When discourse met null subjects

! ã Ángel Luis Jiménez-Fernández. Borealis: An International Journal of Hispanic Linguistics, 2016, 5 / 2. pp. 173- 189. http://dx.doi.org/10.7557/1.5.2.3727 This is an Open Access Article distributed under the terms of the Creative Commons Attribution License https://creativecommons.org/licenses/by/4.0/legalcode) which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. ! WHEN DISCOURSE MET NULL SUBJECTS Ángel Luis Jiménez-Fernández Universidad de Sevilla ABSTRACT. In this paper a novel view of the distribution of null subjects in Spanish is explored in which the discourse category of the antecedent of pro is crucially paid attention to. Implementing Frascarelli’s (2007) work, I propose that the relevant interpretative condition to meet is for the null subject to be coreferential with an Aboutness-Shift Topic, which must be either an explicit one or a null (i.e. silent) double in the local domain where pro occurs. When silent, this antecedent can refer back to any type of discourse category in a previous clause. This implies an update of the information provided in the sentence containing pro. The analysis is supported by an experiment run among native speakers of Spanish. Keywords. null subject, types of topics, types of foci, silent double RESUMEN. En este artículo se explora un novedoso modo de ver la distribución de sujetos nulos en español, basado en la categoría discursiva que tiene el antecedente de pro. Implementando el trabajo de Frascarelli (2007), propongo que la condición interpretativa que indispensablemente debe cumplir un sujeto nulo es el que sea correferente con un Tópico Oracional, el cual debe aparecer explícitamente o tener un doble nulo (o mudo) en el dominio local donde se encuentra pro. Cuando este antecedente es mudo, puede éste referirse cualquier tipo de categoría discursiva en la ración precedente. Esto implica una actualización de la información que se da en la oración donde aparece pro. El análisis presentado se apoya en un experimento llevado a cabo entre hablantes nativos de español. Palabras clave. Sujeto nulo, tipos de tópicos, tipos de focos, doble mudo. 1. Introduction1 In this work I explore the Information Structure conditions for licensing null subjects in Spanish. In particular, I discuss the connection of pro with the type of topic (Aboutness-Shift Topic, Contrastive Topic and Given or Familiar Topic in the sense of Frascarelli and Hinterhölzl 2007, Bianchi and Frascarelli 2010) and the type of focus (Information Focus and Contrastive Focus; cf. Belletti 2004) which serves as its antecedent. Traditionally, Generative Grammar has paid attention to the agreement conditions of NSs in terms of φ-features and its correlation with the rich morphological import of V (Chomsky 1981; Jaeggli 1982; Rizzi 1986), which is a hot issue even today (Barbosa 2011a,b; Biberauer 2010; Pešková 2014; a.o.). Another hallmark in the !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 1 Earlier versions of this paper have been presented at the GLUE 2 meeting in University of Rome 3 and at the CECIL 4 conference in John Paul II Catholic University of Lublin. I thank the audiences there for their fruitful comments. In particular, I am grateful to Mara Frascarelli for our constant discussion on the issues treated here. I am also thankful to Gigi Andriani, Roberta D’Alessandro and Geraldine Quartararo for the Italian data. The research in this paper has been funded by the project FFI2013-41509-P of the Spanish Ministry of Economy and Competitiveness. Finally, I am grateful to the two anonymous reviewers of Borealis for their insightful comments which definitely have improved the work. ÁNGEL LUIS JIMÉNEZ-FERNÁNDEZ ! 174 literature has been the distinction between consistent, partial and radical pro-drop (Barbosa 2013; Miyagawa 2013; Roberts 2010; Duguine 2014; a.o.), to capture null subjects in agreementless languages (Âmbar 2014). See D’Alessandro (2014) for a detailed state of the art and hints about the future. Null subjects have been traditionally classified into two types, namely referential and non-referential null subjects. In this paper, I discuss the interpretation of referential pro. More precisely, I discuss the distribution of third person referential pro. Thus the analysis I put forth does not say anything about non-referential pro, such as the one we can find with weather verbs such as llover ‘rain’ (pro Estuvo lloviendo todo el tiempo ‘It was raining all the time’). Here the null subjects is just an expletive. On expletive pro, see Svenonius (2002). On the other hand, first and second person NSs are contextually salient and, as a reviewer points out, they need no antecedent. I do not deal with first and second person pro, and leave aside the question of whether they do or do not have an antecedent. To my knowledge, not much heed has been given to the discourse conditions of NSs. Exceptions are just a few. Grimshaw and Samek-Lodovici (1998: 195) mention the topic discourse status of the antecedent of pro, but offer no further elaboration. Camacho (2013: 146 and ff.) argues that null subjects are topic-oriented and connects their availability with the fact that the relevant language is topic-oriented. Frascarelli (2007: 694) claims that “[a] thematic NS is a pronominal variable, the features of which are valued (i.e., ‘copied through matching’) by the local Aboutness-Shift Topic.” To illustrate the connection between a referential null subject and its Aboutness- Shift Topic antecedent, consider (1), where the underlined constituent is the antecedent: (1) Un empresario diferentei es arrestado cada día. ¡Y proi se a businessman different be.3sg arrested each day And SE declara inocente! declare.3sg innocent ‘A different/some businessman is arrested every day. And he declares himself innocent!’ In (1) the null subject refers back to the businessman mentioned in the first clause, and this is the only interpretation of pro in this context. If referential pro is licensed when coreferent with the local Aboutness-Shift Topic, its antecedent must be a [+specific] DP (cf. also Frascarelli to appear). Consequently, pro cannot co-refer with a [-specific] DP since AS-Topics always get a specific interpretation. However, Spanish brings about a certain conspiracy since, as I will show, the antecedent of pro may be other types of topic (Given and Contrastive topics), and even a focused element. Moreover, in Spanish the antecedent of pro can be [+ specific] or [-specific], as illustrated in (2): (2) Muchos profesoresk se han puesto de huelga, y prok han ..... dejado many professors SE have.3pl put of strike and have.3pl left de dar clases. from to.give classes ‘Many professors have gone on strike, and have stopped teaching their classes’ WHEN DISCOURSE MET NULL SUBJECTS ! ! ! ! 175 Quantified subjects such as muchos profesores in (2) are taken to be ambiguous, yielding a specific or non-specific reading (Suñer 2003), yet they can qualify as antecedent of pro. I follow Suñer (2000) and Di Tullio and Zdrojevski (2008) in taking [+ specific] to refer to an identifiable entity, already mentioned in the previous context. On the other hand, definiteness is a grammatical concept which is connected with the presence of a definite or indefinite article. In this perspective, a DP such as un famoso niño ‘a famous kid’ is dubbed by Di Tullio and Zdrojevski (2008: 27) as [+ specific, - definite]. On the other hand the DP el niño que termine primero ‘the child who finishes fist’ is regarded as [-specific, +definite]. Note the use of subjunctive in the latter case, which is taken to be compatible with only non-specifics. My working hypotheses are four: (i) In Spanish the antecedent of pro must simply be a topic, regardless of the type (Aboutness-Shift, Contrastive or Given Topic; hereafter AS-Top, C-Top or G-Top, respectively). (ii) Spanish is quite permissive with the coreference of NS. Hence, other types of discourse categories (Contrastive Focus, henceforth CF) can serve as antecedent for pro, as long as they contain the feature [+ given], meaning that their reference is mentioned in (or inferred from) the previous context. (iii) As a consequence of 2, Information Focus (IF) should not be allowed as an antecedent. (iv) All discourse categories (except IF) may qualify as antecedent irrespective of their [+/-specific] interpretation. The proposal I want to put forth is that pro is coreferential with any discourse category, which is resumed in the local domain of pro via a null AS-Top, and this silent double changes the specificity feature of the fake antecedent, in that in the procontaining clause the real AS-Topic antecedent is specific given that it has been mentioned before in the context.. In elaborating this proposal, I try to answer the following questions: 1) If G-Tops can be antecedents, can other types of topics (C-Top) be antecedents as well? 2) With respect to focus, can a focused DP function as antecedent of a NS? 3) Provided that NSs convey given information and that given information presupposes specificity (Leonetti 2004, 2008; Jiménez-Fernández and Spyropoulos 2013), can the antecedent be really [-specific]? 4) What evidence have we got that there is a null AS-Top serving as antecedent of pro when there is an intervening G-Top? The paper is organised as follows. In section 2 the data from the survey is presented classified depending on the type of topic or focus that the null subject is coreferential with; in addition, the methodology used in the survey is briefly explained and the results of the test are discussed. In section 3 I deal with the connection between specificity and the interpretation of null subjects, focusing on indefinite subjects and their ambiguity in terms of specificity. Section 4 presents my analysis arguing that the licensing condition that null subjects have to meet is to have a local explicit or null AS-Top as antecedent. Being a topic, this antecedent is always specific. Section 5 provides evidence supporting the existence of null AS-Tops, based on Differential Object Marking. Finally, section 5 summarises the main findings. ÁNGEL LUIS JIMÉNEZ-FERNÁNDEZ ! 176 2. Types of topics and foci: data, statistics and problems 2.1. Methodology In this section I present the data which have been used in my analysis alongside a brief description of the types of topics and foci taken into account. I have made a test for Spanish informants. This experiment was given in written form, where the distinct tokens were conveniently randomised. The task was carried out mostly face-to-face, although some informants were asked to fill in the survey via e-mail. The total number of participants in the experiment has been 56, from different areas of Spain. Speakers were asked to express their judgments as “OK” (full acceptance/grammaticality), “??” (marginal, but still acceptable) and “NO” (unacceptable/ungrammatical). They were warned about connecting the subject of the relevant clause with the specific discourse category in the previous sentence, which was underlined, so as to avoid other possible interpretations that the null subject may have. Speakers were confronted with data such as those in (4-10). In these data the first sentence establishes the context which will induce a particular discourse interpretation of the relevant DP as antecedent of the null subject (G-Top, C-Top, IF and CF). For each discourse type I provide just one example, but speakers had two to make sure that their judgements were consistent (hence they were given two examples for each discourse category). Since one of the properties of topics is that they are marked as [+specific] (Leonetti 2008; Frascarelli 2007), the examples include specific DPs as topics, but they could be definite or indefinite (Alexopoulou and Folli 2011; Jiménez-Fernández and Spyropoulos 2013). 2.2. Types of topics and null subjects In this section three types of topics are distinguished based on the different properties assigned to each class by Frascarelli & Hinterhölzl (2007), Bianchi & Frascarelli (2010), Frascarelli & Jiménez-Fernández (2016), among others. It should be stated that the main distinguishing property for each type is prosody. For AS-Top we have a complex L*+H tone pitch accent; for C-Tops a H* pitch accent is detected; and finally, a L* tone is associated with G-Tops. However, intonation is out of the scope of this work. See the references above for the prosodic properties. Syntactically, Frascarelli (2007) has claimed that the three types of topics sit in different designated positions in the CP system. On the other hand, Jiménez- Fernández & Miyagawa (2014) have argued that A-Tops move to CP, whereas C- Tops and G-Tops move to (possibly multiple) specifiers of TP. Since the main goal of this paper is to analyse the interpretation of null subjects, I will focus on the interpretative properties that set apart the three types of topics. 2.2.1. Aboutness-Shift Topic An AS-Top connects Reinhart’s (1981) aboutness (i.e. the sentence Topic) with the property of being newly introduced or reintroduced and changed to (hence, the shift). It is what Lambrecht (1994) described as what the sentence is about. The AS-Top is illustrated in (3) by the DP tu hermano ‘your brother’ (underlining is used for the relevant discourse category; the subindex marks the coreference between the NS and its antecedent): WHEN DISCOURSE MET NULL SUBJECTS ! ! ! ! 177 (3) Has estado hablando de Juan durante horas…. ¿Y tu hermanok have.2sg been talking of John during hours and your brother cómo está? Siempre prok está de viaje... how be.3sg always be.3sg of trip ‘You have been talking about John for hours…. (as for) your brother, how is he? He is always travelling’ The relevant DP tu hermano shows a shift of topic with respect to the previous context, but no contrast at all is established between this DP and Juan. As opposed to C-Tops, AS-Tops do not involve a set of different members which are contrasted between them with respect to a given proposition. Therefore, in the preceding example Juan and the speaker’s brother may be member of the group of friends that both interlocutors share, but this does not mean that they are somehow compared. Since there is no controversy about the role of AS-Tops in licensing NSs, this kind of discourse category has not been tested among informants. 2.2.2. Contrastive Topic Marking an element as C-Top is used to divide a complex proposition into a conjunction of more simple propositions in which a predicate applies separately to each member of a salient set. In the survey I included examples showing coreference with an indefinite C-Top as in (4) and with a definite C-Top as in (5) –note that the relevant examples are preceded by a sentence which establishes the specific context that is pursued: (4) A: ¡¿Te has enterado de cuántos regalos ha hecho Juan reflx have.2sg heard of how.many presents have.3sg made Juan a Jimena por su cumpleaños?! to Jimena for her birthday ‘Did you hear how many presents Juan gave Jimena for her birthday?!’ B: Bueno, todos los regalos, no sé. Algunas joyask las está well all the presents not know.1sg some jewels them be.3sg estrenando hoy en su fiesta (del resto de regalos, wearing for the first time today in her party of.the rest of presents, ni idea). prok le sientan muy bien. no idea her suit.3pl very well ‘Well, I don’t know about all the presents. Some jewels she’s wearing for the first time today at her party (I don’t know about the other presents). They suit her very well’ (5) El programa de Lingüística me parece demasiado exigente este año, the programme of Linguistics me seem.3sg too demanding this year y por varios motivos… el examen final tiene demasiados apartados; and for several reasons the exam final have.3sg too.many sections y los trabajos semanalesk los han cambiado por lecturas and the assignments weekly them have.3pl changed for readings obligatorias. prok eran más fáciles para los alumnos. Obligatory be.past3pl more easy for the students ‘This year’s programme in Linguistics is rather demanding, and for several reasons… The final exam has too many sections; and the weekly assignments have been changed for obligatory readings, and they are easier for students’ ÁNGEL LUIS JIMÉNEZ-FERNÁNDEZ ! 178 Concentrating on the first example, the DP algunas joyas ‘some jewels’ is a member of the set identified as Jimena’s presents. Other members of the set are only implicitly inferred. However, in the second example the set represented by this year’s programme in Linguistics has two explicit members, namely the final exam and the weekly assignments. In both cases, a clear contrast is established, which makes the relevant topic qualify as a C-Top. 2.2.3. Given Topics G-Tops are employed either for continuity with respect to the current sentence Topic or to resume background information. In the survey G-Tops were also indefinite (6) or definite (7): (6) A: Me encantaría una ensalada de tomates para la cena. me would.please.3sg a salad of tomatoes for the dinner ‘I’d love a tomato salad for dinner’ B: Pues, no te olvides de comprar en el mercado. Varios tomatesk well not you forget.2sg of to.buy in the market several tomatoes los he tenido que tirar a la basura. prok estaban them have.1sg had that to.throw to the rubbish were.3pl podridos. spoilt ‘Well, don’t forget to buy in the market. I had to throw several tomatoes to the rubbish bin. They were spoilt’ (7) A: Espero que la cena esté lista ya. Estoy muerto de hambre. hope that the dinner be.3sg ready already be.1sg dead of hunger ‘I hope dinner is ready. I am starving’ B: La cena la he preparado yo ya. prok está ya servida the dinner it have.1sg prepared I already be.3sg already served on the table en la mesa. ‘I have already prepared dinner. It is on the table.’ As topics, both C-Tops and G-Tops imply information shared in the context. For example, in (6) the DP varios tomates ‘several tomatoes’ is resuming the information conveyed by speaker A. Thus, it is given information. However, no contrast is established. There is actually no need to continue this conversation with a sentence about the rest of tomatoes. It may simply be the case that there were no other tomatoes. The fact that G-Tops convey purely shared information (and the absence of any contrast whatsoever) is even clearer in (7), where the DP la cena ‘dinner’ in B is repeating exactly the same entity used in A. 2.3. Types of focus and null subjects In this section I discuss the two types of foci I have concentrated on, namely Contrastive Focus and Information Focus. Intonationally, they are different (see Vanrell & Fernández Soriano (2013) on the intonation of focus in Spanish). Syntactically, Rizzi (1987) made a distinction between CF, moving to the left periphery, and IF staying low in the structure. The latter could be in situ or moved to a designated position in a low periphery (Belletti 2004). However, recent research has found out that in some languages such as Spanish CF can also occur postverbally (Ortega-Santos 2013, 2016), so the focussed element moves to FocP in the left WHEN DISCOURSE MET NULL SUBJECTS ! ! ! ! 179 periphery and the rest of the sentence undergoes remnant movement to a position preceding Foc. In addition, IF has also been shown to target the left periphery in at least some Spanish varieties such as Andalusian Spanish (Jiménez-Fernández 2015b) or Spanish in the Basque Country (Vanrell & Fernández-Soriano 2013). In this work, I have tested fronted CF and in-situ IF. Given that I focus on interpretative properties, I leave the intonation and syntax of focus aside. 2.3.1. Contrastive Focus2 Following Zubizarreta (1998) and Jiménez-Fernández (2015), I take CF to denote a constituent which is asserted in clear opposition with another previously mentioned entity. Example (8) illustrates the coreference of a NS with CF: (8) A: Ana es una crack de la lingüística. Ha publicado en revistas Ana be.3sg an expert of the Linguistics have.3sg published in journals de alto prestigio internacional en este año. of high prestige international in this year ‘Ana is a star in Linguistics. She has published in high-quality international journals this year.’ B: ¡De eso nada! En revistas internacionales no. En revistas nacionalesk of that nothing in journals international not in journals national ha publicado. Y además, prok no están indexadas en have.3sg published and besides not be.3pl indexed in ninguna base de datos. no database ‘No way! She hasn’t published in international journals. She has published in national journals. Moreover, they are not indexed in any database.’ As an anonymous reviewer notes, B’s reply may perfectly leave the fronted CF in situ (Ha publicado en revistas nacionales ‘She has published in national journals’). However, I did not test in-situ CF here. My intuition is simply that the position of CF (in the left or right periphery) does not alter the compatibility with null subjects. 2.3.2. Information Focus IF denotes purely new information (Zubizarreta 1998; Belletti 2004). According to Krifka (2006), question-answer congruence requires the Focus in the answer to satisfy the information request in the interrogative phrase of the question, so the backgrounds should be identical in the question and in the answer. The relevant element in the answer constitutes new information, as illustrated in (9) with a NS referring back to a focused indefinite DP and in (10) with a NS coreferential with a focused definite DP: (9) Se esperaba que sólo unos cuantos políticos estuvieran presentes SE waited.3sg that only a few politicians were.3pl present en la rueda de prensa. Y, ¿sabes quiénes asistieron? in the circle of press and know.2sg who attended.3pl !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 2 In this paper I concentrate on two types of focus, namely contrastive focus and information focus. For a full typology of focus, see Jiménez-Fernández (2015a,b). ÁNGEL LUIS JIMÉNEZ-FERNÁNDEZ ! 180 Vinieron algunos miembros del gobiernok. ¡prok No han apoyado came.3pl some members of.the government not have.3pl supported la decisión del presidente! the decision of.the president ‘Only a few politicians were expected to be present at the press conference. And, you know who attended? Some members of the government were present. They haven’t supported the President’s decision!’ (10) A:El pastel que ha hecho tu madre está buenísimo!! Y también the pie that have.3sg made your mother be.3sg very.good and also la tarta! ¿Quién la ha hecho? the cake who it have.3sg made ‘The pie your mother made is fantastic! And the cake as well! Who cooked it?’ B:¡No vas a creerlo! María, que se tomó un trozo, me dijo not go.2sg to to.believe.it Maria who SE took.3sg a piece me told.3sg que la había hecho la madre de Fernandok . ¡Siempre prok ha that it had.3sg made the mother of Fernando always have.3sg odiado el dulce! hated the sweet ‘You won’t believe it! Mary, who had a piece, told me that Fernando’s mother made it. She has always hated sweet stuff!’ 2.4. Results and discussion To answer the questions raised on the interpretation of NSs in section 1, now I turn to the results of speakers’ grammaticality judgements. Figures are based on the informants’ positive responses. The raw number of participants with positive answers is given in between brackets. In Table 1 I show the coreference between a NS and either a G-Top or a C-Top. Table 1. Coreference of pro and Topics Coreference with a TOPIC DP Preverbal [-def] G-Top Preverbal [+def] G-Top Preverbal [-def] C-Top Preverbal [+def] C-Top 83% (46) 83% (46) 91% (51) 100% (56) Both G-Tops and C-Tops are licit antecedents for pro. This is no surprise in light of the specific nature of topics in general and the constraint suggested earlier that NSs require a specific antecedent. Let us see the results in connection with foci. As observed in Table 2, both CF and IF are possible antecedents for pro. This is unexpected if specificity is associated with given information since it is clear that at least IF conveys new information (or at least the use of IF implies a new relation between the focused constituent and the rest of the sentence). WHEN DISCOURSE MET NULL SUBJECTS ! ! ! ! 181 Table 2. Coreference of pro and focus Coreference with focused DP Postverbal [-def[ IF Postverbal [+def] IF Fronted CF 83% (46) 83% (46) 96% (54) The following conclusions are drawn from the survey: 1) Alongside AS-Tops, G- Tops and C-Tops can be interpreted as the antecedent of a NS, regardless of the [± def] feature, which justifies the validity of my working hypotheses 1, 2 and 5. 2) CF is also construed as coreferential with the NS, thereby confirming the soundness of my working hypothesis 3. 3) Finally, IF is also a possible antecedent, contrary to the expectations expressed in working hypothesis 4. Speakers tend to accept coreference when sharing the feature [+given] with the antecedent, which is present in the featural set of all discourse categories except that of IF. As stated earlier, the [+ given] feature is connected with specificity (i.e. mentioned in previous discourse, in Suñer’s 2000 view). Asked whether the null subjects in these sentences are specific regardless of the specific/non-specific character of the antecedent (i.e. whether they were known in the context), informants gave a positive answer. From this it follows that the interpretation of pro depends on the availability of a specific antecedent (much in line with Frascarelli to appear; see also Âmbar 2014). The problematic key factor seems to be the unavailability of [+given] in IF. To solve this mismatch, what I will propose below is that the [- specific] feature of the antecedent turns into [+ specific] in a null AS-Top to license the NS in its local domain. In other words, in all cases (including IF) the NS establishes its co-reference with the local (silent) A-Top, hence its [+ specific] feature. It should also be clear that the antecedent of the NS in a previous sentence is not claimed to be dislocated or focused. However, in the sentence where pro occurs a null double of the relevant DP is dislocated in the Left Periphery as an AS-Top. 3. On indefinites, specifics and the interpretation of pro Indefinites can be ambiguous between a [+specific] reading and a [-specific] one (Ihsane and Puskás 2001), as shown in (11), from Frascarelli (2007: 715) for Italian, and (12), from Suñer (2003: 345): (11) Un poliziotto stava a guardia di ogni angolo. a policeman be.3sg at guard of every corner (12) Un vigilante montaba guardia en cada esquina. a policeman stand.3sg guard in each corner ‘A policeman guarded each corner’ (ambiguous) This interpretive divide is connected with the syntactic position occupied by the indefinite DP. Concentrating on subjects, much research has been carried out on the precise position of pre-verbal subjects in Romance languages. Basically, we find two mutually exclusive trends (see Sheehan 2006 for a full review), namely those linguists who have argued that pre-verbal subjects target Spec-TP, an A-position (Suñer 2003; Cardinaletti 1997; among others); and those who have claimed that these subjects are placed in the CP-domain, sitting in an A’-position (Barbosa 1995, 2000; Alexiadou and Anagnostopolou 1998; Frascarelli 2007; among others). ÁNGEL LUIS JIMÉNEZ-FERNÁNDEZ ! 188 Camacho, J. (2013). Null subjects. Cambridge, Cambridge University Press. https://doi.org/10.1017/cbo9781139524407 Cardinaletti, A. (1997). Subjects and clause structure, in L. Haegeman (ed.), The new comparative syntax. London and New York, Longman, pp. 33–63. Chomsky, N. (1981). Lectures on government and binding. Dordrecht, Foris. D’Alessandro, R. (2014). The null subject parameter: Where are we and where are we headed? Ms., University of Leiden. Available at http://ling.auf.net/lingbuzz/002159 Di Tullio, Á. & P. Zdrojewski. (2008). Notas sobre el doblado de clíticos en el español rioplatense: asimetrías entre objetos humanos y no humanos. Filología XXXVIII, pp. 13-44. Duguine, M. 2014. Null arguments and linguistic variation: A minimalist analysis of pro-drop. PhD. dissertation, Universidad del País Vasco/Université de Nantes. Erteschik-Shir, N. (2006). What’s what?, in G. Fanselow, C. Fery, M. Schlesewsky & R. Vogel (eds.), Gradience in grammar. Oxford, Oxford University Press, pp. 317–335. https://doi.org/10.1093/acprof:oso/9780199274796.003.0016 Frascarelli, M. (2007). Subjects, topics and the interpretation of referential pro. An interface approach to the linking of (null) pronouns. Natural Language and Linguistic Theory 25, pp. 691–734. https://doi.org/10.1007/s11049-007-9025-x Frascarelli, M. (in press). The interpretation of pro in consistent and partial NS languages: A comparative interface analysis, in F. Cognola (ed.), Understanding Pro-Drop. Oxford, Oxford University Press. Frascarelli, M. & R. Hinterhölzl. (2007). Types of topics in German and Italian, in K. Schwabe & S. Winkler (ed.), On Information Structure, Meaning and Form. Amsterdam, Benjamins, pp. 87–116. https://doi.org/10.1075/la.100.07fra Frascarelli, M. & Á.L. Jiménez-Fernández. (2012). Specificity effects at the discourse-grammar interface. Paper presented at Going Romance, Leuven. December 6-8. Frascarelli, M. & Á.L. Jiménez-Fernández. (2013). Fronting contrast at the interfaces. Paper presented at CamCoS 2, Cambridge, May 9-11. Grimshaw, J. & V. Samek-Lodovici. (1998). Optimal subjects and subject universals, in P. Barbosa, D. Fox, P. Hagstrom, M. McGinnis & D. Pesetsky (eds.), Is the Best Good Enough? Optimality and Competition in Syntax. Cambridge (Mass.), MIT Press, pp. 193–219. Ihsane, T. & G. Puskás. (2001). Specific is not definite. Generative Grammar in Geneva 2, pp. 39–54. Jaeggli, O. (1982). Topics in Romance Syntax. Dordrecht, Foris. Jiménez-Fernández, Á.L. (2015a). Towards a typology of focus: Subject position and microvariation at the discourse–syntax interface. Ampersand: An International Journal of General and Applied Linguistics 2, pp. 49–60. https://doi.org/10.1016/j.amper.2015.03.001 Jiménez-Fernández, Á.L. (2015b). When focus goes wild: Syntactic positions for information focus. Linguistics Beyond and Within 1, pp. 119-133. Online at: http://lingbaw.com/2015/Ángel-L.-Jiménez-Fernández Jiménez-Fernández, Á.L. & S. Miyagawa. (2014). A feature-inheritance approach to root phenomena and parametric variation. Lingua 145, pp. 275–302. https://doi.org/10.1016/j.lingua.2014.04.008 Jiménez-Fernández, Á.L. & V. Spyropoulos. (2013). Feature inheritance, vP phases and the information structure of small clauses. Studia Linguistica 67 (2), pp. 185–224. https://doi.org/10.1111/stul.12013 WHEN DISCOURSE MET NULL SUBJECTS ! ! ! ! 189 Krifka, M. (2006). Association with focus phrases, in V. Molnár & S. Winkler (eds.), The Architecture of Focus. Berlin/New York, Mouton de Gruyter, pp. 105–36. https://doi.org/10.1515/9783110922011.105 Lambrecht, K. (1994). Information Structure and Sentence Form. London, Cambridge University Press. https://doi.org/10.1017/cbo9780511620607 Lahousse, K. (2013). Binding at the syntax-information structure interface, in Victoria Camacho-Taboada, Ángel L. Jiménez-Fernández, Javier Martín-González and Mariano Reyes Tejedor (eds.), Information Structure and Agreement. Amsterdam, John Benjamins, pp. 141–174. https://doi.org/10.1075/la.197.05lah Leonetti, M. 2004. Specificity and differential object marking in Spanish. Catalan Journal of Linguistics 3, pp. 75–114. Ortega-Santos, I. (2013). Corrective focus at the right edge in Spanish. Lingua 131, pp. 112–135. https://doi.org/10.1016/j.lingua.2013.02.007 Ortega-Santos, I. (2016). Focus-Related Operations at the Right Edge in Spanish: Subjects and Ellipsis. Amsterdam, John Benjamins. https://doi.org/10.1075/ihll.7 Miyagawa, S. (2013). Surprising agreement at T and C. Ms., MIT. Pešková, A. (2014). Sujetos pronominales en el español porteño: Implicaciones pragmáticas en la interfaz sintáctico-fonológica. Ph.D Dissertation, University of Hamburg. Rizzi, L. (1986). Null objects in Italian and the theory of pro. Linguistic Inquiry 17, pp. 501–558. Roberts, I. (2010). A deletion analysis of null subjects, in T. Biberauer, A. Holmberg, I. Roberts & M. Sheehan (eds.), Parametric Variation. Cambridge, Cambridge University Press, pp. 58–87. Sheehan, M. (2006). The EPP and Null Subjects in Romance. Ph.D Dissertation. University of Newcastle. Spyropoulos, V. & A. Revithiadou. (2009). The morphology of PAST in Greek. Studies in Greek Linguistics 29, pp. 108–122. Suñer, M. (2000). Object shift: Comparing a Romance language to Germanic. Probus 12, pp. 261-289. Suñer, M. (2003). The lexical preverbal subject in a Romance null subject language. Where are thou?, in R. Núñez-Cedeño, L. López & R. Cameron (eds.), A Romance Perspective on Language Knowledge and Use. Amsterdam/Philadelphia, John Benjamins, pp. 341–357. Svenonius, P. (2002). Subjects, Expletives and the Extended Projection Principle. Oxford, Oxford University Press. Vanrell, M.M. & O. Fernández-Soriano. (2013). Variation at the interfaces in Ibero- Romance. Catalan and Spanish prosody and word order. Catalan Journal of Linguistics 12, pp. 253-282. Zubizarreta, M.L. (1998). Focus, Prosody and Word Order. Cambridge (Mass.), MIT Press.