diagnostics in the literature (Bosque 2017; Masullo 2017; etc.) that enable us to tell exclamatives apart is also possible when it comes to «La de…» structures. In addition to sharing the semantic, pragmatic and prosodic properties of other prototypical exclamative constructions, the pattern with the feminine article under study: (i) can be embedded as the complement of verbs such as ver, mirar, fijarse and non-factive volitive verbs such as encantar/sorprenderse (18); (ii) cannot cooccur with other elements which, according to the standard treatment in the bibliography, occupy the same syntactic position (i.e., specifier of FocP or CP) (19); (iii) cannot be used in sentences whose illocutionary force is jussive, dubitative or desiderative (20);3 (iv) admits a resultative or consecutive sequel or coda (21); (v) can function as the argument of predicates such as increíble, both as subjects and as the adjacent adjectival complement (22).4

(18)	Prototypical exclamatives:
	a.	¡Mirá cuánta gente que vino a la fiesta!
		‘Look how many people came to the party!’
	b.	*¡Me temo cuánta gente que vino a la fiesta!
		‘*I’m afraid how many people came to the party!’
	«La de…» exclamatives:
	c.	¡Mirá la de gente que vino a la fiesta!
		‘Look how many people came to the party!
	d.	*¡Me temo la de gente que vino a la fiesta!
		‘*I’m afraid how many people came to the party!’

(19)	Prototypical exclamatives:
	a.	*¿¡Qué vino cuánta gente que compró!?
		‘*What a wine how many people bought!?’
	«La de…» exclamatives:
	b.	*¿Qué chicos la de gente que vieron en el parque!?
		‘*What boys how many people saw in the park!?’

(20)	Prototypical exclamatives:
	a.	*¡Tal vez cuánta gente que vino!
		‘*Maybe how many people came!’
	«La de…» exclamatives:
	b.	*¡Tal vez la de gente que vino!
		‘*Maybe how many people came!’

(21)	Prototypical exclamatives:
	a.	¡Qué caro que está todo que somos los únicos que tienen auto!
		‘How expensive everything is that we are the only ones with a car!’
	«La de…» exclamatives:
	b.	¡La de gente que vino que nos quedamos sin alcohol a la medianoche!
		‘So many people came that we’ve run out of alcohol at midnight!’

(22)	Prototypical exclamatives:
	a.	¡Es increíble cuánta gente vino a la fiesta!
		‘It’s incredible how many people came to the party!’
	«La de…» exclamatives:
	b.	¡Es increíble la de gente que vino a la fiesta!
		‘It’s incredible that so many people came to the party!’
	c.	¡La de gente que vino a la fiesta es increíble!
		‘That so many people came to the party is incredible!’

The empirical evidence offered above bears out the hypothesis that the structures introduced by «La de…» are analogous to true wh- exclamatives. They do not constitute ‘covert’ (Masullo 2017) or ‘total’ or ‘secondary’ exclamative clauses (i.e., structures whose intonation and proper interpretation of the exclamative illocutionary force associated with it are the linguistic markers of exclamative import (Bosque 2017, p. 7)) but rather structures whose semantic, syntactic, pragmatic and prosodic features coincide with the so-called genuine or partial exclamatives, i.e., those which contain an overt exclamative operator.

2.3. A ‘Relative’ Relative Clause

In the previous subsection, we highlighted the resemblance between «La de…» and other types of partial exclamatives. A dissimilarity which was not mentioned, however, is that in the «La de…» pattern the presence of que (‘that’) is mandatory (23), while in (24) it is optional. In our view, this discrepancy stems from the hybrid and chimeric status of the construction, because, even though it displays all the characteristics of exclamative clauses, its syntactic structure is that of a DP.

(23)	¡La	de vino *(que)	tomamos	anoche!
	Art.Fem.Sg	of wine that	drank	last night
	‘How much wine *(that) we drank last night!

(24)	¡Cuánto	vino	(que)	tomamos	anoche!
	How-much	wine	that	drank	last night
	‘How much wine (that) we drank last night!

Some authors such as Bosque (2017), Brucart (1999), RAE-ASALE (2009) or Masullo (2012) contend that the underlying configuration of definite article degree exclamatives corresponds to a clause. From this perspective, la de vino would be the direct object of the verb tomamos, and would be displaced to the initial position by means of a process of prolepsis or operator movement. Accordingly, the word que (‘that’) would be a complementizer or subordinating conjunction, as the form que in (24) is generally analyzed (see Brucart 1999; Plann 1984, etc.). In this paper, and along the lines of Torrego’s (1988) and Portner and Zanuttini’s (2005) proposals, we will defend the divergent view that the string is a DP that contains a subordinate relative clause, which is most of the times introduced by the relativizer que, although other relativizers are possible. We will now proceed to outline the arguments in favor of this analysis.

First and foremost, the strongest piece of evidence to support our hypothesis is that it is possible to find examples in which the pseudopartitive construction is followed by a clause headed by other simple relativizers—e.g., cuyo (‘whose’) and donde (‘where’) (25a–b)—or else by complex relativizers with prepositions, as in (25c–d).5 Note the correspondence with the sentences in (26), which include canonical relative clauses.

(25)	a.	La de gente cuya vida no vale nada
		‘How many people whose life is worthless!’
	b.	¡Increíble la de lugares donde aún se permite fumar en el interior de los bares!
		‘It’s incredible the number of places where it’s allowed to smoke indoors!’
	c.	¡La de chongos con los que chateé anoche en Scruff!
		‘The guys with whom I chatted yesterday on Scruff!’
	d.	¡¡La de sitios en los que has estado en estos 30 años de profesión!!
		‘The places in which I’ve been in my 30-year career!’

(26)	a.	Me entristece la gente cuya vida no vale nada
		‘I’m sad about people whose life is worthless.’
	b.	Han multado los lugares donde aún se permite fumar en el interior de los bares.
		‘They have fined the place where it’s allowed to smoke indoors.’
	c.	Los chongos con los que chateé anoche en Scruff me bloquearon.
		‘The guys with whom I chatted yesterday on Scruff blocked me.’
	d.	Le mostré los sitios en los que he estado en estos 30 años de profesión.
		‘I showed them the places in which I’ve been in my 30-year career’

Additional evidence comes from case-assignment phenomena. In some Romance languages, relative clauses have been shown to display some of the possibilities of case assignment found in long- and short-distance syntactic dependencies, be it upward or downward case attraction. Consider the following paradigm taken from Agulló (2023).

(27)	a.	Hay en sitios que se hace una morcilla un poco más dulce (COSER, Etxauri, Navarra)
		‘There are places in which a sweeter blood sausage is made’
	b.	Hay a gente que la echan por cualquier tontería (CORPES XXI, PRESEGAL, Spain).
		‘There are people who are fired for nothing’
	c.	Hay a personas que hay que reñirlas más (COSER, Cigales, Valladolid).
		‘There are people that should be told off more frequently’

In the examples in (27), the underlined prepositions appear after the existential verb haber in an unexpected position, different from the one in which they originate. The locative preposition en (27a), the differential object marker (27b) and the dative preposition a (27c) belong to the CP introduced by que but are spelled out as though they were case marking the pivot of the existential clauses, which, in Spanish, are known to reject DOM, e.g., Hay (*a) muchos edificios nuevos en Buenos Aires; ‘There are many new buildings in Buenos Aires.’ These data can lend support to the claim that the sequence headed by que in the structures under study is a relative clause, as upward case attraction can also be found.6

(28)	a.	¡La de gente a la que has visto!
	b.	¡A la de gente que has visto!
	c.	¡A la de gente a la que has visto!
	d.	¡La de gente que has visto!
		‘How many people you’ve seen!’

In (28a), the DOM preposition of the verb visto appears in its default position within the relative clause, pied-pied by the complex relative la que. In (28b) and (28c), it seems to be case marking the whole DP, occupying the upmost position, although in the latter a copy of the preposition remains in the default position. While a clausal analysis might account for the position of the preposition in (28b), by proposing that a la de gente is the DO of visto and is displaced to the left periphery of the clause, it fails to explain why it might occur before the apparent complementizer and not before the DO in (28b), or in both positions (in Spec-CP and Cº) (28c) or, fundamentally, why it most frequently does not appear altogether, as in (28d). The non-occurrence of the DOM preposition typically found with verbs like ver ‘see’ (29a) follows from the fact that the relativizer que does not need to be case marked (29b), as opposed to complex relativizers (29b), which indeed do. This paradigm can elegantly explain the behavior of a in (28), while a clausal analysis, to the best of our knowledge, would encounter several difficulties in doing so.

(29)	a.	Vi		a	los chicos	en la fiesta.	Son geniales
		Saw		DOM	the guys	at the party	Are great
		‘I saw the guys at the party. They are awesome.
	b.	Los chicos	(*a)	que	vi	en la fiesta	son geniales
		The guys	DOM	that	saw	at the party	are great
	c.	Los chicos	*(a)	los que	vi	en la fiesta	son geniales
		The guys	DOM	The that	saw	at the party	are great
		‘The guys I saw at the party are awesome’

Examples such as (30) with reduced relative clauses—that is, with non-finite verbs—are also attested. This specific example, moreover, indicates that the pattern can be preceded by the preposition con (‘with’), which definitely requires a nominal complement, as shown in (31).

(30)	Te	meás	en verdad	con	la	de	gente	estudiando
	CL.2.Sg	pee	in truth	with	Art.Fem.Sg	of	people	studying
	psicología con 0 capacidad de empatía y análisis.
	Psychology with 0 capacity of empathy and analysis
	‘You won’t believe how many people there are studying psychology without any empathy or analysis’

(31)

a.

¡Con

la casa

que

tiene!

With

the house

that

has

‘In spite of the house that he has!’

b.

Con

lo

que

come

ese muchacho

With

what

that

eats

that guy

‘Despite what that guy eats!’

c.

Ya no se

puede

jugar al

fútbol

con

la

de

giles

Still not Cl

can

play to-the

football

with

Art.Fem.Sg

of

fools

que

hay

acá

That

are

here

‘You can’t play football what with the fools that there are here!

d.

*¡Con

cuánto

come

ese muchacho!

With

how-much

eats

that guy

‘*Despite how much that guy eats!’

e.

*¡Ya

no

se

puede

jugar

al fútbol

con

cuántos giles

Still

not

CL

can

play

to-the football

with

wow-many fools

que

hay acá

that

are here

‘*You can’t play football what with how many fools there are here!’

If the clause introduced by que is a relative clause, it follows that the omission of the relativizer be ruled out, as shown in (23). It is a well-known fact that, in Spanish, as opposed to other languages such as English, it is not possible to find contact relative clauses, i.e., those in which the relativizer is phonologically null. If, alternatively, the whole structure is clausal and que is just a complementizer or subordinating conjunction, the reason why que is mandatory here but optional in the rest of the exclamative clauses such as ¡Cuánto/qué calor (que) hace! (‘How hot it is!) remains unexplained. Furthermore, as is well known with relative clauses, nominal exclamatives can both be stacked and coordinated.

(32)	¡La de gente que conocí que estudia en la Universidad y que vota a Milei!
	‘How many people I know who study at college and vote for Milei!’

Notice that a clausal analysis of a sentence such as (32) cannot smoothly account for the non-movement of the stacked and coordinated que clauses after the main predicator conocí. Let us assume that the whole structure is indeed a CP and that the constituent la de gente que estudia en la Universidad y que vota a Milei allows discontinuous spell out, as is schematically represented in (33). If this were the underlying structure of (32), the non-discontinuous spell out (33b) would also be expected and other similar focalized constructions (33c) could be analyzed in a similar fashion. However, as (33b–c) show, this does not occur.

(33)	a.	[CP [La de gente que estudia en la Universidad y que vota a Milei]C que [TP conocí [la de gente que estudia en la Universidad y que vota a Milei]]].
	b.	??[CP[La de gente que estudia en la Universidad y que vota a Milei]C que[TP conocí [la de gente que estudia en la Universidad y que vota a Milei]]]].
	c.	*La gente, conocí [la gente que estudia en la universidad y que vota a Milei].
	‘The people, I met who study at college and vote for Milei’

Some counterevidence to a DP analysis is found in the fact that these configurations might be pronominalized with the neuter clitic lo or the neuter demonstrative eso, as though they were a sentence:

(34)	A:	¡La de veces que lloré este año!
		‘How many times I cried this year!’
	B:	Ya me lo dijiste ayer eso.
		‘You told me that yesterday’

What seems to be at work here is that the verb of communication decir ‘tell’ typically selects propositional arguments, even so, it can still select DPs as in decir la verdad ‘tell the truth’ or decir mentiras ‘tell lies.’7 The reason why the sequence «La de…» might not be replaced with la or another clitic is that it is headed by a quantifier, and hence lacks referential properties. When the DP contains a referential cantidad as in (14) or (35a) below, agreement with the head noun is indeed possible, but when cantidad is the head quantifier of a pseudopartitive this option is ruled out (35b).

(35)	a.	La cantidadi de inscriptos en el examen ya me lai has dicho
		‘The number of students enrolled in the test, you have already told me’
	b.	¡(Una) cantidad de veces te lo/*la he mencionado!
		‘I’ve mentioned this so many times!’

Another argument in favor of the hypothesis that these sentences entail some wh-movement, as that found in the relative clauses in (36), comes from Torrego’s (1988) observation that nominal exclamatives (37a) license parasitic gaps due to the presence of the wh- operator, which, in our case (37b), would be the relativizer que heading the dependent clause.

(36)	María me dio el libro quei Juan devolvió ___i sin mirar ___i.
	‘Maria gave me the book that John returned without reading’

(37)	a.	¡Los libros quei Juan devolvió ___i sin mirar ___i!
		‘The books that John returned without reading!’
	b.	¡La de libros quei Juan devolvió ___i sin mirar ___i!
		‘How many books that John returned without reading!’

Let us suppose that in (37b) the underlying structure is clausal and that la de libros is a constituent that moves to a Focus position before the complementizer que. If this were the case, the sentences in (38) should be ungrammatical, for the simple reason that, as is assumed in the bibliography (Rizzi 1997), foci do not tend to precede topics, as those italicized in the examples below.8

(38)

a.

¡La

de

especialistas

que,

sobre

ese

tema,

sabe

poco y nada!

Art.Fem.Sg

of

specialists

that

about

that

topic

knows

little and nothing

‘How many specialists know very little about that topic!’

b.

¡La

de

hospitales

que,

a

la

hora

de

una

urgencia,

están

cerrados!

Art.Fem.Sg

of

hospitals

that

to

the

hour

of

an

emergency

are

closed

‘How many hospitals are closed in an emergency!’

A plausible objection to our analysis could be that these sentences reject other foci after que, in view of the fact that these are not recursive within the same clausal domain.

(39)

a.

*¡La

de

temas

que

A

JUAN

le

encantan

Art.Fem.Sg

of

songs

that

to

John

CL

loves

‘How many specialists know very little about that topic!’

b.

*¡La

de

veces

que

MARÍA

se

lastimó,

no

Juan.

Art.Fem.Sg

of

times

that

Mary

CL

hurt

not

John

The ungrammaticality of (39) has nothing to do with the lack of foci recursivity, however, but with the fact that the clause headed by que is a relative clause. As pointed out by Leonetti and Escandell-Vidal (2017), restrictive relatives provide descriptive information that facilitate the identification of the referent of the NP, operating as predicates that narrow down the denotation of the antecedent. In this sense, they contribute presupposed, non-controversial information, which is part of the participants’ common ground. Even though these clauses might contain topics, as revealed by (38), they constitute syntactic environments which lack any internal informational domain, as they are associated with assertions and at-issue topics.

In connection with this point, a quirky feature of the relative clauses under consideration is their similarity to those clauses found in cleft sentences. In both cases, in (40) and (41), it is possible to identify, on the one hand, a preposed focalized constituent, written in capital letters, accented and with a higher pitch, and, on the other, a relative clause in final position, whose information and content are presupposed and taken for granted. In both cases, what is more, complex relative pronouns can be used.

(40)

a.

¡La

de

PROBLEMAS

con

los

que

tuve

que

lidiar!

Art.Fem.Sg

of

problems

with

the

that

had

that

deal

‘How many problems I had to deal with!’

b.

Fueron

ESOS

PROBLEMAS

con

los

que

tuve

que

lidiar,

no

estos

Were

those

problems

with

the

that

had

that

deal

not

these

‘It was those problems I had to deal with, not these’

Another contact point between cleft sentences and our data is that the relative clauses they are made up of are not prototypical. In the case of cleft sentences, as Di Tullio (2014) claims, the anomaly is concerned with the fact that they can modify a proper noun with unique reference (e.g., Fue A JUAN que echaron de la oficina, ‘It was John that was fired from the office’) and even prepositional phrases, among other categories which are not nominal (e.g., Fue EN LA FIESTA que encontraron el cuerpo, ‘It was at the party that the body was found’). As for «La de…», it deviates from canonical relative clauses in that the latter do not trigger subject–auxiliary inversion, whereas «La de…» apparently does, albeit the oddity of (43) is not straightforward for all speakers:

(42)	a.	Los libros que el profesor me prestó son muy interesantes
		The books that the teacher me lent are very interesting
	b.	Los libros que me prestó el profesor son muy interesantes
		The books that me lent the teacher are very interesting
	‘The books that the teacher lent me are very interesting’

(43)	?¡La	de	libros	que	el	profesor	me	prestó!
	Art.Fem.Sg	of	books	that	the	teacher	CL	lent
	(cf. ¡La de libros que me prestó el profesor!)

Additionally, while most of the time we can dispose of a restrictive relative clause and obtain a grammatical utterance, with «La de…» such an omission is not available.

(44)	Los libros (que el profesor me prestó) son muy interesantes
	‘The books (that the teacher lent me) are very interesting’

(45)	¡La de libros *(que me prestó el profesor) es de no creer!
	‘I can’t believe how many books *(the teacher lent me)’

In this sense, this type of relative clauses resembles emphatic relatives (Plann 1984; Gutiérrez-Rexach 1999; Brucart 1999), whose main characteristic is that their head and que can be replaced with an equivalent exclamative operator. In the examples below, the relative clause cannot be dispensed with, and the subject tends to appear post-verbally when it is not emphatic.

(46)	a.	¡Lo difícil que es el examen! (=¡Cuán difícil es el examen!)
		‘How difficult the exam is!’
	b.	¡Los disparates que dice tu novio! (=¡Qué disparates que dice tu novio!)
		‘What nonsense your boyfriend says!’

These differences are aligned with Carlson’s (1977) proposal that a number of constructions that have the essential prima facie appearance of defining relative clauses, upon careful examination, do not easily fall into the traditional binary classification of relatives.

(47)	Every man there was on the life-raft died (Carlson 1977, p. 521)

The author labelled these constructions amount relatives, as amounts are part of their syntactic and semantic make up. Grosu and Landman (2017) call these and other kindred constructions maximalizing relatives, while Heim (1987) calls them degree relatives. Said structures are argued to quantify over a complete set of entities and encode amounts, cardinalities, durations, weights, distances, etc. Following Grosu and Landman’s classification, our data could be classified as d(egree)-headed relatives, as the relative clause is headed by a degree head, which was originally the noun cantidad and is now the quantifier la (see Section 4). If this line of reasoning is on the right track, the exclamative flavor attributed to these configurations in Section 2.2 can easily be derived. Torrego (1988) observes that the amount reading emerges by dint of a null operator in the relative clause, which moves in LF from a clause internal position to the left periphery and has scope over its antecedent, the DP, which will determine its amount interpretation. In line with Torrego (1988), we also believe that the relation between the amount relative and the antecedent (the pseudopartitive structure described in Section 2.1) is predication, inasmuch as the subordinate clause allows us to narrow down the scope of reference of the possible entities denoted by the antecedent, that is, the extension of the set defined by the noun in the tail of the pseudopartitive.

The subject–predicate relation can be observed not only in the paraphrase of this construction below, but also in others expressing a similar qualifying or quantifying meaning (see Saab 2004; Villalba and Bartra 2010, for a detailed analysis of the latter).

(48)

a.

¡La de libros que compró Juan! → Juan compró muchos libros

‘How many books John bought! → John bought many books’

b.

Me

sorprende

[lo

lindo

de

la

casa]

→

La casa

es

linda

CL1.Sg

surprises

Art.

nice

of

the

house

the house

is

nice

‘It surprises me how nice the house is! → The house is nice’

c.

El

idiota

de

tu

novio

→

Tu

novio

es

un idiota

The

idiot

of

your

boyfriend

Your

boyfriend

is

an idiot

‘That idiot of your boyfriend → Your boyfriend is an idiot

Following Villalba and Bartra’s (2010) analysis of (48b), in (48a) the subject–predicate relation happens to be altered due to information packaging and discourse-related issues: the QP la de libros constitutes the focus of these constructions and the relative clause the background information. Evidence in favor of this comes from the impossibility of associating the relative with focus particles such as solo ‘only’ (49a) and from the ungrammaticality caused by the presence of non-specific quantifiers such as cualquiera (49b) or excess quantifiers such as demasiados (49c), which are known to reject background or topic positions.

(49)	a.	*¡La de libros que sólo compró Juan!
		‘How many books John only bought!’
	b.	?*¡La de libros que lee cualquier estudiante!
		‘How many books any student reads!’
	c.	*¡La de libros que leen demasiados estudiantes!
		‘How many books too many students read!’

All things considered, it is possible to conclude that the clause that restricts the antecedent in the pseudopartitive is ‘relatively’ relative or, at least, different from the relatives found in DPs whose semantic import and force are not exclamative. This anomaly is nothing but an epi-phenomenon of the chimeric and hybrid nature of these constructions, which, as we have attempted to prove throughout this section, are halfway between clauses and noun phrases.

3. The Internal Structure of the DP

Having described the structure introduced by «La de N…», we will now move on to outline some of the theoretical assumptions from the framework adopted in our research, which will pave the way for the morphosyntactic analysis presented in Section 4. Our main focus will be the internal structure of the DP and its split into smaller discourse projections.

Distributed Morphology (Halle and Marantz 1993; et seq.) is a non-lexicalist, decompositional model, whose core tenets, according to Bobaljik (2017), are: (i) syntax-all-the-way-down: the internal hierarchical structure of words is syntactic (complex words are derived syntactically), and (ii) late insertion: the syntax operates on abstract morphemes, defined in terms of morphosyntactic features, and the spell out (realization, exponence) of these abstract morphemes occurs after the syntax, in the Morphological Structure, which is located after the Syntax and before the Phonetic Form.

According to this framework, there exist post-syntactic operations that can alter the externalization of the syntactic structures generated by the computational system in order to comply with language-specific requirements. Agreement phenomena, for example, are usually attributed to these post-syntactic operations, given their strictly morphophonological nature. In this view, the agreement or concord between a determiner and a noun in Spanish basically consists in copying the number and gender features of the latter to the former, post-syntactically, through an operation known as Feature Copying (Embick and Halle 2011).

As regards the internal structure of the DP, we will follow Mare’s (2016) proposal for pseudopartitive constructions, schematized in (50) below.

(50)

Based on previous research on DPs (Kornfeld and Saab 2005; Panagiotidis 2002; etc.), the author divides the projection into three functional heads, each of which houses specific syntactic–semantic features. The first one, D, encodes person and definiteness features. As argued in Section 2.1, pseudopartitives are indefinite, which is why the head carries a [-DEF] feature. NumP includes number information, which, in this case, as the structure is a pseudopartitive, is always [+PLURAL]. This projection has a close tie to quantification, which is why it contains quantifiers in its specifier, as well as epithets, evaluative adjectives and restrictive modifiers, such as the relative clauses discussed in the previous section (Saab 2008; Ticio 2003). The third head is the categorizer n, in which gender and class features such as [ANIMATE] or [HUMAN] can be found.

The representation in (50) captures the meaning of equivalent quantified expressions like muchos chicos (‘many boys’) and the pseudopartitive un montón de chicos (‘a lot of boys’). In both cases, the noun chicos is quantified by muchos/un montón. In the first case, the quantifier muchos lacks inherent gender and number features and must hence value them through agreement with the noun. Conversely, in the pseudopartitive, both the quantifier (montón) and the noun in the tail (chicos) possess their own already valued features, which is the reason why the default preposition de (‘of’) appears, according to Mare, as a copulative element that builds a relation between two constituents when the traditional agreement mechanisms have already taken place. As regards the final position occupied by the quantifier, Mare claims that this raises to Spec-DP, attracted by the [-DEF] feature in the head.

So as to account for the exclamative flavor of the pseudopartitive constructions analyzed in this paper, it is pivotal to complement Mare’s (2016) proposal with a more fine-grained DP whose structure lets us describe discourse-related nuances and meanings. Those alternatives that translate Rizzi’s (1997) Split-CP hypothesis into the DP domain can help us do so, as they mirror said likeness between DPs and CPs.

According to Rizzi (1997, et seq.), the CP can be decomposed into smaller projections in order to account for the analysis of phenomena that give rise to leftward movement (topicalization, focalization, exclamation, etc.). More narrowly, the author suggests that the CP can be split into at least two projections: (i) a Force Phrase (ForP), which is in charge of encoding the illocutionary force of the sentence and is linked to discursive facts; and (ii) a Finiteness Phrase (FinP), which establishes a stronger relation with the Tense Phrase (TP) by virtue of being responsible for determining the inflectional properties of the utterance. In turn, between these two nodes, there are two further optional functional projections: (iii) the Focus Phrase (FocP), which accounts for the interpretation of wh-elements and focalized constituents; and (iii) the Topic Phrase (TopP), which serves as the landing site for topics (old or shared information which is already known in the discourse). The resulting structure is summarized below:

(51)	CP = ForP > (TopP) > (FocP) > FinP

The spirit of Rizzi’s proposal has been extended to the analysis of DPs by several authors (Aboh 2004; Roca 2009; Ticio 2003; to name but a few). In this paper, we will adopt Roca’s (2009) reinterpretation, according to which the D node can also be expanded into four smaller projections, akin to those specified in (51): (i) a FinP, which encodes the definiteness of the DP. In Spanish, for instance, demonstrative pronouns could bear this function; (ii) a FocP, which features wh-interrogative DPs (e.g., qué libros, ‘what books’) or contrastive foci (e.g., Me llevaré MI libro, no el tuyo, ‘I’ll take my book, not yours’), as well as those determiners used for emphasis (e.g., Me compré EL auto, ‘I’ve bought THE car’) or utter exclamations (e.g., ¡LAS DEUDAS que tengo!, ‘The debts that I have!’); (iii) a TopP, which would be in charge of anaphoric readings of definite determiners (e.g., Vi una chica. La chica era muy alta, ‘I saw a girl. The girl was very tall); and (iv) a ForP, an externally oriented category, which inserts the construction into a larger grammatical domain (the sentence or a predicative structure). Roca considers that the determiners in this projection are expletive, as they do not make any relevant semantic or syntactic contribution. Even though the inclusion of a ForP in Roca’s analysis is motivated by the empirical evidence offered in his paper, it will not be included here, for it is tangential for the purposes of our research.9

Having laid out these central assumptions on the structure of nominal expressions, we schematize below the architecture of the DP that will guide the morphosyntactic analysis of «La de..» in the following section of the paper.

(52)

4. The Morphosyntax of «La de»

The conclusion we arrived at in Section 2 was that the categorial status of «La de…» is hybrid due to the fact that it shows the semantic–pragmatic behavior of an exclamation and at the same time the internal syntactic structure of a DP. Building on what was explained in the previous section, we will now turn to the morphosyntactic structure of the pattern. But before doing so, some brief review of its origin is in order.

As anticipated in the introduction, the structure began to be documented as from the late nineteenth century. A large number of authors agree on the fact that the pattern originated from the DP la cantidad de N ‘the quantity/number/amount of N’ (Bosque 2017; Brucart 1999; Fernández Ramírez 1951; RAE-ASALE 2009; Roca 2009; Torrego 1988; etc.). If this is on the right track, it means that the first step in the transformation of the definite article into an exclamative pronoun was the change in category undergone by the noun cantidad, which became a quantifier. Formally speaking, this implies internally merging (moving) the noun categorizer and the root to Spec-NumP, which is where it will obtain the interpretation corresponding to a quantifying noun.10

(53)

The second step in the process consists in the loss of the root and the pseudomarginalization of the structure. Pseudomarginalization (Arias 2021) is a process characteristic of the feminine pronoun la by means of which it loses its status as an article to become a clitic, the root in the nP (cantidad) not being phonologically realized. For example, in the expression juntarla en pala (‘to rake it in/make a lot of money’), although la does not appear with the noun plata (‘money’), it is possible for most speakers to retrieve the root that the clitic is substituting for. In the case of «La de…», pseudomarginalization implied the loss of the root cantidad, which, gradually, as from the late nineteenth century, turned into a null quantifying noun. When asked what «La de..» refers to, speakers unanimously coincide that it comes from «La cantidad de…».11

(54)

The next stage involves the grammaticalization and reanalysis of the structure in (54), which means the direct merging of la as a quantifier and its dissociation from the null quantifying noun cantidad. Evidence in support of our analysis comes from the following examples taken from Twitter, where la appears severed from the noun cantidad by the functional preposition de (‘of’).

(55)

a.

La

de

cantidad

de

señores

de 40 años

que

hay

en Twitter no

Art.Fem.Sg

of

quantity

of

men

of 40 years

that

are

in Twitter not

es

normal.

is

normal

‘The number of forty-year-old men on Twitter is not normal!’

b.

Es

absolutamente

injustificable

la

de

cantidad

de

puntos

que

se

Is

absolutely

unjustifiable

Art.Fem.Sg

of

quantity

of

points

that

CL

dejan

contra

estos

equipos

left

against

these

teams

‘It is unjustifiable how many own goals they scored against these teams!’

Establishing what the internal structure of quantifiers is like in a framework such as Distributed Morphology exceeds the purposes of this paper. It suffices to say that the structure of the quantifier la is similar to the structure of the marginal clitic la, except for the fact that it is not underspecified for definiteness (Arias 2021) but carries a [–DEFINITE] feature, as suggested in Section 2.1. Marginal clitics stand in opposition to regular clitics by virtue of the fact that they are non-referential and do not share the phi-features of their referents. The following example clearly illustrates the contrast between these two types of clitics (see Arias 2021; Bibis and Roberge 2004; Espinal 2009; etc.):

(56)	a.	La		rompiste		en		el	partido
		CL.Fem.Sg		broke		in		the	match
		‘You rocked it in the match!’
	b.	¡Mi	telei!		¡Lai		rompiste!
		My	TV		CL.Fem.Sg		broke
		‘My TV! You’ve broken it!’

As is the case with marginal clitics, the feminine article introducing this exclamative lacks demonstrative or anaphorical content.

Regarding the gender morphology, there are various ways to account for it. One possibility could be that, given its grammaticalization, the quantifier has no gender and would therefore behave like a quantifier such as qué, which is invariant in this respect. Another one could be that it receives its gender feature as proposed in (Arias 2021, 2022), post-syntactically, as the lexicalization of the feature [-ANIMATE]. Either way, the relevant feature that distinguishes the quantifier la from other analogous marginal clitics is [+EXTREME], which is responsible for the evaluative and extreme degree values encoded by the operator. This feature will enable the quantifier to check the [+EXCLAMATIVE] and [+Q] features on FocP and hence trigger its A’-movement, which is characteristic of exclamative clauses.

(57)

Once we have traced the transformative path followed by the quantifier la, we will delve into the derivation of the whole pseudopartitive DP. By way of illustration, we will focus on the morphosyntactic structure of the sentence in (58).

(58)	¡La de boludos que vino a la fiesta!
	‘How many fools came to the party!’

(59)

A key notion in the structure in (59) is that the operator la quantifies over the nP boludos and moves further up in the tree to comply with the features on Finº and Focº. First of all, in tune with Mare’s (2016) research on pseudopartitives, la moves to Spec-FinP to check [-DEFINITE] feature. As can be inferred from the diagram in (59), the quantifier pied-pipes the domain it c-commands for two relevant reasons. The first one has to do with the clitic nature of the quantifier and with the fact that in Spanish wh-elements and prepositions pied-pipe their complements as they move along in the configuration. The second one pertains to the [+Q] feature on Focº, whose main function is focalizing the structure to obtain the exclamative interpretation attested in Section 2.3. The [+Q] feature must target an operator with a [+EXTREME] feature so as to generate the quantificational widening effect of exclamations (Zanuttini and Portner 2003). The requirement of FocP to host a focus which is prosodically highlighted through intonation and nucleus placement clashes with the phonological deficiency of clitics, which is why the PP de boludos must move so that the stress and prominence that signal focalized constituents fall on the bare noun.

On the other hand, the presence of a focus comes along with presupposed information that, in this case, facilitates narrowing down and identifying the antecedent. This provides us with a natural explanation for the obligatoriness of the relative clause, given that its omission would yield an inappropriate or infelicitous sentence from a pragmatic point of view, as it is the constituent which adds information that is already known for the speaker (see Section 2.3 on the focus–background partition of the structure). The position occupied by the relative clause (Spec-NumP) is motivated not only by other authors’ proposals on the distribution of modifiers within the DP (e.g., Saab 2008; Ticio 2003) but also empirically. Notice that determiners, complex relativizers and the conjugated verb in the subordinate clause might obtain different agreement morphemes.

(60)

a.

¡La

de

boludos

con

la

que

tengo que

lidiar!

Art.Fem.Sg

of

fool.Masc.Pl

with

Art.Fem.Sg

that

have that

deal

b.

¡La

de

boludos

con

los

que

tengo que

lidiar!

Art.Fem.Sg

of

fool.Masc.Pl

with

Art.Masc.Pl

that

have that

deal

‘The fools that I have to deal with!’

(61)	a.	¡La	de	boludos	que	vino	a la fiesta!
		Art.Fem.Sg	of	fool.Masc.Pl	that	came.3.Sg	to the party
	b.	¡La	de	boludos	que	vinieron	a la fiesta!
		Art.Fem.Sg	of	fool.Masc.Pl	that	came.3.Pl	to the party
		‘How many fools came to the party!’

Merging the relative clause in Spec-NumP has the advantage of keeping Mare’s (2016) hypothesis for other cases of alternating agreement patterns with pseudopartitives. If we apply her analysis to our data, the structure can bring about alternating agreement patterns because there are two DPs (the quantifier and the bare noun) equally local with respect to the relativizer, which, after valuing its features with one of the DPs, can establish agreement relations with T and the article in the complex relative. The author follows Van Koppen’s (2005) definitions of equally local and c-command:

(62)	Equally local
	Y and Z are equally local to X iff,
	a.	X c-commands both Y and Z.
	b.	The set of nodes that c-command Y is identical to the set of nodes that c-command Z.

(63)	C-command:
	X c-commands Y iff,
	a.	X excludes Y.
	b.	The first node that dominates X, also dominates Y.

An analysis of relative clauses in these terms can account for the fact that nominal relatives allow for so-called reconstruction effects. Reconstruction is obtained when a syntactic constituent is bound by another constituent which does not c-command it, at least after spell out. The relevant examples are shown below, where subindices are used to mark co-reference:

(64)

a.

¡La

de

retratos

de

sí mismoi

que

pintó

Juani!

Art.Fem.Sg

of

portraits

of

himself

that

painted

John

‘How many portraits of himself John painted!’

b.

¡La

de

fotos

suyas{i-j}

que

tiene

Juani!

Art.Fem.Sg

of

pictures

Poss.

that

has

John

‘How many of his pictures John has!’

In (64a), the co-reference between the anaphor sí mismo and Juan appears to be compulsory, while in (64b) the post-posed possessive suyas may or may not be bound by Juan. The structure presented in (59) allows for the referential interpretations described, since before spell out, the anaphor and the possessive are bound by the nominal in the relative clause, which occupies a position higher up in the structure.

5. Conclusions

In this paper, we have attempted to analyze the morphosyntax of «La de…», a pattern which is extensively used in different varieties of Spanish in informal contexts. The structure can be subsumed within a wide range of non-standard, colloquial linguistic phenomena from Argentine Spanish, which have attracted the attention of numerous linguists over the last decades (Arias 2021; Carrió 2021; Di Tullio and Kornfeld 2012; Kornfeld 2016, 2023; Masullo 2017; Resnik 2013; just to mention a few). The examples in (65) are a small sample of the effervescent dynamism in the field, where exclamation and speakers’ subjectivity converge.

(65)	a.	¡Un					loco!
		Art.Indef.Masc.Sg					crazy
		‘How crazy!’
	b.	¡Más		lindo!
		More		cute
		‘So cute!’
	c.	¡Es	de			lindo!
		Is	of			cute
		‘It’s so cute!’
	d.	¡Ni			cómodos			estos	zapatos!
		Neither			comfortable			these	shoes
		‘These shoes are so comfortable!’
	e.	¡Flojas/Altas/Zarpadas/Mansas								tortas!
		Loose/Tall/Incredible/Meek								cakes
		‘Amazing cakes!’

Our main conclusion has been that «La de…» is a pseudopartitive indefinite nominal structure which, nonetheless, fulfills all the requirements of partial exclamative clauses. We have argued against the standard clausal treatment of this construction and in favor of the hypothesis that it is a nominal exclamative containing a relative clause. We have proposed the definite article la became an exclamative operator after undergoing a process of grammaticalization, which is summarized in (66) below (see Arias, forthcoming, for a more detailed and technical description of this process):

(66)	a.	The conversion of the root cantidad into a quantifying noun.
	b.	The pseudomarginalization of the quantifying DP la cantidad, which resulted in the loss of the noun root.
	c.	The change in category of la, which became an exclamative quantifier.
	d.	The internal merge of the quantifier la in the left periphery of the DP, i.e., in Spec-Fin-P and Spec-Foc-P.

There are two pending issues which require further research. The first question that remains to be answered is how «La de…» interacts with other constructions which appear to belong to the same family of grammatical phenomena. What is the relation between «La de…» and other structures with equivalent meaning, such as those introduced qué de or cuántos exemplified in (1) and other definite article degree exclamatives headed by the neuter article lo (e.g., ¡LO que labura ese muchacho!, ‘How much/hard that guy works!’)?

Most importantly, another issue which is worth further exploring in depth is whether canonical exclamative clauses (67) can be analyzed in the same fashion as the nominal exclamatives discussed in this paper, as though they were DPs and not CPs. The examples in (67b–c) with relativizers suggest that the standard treatment of exclamative clauses—according to which the focalized constituent is preceded by an overt or null complementizer—should be reexamined in light of the role of relativization and their relation to other exclamative structures which seem to display a very similar syntax.

(67)	a.	¡Cuántos libros (que) leí este año!
		‘How many books (that) I’ve read this year!’
	b.	¡Cuántos niños cuyos padres se han divorciado!
		‘How many kids whose parents have divorced!’
	c.	It’s incredible how many people whose names we now know!