Reduced Parenthetical Clauses and the syntax-discourse interface

33 Comment Reduced Parenthetical Clauses and the syntax-discourse interface* Orações parentéticas reduzidas de comentário e a Interface sintaxe-discurso Gabriela Matos**8 ABSTRACT Reduced parenthetical clauses (RPCs) in European Portuguese have not been systematically studied until recently. Focusing on comment RPCs, a preliminary classification of the most common verbal predicates in this context is set up for this language, and their pragmatic values established. The distribution of comment RPCs, especially their floating status and the blocking of c-command from the host sentences, indicates that they are syntactically related to their hosts by a paratactic link, as parenthetical modifiers. As for the internal structure of the comment clause, island effects reveal that, in European Portuguese, the argument gap of the verb results from movement of a null category to an A-bar position of the parenthetical clause. Also, adverb placement shows that the verb and the subject occupy high discursive projections in the left periphery of the RPC, suggesting that the postverbal subject is focalized in specifier of FocP. The contrasts between comment RPCs with null and overt subjects shows that in European Portuguese the mitigative value of the RPC is not exclusively drawn from the epistemic, evidential or evaluative predicates, but also relies on the post verbal use of the subject. Keywords: Reduced Parenthetical Clauses, comment parentheticals, Subject placement *The research for this paper has been developed within the projects TAXE and UID/LIN/00214/2019, funded by FCT. I would like to thank all the people who gave me feedback on previous versions of this work, in particular Rui Marques. I am also grateful to the audience the 12th Workshop on Formal Linguistics for their comments on the oral version of this paper. ** Faculdade de Letras da Universidade de Lisboa - Centro de Linguística da Universidade de Lisboa 34 RESUMO As orações parentéticas reduzidas só recentemente têm sido sistematicamente estudadas em português europeu. Com foco nas parentéticas reduzidas de comentário, é estabelecida uma classificação preliminar dos predicados verbais que ocorrem nestes contextos nesta língua. A distribuição destas orações, em especial o seu estatuto flutuante e o bloqueamento de c-comando a partir da frase hospedeira, indicam que estabelecem com o seu hospedeiro uma relação paratática de modificador parentético. Quanto à estrutura interna da oração comentário, efeitos de ilha revelam que, em português europeu, o argumento omitido do verbo resulta do movimento de uma categoria nula para uma posição A-barra da oração parentética. Por seu turno, a colocação de advérbios mostra que o verbo e o sujeito ocupam projeções discursivas altas na periferia da parentética e sugerem que o sujeito pós-verbal ocorre em especificador de FocoP. Os contrastes entre parentéticas reduzidas comentário com e sem sujeitos omitidos mostram que, em português europeu, o valor pragmático mitigativo não decorre apenas dos predicados epistémicos evidenciais ou avaliativos, mas também da realização do sujeito em posição pós-verbal. Palavras-chave: Orações parentéticas reduzidas, parentéticas de comentário, colocação do sujeito. Matos, G: Comment Reduced Parenthetical Clauses and the syntax-discourse interface 35 Revista Letras, Curitiba, ufpr, n. 99, pp. 33-57, jan./jun. 2019. issn 2236-0999 (versão eletrônica) 1. Introduction The designation comment clause appears in Jackendoff (1972) applied to parenthetical clauses involving verbal predicates, such as belief and though, whose main semantic import is to express the speaker’s comment on the related clause, as shown in (1) and (2) for English and European Portuguese. (1) John is, I believe, an excellent violinist. (2) Este estudante leu, penso eu, todos os artigos sobre esse assunto. this student read, think I, all the papers on that subject ‘This student read, I think, all the papers on that subject.’ The term comment parenthetical has been extended to other kind of clauses (e.g. QUIRK et al. 1972, 1985; GIORGI, 2016), as, for instance, to “as-parentheticals” or “-ing parenthetical clauses” (cf. (3)), which are outside the scope of this paper. 36 (3) a. I’m a pacifist, as you know. b. I doubt, speaking as a layman, whether television is the right medium. (QUIRK et al., 1972) Comment parentheticals like those in (1) and (2) are syntactically characterized by the omission of an argument of the verb, usually its complement, which is related to the content of the host clause. In addition, in some languages, for instance in European Portuguese, they typically exhibit subject inversion. These comment parentheticals share their structural shape with direct speech report parentheticals, illustrated in (4), also known as quotative parenthetical clauses (COLLINS, 1997; COLLINS; BRANIGAN, 1997): (4) a. “Don’t turn back!” warned Marcel (COLLINS; BRANIGAN, 1997) b. “O João leu todos os artigos sobre esse assunto”, disse eu. (EP) the John read all the papers on that subject, said I ‘João read all the papers on that subject, I said.’ Comment and quote parentheticals presenting this configuration have also been named Parenthetical Verbs (URMSON, 1952), Reduced Parenthetical Clauses (SCHNEIDER, 2007a, 2007b) or Parenthetical Verb Clauses (KLUCK; DE VRIES, 2015). I will adopt the designation Reduced Parenthetical Clauses (henceforth, RPCs). Comment and quote RPCs have been correlated by their structural similarity, but also by their semantic and pragmatic value (QUIRK et al., 1972, 1985; ROORYCK, 2001; KLUCK; DE VRIES, 2015; GRIFFITHS, 2015). When trying to differentiate them, some authors remark that comment RCPs are speaker oriented (e.g. REINHART, 1983) and typically present subjects in the first person, referring to the speaker. However, even taking these properties into account, the ambiguity between direct reporting speech RPCs and comment RPCs may arise, as in (5). Thus, Kluck and de Vries (2015, p. 105) claim that “a parenthetical verb construction is necessarily an epistemic or evidential comment if the host is not a direct quote”. The examples in (6) illustrate this claim: in (6a), the parenthetical clause reports what was said by Anna in quotation marks, while the parenthetical in (6b) may be interpreted as a speaker’s comment on the whole utterance, meaning according to Anna. (5) It’s late, I admitted. (QUIRK et al., 1985) (6) a. “Bob is a real charmer”, said Anna. b. Bob is a real charmer, said Anna. (KLUCK; DE VRIES, 2015) Matos, G: Comment Reduced Parenthetical Clauses and the syntax-discourse interface 37 Revista Letras, Curitiba, ufpr, n. 99, pp. 33-57, jan./jun. 2019. issn 2236-0999 (versão eletrônica) In this paper, I will focus on comment RPCs in European Portuguese, comparing their behavior in this language variety with English and V2 languages. Several questions emerge in the study of verb comment RPCs, some of them correlated: (i) How do recent RPC typologies account for comment clauses and which verbal predicates may occur in comment RPCs? (ii) What is the nature of the omitted argument of the verb (iii) How is the parenthetical clause connected with its host, given its syntactic behavior and pragmatic value? (iv) What is the internal structure of comment parenthetical clauses, considering the word order patterns allowed and the verbs involved? These questions will guide the current analysis. In section 2, on the basis of the existing literature, I will try to establish the discursive classes related to Comment RPCs and the classes of predicates that may be correlated with them in European Portuguese; in section 3, I will discuss the nature of the omitted argument in the RPC and its correlation with the host clause; in section 4, I will sketch a proposal to account for the structural connection of the parenthetical with the host sentence; in section 5, the internal structure of the parenthetical and its relation with its pragmatic value will be analyzed. In section 6, some final remarks will be presented. 2. The RPC typologies and the classes of verbal predicates in comment clauses In order to account for comment RPCs in European Portuguese, the specific discourse properties that they exhibit and the classes of predicates that usually occur in these clauses must be established. To my knowledge, no such study exists for EP. So, I will review proposals presented for other languages, in particular for English, and try to establish their most usual correlates in European Portuguese. 2.1 Typologies on RCPs Although several studies on RPCs assume the existence of comment clauses (KLUCK; DE VRIES, 2015; GIORGI, 2016), recent typologies on RPCs by Schneider (2007a, 2007b) and Griffiths (2015) do not explicitly refer to them. Schneider (2007a, 2007b) proposes a pragmatic typology taking as main criterion the speaker´s responsibility. He classifies RPCs into three types, the third of which is divided into four subtypes: (7) Type 1 - Phatic RPCs: the speaker invites the addressee to cooperate. Type 2 - Reporting RPCs:the speaker reports something said by another or by him/herself, without reflecting his/her judgment. Type 3 - Mitigative RPCs downgrade speaker’s commitment, by alleviating, removing or sharing his/her responsibility: Subtype 1 mitigates the host propositional content. 38 Subtype 2 mitigates the host propositional content and the speaker’s commitment. Subtype 3 mitigates the speaker’s commitment to the truth of the host clause. Subtype 4 mitigates the speaker’s commitment by attributing knowledge to the hearer or others. Although the expression comment clause is not mentioned, mitigative RPCs are the most closely related to comment RPC. In fact, type 1 deals with phatic expressions and type 2, as defined, applies to quotative parentheticals. The following data, presented by Schneider, illustrate each one of the mitigation subtypes: (8) Mitigation of the host propositional content ça faisait partie de ces disons donnés culturelles ‘That was part of these, let’s say, cultural data.’ (SCHNEIDER, 2007b, p. 245) (9) Mitigation of the host propositional content and the speaker’s commitment Porque si a mata<(d)>o a una mujer será cirurjano de Because if has killed to a woman be.FUT.3SG surgeon for mujeres, digo yo. women say I because if he has killed a woman he must be a surgeon for women, I say.’ (cf. SCHNEIDER, 2007b, p. 246-247) (10) Mitigation of the speaker’s commitment to the truth of the host clause la cosa su cui voglio intervenire su questo spero primo intervento the thing on which want intervene.INF on this hope.1SG first intervention è questo concetto di democrazia. is this concept of democracy ‘The issue I’d like to address in this, I hope, first talk is this concept of democracy.’ (SCHNEIDER, 2007b, p. 247) (11) Mitigation by attributing knowledge to the hearer or others para mí, los tíos, ya ves, son todos amigos mios. For me the guys, now see, are all friends my ‘For me the guys, you see, are all my friends.’ (SCHNEIDER, 2007b, p. 248) Matos, G: Comment Reduced Parenthetical Clauses and the syntax-discourse interface 39 Revista Letras, Curitiba, ufpr, n. 99, pp. 33-57, jan./jun. 2019. issn 2236-0999 (versão eletrônica) Schneider’s proposal to sharply distinguish report/quote from mitigative RPCs is too radical, as we saw by the cases in (5) and (6), and a more flexible approach would be preferable. Griffiths (2015) presents an alternative typology. He assumes that RPCs are definable along two dichotomies: 1) quote vs. speaker’s use, 2) report vs. attitude, characterized as follows: (12) a. Quote RPCs voice utterances over which content the speaker takes no responsibility. b. Speaker RPCs modify propositions used by the speaker to commit a speech act. These RPCs fall into three pragmatic subtypes: 1) Mitigatives, which diminish the speaker’s responsibility with respect to the truth of the host proposition. 2) Speech Act, which states the illocutionary force of the host. 3) Evaluative, which expresses the speaker’s emotional attitude regarding the host. c. Report RPCs describe the action of the speaker or another agent. d. Attitude RPCs expresses the attitude of the speaker concurrent with the speaker’s utterance time. Some of the Griffiths’ examples are presented below: (13) Quote RPCs a. “Orson,” I {think / say}, “must be fired.” b. “Orson” Mank thinks/ says “must be fired.” (GRIFFITHS, 2015, p. 74) (14) Mitigative RPCs a. John will, I believe, be late. b. John will, Mary heard, be late. (GRIFFITHS, 2015, p. 75) (15) Speech act RPCs I will, I {promise / swear / declare}, always love you. (GRIFFITHS, 2015, p. 75) (16) Evaluative RPCs a. My article should, I {hope / pray}, be accepted. b. John will, I regret to say, be late. (GRIFFITHS, 2015 p. 75-76) 40 The values presented in the first dichotomy may, in certain cases, combine with those of the second one: quote may combine with report; and speaker’s use may combine with report or attitude. Thus, the examples (13) combine quote and report values; (14a) reveals speaker’s use and attitude; (14b) evidences speaker’s use and reports values. Trying to establish a correlation between Griffiths’ typology and comment RPCs, we would say that these ones mainly correspond to speaker’s use RPCs, in particular, to the mitigative subtype, which Griffiths assumes to present an epistemic or evidential value, and to the Evaluative subtype, including Speech Act RPCs, when they exhibit an evaluative value. Previous studies on comment predicates in English seem to be compatible with these assumptions. 2.2 Comment predicates Some attempts have been made to establish a list of predicate expressions used in comment RPCs in in English (e.g. QUIRK et al., 1985; BRINTON, 2008). QUIRK et al. (1985) range the expressions occurring in comment RPCs, as opposed to report parenthetical clauses, into four classes, according to the attitude of the speaker with respect to the content expressed by the related clause: (17) (i) Class 1 – speaker’s tentativeness over the truth value of the related clause I believe, I think, I expect, I presume, I suspect, I guess, I assume, I suppose, I consider, I understand, I feel, I hear, I have read, I can see, I may assume, I dare say, I venture to say, one hears, they tell me, they allege, they say (ii) Class 2 – speaker’s certainty over the truth value of the related clause I know, I claim, I see, I remember, I agree, I admit, I’m sure, I’m convinced, I have no doubt. I must say, I must admit, I must tell you, I have to say (iii) Class 3 – speaker’s emotional attitude I hope, I wish, I fear, I regret, I’ m afraid, I regret to say, I’m sorry to say, I’m glad to say, I’m happy to say, I’m pleased to say, I’m delighted to say. (iv) Class 4 – speaker’s claim for hearer’s attention or agreement You know, you see, you realize, you can see, you can know, you may have heard, you must admit. Classes 1 and 2 present epistemic and evidential predicates, class 3 includes evaluative predicates and class 4 involves epistemic and evidential predicates in expressions with a phatic import. Matos, G: Comment Reduced Parenthetical Clauses and the syntax-discourse interface 41 Revista Letras, Curitiba, ufpr, n. 99, pp. 33-57, jan./jun. 2019. issn 2236-0999 (versão eletrônica) Relating this classification to Griffiths’ typology and leaving aside the expressions which present a phatic value, I will tentatively propose a non-exhaustive classification of the verbs that occur in comment RPCs in European Portuguese. (18) (i) Epistemic/Evidential predicates achar ‘believe’, crer ‘believe’, julgar ‘believe’, pensar ‘think’, dizer ‘say’, supor ‘suppose’, suspeitar ‘suspect, calcular ‘reckon’, deduzir ‘deduce’, presumir, ‘presume’, pressupor ‘presuppose’, ouvir ‘hear’. (ii) Evaluative predicates esperar ‘hope’, temer, recear ‘be afraid’, lamentar ‘regret/be sorry’ The following data exemplify these two classes: (19) O conflito ─ penso eu ─ não está superado. the conflict think1Sg I not is overcome ‘The conflict ─ I think ─ is not overcome.’ (20) (…) ele ainda não foi contratado, espero eu. he yet not has been hired, hope I ‘he has not yet been hired, I hope.’ (CETEMPÚBLICO 1.7, v. 9.1) These examples present the most characteristic structure of RPCs in European Portuguese: the object of the verb is missing and the subject occurs in post-verbal position. However, in EP, Subject omission is also frequent in comment RPCs, as in (21): (21) a. O conflito ─ penso ─ não está superado. the conflict think1Sg not is overcome b. ele ainda não foi contratado, espero he yet not has been hired, hope These two aspects of RPCs will be analyzed in the following sections. 3. The nature of the omitted argument The omitted argument of the verb in RPCs denotes the content of the host, both in comment and in quote parentheticals. This raises the question of its nature as a null pronoun that retrieves the content of its host, or a category resulting from movement of the host sentence or a constituent related to the host sentence. Although this problem is common to quote and comment RPCs, I will center the discussion on the second type of parentheticals. 48 Kluck and de Vries show that the inversion is obligatory in Dutch RPCs, a fact that they impute to the V2 requirement: (41) a. *Bob is, (zo) ik vermoed, een echte charmeur. b. *“Bob is”, (zo) Anna zei, “een echte charmeur.” (KLUCK; DE VRIES, 2015, p. 110) In turn, English allows Verb-Subject order with full DP (R-expressions) in quote RPCs, especially in the written language (COLLINS; BRANIGAN, 1997, p. 2), but with pronominal subjects, subject inversion produces marginal results (QUIRK et al., 1985; COLLINS; BRANIGAN, 1997), as shown in (42). Since comment RPCs mainly exhibit pronominal subjects, Verb/Subject Inversion does not typically occur (see (43)): (42) a. I wonder’, said John, ‘whether I can borrow your bicycle.’ b. ? “I’ve lost my keys” said he. (COLLINS; BRANIGAN, 1997, p. 7) (43) There were no other applicants, I believe, for that job. (QUIRK et al., 1985, p. 1112) In European Portuguese, with overt DPs and pronominals, the Verb-Subject order is obligatory in quotative RPCs (see (44)) and the usual order in comment RPCs (cf. (45)) and (46a), although the Subject-Verb pattern is also acceptable with certain predicates, as, for instance, the evaluative verb esperar ‘hope’, especially when the whole RPC is stressed, see (46b): (44) a. A minha pátria é a língua portuguesa - disse Pessoa e digo eu. the my homeland is the language Portuguese – said Pessoa and say I. ‘My homeland is the Portuguese language –said Pessoa and I (also) say it.’ (CETEMPÚBLICO, 1.7, v. 9.1) b. *A minha pátria é a língua portuguesa – Pessoa disse e eu digo. the my homeland is the language Portuguese – Pessoa said and I say. (45) a. Não há espaço para o empenho do escritor (…), penso eu. not there is space for the commitment of the writer think I ‘There is no room for the writer’s commitment, I think’. (CETEMPÚBLICO, 1.7, v. 9.1) b. *Não há espaço para o empenho do escritor (…), eu penso. not there is space for the commitment of the writer I think ‘There is no room for the writer’s commitment, I think’. Matos, G: Comment Reduced Parenthetical Clauses and the syntax-discourse interface 49 Revista Letras, Curitiba, ufpr, n. 99, pp. 33-57, jan./jun. 2019. issn 2236-0999 (versão eletrônica) (46) a. O tribunal decidirá, espero eu, quem tem razão no conflito (…) the court will decide hope I who has reason in.the conflict ‘The law court will decide, I hope, who is right in this conflict.’ (CETEMPÚBLICO, 1.7, v. 9.1) b. O tribunal decidirá, eu espero, quem tem razão the court will decide, I hope, who has reason ‘The court will decide, I hope, who is right. Thus, in European Portuguese, there are differences in the placement of the subject with respect to the verb in quote vs. comment RPCs, and some variation within the latter. This behavior diverges from Dutch, a V2 language, which systematically requires subject inversion, and from English, where the preverbal position of the subject is prevalent in RPCs with full DPs and required with pronominal subjects. I assume that the different patterns of subject placement in RPCs in a language like European Portuguese correspond to syntactic differences between comment and quote RPCs and I believe that these differences are related to their discursive structure and pragmatic meaning. 5.2 Pragmatic, discursive and syntactic differences between comment and quote RPCs Comment and quote RPCs pragmatically differ, since only the former may present a mitigative value, as shown in (47), where dizer ‘say’ occurs both in a quote parenthetical, diz a Bennett ‘Bennett says’, and in a comment RPC, digo eu ‘I say’, that mitigates the speaker’s commitment regarding the quote’s authorship: (47) «E tinha razão», diz a Bennett, digo eu. and has reason said the Bennett, say I and (s)he/I was right, said Bennett, I say.’ (CETEMPÚBLICO, 1.7, v. 9.1) These different values may be explicitly evidenced in the follow up of the mitigative and quote RPCs. While a comment expressing the uncertainty implied by the mitigative use of say is available, it is impossible to doubt that a quote has been expressed by using quotative say: (48) a. … diz a Bennett, digo eu mas não tenho a certeza. says the Bennett, says I, but not have the certainty ‘Says Bennett, I say, but I am not sure.’ 50 b. #*«E tinha razão», diz a Bennett mas não afirma. And has reason says the Bennett, but not state ‘“And (s)he was right”, says Bennett, but she did not state that.’ From a syntactic point of view, comment and quote RPCs also differ. The post verbal subject in comment RPCs occurs in adjacency to the verb and interposition of low or high adverbials produces marginal results, as in (49). This does not happen in quotative RPCs, as shown in (50): (49) a. Não há espaço para o empenho do escritor not there is space for the commitment of the writer 1. * — penso tristemente / francamente eu. think sadly / frankly I 2. √ — penso eu tristemente /francamente. think I sadly / frankly ‘There is no room for the writer’s commitment — I frankly/sadly think (so). b. O tribunal decidirá — {??espero evidentemente eu/√ espero eu evidentemente} the court will decide hope evidently / hope I evidently — quem tem razão. who has reason ‘The law court will decide, I evidently hope (so), who is right.’ (50) “O tsunami devastou a ilha. ” — noticiou claramente o repórter. the tsunami devastated the island announced clearly the reporter ‘”The tsunami devasted the island — the reporter clearly announced.’ The contrasts in acceptability between quote and comment RPCs in European Portuguese show that in a comment RPC adverb interposition between the verb and subject does not occur and suggest that these elements occupy higher functional projections than in quote RPCs. In European Portuguese, the subject position with respect to the verb may vary and determine different informational interpretations (AMBAR, 1992; COSTA, 2004; DUARTE; FIGUEIREDO SILVA, 2016). In this language variety, as in Spanish and Italian, informational focus subjects occur in post-verbal position2, either because they have been raised to FocP in the sentence low periphery (BELLETTI, 2004), or because they stayed in their original position inside vP (COSTA, 2004). The interposing of a low adverbial between the verb in T and the subject is, thus, expected, as shown in (51) for the sentence in (50): (51) Host — [ForcePparent [TopP Øopi [TP noticiouj [AdvP claramente [o reporter tj ti ]]]]] 2 Giorgi (2016, p. 484-485), for a similar proposal for Italian. Matos, G: Comment Reduced Parenthetical Clauses and the syntax-discourse interface 51 Revista Letras, Curitiba, ufpr, n. 99, pp. 33-57, jan./jun. 2019. issn 2236-0999 (versão eletrônica) In contrast, in comment RPCs high adverbs occur after the verb and the post-verbal subject. The cartographic approach to the left periphery of IP and CP, respectively, proposed in Rizzi and Cinque (2016) and Rizzi and Bochi (2017), offers a way to deal with the internal structure of comment RPCs: (52) a. CP[IP Mood-speech act >Mood-evaluative >Mood-evidential >Mod epistemic> TP…] (RIZZI; CINQUE, 2016, p. 149) b. [Force [Top* [Int [Top* [Foc [Top* [Mod(ification) [Top* [Q_emb [Fin [IP … (RIZZI; BOCHI, 2017) Considering that speech act adverbials (e.g. sinceramente ‘sincerely’) and evidential adverbs (e.g. evidentemente ‘evidently’) may follow the post-verbal subject, I consider that, in comment RPCs, the subject moves to spec, FocP, the position of contrastive focus and the verb may raise from T to Fin and then to Force, as represented in (52b). (53) a. Isso até nos atrasou um bocadinho, creio eu sinceramente… that even us delayed a little bit, believe I sincerely ‘That has even delayed us a bit, I believe.’ b. Host  [ForcePparent [Force creioj] [FocP eui [TopP Øop_k [FinP ti [IP sinceramente [ti tj tk Given this configuration, the non-interposition of adverbs between the verb and the subject is explained: these adverbs occur lower in the sentence structure. Notice that the possibility of stressing the whole RPC in sentences with preverbal subjects like (46b), repeated in (54a), also corroborates that contrastive focus is at stake in comment RPCs. Thus, I assume that the parenthetical in (54a) exhibits the structure in (54b): (54) a. O tribunal decidirá, eu espero, quem tem razão the court will decide, I hope, who has reason ‘The court will decide, I hope, who is right. b. Host  [ForcePparent [FocP eui [Focº esperoj] [TopP Øop_k [FinP [IP ti tj tk In (54) both the subject and the verb are interpreted as the contrastive focus and merged in FocP in the left periphery of the RPC, the subject in spec, FocP and the verb in Focº. 52 5.3 Interactions between Syntax-Discourse and Pragmatics in comment RPCs In V2 languages and in English, the mitigative reading of RPCs seems to be only related to the epistemic/evidential and evaluative content of verbs. However, in European Portuguese, the syntactic and discursive structure of the sentence is also relevant to establish the pragmatic value of the comment RPC. I believe that contrastive focus post-verbal subjects contribute to establish the mitigative value of the RPC. In fact, the contrast between comment RPCs with and without an overt subject shows that the overt focused post-verbal subject favors the mitigative interpretation while omitted subjects produce a weakening of the mitigative value (see (55a) vs. (55b) and (56a) vs. (56b)). (55) a. A publicidade alivia a depressão, acho eu the publicity mitigates the depression, believe I ‘Publicity mitigates depression, I think/believe’ (CETEMPÚBLICO, 1.7, v. 9.1) b. a publicidade alivia a depressão, acho. the publicity mitigates the depression believe ‘Publicity mitigates depression, I believe.’ (56) a. O tribunal decidirá, espero eu, quem tem razão. the court will decide, hope I who has reason ‘The law court will decide, I hope, who is right.’ b. O tribunal decidirá, espero, quem tem razão. the court will decide, hope who has reason ‘The law court will decide, I hope, who is right.’ This is corroborated in RPC with lamentar ‘to be sorry’. This verb has a factive meaning and cannot head a mitigative RPC, as shown by the marginality of (57b). In correlation lamentar in its evaluative reading precludes a post-verbal subject, as illustrated in (57a). When a post verbal subject occurs, the parenthetical clause is interpreted as quotative RPC, meaning ‘I say with regret’, as in (57c): (57) a. Não conheço ninguém, lamento. not know.1SG nobody, regret/ I’m sorry ‘I don’t know anybody, I’m sorry.’ (CETEMPÚBLICO, 1.7, v. 9.1) b. #/?? Não conheço ninguém, lamento mas posso estar enganado not know.1SG nobody, regret/I’m sorry but (I) might be wrong Matos, G: Comment Reduced Parenthetical Clauses and the syntax-discourse interface 53 Revista Letras, Curitiba, ufpr, n. 99, pp. 33-57, jan./jun. 2019. issn 2236-0999 (versão eletrônica) c. #Não conheço ning ué m , lamento eu not know.1SG nobody, regret I ‘I don’t know anybody, I regret.’ In sum, focused post-verbal subjects in comment RPCs headed by verbs with no factive content favor the mitigative interpretation. Turning now to comment RPCs with preverbal subjects, which arise with some epistemic and evaluative predicates in European Portuguese, we find that this word order may also consequences in the pragmatic value of the RPC: (58) a. As fotos não tinham a ver com a minha pintura ou desenho, eu acho. the photos not had to see with the my painting, or drawing, I think ‘The photos had nothing to do with my painting or drawing, I think.’ (CETEMPÚBLICO, 1.7, v. 9.1) b. Será difícil, eu receio bem, falar de justiça internacional. will be difficult I fear well, speak of justice international ‘It will be difficult, I’m afraid, to speak of international justice.’ c. O tribunal decidirá, eu espero, quem tem razão the court will decide, I hope, who has reason ‘The court will decide, I hope, who is right. In these sentences, the preverbal subjects express a greater degree of certainty of the speaker in his/her opinion, and the RPC is meant to mainly emphasize this attitude rather than to mitigate its host. This is particularly evident in the sequences that may follow the sentences containing stressed comment RPCs, as illustrated bellow: (59) O tribunal decidirá, {eu acho / eu espero}, quem tem razão, e esta é the court will decide, I believe / I hope, who has reason, and this is a minha posição. the my position ‘The court will decide,{I believe /I hope}, who is right, and this is my position.’ In sum, in European Portuguese the pragmatic value of Epistemic/Evidential and Evaluative RPCs does not exclusively rely on the content of the verbal predicates, but is also determined by the Verb/Subject order patterns. With verbs with no factive content, mitigative interpretations are stressed by post-verbal subjects contrastively focalized. 54 6. Final remarks This paper dealt with a specific type of comment parenthetical clauses, the reduced parenthetical clauses, confronting European Portuguese and V2 languages and English. Two main achievements have been attained: a first attempt to establish the most common predicates that occur in these clauses in European Portuguese and an analysis of their behavior in this language variety. We have seen that European Portuguese RPCs share with Dutch (and English) RPCs the fact that the null argument of the verb is the result of a moved constituent that merges to the left periphery of the parenthetical clause and A’-binds its copy. However, these languages differ with respect to how the discursive and pragmatic values of the comment RPCs are established. Apparently, in V2 languages and in English the pragmatic value of the RPCs is only determined by the classes of predicates that occur in these parentheticals, and the subject – verb order plays no role. In contrast, in European Portuguese, in addition to the verb classes involved, the word order patterns are also relevant to set the discursive and pragmatic value of the comment RPCs: post-verbal focused subjects stress the mitigative meaning of the parenthetical clause. Matos, G: Comment Reduced Parenthetical Clauses and the syntax-discourse interface 55 Revista Letras, Curitiba, ufpr, n. 99, pp. 33-57, jan./jun. 2019. issn 2236-0999 (versão eletrônica) References AMBAR, Maria Manuela. Para uma sintaxe da inversão sujeito verbo em Português. Lisboa: Edições Colibri, 1992. p. 72-73, p. 200-206. BELLETTI, Adriana. Aspects of the low IP area. In: Rizzi, Luigi (ed.). The structure of CP and IP. Oxford: Oxford University Press, 2004. p. 16-51. ISBN 0-19-515949-7. BRINTON, Laurel. The Comment Clause in English: Syntactic Origins and Pragmatic Development. Cambridge: Cambridge University Press, 2008. ISBN 978-0-521-88673-4. CETEMPUBLICO. Available at: https://www.linguateca.pt/CETEMPublico. CINQUE, Guglielmo. Adverbs and Functional Heads. New York/Oxford: Oxford University Press, 1999. ISBN 0-19-551527-9. COLLINS, Chris. Local economy. Cambridge, Massachusetts: MIT Press, 1997. pp. 13-61. ISBN 0-262-53144-5 COLLINS, Chris; BARNINGAN, Philip. Quotative inversion. Natural Language and Linguistic Theory, v. 15, n. 1, pp. 1-41, 1997. CORVER, Norbert; THIERSCH, Craig. Remarks on parentheticals. In: OOSTERDORP, Marc van; AGNASTOPOULOU, Elena (eds.). Progress in grammar: articles at the 20th anniversary of the comparison of grammatical models in Tilburg. Utrecht: Roquade, 2002. COSTA, João. Subject Positions and Interfaces: The case of European Portuguese. Berlin/New York: Mouton de Gruyter, 2004. DUARTE, Inês; FIGUEIREDO-SILVA, Cristina. The Null Subject Parameter and the Structure of the sentence in European and Brazilian Portuguese. In: WETZELS Leo; MENUZZI, Sérgio; COSTA, João (eds). The Handbook of Portuguese Linguistics. Malden, MA: Wiley-Blackwell, 2016. pp. 235-253. ISBN 978-1-11879195-0 EMONDS, Joseph. Parenthetical Clauses. You take the high node and I’ll take the low node. Chicago: Chicago Linguistics Society, 1973. p. 333-347. FORTMANN, Christian. The complement of reduced parentheticals. In: DÉHÉ, Nicole; KAVALOVA, Yordanka (eds.). Parentheticals. Amsterdam/Philadelphia: John Benjamins, 2007. p. 89-119. ISBN 978 90 272 3370 7. GIORGI, Alessandra. Integrated Parentheticals in Quotations and Free Indirect Discourse. In: Capone, Alessandro; KIEFER, Ferenc; LO PIPARO, Franco (eds.). Indirect reports and Pragmatics – Perspectives in Pragmatics, Philosophy and Psychology, 5. Switzerland: Springer, Cham, 2016. ISBN 978-3-319-21394-1. 56 GRIFFITHS, James. Speaker and quote reduced parenthetical clauses. In: SCHNEIDER, Stefan; GLIKMAN, Julie; AVANZI, Mathieu. Parenthetical Verbs, Linguistische Arbeiten 557. Berlin: Mouton De Gruyter, 2015. p. 71-102. ISBN-13: 9783110376036. JACKENDOFF, Ray. Semantic interpretation in generative grammar. Cambridge, Massachusetts: MIT Press, 1972. ISBN 0-262-10013-4. KLUCK, Marlies; DE VRIES, Mark. On V2, gaps, and operators in comment and reporting parentheticals. In: SCHNEIDER, Stefan; GLIKMAN, Julie; AVANZI, Mathieu. Parenthetical Verbs, Linguistische Arbeiten 557. Berlin: Mouton De Gruyter, 2015. p. 103-132. ISBN-13: 978-3110376036. MATOS, Gabriela. Appositive sentences and the structure(s) of coordination. In: TORCK, Danièle; WETZELS, Leo (eds.). Romance Languages and Linguistic Theory 2006. Amsterdam/Philadelphia: John Benjamins, 2009. p. 159-173. ISBN 978 90 272 4819 0. MATOS, Gabriela. Quotative Inversion in peninsular Portuguese and Spanish and in English. Catalan Journal of Linguistics, v. 12, p. 111-130, 2013. ISSN: 1695-6885 (in press); ISSN 2014 – 97 18 (online). QUIRK, Randolph; GREENBAUM, Sidney; LEECH, Geoffrey; SVARTVIK, Jan. A Grammar of Contemporary English. London: Longman, 1972. ISBN 0-582-52444X. QUIRK, Randolph; GREENBAUM, Sidney; LEECH, Geoffrey; SVARTVIK, Jan. A Comprehensive Grammar of the English Language. London: Longman, 1985. ISBN 0-582-51734-9. RAPOSO, Eduardo. On the Null object in EP. In: JAEGGLI, Osvaldo; SILVACORVALÁN, Carmen (eds.). Studies on Romance Linguistics. Dordrecht: Foris Publications, 1986. p. 373-390. ISBN 90 6765 253 9. REINHART, Tanya. Point of view in language - the use of parentheticals. In: RAUCH, Gisa. (ed.). Essays on Deixis. Turbingen: Gunter Narr Verlag, 1983. p, 169-194. ISBN 3878089597. RIZZI, Luigi. The Fine Structure of the Left Periphery. In: HAEGEMAN, Liliane (ed.). Elements of Grammar: A Handbook of Generative Syntax. Dordrecht: Kluwer/Springer, 1997. p. 281-337. ISBN 978-0-7923-4298-4 (book). RIZZI, Luigi. The fine structure of the left periphery. In: HAEGEMAN, Liliane (ed.). Elements of Grammar. Dordrecht: Kluwer, 1997. p. 281–337. RIZZI, Luigi; BOCCI, Giuliano. The Left Periphery of the clause – primary illustrated for Italian. In: The Wiley Blackwell Companion to Syntax. edition II. Hoboken: Blackwell Publishers, 2017. ISBN: 9781118358726 (print) 978 1118358733 (online). Matos, G: Comment Reduced Parenthetical Clauses and the syntax-discourse interface 57 Revista Letras, Curitiba, ufpr, n. 99, pp. 33-57, jan./jun. 2019. issn 2236-0999 (versão eletrônica) RIZZI, Luigi; CINQUE, Guglielmo. Functional categories and syntactic theory. In: LIBERMAN, Mark; PARTEE, Barbara (eds.). Annual Review of Linguistics – vol. 2. 2016. p. 139–63. ISBN 011415-040827. ROORYCK, Johan. Evidentiality, part I. Glot International, vol. 5, n. 4, p. 125-133, 2001. ROSS, John. Slifting. In: GROSS, Maurice; HALLE, Morris; SCHÜTZENBERGER, Marcel-P. (eds.). The Formal Analysis of Natural Languages – Proceedings of the first international conference. The Hague: Mouton, 1973. p. 133-169. SANTOS, Diana; ROCHA, Paulo. Evaluating CETEMPúblico, a free resource for Portuguese. In: Proceedings of the 39th Annual Meeting of the Association for Computational Linguistics (Toulouse, 9-11 July 2001), p. 442-449, 2001. SCHNEIDER, Stefan. Reduced Parenthetical Clauses as Mitigators. Amsterdam/ Philadelphia: John Benjamins, 2007a. ISBN 978 90 272 2301 2. SCHNEIDER, Stefan. Reduced parenthetical clauses in Romance languages. In: DÉHÉ, Nicole; KAVALOVA, Yordanka (eds.). Parentheticals. Amsterdam/ Philadelphia: John Benjamins, 2007b. p. 237-258. ISBN 978 90 272 3370 7. STEINBACH, Markus. Integrated parentheticals and assertional complements. In: DÉHÉ, Nicole; KAVALOVA, Yordanka (eds.). Parentheticals. Amsterdam/ Philadelphia: John Benjamins, 2007. p. 53-87. ISBN 978 90 272 3370 7. URMSON, James Opie. Parenthetical verbs. Mind, v. 61, n. 244, p. 480-496, 1952.