A Cognitive Approach to Adverbial Subordination in European Portuguese: The Infinitive, the Clitic Pronoun Se and Finite Verb Forms 1443829285, 9781443829281

Citation preview

A Cognitive Approach to Adverbial Subordination in European Portuguese

A Cognitive Approach to Adverbial Subordination in European Portuguese: The Infinitive, the Clitic Pronoun Se and Finite Verb Forms

By

Rainer Vesterinen

A Cognitive Approach to Adverbial Subordination in European Portuguese: The Infinitive, the Clitic Pronoun Se and Finite Verb Forms, by Rainer Vesterinen This book first published 2011 Cambridge Scholars Publishing 12 Back Chapman Street, Newcastle upon Tyne, NE6 2XX, UK British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Copyright © 2011 by Rainer Vesterinen All rights for this book reserved. No part of this book may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior permission of the copyright owner. ISBN (10): 1-4438-2928-5, ISBN (13): 978-1-4438-2928-1

TO MY FAMILY GRÖNBLAD AND VESTERINEN

TABLE OF CONTENTS

Acknowledgements .................................................................................... ix Chapter One................................................................................................. 1 Introduction Research Questions................................................................................ 4 Some Earlier Proposals ........................................................................ 11 The Plain vs. the Inflected Infinitive.................................................... 12 The Clitic Pronoun Se .......................................................................... 15 Finite and Infinitive (adverbial) Clauses.............................................. 18 Synthesis .............................................................................................. 20 The Model of Cognitive Grammar....................................................... 21 Grammatical Structure ......................................................................... 23 A Usage Based Model ......................................................................... 25 Subordination in Cognitive Grammar .................................................. 26 Semantic Structure ............................................................................... 29 Encyclopedic and Contextual Semantics ............................................. 29 Profile and Base ................................................................................... 30 Trajector and Landmark....................................................................... 31 Working Hypotheses............................................................................ 33 Material and Method............................................................................ 35 Chapter Two .............................................................................................. 39 The Degree of Prominence in Infinitive Adverbial Clauses: Cognitive Aspects Introduction.......................................................................................... 39 The Relation Between Trajector and Landmark .................................. 42 Syntactic Relation ................................................................................ 43 Semantic Relation ................................................................................ 47 The Prominence Scale.......................................................................... 51 Analysis ............................................................................................... 54 The Pragmatic Approach ..................................................................... 56 The Cognitive Approach...................................................................... 62 Spoken Language................................................................................. 68 Written Language................................................................................. 73 Conclusion ........................................................................................... 76

viii

Table of Contents

Chapter Three ............................................................................................ 79 The Clitic Pronoun Se: A Generic Trajector Introduction.......................................................................................... 79 The Clitic Pronoun Se .......................................................................... 81 Function and Semantic Meaning.......................................................... 81 Defocalization and Focus Displacement .............................................. 84 Se as an External Force ........................................................................ 85 Analysis ............................................................................................... 87 The Degree of Transitivity................................................................... 88 Obvious Referents................................................................................ 95 The Internal Position............................................................................ 97 The Verbal Inflection ........................................................................... 99 Different Conceptualizations ............................................................. 104 Conclusion ......................................................................................... 108 Chapter Four ............................................................................................ 111 Finite Adverbial Clauses: Iconicity, Subjectification and Mental Spaces Introduction........................................................................................ 111 Iconicity ............................................................................................. 115 Subjectification .................................................................................. 118 Mental Spaces .................................................................................... 122 Analysis ............................................................................................. 128 Infinitive Clauses Vs. Indicative Clauses........................................... 128 Infinitive Clauses Vs. Subjunctive Clauses ....................................... 140 Conclusion ......................................................................................... 157 Chapter Five ............................................................................................ 161 Final Considerations Summary............................................................................................ 161 The Cognitive Approach.................................................................... 168 References ............................................................................................... 177 Index........................................................................................................ 187

ACKNOWLEDGEMENTS

The present study is a revised and updated English version of my PhD. thesis in the field of Portuguese linguistics: Subordinação adverbial: um estudo cognitivo sobre o infinitivo, o clítico se e as formas finitas em proposições adverbiais do Português Europeu. The thesis is the result of my time as a student within the National School of Romance Languages (FoRom). I would like to thank all my colleagues within Forom for the creative and stimulating atmosphere during the process that led to the conclusion of the thesis. In particular, I would like to thank my supervisors, Lars Fant and Johan Falk, for their support and comments on different parts of the thesis. Also, my thanks go to Fernando Bermúdez for his insightful comments and to Manne Bylund for his interest in the development of the original thesis, and for his help and encouragement during the work with the present publication. But most of all, I would like to thank Manne Bylund for our friendship. To express thoughts and ideas on a language different from your mother tongue always entails some difficulties. This was certainly the case with the original thesis in Portuguese, and the present publication is not an exception to the rule. In fact, three different languages were involved in the process: a native speaker of Swedish translated–and rewrote–a text that originally was written in Portuguese to English. The process has been stimulating, but it did also create doubts about how to find adequate English expressions. In sum, I would like to thank John Farrow and Gary Watson for efficiently and cheerfully checking my English writing. A special gratitude goes to Augusto Soares da Silva (Universidade Católica de Braga, Portugal). Augusto Soares da Silva was my opponent at the dissertation act of the original thesis and made an excellent contribution. The dissertation act was also the beginning of warm and intellectually stimulating friendship. Finally, I would like to thank my family for their support and love. Family life is a continuous reminder that there is a life beyond adverbial clauses in European Portuguese. Thank you Lisa and Marve for reminding me of that! Stockholm, April 2011

CHAPTER ONE INTRODUCTION

Since the breakthrough of the generative approach to language and language use, there is a tradition within linguistic studies to consider language as an abstract system without taking into consideration the language user and his/her cognitive faculties. Grammar–and syntax in particular–is studied in isolation on the basis of the assumption that it has no relation to semantics. The underlying idea is that semantics is to be found in lexicon and not in language structure. The present study is at odds with this view of linguistic analysis. Thus, in accordance with the perspective of Cognitive Grammar (Langacker 1987, 1991, 2008), I will contend that there is a strong relation between language as a system for creating meaning and human cognitive faculties. Moreover, I will argue that grammar (and syntax) is meaningful in its own right. In this sense, language structure mirrors conceptual structure. This being so, the present book aims at creating a greater understanding about finite and infinitive adverbial clauses in European Portuguese from a cognitive linguistic perspective of language and language use. In particular, I will examine the reason why the native speaker of European Portuguese uses one or another verb form in different types of adverbial clauses. As will be shown throughout the book, the variation between finite and infinitive verb forms in adverbial clauses is highly flexible in Portuguese due to the existence of an inflected infinitive with subject agreement. That is, the Portuguese inflected infinitive shares the feature [person] with the finite verb forms. On the other hand, it lacks the feature [tense]. It is, so to speak, a verb form situated in between the plain infinitive and the finite verb forms with reference to morphological information. The inflection pattern of the inflected infinitive is illustrated in Figure 1:1 below:

2

Chapter One

Figure 1:1. The Portuguese inflected infinitive Ler (‘to read’) Singular 1 ler (ø) 2 leres 3 ler (ø)

Plural 1 lermos 2 lerdes 3 lerem

The inflected infinitive appears in a wide array of grammatical contexts and, in fact, the complete distribution of this particular verb form is not yet totally covered (cf. Scida 2004: 3). However, it is commonly used as a complement to epistemic, declarative, factive, perception and causation verbs in subordinate clauses. With this being the case, the Portuguese speaker can choose between using a finite verb form introduced by the conjunction que (‘that’) or simply exclude the conjunction and use the infinitive instead of the finite verb form. This is also the case for impersonal subject clauses, e.g., it is necessary to buy, and for the theme of the present book–adverbial clauses introduced either by a preposition or a conjunction. Let us consider some typical examples displaying the variation between finite and infinitives in subordinate contexts: 1.

a)

Penso não dizerem a verdade say-INF. 3P.P the truth Think-PRES: 1P.S not Penso que não dizem a verdade not say-PRES: 3.P.P the truth Think-PRES: 1P.S that ‘I think they are not telling the truth’ [epistemic verb: ‘to think]

b) Afirmo não dizerem a verdade say-INF: 3P.P the truth Declare-PRES: 1P.S not Afirmo que não dizem a verdade not say-PRES: 3.P.P the truth Declare-PRES: 1P.S that ‘I declare that they are not telling the truth [declarative verb: “to declare”]

c) Lamento

não dizerem a verdade say-INF: 3P.P the truth Regret-PRES: 1P.S not Lamento que não digam a verdade not say-PRES.SUBJ: 3P.P the truth Regret-PRES: 1P.S that ‘I regret that they are not telling the truth’ [factive verb: “to lament”]

d) Vejo os miudos brincarem See-PRES: 1P.S the children play-INF: 3P.P Vejo que os miudos brincam the children play-PRES: 3P.P See-PRES: 1P.S that ‘I see the children playing [perception verb: “to see”]

Introduction

3

e) Faz os alunos estudarem Make-PRES: 3P.S the pupils study-INF: 3P.P Faz com que os alunos estudem with that the pupils study-PRES.SUBJ: 3P.P Make-PRES: 3P.S ‘She/he makes the pupil study’ [causation verb: “to make”] f) É preciso Be-PRES: 3P.S necessary É preciso Be-PRES: 3P.S necessary ‘It is necessary that they “to be necessary”]

dizerem a verdade say-INF: 3P.P the truth que digam a verdade that say-PRES.SUBJ: 3P.P the truth tell the truth [impersonal subject clauses:

g) Faz tudo para everything to Do-PRES: 3P.S Faz tudo para everything to Do-PRES: 3P.S a verdade the truth ‘He/she makes a great effort to make purpose clause: “to”]

dizerem a verdade say-INF: 3P.P the truth que digam that say-PRES.SUBJ: 3P.P them tell the truth’ [adverbial

The examples displayed above illustrate the astounding possibilities for using the infinitive in Portuguese. While many other languages require the use of a finite verb form in subordinate clauses with a non co-referential or a new subject, Portuguese displays an alternative structure with the inflected infinitive. This is also a particular feature of the Portuguese language among the family of national Romance languages. For example, causal adverbial clauses can be introduced either by the preposition por (‘through’) followed by the inflected infinitive or by the conjunction porque (‘because’) followed by a finite verb form. Examples (2a-e) below show that the other national Romance languages behave differently: 2.

a)

O menino está triste por sairmos through go out-INF: 1P.P The boy be-PRES: 3P.S sad O menino está triste porque saímos because go out-PRES: 1P.P The boy be-PRES: 3P.S sad ‘The boy is sad because we are going out’ [Portuguese]

b) Le garçon est trist parce qu’ on sort because one go out-PRES: 1P.S The boy be-PRES: 3P.S sad ‘The boy is sad because we are going out’ [French]

4

Chapter One c) Il ragazzo è triste perché usciamo because go out- PRES: 1P.P The boy be-PRES: 3P.S sad ‘The boy is sad because we are going out’ [Italian] d) El muchacho está triste porque The boy be-PRES: 3P.S sad because ‘The boy is sad because we are going out’ [Spanish]

salimos go out-PRES: 1P.P

e) El muchacho esta triste por salir nosostros through go out-INF. we The boy be-PRES: 3P.S sad ‘The boy is sad because we are going out’ [Spanish]

Thus, the other national Romance languages prefer–or require–the insertion of a finite adverbial clause in these contexts while Portuguese permits a variation between infinitive and finite verb forms. As can be seen in example (2e.), however, Spanish may permit adverbial structures like: por salir nosotros. On the other hand, examples of this kind do not reflect standard Spanish and are often considered as “regional variants” or as cases of “not standard Spanish” (cf. De Mello 1995). This is not the case with infinitive structures in Portuguese. On the contrary, the infinitive is highly productive in a wide array of subordinate contexts. The present book will focus on a particular grammatical context, namely the variation between infinitive and finite adverbial clauses. The possibility to choose between different verb forms in this grammatical context indeed raises questions about the nature of this variation and about the reasons for choosing one or another verb form. These questions will be examined in more detail in the following section.

Research Questions As referred to above, the variation between infinitive and finite adverbial clauses in Portuguese is related to the existence of the inflected infinitive. However, the fact that there are two different categories of infinitives–the plain infinitive and the inflected infinitive–subsumes an even more complex situation. On the one hand, there is a feasible variation between the plain and the inflected infinitive in some grammatical contexts. On the other hand, there is a variation between the inflected infinitive and finite verb forms in other grammatical contexts. Hence, the adverbial clause may include the plain infinitive, the inflected infinitive or a finite verb form:

Introduction

5

3.

eu detesto pessoas, pá, que andem à boy, that go-PRES.SUBJ: 3P.P to the I detest-PRES: 1P.S people, procura de empregos e que não se that not CLITIC.PRON. search of work-PL. and movam para procurar emprego to search-INF. work move- PRES.SUBJ: 3P.P [Pfu: 140] 1 ‘I detest unemployed persons that do not make any effort to find work’

4.

tenho um processo de darlhes um bocadinho process of give-INF. them a little bit have-PRES: 1P.S a de limão para eles picarem com um garfo no with a fork on the of lemon to they prick-INF: 3P.P limão lemon [Pfu: 85] ‘I usually give them a bit of lemon so that they can prick it with a fork’

5.

Inclusivamente chegavase ao apuro de CLITIC.PRON. to the difficulty of Also come-IMP: 3P.S ter que pedir autorização, se fosse that ask for-INF. authorization, if be-IMP.SUBJ: 3P.S have-INF. necessária uma operação ou qualquer coisa para que a necessary a operation or something to that the criança fosse operada… child be-IMP.SUBJ: 3P.S operated… [Pfu: 85] ‘We also encountered the difficulty of having to ask for permission in order to get an operation for the child, if it would be necessary’

Examples (3-5) differ regarding the verb form in the adverbial clause. Example (3) displays the plain infinitive, without any morphological markings of the feature [person]. That is, the infinitive procurar does not express an explicit subject, but shares the same subject with the main clause. This situation implies a co-referential subject relation between the main clause and the adverbial clause. Another pattern is observed in the two following examples where the subject is overtly expressed in the adverbial clause. In (4) the subject is expressed by the inflected infinitive form picarem (the third person plural inflection), and in (5) it is expressed by the 1

The examples are selected from different corpora of natural spoken and written European Portuguese: Português Falado (Pfa), Português Fundamental (Pfu) and Diaclav (name of the newspaper). The corpora will be presented more in detail in the method section.

6

Chapter One

finite verb form fosse in the subjunctive mood. In these latter cases (4-5), the subject in the adverbial clause is non co-referential with the subject of the main clause. Thus, examples (3-5) demonstrate some rather obvious morphological differences between the plain infinitive, the inflected infinitive and the finite verb forms. The plain infinitive does not express explicit subject marking, while the opposite holds for the inflected infinitive. On the other hand, the inflected infinitive shares the feature [person] with finite verb forms, but does not indicate tense. It is the finite verb forms that express this feature, together with the feature [mood]. This being so, the plain infinitive, the inflected infinitive and the finite verbs represent a continuum of morphological marking: Figure 1:2. The morphological marking of the plain infinitive, the inflected infinitive and finite verb forms. Plain infinitive [-person], [-tense] Inflected infinitive [+person], [-tense] Finite verb [+person], [+tense], [+mood]

The parameter [+/-person] could contribute to an explanation in which the uses of the plain and the inflected infinitive is basically a referential issue. In other words, one could argue that the plain infinitive is appropriate in co-referential contexts and that the inflected infinitive, in contrast, is used to indicate that the subject of the adverbial clause is non co-referential with the main clause subject. The following examples (6-7) illustrate that this explanation is not entirely correct: 6.

É interessante notar como pessoas ligadas Be-PRES: 3P.S interesting note-INF. how persons connected a vários partidos souberam dar as mão hand to different parties know-PRET: 3P.P give-INF. the para encontrar soluções para Leiria e a sua região Leiria and its region to find-INF. solutions to [Diário de Leiria-N2989-2] ‘It is interesting to observe that persons connected to different parties cooperated to find solutions for Leiria and the region’

Introduction 7.

7

e era assim que eles iam a pé like this that they go-imp: 3p.p to foot and be-IMP: 3P.S de alco[baça], da nazaré para alcobaça para from alco[baça], from the nazaré to alcobaça to se entregarem à cadeia, não eram CLITIC. PRON. give up-INF: 3P.P to the prison, not be-IMP: 3P.P acompanhados por guardas nem nada accompanied by guards nor nothing [Pfu: 291] ‘And the story is that they went by foot from Nazaré to Alcobaça to give themselves up to the prison, they were not accompanied by guards or by anyone’

Accordingly, the question of co-referentiality is not fundamental in order to understand the variation between the plain and the inflected infinitive. It is true that the use of the plain infinitive is prototypical for co-referential contexts like (6), but not even this statement is entirely correct. It is, for example, possible to use the plain infinitive in cases where the indirect object of the main clause turns to be the subject of the adverbial clause, e.g., deram-lhe dinheiro para comprar um carro (‘they gave him/her money to buy a car’). But perhaps even more interesting is the fact that examples like (7) are highly productive. The inflected infinitive is not only used in non coreferential contexts, but do also occur in co-referential contexts. Against this background, I will try to find a plausible explanation for the variation between the plain and the inflected infinitive in co-referential contexts. This will be done in Chapter 2. Another question that will be dealt with is related to the distinction between the plain and the inflected infinitive and the use of the clitic pronoun se. The examples given so far (3-7) do not raise any problems regarding the distinction between the plain and the inflected infinitive given the lack of inflection on the infinitive in contexts where the main clause has plural subjects. 2 But this distinction is more hazardous to make in contexts where the infinitive has a zero inflection–in the first and third person singular. The traditional solution to this problem has been to claim that the infinitive is inflected if it refers to a Nominal Phrase (NP) or if it can be preceded by a pronominal subject (cf. Maurer 1968). Let us consider some typical cases:

2

The infinitive procurar in (3), for example, does not agree with the main clause subject expressed by the reflexive se movam in the subjunctive third person plural mood.

8

Chapter One 8.

para aprender a doutrina para ir a minha primeira go-INF. to my first to learn-INF. the doctrine to comunhão, eu sei la quanto Holy Communion, I know-PRES: 1P.S there much tempo andei time go-PRET: 1P.S [Pfu: 83] ‘I do not know how much time it took for me to learn the doctrine and to attend my first Holy Communion’

9.

nem sei tampouco para onde caminhará isto, to where walk-FUT: 3P.S this nor know-PRES: 1P.S neither apesar de eu ser um jovem e ter ideias young and have-INF. ideas in spite of I be-INF. a avançadas advanced [Pfu: 148] ‘nor do I know what this will lead to, despite the fact that I am young and have advanced ideas’

The common denominator of these two cases is that the action expressed in the adverbial clause is related to a certain element. In (8), the infinitive refers to the pronominal subject eu (‘I’) and the verbal inflection sei (‘I know’) in the main clause. Moreover, it would be possible to insert the pronominal subject in the adverbial clause, e.g., para eu aprender…para eu ir (‘so that I learn…so that I go’). This reasoning implies that the infinitive should be regarded as an inflected one. The overtly expressed subject in the subsequent case (9), on the other hand does not leave any space for argument–the infinitive in the adverbial clause is inflected. One problem that arises from this explanation is that it only seems to cover two different infinitive contexts: impersonal subject clauses and subordinate clauses with plural subjects. For instance, example (6), e.g. é interesante notar (‘it is interesting to note’) would be categorized as a plain infinitive because of the lack of inflection and of the impersonal context. Equally, (3) is considered a plain infinitive given the lack of inflection of the infinitive in a plural context. The traditional explanation also creates a problem related to the different uses of the infinitive in adverbial clauses. Apparently, the infinitive can be preceded by personal pronouns which make the notion of the inflected infinitive more salient. The question, though is: what happens when the pronoun designates an impersonal subject? This is often the case with the clitic pronoun se. Let us consider two cases that differ substantially in this respect:

Introduction

9

10. De acordo com o vidreiro, «é preciso um Of agreement with the glassmaker, be-pres: 3p.s necessary one mês para se conseguir fazer uma peça CLITIC.PRON. achieve-INF make-INF. one piece month to com perfeição, e outros cinco, para se aprender CLITIC.PRON. learn-INF. with perfection, and others five, to seis modelos diferentes». six models different [Diário de Aveiro-N0571-1] ‘According to the glassmaker, it takes one month to learn how to make one perfect piece and five more to learn how to make six different models’ 11. Já demos Already give-PRET: 1P.P fazerem os estudos... make-INF: 3P.P the studies… projecto quanto custa, qual project much cost, which [Diário de Coimbra-N2512-1] ‘We have already indicated that and find the best location for it’

indicação para se indication to CLITIC. PRON. para vermos em termos de to see-INF: 1P.P in terms of a melhor localização. the best location. they should study the costs of the project

Although the examples displayed above (10-11) show some similarities, they differ in terms of morphological marking. This difference has some immediate consequences for the interpretation of the pronoun se. The plain infinitive in (10), for example, increases the inclination to consider se as an impersonal subject. That is, se designates an active initiator of the process described in the adverbial clause. In contrast, the verbal inflection in (11), referring to the NP (os estudos), confers the subject role to this participant, and this has the effect of decreasing the notion of an active agent. Nonetheless, it is obvious that the studies cannot be done without the presence of an active participant. In order to understand this dilemma, the semantic meaning of the structure PREPOSITION + SE + INFINITIVE needs to be further analysed. This will be the topic of Chapter 3. The following question is related to the variation between inflected infinitive adverbial clauses and finite adverbial clauses. As has been shown, both structures include morphological marking of the feature [person]. Nonetheless, it is only the finite adverbial clause that situates the adverbial process in a specified time. In other words, it is only the finite clauses that determine this process in relation to the time of the communicative event:

10

Chapter One 12. e às onze e meia tenho de o pôr of the put-INF. and to the eleven and half have-PRES: 1P.S (…) para (…) para as raparigas, para elas ao (…) to (…) to the girls, to they to the depois terem lá comer ao meio-dia to the middle day later have-INF: 3P.P there eat-INF. [Pfu: 87] ‘At half past eleven I have to put the food on so that the girls have something for lunch 13. quando fui para o curso de direito, fui the course of law, go-PRET: 1P.S when go-PRET: 1P.S to na convicção de que auxiliar a justiça, ou contribuir in the conviction of that help-INF the justice, or contribute-INF para que se fizesse justiça no CLITIC. PRON. make-IMP.SUBJ: 3P.S justice in the to that mundo era um ideal bastante elevado a ideal rather high world be-IMP: 3P.S [Pfu: 97] ‘when I attended law school, I went there with the conviction that I could contribute to justice in the world; this was a rather high ideal for me’

The verbal inflection of the infinitive adverbial clause (12) gives no temporal indication of the described process, while this is done in the subsequent example (13). In this particular case, the preterite tense in the subjunctive mood has the effect of dislocating the act of making justice in the world to a time prior to the communicative event. Therefore, it would be tempting to argue that the variation between infinitive and finite adverbial clauses first and foremost depends on the parameter [+/-tense]. The use of a finite adverbial clause would then be motivated by the speaker’s need to locate the adverbial process temporally in relation to the communicative event, while the infinitive adverbial clause would be used in situations where there is no such need. However, I believe that there may also be other factors involved in the variation between these two structures. Certainly, the temporal dimension is important to reconsider, but it seems that this dimension does not reveal the full complexity of the variation between infinitive and finite adverbial clauses. In Chapter 4, this question will be studied more thoroughly. To summarize, the research questions can be stated as follows: •

What motivates the variation between the plain and the inflected infinitive in co-referential Portuguese adverbial clauses?

Introduction

• •

11

What is the meaning and function of the clitic se in Portuguese adverbial clauses? What motivates the variation between inflected infinitive adverbial clauses and finite adverbial clauses?

My impression is that these questions by tradition have received a rather formal treatment. There has been a tendency to study the variation between the Portuguese infinitive and finite verb forms by formulating abstract rules in order to explain the occurrence of the inflected infinitive. This tendency is most likely to be found within the Generative Grammar approach to linguistic analysis. Traditional grammarians, on the other hand, have often contributed with pure descriptions of the variation without aiming too hard to find an explanation (cf. Bechara 1999; Brito 1995; Caetano Silveira et al. 1994; Cunha and Cintra 1984; Maurer 1968; Perini 1977; Raposo 1975, 1987). It is not my intention to challenge the studies referred to above. On the contrary, they formulate important and insightful contributions that increase our understanding about the variation between the Portuguese infinitive and finite verb forms. Nevertheless, I believe that the issue also needs to be addressed from a perspective that considers the cognitive processes that may come into play in producing one or another adverbial structure, and from a perspective that conceives of language as a tool for creating meaning –not only from a lexical standpoint but also from a syntactical one. Therefore, the perspective that I will adopt in the present study is based on theories within the cognitive linguistics paradigm, and on the model of Cognitive Grammar in particular (Langacker 1987, 1991, 2008). As an alternative to the traditional view, I will try to establish a relation between different subordinate adverbial structures and their different meanings. In the following section, though, the traditional way of looking at these questions will be reviewed.

Some Earlier Proposals It is indeed impossible to provide a review that fully covers everything that has been written on the variation between different adverbial structures in Portuguese. The following review will therefore focus on the most important and representative contributions on the matter. The review will further reveal a tendency within Portuguese linguistics to study the infinitive and finite verb forms from a purely formal or descriptive perspective. First, the variation between the plain and the inflected infinitive will be reviewed. This review will be followed by the use of the clitic

12

Chapter One

pronoun se. Finally, focus will be put on earlier proposals regarding the variation between the inflected infinitive and finite verb forms.

The Plain vs. the Inflected Infinitive Cunha and Cintra (1984) comment on the use of the plain and the inflected infinitive–the impersonal and the personal infinitive in their terminology– by saying that it is one of the most controversial questions in Portuguese syntax. They conclude that the rules created by grammarians to explain the use of one or another infinitive form have always been insufficient or unrealistic. Instead, the authors point at non grammatical factors in order to find an explanation to the problem: In fact, during the different stages of the Portuguese language, writers have never looked upon this question by basing themselves on reasons of purely grammatical logic, but in the act of choosing they have always been influenced by motives related to stylistics, such as the rhythm of the phrase, the emphasis of the uttering, the clarity of the expression (Cunha and Cintra 1984: 482) 3

Further, the authors agree with Bechara (1999) considering that the variation between the two infinitive forms may be motivated by the propensity to empathize different aspects of the adverbial clause. Bechara (1999) explains the difference in the following way: “The plain infinitive reveals that our attention is drawn especially to the verbal action; the function of the inflection is to insist on the personal subject” (Bechara 1999:285-286). 4 In other words, the plain infinitive is used in order to emphasize the action, while the inflected infinitive emphasizes the agent of the action. This is illustrated by the subsequent example:

3

“Em verdade, os escritores portugueses das diversas fases da língua portuguesa nunca se pautaram, no caso, por exclusivas razões de lógica gramatical, mas viramse sempre, no acto da escolha, influenciados por ponderáveis motivos de ordem estilística, tais como o ritmo da frase, a ênfase do enunciado, a clareza da expressão” (Cunha and Cintra 1984: 482). 4 ”O infinitivo sem flexão revela que a nossa atenção se volta com especial atenção a ação verbal; o flexionamento serve de insistir na pessoa do sujeito” (Bechara 285286).

Introduction 14. Estudamos Study-pres: 1p.p

a) para to

vencer win-INF

13 na vida in the life

b) para vencermos na vida to win-INF: 1P.P in the life ‘We study in order to be successful in life’

Thus, the speaker expressing (14a) is more concerned with the verbal act of “being successful”, while the person uttering (14b) puts emphasis on the subject of the verbal act. This explanation may very well be true. The problem, however, is that it does not say anything about the underlying reasons for choosing one or another expression. The question that remains is: why would a given speaker feel a need to emphasize one or another facet of the adverbial process? Maurer (1968) elaborates three rules to facilitate our understanding of the Portuguese infinitive. The rules are created in order to explain the obligatory use of the plain infinitive and the inflected infinitive, and to show the contexts that allow a variation between the two infinitives. The first rule refers to the use of the plain infinitive and to the premises that determine its adequate use. The second rule concerns the inflected infinitive and its obligatory use in certain contexts. Finally, the third rule explain the contextual premises that permit a variation between the plain and the inflected infinitive. The rules are listed below: A. When the infinitive is rigorously impersonal, that is, when the action expressed by it does not refer to a determined agent either expressed by the proper infinitive, or indicated in the context, it is always invariable (impersonal). B. When the infinitive is clearly personal, that is, when it has a proper subject –expressed or not – it is always inflected, no matter if the subject is identical or not with the main verb subject. C. When the infinitive, despite not having a proper subject – therefore having an impersonal use in other romance languages– expresses an action that is nonetheless put forward by an agent known from the context, it may be inflected or invariable, although there is often a preference for the use of one or another form of the infinitive (cf. Maurer 1968: 135, 145, 153).

A question that arises from these rules is how to distinguish rule B from rule C. In fact, it is difficult to recognize the difference between infinitives with a proper subject that may or may not be expressed from infinitives that have no proper subject but express an action put forward by an agent known from the actual context. This makes it hard to fully comprehend the application of the rules listed above.

14

Chapter One

In his analysis of the variation between the plain and the inflected infinitive in adverbial clauses, Maurer (1968) concludes that the latter is more commonly used in this grammatical context. The reason for this, he claims, is that the insertion of a linguistic unit (the preposition) between the main verb and the infinitive makes the adverbial clause more independent of the main clause. That is, the greater formal distance between the main verb and the infinitive leads to a tendency to use the inflected infinitive form: …when the proper infinitive constitutes a type of clause…it acquires a greater autonomy in its relation to the main verb and, obviously, this is expressed by the inflected form of the infinitive in a language that, like the Portuguese, possesses this syntactic resource (Maurer 1968: 170). 5

Another interpretation of the inflected infinitive is put forth by Perini (1977). From a generative perspective of language the author formulates a rule called the double inflection filter (‘o filtro de dupla desinência’). According to the author, the rule marks all surface structures with the inflected infinitive as grammatically incorrect if they are preceded by a verb with identical subject marking: The double inflection filter All surface structures containing an inflected IP (Infinitive Phrase) that is identical with another verb having identical inflection should be marked as ungrammatical, except in cases with a NP preceding the co-referent subject of the infinitive, given that there is no other verb between the NP and the infinitive (cf. Perini 1977: 107).

This being so, an example like: os jardineiros vieram para ajudarem (‘the gardeners came to help’) is considered ungrammatical. In the transformation from the deep structure to the surface structure, the IP loses its inflection when it passes the double inflection filter. There is, so to speak, no reason to mark a co-referential subject in the IP. This argument is fundamentally based on the assumption that both infinitives–the plain and the inflected–have a subject in the deep structure, and that the subject may be “turned off” or “removed” by different rules and syntactic processes (cf. Raposo 1975). It is interesting to note, though, that the violation of the

5

“…no momento em que o infinitivo constitui uma oração…êle adquer uma autonomia maior em relação ao verbo de que depende, o que se exprime, naturalmente, pela forma flexionada em uma língua que, como o Português, possui esse recurso sintático” (Maurer 1968:170).

Introduction

15

double inflection filter is more acceptable if there is a greater linear distance between the main verb and the infinitive (cf. Perini 1977: 85). Raposo (1987) studies the distributional properties of the Portuguese infinitive from the theoretical framework of government and binding. The hypothesis put forward is that clauses containing the inflected infinitive belong to the NP category and therefore may be attributed case from an external case assigner (verb, preposition or inflection). This is made possible by two parameters: the inflection parameter and the null subject parameter. The first parameter holds that Portuguese, in contrast to many other languages, possesses the option to have subject agreement without expressing the feature [tense]. The subsequent parameter posits that the subject agreement leads to the possibility of assigning nominative case to a lexical subject if the infinitive itself is specified for case (cf. Raposo 1987: 92). These findings have in later stages of generativist theory been modified by Caetano Silveira et al. (1994). Basing themselves on the model of principles and parameters the authors claim that the functional node of the NP is the determiner. Therefore, the inflected infinitive should be regarded as a DP and not a NP.

The Clitic Pronoun Se Traditional grammarians often consider the clitic pronoun se as not one but two different pronouns. Together with a transitive verb this pronoun is a passive marker, and with an intransitive verb (or with a transitive verb used intransitively) it is a symbol representing an impersonal subject. Cunha and Cintra (1984: 308-309) expose the following cases in order to explain the difference: 15.

Transitive verbs: se as a passive marker. a) ouvese ainda o toque de rebate CLITIC. PRON. still the signal of alarm hear-PRES: 3P.S ‘The alarm signal is still ringing’ b) fezse make-PRET: 3P.S CLITIC. PRON. ‘It became silent again’

de of

novo new

silêncio silence

16

Chapter One 16.

Intransitive verbs: se as a symbol for an undetermined subject. vivese ao ar livre, comelive-pres: 3p.s CLITIC. PRON. to the air free, eat-PRES:3P.S . se ao ar livre, dormese CLITIC. PRON. to the air free, sleep-PRES:3P.S CLITIC.PRON. ao ar livre to the air free ‘one lives in the open air, one eats in the open air, one sleeps in the open air’

The partition illustrated in (15-16) becomes more salient in contexts with a transitive verb and a plural NP. In these cases, the NP is accompanied by the plural inflection of the verb in third person, e.g. vendem-se casas (‘houses are sold). This reflects a situation in which a passive participant, often with the feature [-animate], becomes the clausal subject. Consequently, the notion of an active participant in the described event tends to diminish. From a prescriptive perspective of language and language use, Cunha and Cintra (1984: 309) comment on the use of se as a passive marker by saying that the verbal agreement with plural NPs is obligatory. Hence, the use of a verb with singular inflection in plural contexts is to be avoided (Cunha and Cintra 1984: 309). Bechara (1999) follows the analysis cited above. However, he also admits a certain evolution of the structure V + SE + NP and the use of a verb with singular inflection in plural contexts. This evolution is illustrated below: 17.

a.

Vendemse Sell-PRES: 3P.P CLITIC.PRON. (‘houses are sold’)

b.

Vendemse casas Sell-PRES: 3P.P CLITIC.PRON. houses (‘Someone is selling houses’)

c.

Vendese casas Sell-PRES: 1P.S CLITIC.PRON. houses (‘Someone is selling houses’)

casas houses

Thus, an example like (17c.) is the final path in a natural evolution in which a given speaker conceptualizes the clitic pronoun se as a symbol that designates an undetermined subject. The author concludes that: “the genuine literary language requires ‘vendem-se’ (…). But both structures are

Introduction

17

correct, and the singular case is not, as has been demonstrated, a modification of the plural case. They only represent two different stages in the evolution (Bechara 1999: 178). 6 In his analysis of impersonal structures with plural inflection on the verb, e.g. quebraram meu relógio (‘someone broke my watch’), Perini (1998) claims that the elaboration of three semantic rules increases our understanding about these structures. Further, he connects these rules to the clitic pronoun se and the elaboration of a fourth rule. The rules are given in Figure 1:3: Figure 1:3. Perini’s semantic rules. Rule 1. The indirect object is the patient. Rule 2. The attribute (with the preposition ‘with’) is the instrument. Rule 3. The subject is agent, instrument or patient (in this order). Rule 4. Introduce a semantic representation of an undetermined agent in phrases with a NP in third person without a subject (cf. Perini 1998: 266, 270).

On the basis of these rules, Perini analyses the variation between singular and plural verbal inflection in contexts with a plural NP. In cases without verbal agreement, the author says, it is indeed possible to make an impersonal reading of the pronoun se. First, Rule 1 creates a correspondence between the grammatical role indirect object and the semantic role patient. Thereafter, Rule 4 introduces the representation of an undetermined agent, embodied by the clitic se. However, Perini recognizes the difficulty of extending the analysis to contexts with verbal agreement in plural. That is, he does not consider se to represent an undetermined agent in cases where Rule 4 cannot be applied (cf. Perini 1998: 272). Ikeda (1977) applies a transformational perspective on se and claims that it is always generated in the base as a constituent of V(erb), having the function as a “logical reducer”. In grammatical contexts with intransitive verbs it reduces the notion of the direct object, and in contexts with transitive verbs the subject role is reduced. Accordingly, the distinction between the passive se and the impersonal se is inadequate because it does not capture the entire meaning of this pronoun and its function. Ikeda 6

”A genuína linguagem literária requere ’vendem-se’ (…). Mas ambas as sintaxes são corretas, e a primeira não é absolutamente, como fica demonstrado, modificação da segunda. São apenas dois estágios diferentes da evolução” (Bechara 1999: 178).

18

Chapter One

further recognizes that the analysis raises some problems regarding the interpretation of se as an undetermined agent (Ikeda 1977: 62). A similar perspective is taken by Milanez (1982). Nevertheless, a distinction is made between indeterminateness and indefiniteness. The first term is reserved for human participants that cannot be quantitatively restricted. Therefore, the use of the clitic pronoun se as an undetermined subject implies a participant with the feature [+H universal]. On the other hand, it is claimed that the structure NP + V (PLURAL) + SE does not include a human agent, and this claim certainly creates some confusion related to the analysis. Finally, Nunes (1990) maintains the division between the passive and the impersonal se. From the theoretical framework of Government and Binding, it is concluded that the pronoun se absorbs the internal argument role and receives accusative case in structures with plural marking on the verb. Subsequently, the plural NP receives nominative case. In structures without plural marking on the verb the situation is inverted (cf. Nunes 1990: 31-40).

Finite and Infinitive (Adverbial) Clauses The variation between finite and infinitive adverbial clauses is rather unexplored by traditional grammarians. Cunha and Cintra (1984) explain that the variation is possible and provide a list of its distribution. Bechara (1999) proceeds in the same manner, but comments that the infinitive may contribute to a more precise and elegant style (Bechara 1999: 514). On the other hand, Raposo (1975), analysing the broad-spectrum variation between finite and infinitive verb forms from a transformational perspective, analyses the infinitive as a complementizer inserted at the same syntactic level as the complementizer que. This process is denominated the complementizer insertion (‘inserção do complementador’). According to Raposo, there are two reasons to consider the infinitive to belong to the same category as the complementizer que. One is the inexistence of the structure que + infinitive. The other is based on the fact that there is a complementary distribution between the structures introduced by que and the structures introduced by the infinitive (cf. Raposo 1975: 22). From the same linguistic perspective, Perini (1977) formulates the insertion of que rule (‘inserção de que’). Essentially, the rule applies to subordinate non co-referential contexts and it specifies that the infinitive should be replaced by a finite verb form introduced by que in the transformation from the deep structure to the surface structure if the main and subordinate clauses have different subjects.

Introduction

19

Further, Perini comments on the [+/-tense] parameter. That is, while the feature [tense] is present in the deep structure, it may be omitted in the surface structure: 18. a. b.

Ele disse que He say-PRET: 3P.S that

estava be-IMP: 3P.S

Ele disse estar comendo He say-PRET: 3P.S be-INF. eat-PROG

comendo acarajés eat-PROG. beans

acarajés beans

‘He said (that) he was eating beans’

The reason why the infinitive clause is possible in cases like (18b.) is that it has the same time frame as the main clause. The rule that explains this condition is called tense suppression (‘supressão de tempo’) and it is formed in order to explain the variation between finite and infinitive clauses. Perini recognizes that the tense suppression rule is optional and therefore may not fully provide an explanation for this variation (cf. Perini 1977: 62, 63). Raposo (1987) builds on his earlier works and is more concerned with giving a structural explanation to why inflected infinitive clauses may occur in grammatical contexts that usually are restricted to finite clauses. The idea put forth is that the inflected infinitive is raised to complement position and therefore can be assigned case by an external case assigner: Crucially, in an important subset of inflected infinitival complements, Case is available for Agr in Infl only if Infl raises to Comp position (more accurately, to the head of CP) where a governor (and Case assigner) external to the embedded clause may successfully assign Case to it (Raposo 1987: 85).

The formal differences between finite and infinitive subordinate structures are most accurately described by the predicate lamentar (‘to regret’) and the following complement structure. These formal differences are illustrated below: 19. a. Nós lamentamos [CP que eles We regret-PRES: 1P.P [CP that they-MASC. recebam pouco dinheiro] little money] receive-PRES.SUBJ: 3P.P ‘We regret that they receive so little money’

20

Chapter One b.

Nós lamentamos [Nmax = IP eles We regret-PRES: 1P.P [Nmax = IP they-MASC. [I’ [I Agr] ter recebido pouco dinheiro]] [I’ [I Agr] have-INF. receive-PART. little money]] ‘We regret that they have received so little money’ (cf. Raposo 1987: 97, 101).

Thus, the finite subordinate structure (19a.) is a Complement Phrase. On the other hand, Raposo claims, the projective properties of X’-theory posit that the inflection of the infinitive is the head of the IP and therefore a maximal projection of [+N]. This is shown in example (19b.). Raposo offers a brief description of the properties of infinitive adverbial clauses, adjunct clauses in his terminology, in which he explains that nominative case is assigned to the inflected infinitive by the preposition. The assumption is that the category P(reposition) subcategorizes for maximal projections of N and therefore assigns nominative case to the inflected infinitive. The formal description given for infinitive adverbial clauses is [main clause [PP] [IP]]. Unfortunately, Raposo (1987) does not offer any formal description of finite adverbial clauses.

Synthesis The studies reviewed above contribute–in one way or another– to shedding light on the distribution and variation between the plain infinitive and the inflected infinitive, over the meaning of the clitic pronoun SE and over the variation between finite and infinitive clauses. Therefore, I will not take issue with Cunha and Cintra (1984) or to Bechara (1999) and the claim that the variation between the plain and the inflected infinitive is motivated by stylistic factors. Further, I fully sympathize with Maurer (1968) and Perini (1977) regarding their claim that a greater linear distance between the main verb and the infinitive may contribute to the use of the inflected infinitive in co-referential contexts. Finally, studies within the generative framework furnish a structural description of the variation between finite and infinitive clauses. However, I believe that the traditional way to handle these questions somehow misses the main point: what are the underlying reasons for choosing one or another subordinate structure? Likewise, should one accept the claim that se only represents an impersonal subject with singular NPs and in cases that do not show verbal agreement with the plural NP? Does the notion of an active subject of the described process totally disappear in cases with verbal plural agreement?

Introduction

21

The review offered above illustrates that the traditional way of analysing these questions tends to the formulation of more or less complex rules, often with focus on purely formal and structural matters. Obviously, the formal properties of the variation between finite and infinitive clauses constitute the main interest in the generative framework and semantic considerations are often neglected. This procedure is in tune with the theoretical framework: in the generative approach to language–a framework that posits an autonomous syntax–semantics is first and foremost a question related to the lexicon and not to different syntactic structures. Therefore, I will attempt to furnish an explanation to the questions exposed above from a cognitive perspective of language, and in particular from a Cognitive Grammar perspective. The model of Cognitive Grammar (Langacker 1987, 1988a-d, 1991, 2008) is a usage based model that is opposed to the traditional view of grammar as an autonomous and abstract system of rules without semantic meaning. Instead, the model posits a relation between grammar and semantics in which grammar is understood as being a linguistic convention. The linguistic units of grammar create a form-meaning relation that is based on conventional language use. Further, grammar is conceived of as an integral part of human cognition, and as such it may be explained by general cognitive abilities. The model of Cognitive Grammar will be studied in more detail in the following section.

The Model of Cognitive Grammar Cognitive Grammar is a forceful reaction to the generativist tradition and the claim that grammar is an autonomous and universal system of rules without semantic meaning. Langacker (1987) articulates three fundamental claims regarding language structure and meaning: 1. Semantic structure is not universal; it is language-specific to a considerable degree. Further, semantic structure is based on conventional imagery and is characterized relative to knowledge structures. 2. Grammar (or syntax) does not constitute an autonomous formal level or representation. Instead, grammar is symbolic in nature, consisting or the conventional symbolization of semantic structure. 3. There is no meaningful distinction between grammar and lexicon. Lexicon, morphology and syntax form a continuum of symbolic structures, which differ along various parameters but can be divided into separate structures only arbitrarily (Langacker 1987: 2-3).

These claims differ fundamentally from the generative approach to language. The first claim rejects the idea that semantic structure is

22

Chapter One

universal. In contrast, semantic structure may differ from one language to another. The reason why this happens is rather obvious: different languages have different way to express a given situation. Thus, the conventional imagery of a language is shaped by the way in which the speakers of that language conceptualize a given situation. Being so, there is a relation between knowledge structures and semantic structures. Further, grammar (or syntax) is not without meaning. Being related to semantic structures by conventional symbolization, grammar is meaningful by itself and thus serves to express semantic meaning. In fact, the third claim puts emphasis on the difficulty of drawing a sharp line between grammar and lexicon. Grammar, lexicon and morphology are symbolic structures with inherent meanings. Consequently, different grammatical structures may match different conceptualizations of a given situation. Also, different languages may differ in the way in which they conceive of a certain situation. Examples (20a.21b.) below illustrate these tendencies: 20. a)

b)

21. a)

b)

Mandeilhe uma CLITIC.PRON: 3P.S a Send-PRET: 1P.S ‘I sent him/her a letter’ Mandei uma Send-PRET: 1P.S a ‘I sent a letter to him’ Tenho Have-PRES: 1P.S ‘I have a hunger’

carta letter

para to

carta letter

ele him

fome [Portuguese] hunger

I am hungry [English]

Both (20a.) and (20b.) describe the situation of sending a letter to someone. Nonetheless, they show some structural differences. In (20a.), the two participants (lhe and carta) are juxtaposed, while the preposition para is inserted between them in (20b.). From a Cognitive Grammar perspective, then, these structural differences designate different ways of conceptualizing the situation. The juxtaposition and linear order of lhe and carta in (20a.) designates a possessive relation between the receiver and the letter. It focuses, so to speak, on the final result in which the letter has arrived with the receiver. The use of the preposition para in (20b.), on the other hand, renders more saliency to the path traversed by the letter. The following examples (21a.-21b.) both express the event of being hungry. They differ, however, relative to the verb used in order to express

Introduction

23

this experience. A native speaker of English uses the verb to be to describe the experience, while a Portuguese speaker uses the possessive verb ter (‘to have’). Cognitive Grammar claims that the different verbs evoke different conceptualizations regarding the notion of hunger (cf. Langacker 1987: 39, 47). In (21a.-b.) the conceptual differences are highly related to notions like association, on the one hand, and possession, on the other.

Grammatical Structure The basic structure of grammar is reduced to include three different structures: semantic, phonological and symbolic structures. The symbolic structure is bipolar and constitutes the relation between the semantic and the phonological structures. Grammar is first and foremost equated with linguistic convention that sanctions the grammatical structures used in any particular usage event. The usage event represents a conceptualization or a vocalization of a grammatical/semantic structure. The coding takes place between the sanctioning structure and the target structure. Figure 1:3 illustrates the basic structure of the model: Figure 1:3. Grammatical structure (Langacker 1987: 77) Semantic Space

Grammar (linguistic convention) Symbolic unit Semantic unit

Usage Event cod

sym

sym Phonological Space

Conceptualization

cod Phonological unit Sanctioning Structure

Vocalization

Target Structure

24

Chapter One

Grammar is understood as linguistic convention and the Sanctioning Structure. Within Grammar, symbolization takes place between a Semantic unit and its corresponding Phonological unit. The symbolization between the Semantic and the Phonological Space constitutes a Symbolic unit. This implies a direct and bipolar relation between semantic meaning and phonological representation. In the Target Structure (the Usage Event), the conceptualizer gives meaning to an event, a process or a sensation and by means of symbolization he/she may proceed to the vocalization of his/her conceptualization. The Symbolic Units in the Sanctioning Structure furnish the conceptualizer with the coding elements for the conceptualization, and for a vocalization of the same. Thus, a typical characteristic of the model of Cognitive Grammar is that it puts focus on the relation between grammar and semantics. First, the bipolar relation between the semantic and the phonological space implies that grammar–including syntax–is meaningful and does not represent a meaningless autonomous entity. In other words, grammar is the “structuring and symbolization of semantic content” (Langacker 1987: 12). Secondly, the claim that grammar is equated with linguistic convention, i.e., that the model includes language use, identifies the relation between language, language use and contextual uses of language. Meaning is not merely something in the air, but is created from our experience as human beings living in this world. Accordingly, the traditional claim that grammar (and syntax) constitutes an autonomous component is overtly rejected. This position is perhaps best illustrated by the difficulty to analyse the sentence “colorless green ideas sleep furiously”; a well-known sentence conjured in order to illustrate the autonomous nature of grammar (cf. Chomsky 1957: 15). The Cognitive Grammar analysis of grammatical organization considers the lack of coherent semantic content in this sentence to be problematic:

Introduction

25

Figure 1:4. Grammatical organization in Cognitive Grammar (Langacker 1988a: 45) ??? colorless green ideas sleep furiously

COLORLESS colorless

???

???

colorless green ideas

sleep furiously

GREEN

IDEAS

SLEEP

FURIOUSLY

green

ideas

sleep

furiously

The Cognitive Grammar analysis of grammatical organization divides relations into component structures and composite structures. The bipolar relation between the semantic pole and the phonological pole is represented with capital letters for the semantic pole and lower-case letters for the phonological pole. The component structures, at a lower level, constitute the linguistic units that form the composite structure at a higher level. Figure 1:4 indicates that the bipolar relation between the semantic pole and the phonological pole renders the meaning anomalous. In fact, it is only at the lowest component level that a correspondence is made between the semantic pole and the phonological pole. At the following level, both the structure colorless green ideas and sleep furiously are “ruled out” because of their lack of semantic meaning. The Cognitive Grammar analysis of the sentence also questions whether it provides evidence for an autonomous grammar. One could easily argue that the sentence merely demonstrates that languages have restrictions regarding the combination of different grammatical classes (e.g. N, V, ADJ, ADV, NP), and that human beings are capable of organizing these grammatical classes even when they fail to create a coherent semantic meaning (cf. Langacker 1988a: 43-44).

A Usage Based Model The usage based model of Cognitive Grammar posits a relation between linguistic convention (i.e., grammar) and extra linguistic knowledge that may be relevant in actual language use. That is, every act of creating or understanding the meaning of a linguistic expression is based on a combination of different general abilities. One fundamental assumption is

26

Chapter One

therefore that communication (or conceptualization) is a problem-solving activity in which the speaker/conceptualizer makes use of different kinds of knowledge: Figure 1:5. Resources in conceptualization/vocalization (cf. Langacker 1988a: 14).

symbolic units (grammar) general knowledge knowledge of context communicative objectives esthetical judgements etc. resources

coding problem-solving categorization

conceptualization vocalization

usage event

Thus, semantic meaning is not only related to symbolic units. Instead, there is a correlation between symbolic units and general human abilities, such as: general knowledge, knowledge of context, communicative objectives etc. Symbolic units are understood as resources used in the act of expressing meaning, but meaning is further construed on the basis of our general knowledge of the world and on the contextual situation. That is, the compositional value of a symbolic unit tends to be far more schematic and vague than the meaning the language user renders it in an actual usage event. This tendency is, for example, salient in the creation of novel metaphors, e.g. cabbage harvester as a metaphor for guillotine. It is also salient in creating and understanding new expressions. A term like pencil sharpener, for instance, is conventionalized to have the meaning of “a mechanical devise that sharpens pencils. However, before the process of conventionalization–through context–it could also mean “something or someone that sharpens pencils” (cf. Langacker 1988a:14-16).

Subordination in Cognitive Grammar From a traditional perspective of language, subordination has often been described as structural dependency. That is, while the main clause in prototypical cases is able to stand alone, the subordinate clause cannot occur in isolation. For example, the adverbial expression because it is raining cannot stand alone, but requires a main clause for the elaboration of a complete linguistic expression. Another view is that subordinate structures are embedded in the main clause structure. That is, an example

Introduction

27

like I came to rescue you is divided into the following structure: [I came [to rescue you]]. (cf. Cristofaro 2003). The cognitive approach to language emphasizes the relation between semantic structure and grammatical structure. Being so, the term subordination is equated with conceptual subordination. The relation between the main clause and the subordinate clause is therefore conceived of as an asymmetrical relation where the main clause elaborates the head of the expression. In a multi-clausal expression, the main clause is the focused part of the composite structure. The subordinate clause, on the other hand, can be defined as a clause whose profile is overridden by the main clause’s profile. In a coordinate structure, e.g. Mary came and John went home both processes are focused on. On the other hand, the subordinate structure does not usually receive the same degree of attention as does the main clause. A structure like I heard she cried designates the act of hearing, and not the one of crying (cf. Langacker 1991: 436). This tendency also holds for adverbial relations, e.g. she cried before I left. The expression designates first and foremost the act of crying and not the act of leaving. The examples discussed above may further illustrate some of the conceptual differences between complement structures and adverbial relations. The traditional assumption that subordinate clauses are dependent, i.e., that they cannot occur in isolation, may be difficult to sustain in relation to structures like I heard she cried. In fact, it is the main clause I heard that seems to be conceptually more dependent. Obviously, verbs like see, hear, know, think and other complement taking verbs may occur in isolation, but they do often evoke the question what the subject sees, hears or knows. This reflects an asymmetrical A(utonomy)/D(ependence) relation. In fact, the complement structure is in this case an autonomous substructure that elaborates the dependent head (the main clause). In adverbial relations, on the other hand, the main clause is the conceptually autonomous linguistic unit. It is the adverbial clause that shows some difficulties in standing alone. Thus, adverbial relations like she cried before I left show a totally different pattern regarding the A/D asymmetry. The adverbial clause is functioning as a modifier in its relation to the main clause and is, thus, more dependent on the main clause. The difference between complement and adverbial clauses in terms of A/D asymmetry is shown below:

28

Chapter One

Figure 1:6. A/D alignment for complements and modifiers (cf. Langacker 2008: 203). A) Complement construction Head

Complement

e-site

profile

Dependent

Autonomous

B) Modifier construction Head

profile Autonomous

Modifier

e-site Dependent

To start with, the head is the most prominent linguistic unit of the relational predication. This is indicated by bold lines in Figure 1:6. Recall that expressions like I heard she cried and she cried before I left focus on the processes described by the main clause, and not on the subordinate processes. On the other hand, the complement and modifier constructions differ regarding notions like autonomy and dependence. The head of the complement construction is an elaboration site (e-site) for the full linguistic expression, because complement taking verbs are often conceptually dependent on a complement for the creation of an elaborated expression. With modifiers, however, the situation is reversed. The head tends to be conceptually autonomous, and the modifier is conceptually dependent. That is, the modifier is the e-site for the full linguistic expression. A basic conceptual difference between complements and modifiers, then, is that the complement is a salient substructure that elaborates the head, while the modifier is a salient substructure that gets elaborated by the head. The modifying character of adverbial clauses may further explain the tendency to conceptualize the relation between the main and the adverbial clause in terms of figure/ground organization (cf. Talmy 1978, 2000). The figure/ground distinction is generally made on perceptual grounds and is said to reflect the human tendency to focus attention on one or another entity in a given scene. If a person, for instance, is entering a restaurant, it is likely that this situation is coded by the main clause He/She is entering the restaurant. Further, if I want to explain the reason why he/she enters the restaurant, this is generally done by an adverbial expression like because he/she is hungry. This coding reflects the figure ground organization in adverbial expressions. The main clause (the figure) tends to receive primary attention, while the adverbial clause (the ground) provides additional information about the main clause and/or its participants.

Introduction

29

To sum up, Cognitive Grammar declines to handle grammar as an autonomous component without semantic meaning. In contrast, it recognizes the difficulty to draw a sharp line between morphology, lexicon and syntax. These constructs are conceived of as symbolic units conveying semantic content. This relation is made explicit by the claim that there are only three different structures: semantic structure, phonological structure, and the symbolization between the structures. Finally, Cognitive Grammar highlights the relation between linguistic and extra-linguistic competence in the creation of meaning. The usage based model of Cognitive Grammar sees conceptualization as a problem-solving activity where the language user not only makes use of symbolic units, but also of his/her general knowledge. In the following section, the encyclopedic view of language and meaning will be examined more in detail.

Semantic Structure Encyclopedic and Contextual semantics Cognitive Grammar contends the idea that there is a relation between grammar and semantics. This idea is further emphasized in the claim that there is a symbolic relation between the semantic and phonological poles. As to the claim that our understanding of linguistic expressions is encyclopedic, Langacker (1987) discusses the concept [BANANA] and its location in a number of domains. First, the concept evokes a spatial/visual domain that characterizes its shape, a color domain and a taste domain. But it also evokes other more abstract domains that are related to our encyclopedic knowledge of bananas: that they grow in trees, that they come from tropical areas and so on. These different domains contribute to the understanding of the concept [BANANA]. As a result, it is problematic to draw a precise line between linguistic and extra linguistic knowledge in determining the semantic value of the concept (cf. Langacker 1987: 154). In the same manner, contextual knowledge of a situation is essential in order to comprehend the meaning of a linguistic expression. For example, the expression the cat is on the mat has a number of readings that differ according to the context in which it is expressed. In the prototypical case, it designates the event of a cat sitting or lying on a mat that is spread out on the floor. Nonetheless, there are various other possible readings. One is that there is an exhibition fight going on between a wrestler and a tiger. In this case, the expression may be uttered to point out that the wrestler has managed to defeat the cat. Another is that a person working in a mat factory has finished the work of decorating a mat with a picture of a cat. These are

30

Chapter One

only some of the possible meanings of the expression the cat is on the mat. The main point is that it has different meanings in different contexts (cf. Langacker 1987: 155). Another basic claim of Cognitive Grammar is that there is a relation between cognitive processes and the way in which we conceptualize linguistic meaning. That is, the ability to compare, contrast and structure different situations, or events, has a direct impact on how we choose to describe the situation. One example of this was shown in (20a-b) where the event of sending a letter to someone can be conceptualized in different ways and focus either on the final result or on the path traversed by the letter.

Profile and Base Another, and perhaps a more fundamental illustration of the human capacity for comparing, contrasting and structuring things and/or processes in the surrounding world, is the human capacity of structuring linguistic expressions into different domains at different levels. The underlying idea is that semantic meaning is not in created in isolation. Rather, a concept presupposes other concepts for its characterization. For example, any linguistic expression evokes certain domains as its conceptual base and the linguistic expression is the profile of these domains. Thus, the linguistic expression is said to designate a prominent substructure in a more basic domain, and this basic domain is indispensable for the understanding of the linguistic expression. For example, an expression like flat evokes the notion of a block of flats; and the expression finger evokes the notion of a hand. Langacker (1987: 184) illustrates the relation between profile and base with the concepts arc and circle: Figure 1:7. Profile and Base of the concepts ARC and CIRCLE.

circle ARC

space CIRCLE

Introduction

31

The cognitive plausibility for the relation between profile and base is, thus, that the concept [ARC] presupposes the conception of a circle. Being so, the concept [CIRCLE] serves as the base and an abstract domain for the higher order concept [ARC]. On the other hand, the concept [CIRCLE] also needs to be characterized in relation to some domain. This domain is space, since we can hardly conceptualize anything without some representational space that creates the potential for spatial relationships. Space is therefore considered to be a primitive representational field and a basic domain (cf. Langacker 1987: 148). The asymmetric relation between profile and base and between abstract and basic domains is perhaps best illustrated by the concept [BODY] and its relation to other body parts. The concept is a configuration in space and has the basic domain [SPACE] as its base. Further, it serves as an abstract domain, and the base, for the concept [ARM]. This concept, then again, is indispensable for the understanding of the concept [HAND] and constitutes the abstract domain for this concept. Further, [HAND] is the abstract domain for the concept [FINGER] which is the domain for [KNUCKLE] and [NAIL] (cf. Langacker 1993: 9).

Trajector and Landmark In Cognitive Grammar, the relation between the trajector and the landmark is a linguistic manifestation of the figure/ground organization. The relation reflects the subject/object distinction (but not only, cf. Chapter 2) and is often made on perceptual grounds. Thus, there is a tendency for the trajector to be equated with the participant that is moving – or that is conceptually moveable – in the scene. The landmark, on the other hand, is often a reference point or provides the background scene for the trajector. Also, the trajector is prototypically a well delimited and discrete participant occupying the initial position of the linguistic expression. This tendency is illustrated in the following examples: 22. a) The man [Tr] bought a house [Lm] in Estoril b) ? The Man [Lm] bought a house [Tr] in Estoril c) Mike [Tr] is walking down the street [Lm] d) The car [Tr] is in front of the house [Lm] e) ? The house [Tr] is in front of the car [Lm]

32

Chapter One

The difference between (22a) and (22b) is rather obvious and reflects the tendency to choose a well delimited and discrete entity as the trajector of the event. In fact, regarding trajector/landmark alignment, the peculiarity of (22b) rests on the fact that: (1) the least delimited participant has trajector status, (2) this participant does not occupy the initial position and (3) it is the passive participant in the described event. Example (22c) reflects the tendency to confer trajector status to the moving participant in a given scene. That is, if we see someone or something moving it is likely that this participant will attract our attention more than the background in which he/she/it is moving. Finally, the difference between (22d) and (22e) is that the car is the participant that is conceptually moveable, while the house is a more stable participant. Therefore, to confer trajector status to the house in (22e) creates a somewhat odd situation in which the firm object is located in relation to the object that is moveable. In other words, the house is a reference point for the car. As Silva (1997: 81) puts it, there is a strong tendency to conceptualize the most stable element as the landmark of an event. This remark does not mean that there is no flexibility in conferring trajector and landmark status to different participants in a given scene. An example is found in active and passive sentences. The difference between the cat chased the rat and the rat was chased by the cat, for example, illustrates the idea that the language user is able to conceptualize an event in alternate ways. In Cognitive Grammar, this ability is called imagery and it reflects the basic claim that grammar is meaningful. The active and the passive sentences above are not considered to be synonymous. In a perceptual process, the conceptualizer may move focus and concentrate either on the cat or on the rat. The shift of trajector and landmark status has the effect of putting primary focus on one or another participant in the event. Thus, Cognitive Grammar claims that semantic structure is equated with conventional imagery (cf. Langacker 1987: 111). In conclusion, Langacker’s (1987; 1988: a-d; 1991; 2008) model of Cognitive Grammar highlights the relation between human cognitive processes and language structure. Therefore, Cognitive Grammar is also at variance with the idea that it is possible to deal with semantic meaning without making reference to encyclopedic or contextual knowledge. In contrast, it posits a strong relation between the conceptualizer’s internal capacities and the surrounding world. The ability to contrast, compare and structure things and processes in the outside world has a direct effect on semantic structures. The relation between semantics and grammar is explicitly spelled out by the claim that there is a bipolar relation between

Introduction

33

the semantic pole and the phonological pole: grammar is the linguistic manifestation of semantic structures. Further, the encyclopedic view of semantics entails that every linguistic expression evokes various domains. These domains are indispensable for a full understanding of the linguistic expression. The linguistic expression– the profile–is a prominent substructure in some domain. The domain, on the other hand, functions as the base for the semantic meaning of the linguistic expression. Another asymmetrical relation is identified between the two most prominent participants in a relational predication. The trajector, often equated with the clausal subject, receives focal prominence in its relation to the landmark. The latter is often a reference point for the trajector. The relation between the trajector and the landmark is considered to be a linguistic manifestation of the perceptual relation between figure and ground. Hence, the model of Cognitive Grammar emphasizes the relation between perception and language structure. In the following section, I will relate the model of Cognitive Grammar to the hypothesis put forward regarding different adverbial structures in European Portuguese.

Working Hypotheses The working hypotheses of the present study are based on the following claims: (1) language use is a problem-solving activity, (2) grammar–or phonological structure–is essentially a means for expressing semantic meaning, and (3) semantic meaning is encyclopedic and contextual. Furthermore, concepts like trajector and landmark, prominence, domain and imagery are essential in order to attain a greater understanding about different adverbial structures in European Portuguese. Thus, in the search for an explanation to the variation between adverbial clauses with the plain or the inflected infinitive in co-referential contexts, I will focus on the fact that they may or not present an elaborated trajector. Instead of postulating deep and surface structures, I will examine the actual phonological representation of the adverbial structures and the use of a more or less prominent adverbial clause–that is, with or without phonological markings of a trajector. Additionally, the use of an elaborated trajector will be understood of as a kind of problem-solving activity. Thus, I believe that the use of an elaborated trajector may be motivated by contextual factors that may create a cognitive need to highlight the trajector of the adverbial clause, even when it is co-referential with the main clause trajector. The working hypothesis, then, is that the elaborated trajector in co-referential contexts is a strategy used to put focus on the trajector when

34

Chapter One

the notion of this participant is diminishing in the context. In sum, the working hypothesis underscores the idea that the inflected infinitive is a resource instead of formulating abstract rules about it (cf. Maurer 1968; Perini 1977: Raposo 1975, 1987). Turning to the use of the clitic pronoun se as an undetermined subject, I will consider this participant a generic trajector. That is, the use of constructions with se that agree or do not agree with the (subject) NP in plural contexts reflects conventional imagery and our capacity to formulate expressions in accordance with the way in which we conceptualize them. Consequently, the issue at stake is that there is no definitive line between the passive and the impersonal reading of this pronoun. Rather, there are contextual features that will contribute to one or another reading of se and its semantic meaning. Among these features, I will analyse the structural position of the linguistic units se and NP, the kind of action in the described event, the kind of (subject) NP and the existence of an obvious referent to se. Thus, I believe there to be more features than the verbal agreement to consider in finding the function and semantic value of this pronoun. Finally, the variation between finite and infinitive adverbial clauses is an issue that remains rather neglected within Portuguese linguistics. On the one hand, the traditional grammars do not study the motivation for this variation, but mention the possibility to use one or another verb form. On the other hand, the generativist approach to the issue is to formulate the syntactic operation behind the variation. That is, the semantic aspects of the variation are strongly neglected. This procedure is clearly in contrast with the theoretical perspective taken in the present study. Consequently, I will argue that the structural differences between finite and infinitive adverbial clauses entail different meanings. Further, I will show that the semantic differences are based on the morphological differences between the finite and the infinitive verb forms. That is, the finite verb forms are conceptually more complex than the infinitive. Together with a greater formal difference between the main verb and the subordinate finite verb, this leads to what I will refer to as iconic linguistic principles–formal complexity corresponds to conceptual complexity, and formal distance corresponds to conceptual distance (cf. Haiman 1980, 1985). The working hypothesis that follows is that the finite adverbial clause is conceptually more complex than the infinitive adverbial clauses. Essentially, the higher degree of complexity in the finite adverbial clause contributes to a dislocation from one domain of origin to other domains, while the infinitive adverbial clause is more restricted in this sense and stays in the domain of origin. I will identify this phenomenon as a

Introduction

35

higher degree of subjectification in the finite clauses than in the infinitive ones (cf. Langacker 1990, 1999, 2003, 2006). The hypothesis underlying the research questions and hypothesis mentioned above is that a cognitive approach to Portuguese adverbial clauses may provide a natural and more plausible explanation to issues that–by tradition–have received “formal treatment”. That is, instead of formulating rules with a higher or lesser degree of formal complexity, or postulating deep and surface structures in the attempt to deal with the problem, the cognitive approach to adverbial clauses may find a natural and plausible explanation to the semantic and conceptual differences between different adverbial structures in Portuguese. That is the main hypothesis of the present analysis.

Material and Method The adverbial clauses that allow a variation between finite and infinitive verb forms express the semantic meaning of cause, temporality, manner, concession and purpose in their relation to the main clause. The infinitive adverbial clauses are commonly introduced by a preposition, while the finite clauses are introduced by a conjunction. The formal differences between the linguistic units introducing the adverbial clauses are given in Figure 1:7 below: Figure 1:7. Linguistic units introducing finite and infinitive adverbial clauses. Preposition (+ infinitive)

Conjunction (+ finite verb)

A) CAUSE

por (‘through’)

porque (‘because’)

B) TEMPORALITY

antes de (‘before’) depois de (‘after’) até (‘until’) ao (‘when’)

antes que (‘before’) depois que (‘after’) até que (‘until’) quando (‘when’)

C) CONCESSION

apesar de (‘even if’)

embora (‘even if’)

D) MANNER

sem (‘without’)

sem que (‘without’)

E) PURPOSE

a fim de (‘in order to’) para (‘to’)

a fim de que (‘in order to’) para (‘to’)

36

Chapter One

Figure 1:7 illustrates the parallelism between the linguistic units introducing the finite and infinitive adverbial clauses. In most cases, there is a formal similarity between them. The difference is that the preposition is followed by the conjunction que (‘that’) in the finite adverbial clauses. Two structures diverge from the general pattern: ao/quando and apesar de/embora. The motive for including these types in the analysis is that they frequently are conceived of as synonyms. Concerning the causal adverbial clauses, it could further be argued that clauses introduced by visto/visto que (‘seen/seen that’) and dado/dado que (‘given/given that’) ought to be included in the study. The reason for not including them is that they do not express a causal relation between the main and the adverbial clauses, but rather an explicative relation (cf. Vogt 1976-77: 144). The material that will be used in the analysis is extracted from four different corpora of spoken and written Portuguese: • • • •

Português Fundamental [Pfu] Português Falado [Pfa] LINGUATECA (DiaClav) [indication of newspaper] JORNAL-ELAN [Diário de Notícias]

Português Fundamental and Português Falado are corpora of spoken European Portuguese, while LINGUATECA and JORNAL-ELAN consist of written European Portuguese extracted from news papers. Hence, the corpora furnish representative and natural examples from both written and spoken European Portuguese. Further, they cover a reasonably extensive period of spoken and written language use. The Português Fundamental corpus presents linguistic material from the period 1970-1974. The Português Falado corpus recollects material from the decades 1970-1990. On the other hand, the two corpora of written newspaper prose are more recent. The JORNAL-ELAN corpus dates from 1996-1997, and LINGUATECA is from 1999-2000 (cf. Nascimento 1987; Casteleiro and Nascimento 2001; JORNAL-ELAN: E2>E3. The succession of events changes their degrees of activation in the active memory of the speaker. If 11

At this point, it should be clarified that the studies on referential expressions cited in this section are not conducted within the framework of Cognitive Grammar. However, they focus on the cognitive aspect of referential expression.

The Degree of Prominence in Infinitive Adverbial Clauses

63

he/she is commenting upon E1, the event (and the referents) is in his active memory. The limited human capacity to store information in the active memory, however, entails that E1 disappears from the active memory when new information enters (E2). Obviously, E2 will disappear from the active memory when E3 is the focus of attention. This being so, the use of one or another referential expression in discourse depends on the status of the referent in the speaker’s active memory. The results of Tomlin and Pu (1991) are corroborated by data furnished by Cowan (1988). Test results indicate that the human capacity to store information and to give it focal attention is reduced to three units or 1.5-2.0 seconds (cf. Cowan 1988: 166; Tomlin and Pu 1991: 161). Chafe (1987, 1994) is in accordance with Tomlin and Pu (1991) and the claim that referential expressions are related to cognitive processes. The author makes a distinction between three cognitive statuses of the referent: active, semi-active and inactive. The cognitive status of the referent is related to the time parameter and the activation cost. The referent that is active in the moment when the referential expression is uttered (t2) is active, semi-active or inactive in a preceding time (t1). The referent that is active in (t1) is considered given information in (t2). The semi-active referent in (t1) represents accessible information in (t2), while the inactive referent in (t1) is new information in (t2). The difference is illustrated in Figure 2:10: Figure 2:10. The cognitive status of the referent (cf. Chafe 1994: 73).

(T1) Active Semiactive Inactive

(T2) Given

Active

Accessible New

In terms of cognitive processing, then, the active referent in (t1) requires less effort to be evoked in (t2) than a semi-active or inactive referent. This pattern is further related to different referential expressions. Being given information, the active referent may be evoked with a less prominent

64

Chapter Two

referential expression. On the other hand, semi-active and inactive referents are accessible and new information respectively. Therefore, they require a more prominent referential expression in order to be evoked. That is, the active referent constitutes the element that has the focal attention of the speaker and the hearer, which entails a lower cognitive cost and a less prominent referential expression. The inactive referent is not present in the short-term memory and requires the most prominent referential expression. Finally, the semi-active referent is a peripheral referent that has lost focal attention in the context because of a deactivation (cf. Chafe 1987: 25-36). Ariel (1991, 1994, 1996) analyses referential expressions in relation to the term accessibility. The more accessible the referent, the less prominent is the referential expression. Another feature is that terms like co-referential and non co-referential are secondary. The main issue is the cognitive process of the speaker when he/she evokes the referent and not the question whether the referent is co-referential or not. Ariel’s (1996) argument is as follows: …the primary function of the various referring expressions, I claim, is to mark different degrees of accessibility in memory. Thus, in her choice of referring expression, the speaker does not indicate to the addressee whether a particular expression is to be interpreted co-referentially or disjointly. Rather, she signals how accessible to him the mental entity represented by the referring expression is (according to her best judgment) (Ariel 1996: 15).

Obviously, there is a tendency to use a more prominent referential expression in non co-referential contexts, but this is not the main issue. Coreferential cases may occur with prominent referential expressions and non co-referential referential expressions may have a higher or a lower degree of prominence. According to Ariel (1991, 1994, 1996) the use of referential expressions reflects a more basic cognitive process–the one of accessibility. The examples displayed below illustrate this basic cognitive process: 63. O meu ritmo cardíaco acelerava para a casa the place The my rhythm heart accelerate-IMP: 3P.S to dos cento e setenta, parado. recorri of the one hundred and seventy stand-PART. turn to-PRET: 1P.S várias vezes à urgência do Hospital da Covilhã, many times to the emergency of the hospital of the Covilhã, onde me administravam uma injecção intravenosa, injection intravenous, where me administer-IMP: 3P.P an para o coração voltar ao normal to the heart return-INF. to the normal [Pfa: Um Mal Desconhecido]

The Degree of Prominence in Infinitive Adverbial Clauses

65

‘My pulse accelerated to one hundred and seventy at rest. I often visited the hospital in Covilhã where they gave me intravenous injections to get my pulse back to normal’ 64. eu, por exemplo, gostei muito de Paris por ver I for example, like-PRET: 1P.S much of Paris for see-INF. muitos portugueses, por encontrar portugueses many Portuguese people, for meet-INF. Portuguese people [Pfa: As Grandes Cidades] ‘I, for example, really liked Paris because I saw and met many Portuguese people’

From a cognitive perspective of referential expressions, it is possible to explain the reason why the speaker uses a referential expression that is more prominent in (63) than in example (64)–there is a difference in accessibility. The antecedent o meu ritmo cardíaco in (63) is distanced from its referent (o coração) and has therefore a low degree of accessibility. Further, the antecedent gives evidence of a rather vague relation to the referent, a relation that is articulated metonymically. On the other hand, the referential relation in (64) is of a more direct character. The referent is accessible and the referential expression is therefore less prominent. In order to systematically analyse the various kinds of referential expressions, Ariel (1991) establishes a hierarchy of accessibility markers. The most accessible referent is marked with zero (null subject, gap), and the referent with the lowest degree of accessibility is marked with full name + modifier. The hierarchy of accessible markers is illustrated below:

66

Chapter Two

Figure 2:11. The hierarchy of accessible markers (Ariel 1991: 449). LOW ACCESSIBILITY Full name + Modifier Full name Long definite description Short definite description Last name First name Distal demonstrative (+Modifier) Proximal demonstrative (+Modifier) Stressed pronoun + Gesture Stressed pronouns Unstressed pronouns Zero HIGH ACCESSIBILITY

Although the hierarchy of accessible markers is established on the basis of English and Hebrew, it captures some rather general tendencies – the prominence of the referential expression increases as the degree of accessibility descends. Furthermore, the hierarchy is established on natural language use, and that the author emphasizes the differences that may exist from one language to another. 12 Ariel (1991: 445; 1994: 28; 1996: 22) suggests that the formal distance between the antecedent and the referent, the prominence of the antecedent and competition between potential antecedents are important features in the choice of a referential expression. In sum, the cognitive approach to referential expressions focuses on the level of attention of the referent in the speaker’s mental representation and on the degree of accessibility of the referent. This being so, the reanalysis of examples (58-59) may provide a plausible explanation to “marked” referential expressions in co-referential contexts. Below, these examples (58-59) are renumbered to (65-66):

12

Portuguese displays a fundamental different pattern regarding the pronouns and the use of demonstratives. Further, verbal inflection is often used referentially.

The Degree of Prominence in Infinitive Adverbial Clauses

67

65. o

linho era, com rapazes outras vezes e as with boys other times and the the flax be-IMP: 3P.S raparigas, era todo amassado, debulhado, todo, all smash-PART. thresh-PART., all girls, be-IMP: 3P.S am[…], para ele amassar, para ele tombar, para ele he fall-INF. to he sm[…], to he smash-INF. to ficar tombado be-INF. fall-PART. [Pfa: O Linho] ‘The boys and girls took turns at crushing and threshing the flax after which they broke it up and let it fall to the ground’

66. os conjuntos que existem cá todos eles here all them the groups that exist-PRES: 3P.P tentam estar sempre, eh, em cima do, do be-INF. always, eh, on over the, the try-PRES: 3P.P top e aprenderem os primeiros classificados, first classify-PART top and learn-INF: 3P.P the aprenderem da melhor maneira para agrad[…], way to satis[…] learn-INF: 3P.P of the best para agradarem ao, aos jovens to the, to the youth to satisfy-INF: 3P.P [Pfa: Um Meio Pequeno] ‘The groups around here always try to be up to date and try to learn the biggest hits to satisfy the young audience’

In the examples above, the antecedent is not in the active (short-term) memory of the speaker. In (65), the first anaphoric expression (ele) is preceded by three short pauses. In the transcription the pauses are marked by commas. Further, there is a partial repetition preceding the anaphoric expression. This is represented by the sequence am[…], in which the parenthesis represents an expression that is not clearly heard. Finally, the second adverbial clause (para ele tombar) is preceded by an incomprehensible vocalization and the third adverbial clause is preceded by a long pause (.). In example (66), the partial sequence para agrad[…] and a short pause (,) precede the anaphoric expression (the verbal inflection). Consequently, the lack of cohesion between the antecedent and the coreferent may explain the use of a more prominent referential expression in the adverbial clause. Therefore, I sympathize with the cognitive approach and the hypothesis that the choice between different referential expressions is, first and foremost, related to the speaker’s cognitive processes. The distinction between co-reference and non co-reference is in this sense explained by the level of attention of the referent and the accessibility of the referent. The

68

Chapter Two

tendency to use a more prominent referential expression in cases of non coreference may, thus, be explained by cognitive processes. These cognitive processes are also able to furnish an explanation for the use of prominent referential expressions in co-referential contexts. This issue will be further analysed in the following sections.

Spoken Language The claim made in this section is that the use of one or another type of adverbial clause is strongly dependent on the cognitive need to focus on the trajector of the adverbial process. The underlying factors that create this need are divided into two different features of language and language use. The first is related to the structural properties of a linguistic expression, that is, to the form of the linguistic expression. The second is related to the usage event and to the actual planning of discourse. Obviously, these two factors are related. Cognitive Grammar postulates a relation between grammar and semantics and further recognizes the difficulty of separating semantics from pragmatics (cf. Langacker 1987: 154). Thus, Cognitive Grammar does not separate grammar from discourse and the planning of discourse. The structural properties that create a weaker link between the trajector of the main clause and the co-referent in the adverbial clause are found in the linguistic context. The increased number of linguistic units between the antecedent and the co-referent creates a structural distance between them. Moreover, the existence of different potential antecedents to the adverbial trajector, expressed by the number of NPs in the linguistic context, creates competition. Finally, the syntactic position of the main and the adverbial clauses, expressed by anaphoric and cataphoric referential relations, may have a certain impact on the choice of adverbial clause. The usage event factors are related to the speaker’s planning of the ongoing discourse. In actual discourse, the speaker often feels a need to pause, repeat or reformulate the linguistic expression in order to find the linguistic expression that matches his/her conceptualization. These pauses, repetitions and reformulations create a weaker degree of cohesion between the trajector of the main clause and the co-referent. A classification of the structural and the usage event factors is offered in Figure 2:12:

The Degree of Prominence in Infinitive Adverbial Clauses

69

Figure 2:12. Structural properties and Usage Event properties. Structural properties Distance: The number of linguistic units between the main clause trajector and the adverbial co-referent. Competition: The number of NPs in the linguistic context. Syntactic position: Anaphoric or cataphoric referential relation. Usage Event properties: Cohesion: pauses, repetitions and reformulations.

The structural properties are more concrete than the usage event factors, being present in the linguistic context. The increased number of linguistic units between the main clause trajector and the co-referent, the existence of various NPs in the linguistic contexts and the internal syntactic position of the main and the adverbial clauses may create a cognitive need to focus on the adverbial trajector. The usage event factors, on the other hand, are more abstract. The planning of discourse entails a cognitive process in which the speaker seeks the accurate expression for his/her conceptualization. The manifestation of this cognitive process is noticed in pauses, repetitions and reformulations. The following examples illustrate the structural properties that may explain the use of a more prominent adverbial clause: 67. eu, eu eu, estive, eu estive na Bélgica I, I, I, be-PRET: 1P.S I be-PRET: 1P.S in the Belgium há, em oitenta e seis, oitenta e sete, com eighty and six, eighty and seven, with be-PRES: 3P.S in dezoito, dez[…], dezoito, dezanove anos, quando eighteen, ten[…], eighteen, nineteen years, when vim para cá odiava isto. porque here hate-IMP: 1P.S this because come-PRET: 1P.S to não tinha nada, não havia nada nothing not have-IMP: 3P.S nothing, not be-IMP: 3P.S – apesar de eu na Bélgica estar numa aldeiazinha, village-DIM., – even of I in the Belgium be-INF. in a coitadinha, de três mil habitantes poor-DIM., of three thousand inhabitants [Pfa: Um Meio Pequeno] ‘I was in Belgium in 1986 and 1987, when I was between eighteen and nineteen years old. When I came here, I hated it, because nothing ever happened–even though I had lived in a small, poor village with three thousand inhabitants in Belgium’

70

Chapter Two

68. era o Francisco Pires, que veio da be-IMP: 3P.S the Francisco Pires, that come-PRET: 3P.S from the Benfica. foi um grande atleta, um jogador great athlete, a player Benfica. be-PRET: 3P.S a famoso. e há um suborno aqui na here in the famous. and be-PRES: 3P.S a bribery Famalicão, que é com o Oriental, que Oriental, that Famalicão, that be-PRES: 3P.S with the no jogo que vem aqui fazer (…) in the game that come-PRES: 3P.S here make-INF. (…) foram ter com ele e deramhave-INF. with him and give-PRET: 3P.P go-PRET: 3P.P lhe cinco mil escudos para ele se CLITIC.PRON. him five thousand escudos to him vender sell-INF. [Pfa: Desporto e Dinheiro] ‘It was Francisco Pires, who came from Benfica. He was a great athlete, a famous player. And there has been bribery here in Famalicão involving Oriental, that will play here (…) they went to see him and gave him five thousand escudos in bribes’ 69. depois do linho estar ripado, o linho era the flax be-IMP: 3P.S after of the flax be-INF. ripe-PART., estendido em currais de reses, que havia be-IMP: 3P.S stretch-PART. in corral-PLUR. of cattle-PLUR., that muito curral de reses much corral of cattle [Pfa: O linho] ‘After the ripening of the flax, it was stretched in corrals, there were many corrals’

The use of a prominent adverbial clause in the examples displayed above is not accidental. Rather, it finds an accurate explanation in the structural properties of the linguistic expression and in the increased distance between the main clause trajector and the adverbial trajector. In (67), there are two causal clauses inserted between the main clause and the adverbial clause, e.g., porque não tinha nada, não havia nada. In the following example (68), the linguistic context offers various potential antecedents to the trajector of the adverbial clause: foram, deram, ele and lhe. And, finally, the cataphoric relation in (69) may explain the emergence of a prominent adverbial clause. Thus, the common denominator for examples (67-69) is that the need to focus on the adverbial trajector has a structural explanation.

The Degree of Prominence in Infinitive Adverbial Clauses

71

Nonetheless, the structural property of examples (67-69) combines with factors that are related to the actual usage event and the planning of discourse. In (67), for example, a number of pauses between the main clause and the adverbial clause cause a weaker cohesion between the two clauses. The first pause appears between the main clause (odiava isto) and the first causal clause (porque não tinha nada). Moreover, there is another short pause between the causal clauses. The weakened cohesion is also present in the following examples: 70. e ele dali é que se punha put-IMP: 3P.S and he from there be-PRES: 3P.S that CLITIC.PRON em pé outra vez. chegava à fim de muito to the end of much in foot another turn. arrive-IMP: 3P.S tempo ele iase pondo em pé. CLITIC.PRON put-PROG. in foot time he go-IMP: 3P.S até ele espigar e dar semente outra vez and give-INF. seed another turn until he fully grow-INF. [Pfa: O linho] ‘And from there the flax flowered once again. After a long time, it successively flowered once again until giving seed once again’ 71. pois. mas eu até penso que sim, que eles yes, that they well. but I even think-PRES: 1P.S eram pessoas para, para irem assim também to, to go-INF: 3P.P thus also be-IMP: 3P.P people à missa to the mass [Pfa: Na Base Militar de Beja] ‘Well, I even believe that they were people who used to go to church’

In these cases, it is difficult to find a structural explanation for the use of a more prominent adverbial clause. The formal relation between the main clause and the adverbial clause is rather strong. In (70), there is a direct relation between the main clause (ele ia pondo-se em pé) and the adverbial clause (até ele espigar). In example (71), the adverbial conjunction (para) occurs twice, but no other linguistic unit is inserted between the clauses. On the other hand, examples (70-71) display features related to the usage event and the planning of discourse. The long pause in (70), represented by (.), the short pause (,) and the repetition in example (71) indicate that the speaker is searching for the accurate linguistic expression for the conceptualization. Further, the process of matching a conceptualization with an accurate linguistic expression creates a weakened cohesion between the main clause and the adverbial clause. It also creates a lesser degree of cohesion between the main clause trajector and the co-referent in the adverbial clause. This

72

Chapter Two

being so, the use of an elaborated trajector in the adverbial clause is not difficult to explain. The structural and usage event factors correlate in the process of making the trajector of the main clause less accessible in the active short-term memory of the speaker. Therefore, the adverbial clause requires a more prominent referential expression. The correlation between the structural properties and the usage event properties is illustrated in Figure 2:13: Figure 2:13. The correlation between structural properties and usage event properties. Main clause Structural properties: Structural distance Number of NPs Formal position

Adverbial clause

Main clause

Usage Event properties: Pauses Repetitions Reformulations

Figure 2:13 illustrates some of the structural and usage event properties that may explain the reason why an elaborated trajector emerges in adverbial clauses. The structural properties distance and competition precede the adverbial clause and create a cognitive need to emphasize the participant of the adverbial process. That is, the increased structural distance between the main clause trajector and the adverbial clause creates a need to reintroduce the trajector in the adverbial clause. Moreover, the increased number of potential antecedents in the linguistic contexts creates the same need. In order to avoid trajector confusion, the speaker chooses to express the trajector of the adverbial process overtly. The inversion of syntactic position between the main and the adverbial clauses reflects the same tendency. In cases where the adverbial clause precedes the main clause, it is likely that the focal participant is not in the active short-term memory of the speaker, especially if this participant is introduced in the linguistic context as a new referent. Evidentially, the planning of the ongoing discourse co-occurs with the actual discourse. Therefore, it is highly expected that the speaker will need to pause, repeat or reformulate him/herself during this process. These

The Degree of Prominence in Infinitive Adverbial Clauses

73

interruptions create a weaker cohesion between the main clause and the adverbial clause and, as a result, the speaker may need to put emphasis on the adverbial trajector even though it is the same trajector as the main clause trajector. Put another way, the adverbial process needs a referential reactivation. As indicated above, the structural properties and the usage event properties may co-occur. There are, however, hierarchical differences. The structural properties of a linguistic expression are utterly related to–and depend on–the message that the speaker wishes to communicate. Thus, a formally complex linguistic expression, consisting of various competing antecedents or inserted clauses, may influence the speaker’s propensity for pauses and on the need to reformulate or repeat some elements of the linguistic expression. Needless to say, it is not very likely that the pauses, the reformulations and the repetitions have any impact on the structural properties of the linguistic expression. In sum, the examples examined above corroborate the claim that the cognitive approach to reference and referential expressions may explain occurrences of a prominent adverbial clause in co-referential contexts. Studies within the pragmatic perspective, on the other hand, lead to a non co-referential reading of these examples. In the following section, I will show that the cognitive approach not only is valid for natural spoken language, but also for written language.

Written Language The cognitive need to use a prominent adverbial clause in a co-referential context in natural spoken language also holds for cases of written language. That is, if the referent of the trajector of the adverbial clause has a low degree of accessibility in the active short-term memory, the reactivation of this participant in the adverbial clause is natural. In contrast, the use of an adverbial clause without an elaborated trajector emerges in cases without the cognitive need to highlight the trajector. This latter tendency is exemplified below: 72. Mesmo as próprias direcções Same the proper managements Ambição para ter uma Ambition to have-INF. a [Diário de Coimbla-N2419-1] ‘The management themselves do not even high-standard team’.

não not equipa team

revelam show-pres: 3p.p de alto nível of high level

show any ambition to have a

74

Chapter Two

Example (72) does not display any cognitive factor that would justify the use of a prominent adverbial clause with an elaborated trajector. The antecedent in the main clause is fully accessible: there is no formal distance between the clauses and there is no competition between potential referents to the adverbial trajector. Thus, there is no cognitive need for an elaborated trajector in the adverbial clause. The following occurrences, however, exemplify the same cognitive need as the one existing in natural spoken language: 73. deixarão

de ter que percorrer mais de dez traverse-INF. more of ten stop-FUT: 3P.P of have-INF. that quilómetros até à sede do concelho para kilometers until to the seat of the county to frequentarem a escola the school frequent-INF: 3P.P [Diário de Aveiro-N4044-11] ‘they will no longer need to travel more than ten kilometres, to the county centre, to attend school’.

74. enquanto houver possibilidades matemáticas de lá mathematical of there while be-FUT.SUBJ. possibilities chegar vamos lutar até à exaustão arrive-INF. go-PRES: 1P.P fight-INF. until to the exhaustion dentro de todas as nossas forças, para the our-PLUR. strength, to inside of all-PLUR. conseguirmos aquilo que de facto nestas ultimas that-DIST. that of fact in these last achive-INF: 1P.P semanas perdemos weeks lose-PRET: 1P.P [Diário de Coimbra-N2908-1] ‘As long as there is a small chance of succeeding, we will fight with all our strength to achieve what we have lost in the last weeks’. 75. Chegaram no início deste mês a Coimbra in the beginning of this month to Coimbra Arrive-PRET: 3P.P para, no Hospital Militar, receberem as próteses to, in the Hospital Military, recieve-INF: 3P.P the prosthesis de membros perdidos entre estilhaços de bombas. between splinter of bombs. of limb lose-PART. [Diário de Coimbra-N1764-1] ‘They came at the beginning of this month to Coimbra Military Hospital to receive prosthetic limbs to replace those they had lost due to the splinter bombs’.

The Degree of Prominence in Infinitive Adverbial Clauses

75

Examples (73-75) represent expressions in which the insertion of linguistic units creates a greater formal distance between the main clause trajector and its co-referent in the adverbial clause. In (73), the formal distance is due to the insertion of the NP dez quilómetros and the adverbial locative até a sede do concelho. It is also noteworthy that there are two occurrences of the plain infinitive, e.g., ter and percorrer between the main clause verb and the adverbial clause. The next example (74) displays the same pattern. The adverbial phrase dentro das nossas forças creates a formal distance between the antecedent and the co-referent. Finally, the same pattern prevails in (75) where the locative phrase no Hospital Militar effectuates the greater formal distance. As a result, the use of an elaborated trajector in the adverbial clause is not really surprising. Rather, it is a strategy to reactivate a participant that is not accessible in the short-term memory. It is also possible to come across examples where a pause (the comma (,)) seems to trigger the use of an elaborated adverbial trajector: 76. Os nossos atletas precisam de need-PRES: 3P.P of The our-PLUR. athletes aos das outras equipas, para to the of the other teams, to competir compete-INF. [Diário de Aveiro-N1540-1] ‘Our athletes need the same conditions as the compete’.

condições idênticas conditions identical poderem be able to-INF: 3P.P

other teams to be able to

Hence, the pause creates a lesser degree of cohesion between the main clause trajector and the trajector of the adverbial clause. Subsequently, this may create a cognitive need to express a more prominent adverbial clause, i.e., an adverbial clause with an elaborated trajector. Moreover, there is a competition between various potential antecedents in the example, that is, os nossos atletas and os das outras equipas. It is not too hazardous to claim that this feature contributes to the elaborated trajector in the adverbial clause. To conclude, it is reasonable that the same factors that contribute to the use of an elaborated adverbial trajector in natural spoken language also may explain the occurrence of this participant in written language. A formally complex sentence triggers the use of an elaborated trajector by the fact that this participant is not at the centre of attention, it is not accessible in shortterm memory. Further, the use of commas creates a lesser degree of cohesion between the antecedent and the co-referent and, therefore, the

76

Chapter Two

strategy is to overtly mention the adverbial trajector. Otherwise, this participant may disappear in the context.

Conclusion The subject of the present chapter has been to find a plausible explanation for the use of an adverbial clause with an elaborated trajector in coreferential contexts. In order to do so, the first assignment was to find the proper conceptual meaning for the trajector and to explore the relation between this participant and the landmark. In accordance with Langacker (1987, 1991), the relation was defined as a linguistic manifestation of the perceptual figure and ground asymmetry. The trajector is the primary figure in a predicative relation and the landmark is the secondary figure in a predicative relation. A prominence scale regarding infinitive and finite adverbial clauses was further established. The underlying idea in the creation of the scale was that the adverbial clause increases in prominence according to the morphological information about the trajector of the clause. An adverbial clause without morphological information about the trajector has consequently a lesser degree of prominence than a clause with an elaborated trajector. The morphological information corresponds to explicit mention of the trajector by the use of a personal pronoun and/or verbal inflection on the infinitive. The mention of an elaborated trajector in co-referential contexts was related to anaphoric expressions. First, the pragmatic approach to anaphoric expressions was analysed. The analysis revealed that pragmatic principles – such as Levinson’s principles for anaphoric expression – are not able to explain the use of a prominent adverbial clause in co-referential contexts. Rather, they imply a non co-referential relation between the antecedent and the co-referent. The same pattern was at hand in the analysis of the Relevance Theory (Sperber and Wilson 1990, 1995). That is, the use of an elaborated trajector in co-referential contexts is considered irrelevant because it does not furnish any contextual effects. This being so, the cognitive explanation for anaphoric expressions proved to be more accurate. In fact, the cognitive approach considers the distinction between co-reference and non co-reference to be secondary. The main issue is related to the accessibility of the referent and not to the question whether this participant is co-referential or not. Thus, the use of an elaborated trajector in co-referential contexts may find an explanation in terms of accessibility, activation and attention. In other words, the uttering of prominent adverbial expression is explained by the cognitive need to put focus on a participant that is disappearing in the context, a participant that is

The Degree of Prominence in Infinitive Adverbial Clauses

77

not in the speaker’s short-term memory. In short, the use of an elaborated trajector is a strategy to maintain the notion of this participant. This being so, the choice of one or another referential expression reflects the claim made in Cognitive Grammar that language and language use is a problem solving activity. Finally, the creation of the prominence scale revealed the difficulty of interpreting the function of the clitic pronoun se in the structure with verbal inflection (3p.p). This is evidently a problem that merits further analysis. The issue is considered in more detail in the following chapter.

CHAPTER THREE THE CLITIC PRONOUN SE: A GENERIC TRAJECTOR

Introduction The prominence scale elaborated in the previous chapter revealed the difficulty of grasping the meaning and function of the clitic pronoun se. This problem will be further analysed in the present chapter. Thus, the question is whether the clitic pronoun se only may be conceptualized as a generic trajector, i.e., an undetermined and active agent, in structures lacking another participant or lacking verbal agreement with this participant. In other words, is the conceptualization of an active participant totally missing in cases where there is verbal agreement with the passive participant? The following examples illustrate the problem: 77. As vendas de pipocas para se fazer em CLITIC.PRON. make-INF. in The sale of popcorn to microondas representam hoje um negócio próximo today a business near microwaves represent-PRES: 3P.P dos 400 milhões de dólares. of the 400 millions of dollars. [Diário de Notícias, 19970421] ‘The market for microwave popcorn is today almost 400 million dollars’. 78. De resto, para o autarca é claro que Of remainder, to the Mayor be-PRES: 3P.S clear that ainda é cedo para se estarem a early to CLITIC.PRON. be-INF: 3P.P to be-PRES: 3P.S fazer hipotéticos cenários autárquicos. non-governmental. make-INF. hypothetic scenes [Diário de Aveiro-N1080-1] ‘Apart from that, the Mayor firmly believes that it still is too early to make hypothetical scenarios about autonomy’.

80

Chapter Three

Traditionally, the clitic pronoun se is considered an impersonal subject in (77), and a passive marker in (78). That is, there is a notion that se represents an active and generic subject in cases where the verbal agreement does not agree with the passive participant in the described event. On the other hand, the notion of an active initiator of the described event disappears in cases like (78). If the verb agrees with the passive participant (the logical object), subject status is conferred to this participant. Consequently, the notion of an active and generic agent fades away. The position taken in the present study, however, is that there are cases in which an absolute distinction between an impersonal and passive reading of se is problematic. The question should not be considered in absolute terms. Rather, the function and meaning of se is a gradient phenomenon, depending on the context in which it occurs. That is, the mere fact that the plural NP has subject status does not actually make the notion of an active and generic initiator of the described event disappears. The question is even more subtle when considering the singular cases. The formal treatment of giving subject status to the plural NP is actually difficult to make in singular cases, given the neutrality of the verbal inflection. Accordingly, the interpretation of singular cases with se is utterly context-dependent. Below, is a typical occurrence with the structure SE + SINGULAR NP: 79. É de um grande irrealismo supor of a great unreality suppose-INF. Be-PRES: 3P.S -se que se pode governar CLITIC.PRON that CLITIC.PRON. be able to-PRES: 3P.S govern-INF. sem se conhecer a realidade without CLITIC.PRON. know-INF. the reality [Diário de Notícias, 19970426] ‘It is unrealistic to suppose that it is possible to govern without knowledge of the political realities’.

The adverbial structure sem se conhecer a realidade seems to evoke an active and generic interpretation, and not a passive one. Further, a reflexive reading seems highly unrealistic, given the existence of the non-animate NP a realidade. That is, although the verb conhecer may have reflexive use, e.g., to know myself, this is not the case in (79). Hence, the propensity for interpreting se as a generic trajector depends heavily on the contextual nature of the described event. I would propose four features of main interest: (1) the respective position of se and NP, (2) the existence of an obvious referent to se, (3) the type of verb and finally (4) the type of NP.

The Clitic Pronoun Se: A Generic Trajector

81

Obviously, the verbal inflection is relevant in the attempt to make a semantic and functional interpretation of the clitic pronoun se. However, I strongly believe that the features mentioned above may have a certain impact on how we interpret structures with se. That is, these features may improve our understanding of a delicate and most subtle issue, not only regarding singular cases with se but also in relation to structures with a plural NP.

The Clitic Pronoun Se The present section will begin with some introductory considerations regarding the function and semantic meaning of se. Thereafter, the effect of defocalization and focus displacement will be analysed. Finally, the propensity for conceptualizing se as an external force will be discussed.

Function and Semantic Meaning The clitic pronoun se holds a rather exceptional position in the pronoun category. In fact, one of the most typical features of this pronoun is related to the vague and indefinite character that it displays. This vagueness is, among other things, visible in the lack of gender and number features. While other personal pronouns have different forms expressing gender and number, se is invariable in this respect: Figure 3:1. Gender and number of pronouns in Portuguese. Masculine

Feminine

Singular

Plural

Nominative

ele

ela

ele/ela

eles/elas

Accusative

o

a

o/a

os/as

lhe

lhe

lhe

lhes

Dative

Consequently, personal pronouns in the nominative and accusative case express gender and number. Further, dative case pronouns furnish a distinction between singular and plural. The clitic se, on the other hand, does not express gender, or number. Clearly, this feature contributes to the undetermined and vague nature of this pronoun. Another feature of this pronoun is its multi-functionality. In Cognitive Grammar terms, se not only functions as a generic trajector but also as a

82

Chapter Three

landmark. This latter function is typical for the use of se as a reflexive marker. The following examples (80-83) illustrate some standard uses of this pronoun: 80. e está provado e mais que provado que and more than proved that and be-PRES: 3P.S proved os árbitros que se vendem CLITIC.PRON. sell-PRES: 3P.P the referees that [Pfa: Desporto e Dinheiro] ‘And it has been proved beyond any reasonalble doubt that the referees are bribed’ 81. vamos procurar que junto dos nossos colegas search-INF. that near of the our colleagues go-PRES: 1P.P possa existir alguma coordenação para some coordination to be able to-PRES.SUBJ: 3P.S se desenvolverem iniciativas CLITIC.PRON. develop-INF: 3P.P initiatives [Diário de Aveiro-N0778-1] ‘We will try to establish some form of coordination with our colleagues to develop initiatives’ 82. Mas os reparos estenderam -se ainda à But the criticism extend-PRET: 3P.P CLITIC.PRON even to the falta de estruturas para se instalar máquinas CLITIC.PRON install-INF. machines absence of structures to de cortar caldo-verde of cut-INF. broth-green [Diário de Aveiro-N1975-1] ‘But the criticism extended to even include the lack of structures necessary to be able to install the machines that prepare the potato and cabbage broth’

The first example (80) illustrates the reflexive use of se. The subject of the verb vendem-se (‘sell themselves’) is also the direct object of the verb. The subsequent example (81) represents a structure that is similar to the reflexive construction. The difference, however, is that the plural NP is nonanimate. Consequently, the reflexive interpretation is ruled out. It is, in fact, not very likely that the initiatives “develop themselves”. The traditional claim is that se is a passive marker in this example. Finally, example (82) illustrates the impersonal use of se. The verb does not agree with the plural NP, thus giving rise to the conceptualization of se as the subject of the clause–a generic trajector in the terminology of the present analysis.

The Clitic Pronoun Se: A Generic Trajector

83

The examples displayed above (80-82) also illustrate one of the most obvious functions of se. As García (1975: 7) eloquently points out, this clitic creates a kind of referential short-circuit. In the reflexive and passive use, se blocks the insertion of participants other than the grammatical subject. The message is that there are no other participants. In the generic use, se has a defocalization effect on the active participant of the described event. Instead of pointing out the specific trajector, se conveys the message that this participant is someone or everyone, i.e., that the trajector is unspecified. García (1975: 67) further connects the vague and unspecified character of se to the distinction between what se refers to as l and s forms. The former include the subject forms ele, ela, and the dative forms lhe and lhes. The latter is related to se, the possessives seu, sua and the prepositional form si. In her opinion, the semantic difference between the l and s forms is explained in terms of deixis, defined in the following manner: “the force with which the hearer is instructed to seek the referent of the pronoun. HIGH DEIXIS tells him: ‘Find the person referred to!’ – an instruction that is appropriate where the antecedent of the pronoun is not obvious, or is relatively hard to find or should under no circumstances be missed. LOW DEIXIS tells him: ‘Seek neither hard nor far for the third person referred to’, – an instruction appropriate where the antecedent is obvious, easy to find, or where it does not much matter if the referent is identified or not (as in the case of the ‘impersonal’ se, which in terms of specific reference can perfectly well be hanging) (García 1975: 65).

Thus, the difference in referential terms is that the s forms, being of low deixis, instruct the hearer that there is no specific reason to search for the referent. The reason for not doing so is that the referent is obvious, easy to find or does not really matter. The l forms, on the other hand, instruct the hearer to search for the referent, either because he/she is hard to find or because he/she should not be missed. This being so, the use of se can be understood of as a strategy of defocalization and focus displacement. Instead of putting the initiator and energy source of an event on “stage”, the speaker may choose to “downgrade” the notion of this participant. A common denominator for the different uses of se is that it occurs in contexts where the expression deviates from the prototypical conceptualization of events. This is the case with the reflexive use of se in which the same participant acts like the grammatical subject and direct object. It is also the case in the passive construction where the passive participant recieves subject status. Finally, the use of se as a generic trajector corresponds to a deviation from the tendency that the trajector prototypically corresponds to

84

Chapter Three

a well delimited and concrete participant (cf. Chapter 2). This latter strategy will be the topic of the subsequent section.

Defocalization and Focus Displacement The strategy of defocalization and focus displacement is, first and foremost, dependent on three different factors: the use of se instead of an identified trajector, the verbal agreement with a passive and non-animate NP and, finally, the internal position of the relational participants se and NP. Evidently, the description of an event includes a variation of these factors which contributes to a weakening of the notion of an active participant to a higher or lesser degree. In the unmarked case, the referential issue is non-problematic. A purpose adverbial expression like: para ela abrir a porta (litt: to she open the door) or para abrirmos a porta (litt: to we open the door) designates a specified referent in the described event. However, the lack of an obvious referent in the described event may or may not include focus displacement. The following examples (83-85) illustrate defocalization without focus displacement: 83. para se abrir as portas CLITIC.PRON. open-INF. the doors to ‘for someone to open the doors’ 84. para abrirem as portas the doors to open-INF: 3P.P ‘for someone to open the doors’ 85. para as pessoas abrirem as to the persons open-INF: 3P.P the ‘for the people to open the doors’

portas doors

The examples offered above coincide in the lack of referentiality. In the first example (83), the use of se instead of a lexical subject or a subject pronoun creates a defocalization of the referent. The same pattern is at issue in (84). In this case, the verbal inflection in third person plural is similar to the English use of they in impersonal expressions like: they say that it is going to rain tomorrow. In example (85), finally, the expression as pessoas (‘people’) illustrates another common strategy of defocalization. The change of focus represents one further step in the process of downgrading the notion of an active agent. Instead of putting focus on the active participant, it is directed towards other participants in the conceived

The Clitic Pronoun Se: A Generic Trajector

85

event. The focus displacement is achieved by conferring subject status to other participants than the actual energy source of the event, either by means of verbal inflection or by verbal inflection and a positional change of the participants: 86. para se abrirem as CLITIC.PRON. open-INF: 3P.P the to ‘for the opening of the doors’ 87. para as portas se to the doors CLITIC.PRON. ‘for the doors to open’

portas doors

abrirem open-INF: 3P.P

Example (86) illustrates the ambiguity of passive constructions. In fact, there may be two alternative interpretations of the described event. Obviously, the verbal inflection, conferring subject status to the doors, has the effect of putting focus on this participant. Nonetheless, the subject position of se also makes it possible to conceptualize this participant as a generic trajector. In the following example (87), however, the change of internal position between the doors and se seems to make the focus displacement even stronger. The initial position of the NP (‘the doors’) cooperates with the plural inflection in conferring a high degree of topicality to the NP (cf. Givón 1992). In conclusion, the vague and undetermined character of se makes it open to conceptual ambiguity. Further, this vagueness has a certain effect on the degree of prominence attributed to se. In cases where se is conceptualized as a generic trajector, it seems to gain prominence. If the conceptualizer, on the other hand, does not conceptualize se as a generic trajector, focus and prominence is moved to other participants of the described event. The verbal inflection and the internal position of the clitic se and the NP play an important role in this process. Nonetheless, the subsequent section will demonstrate that the way in which se is conceptualized reflects the tension between imagery, i.e., the ability to describe a situation in different ways for communicative purposes, on the one hand, and human reasoning about the occurrence of events, on the other.

Se as an External Force Among others, Maldonado (1992: 233) comments upon the passive and impersonal uses of se as extensions from its basic reflexive meaning. Thus, the initial meaning of se as a reflexive marker gave rise to what is referred to as Middle Voice. Quesada (1998: 13) defines this phenomenon as: “the

86

Chapter Three

tension between the tendency to keep a transitive relation (agentivity) and the desire to highlight the patient role of the subject” (Quesada 1998: 13). Certainly, this tension creates a certain ambiguity related to the meaning of the clitic se. If the conceptualizer considers the event to be transitive, the tendency to conceptualize se as a generic and active trajector increases. If the conceptualizer does not, the passive reading is preferred. This being so, examples like the following one are highly ambiguous: 88. Antes de se apresentarem projectos de lei Before of CLITIC.PRON. present-INF: 3P.P projects of law que demoram poucos minutos a ser feitos, minutes to be-INF. make-PART. that delay-PRES: 3P.P few é preciso começar por elaborar um necessary begin-INF through elaborate-INF. a be-PRES: 3P.P estudo study [Viseu Diário-N2393-1] ‘Before the presentation of legal projects, it is necessary to begin with the elaboration of a study’

According to Maldonado (1992: 242) there are two basic factors to consider in the choice between a passive and an impersonal reading of examples like (88). The first is related to the nature of the theme (projectos) and its capacity to bring on the event. The second is related to the nature of the verb, that is, if the verb describes actions that can be accomplished by a non-animate participant. Evidently, there are non-animate entities capable of initiating changes: a plate may break because it is old, and the natural world consists of powerful elements able to initiate a great number of processes. The main issue, however, is if the non-animate NP (singular or plural) is capable of initiating the process described by the verb. The difference between NPs capable of such actions and NPs that are not is illustrated in Figure 3:2:

The Clitic Pronoun Se: A Generic Trajector

87

Figure 3:2. External or internal energy of events (cf. Maldonado 1992: 243).

A) External force

B) Internal force

Thus, the way in which an event is conceptualized depends on factors related both to the form of the linguistic expression describing the event and to common human experience about events and processes in the world. Certainly, factors as verbal agreement and internal position create a focus displacement in the described event. This effect is a direct manifestation of the claim made by Cognitive Grammar about the human capacity to create a linguistic expression in accordance with the communicative purpose. On the other hand, however, an encyclopedic knowledge of events and processes has a crucial role in how to conceptualize se. Therefore, the hypothesis put forward is that the tendency to conceptualize se as a generic trajector increases when the described event depends on an external force in order to be accomplished.

Analysis The present analysis aims at highlighting some of the factors that may lead to one or another interpretation of adverbial structures with se and a plural NP. Special attention will be given to the degree of transitivity in the described event, to the possibility to determine the number of participants in the event and to the internal position of these participants. Further, occurrences with an obvious referent to se will be studied, as the effect of the verbal inflection. Finally, different structures with se will be analysed in the light of the conceptual differences they convey.

88

Chapter Three

The Degree of Transitivity In order to study the degree of transitivity in the structure with se and a plural NP, the present analysis will make use of the transitivity parameters elaborated by Hopper and Thompson (1980). The most obvious feature of transitivity is the transfer of energy from one participant to another, but other features such as kinesis, aspect, volition and others influence the degree of transitivity in an event. The parameters are listed below: Figure 3:3. Transitivity parameters (Hopper and Thompson 1980: 252). TRANSITIVITY

HIGH

LOW

PARTICIPANTS

2 or more participants (A and O)

1 participant

KINESIS

action

non-action

ASPECT

telic

atelic

PUNCTUALITY

punctual

non-punctual

VOLITIONALITY

volitional

non-volitional

AFFIRMATION

affirmative

negative

MODE

realis

irrealis

AGENCY

A high in potency

A low in potency

AFFECTEDNESS OF O

O totally affected

O not affected

INDIVIDUATION OF O

Individuation of O

O non-individuated

The most obvious condition for a transitive relation is the existence of two participants: the agent (A) and the object (O). Further, there is a distinction made between action and non-action in the sense that verbs expressing states rather than actions display a low degree of transitivity. The nature of the action also influences the degree of transitivity. Telic action in which there is an endpoint has a higher degree of transitivity than atelic action. The difference can be seen in expressions like: I ate it up and I am eating it (cf. Hopper and Thompson 1980: 252). In addition, punctual actions display a higher degree of transitivity than actions that are nonpunctual. Thus, verbs with a punctual aspect (kick, break, give, put etc.) are considered to increase the transitivity of an event. Finally, the degree of transitivity increases if the action is conceived of as intentional, affirmative and has the status of being real. That is, if the action really has occurred.

The Clitic Pronoun Se: A Generic Trajector

89

Moreover, there is a distinction made between agents with a high or low degree of potency. Generally, an animate agent has a higher degree of potency than a non-animate agent. Thus, John gave me an idea is more transitive than the painting gave me an idea. In the latter case, there is no actual action involved but a mental/internal one. As a final point, Hopper and Thompson (1980) highlight the character of the object (or patient). An affected object presupposes a higher degree of transitivity than an object that is not affected by the action. And so does an individuated object (cf. the distinction made between linguistic categories in Chapter 2.) Obviously, the parameters discussed above are highly inter-relational. For example, the affectedness of the object is related to the action expressed by the verb. A verb expressing a telic and punctual action tends to affect the object to a higher degree than an atelic and non-punctual verb. This is, for example, the case in the distinction between I gave her and I like her. In the latter case, the atelic and non-punctual character of the verb coincides with a lesser degree of affectedness of the object. This is also the case in expressions describing the non-occurrence of an action, like: I did not kick him (negative, irrealis). Relating the Transitivity parameters to Portuguese adverbial clauses with se it is pertinent to first identify some occurrences that do not exhibit any transitivity. The following reciprocal and reflexive examples are prototypical cases in point: 89. As pessoas cruzavam -se e The persons cross-IMP: 3P.P CLITIC.PRON. and Reencontravamse, muitas delas, após mais meet again-IMP: 3P.P CLITIC.PRON. many of them, after more de 25 anos sem se verem CLITIC.PRON. see-INF: 3P.P of 25 years without [Diário de Leiria-N2774-2] ‘The people met up, many of them after more than 25 years of not seeing each other’ 90. Depois de duras negociações, o guarda-redes Costinha e After of hard negotiations, the goal keeper Costinha and o médio Tiago decidiram recusar algumas the midfielder Tiago decide-PRET: 3P.P refuse-INF. some propostas para se manterem em Espanha CLITIC.PRON. remain-INF: 3P.P in Spain offers to [Diário de Leiria-N2886-1] ‘After tough negations, the goal keeper Costinha and the midfielder Tiago decided to turn down the offers and stay in Spain’

90

Chapter Three

In terms of transitivity, perception verbs like ver (‘to see’) express input rather than output. The conceptualizer perceives an event in the surrounding world and does not actually perform an action. Thus, the use of se in (89) has a reciprocal meaning (‘each other’). That is, the adverbial expression sem se verem does not express a generic trajector but the event of not seeing each other. In (90), the reflexive use of the verb manter (‘to maintain’, ‘to support’ etc.) expresses the idea of staying. The clitic se has the effect of changing the meaning of the verb and it further impedes the insertion of other participants than the verbal subjects. In contrast, however, the subsequent examples (91-92) show evidence of transitivity. Moreover, it is noteworthy that the prepositions introducing the adverbial clause seem to have semantic features that increase the degree of transitivity: 91.

Diamantino Lopes apelou à necessidade de Diamantino Lopes appeal-PRET: 3P.S to the need of passar por cima de querelas partidárias’ para through over of disputes party to pass-INF. se resolverem os verdadeiros problemas CLITIC.PRON. solve-INF: 3P.P the real problems [Diário de Aveiro-N0368-1] ‘Diamantino Lopes appealed to the need to put aside inter-party disagreements in order to solve the real problems’

92. Parece -lhes que os Xutos e Pontapés nunca the Xutos and Pontapés never Seem-PRES: 3P.S them that mais sobem ao palco, onde o cenário, cheio more go up-PRES: 3P.P to the stage, where the scenery, full de dados azuis já está coberto com uma with a of dice blue already be-pres: 3p.s cover-PART. pequena nuvem de fumo. Assobios de impaciência até small cloud of smoke. Whistles of impatience until se apagarem as luzes CLITIC.PRON. turn of-INF: 3P.P the lights [Diário de Noticias, 19970426] ‘It seems that Xuto e os Pontapés are not going to return to the stage, where the scenery, so full of blue dice, is already covered with small clouds of smoke. Impatient whistling continues until the lights are switched off’

The common denominator for examples (91-92) is that the verb in the adverbial clause designates an action that is completed (telic). The acts of solving a problem (91) and switching the lights off (92) express a changed situation–from one state to another. Further, the plural NPs are individuated.

The Clitic Pronoun Se: A Generic Trajector

91

In example (91), the NP as luzes includes a definite article and the same holds for the NP os verdadeiros problemas in the subsequent example. Finally, the preposition para expresses a purpose and volitionality, while the preposition até is punctual. That is, the prepositions seem to share some inherent characteristics that correspond to a higher degree of transitivity. The following case, an adverbial clause introduced by the temporal depois de, designates an action that is seen as completed: 93. O porta-voz do MNE salientou que o reatamento The spokesman of MFA stress-PRET: 3P.S that the resumption de relações diplomáticas entre os dois países of relations diplomatic between the two countries ocorre depois de se terem after of CLITIC.PRON. have-INF: 3P.P occur-PRES: 3P.S concretizado três ‘factores importantes’: Timor-Leste important’: Timor-Leste achieve-PART. three ‘factors estar sob administração da ONU, saída be-INF. under administration of the UN, depart-PART. do território de todos os militares indonésios e of the territory of all the militaries Indonesian and a libertação de todos os presos políticos timorenses the liberation of all the prisoners political Timorese [Diário de Leiria-N1351-1] ‘The spokesman of the MFA stressed that the resumption of diplomatic relations between the two countries had occurred after the fulfilment of three important conditions: that Timor-Leste is under the administration of the UN, the departure of the Indonesian military and the liberation of all political prisoners’

Example (93) describes actions that have the status of real occurrences. Furthermore, the detailed description of the three conditions accomplished: (1) that Timor-Leste is under the administration of the UN, (2) the departure of the Indonesian military and (3) the liberation of all political prisoners contributes to an increase in the transitivity of the described events. It is not likely that these processes occurred without the transmission of energy from active participants. Thus, the conclusion is that the structures with se and a plural NP very well may provide characteristics that are compatible with a high degree of transitivity. One problem that arises, however, is how to analyse the number of participants in the described event. In fact, both se and the plural nonanimate NP are rather weak candidates for assuming the subject role. This being so, it is interesting to consider the claim made by Quesada (1998) that structural restrictions explain why the plural NP cannot be the subject of

92

Chapter Three

passive sentences in Spanish. These restrictions are: (1) it is impossible to have subject agreement with a plural animate NP that is specific, (2) the plural NP does not control the event and (3) it is not possible to insert an agent into the event. The following Spanish examples illustrate the point made by the author: 94. * Se CLITIC.PRON

95. * Se CLITIC.PRON

a to

matan a los criminales kill-PRES: 3P.P to the criminals buscan search-PRES: 3P.P

empleadas y employees and

buscan search-PRES: 3P.P

empleadas employees

comienzan begin-PRES: 3P.P

trabajar work-INF.

96. * Se CLITIC.PRON.

por through

los the

patrones employers

Example (94) illustrates the impossibility to have verbal agreement with an animate and specific plural NP. In these cases, the verbal inflection has to be in third person singular–which essentially increases the notion of the se as a generic trajector. The ungrammaticality of the following example (95) is due to the fact that the deleted subject NP in the second conjunct has to be identical with the subject of the first. This is not the case, and therefore it is not grammatical. 13 Further, this entails that se may function as an unspecified searcher, i.e., a generic trajector. Finally, the fact that it is impossible to attach an agent by means of a por-phrase in (96) implies that there already is an unspecified agent in the described event (cf. Quesada 1998: 18-19). Even so, the clitic se is also a poor controller. Example (97) gives evidence of this feature: 97.

13

...e

depois …and after tinha have-IMP: 3P.S areia

para to que that

se CLITIC.PRON.

se CLITIC.PRON.

sair leave-INF. atravessar cross-INF.

dali from there aquele bocado de that bit of

In order to make the sentence grammatical it is necessary to insert a NP in the second conjunct: Se buscan empleadas e ellas comienzan a trabajar. In this sense, the sentence is not comparable with a sentence like: Pedro vuelve a casa y comienza a preparar la comida (‘Pedro returns home and begins to prepare the supper’). In the latter sentence, the subject of the first sentence also controls the second sentence.

The Clitic Pronoun Se: A Generic Trajector

93

sand [Pfu: Um Sonho] ‘And afterwards, to leave from there one had to cross that bit of sand

That is, in order for se to maintain the subject role, it is necessary to insert this participant not only in the adverbial clause, but also in the main clause. Hence, the clitic pronoun se and the plural NP share the feature of being rather weak subject candidates. Thus, a phenomenological perspective on the issue offers an alternative analysis. 14 Given the fact that both the plural NP and se are non-typical subjects, the tendency to attribute subject status to one of them may be related to general human experience of event structure. In accordance with Maldonado (1992), the claim made in the present analysis it that the conceptualization of se as a generic trajector correlates with the type of action that the verb designates. If the verb in question designates an action that typically includes an agent, the propensity for conceptualizing se as a generic trajector will increase: 98.

É igualmente inconcebível que para likewise incredible that to Be-PRES: 3P.S se arranjarem 2 quilómetros de via CLITIC.PRON. prepare-INF: 3P.P 2 kilometres of road se ande já desde há CLITIC-PRON. walk- PRES.SUBJ: 3P.S already from exist-PRES: 3P.S oito meses a abrir e tapar buracos cover-INF. holes eight months to open-INF. and [Viseu Diário-N3024-1] ‘Likewise, it is incredible that they have spent eight months opening and covering holes to repair 2 kilometres of road’

99. ...antes de se tomarem decisões deve ...before of CLITIC.PRON. take-INF: 3P.P decisions must -se ter certeza do que se quer CLITIC.PRON have-INF. certainty of what CLITIC.PRON. want-INF. [Diário de Coimbra-N3114-1] ‘Before making decisions one must be sure of what one wants’ 100. Segundo According ouvirem 14

o jovem alvejado, antes de se before of CLITIC.PRON. the young boy target-PART os disparos, o ‘Caroceiro’ fez

See Langacker (2008: 31) for the relation between Cognitive Grammar, conceptualization and the phenomenological and processing levels of conceptualization.

94

Chapter Three the shots, the ‘Caroceiro’ make-PRET: 3P.S hear-INF: 3P.P uma ameaça a threat [Viseu Diário-N2218-2] ‘According to the young boy that was hit, ‘Caroceiro’ made a threat before the shots were heard’

The examples (98-100) displayed above differ regarding the action they designate. In the first example (98), the adverbial clause para arranjarem 2 quilómetros de via designates physical action, while the action in the subsequent example–the one of solving problems–is mental in character. Finally, example (100) illustrates a perceptual event, designated by the verb ouvirem. Consequently, these examples form a scale according to the degree of action in the event: physical action, mental action and perceptual action. In fact, these differences are likely to influence our inclination to make one or another reading of the clitic se. If the described event entails a high degree of activity, like the physical activity in example (98), it is not too risky to claim that the propensity to consider se as a generic trajector increases. Likewise, the act of making decisions in example (99) requires a “problem solver”. Therefore, it is likely that se symbolizes this participant. The third example (100) deviates from this pattern. In fact, it designates a rather ambiguous situation. As commented upon in relation to example (89), perception verbs seem to display a low degree of transitivity. The issue is also related to the direction of the sound waves in the conceived event. A plausible hypothesis is that the sound is the moving entity in example (100). Thus, the notion of a generic trajector may decrease. Further, actions described by verbs like comer, tomar, beber, comprar etc. facilitate a benefactive interpretation of the event. As commented upon above, the benefactive use of se is often considered to be an extension of the reflexive se. The main difference is that reflexives impede the insertion of a direct object, while the benefactive se impedes the insertion of a third participant different from the grammatical subject, that is, a dative object. In this sense, the benefactive se has possessive features and further it tends to designate an action that has positive consequences for the subject: 15 101. No Japão, para In the Japan, to tem que have-PRES: 3P.S that o estacionar the park-INF. 15

se

comprar um buy-INF. a se ter CLITIC.PRON. have-INF.

CLITIC.PRON.

carro, car, espaço para space to

See Maldonado (1992) and Quesada (1997) for detailed discussions on the matter.

The Clitic Pronoun Se: A Generic Trajector

95

[Diário de Notícias, 19970420] ‘In Japan, when buying (yourself) a car one is going to need a lot of space to park it’

Even though a benefactive reading is highly probable, example (101) is ambiguous. In the benefactive interpretation, the meaning is to buy oneself a car. On the other hand, the generic interpretation is also possible. That is, the example may also have the meaning to someone buy a car. The generic reading finds further support in the main clause event tem que se ter espaço, where se without any doubt designates a generic trajector. Further, the verbs offering a benefactive reading often designate events that presuppose a subject. Accordingly, example (101) offers a delicate problem. A reasonable interpretation, however, is that the generic trajector is expressed by the infinitive form (comprar). 16 From the phenomenological perspective, the character of the NP plays another important role. If the NP is not very likely to display the typical features of an active participant, the more likely it is that the conceptualizer will conceive of the situation as having more than one participant. Obviously, the NPs displayed in the examples of the present section (89101) share more features with a typical object than with the typical subject (cf. Givón 1995). This being so, the propensity for conceptualizing se as a generic trajector is comprehensible – not only in the singular cases, but also with a plural NP.

Obvious Referents In previous sections, the non-animate plural NP was said to increase the notion of se as a generic trajector. In the present section, some special cases with an obvious referent to se will be analysed. Prototypically, the feature that makes referentiality possible is the animacy of the NP. The animacy may be explicit, i.e., the actual mention of an animate referent, or it may be expressed in metonymic fashion. First, a case of explicit mention of the referent:

16 That is, infinitive expressions tend to evoke a non elaborated generic trajector. In a case like: to watch television is boring, for example, the infinitive form watch evokes a non elaborated experiencer. The issue in the present chapter, however, is related to se and the question as to whether this pronoun symbolizes an elaborated generic trajector

96

Chapter Three 102. …as mulheres reúnem -se ao fim CLITIC.PRON. to the end …the women-PLUR. meet-PRES: 3P.P da tarde no largo da aldeia, partindo of the afternoon in the square of the village, go out-PROG. depois para um restaurante da região, onde fazem after to a restaurant of the region, where make-PRES: 3P.P ‘uma grande festança’, antes de se deslocarem ‘a great party’, before of CLITIC.PRON. dislocate-INF: 3P.P para uma discoteca to a disco [Diário de Coimbra-N1364-1] ‘In the afternoon, the women meet in the village square, after which they go to a typical local restaurant where they kick the party off before going on to the disco’

It is difficult to find another reading than the reflexive one in cases like (102). The main reason for this is that there is an animate NP capable of introducing energy in the event. Likewise, this participant is well defined and concrete. 17 This being so, it is pertinent to remember the notion of short-circuit–the effect of se is to exclude the notion of other participants than the subject and to put focus on this particular participant. Other cases with se display a referential relation that is not that obvious. Nonetheless, they correspond to the reflexive use of se as instances of metonymic reference. Although the NP is non-animate, it implies animate participants. The following two examples illustrate this phenomenon: 103. …as empresas associadas passam a dispor to have-INF. …the companies associated pass-PRES: 3P.P de um período alargado para se CLITIC.PRON. of a period extended to adoptarem à legislação vigente adapt-INF: 3P.P to the legislation current [Diário de Aveiro-N0678-1] ‘The associated companies will have an extended period to adapt to the current legislation’ 104. De acordo Of agreement licenciaturas degrees entregues os 17

com Salgado with Salgado em engenharia in engineering dossiers de 70

Barros, existem 150 150 Barros, exist-PRES: 3P.P e já foram and already be-PRET: 3P.P para se

In this sense, it is interesting to compare (102) with (95-96) where the reflexive reading is not at issue.

The Clitic Pronoun Se: A Generic Trajector

97

CLITIC.PRON delivered the files of 70 to submeterem à avaliação to the evaluation submit-INF: 3P.P [Diário de Notícias, 19970422] ‘According to Salgado Barros, there are 150 degrees in engineering and 70 files that have already been delivered for evaluation’

Even though, the plural NPs in (103-104) are non-animate, they imply animate participants. Example (103) includes a metonymic relation between companies and the people that work in those companies. That is, most companies generally entail animate participants. This being so, the example expresses the notion of human activity. In the following example (104), the metonymic relation is of the type: degrees > files. This metonymic chain indicates a high degree of topicality. Moreover, it includes reference to animate participants. This being so, examples (103-104) are more related to the reflexive use of se than to the conceptualization of a generic trajector.

The Internal Position The present section will deal with the internal position of the NP and se. The claim has been made that the internal position of the NP and se may determine the way in which these linguistic units are conceptualized. Prototypically, the initial position of a clause is reserved for the clausal subject, while less prominent participants in the predicative relation have other positions. The reason for this tendency seems to correspond to natural cognitive processes and to the iconic principle of linear ordering. That is, the description of a process tends to coincide with the way in which the process evolved (cf. Haiman 1985). Langacker (1991: 292-293) relates this tendency to the metaphors energy source and energy sink. In an action chain, the first participant is usually the one who introduces the energy into the event, while the affected participant occupies a position that is downstream in the linear ordering. The prototypical ordering of events is illustrated in Figure 3:4:

98

Chapter Three

Figure 3:4. Prototypical action chain.

tr

lm

time

Figure 3:4 illustrates a rather obvious phenomenon. In order for an action to be set in motion, there is a need for some initial activation. This being so, the tendency for the trajector to correspond to the energy source of an action subsumes that this participant occupies the initial position in the description of the event. Further, the initial position equates with a high degree of prominence. Apart from being the preferred position for the energy source of an action chain, it is also the position occupied by participants with a high degree of topicality (cf. Givón 1990). Consequently, the internal position between the NP and se may influence the propensity for conceptualizing the latter as a generic trajector. If se occupies the initial position, the propensity for considering this linguistic unit as a symbol for a generic trajector increases: 105. ...antes de se tomarem decisões deve ...before of CLITIC.PRON. take-INF: 3P.P decisions must -se ter certeza do que CLITIC.PRON. have-INF. certainty of the what se quer CLITIC.PRON. want-INF. [Diário de Coimbra-N3114-1] ‘Before making decisions, one must be sure of what one wants’

In the same manner, the post verbal position of the NP (as decisões) increases the motivation for considering this participant as an energy sink, that is, a participant that needs external energy in order to be accomplished. Additionally, the passive, undetermined and abstract feature of the NP corroborates the impression that this participant is not a proper candidate for the subject role.

The Clitic Pronoun Se: A Generic Trajector

99

In fact, Maldonado (1992: 285) elaborates a test in which the plural NP is moved to subject position in order to verify if it has subject status. The following example is a modification of (105): 106. ...antes de decisões ...before of decisions * ‘Before decisions take’

se CLITIC.PRON.

tomarem take-INF: 3P.P

Undoubtedly, the initial position of the plural NP produces a semantic conflict. The problem is related to the undetermined and abstract feature of the plural NP and the supposed subject status of this participant. In other words, the low degree of prominence of the NP is not well-matched with the subject position. The subject position is reserved for better suited candidates: the initiator of the event or a well defined participant with a high degree of topicality. In fact, this is the case in the subsequent example: 107. No caso de a final não se disputar In the case of the final not CLITIC.PRON. play os espectadores serão reembolsados the audience be-FUT: 3P.P repaid [Diário de Notícias, 19970421] ‘If the final is not played today, the audience will get refunds’

hoje, today,

Thus, the NP fulfills the requirement for a subject. In this particular case, the initial position of the NP does not create any semantic conflict. Rather, the high degree of topicality of the NP has the effect of decreasing the notion of the actual participants of the game. This being so, the passive (or medial) interpretation is more likely than the generic one. In sum, the internal position of the NP and se has the effect of directing the attention to one or another of these participants.

The Verbal Inflection In the search for a plausible explanation for structures displaying verbal agreement with a non-animate plural NP, it is pertinent to recall the function of the clitic se in these structures. In most cases, the communicative purpose is to reduce the notion of a specific trajector. Thus, the use of the clitic se exemplifies the claim made in the model of Cognitive Grammar that the conceptualization of events may depend on how the speaker wishes to structure his/her message. If the speaker feels the need to focus on the adverbial trajector, he/she will use an overt trajector in the clause. If the

100

Chapter Three

speaker prefers to not focus on this participant, the alternative is to use the clitic se: 108. ‘Ainda durante a manhã lhe telefonei para ‘Even during the morning him telephone-PRET: 1P.S to casa para ele vir almoçar comigo mas with me but home to he come-INF. have lunch-INF. ele não respondeu’… o seu pai tinha the her father have-IMP: 3P.S he not answer-PRET: 3P.S… o hábito de ir ao cemitério aos domingos the habit of go-INF. to the cemetery to the Sundays [Diário de Aveiro-N3930-4] ‘During the morning, I called him at his home to invite him for lunch but he did not answer…Her father used to go to the cemetery on Sundays’ 109. Madail foi claro ao lembrar que, apesar de Madail be-PRET: 3P.S clear to the recall-INF. that, despite of se ter procedido a uma revisão… não CLITIC.PRON have-INF. proceed-PART. to a revision… not se teve em conta a evolução da CLITIC.PRON. have-PRET: 3P.S in calculation the evolution of the própria modalidade proper mode [Diário de Notícias, 19970427] ‘Madail was clear when he thought back that, despite of having proceeded to a revision…the evolution of the proper mode had not been taken into account’

Example (108) displays an overt trajector in the adverbial clause (ele). Further, there is an explicit reference between the subject pronoun and other ways in which this participant is expressed: lhe, o pai. On the other hand, example (109) is much vaguer in referential linking. It is possible that there is a referential relation between Madail and se. But this relation is not overtly spelled out. An alternative way to interpret se is that Madail was not involved in the revision, or that the revision was conducted by Madail and other participants. In sum, se blurs the issue of referentiality. Thus, the speaker chooses to point out that the revision was done and to omit the mention of the participants of this process. The difference between a focused and a defocused trajector is illustrated in Figure 3:5:

The Clitic Pronoun Se: A Generic Trajector

101

Figure 3:5. Focused and defocused trajector.

tr

lm

tr

a) Focused trajector

lm

b) Defocused trajector

As indicated in Figure 3:5, the focused trajector has a higher degree of prominence than the generic trajector. This difference is illustrated by the bold line of the circle representing the trajector. The higher degree of prominence further entails that the trajector is the focus of attention in the described event, while the landmark has a secondary degree of prominence. The use of a generic trajector (represented by the triangle) does not change the asymmetric relation between the trajector and the landmark. Nonetheless, the weakened prominence of the generic trajector has the effect of increasing the relative prominence of the landmark. This being so, the focus displacement effectuated by verbal agreement with a plural NP constitutes an extension in the defocalization process of the generic trajector. The verbal inflection has the effect of conferring a higher degree of prominence to a specific participant in the predicative relation. Three different situations are at hand: 110. É evidente que para se saber clear that to CLITIC.PRON. know-INF. a Be-PRES: 3P.S a realidade do País tem que haver that exist-INF. the reality of the country have-PRES: 3P.S uma observação sobre a mesma an observation over the same [Diário de Notícias, 19970426] ‘It is obvious that in order to know the reality of the country one has to observe it’ 111. Eu penso que that I think-PRES: 1P.S coisas é o the things be-PRES: 3P.S [Viseu Diário-N0110-1]

para to dia day

se CLITIC.PRON.

fundamental fundamental

resolver as solve-INF. the

102

Chapter Three ‘I believe that this is the crucial day to get things solved’ 112. Diamantino Lopes apelou à necessidade de of Diamantino Lopes appeal-PRET: 3P.S to the need ‘passar por cima de querelas partidárias’ para Through over of disputes party to pass-INF. se resolverem os verdadeiros problemas CLITIC.PRON. solve-INF: 3P.P the real-PLUR. problems [Diário de Aveiro-N0368-1] ‘Diamantino Lopes appealed to the need to put aside inter-party disagreements in order to solve the real problems’

Examples (110-112) illustrate how the verbal inflection confers prominence to different participants in the described event. The first example (110) is neutral regarding verbal inflection. That is to say, the verbal inflection agrees both with the NP and with se. Consequently, (110) depends highly on other contextual factors for a proper interpretation. In the following case (111), the verbal inflection agrees with se. Therefore, a generic interpretation of this pronoun is most likely. Finally, example (112) provides a case of verbal agreement with the plural NP. Although it is highly unlikely that the problems will be solved without some input of energy, the notion of the energy source is minimal. Figure 3:6 provides a graphic illustration of the verbal inflection in (110-112): Figure 3:6. Verbal agreement with different participants.

se saber

ø

a realidade

se resolver

as coisas

se resolverem

os verdadeiros problemas

Consequently, the verbal agreement has the effect of creating different kinds of conceptualizations–with an active generic trajector to a higher or lesser degree. The structure with neutral verbal inflection (ø) gives evidence of a more prominent trajector than the structure that agrees with the plural NP. Obviously, the clitic se gains the highest degree of prominence in the structure where the verbal inflection does not agree with the plural NP. Therefore–and once again–the existence of the different structures in Figure 3:6 illustrates the possibility to choose a linguistic expression that matches the speaker and his communicative purposes. If the speaker feels

The Clitic Pronoun Se: A Generic Trajector

103

the need to avoid focussing on a specific participant in the event, he/she may choose a structure with the clitic se. Moreover, if the defocalization of the energy source is not sufficient, he/she may dislocate the focus from the energy source to the passive participant in the event. Obviously, this is made by means of verbal agreement. Bearing general human experience about how events and processes are brought about, however, the question is if the notion of a generic trajector totally disappears in the structure that agrees with a plural NP. The question is rather subtle and it is not easy to find a simple answer. As mentioned earlier, the problem is related to two related topics: language as a means for expressing conceptual content and human experience about how the world is shaped. The position taken in the present analysis is to admit the importance of both these factors. The common denominator for these topics is the relation between a conceptualizer and the linguistic expression, but from two different perspectives: the speaker’s and the listener’s. Obviously, the speaker may construe a linguistic expression in accordance with the conceptual content that he/she wishes to convey. The verbal agreement with a plural NP is a prime example of this phenomenon. But, the listener may interpret the message in accordance with his/her experience of events in the world and conceive of se as a generic trajector in spite of the plural inflection. Further, the occurrence of structures like example (111), that is, cases that in traditional grammar are said to be avoided (cf. Chapter 1), corroborates the impression that se may symbolize a generic trajector–even if it is a weakened one. This tendency is illustrated in Figure 3:7: Figure 3:7. Generic trajector with and without plural inflection.

tr

lm

a) [-] plural inflection

tr

lm

b) [+] Plural inflection

As the notion of a generic trajector decreases, the focus displacement confers a higher degree of prominence to the logical object (the landmark).

104

Chapter Three

In Figure 3:7b, this is illustrated by the dotted circle surrounding the generic trajector. In contrast, Figure 3:7a is identical to Figure 3:5b. In structures without plural inflection on the verb, the generic trajector has a higher degree of prominence than the relational landmark. This being so, the verbal inflection is another factor to take into consideration in the reading of adverbial structures with the clitic pronoun se. The following section will deal with different adverbial structures with the clitic pronoun se. These structures will illustrate how the verbal inflection and the position of the participants cooperate in creating different conceptualizations.

Different Conceptualizations The structures analysed in this section provide evidence for a continuum in the conceptualization of the clitic pronoun se as a generic trajector. From the overt manifestation of a generic trajector, the notion of this participant will gradually fade away. The structures that come into play are exposed in 113a-d: 113.

A) PREPOSITION + SE + INFINITIVE: 3P.S + PLURAL NP (NON-ANIMATE) B) PREPOSITION + SE + INFINITIVE: 3P.S + SINGULAR NP (NON-ANIMATE) C) PREPOSITION + SE + INFINITIVE

: 3P.P + PLURAL NP (NON-ANIMATE)

D) PREPOSITION + SN PLURAL (NON-ANIMATE) +SE + INFINITIVE

: 3P.P

The structures displayed above represent different conceptualizations of the clitic pronoun se in Portuguese adverbial clauses. Structure (113a) is an obvious example of se as a generic trajector. First, this participant occupies the subject position. Second, the verbal inflection does not agree with the plural NP. Structure (113b) is neutral regarding verbal agreement. Thus, the interpretation of se depends on other contextual factors, for example: verb type and type of NP. In structure (113c), the plural inflection of the verb entails a focus displacement. However, the subject position of se makes a generic reading possible. Finally, the generic reading is hard to maintain in structure (113d). Both the plural inflection and the subject position of the NP deviates the focus from an external energy source. In what follows, natural examples will be studied in order to examine the way in which se, the verbal inflection and the internal position of the participants influence the conceptualization of events. Special attention will

The Clitic Pronoun Se: A Generic Trajector

105

be given to the contextual participants–and possible referents to se. The first example represents structure (113a): 114. MM: Estamos na presidência Portuguesa da União of the Union MM: Be-PRES: 1P.S in the presidency Portuguese Europeia e, penso que esta é that this be-PRES: 3P.S European and, think-PRES: 1P.S a altura ideal para se remodelar os CLITIC.PRON. improve-INF. the the time ideal to programas da Juventude para a Europa programs of the youth to the Europe [Diário de Aveiro-N2063-1] ‘We hold the presidency of the European Union and I believe that this is the ideal moment to improve the Youth programs of the European Union’

Example (114) displays three possible referents to se: the speaker (MM, penso), we (estamos) and the European Union. And, in fact, it is not easy to verify who se refers to. On the one hand, it is obvious that the speaker shares some interest in the improvement of the Youth program, given that he says that this is the moment for doing so. Moreover, it is possible that se refers to ‘we’ (estamos). The fact that Portugal holds the presidency of the European Union may strengthen their impact on the improvement of Youth programs. Nonetheless, it is not likely that the issue is restricted only to the Portuguese people. This being so, the use of se seems to include all the members of the European Union. In sum, se decreases the role of the speaker in the described event. The subsequent example (115) displays the structure with se in subject position and neutral inflection on the verb: 115. “O início da Conferência Inter- Governamental (CIG) “The opening of the Conference Inter- Governmental (CIG) é uma prioridade de todos os ‘Quinze’ Estadosa priority of all the ‘Fifteen’ Statebe-PRES: 3P.S membros, porque temos de racionalizar o of rationalize the members, because have-PRES: 1P.P funcionamento das instituições europeias antes de function of the institutions European before of se proceder ao alargamento da União CLITIC.PRON. proceed-INF. to the enlargement of the Union Europeia European [Diário de Leiria-N1125-1] ‘The opening of the Inter-Governmental Conference is a priority for all State-Members of the European Union, because we need to rationalize the

106

Chapter Three function of all European institutions before proceeding to the enlargement of the European Union.’

It is noteworthy that example (115) displays a metonymic referential relation. The speaker begins by referring to the fifteen state-members. Thereafter, he/she uses the verb temos (‘we have’) in order to refer to the fifteen state-members. This certainly reflects a personal involvement in the matter. That is, the speaker gives evidence of a personal interest to participate in the process of rationalizing the functions of the European institutions. Finally, however, the insertion of the clitic se withdraws this personal involvement by defocusing the actual participants of the process. In the following example (116), the structure with plural inflection creates a focus displacement. Even so, it is interesting to note that the verb (desenvolverem) implies an external energy source. Further, the final position of the NP makes the notion of a generic trajector possible. The way in which the structure is interpreted depends on the tension between the verbal inflection, on the internal position of the participants and on the described process: 116. A Associação Académica de Aveiro vai estar The Association Academic of Aveiro go-PRES: 3P.S be-INF. este fim-de-semana numa reunião com as suas congéneres this weekend in one meeting with the their counterparts associativas e vamos procurar que junto dos together of the associative and go-PRES: 1P.P search-INF. that nossos colegas possa existir alguma exist-INF. some our colleagues can-PRES.SUBJ: 3P.S coordenação para se desenvolverem iniciativas CLITIC.PRON. develop-INF: 3P.P initiatives coordination to [Diário de Aveiro-N0778-1] ‘The Aveiro Academic Association will be meeting their associated counterparts this weekend and, together with our colleagues, we will try to find some coordination in order to develop initiatives.’

There are four potential referents to se in the example above: Aveiro Academic Association, the associated counterparts (also expressed by our colleagues), the total of these potential referents (vamos) and the passive participant. Moreover, the inflected infinitive designates an action: the one of developing. The context implies that this action will be accomplished by the total of the referents: vamos procurar junto dos nossos colegas (‘we will try together with our colleagues). However, the use of se and the plural inflection on the infinitive draw the attention from this fact. The event is

The Clitic Pronoun Se: A Generic Trajector

107

presented without any reference to the participation of an external energy source. Alternatively, this energy source is se. The final structure to be analysed represents indeed the final step in the process of downgrading the notion of an external force. In effect, the notion of a generic trajector seems to vanish completely: 117. Da mesma opinião é Maria Afonso, sua his Of the same opinion be-PRES: 3P.S Maria Afonso, esposa que afirmou haver ‘mais espaço para os the wife that affirm-PRET: 3P.S be-INF. ‘more space to carros não se atrapalharem antes de CLITIC.PRON. interfere-INF: 3P.P before of cars not porem o lancil na estrada the curb in the road put-INF: 3P.P [Diario de Leiria-N0765-2] ‘Maria Afonso, his wife, is of the same opinion. She stated that there was more space so that the cars would not cause an obstruction before they put the curb on the road.’

In fact, this structure has the effect of topicalizing the plural NP. That is, the speaker is only concerned with the cars and with the consequences that the curb has for the cars (and for the traffic in general). Consequently, the trace of a generic trajector is minimal. Obviously, the initial position of the NP and the plural inflection are highly relevant factors in the topicalization process. Examples (114-117) illustrate how different structures with se make the notion of an explicit agent disappear gradually in Portuguese infinitive Adverbial clauses. Evidently, the insertion of se instead of a specific and explicit trajector causes a defocalization–the issue of referentiality is blurred. Further, the verbal inflection and the internal position of se and the plural NP contribute to different ways of conceptualizing the adverbial process. This being so, there seems to be a pattern regarding the prominence of the pronoun se as a generic trajector in Portuguese infinitive adverbial clauses:

108

Chapter Three

Figure 3:8. The prominence of se as a generic trajector. HIGH PREPOSITION + SE + INFINITIVE: 3P.S + PLURAL NON-ANIMATE NP PREPOSITION + SE + INFINITIVE: 3P.S + SINGULAR NON-ANIMATE NP PREPOSITION + SE + INFINITIVE: 3P.P + PLURAL NON-ANIMATE NP LOW PROMINENCE

To begin with, the structure PREPOSITION + SE + INFINITIVE: 3P.P + NP displays a rather weak generic trajector. The verbal inflection has the effect of putting focus on the non-animate NP. Further, the prominence of se as a generic trajector increases in the structure PREPOSITION + SE + INFINITIVE: 3P.S + SINGULAR NON-ANIMATE NP. That is, the neutrality of the verbal inflection (ø) makes the reading of this structure utterly dependent on other contextual factors. Finally, the role of se in the structure PREPOSITION + SE + INFINITIVE: 3P.S + PLURAL NON-ANIMATE NP is certainly equated with the one of a generic trajector, given the lack of plural inflection on the infinitive. This being so, the structures analysed above share the feature of being possible manifestations of se expressing a generic trajector, even though the notion of this participant may be salient to a higher or lesser degree. The analysis has also shown that the use of se as a generic trajector is frequently a communicative strategy in order to diminish the energy source of a given process. This strategy may depend on a communicative need to highlight the process and not the participants of the same. But, it may also be a strategy for diminishing the responsibility of the energy source. PLURAL NON-ANIMATE

Conclusion The present chapter has questioned the reason for interpreting the clitic pronoun se as a passive marker in some structures and as an impersonal subject in others. Thus, it was suggested that se symbolizes an external energy source, i.e., a generic trajector, not only in the impersonal readings but also in the cases that traditionally have been labelled as passives. The reason for this is that contextual features, apart from the verbal inflection, influence the propensity to interpret se as an external energy source.

The Clitic Pronoun Se: A Generic Trajector

109

Reference was made to Garcia (1975) and the short-circuit mechanism of se. In the impersonal reading, se impedes the insertion of a definite, well defined initiator of the described event. In the reflexive reading, the insertion of se has the effect of impeding the notion of other participants than the grammatical subject. That is, the subject and the direct object correspond to the same participant. The tension between structural and phenomenological interpretations of se was discussed. From the structural perspective, the use of se represents a defocalization effect. Instead of putting focus on an explicit and defined trajector of the adverbial event, this participant is conceived of as somebody or someone. Further, the plural infinitive inflection corresponds to a change of focus, i.e., a focus displacement. The verbal inflection puts the focus on the non-animate plural NP to the extent that the notion of an external energy force diminishes. From the phenomenological perspective, two features were primarily considered: the type of verb and the type of NP. If the verb describes a process that requires external force in order to be accomplished, the propensity for a generic reading of se increases. If the NP is conceptualized as a participant capable of introducing the required energy, on the other hand, this propensity decreases. Thus, the way in which the clitic pronoun se is conceptualized depends on both structural and phenomenological aspects. In the analysis, the infinitive adverbial clauses with se were first considered in relation to the transitivity parameters elaborated by Hopper and Thompson (1980). The analysis revealed that these structures exhibit features compatible with the notion of transitivity. Also, the prepositions introducing the adverbial clauses share inherent features that increase the degree of transitivity in the described event. The preposition para (‘to’) exhibits volitional features and até (until) has a punctual aspect. Finally, the prepositional expression depois de tends to designate adverbial processes that are understood as being concluded. The problem of analysing the number of participants in structures with se and a plural NP was related to the vagueness of this pronoun. Following Quesada (1998), however, the analysis verified that structural restrictions explain why the plural NP cannot be the subject of “passive sentences”: (1) it is impossible to have subject agreement with an animate plural NP that is specific, (2) the plural NP does not control the event and (3) it is impossible to insert an agent in the event. Thus, the conclusion is that se symbolizes a participant even in structures with a plural NP. The reflexive meaning of se was discussed in relation to the existence of an obvious referent. The analysis focused on the existence of a defined and definite animate participant in the context. The existence of a participant

110

Chapter Three

with these features increases the tendency to make a reflexive reading of se. Moreover, the reflexive reading is highly possible in metonymic referential relations. That is, institutions, companies and other localities that are connected with human activity tend to give a reflexive reading of the structures with se and a non-animate NP. The concluding part of the analysis dealt with the structural aspects of infinitive adverbial clauses with se. Thus, priority was given to the internal position of se and NP and to the effect of the verbal inflection. It was shown that the initial position, i.e., the subject position prototypically equates with the position of the energy source. Consequently, the tendency to conceptualize se as a generic trajector tends to increase when it occupies the initial position. On the other hand, the reversed situation creates a focus displacement. A non-animate plural NP in subject position makes the notion of an external energy source disappear–the focus is put on the NP. Likewise, the verbal inflection has influence on the degree of prominence of se as a generic trajector. If the verbal inflection does not agree with the plural NP, se gains a higher degree of prominence. In the opposite case, the plural inflection contributes to diminish the notion of se as a generic trajector. Finally, the neutral cases are highly context dependent, given that the null inflection does not direct focus to either participant. In conclusion, the way in which se is conceptualized depends on a wide array of contextual factors. Therefore, the statement that it is a passive marker in some structures and an impersonal subject in others is problematic. Instead of creating absolute boundaries between the passive reading and the impersonal reading of se, it is necessary to extend the analysis to not only include the verbal inflection. In fact, the analysis of the present chapter revealed the possibility to consider se as a generic trajector also in cases with plural inflection. This being so, the meaning of se may be a gradual phenomenon. That is, depending on the context, the notion of this participant as a generic trajector may be more or less prominent. In accordance with the Cognitive Grammar perspective of language and linguistic analysis, thus, the analysis offered in this chapter exemplifies the difficulty to create clear-cut boundaries between linguistic categories.

CHAPTER FOUR FINITE ADVERBIAL CLAUSES: ICONICITY, SUBJECTIFICATION AND MENTAL SPACES

Introduction The present chapter will put focus on the variation between finite and infinitive Portuguese adverbial clauses with an elaborated trajector. The primary concern is to understand this variation and to find a conceptual explanation for it. In doing so, I will relate the alternation between finite and infinitive clauses to phenomena like iconicity, subjectification and mental spaces. The hypothesis is put forward that formal differences between different adverbial structures reflect conceptual differences. The formal differences are illustrated in examples (118-119): 118. ela que venha de manhã, porque morning, CAUSE:CONJ. she that come-PRES.SUBJ: 3P.S of depois pode não me apanhar cá… me catch-INF. here… after be able to-PRES: 3P.S not eu ando agora numa vida ambulante now in a life mowing I walk-PRES: 1P.S [Pfu: 159] ‘she has to come in the morning, because she will not catch me here later…I am moving around a lot nowadays’ 119. eu acho que é mais por, that be-PRES:3P.S more through, I think-PRES: 1P.S por serem bonitos, por serem CAUSE:PREP be-INF: 3P.P beautiful-PLUR. CAUSE:PREP. be-INF: 3P.P diferentes different-PLUR. [Pfa: saber vender] ‘I think that it is mostly because they are beautiful, because they are different’

112

Chapter Four

One formal difference pertains to the linguistic unit(s) that introduces the adverbial clause. As can be verified in (119), the infinitive clause is introduced by a preposition, forming the structure PREPOSITION + (NP) + INFINITIVE, in which the use of the nominal phrase is optional. On the other hand, example (118) illustrates that finite clauses are introduced by a conjunction. In the causal clauses offered above, the difference is related to the use of the preposition por in the infinitive clause and the conjunction porque in the finite clause. Figure 4:1, below, furnishes an illustration of linguistic unit(s) introducing Portuguese adverbial clauses: Figure 4:1. Linguistic units introducing finite and infinitive adverbial clauses. FINITE CLAUSES:

INFINITIVE CLAUSES:

CONJUNCTION: quando, porque, embora

PREPOSITION: por, para, sem, até

ADVERB + CONJUNCTION: antes que, depois que

ADVERB + PREPOSITION: antes de, depois de

PREPOSITION + CONJUNCTION: para que, sem que, até que

PREPOSITION + DETERMINANT: ao

The general pattern holds that the linguistic units that introduce a finite clause display a formally more elaborated structure, either consisting of a subordinating conjunction, the structure ADVERB + CONJUNCTION or the structure PREPOSITION + CONJUNCTION. This being so, the introducing unit creates a boundary between the main clause and the adverbial clause. Infinitive adverbial clauses are introduced by a formally weaker structure: a preposition, the structure ADVERB + PREPOSITION or the structure PREPOSITION + DETERMINANT. As will be shown in the subsequent analysis, these linguistic units tend to create a stronger link between the main clause event and the adverbial clause event, and they also present a higher degree of dependency of the latter. Another formal difference is related to verbal inflection. On the one hand, it is interesting to note that the inflection of the Portuguese infinitive makes it apt for not only co-referential contexts, but also non co-referential contexts. For example, examples (118-119) display a reversed referential pattern. In (118), the trajector of the finite adverbial clause (expressed by

Finite Adverbial Clauses: Iconicity, Subjectification and Mental Spaces

113

pode) is co-referential with the main clause trajector (ela). In the following example (119), in contrast, there is no co-referentiality between the main clause trajector and the adverbial trajector. This feature indicates that the question of referentiality does not determine the variation between finite and infinitive adverbial clauses in Portuguese (cf. Chapter 2). Therefore, the most striking difference between finite and infinitive clauses is related to the temporal dimension of the described event. Whereas the inflected infinitive only provides information about person and number, finite verb forms add this dimension to the described event. This difference is illustrated in Figure 4:2: Figure 4:2. Temporal dimension in finite and infinitive adverbial clauses.

a) [+] temporal dimension

b) [-] temporal dimension

Thus, the tense inflection of the finite verb forms has the effect of profiling the temporal aspect of the described event. In Figure 4:2a, the profiling is illustrated by the bold line on the temporal axis (the arrow). The lack of tense in the infinitive inflection, on the other hand, indicates that the described event is viewed in a holistic fashion. That is, instead of grounding the event in a specific time frame, as the finite clause does, the infinitive clause describes the event in a more general, or generic, manner. In Figure 4:2, this is illustrated by the bold line of the surrounding square. In this sense, the infinitive does not describe the instantiation of an event, but more accurately a type of event (cf. Silva 2005, 2008). Examples (118-119) offer a partial demonstration of the formal and semantic differences between finite and infinitive adverbial clauses in Portuguese, designating tense–or lack of tense–in the described adverbial event. Moreover, there is another highly relevant feature, namely the mood alternation. This difference is shown in the following examples:

114

Chapter Four 120. a mãe é que deve educar ought-PRES: 3P.S educate-INF. the mother be-PRES: 3P.S that os filhos da mesma maneira e preparar os the children of the same way and train-INF. the rapazes também para que eles não sejam PURPOSE:CONJ. they not be-PRES.SUBJ: 3P.P boys also os senhores the masters [Pfu: 303] ‘It is the mother’s duty to educate her children in the same spirit so that the boys do not rule the family’ 121. é pequenita! a horta é pequenita, small! the garden be-PRES: 3P.S small, be-PRES: 3P.S é enfim, é só para a ultimately be-PRES: 3P.S only PURPOSE:PREP. the be-PRES: 3P.S gente se entreter people CLITIC.PRON. entertain-INF. [Pfu: 225] ‘It is small! The garden is small, it is only really there to make people happy’

Apart from finite adverbial clauses introduced by quando (‘when’) and porque (‘because’), which generally are followed by the indicative mood, the most common pattern is to have the verb in the subjunctive mood. The analysis will show that the mood alternation is highly relevant in the search for a conceptual explanation regarding infinitive and subjunctive adverbial clauses in Portuguese. This being so, it is pertinent to point out the mood alternation of the finite clauses. In general, there is a tendency for the subjunctive mood to occur after embora, sem que, para que, antes que, até que and depois que, while quando and porque are followed by the indicative mood. However, the mood choice is sensitive to the time frame of the adverbial event. This is particularly relevant in relation to temporal adverbial clauses introduced by quando, depois que and até que. If these clauses describe an event in the past, the most common mood choice is the indicative. But, when the described event has a projection into the future, the subjunctive mood is chosen. In these cases, quando and depois que are followed by the future subjunctive mood, while até que still takes the present subjunctive mood. Adverbial clauses containing the subjunctive mood in the present maintain the same tense in cases of future projection. From a cognitive linguistic approach to language and linguistic analysis, the formal differences between finite and infinitive adverbial clauses

Finite Adverbial Clauses: Iconicity, Subjectification and Mental Spaces

115

pointed out above entail conceptual differences. The existence of two different adverbial structures is not accidental. Rather, there must be some communicative differences between them. Bearing theories of linguistic iconicity in mind, the initial claim in what follows is that the formal–or structural–differences between finite and infinitive adverbial clauses in Portuguese reflect conceptual differences. Consequently, I will show that the formally more complex structure (the finite clause) is conceptually more complex than the infinitive clause. Further, I will connect this higher degree of conceptual complexity to the notion of subjectification. That is, the finite clauses designate the conceptualizer’s perspective and/or propositional attitude towards the event described in the adverbial clause. This phenomenon will be attested in the dislocation from one mental space, the mental space of origin, to another mental space that is related to the conceptualizer’s mental reasoning about the adverbial event. Thus, finite adverbial clauses not only designate an adverbial relation with the main clause event, but they also designate the conceptualizer’s reasoning about the adverbial event. Obviously, notions like iconicity, subjectification and mental spaces need to be clarified in more detail. Thus, before the analysis, I will try to furnish an adequate explanation of these concepts.

Iconicity The claim that language is iconic in character is not uncontroversial. For example, there is a strong tradition within the generative approach to linguistic analysis to distinguish between two different levels of linguistic representation: deep structure and surface structure. Subsequently, the semantic meaning of a linguistic expression is not found in the actual linguistic expression, but in a deeper representation of the expression. Seen from this perspective, semantic analysis is not quite that simple. The meaning of any linguistic expression is hidden in the deep structure and not in the actual linguistic expression (the surface structure). Further, the postulation of an autonomous syntax implies that meaning is related to lexicon and not to syntax (cf. Chomsky 1995, 2002). 18 It is also common 18 In fact, the notion of different levels of representation is still dominant in the generative approach to linguistic analysis: “We assume that the design of language provides a variety of symbolic systems (levels of representation)…, including the level of Phonetic Form and the level of Logical Form, specifying aspects of sound and meaning, respectively, insofar as they are linguistically determined. Another is the level of D-Structure, which relates the computational system and the lexicon” (Chomsky 1995).

116

Chapter Four

knowledge that any given language consists of signs that do not necessarily resemble the referent they designate. The relation between the sign and its referent is in most cases arbitrary. In view of this fact, it may seem contradictory to claim that languages are iconic in character. However, Haiman (1980: 515) solves this apparent contradiction by making a distinction between imagic and diagrammatic iconicity. An iconic image displays a certain resemblance with its referent; some prototypical cases would be a statue, a painting, a photograph or–in the case of human languages–onomatopoeic words. An iconic diagram, in contrast, is a systematic arrangement of signs that do not resemble their referent, but whose relationship with each other reflects the relationships of their referents. Hence, there is a relation between human language and diagrams. Linguistic symbols do not need to share any resemblance with the concept they represent. Nevertheless, the way human speakers organize linguistic symbols (the grammar) is iconic in nature: every symbol has a counterpart in the concept that it represents (cf. Haiman 1985). 19 In sum, grammar– including syntax– reflects conceptual structure. The same author (Haiman 1985) presents two hypotheses related to the notion of diagrammatic iconicity. The first one, called the isomorphism hypothesis, states a one-to-one correspondence between a linguistic sign and its referent. The second one, known as the motivation hypothesis, states that the formal differences between two contrasting forms reflect their semantic differences: The Isomorphism Hypothesis Different forms will always entail a difference in communicative function. Conversely, recurrent identity of form between different grammatical categories will always reflect some perceived similarity in communicative function. The Motivation Hypothesis Given two minimally contrasting forms with closely related meaning, the difference in their meaning will correspond to the difference in their form (Haiman, 1985: 19-20).

The isomorphism hypothesis rejects the idea of full synonymy. Even though many linguistic signs share some semantic features, they will always entail some differences in communicative function. One example of this is found in the occurrence of different words designing a human being with 19

Hopper and Thompson (1985), Kortmann (1997), Smith (2002) and Wilcox (2004) furnish additional analyses in the field of linguistic iconicity.

Finite Adverbial Clauses: Iconicity, Subjectification and Mental Spaces

117

the feature [+MALE / +YOUNG] in English. Certainly, the words boy, lad and guy signal communicative differences with respect to how we perceive the referent. It is also noteworthy that the second part of the isomorphism hypothesis may account for cases of polysemy. In cases where a form has more than one meaning, these meanings are related. In addition, the motivation hypothesis establishes a relation between linguistic forms and diagrams. If each part of a diagram represents a certain part of a concept, it is likely that each part of a linguistic form represents a certain part of a conceptualization. Lakoff and Johnson (1980) capture the notion of iconicity with the metaphor [LINGUISTIC UNITS ARE CONTAINERS]. The metaphor furnishes the basis for the two expectations: more of form is more content and closeness is strength of effect. The first expectation represents an iconic relation in which the addition of linguistic units correlates with more content. The second expectation emphasizes the relation between formal and conceptual distance. The authors illustrate these expectations with the following examples: 122. 123. 124. 125.

He ran He ran and ran and ran and ran Mary doesn’t think he will leave until tomorrow Mary thinks he won’t leave until tomorrow

The conceptual difference between examples (122) and (123) is rather obvious. The first sentence (He ran) indicates that the subject was involved in the action of running, but does not provide any specification about the quantity or length of the activity. The reduplication of the verb in example (123), on the other hand, implies an extended activity. It conveys the idea that the subject ran considerably more than the subject in (122). In example (124), the negative transportation places the negation further away from the predicate that it logically negates (leave) than in (125). This negative transportation has the effect of creating a weaker kind of negation (Lakoff and Johnson, 1980: 129). Another iconic relation is confirmed in the notion that a greater formal distance between cause and effect often reflects a greater conceptual distance between cause and effect. The following examples are representative illustrations: 126. 127. 128. 129.

Sam killed Harry Sam made Harry die Sam caused Harry to die Sam brought it about that Harry died

118

Chapter Four

In these examples, a successively greater formal distance between the causal predicate and the one of effect designates a more indirect (or diffuse) causation. In (126), the juxtaposition SN + V + SN designates a direct causal relation, while the greater distance between these linguistic units in (129), implies an indirect causal relation. Langacker (1987) comments thus on this phenomenon: “Such paradigms indicate that greater phonological distance between a predication of cause and one of effect correlates with greater conceptual distance in the domains of cause-effect linkages” (Langacker 1987: 181). Examples (122-129) illustrate the iconic relation between linguistic form and diagrams by showing that there is a certain relation between formal complexity and distance, on the one hand, and conceptual complexity and distance, on the other hand. In the same way as every part of a diagram represents a concept, the way in which we organize linguistic expressions reflect a certain conceptualization. A formally more complex structure tends to be conceptually more complex than a structure with a lesser degree of formal complexity. Further, formal distance tends to correlate with conceptual distance. In sum, the notion of iconicity rests on the assumption that linguistic structures can be understood as diagrams. Thus, the isomorphism hypothesis posits a one-to-one correspondence between form and meaning, and the motivation hypothesis defends the idea that formal differences reflect conceptual ones. If we relate these hypotheses to infinitive and finite adverbial clauses, the result is that these clauses cannot be regarded as synonymous. First, one form corresponds to one meaning. Secondly, the formal differences must reflect certain conceptual differences.

Subjectification Subjectification as a linguistic phenomenon has primarily been analysed from two different–and in some cases competing–perspectives in cognitive and functional linguistics. Traugott (1989, 1995, 1996, 1999) and Langacker 1990, 1999, 2000, 2003, 2006) are probably the two most important contributors to the understanding of this phenomenon. They also differ in fundamental aspects regarding how subjectification is to be understood, although they share some general insights as to how to comprehend it. Traugott focuses on the diachronic evolution of subjectification and defines it as “the historical pragmatic-semantic process whereby meanings become increasingly based on the speaker’s subjective belief state or attitude towards what is said…” (Traugott 1996: 185). First and foremost,

Finite Adverbial Clauses: Iconicity, Subjectification and Mental Spaces

119

the notion of subjectification is equated with the speaker’s subjective judgement and propositional attitude towards the proposition. Traugott’s framework emphasizes three tendencies involved in the historical path towards subjectification: Tendency 1: Meanings based on the external described situation > meanings based on the internal (evaluative, perceptual, cognitive) described situation. Tendency 2: Meanings based on the external or internal described situation > meanings based on the textual and meta-linguistic situation. Tendency 3: Meanings tend to become increasingly based on the speaker’s subjective belief/state toward the proposition (cf. Traugott, 1989: 34-35)

An example of Tendency 1 is the shift in the meaning of boor (farmer > crude person). Another example is illustrated by the Old English felan that initially only had the meaning “touch” and did not acquire a perceptual meaning until late Old English. Both cases imply a shift from a concrete domain to an abstract domain. Tendency 2 includes the development of morphological and lexical forms into connectives coding the textual situation and the shift from mental-state to speech-act verb meaning, i.e. observe (perceive > state that). Tendency 3 is exemplified by the shift of the temporal adverb while into a concessive, and the development of the action verb go to have a temporal meaning of immediate planned future (cf. Traugott 1989: 34-35). 20 Langacker (1990), analysing the notion of subjectification mainly from a synchronic point of view, defines it as “the realignment of some relationship from the objective axis to the subjective axis” (Langacker 1990:13). This definition relates to the asymmetry between the ground (in the subjective axis) and the object of conception (in the objective axis) as a matter of construal and perspective, dimensions that include notions like vantage point and viewing arrangement. The term ground is used for the speech event, i.e., the speaker, the hearer and the time and place of speaking. The speaker is the primary subject of the conceptualization (the primary conceptualizer). The ground is said to be subjectively construed when left “offstage” as an unprofiled facet of the speaker’s conceptualization. The object of conception is objectively construed if it is well delimited from the 20

In later works, Traugott and Dasher (2001) and Traugott (2003) introduce the notion of intersubjectification, defined as “the semasiological process whereby meanings come over time to encode or externalise implicatures regarding SP/W’s attention to the ‘self’ of AD/R” (Traugott 2003: 129-130). See also Verhagen (2005) on the matter.

120

Chapter Four

ground. A finite clause has the effect of maximizing the asymmetry between the subject and the object of conception (cf. Langacker 1990: 14-15). The relation between grounding and subjectification in finite and infinite events is shown in Figure 4:3: Figure 4:3. Grounding and subjectification.

Objective scene

Objective scene

tr

lm

g a) Peter sees the children playing

tr

lm

g b) Peter can see that Mary is tired

The fundamental difference between (a) and (b) in terms of grounding is that the finite complement in (b) bears some relationship to the ground. The finite verb creates a relation to the time and place of speaking. Further, the described event, e.g., Mary is tired, is delimited from the ground, displaying an asymmetric relation between the ground and the event. Consequently, both the main clause event and the complement event are grounded. On the other hand, the infinite progressive construction (the children playing) is not capable of creating such a relation to the ground. However, the infinitive event is grounded by the main clause. The conceptual consequences pertain to the effect of subjectification. The fact that the infinitive event is grounded by the main clause entails that the linguistic expression describes the event from the perspective of the main clause trajector. It designates what the main clause trajector is witnessing. Thus, the main clause subject equates with the conceptualizer of the event described in the objective scene. The ground is still the primary conceptualizer of the linguistic expression, but the main clause trajector is the conceptualizer of the infinitive event. In Figure 4:3b, in contrast, there is a change in perspective. Both Peter can see and Mary is tired bear a relation to the ground. As the ground describes the mental process of the main clause trajector, it is mentally scanning between these events. Thereafter,

Finite Adverbial Clauses: Iconicity, Subjectification and Mental Spaces

121

the ground reaches a conclusion regarding the mental process of the main clause trajector. By attributing a mental process to the main clause trajector, however, the conceptualizer provides evidence for his/her own inferential process. Movement in the objective scene has been replaced by the mental process, and thus, movement of the ground. 21 Accordingly, the effect of grounding seems to display a rather straightforward relation with the conceptualizer’s conception of reality. In fact, this also holds for verbs of propositional attitude. Verbs like know, believe, suppose, think and realize take a finite complement. In cases where there is a possibility to use both finite and infinitive complements, like with perception verbs such as see, hear and feel, the finite complement does not actually describe perception, but mental processes that are based on perception (cf. Langacker 1991; Vesterinen 2007). Further, the effect of grounding is related to the notion of event integration (cf. Givón 1993, 2000). The basic idea of event integration is that finite subordinated structures are more loosely connected to the main clause event, describing an event that has its own temporal profile. In contrast, formally weaker structures provide evidence for a stronger relation with the main verb event. In the case of infinitives, the structure MAIN CLAUSE VERB + INFINITIVE evokes the notion of a single and complex event. Clearly, this is an effect of the atemporal feature of the infinitive. In later works, Langacker (2000, 2003, 2006) redefines subjectification as semantic bleaching or fading away. The mental scanning of the ground is there all along and emerges – or becomes more evident – when the object of conception “is no longer there to mask it” (Langacker 2006: 21). Put another way, the mental scanning of the ground is constantly present, but becomes more obvious when the described situation lacks objective movement. This is in accordance with the situation described in Figure 4:3. The objective scene lacks motion. Objective movement in (a) has been replaced by subjective movement in (b). It is the ground, as a subject of conception, which mentally scans between the main clause event and the complement event in order to reach a conclusion about the complex event. As 21

It could be argued that finite adverbial clauses provide evidence for objectification given that the feature [tense, person, mood] furnishes a more detailed description of the objective scene (cf. Achard 1998; Silva 2005). However, this feature also creates a higher degree of asymmetry between the ground and the object of conception. See Pit (1997) for a detailed examination of subjective and objective constructions. Further readings include Nuyts (2001); Pander Maat and Sanders (2000, 2001); Pander Maat and Degand (2001); De Smet and Verstraete (2006) and Verhagen (1995, 2005).

122

Chapter Four

Athanasiadou et al. (2006: 3) point out, this shift constitutes a change in the locus of activity. It is now the conceptualizer, and not the main clause subject, who is the active participant. This phenomenon is a fundamental feature of subjectification, and it involves a loss of subject control and a shift in domain from an active subject to the conceptualizer (cf. Langacker 2000). In sum, the fundamental difference between Traugott and Langacker seems to boil down to that between conceptual content and the way in which an event is construed. For Traugott, subjectification is related to the speaker’s propositional attitude towards the event. This also holds for Langacker, but the central claim is that the conceptualizer is subjectively construed. In this sense, grounding profiles the described event, and not the relation between the ground and the described event. Despite these differences both approaches will be considered in the present analysis. To begin with, it is likely that Tendency 1 and Tendency 3 in Traugott’s framework are highly relevant features in the variation between infinitive and finite adverbial clauses. This will be confirmed in a higher degree of internalization and personal beliefs in the finite clauses than in the infinitive ones. In the second place, the claim made in this chapter is that finite adverbial clauses in many cases display a displacement from one mental space to another. In this sense, the conceptualizer shows evidence of personal inferences about the adverbial event. This subsumes that a subjectively construed conceptualizer is mentally scanning between two events. This is understood to be Langacker’s view of subjectification.

Mental Spaces The theory of Mental Spaces (Fauconnier 1994) departs from the hypothesis that different domains are construed in our mental representation when we are engaged in activities such as thought, communication and other cognitive processes. In the ongoing discourse, the topic of our concern, linguistic expressions create domains, and connections between these domains, in our mental representation of the subject matter. Fauconnier (1994) explains this phenomenon as follows: …when we engage in any form of thought, typically meditated by language (for example, conversation, poetry, reading, storytelling), domains are set up, structured and connected. The process is local: A multitude of such domains–mental spaces–are constructed for any stretch of thought, and language (grammar and lexicon) is a powerful means (but not the only one) of specifying or retrieving key aspects of this cognitive construction (Fauconnier 1994: xxvii)

Finite Adverbial Clauses: Iconicity, Subjectification and Mental Spaces

123

The fundamental claim is that one domain, i.e., mental space, has a counterpart in some other domain. The connection between these mental spaces is completed by a pragmatic function called the identification principle. Thus, reference to one object may be obtained by a pragmatic function that creates a link to it in terms of another object: Identification Principle If two objects (in the most general sense), a and b are linked by a pragmatic function F (b = F (a)), a description of a, da, may be used to identify its counterpart b (Fauconnier 1994: 6).

In metonymic cases, for example, the pragmatic function creates a trigger that corresponds to the target of the linguistic expression. That is, a correspondence is created by the identification principle so that there is a pragmatic link between two different linguistic expressions. Below, this is illustrated by a typical case in which an employee at a restaurant refers to a client at the same restaurant: 130. The cheese omelette has made a complaint about the food

Obviously, the cheese omelette is not the one making a complaint about the food. Instead, it is the speaker who creates a metonymic link between the client, on the one hand, and the food that he/she ordered, on the other. Hence, the linguistic expression establishes a pragmatic relation between the trigger (the omelette) and the target (the client). In this sense, the dislocation from one mental space to another is accomplished by a pragmatic function that connects the food with the client (F in the Identification Principle). The dislocation from the mental space of origin to a new mental space is illustrated in Figure 4:4:

124

Chapter Four

Figure 4:4. Pragmatic function and mental spaces. F (connector)

a a = trigger (the omelette).

b b = target (the client).

In Figure 4:5, the trigger is equated with the omelette while the target corresponds to the client. Subsequently, the actual context in which example (130) is uttered, i.e., the restaurant, creates a relation between the omelette and the client. The target is equated with the trigger. In correspondence with the Identification Principle, the target (b) is the counterpart of the trigger (a) in a metonymic relation where the description of a (da) may be used to identify b. The linguistic expressions that guide the interlocutors from one original mental space to another space are referred to as space builders (cf. Fauconnier 1994). Some typical space builders are found in temporal expressions guiding the discourse participants from the present to another temporal space. Other examples are found in adverbs like probably, maybe etc. that guide the discourse participants to mental spaces that have a weaker link to what is known about the course of events in the world. Also, conditionals like if I had…and other similar expressions function as space builders by creating mental spaces that guide the participants in the actual discourse. The common denominator for these linguistic expressions is that they guide the interlocutors from one mental space, the mental space of origin (also called the parent space), to another mental space. In the cases mentioned above, the mental space of origin is related to the reality. Examples (131-132) demonstrate this dislocation from reality to another mental space: 131. In 1969, the gray-haired lady was blond 132. If I where rich, my VW would be a Ferrari

In both examples, the space builder creates a dislocation to another mental space than the parent space of origin, equated with reality. In (131), for example, the temporal expression In 1969 guides the interlocutors to

Finite Adverbial Clauses: Iconicity, Subjectification and Mental Spaces

125

another time than the one of the ongoing discourse. In the following example (132), the conditional phrase If I where rich creates a hypothetical space that differs fundamentally from the mental space of origin. Nonetheless, there is always a relation between the mental space of origin and the new mental space. If not, an expression like (131) would imply that the gray-haired lady was both gray-haired and blond in 1969. Figure 4:5 may illustrate the relation between the mental space of origin and the new mental space: Figure 4:5. The relation between mental spaces (examples 131-132).

(131) X1: The gray-haired lady

X1 Reality (now)

X2

X2: The blond lady

Temporal mental space (1969)

(132)

X1: VW

X1 Reality (not rich)

X2

X2: Ferrari

Hypothetical mental space (rich)

Thus, the space builder dislocates an object in the present (X1) from the reality to a new mental space in which the original object is identified with another object (X2), or with the same object having different features. Consequently, the gray-haired lady that belongs to present reality is identified as the blond lady in 1969. A proper interpretation would in this case be: the lady that nowadays is gray-haired was blond in 1969. The following example (132) displays a closely related but slightly different pattern. In this case, an object in reality (the VW) has become a different object in a hypothetical mental space–a Ferrari. Nonetheless, there is a common denominator in the sense that both objects belong to the same category (cars). The examples displayed above illustrate a situation in which there is a relation between the original mental space and the new mental space. That is, an object in the created mental space (X2) is identified with an object in the mental space of origin (X1). However, there are cases that contradict

126

Chapter Four

this tendency, and these cases seem to be utterly related to the semantic differences between the indicative and subjunctive mood. That is, mood alteration may indicate whether or not an object belongs to the conceptualizer’s conception of reality. Sweetser (1996: 113) offers two examples in French to explain the difference between relative clauses in the indicative and subjunctive mood: 133. Je veux qu’elle mette une robe que est belle [indicative mood] 134. Je veux qu’elle mette une robe que soit belle [subjunctive mood) ‘I want her to put on a dress that is beautiful’

According to Sweetser (1996), the mood alteration in this grammatical context is related to how the speaker conceives the described object. If the speaker has a particular dress in mind, a dress that exists in the speaker’s conception of reality, the indicative mood is chosen. The indicative mood signals that the dress exists in the speaker’s original mental space (the reality) and in a want or wish space. On the other hand, the subjunctive is chosen when the dress does not actually exist in the speaker’s reality, but only in a want or wish space. That is, the speaker has no particular dress in mind. He/she only wants it to be beautiful. Another related topic is found in Mejías-Bikandi (1996). The author formulates a mental spaces analysis of mood alteration in Spanish, and introduces the term Space Accessibility to capture the semantic distinction between the indicative and subjunctive mood. The hypothesis put forward is that the subjunctive diminishes the accessibility for an object or a presupposition in the parent space (the original space). The indicative mood, on the contrary, signals accessibility to the parent space. Mejías-Bikandi (1996: 159-160) illustrates the difference in Spanish clauses introduced by tal vez (maybe): 135. Tal vez su hijo está en la cárcel [indicative mood] 136. Tal vez su hijo esté en la cárcel [subjunctive mood] ‘Maybe his/her son is in prison’

Mejías-Bikandi (1996: 159-160) relates the difference between (135) and (136) to presupposed and not presupposed objects in the real world. Thus, the indicative in example (135) signals that the person (indicated by the possessive su (‘his/her’)) has a son. That is, the son is presupposed by the speaker and exists in his/her parent space. The subjunctive mood, in contrast, is used to indicate the contrary. By using the subjunctive mood, the speaker conveys the message that there may not be any actual son. That is,

Finite Adverbial Clauses: Iconicity, Subjectification and Mental Spaces

127

the son is not presupposed by the speaker, and therefore he has a lesser degree of accessibility to the parent space. 22 The position taken in the present analysis, however, is not in terms of this analysis. Rather, I will argue that the subjunctive mood signals a lesser degree of control over the described event, either in relation to how the conceptualizer evaluates the event in terms of reality or in relation to his/her capacity to actively manipulate the event. Therefore, examples (135-136) would be better explained by the conceptualizer’s attitude towards the proposition (‘he is in prison’) and not by the presupposition (‘he/she has a son). The indicative mood is chosen to signal a higher degree of commitment to the proposition, while the subjunctive mood signals a lesser degree of control over the described event. The proposition he is in prison does not belong to the conceptualizer´s conception of reality. In sum, the indicative clause has a lesser degree of hypothetical force than the subjunctive clause: Figure 4:6. The dislocation from parent space to a hypothetical space with a lesser degree of control (clauses introduced by tal vez). a) Indicative mood

Parent space

Hypothetical space

b) Subjunctive mood

Parent space

Hypothetical space

Obviously, clauses that are introduced by tal vez will always have some degree of hypothetical force. Nonetheless, this feature seems to be more forceful in cases where the subjunctive mood is chosen. The most appealing interpretation of the combination TAL VEZ + INDICATIVE is that the speaker is almost certain of the veracity of the proposition, and that tal vez is a communicative strategy to mitigate the proposition, i.e., the speaker is convinced that the proposition is true, but does not want to sound too 22

Mejías-Bikandi (1993, 1995) introduces the notion of Presupposition Inheritance and connects it with the semantics of the indicative and subjunctive mood. In accordance with his analysis, the main difference between the indicative and the subjunctive mood is that of presupposed and not presupposed objects in the described event.

128

Chapter Four

rigorous about it. On the other hand, the subsequent analysis will show that the subjunctive mood often is chosen in grammatical contexts of doubt. Therefore, it is not surprising that it occurs with–and is the most common mood–in clauses introduced by tal vez. In Figure 4:6, the difference in hypothetical force is indicated by the type of arrow that goes from the parent space to the hypothetical space. The dotted arrow indicates a lesser hypothetical force.

Analysis The outline of the analysis is as follows. First, I will analyse the variation between indicative and infinitive adverbial clauses. Thus, focus will be put on clause that are introduced either by por or porque (‘because’) and by ao or quando (‘when’). Thereafter, the analysis will focus on the variation between subjunctive and infinitive adverbial clauses. As pointed out in the introduction, this variation occurs mainly in clauses that are introduced by embora (apesar de), sem (que), para (que), antes (de, que), até (que) and depois (de, que). At this juncture, it is pertinent to repeat the initial hypothesis: a higher degree of morphological complexity in the finite clauses will mirror a higher degree of conceptual complexity. Thus, the difference between finite and infinitive adverbial clauses in Portuguese represents a prime example of linguistic iconicity. Further, the conceptual differences will be explained in terms of subjectification and mental spaces. The finite clauses provide evidence for the conceptualizer’s propositional attitude and/or mental scanning between the events described in the main and adverbial clause. This will be manifested in the dislocation from the parent space of origin to an adverbial mental space related to the conceptualizer’s perspective on the adverbial event.

Infinitive Clauses vs. Indicative Clauses The fact that the conjunction when may signal different meanings is widely known in the literature. Among others, Carlson (1979) analyses the atemporal generic reading of this conjunction, illustrating thus the way in which clauses introduced by when may be used roughly in the same manner as relative clauses. For example, expressions like dogs are expensive when they are of a rare breed has roughly the same semantic meaning as a relative clause, e.g., dogs that are of a rare breed are expensive. This observation is also made by (Declerck 1996a), who presents a thorough typological analysis of when and its different uses. In further

Finite Adverbial Clauses: Iconicity, Subjectification and Mental Spaces

129

studies, the author (Declerck 1996b) focuses on the function of when as, what she calls, a free relative adverb, comparing the time frame of the adverbial clause, the main clause and the moment of enunciation. Havu (1997) offers an extensive analysis of four different uses of this conjunction in Spanish (cuando) paying special attention to the iterative use of this conjunction, and its relation to the conditional si (‘if’). Finally, Carecho (1996) analyses the temporal profile of the Portuguese quando. The author concludes that this conjunction is not only used to describe events–or processes–but also to describe static situations. The common denominator for the studies mentioned above is that they share the insight that when/cuando/quando have a wide array of uses, and in particular that these uses often are related to a temporal complexity between the main clause and the adverbial clause. And, it is this temporal complexity that is noteworthy in the variation between clauses introduced either by ao or by quando in Portuguese. While quando is adequate for use in contexts of temporal complexity, ao seems to be highly restricted to punctual and sequential uses. In other words, while the use of quando is not merely restricted to punctual and sequential uses, this seems to be the case with ao. The following examples may illustrate the case in point: 137. quando cheguei à quinta, o caseiro… TEMP:CONJ. arrive-PRET: 1P.P to the farm, the house-keeper… coziame lá as batatas there the potato cook-IMP: 3P.S me [Pfa: Bruxedos] ‘When I arrived at the farm, the house-keeper cooked me some potatoes’ 138. de maneira que, às of manner that, to the exactamente para vir exactly to come-INF. ver um grande bando de flock of see-INF. a big [Pfa: Boa Pontaria] ‘So, in the small hours, when I ducks’

tantas, quando estava many, TEMP:CONJ. be-imp: 1p.p embora, comecei a away, begin-PRET: 1P.S to patos ducks was about to leave, I saw a big flock of

139. Jaime Barbosa Tinto… entristecese quando Jaime Barbosa Tinto… be sad-pres: 3p.p CLITIC.PRON. TEMP:CONJ. recorda que a sua casa foi destruída, the his house be-PRET: 3P.S destroy-PART. remember-PRES: 3P.P that acabando a travar as lágrimas ao dizer tears TEMP:PREP. say-INF. cease-PROG. to stop-INF. the

130

Chapter Four que a sua família está separada that the his family be-PRES: 3P.S separated [Diário de Coimbra-N1048-1] ‘Jaime Barbosa Tinto, becomes sad when he remembers that his house was destroyed, he can’t stop his tears from falling when he explains that his family is torn apart’

It is interesting to note different aspectual descriptions in examples (137138). In (137) the temporal clause quando cheguei à quinta designates the trajector’s arrival to the farm. Hence, the clause has a punctual aspect. On the other hand, the following example with quando (138) is more durative in character, being used as a background for the main clause event–the trajector sees the flock of ducks when he is about to leave. That is, the main clause event is located temporarily inside the not yet concluded adverbial event. This being so, examples (137-138) illustrate that clauses introduced by quando may have different aspects: punctual and durative. This aspectual complexity is not at hand in the clauses introduced by ao. Example (139) seems to represent a typical case. The infinitive event is more punctual in character and it co-occurs with the event of “letting the tears fall”. In order to find a plausible explanation for the difference between examples (137-138), on one hand, and example (139), on the other, it is relevant to consider the effect of grounding. The finite verb has a temporal profile. The preterite tense tends to designate events that are concluded, while the opposite is true for the imperfect tense. 23 The infinitive does not furnish any such information. Instead, the event is grounded by the main clause verb (recorda). Consequently, the only possible interpretation of (139) is that there is a punctual relation between the progressive event and the infinitive event. The event of “letting the tears fall” occurs at the same time as the event of “saying that the family is shattered”. Also, it is more than likely that the progressive clause has larger temporal extension than the infinitive clause. The difference between the examples discussed above is made more explicit by inserting an infinitive adverbial clause in (138). As the original 23

This is particularly salient with verbs like conhecer (‘to know (someone’)) and saber (‘to know (something’)). The imperfect tense of conhecer indicates that the subject knew a person, while the preterite tense has the meaning of learning to know a person. That is, the preterite tense describes the concluded process of getting to know a person, and the imperfect tense describes the static event of knowing a person. Likewise, the imperfect tense of saber means that a person knew something, but the preterite tense has the punctual meaning of getting to know something. Obviously, when you know someone or something, this implies a more durative aspect than the event of getting to know someone or something.

Finite Adverbial Clauses: Iconicity, Subjectification and Mental Spaces

131

example was more extensive in character, including the main clause event in it temporally, this is not the case with the infinitive clause. In fact, the substitution creates an expression that is somewhat strange: 140. de maneira que, às of manner that, to the exactamente para vir exactly to come-INF. ver um grande bando a big flock see-INF. [Pfa: Boa Pontaria] ‘So, in the small hours, when I ducks’

tantas, ao estar be-INF. many, TEMP:PREP. embora, comecei a away, begin-PRET: 1P.S to de patos of ducks was about to leave, I saw a big flock of

In fact, the infinitive clause, e.g., ao estar exactamente para vir embora seems to create problems in the context of (140). The event of being about to leave (estar para vir embora) is not entirely compatible with the semantics of AO+INFINITIVE. This creates a semantic conflict between the description of an event that is punctual and not concluded at the same time. Further, the linguistic expression designates two punctual events: the one of being about to leave and the event of seeing a flock of ducks. Rather than including the main clause event in the adverbial event, there are two punctual events co-occurring. Figure 4:7 captures some general tendencies regarding the aspect of clauses introduced by ao and quando: Figure 4:7. Aspectual tendencies and relation to main clause with ao and quando. A) Punctual included in B) Punctual not included in the main clause the main clause event

Ao Quando

Ao Quando

Main clause

C) Durative/Iterative including the main clause

Quando

Main clause

Main clause

Figure 4:7 illustrates the tendency for the infinitive clause to have a punctual aspect, while the finite clause may display punctual, durative and iterative aspect. In the relation between main clause and temporal clause, there may be inclusion so that one event is included in the time frame of the other event. Also, the relation may be sequential. That is, the main clause

132

Chapter Four

event and the temporal event are presented as two events that have a relation in which one event precedes the other. Figure 4:7 further shows that the temporal clause introduced by quando may include the main clause event. These general tendencies are strongly related to the aspect of the verb and the tense inflection in finite clauses and to the absence of tense in the infinitive ones. The pattern is shown below: A) PUNCTUAL AND INCLUDED IN THE MAIN CLAUSE EVENT: INFINITIVE + IMPERFECT TENSE: ao entrar eu, todos já estavam lá PRETERITE TENSE + IMPERFECT TENSE: quando entrei, todos já estavam lá ‘When I entered, everybody was there already’

B) PUNCTUAL NOT INCLUDED IN THE MAIN CLAUSE EVENT (SEQUENTIAL): INFINITIVE + PRETERITE TENSE: Ao entrar, vi que todos já estavam lá PRETERITE TENSE + PRETERITE TENSE: Quando entrei, vi que todos estavam lá ‘When I entered, I saw that everybody was there already’

C) DURATIVE/ITERATIVE INCLUDING THE MAIN CLAUSE EVENT: IMPERFECT TENSE + PRETERITE TENSE: Quando estava lá, vi o Jorge entrar When I was there, I saw George coming in’ IMPERFECT TENSE + IMPERFECT TENSE: Quando eu estudava, bebia muito café When I was studying, I used to drink a lot of coffee’

Obviously, the pattern displayed above does not illustrate the full complexity of the aspectual differences between clauses introduced by ao and quando. They do, however, supply a basic illustration of the most fundamental ones. The clauses that are introduced by quando have a far more extensive use than the ones introduced by ao. In fact, the occurrence of this type of clause in temporal contexts of more complexity seems to provide the basis for an extension to contexts where the finite adverbial clause designates a condition for the occurrence of the main clause event (cf. Dancygier and Sweetser 2000). This condition may have durative or iterative aspect. The main point is that clauses with quando may guide the interlocutor to mental spaces that I will refer to as habitual or generic circumstantial:

Finite Adverbial Clauses: Iconicity, Subjectification and Mental Spaces

133

141. quando o tempo estava assim menos mal, TEMP:CONJ. the weather be-IMP: 3P.S like less bad andávamos o dia todo; quando começava a all; TEMP:CONJ. begin-IMP: 3P.S to walk-IMP: 1P.P the day puxar mais o inverno, o navio chamava para the winter, the ship call-IMP: 3P.S to pull-INF. more bordo board [Pfu: 109] ‘when the weather was not so bad, we walked (free) all day long; when the winter came, the ship called us on board’ 142. nós só realmente damos, damos interesse, give-PRES: 1P.P interest we only really give-PRES: 1P.P, temos o verdadeiro interesse por uma criança real interest for a child have-PRES: 1P.P the quando ela é nossa TEMP:CONJ. she be-PRES: 3P.S our [Pfu: 193] ‘We only have genuine interest in a child when it is our own child’

It is interesting to note that the temporal profile of quando combines with other features in the examples displayed above (140-141). In effect, the temporal clauses show a certain affinity with conditional clauses. In (141), the clauses introduced by quando are premises for the occurrence of the main clause event: good weather results in walking free, and winter time is equated with work on the ship. Further, the iterative aspect implies that this situation is habitual. On the other hand, example (142) is more generic and circumstantial. The condition for sharing real interest in a child is that it is our own. In both examples, however, the conjunction quando introduces a clause that is not merely temporal. Instead it guides the interlocutor from a temporal parent space of origin to mental spaces with habitual and generic circumstantial features:

134

Chapter Four

Figure 4:8. Quando as a space builder.

Quando Habit space

Quando Ao Quando Temporal parent space

Generic circumstantial space

In sum, the conjunction quando functions as a space builder indicating that the clause is not merely to be understood in temporal terms. Instead, the clause is often a premise for the occurrence of another event–in most cases the main clause event. In cases of iterative aspect, the space builder guides the interlocutor to a habit space. That is, always when the event in the conjunction clause is brought about, it entails the event of the main clause. In durative contexts, quando tends to guide the discourse participants to a generic circumstantial space. This phenomenon was confirmed in example (142) and is further comparable with the example offered by Carson (1979) above. There is, so to speak, a certain affinity between this type of clause with quando and generic relative clauses. Figure 4:8 also highlights the fact that temporal clauses introduced by ao tend to be more restricted to the temporal parent space. The dislocation from a temporal mental space of origin to new mental spaces is further consistent with subjectification. Instead of describing a temporal relation between two events, the conceptualizer connects two events as being related in a habitual or generic circumstantial manner. That is, one event entails the other. This connection is ultimately related to the conceptualizer’s reasoning and propositional attitude towards events in the outside world and it also implies the conceptualizer’s mental scanning between the events described in the main clause and the adverbial clause. The conceptual differences confirmed above are moreover present in causal clauses introduced by the preposition por and the conjunction porque. As will be seen, the finite clauses display a more complex causality than the infinitive clauses. Prior analyses on the matter have frequently explained these conceptual differences by making a distinction between semantic and pragmatic causality (cf. Oversteegen 1997; Pander Maat and

Finite Adverbial Clauses: Iconicity, Subjectification and Mental Spaces

135

Delgand 2001; Pander Maat and Sanders 2000). Hence, the semantic causality is considered to be a question between events taking place in the world, while the pragmatic causality is related to the speaker’s conclusions about cause and effect in the outside world. Portuguese linguistics has shown a propensity to omit the pragmatic account of causality. Instead, the matter has first and foremost been related to the objectivist notion of causes and effects in the outside world. Nonetheless, Vogt (1976-1977) identifies a certain ambiguity concerning the conjunction porque. In a comparative analysis between por and porque, the author reaches the conclusion that the first designates a causal relation, while the latter is more related to the pragmatic uses mentioned above. In particular, Vogt (1976-1977: 152) notes the semantic affinity between porque and the adverbial conjunction pois (‘as’, ‘since’, ‘so’, ‘because’). In other words, porque does not designate an actual cause, but the speaker’s justification or explication of what is said in the main clause. The view taken in the present analysis is in accordance with the findings cited above. The difference, though, is that I will connect the use of porque to inferential processes in the conceptualizer (often equated with the speaker). That is, the use of porque is related to reasoning about causal relations in the world, while por tends to designate visible causal relations in the outside world. The distinction is subtle and there is no clear-cut division to be made between the different structures. However, the following examples may illustrate a certain difference: 143. a fruta quando era de aqui do mondego of here of the mondego the fruit when be-IMP:3P.S e destas zonas, se comprava mais CLITIC.PRON. buy-IMP: 3P.S more and of these regions, barata porque eram os próprios produtores CAUSE:CONJ. be-IMP: 3P.P the self producers cheap que vinham vender à praça sell-INF. to the square that come-IMP: 3P.P [Pfu: 104-105] ‘when the fruit came from round here, from Mondego and the nearby regions, you could buy it cheaper because it was the producers themselves that came to the square to sell it.’ 144. no outro dia ia um aluno a pupil in the other day go-IMP: 3P.S corredor a… cantar, e uma das one of the hall to… sing-INF., and ia à minha frente chama- o front call him go-IMP: 3P.S to the my

meu no my in the senhoras que ladies that e and

136

Chapter Four passalhe um grande descasque por a big reprimand CAUSE:PREP pass-PRES: 3P.S him ele ir a cantar. he go-INF. to sing-INF. [Pfu: 123] ‘the other day, one of my students was singing in the hall and one of the ladies walking in front of me called him over and gave him a reprimand for singing’

It could be argued that these cases display the same kind of causal relation between the described events. In (143), the causal relation would be PRODUCERS SELL FRUIT > LOWER PRICES, and in (144) the relation is: SINGING IN THE HALL > REPRIMAND. Nonetheless, there is a subtle difference between them. The causal event described by the infinitive clause is more direct and does not require the same amount of reasoning about the causal relation. The conceptualizer witnesses the act of singing and the subsequent reprimand. Therefore, there is no need to reason about the causal relation between the events. The finite clause, on the other hand, provides evidence for reasoning and inferential processes. It may be true that the lower price of the fruit was caused by the fact that the producers sold it. But, it is the conceptualizer that reaches this conclusion by reasoning about why the fruit was cheaper–and not by witnessing a causal relation in the outside world. In sum, the conceptualizer connects two separate events in the outside world as having a causal relation. 24 The inferential nature found in the finite causal clauses is even more salient in examples (145-146) below: 145. É evidente que a grande compra vai Be-PRES: 3P.S evident that the big purchase go-PRES: 3P.S ser feita porque cada família tem, CAUSE:CONJ. every family have-PRES: 3P.S be-INF. make-PART. no mínimo, uma viatura in the least, one trolley [Diário de Aveiro-N0846-1] ‘Clearly, a lot of shopping will be done because every family has at least one trolley’

24

In fact, the difference may be explained in terms of direct and indirect causal relations. The infinitive clause frequently designates a direct causal and temporal relation with the main clause event, while the finite clause describes a more indirect causality where the two events may have different temporal profiles. This is also manifested by the mental character of indirect causal relations (cf. Vesterinen 2008, 2010a)

Finite Adverbial Clauses: Iconicity, Subjectification and Mental Spaces

137



146. Segundo um dos ciclistas presente nesta iniciativa According one of the bicyclists present in this initiative “vamos ter com o presidente porque “go-PRES: 1P.P have-INF. with the chairman CAUSE:CONJ. vão ser apresentadas as novas bicicletas…” be-INF. present-PART. the new bicycles…” go-PRES: 3P.P [Diário de Aveiro-N0304-1] ‘According to one of the bicyclists present at the event: “we will meet the chairman to be presented with the new bicycles”’

The causal relations in examples (145-146) are without doubt of a more diffuse character than the one confirmed in example (144). The causal relation in (145) is EVERY FAMILY HAS AT LEAST A TROLLEY > A LOT OF SHOPPING WILL BE DONE and (146) displays a causal relation between THE BICYCLES WILL BE PRESENTED and WE WILL MEET THE CHAIRMAN. Obviously, the diffuse and indirect causal relation of these cases is related to inferential processes and to what the conceptualizer knows about events in the outside world. Thus, the conceptualizer infers the cause of the main clause event from what he considers to be the normal course of events in the world. Figure 4:9 provides an illustration of this inferential process: Figure 4:9. Inferred causal relation (examples 145-146).

145)

146)

Every family has a trolley

The bicycles will be presented

A trolley is used for big purchases

The chairman will be there

A lot of shopping will be done

We will meet the chairman

Figure 4:9 shows that inferential processes like (145-146) are dependent on implied knowledge (shown in the circle in the middle). Consequently, the conceptualizer in (145) connects the fact that every family has a trolley with previous knowledge that a trolley is appropriate for bigger purchases.

138

Chapter Four

Thereafter, the conceptualizer concludes that a lot of shopping will be done. Likewise, the conceptualizer in (146) knows that the chairman will attend when the new bicycles are presented. Thus, he/she knows that there is a good chance to meet him there. The inferential process described in Figure 4:9 therefore implies a more diffuse causal relation that the one attested in example (144). The causality is mental in character and not directly related to causes and effects in the outside world. In fact, the fundamental cause to “a lot of shopping” is first and foremost related to people’s needs or wishes, and the meeting with the chairman is probably caused by some prior need or desire to meet with him. This being so, a question to be raised is if the inferential process found in the finite causal clauses may be found in the infinitive ones. The question may be answered by a substitution test. The following examples are small modifications of examples (145-146) into infinitive causal clauses: 147. ? É evidente que a grande compra vai ser feita por cada família ter, no mínimo uma viatura [preposition: por] 148. ?“vamos ter com o presidente por irem ser apresentadas as novas bicicletas” [preposition: por]

The reason why these examples seem less natural than the finite ones is that por is related to direct causal relations that are grounded in the outside world rather than in the conceptualizer. Recall that finite clauses create a relation to the ground and that this feature constitutes the basis for mental processes. Consequently, the examples displayed above do not combine with reasoning and inference. This leads to a semantic incompatibility between an inferential context, on the one hand, and a causal clause that designates a direct causal relation, on the other. Further, it is likely that the semantic incompatibility discussed above may be explained by the metaphorical notion that por is equated with an entrance through which one proposition may be explained by another (cf. Bermúdez 2003: 247-248). Although this explanation may seem somewhat poetic in nature, it certainly reflects the proper meaning of this preposition. It also accommodates with the claim made in the present analysis. If por equates with an entrance between two propositions, it provides evidence of a direct causal relation between these propositions. Thus, the metaphor reflects the iconic nature of language. 25 25

Lunn (1987: 39) comments upon the relationship established by por with the following words: “The central relationship defined by por is a passage through a bounded space. That is, a trajector is understood to penetrate a landmark introduced

Finite Adverbial Clauses: Iconicity, Subjectification and Mental Spaces

139

Consequently, it is appropriate to attribute the conceptual differences between finite and infinitive causal clauses to iconic principles: formal complexity matches conceptual complexity, and formal distance reflects conceptual distance. The finite causal clause is not primarily related to purely causal relation, but to a mental and inferred causality. Thus, the conjunction porque should not only be considered a causal conjunction. It is also a space builder guiding the discourse participants from the parent space–the causal space–to a causal mental space of inference. The dislocation from the parent space to the mental space of inference is illustrated in Figure 4:10: Figure 4:10. Porque as a space builder.

Porque Por

Porque Inferential mental space

Causal parent space

Figure 4:10 describes the conceptual differences between finite and infinitive causal clauses introduced by por and porque in a mental space analysis. The finite clause, introduced by porque is not restricted to the causal parent space. In fact, it is more closely related to the inferential mental space. The infinitive causal clause, on the other hand, does not dislocate from the parent space. Instead, it provides evidence for a close relation between the events described in the main clause and the causal clause. The fundamental difference between the clauses is found in the grounding–or lack of grounding–of the causal event. The lack of grounding of the causal event implies a closer relation to the main clause event. The causal event is grounded by the main clause and, consequently, it shares the same temporal profile. The finite clause is more independent in relation to the main clause and has its own temporal profile. Therefore, it also creates a by por”. Consequently, the notion of por as an opening between two entities or events is rather entrenched.

140

Chapter Four

relation to the ground and lends itself to mental processes and acts or reasoning. In sum, the grounding of the event contributes to a change of perspective. Instead of designating a causal relation that is visible in the outside world, the finite clause designates the conceptualizer’s inferences about causal relations in the outside world. Thus, the dislocation from a parent space to an inferential mental space also involves subjectification.

Infinitive Clauses vs. Subjunctive Clauses There is a rather homogeneous notion that the subjunctive mood adds a semantic dimension to the clause in which it occurs, and that the infinitive has a “weaker” semantic meaning. That is, the difference in conceptual content between infinitive and subjunctive clauses is closely related to the semantic meaning of the verb form used in the clause. However, there is no standardized answer to be found for the semantic meaning of the subjunctive mood. Some scholars relate the subjunctive and its meaning to notions like non assertion, while others claim there to be a relation between the relevance of the proposition, on the one hand, and the choice of the subjunctive mood, on the other. Terrell (1995) makes a distinction between the indicative and the subjunctive mood in Spanish affirming that the first is selected in asserted clauses while the second occurs in clauses that are not asserted, but presupposed by the speaker. In order to explain the occurrence of presupposed clauses with the indicative mood, the author distinguishes between strong presupposition (the subjunctive mood) and weak presupposition (the indicative mood). Strong presupposition includes factive predicates like deplorar (‘deplore’), sentir (‘feel, be sorry’), resentir (‘resent’) and lamentar (‘regret’), and impersonal expression of the type es triste que (‘it is sad that’) and others. Weak presupposition is associated with verbs like descubrir (‘discover’), ver (‘see’), revelar (‘reveal’) and expressions like darse cuenta de (‘realize’). The analysis offers an interesting starting point for the mood choice, but does not actually explain why a distinction is made between weak and strong presupposition (cf. Terrell 1995). 26 26

See Vesterinen (2010b) for a conceptual approach to the distinction between clauses displaying weak and strong presupposition. The variation between the indicative and the subjunctive mood in these clauses is argued to depend on factors like input and output. The indicative is used in contexts where the conceptual content of the described event enters the conceptualizer’s dominion of reality, while the subjunctive is used in contexts where the conceptualizer comments upon an event outside his/her dominion of active control.

Finite Adverbial Clauses: Iconicity, Subjectification and Mental Spaces

141

The notion of non assertion is also present in Tlâskal (1984), Marques (1995) Oliveira (2001) and Travis (2001). Tlâskal (1984) analyses the subjunctive mood in contexts of non-reality and claims that the semantics of the subjunctive are associated with this feature. Oliveira (2001) relates the non-assertive feature of the subjunctive mood to a subjective distance between the speaker and the described event. By using the subjunctive mood, the speaker signals a mental distance to the described event. Marques (1995), on the other hand, makes a distinction between indicative clauses describing the epistemic attitude of the speaker and subjunctive clauses describing other modalities. Finally, Travis (2001) introduces the terms anti-assertive and anti-cognitive in order to capture the semantic meaning of the subjunctive. The first meaning (anti-assertive) is related to the speaker’s reluctance to comment upon future events. The second meaning, the anticognitive meaning in the terminology of the author, signals that the speaker ignores the veracity of the described event. Among others, Lunn (1987) and Ferreira (1984) analyse the relation between the subjunctive mood and the degree of relevance of the clause in which it occurs. Lunn (1987) identifies a low degree of relevance in Spanish imperfect clauses with the subjunctive mood. According to the author, the imperfect subjunctive is used in contexts where the content of the clause is known. That is, the subjunctive clause signals known information with a low degree of relevance. This approach is totally opposed to the analysis offered by (Ferreira 1984). Instead, the subjunctive mood is understood as the “element carrying principal information” (Ferreira 1984: 291). 27 Although the contributions mentioned above provide important insights about the semantics of the subjunctive, they seem to deal with the issue in a somewhat restricted manner. Instead of seeking the semantics of the subjunctive by taking the whole array of contexts in which it occurs into consideration, they focus on restricted contextual occurrences. Thus, notions like presupposition, non-reality, subjective distance and relevance are able to explain some occurrences of the subjunctive, while they fail to explain other occurrences. Further, the distinction between weak and strong presupposition describes the mood alternation, but it does not explain it. The following examples are therefore difficult to explain: 149. vejo que a nossa integração europeia funcionou bem (indicative mood) [Viana Diário -N2758-1] ‘I see that our European integration has worked well’

27

My translation: ”o elemento portador de informação principal” (Ferreira 1984: 291).

142

Chapter Four 150. este responsável lamenta que a autarquia não dê o apoio (subjunctive mood) [Diário de Aveiro-N1072-1:]. ‘this contributor regrets that the organization does not give any support’ 151. É preciso que as crianças aprendam a brincar (subjunctive mood) [Diário de Aveiro-N3324-4] ‘It is necessary that children learn to play’ 152. É preciso apurar as causas que levam as crianças largar a escola (subjunctive mood) [Diário de Aveiro-N0288-12] ‘We need to ascertain the causes leading children to leave school’

First, the mood alternation in examples (149-150) is not really explained by using the terms weak and strong presupposition. The explanation is probably more related to the conceptual content of the main clause verb. Second, example (150) designates an event that is factive and presupposed by the speaker. Thus, there is no meaning to associate the subjunctive mood exclusively to non-reality. Further, examples (151-152) designate events that seem to be of great concern for the speaker. Thus, the claim that the subjunctive mood signals a low degree of relevance or subjective distance is not easily understood. Rather, I will propose that the subjunctive is used in two different contexts, and that these contexts are related to the notion of dominion. In a seminal analysis of the subjunctive mood, Maldonado (1995) introduces the term in order to capture the semantics of the subjunctive. That is, the subjunctive mood is used in contexts where the conceptual content is outside the conceptualizer’s dominion. Figure 4:11 illustrates this situation: Figure 4:11. Events located inside or outside the conceptualizer’s dominion.

c

e

c e

A) Inside

B) Outside

Finite Adverbial Clauses: Iconicity, Subjectification and Mental Spaces

143

In Figure 4:11, the circle (c) represents the conceptualizer. This participant is equated with the main clause trajector. The circle (e) represents the described event to which the conceptualizer has a certain relation. If the event is located within the conceptualizer’s dominion, the relation in Figure 4:11a prevails. On the other hand, the event may also be located outside the conceptualizer’s dominion. In this case, the relation is described by Figure 4:11b. Recall that the main clause trajector equates with the conceptualizer of the event described in the subordinate clause. The ground, however, is the conceptualizer of the linguistic expression as a whole. Maldonado (1995) connects the events described in Figure 4:11 to the notion of control and to the conceptualizer’s ability to actively control the described event in terms of elaborated reality (cf. Achard 1998). The latter term refers to the dynamic nature of reality and comprehends not only events verified in the past (the basic reality), but also the probability that future events will occur (projected reality). That is, the way in which things have evolved until the present time makes it possible for us to predict some future events with some degree of certainty–and to exclude other events from our conception of reality (cf. Achard 1998). Although Maldonado captures the essential meaning of the subjunctive mood, the analysis is utterly related to the conceptualizer’s conception of reality, which leaves out another facet–the conceptualizer’s ability to actively manipulate the described event. As Langacker (2004: 537) eloquently points out: “being alive is to function as actor in control cycles, interacting with the environment to gain control over certain facets of it”. In fact, a lot of events in our daily life can actually be influenced and manipulated. If I am hungry, I can easily choose to do something about it, if I am tired, I can take a nap etc. Socially, we establish stable relations with certain rules and expectations that increase the notion of control. That is, in the same way that the history of events makes it possible to predict or exclude future events, the social organization of human life entails that we have the possibility to exert control over some events and a lesser degree of control over others. Thus, the position of the present analysis is that an event may be located outside the conceptualizer’s dominion for two different reasons: (1) the event does not belong to the conceptualizer’s conception of reality; (2) the conceptualizer is not able to actively manipulate the event. In both cases, the conceptualizer has a lesser degree of control over the described event. A few expressions may explain the difference. In the expression Não creio que o João venha hoje (‘I do not believe that John is coming today’), the described event does not belong to the conceptualizer’s conception of reality. Further, the described event is also outside the conceptualizer’s

144

Chapter Four

dominion of active control. That is, both conditions for the event being outside the conceptualizer’s dominion are satisfied. An event that is outside our conception of reality is also outside our dominion of active control and manipulation. In contrast, the expression É pena que ele não venha hoje (‘It is a pity that he is not coming today’) designates an event that is inside the conceptualizer’s epistemic dominion. However, the conceptualizer has a minor possibility to actively influence or manipulate the event–it is located outside the dominion of active control. If we relate the notions of dominion and control to infinitive and subjunctive adverbial clauses in Portuguese, and to the mental spaces analysis, it is interesting to note that these clauses seem to differ in the degree of control over the described event. In fact, infinitive clauses designate events with a higher degree of control, while subjunctive clauses often dislocate to a mental space with a lesser degree of control. Let us consider some purpose clauses: 153. Outro aspecto importante para explicar a importância Another aspect important to explain the importance do ritmo na vida de um ser humano, of the rhythm in the life of a be-INF. human, prendese com o facto de todas as with the fact of all the hold-PRES: 3P.S CLITIC.PRON mães, tanto as, dextras como as canhotas, mothers both the right-handed as the left-handed, virarem a cabeça do bebé para o ‘lado the head of the baby to the ‘side turn-INF: 3P.P do coração’ quando os põem ao colo, of the heart’ when them put-PRES: 3P.P to the lap, para eles se sentirem confortáveis PURPOSE:PREP. they CLITIC.PRON. feel-INF: 3P.P comfortable ao ouvirem as batidas daquele órgão humano organ human when hear-INF: 3P.P the beats of that [Diário de Leiria N1649-1] ‘Another important aspect of the role that rhythm plays in our lives is related to the fact that all mothers, both right-handed and left-handed, turn their baby’s head to the side where the heart is when they put them on their lap so that they feel comfortable hearing the mother’s heart-beat’. 154. é be-PRES: 3P.S monta, pois weight, for para que

claro que, clear that, imediatamente immediately se

como são coisas de pequena as be-pres: 3p.p of small insistem connosco insist-PRES: 3P.P with us pague ao terceiro

Finite Adverbial Clauses: Iconicity, Subjectification and Mental Spaces

145

PURPOSE:CONJ.

CLITIC.PRON. pay-PRES.SUBJ: 3P.S to the third os prejuízos que teve ou qualquer coisa e or anything and the dammage that have-PRET: 3P.S tal, por causa da polícia, para que não tome such, for cause of the police, PURPOSE:CONJ. not take-PRES.SUBJ: 3P.S conta da ocorrência account of the occurrence [Pfu: 165] ‘Obviously, being such small sums, they always insist on paying the damages, whatever they may be, to the third party, so that the police do not get involved in the matter’

The overall impression of (153-154) is that they differ substantially in how the main clause subject conceives of the described event in the purpose clause. On the one hand, example (153) designates an event over which the main clause subject has control. It is a fact that the rhythm of the heart-beat has a calming effect, so a well known strategy for calming a baby is to put the baby’s head near the mother’s heart. On the other hand, example (154) displays a lesser degree of control over the event described in the purpose clause. The main clause trajector has to persuade the speaker in order to achieve the purpose clause event, that is, he/she insists on paying the losses to the third party. Further, the first purpose clause is related to the other one, expressing the wish that the police do not get involved. In both cases, the main clause trajector depends on the action of another participant (the speaker) in order to achieve the events described in the purposes clauses. In this sense, it is noteworthy that the negation of the purpose clause sounds rather natural in (154) but seems to cause more problems in (153). The negations are shown in examples (155-156): 155. They always insist on paying the damages, or what it may be, to the third part so that the police do not get involved in the matter. But, we do not let them pay and we always inform the police. 156. Another important aspect of the role that rhythm plays in our lives is related to the fact that all mothers, both right-handed and left-handed, turn the baby’s head to the side where the heart is when they put them on their lap so that they feel comfortable hearing the mother’s heart-beat. ? But they do not feel comfortable.

The negation of the two purpose clauses in (155) does not cause any semantic concerns. In fact, it is most likely that these events are negated in real life. The reversed situation would actually imply that the speaker is breaking the law by not informing the police about the accident. Further, the subjunctive mood in the actual context of (155) is not surprising. Rather, it

146

Chapter Four

designates that the main clause subject has a restricted degree of control over the described event. In contrast, the negation of the infinitive purpose clause in (156) is more troublesome. The intension of the discourse is to highlight the importance of rhythm in human life. The importance is exemplified by making reference to the relation between the mother’s heartbeat and the baby’s feeling of comfort and content. Therefore, the negation of (156) creates a semantic conflict–the positive effects expressed by the speaker are also negated. The relation between verb form and conceptual content in the examples above is not accidental. The subjunctive mood is used in a context where the main clause subject has a lesser degree of control over the event described in the purpose clause while the infinitive purpose clause designates a higher degree of control. Also, the relation is iconic. A higher degree of formal complexity matches a higher degree of conceptual complexity, and a higher degree of formal distance between the clauses reflects conceptual distance. That is, the greater distance between the main clause and the purpose clause mirrors a situation in which the main clause subject has a more restricted control over the adverbial event. This is also true for the following temporal clauses: 157. o líder comunista Guennadi Siuganov tem the leader communist Guennadi Siuganov have-PRES: 3P.S reiterado que não reconhecerá os resultados not recognize-FUT: 3P.S the results repeat-PART. that oficiais das eleições, antes que os observadores do observers of the official of the elections, TEMP:CONJ. the PC verifiquem minuciosamente as cópias dos the copies of the PC verify-PRES.SUBJ: 3P.P thoroughly protocolos. protocols. [Diário de Aveiro-N2435-1] ‘the communist leader Guennadi Siuganov has repeated that he will not recognize the results of the elections before the copies of the protocols have been thoroughly verified.’ 158. O bebé The baby permanecer remain-INF. Menores, Juvenile, assistente assistant o caso

encontrase bem e vai go-PRES: 3P.S find-pres: 3p.s CLITIC.PRON. well and no hospital até que o Tribunal de in the hospital TEMP:CONJ. the Court of a Administração Regional de Saúde e a the Administration Regional of Health and the social encontrem uma solução para a solution to social find-PRES.SUBJ: 3P.P

Finite Adverbial Clauses: Iconicity, Subjectification and Mental Spaces

147

the case [Diário de Aveiro-N1807-1] ‘The baby is well and will remain at the hospital until the Juvenile Court, the Regional Health Administration and the social worker find a solution to the case’ 159. Gostaria de comprar casa na Figueira da Foz, Like-COND: 1P.S of buy-INF. house in the Figueira da Foz mas melhor ainda era ganhá -la de presente, win-INF. it of gift, but better even be-IMP: 3P.S porque caso contrário só depois que o vinho the wine because case contrary only TEMP:CONJ. do Bartolo sair of Bartolo go out-FUT.SUBJ: 3P.S [Diário de Coimbra-N2452-1] ‘I would like to buy a house in Figueira da Foz, but it would be even better to get it as a gift, because otherwise it will not be possible until after I have released the Bartolo wine onto the market’

The common denominator for the examples displayed above is that they designate future events, i.e., events that have not been concluded. In (157), the communist leader expresses a condition for recognizing the result of the elections. In the following example (158), the solution to the described problem is not found as the linguistic expression is uttered. Finally, example (159) describes the future release of the new Bartolo wine. Even though the main clause subjects want the described events to occur, they do not belong to their projected reality. The use of the subjunctive mood in these cases is thus related to a lesser degree of control over the described event. The use of different verb forms describing completed events may add more clarity in order to illustrate the conceptual differences attested above. In the following cases (160-162), the adverbial event is obviously brought about, but the use of one or another verb form seems to alter the perspective on the described event: 160. …tenho ouvido dizer que há that be-PRES: 3P.S …have-PRES: 1P.S hear-PART. say-INF. pessoas e tenho colegas minhas que persons and have-PRES: 1P.S colleagues mine that tiveram grandes dificuldades antes de conseguirem difficulties TEMP:PREP. achieve-INF: 3P.P have-PRET: 3P.P big arranjar um lugar, um bom lugar arrange-INF. a place, a good place [Pfu: 198]

148

Chapter Four ‘I have heard that there are such people, and I have colleagues that went through a lot of difficulties before finding a place, a good place’ 161. O fundador do parque, Manuel Leal, não esconde The founder of the park, Manuel Leal,not hide-PRES: 3P.S alguma tristeza por ter que ‘abrir mão’ de ‘open-INF. hand’ of some sadness for have-INF. that algo que tem estado no centro da something that have-pres: 3p.s be-part. in the centre of the sua vida depois que regressou de Moçambique, nos TEMP:CONJ. return-PRET: 3P.S of Mozambique in the his life anos 70 years 70 [Diário de Aveiro-N3767-1] ‘The founder of the park, Manuel Leal, does not hide his sadness at being forced to beg for what has been his mission since he returned from Mozambique in the seventies’ 162. Muitas mulheres, crianças e homens, alguns com as Many women, children and men, some with the mãos no ar, saíram do avião pela through the hands in the air, leave-PRET: 3P.P of the aircraft porta de trás, antes que a escada fosse be-IMP.SUBJ: 3P.S door of back, TEMP:CONJ. the ladder encostada ao avião lean-PART to the aircraft [Diário de Leiria-N1732-1] ‘Many women, children and men, some with their hands in the air, left the aircraft through the back-door before the ladder had been leant against the aircraft’

The infinitive clause in (160) designates an event from the perspective of the actual time of the uttering. The adverbial event is grounded by the finite clause and is thus located within the participant’s epistemic dominion as a part of his/her basic reality. That is, the event of finding a place is located within the dominion of the relative clause trajector (the colleagues). The subsequent adverbial event (161) is also a part of the conceptualizer’s epistemic dominion. The difference, however, is that the adverbial event is grounded by the indicative, which alters the perspective taken on the event. The perspective of the speaker is emphasized. Example (162) is more complex. Obviously, the conceptualizer knows that the ladder eventually was leant on the aircraft, but the subjunctive mood alters the scene and presents it from a perspective where this event was not concluded. Thus, the imperfect subjunctive adds a lesser degree of control to the described event

Finite Adverbial Clauses: Iconicity, Subjectification and Mental Spaces

149

by a temporal dislocation to a time that is prior to the realization of the adverbial event. The conceptual differences noted in the examples above may seem speculative and built on purely intuitive grounds. In fact, it could be argued that there are no conceptual differences at all between finite and infinitive clauses. In this sense, the infinitive clauses are surface manifestations of a deeper structure including tense (cf. Chapter 1). It could also be argued that the differences are not explained by the dislocation from a parent space to a mental space of a lesser degree of control. There are, however, two features that motivate the position taken in the present analysis. The first one pertains to the semantic differences between the infinitive and the subjunctive, and the other to the semantics of prepositions and conjunction. To begin with, the semantics of the infinitive are closely related to that of nouns. This is, for example, seen in the terminology of the infinitive– frequently referred to as a nominal verb form. It is also noted in the holistic and atemporal feature of the infinitive (cf. Langacker 2000). Therefore, a closer look at the semantics of infinitive adverbial clauses reveals some affinities with nouns, even though the first designate processes and the others designate things. The following examples (163-164) may facilitate the argumentation: 163. João Inês Vaz (…) terá já João Inês Vaz (…) have-FUT: 3P.S already impedir a manifestação, garantindo impede-INF. the manifestation, assure-PROG. que a GNR vai continuar that the GNR go-PRES: 3P.S continue-INF. tempo necessário para a execução

admitido não admit-PART. not ao mesmo tempo to the same time no local o in the local the dos trabalhos

time necessary PURPOSE:PREP. the execution of the work [Viseu Diário-N2203-1] ‘João Inês Vaz…has already promised not to impede the demonstration while at the same time confirming that GNR will stay on in the premises during the time that will be necessary for the fulfillment of the work’ 164. Os hospitais da Universidade de Coimbra foram The hospitals of the University of Coimbra be-PRET: 3P.P escolhidos para iniciarem este projecto initiate-INF: 3P.P this project chose-PART. PURPOSE:PREP. [Diário de Coimbra-N3714-1] ‘The hospitals of the University of Coimbra were chosen to initiate this project’

150

Chapter Four

The noun phrase in (163), e.g., the fulfillment of the work, does not signal any doubt about the realization of the event, but presents it as a fact. This effect is a typical feature of the nominalization. Instead of describing a process that may be interrupted, it designates a thing as if it already existed in the world. In the same manner, the holistic and atemporal feature of the infinitive make it neutral to the question as to whether the described process will be accomplished. Recall that a typical feature of the infinitive is that it describes a type of process and not an instantiation of a process. This seems to be the case also in adverbial contexts. Thus, the infinitive seems to be dependent on another verb in order to describe processes that are observable in the outside world. This is particularly salient in the structure PERCEPTION VERB + INFINITIVE COMPLEMENT. In these cases, it is also noteworthy that the semantics of the main clause verb presupposes the factivity of the complement event. The event is located inside the conceptualizer’s epistemic dominion: 165. Vi duas senhoras serem atacadas [Diário de Aveiro-N2240-1] ‘I saw two ladies be-INF: 3P.P attacked’ 166. Senti a minha cama tremer [Diário de Leiria-N0991-1] ‘I felt my bed shake-INF: 3P.S 167. Ouviu populares imputarem a prática do crime a alguém [Diário de Coimbra-N0859-1] ‘He/she heard people attribute-INF: 3P.P the crime to someone’

In the examples displayed above (165-167), the structure PERCEPTION + INFINITIVE COMPLEMENT designates a direct perceptual relation between the main clause subject and the event described by the complement. The semantics of the structure PERCEPTION VERB + INFINITIVE COMPLEMENT implies that the complement event is actually experienced, and it also implies that the event is located inside the conceptualizer’s epistemic dominion–it belongs to the conceptualizer’s elaborated reality. The direct perceptual relation is an effect of the grounding of the event. The event described by the infinitive complement is grounded by the main verb. Thus, the infinitive structure designates the main clause subject’s perceptual experience. The fact that the perception verbs see, feel and hear also take complements in the indicative mood makes it tempting to also attribute a high degree of control to events described by this verb form. However, the indicative complement creates a relation to the ground and therefore it VERB

Finite Adverbial Clauses: Iconicity, Subjectification and Mental Spaces

151

designates not only the main clause subject’s experience, but also the conceptualizer’s perspective on the described event. Further, the grounding of indicative complements does not merely designate perception. Rather, the relation between the main verb and the complement event is mental. The conceptualizer attributes a mental process to the main clause subject. The perception verb see illustrates some semantic differences between the infinitive, the indicative and the subjunctive: 168. ? O João vê os filhos estarem cansados [infinitive mood] 169. O João vê que os filhos estão cansados [indicative mood] 170. * O João vê que os filhos estejam cansados [subjunctive mood] ‘John sees that the children are tired’

The reason why example (168) sounds unnatural has to do with the meaning of the main clause verb in the actual contexts. In fact, it is difficult to see the process of someone being tired. It is more likely that the main clause subject infers the state of the children on the basis of some perceptual experience. A natural case would be that he/she sees the children yawning, or lying on the sofa. Hence, see does not describe direct perception, but inferential processes drawn on perceptual processes in the context of (168170). Therefore, the indicative complement in (169) is more appropriate, given that it designates not only perception but mental processes. The mental character of the relation implies a more indirect relation between the main clause event and the complement event, and also a lesser degree of control. That is, even though the event is located inside the conceptualizer’s epistemic dominion, it is so to a lesser degree than in the case of direct perception. Finally, the impossibility of using a subjunctive complement is consistent with the semantics of this mood. There is a semantic conflict between a context of control and a mood that signals a lack of control. In effect, this latter feature is normally at hand in contexts with the subjunctive mood. Some typical contexts are offered below: 171. Quero que a Joana saia comigo the Johanna go out-PRES.SUBJ: 3P.S with me Want-PRES: 1P.P that ‘I want Johanna to go out with me’ [volitional context]

172. É possível que a Joana venha possible that the Johanna come-PRES.SUBJ: 3P.S Be-PRES: 3P.S ‘It is possible that Johanna will come’ [epistemic context/doubt]

152

Chapter Four 173. É triste que a Joana não venha sad that the Johanna not come-PRES.SUBJ: Be-PRES: 3P.S 3P.S ‘It is a pity that Johanna is not coming [factive context/feeling] 174. Mando que a Joana venha that the Johanna come-PRES.SUBJ: 3P.S Order-PRES: 1P.P ‘I order Johanna to come’ [causal context/command] 175. Duvido que a Johana venha THAT THE JOHANNA come-PRES.SUBJ: 3P.S Doubt-PRES: 1P.S ‘I doubt that Johanna will come’ [epistemic context/doubt] 176. Não creio que a Joana venha JOHANNA come-PRES.SUBJ: 3P.S Not believe-PRES: 1P.S THAT THE ‘I do not believe that Johanna will come’ [epistemic context/doubt]

The examples displayed above share the feature of describing events with a low degree of control over the event described in the subjunctive clause. The volitional context of (171) designates the main clause trajector’s wish that the subordinated event will happen, but it is outside this participant’s control. The final decision is made by the trajector of the subjunctive clause. Moreover, it is interesting to note that the co-referential counterpart, e.g., quero sair (‘I want to go out’), takes the infinitive. The most appealing explanation for this tendency is that we control our own actions more than we control another person’s actions. The following examples display a similar pattern. In example (172), the main clause designates an epistemic context of doubt and, consequently, the subjunctive mood follows in the subordinate clause. In (173), the sadness of the main clause trajector is related to the factive event described in the subordinate clause–an event that the main clause trajector cannot influence actively. Further, the command in (174) does not determine that the trajector of the subordinate clause will obey. The main clause trajector can only hope that it will happen. Finally, the expressions of doubt in examples (175-176) signal that the subordinate events do not exist in the elaborated reality of the trajector. This being so, the subjunctive mood occurs in the subordinate clause. As for these two latter examples, it is revealing that the reversed expressions (‘I do not doubt’ and ‘I believe’) take the indicative mood. The mood alternation in relative clauses provides another case in point. The subjunctive mood occurs in cases where the object referred to does not exist in the conceptualizer’s conception of reality, while the opposite holds for the indicative mood:

Finite Adverbial Clauses: Iconicity, Subjectification and Mental Spaces

153

177. Procuramos uma pessoa que fala Português [indicative mood] ‘We are looking for a person that speaks Portuguese’ 178. Procuramos uma pessoa que fale Português [subjunctive mood] ‘We are looking for a person that speaks Portuguese’

The difference between examples (177) and (178) is that the person in the first relative clause (177) is known by the speaker. The indicative mood signals that the person exists in the elaborated reality of the conceptualizer. The conceptualizer is searching for a specific person, and that person speaks Portuguese. On the other hand, the conceptualizer in (178) does not know the person referred to. He/she is only searching for a person who is able to speak Portuguese. It does not matter who that person is. Finally, the structure with the factive verb LAMENTAR + COMPLEMENT provides evidence for an interesting variation. A quantitative analysis comprising 221 cases of this verb in third person singular exhibits merely one occurrence of the infinitive in non co-referential contexts. In contrast, 220 cases take the subjunctive complement. The prevailing pattern is: 179. …lamenta sentirse obrigado a …regret-PRES: 3P.S feel-INF. CLITIC.PRON. oblige-PART. to tornar público este esclarecimento this clarification turn-INF. public [Diário de Coimbra-N3700-1] ‘He/she regrets that he/she feels the obligation to make this clarification public’ 180. Jorge Sampaio lamenta que a maioria das the majority of the Jorge Sampaio regret-PRES: 3P.S that Pessoas apenas dê atenção às florestas Persons only give-PRES.SUBJ: 3P.S attention to the forests quando há fogos when be-PRES: 3P.S fire [Diário de Coimbra-N2718-1] ‘Jorge Sampaio regrets that most people are only concerned about the forests when there is a fire’

Obviously, the pattern reveals a conceptual difference between the competing structures. The subjunctive is chosen in non co-referential contexts. These contexts also display a lesser degree of control over the described event, given the fact that they entail the involvement of another person. The infinitive, on the other hand, is used in co-referential contexts. That is, the infinitive tends to occur when the main clause trajector is the one responsible for the complement event. This being so, the infinitive

154

Chapter Four

correlates with contexts of a higher degree of control over the described event. The conceptual differences confirmed above are also a result of the component that introduces the clause. Prepositions are frequently used to put objects together, while conjunctions divide events into different levels in a clause binding hierarchy. The tendency for prepositions to create interconnections is predominant in the vast number of compound words linked together by structures like those below: 181.

A) B) C) D) E)

Máquina para escrever Casa de banho Dona de casa Pano de chão Feito à mão

> > > > >

Typewriter Bathroom Housewife Floor cloth Handmade

The tendency is also noteworthy in many possessive constructions. The possessive relationship between the owner and the owned object is frequently formed by the structure NOUN + PREPOSITION + NOUN and the conceptual content of these constructions is, in its most abstract sense, that two entities are located in the same cognitive domain (cf. Langacker 1991). Thus, possessive expressions like a casa do Pedro (‘Peter’s house’), o carro da Joana (‘Johanna’s car’) etc. designate a relation in which the owner and the possessed object are located in the same domain. Moreover, they designate relations of control–the possessing entity controls the possessed object. These conceptual consequences are also noteworthy in the variation between concessive clauses introduced either by the preposition apesar de (litt. ‘In weight of’) or by the conjunction embora (from em boa hora, ‘In good time’). The canonical relation between the main clause and the concessive clause is formulated as although P, Q (cf. Verhagen 2000; König and Siemund 2000). However, infinitive concessive clauses introduced by a preposition tend to be entrenched in factive and visible relations, while finite clauses display hypothetical features: 182. …o que está realmente à vista, to the sight …the that be-PRES: 3P.S really muito bem, tá muito limpado, very good, be-PRES: 3P.S very clean-PART., muito arranjadinho – apesar das paredes walls very arrange-PART.DIM. – CONC:PREP. de ser arranjados of be-INF. arrange-PART [Pfu: 164]

tá be-PRES: 3P.S tá be-PRES: 3P.S precisarem need-INF: 3P.P

Finite Adverbial Clauses: Iconicity, Subjectification and Mental Spaces

155

‘What I can see is really nice, it is very clean, it is very tidy, even though the walls need to be fixed’ 183. …os aveirenses embora ultrapassem os Surpass-IMP.SUBJ: 3P.P the …the Aveiro team CONC:CONJ. adversários de hoje, na tabela classificativa, Opponents of today, in the table league, ficam com o mesmo número (31) de pontos. stay-PRES: 3P.P with the same number (31) of points. [Diário de Aveiro-N0110-7] ‘Even if Aveiro win against their opponents today, they will stay at the same position in the league table (31).’

Examples (182-183) illustrate this difference. The infinitive concessive clause in (182) describes a fact that is visible in the outside world. The conceptualizer, equated with the speaker, has a direct access to the events commented upon in the concessive relation: the place is nice, although the walls need to be fixed. On the other hand, the conceptualizer of example (183) creates a concessive relation that is more related to reasoning about possible future events than to events that are more concrete. In short, the conceptualizer creates a relation that is based on his own hypotheses (cf. Vesterinen Forthcoming) The mental character found in (183) may also explain occurrences in which there seem to be a logical opposition between the conceptual content of the main clause and the concessive clause: 184. efectivamente, aqui neste hotel gosto de trabalhar, actually, here in this hotel like-PRES: 1P.S of work-INF. embora seja um trabalho chato CONC:CONJ. be-PRES.SUBJ: 3P.P a job boring [Pfu: 277] ‘Actually, I like working at this hotel, even though it is a boring job’

In fact, example (184) may create confusion given that the concessive relation seems to be of a kind where the opposition between the two clauses is far too big to be coped with. That is, how can a person like his work if he actually thinks that it is boring? Obviously, the relation is related to mental experiences. This being so, one interpretation of the concessive relation is that the conceptualizer of (184) actually thinks that the work is boring but likes it for other reasons. In this case, the event described in the concessive clause is outside the conceptualizer’s dominion of active control–he/she cannot change the conditions of the work. Another possible interpretation is that the conceptualizer, in fact, does not consider the work to be boring, but

156

Chapter Four

knows that this is the general opinion about it. In this case, the conceptual content is located outside the conceptualizer’s conception of reality. Finally, the conceptual differences confirmed up to this point are also visible in relation to adverbial manner clauses that are introduced either by the preposition sem or by the conjunction sem que. A comparison between finite and infinitive clauses is highly illustrative: 185. …é que a autarquia não aproveitou the organization not utilize-PRET: 3P.S …be-PRES: 3P.S that o mês de paragem da instituição para concluir the month of stop of the institution to complete-INF. as obras, sem causar transtornos ao normal cause-INF. disorder to the normal the works, MANNER:PREP. funcionamento do Centro e, claro, sem que as function of the Centre and, of course, MANNER:CONJ. the crianças ficassem expostas ao perigo children be-PRET.SUBJ: 3P.P expose-PART. to the danger [Diário de Aveiro-N0757-1] ‘the organization did not use the holiday month to complete the work, thus causing disorder in the normal function of the Centre and, of course, leading to the children being exposed to danger’

The first occurrence of the manner clause implies that the main clause trajector (the organization) has a possibility to actively influence the event of creating disorder in the daily work of the institution. The disorder would have been avoided if they had chosen to complete the work during the holiday month. The second manner clause, however, alters the perspective and designates the conceptualizer’s concern for the children being exposed to the danger implied in the works. The inclusion of a new participant in the event (the children) has the effect of diminishing the main clause subject’s degree of control. In fact, the main clause participant is not the only one responsible for the event described in the manner clause, but the children are also active participants. Thus, the conceptualizer creates an event in which the main clause trajector loses control. In sum, the analysis presented above provides an explanation for the conceptual differences between infinitive and subjunctive adverbial clauses in Portuguese. The infinitive clauses designate a direct adverbial relation with a high degree of control, while the subjunctive clauses are more complex in character, dislocating from the parent adverbial space to a mental space with a lesser degree of control. In this sense, the conceptual (or semantic) differences between prepositions, on the one hand, and conjunctions, on the other, play a fundamental role. While prepositions tighten the relation between the main clause and the adverbial clause, this is

Finite Adverbial Clauses: Iconicity, Subjectification and Mental Spaces

157

not the case with conjunctions. Rather, the latter may be referred to as space builders that guide the interlocutors to a mental space of lesser control. The dislocation from a parent adverbial space to an adverbial mental space with a lesser degree of control is shown in Figure 4:12: Figure 4:12. Conjunctions as space builders.

CONJUNCTIONS INTRODUCING SUBJUNCTIVE CLAUSES

Subjunctive adverbial clause

Infinitive adverbial clause Adverbial space

Lesser degree of control

Thus, conjunctions that introduce subjunctive clauses function as space builders. The conjunction guides the interlocutors to a mental space with a lesser degree of control. The degree of control in the adverbial clause is related to the semantics of the infinitive and the subjunctive, respectively. The holistic and atemporal feature of the infinitive tightens the bond between the main clause and the adverbial clause event. Further, this bond is reflected in the use of a preposition that introduces the adverbial clause. The semantics of the subjunctive, on the other hand, is closely related to a lesser degree of control. The grounding of the event described by this mood alters the perspective, which has the effect of diminishing the responsibility of the main clause subject. Instead, it provides evidence for the mental processes of the primary conceptualizer, i.e., the ground. Finally, the lesser degree of control equates with an event that is outside the dominion of the main clause trajector.

Conclusion The aim of the analysis presented above was to find a conceptual explanation for the occurrence of finite and infinitive adverbial clauses in Portuguese. The initial claim was that these structures may be explained by

158

Chapter Four

the iconic nature of language. In this sense, it was argued that formal complexity would match conceptual complexity, and that formal distance would match conceptual distance. Further, it was argued that the iconic nature of language would provide the basis for the phenomenon of subjectification, manifested in the dislocation from a parent space of origin to a mental space with a higher degree of conceptual complexity, including the mental processes of the primary conceptualizer (the ground). The subsequent analysis showed that the conceptual differences between finite and infinitive adverbial clauses may indeed be explained by iconic principles and notions like subjectification and mental spaces. The conceptual difference between infinitive and indicative clauses was related to a dislocation of the latter to mental spaces that was related to the conceptualizer’s mental processes. Temporal clauses introduced by the conjunction quando are not restricted to temporal contexts, but the conceptual content of the clause may dislocate to mental spaces of habitual or generic circumstantial features. Likewise, causal clauses introduced by the causal conjunction porque do not describe purely causal relations in the outside world, but the conceptualizer’s way of conceiving causal relations. That is, the finite clause dislocates to a mental space of inferential causal relations. Infinitive clauses, in contrast, are more restricted, describing temporal and causal relations. The conceptual content of these clauses stays in the parent space of origin. The conceptual differences between infinitive and subjunctive clauses display a similar pattern. The formally weaker infinitive clauses stay in the original adverbial mental space, while the subjunctive clauses dislocate to a mental space of a lesser degree of control. In this sense, the latter are located outside the dominion of the main clause trajector. When the main clause trajector is equated with the primary conceptualizer, the described event is located outside this participant’s dominion. This being so, the analysis offered in the present chapter provides evidence for a relation between linguistic iconicity and subjectification in Portuguese adverbial clauses. From Traugott’s perspective, a higher degree of formal complexity in the finite clauses corresponds to internalization and propositional attitude. The speaker does not describe events in the outside world. Rather, the conceptual content of the linguistic expression designates the speaker’s perspective on events in the outside world. That is, instead of designating events in the outside world, the linguistic expression designates the speaker´s thoughts and reasoning about events in the outside world. From Langacker’s perspective, this subsumes the mental scanning between two events that may be, but does not have to be, related in the outside world. Further, the effect of grounding is exactly that of creating a relation

Finite Adverbial Clauses: Iconicity, subjectification And Mental Spaces

159

between the ground–as the primary conceptualizer–and the described process. Finally, the grounding of the event entails that the ground is subjectively construed.

CHAPTER FIVE FINAL CONSIDERATIONS

The aim of the present study was to apply a cognitive perspective to the syntax and semantics of adverbial clauses in European Portuguese. In particular, three specific research questions were addressed: the variation between the plain and the inflected infinitive, the function and semantic meaning of the clitic pronoun se in infinitive adverbial clauses and, finally, the variation between infinitive and finite adverbial clauses. The initial claim was made that a cognitive linguistic perspective in general–and the Cognitive Grammar perspective in particular–would be able to provide a conceptually grounded answer to these three research questions.

Summary In the introductory chapter, it was shown that Portuguese, unlike other European national languages, admits adverbial clauses to be sub-divided into three different structures with three different verb forms: plain infinitives, inflected infinitives and finite verb forms. In order to understand the variation between these different verb forms, three central questions were raised. The variation between the inflected and the plain infinitive, on the one hand, and between the infinitive and a finite verb form, on the other hand, is an issue that has already been the subject of several studies. In the present study, however, it was argued that the variation cannot be related to subject co-reference. In fact, it was confirmed that there are many cases in which the inflected infinitive is used in spite of a co-referential subject in the main and the adverbial clause. This linguistic fact gave rise to the first research question: what motivates the variation between the plain and the inflected infinitive in co-referential adverbial clauses? This question, however, was closely related to another issue. The fact that the Portuguese inflected infinitive does not have any morphological markings in the first and third person singular implies a difficulty to distinguish it from the plain infinitive. This problem has traditionally been solved by arguing that the infinitive is inflected if it has a “logical” subject

162

Chapter Five

in the context, and if it is possible to place a nominal subject before the infinitive. In cases like: vou sair para (eu) comprar, the reasoning goes that the infinitive is inflected because it has a logical subject which is expressed by the finite verb in the main clause, and that it is also possible to use a nominal subject in the embedded adverbial clause. Nevertheless, it was shown that this explanation only covers the use of the infinitive in singular contexts, and not in cases like: vamos sair para comprar, where it is also obvious that the action expressed by the infinitive refers to the subject in the main clause. A similar case is the function and use of the clitic se with intransitive verbs. In these contexts, it is traditionally stated that se signals an undetermined subject. But, what happens in the context of a transitive verb or in cases where the inflected form of the infinitive is in the plural? This was the second question raised in the introductory chapter. The third research question was related to the variation between infinitive and finite adverbial clauses. On the one hand, there is a similarity between the clauses in the sense that both signal the presence of a subject. On the other hand, the finite adverbial clauses also exhibit the feature [+ tense/mood]. This difference raised the question as to whether the variation between the infinitive and the finite adverbial clause is related to the need for expressing tense, or whether there may be other reasons for using one or another adverbial clause. After a preliminary discussion concerning the research questions, some earlier research related to these issues was examined. It was concluded, however, that the prevailing perspective in treating these issues has been purely descriptive and formalistic. Traditional grammar has described the different uses of infinitive and finite adverbial clauses without any serious attempt to explain them. Nor does the traditional grammar try to explain the different uses of the clitic pronoun se in its various adverbial contexts. Studies conducted from a generativist perspective, on the other hand, have aimed at setting up specific rules and transformations in order to account for these phenomena. Thus, generative approaches, no more than traditional approaches, were seen as capable of providing a satisfactory answer to the questions raised in the present study. Instead, it was argued that these would be more properly investigated from a cognitive linguistic perspective, which takes its point of departure in Langacker’s (1987, 1991, 2008) model of Cognitive Grammar. Accordingly, the following working hypotheses were formulated: •

What motivates the variation between the plain and the inflected infinitive in co-referential adverbial clauses?

Final Considerations

• •

163

What is the meaning and function of the clitic se in adverbial clauses? What motivates the variation between inflected infinitive adverbial clauses and finite adverbial clauses?

The variation between the plain and the inflected infinitive in adverbial same-subject clauses was the topic of Chapter 2. It was argued that the choice of one or the other type of adverbial construction is closely related to the need to encounter an adequate anaphoric expression, and that the use of an adverbial clause with an inflected infinitive might be explained by a cognitive need to highlight the subject of the adverbial clause. To this end, Langacker’s (1987) concepts of trajector and landmark were used. The grammatical and semantic relation between the trajector and the landmark of a predicative relation was examined. It was shown that the trajector, in prototypical cases, is associated with the subject of the clause, while the latter conforms to the direct object. Both clausal participants have a high degree of prominence, and constitute the focal participants of a described event. However, the fact that the trajector, in most cases, is associated with the most active participant, i.e., the one which represents a force that is capable of changing an event, and the one which is perceptually more salient, entails a higher degree of cognitive salience/prominence in this participant than in the landmark. The latter participant, on the other hand, can be understood as a reference point in relation to the trajector. The concept of elaborated trajector plays a fundamental role in the elaboration of a prominence scale for adverbial clauses. In accordance with the prominence scale, the adverbial clause gains prominence when morphological information about its trajector is being signalled. This can happen by means of the verbal inflection of the infinitive, or by the insertion of a nominal subject in the adverbial proposition. In other words, a phrase such as: para comprar (‘in order to buy’) exhibits a low degree of prominence, while similar phrase with a lexical subject and verbal inflection like: para o Rui e a Ana comprarem (“in order [for] Rui and Ana to buy) stands out as more prominent. In other words, the prominence scale showed that the use of nominal subjects and verbal inflection makes the presence of a trajector in the adverbial clause more salient. The notion of various degrees of prominence in the infinitive adverbial clause was consequently referred to in the empirical analysis, in which a comparison between a pragmatic and a cognitive approach to anaphoric expressions was also made. With regard to the pragmatic approach, it was argued that neither Levinson’s (1987, 1991) principles, nor Sperber and Wilson’s (1990, 1995) Relevance Theory, can fully explain why the more

164

Chapter Five

prominent option will be used in co-referential contexts, considering that this option will lead to a non co-referential interpretation instead of a coreferential one. In contrast, studies on anaphoric expressions conducted within a cognitive approach focus on referents which are not in the centre of focal attention, referent accessibility, and referential activation (cf. Tomlin and Pu, 1991; Ariel, 1991, 1994, 1996; Chafe, 1987, 1994). The latter studies provide important evidence for a correlation between a low degree of referential accessibility and a prominent anaphoric expression. In order to examine some of the factors that may cause a cognitive need to use a prominent adverbial clause, a distinction was established between usage event properties and structural properties. The former are understood as related to the actual planning of discourse, whereas the latter are associated with purely linguistic context. Usage event properties can for example, be pauses, repetitions, and reformulations which reduce the degree of cohesion between the trajector of the main clause and the co-referential adverbial trajector. The structural properties may also produce a reduced degree of cohesion, but in such cases, it is the linguistic units that appear between the main clause trajector and the one of the adverbial clause which result in a weakened cohesion effect. Another utterance factor of importance is when a cataphoric relationship occurs between the trajector of the superordinate and the adverbial clause. In accordance with the cognitive perspective of the present study, it was also explained that there is a strong relation between usage based properties and structural properties. In sum, it was concluded that any of these factors may give rise to the cognitive need for choosing a more prominent adverbial clause. In the following chapter, Chapter 3, the function and semantic meaning of the pronoun se was analysed in different grammatical contexts. It was argued that the classical categorization of se as an “impersonal subject” with intransitive verbs, and as a “passive marker” with transitive verbs, does not take into account a wide array of contextual factors that can lead to one or another interpretation of this clitic. Further, it was argued that the vagueness of this pronoun, among other things reflected in the fact that it does not signal case nor gender, makes the interpretation of se dependent on various contextual factors. A special concern was given to the notion of transitivity. It was claimed that a certain degree of transitivity implies that this clitic may be interpreted as a generic trajector, i.e., an external force that induces a change of an event, even in cases with infinitive inflection in the third person plural. Further, different strategies for defocalization and focus displacement were examined. Thus, it was hypothized that the position of the components (the NP and the clitic se) and the infinitive inflection in third person plural, may

Final Considerations

165

contribute to one or another interpretation of se. In those cases where this clitic is interpreted as a generic trajector, the lack of referentiality creates a defocalization of this participant. In other cases, when the noun phrase is inserted in the initial position, or when the verbal inflection is in the third person plural, a focus displacement takes place. The change of focus weakens the idea of a generic trajector. Nonetheless, it was argued that other contextual factors may be involved in the conception of a generic trajector. In accordance with Maldonado (1992), two basic conditions that can influence the notion of se as a marker of an external force were examined: (1) if the theme itself is capable of inducing an event and (2) if the verb describes an action that can be realized by a non-animate participant. Thus, it was claimed that the idea of se as a marker for a generic trajector increases when the action described by the verb presupposes an external force. On the other hand, it was also argued that the verbal inflection in third person plural with a non-animate NP in initial position would represent communicative strategies for diminishing the role of an external generic trajector. A wide array of adverbial clauses with se was examined with regard to Hopper and Thompson’s (1980) transitivity parameters, and thus it was confirmed that all of these cases present a certain degree of transitivity. Special attention was given to the fact that certain adverbial prepositions (para [‘in order to’] and até [‘until’]) have semantic properties that seem to increase the degree of transitivity. Further, the number of participants in the event described in the adverbial clause, the position of the various components involved, the existence of an obvious referent of se, and the effect of the infinitive inflection in the third person plural was examined. It was shown that the notion of one or more participants being included in the adverbial event depends heavily on the type of action described by the verb. A verb which describes a physical action seems to increase the need to conceptualize se as a participant in that event, whereas this need seems to diminish gradually if the verb describes a mental action or a perceptual experience. Accordingly, a relation between phenomenological aspects of conceptualization and language use was established. The analysis further provided evidence for the claim that the initial position of a component has an impact on how the described event is interpreted. The idea of the initial position being the starting point for any action chain was claimed to explain the tendency to conceptualize the clitic se as a generic trajector when it occupies this position. Moreover, it was shown that a non-animate NP and the third person plural inflexion gradually diminish the idea of a generic trajector being present in the adverbial clauses. Finally, it was concluded that the contextual factors studied may indeed have some influence on our

166

Chapter Five

tendency to conceptualize the clitic pronoun se as a marker for a generic trajector. This being so, a distinction was made between four different structures containing PREPOSITION + INFINITIVE + SE + NP (-ANIMATE) in which the notion of a generic trajector is present to a larger or a lesser degree. The variation between infinitive and finite adverbial clauses was analysed in Chapter 4. The notion of iconicity (Haiman 1980, 1985) played a fundamental role. It was argued that a higher degree of morphosyntactic complexity in the finite adverbial clauses, and a greater formal distance between the main clause event and the adverbial event, can be matched with a greater conceptual complexity and distance. It was also argued that a higher degree of conceptual complexity in the finite adverbial clauses can be explained by Subjectification and Mental Spaces. Subjectification was to be understood both from Traugott’s and Langacker’s frameworks. From the perspective of the latter, this may be expressed by a finite verb that connects the linguistic expression directly to the ground. It was claimed that this phenomenon only represents an initial step. The displacement of the finite adverbial clause’s propositional content from a paternal mental space of origin to another mental space related to reasoning and inference is the second step. The first part of the analysis dealt with the variation between infinitive and finite adverbial clauses that are introduced either by ao/quando or by por/porque. A difference between adverbial clauses introduced by ao or quando was detected to the effect that the former can only be used in punctual contexts, giving rise to a restricted temporal use. Adverbial clauses that are introduced by quando, on the other hand, have a wider array of uses, and can occur in generic-circumstantial and habitual contexts. In these types of contexts, the adverbial clause either expresses a circumstance related to the realization of an event in the superordinate clause, or describes a condition for a habit expressed in it. This being so, it was claimed that adverbial clauses introduced by quando often may lead to the displacement of the propositional content from a temporal mental space of origin to a mental spaces of a generic-circumstantial or habitual type. Likewise, it was shown that adverbial clauses introduced by the preposition por are heavily restricted to prototypical causal relations between the main clause and the adverbial clause. In contrast, this is not the case in adverbial clauses introduced by porque. In several contexts, the relation between the main clause and adverbial clause is of an inferential type. That is, the causal relation between the clauses is not to be detected in the relation between them, but is to be inferred from previous knowledge. The difference between the infinitive and the finite adverbial clauses may

Final Considerations

167

therefore, in the case of por and porque, be explained by the displacement of the propositional content of the latter to an inferential mental space. In contrast, the former stays in a causal mental space of origin. The second part of the analysis focused on adverbial clauses in the subjunctive mood. After examining earlier research on the semantics of the subjunctive mood, it was concluded that notions like non-assertion, distance, relevance, reality (i.e., notions that often have been related to the semantics of the subjunctive mood) even if they grasp some of the meaning of this mood, do not cover the full range of its use. Following Maldonado (1995), it was proposed that the subjunctive designates events that are outside the main clause trajector’s dominion of active control. This loss of control was said to be manifested in two different ways: (1) the content of the adverbial clause does not correspond to the conceptualizer’s elaborated reality, and (2) this participant is not capable of exerting influence on the described event. An examination of different types of adverbial clauses in fact showed that the concept of control can indeed explain the variation between infinitive and finite subjunctive adverbial clauses. In discussing why the finite subjunctive clauses, and not the infinitive ones, exhibit a lesser degree of control, three factors were addressed: (1) the semantics of the subjunctive mood and the infinitive, (2) the occurrence of the subjunctive in non coreferential contexts with factive verbs, (3) the tendency for conjunctions to separate clauses in a clause binding hierarchy, while prepositions frequently are used to put different components together. These factors were further examined. It was claimed that the infinitive, being a nominative verb form, presents a holistic view of an action and therefore focalises its result, which leads to a higher degree of control over the described event. Another typical feature that was commented upon was the relation between infinitives and perception verbs. This relation could imply that the event described by the infinitive belongs to the main clause trajector’s conceived reality and, thus, is located inside his/her dominion. The analysis revealed that the subjunctive mood is used in contexts where the event is located outside this participant’s dominion. In addition, it was also shown that there is an overwhelming tendency to use the subjunctive verb forms in non co-referential contexts with factive verbs. This use was explained by the fact that these contexts presuppose a low degree of control. Finally, there is a difference in the components that serve to introduce the adverbial propositions: a conjunction separates the two clauses, while a preposition unites them. This was shown in the vast number of compound words and possessive relations expressed by prepositions.

168

Chapter Five

In accordance with the preceding analysis, it was further claimed that the difference between infinitive and subjunctive adverbial clauses can be explained by the displacement of the propositional content of the latter to a mental space possessing the feature [- control]. This being so, the present study may provide the following conclusions regarding the different adverbial structures analysed: • The use of prominent infinitive adverbial expression reflects a cognitive need to put focus on a participant that is disappearing in the context, a participant that is not in the speaker’s short-term memory. In short, the uttering of an elaborated trajector is a strategy to maintain the notion of this participant. If there is no such need, the tendency is to use the plain infinitive. • The tendency to conceptualize the clitic pronoun se as a generic trajector in infinitive adverbial clauses does not only depend on the verb inflection. In fact many other contextual factors are capable of resulting in one or another interpretation of this clitic. • The semantic differences between infinitive and finite adverbial clauses may be explained by a higher degree of subjectification in the latter ones. Further, the higher degree of subjectification is related to iconic principles and is manifested through the displacement from a mental space of origin to mental spaces that display a more complex relation between the main clause event and the adverbial event. Arriving at this point, a final question still remains. In the introductory chapter, it was claimed that a cognitive linguistic approach to different adverbial structures in European Portuguese would contribute to a greater understanding of their semantics. That is, where traditional and generative linguistics had partly failed to explain the variation between finite and infinitive structures, and the function of the clitic pronoun se in these structures, the cognitive approach would provide a plausible explanation. Also, this explanation would be built on semantic concerns and not on abstract formalizations or specific rules and transformations. This issue is to be dealt with in the subsequent section.

The Cognitive Approach The Sausurrean dichotomy between langue and parole has had a strong impact on modern linguistic analysis. The tendency during the first half of

Final Considerations

169

the twentieth century was to focus on language as a social object. That is, the centre of attention was related to language as a communicative object taking place in the world. Nonetheless, the pioneer works of Chomsky (cf. Chomsky 1957, 1964) implied a change of focus. Instead of focusing on the communicative side of the Saussurean dichotomy, language was conceived of as a formal system with computational features. Thus, the conception of grammar as an autonomous component without semantic meaning was established in the outset of the generative framework: Despite the undeniable interest and importance of semantic and statistical studies, they appear to have no direct relevance to the problem of determining or characterizing the set of grammatical utterances. I think that we are forced to conclude that grammar is autonomous and independent of meaning…(Chomsky 1957: 17).

According to Chomsky (1957), the fundamental aim in linguistic analysis was to separate all the ungrammatical sequences of a language from the grammatical ones, and thereafter create formal models and theories regarding the syntactic structure of that particular language. The linguistic analysis was said to meet descriptive adequacy when it correctly described the implicit knowledge that adult speakers of a language had in order to speak that particular language. In short, the focus on structure and linguistic competence, rather than on meaning, implied that the notion of syntactic structure as a tool for expressing different meanings was highly dismissed. Meaning was to be found in lexicon, and not in syntactic structures. The requirement for descriptive adequacy was the initial step towards another related question–the one of language acquisition and explanatory adequacy. The latter is achieved when the described grammar of a particular language is derivable from two components: the Universal Grammar and the exposition to language during the language acquisition period (cf. Chomsky 2002). The Universal Grammar equates with the theory of the initial cognitive state of language knowledge of the mind at birth. The reconciliation between descriptive and explanatory adequacy, however, was a far more difficult task than expected. The tension was related to a need for new analytical tools in order to meet the requirements of descriptive adequacy, on the one hand, and the claim of a Universal Grammar, on the other. The expansion of the data (new phenomena in natural languages) was opposed to the postulation of a strong cross-linguistic uniformity. The solution to this apparent problem was found in the theory of principles and parameters. Although earlier versions of Generative Grammar had postulated the existence of a Universal Grammar, they considered it to be a system of language specific rules. In order to account for conflicting

170

Chapter Five

linguistic data, transformational and language specific rules were created. Differences between languages in the surface structure, i.e., the actual linguistic manifestation, were not to be found in the deep structure of the sentence. The problem, however, was how to explain the way in which the language user could induce the language specific rules within the guidelines provided by Universal Grammar. The theory of principles and parameters offered another explanation. Universal Grammar is a system of universal parameters under which no language specific rule system is postulated. Instead, different language structures are computed directly by Universal Grammar under particular parametric choices. For example, the fact that languages may differ in word order is explained by the postulation of empty slots and overt and covert movement. The theory of principles and parameters furnishes the base for the minimalist program, Chomsky’s latest contribution to linguistic analysis (cf. Chomsky 1995). One fundamental aim of the minimalist program is to develop the transformational notions economy of derivation and economy of representation. The tendency is towards more uniformity regarding the derivation of syntactic structures. The earlier notion of surface structure and deep structure is replaced by Phonetic Form (PF) and Logical Form (LF). The revision offered above does not pretend to offer an exhaustive summary of the framework of Generative Grammar. However, it captures some specific features of it. First, the framework of Generative Grammar offers a decontextualised vision of linguistic analysis. Certainly, Chomsky has a point when he states that: “language is not properly regarded a system of communication. It is a system for expressing thought, something quite different” (Chomsky 2002: 76). The statement refers to what Generative Grammar denominates the I-language (internal, individual and/or intentional) and captures the idea that language is a tool for expressing thoughts. However, the framework of Generative Grammar goes one step further, claiming that the I-language is independent of other elements in the world (cf. Chomsky 1995: 15). The consequence is that language is to be analysed in its “pure form”, without paying attention to the context in which it occurs. This vision of linguistic analysis neglects two crucial issues. The first one is related to the fact that language, in fact, is not only a system for expressing thoughts. It has also a communicative function. This facet of linguistic analysis is largely omitted in the framework of Generative Grammar. The second issue bears a relation to the claim that language is a system for expressing thoughts. This claim is not only found in the Generative Grammar, but in many other current linguistic theories. Cognitive Grammar is one of them (cf. Langacker 1987, 1991). A related view is found in the works of Slobin

Final Considerations

171

(cf. Slobin 1996) and the hypothesis of linguistic relativism. The crucial issue, however, is that Generative Grammar does not seek any relation between the conceptualizer and the conceptualization. That is, what is the idealized speaker of a language thinking about? According to Chomsky, the I-language is independent of elements in the world. The conclusion is that the generative approach to linguistic analysis does not need any context. The question of decontextualization is, in first place, related to the requirement of descriptive adequacy in Generative Grammar, even though it also relates to the one of explanatory adequacy. In order to analyse syntactic structure–and the linguistic knowledge of an idealized speaker–there is no need for context. The most striking feature of the explanatory adequacy, on the other hand, is the search for linguistic uniformity. In order to explain why children learn languages, and why the process is so rapid, the generative hypothesis is that we have an innate Universal Grammar–the initial state of birth. The problem is that the different languages in the world present different syntactic structures. Thus, the hypothesis of a Universal Grammar tends to minimize the structural differences between languages: The child interprets the incoming linguistic data through the analytic devices provided by Universal Grammar, and fixes the parameters of the system on the basis of the data…Acquiring a language thus means selecting, among the options generated by the mind, those which match experience, and discarding other options. So, acquiring an element of linguistic knowledge amounts to discarding the other possibilities offered a priori by the mind; learning is then achieved by forgetting… (Chomsky 2002: 16).

In sum, the structural differences that may occur between languages in overt linguistic manifestations are not to be seen in a deeper representation of the same. Universal Grammar provides different parametric choices, and as the child learns the language of contact, the other parametric choices are forgotten. Thus, the apparent diversity of language structure is eliminated by the creation of abstract and deeper levels of representation. Obviously, this approach opens up for a number of questions. The most troublesome one is related to the possibility to falsify the hypothesis of Universal Grammar. If the claimed linguistic uniformity is hidden at a deeper level of representation, that is, if linguistic uniformity is not to be seen on the surface: how may one corroborate or falsify the hypothesis? In fact, recent criticism of Universal Grammar has pointed out the fact that the components of Universal Grammar do not all apply universally (cf. Evans and Levinson 2009). The response to this criticism has been that one cannot see them on the surface level, or that they do not need to apply universally.

172

Chapter Five

The first response illustrates the problem of corroboration/ falsification. The second response leads to the question whether Universal Grammar has any explicatory adequacy, or if it has become an empty concept. The cognitive approach to linguistic analysis offers an alternative to the generative paradigm. To begin with, a fundamental claim is that grammar is symbolic in nature. Lexicon, morphology and syntax represent a continuum of symbolic structures with semantic meaning. Further, the relation between a phonological pole and a semantic pole underscores the idea that there is a direct relation between phonological representation and semantic meaning. The object of linguistic analysis is meaning, equated with conceptualization, and it is directly represented in the symbolization between the phonological pole and the semantic pole. Instead of postulating deep structures or Logical Forms, the analysis focuses on concrete manifestations (in the mind or in the world). Therefore, the enterprise is to analyse structural differences and to find out how these differences may relate to semantic meaning. The term construal captures the basic notion that an event may be conceptualized in different ways, and that the difference is reflected in the construal of the conceptual content. That is, grammar is not only a formal system; it is also a means for describing the conceptualizer’s way of conceiving an event or a situation. If grammar provides different structures to describe an event, these structures must entail different meanings. They must not represent overt manifestations of a semantic meaning that is hidden at a deeper level of representation. Furthermore, the term conceptualizer highlights the relation between meaning and the construction of meaning. That is, the Cognitive Grammar approach distinguishes a relation between the conceptualizer, on the one hand, and the conceptualized entity on the other. In accordance with this perspective, meaning is not just a “thing” out there in the world. It is something that humans create in their daily life. Obviously, the approach is totally opposed to the objectivist tradition, in which facts of the world are independent of “belief, knowledge, perception, modes of understanding, and every other aspect of human cognitive capacities” (cf. Lakoff 1987). In this sense, the Cognitive Grammar perspective provides a recontextualization process. Language is related to language users and cognitive processes. The contextual approach is overtly spelled out in Langacker’s (1987) model of Cognitive Grammar. As was commented upon above, symbolization occurs between the semantic and phonological poles. This symbolization, however, merely constitutes a certain part of the model–the relation between semantic meaning and phonological representation. Another aspect of the model is represented by the codification between Grammar, equated with

Final Considerations

173

linguistic convention, and the usage event. The codification is realized both in the semantic pole and the phonological pole. It may be purely conceptual, without any physical manifestation, or result in the overt vocalization of the conceptual content. Therefore, the model accounts not only for individual conceptualization (in the mind), but also for the overt vocalization of conceptual content. The term usage event covers them both, bridging the gap between langue and parole, on the one hand, and competence and performance, on the other (cf. Geeraerts 2006). Evidently, the analysis offered in previous chapters of the present study is very much committed to the Cognitive Grammar approach to linguistic analysis. Instead of considering grammar as an autonomous component without meaning, the fundamental aim was to explain the occurrence of different adverbial structures in European Portuguese in terms of cognitive processes and conceptualizations. In this sense, the analysis underscored the importance of taking the actual context of the linguistic manifestation into account. For this reason, the object of analysis, adverbial clauses in European Portuguese, was studied on the basis of actual occurrences. In sum, the analysis was committed to the functions and semantics of adverbial clauses in relation to the contexts in which these clauses occurs. The cognitive approach thus leads to new findings about adverbial clauses in European Portuguese. Regarding the variation between the plain and the inflected infinitive, former proposals are strongly committed to the idea that it is, first and foremost, a question of co-referentiality vs. non coreferentiality. It is interesting to note, though, that both Maurer (1968) and Perini (1977) comment the possibility to use the inflected infinitive in cases of a greater linear distance between the main clause verb and the infinitive. However, they do not go any further in explaining why this greater linear distance creates this possibility. Instead, they create rules like the double inflection filter (Perini 1977) or instructions for when to use one or another infinitive (Maurer 1968). Likewise, Bechara (1999) is more concerned with a formal description rather than an explanation for the occurrence of the plain and the inflected infinitive. Moreover, the (formal) pragmatic approach to referential expressions furnishes a decontextualized vision of the subject matter. Levinson (1987) also focuses on the question of co-referentiality vs. non co-referentiality and formulates his principles as instructions to the speaker and the listener. He does not, however, give any motivation for why the discourse participants should follow these instructions. In the same manner, the Relevance Theory (Sperber and Wilson 1995) rules out the use of more prominent adverbial expressions in co-referential contexts given that they do not create any contextual effect.

174

Chapter Five

The analysis presented in Chapter 2 furnished an explanation for the use of the plain and the inflected infinitive in adverbial clauses. To begin with, it was shown that the inflected infinitive, in fact, occurs in co-referential contexts. Thereafter, the analysis focused on the underlying motivation for the use of the inflected infinitive in these contexts. This being so, the inflected infinitive was related to the discourse participants’ need to put focus on the adverbial trajector when the notion of this participant is diminishing due to contextual factors. Special attention was given to notions like referential activation, accessibility and attention. Thus, these notions may explain the cognitive processes of the speaker and the listener, and the need to use a more prominent adverbial clause in co-referential contexts. Obviously, the non co-referential context implies a low degree of referential activation. The main point, however, is that this also may be the case in coreferential contexts. Thus, the cognitive approach to the inflected infinitive furnishes an explanation for its occurrence by considering the contextual factors and cognitive processes that may come into play–not by creating rules. The creation of the prominence scale gives rise to another fundamental issue. The revision of earlier proposals regarding the plain and the inflected infinitive implied a certain difficulty to distinguish between them in the first and the third person singular given the lack of verbal inflection. Thus, earlier analyses tend to the creation of different rules in order to identify the inflected infinitive in these grammatical contexts. However, the problem vanishes completely if we consider the prominence of the infinitive. Instead of making a distinction between the plain and the inflected infinitive, it suffices to consider the Portuguese infinitive in terms of prominence. That is, there is no need to posit two different infinitives, only the fact that it may have a higher or lesser degree of prominence, signaled by inflection and/or an overt noun phrase. The term construal is highly important in relation to the analysis offered in the third chapter. The problem of finding an appropriate interpretation of the function and semantics of the clitic pronoun se was corroborated in the abundance of terms like passive marker vs. impersonal subject, and/or indeterminateness vs. indefiniteness (cf. Cunha and Cintra 1984; Milanez 1982). Thus, it was shown that the occurrence of four different adverbial structures containing se contribute to the notion of this pronoun as a generic trajector to a greater or lesser degree. Further, the analysis of these structures confirmed that the traditional way of looking at se is problematic. There is no basis for making a clear-cut division between se as a passive marker with transitive verbs and as an impersonal subject with intransitive

Final Considerations

175

verbs. In fact, the analysis verified occurrences of structures without plural marking on the verb in plural contexts. From a cognitive approach to linguistic analysis, these structures reflect the relation between linguistic structure and phenomenological aspects. The way in which we conceptualize events in the outside world has an impact on how we describe them. In the analysis, this was related to Maldonado (1992) and to the human propensity for conceptualizing events brought about by internal or external forces. Consequently, the adverbial clauses containing se reflect a certain tension between the claim that se represents a passive marker with transitive verbs, on the one hand, and human conceptualizing processes, on the other. The occurrences of non-inflection in plural contexts reveal a tendency to conceptualize se as a generic trajector also in transitive contexts. That is, the cognitive approach highlights the relation between the conceptualizer of an event and the way in which he/she describes the event. It also accounts for the other contextual features, such as the type of event the adverbial clause designates. Thus, the common denominator for the analyses presented in Chapter 2 and 3 is that they question an assumed dichotomy. The elaboration of the prominence scale in Chapter 2 may unify the notion of the Portuguese infinitive. Given that there is a difficulty to make an accurate distinction between the plain and the inflected infinitive, it would certainly be more natural to consider the prominence of the infinitive. Likewise, the analysis of the clitic pronoun se confirmed the problem of making a distinction between the passive reading and the impersonal reading. An alternative is to study how different structures contribute to evoking the notion of a generic trajector–or to minimizing the notion of it. The introductory chapter revealed a great tendency for applying a Generative Grammar approach in analysing finite and infinitive clauses in Portuguese. Perini (1977) elaborates the insertion of que rule in non coreferential contexts, affirming that the tense feature in the deep structure is omitted in the surface structure with the infinitive. Further, Raposo (1975, 1987) argues that the infinitive belongs to the same category as the complementizer que. Both analyses are based on the idea that the difference between finite and infinitive structures is related to transformations from the deep structure to the surface structure. Obviously, the cognitive approach to linguistic analysis deviates substantially from this vision of language. The present analysis provided evidence for the iconic nature of language: conceptual differences between finite and infinitive adverbial clauses are seen in the structural differences. A fundamental difference was related to the effect of grounding. The fact that finite verbs create a relation to the ground can explain why they are

176

Chapter Five

used to describe mental experiences and reasoning. The grounding of the event was further understood as the foundation for subjectification, i.e., the mental scanning of the conceptualizer and a higher degree of internalization and propositional attitude. Thus, the analysis confirmed some conceptual differences between finite and infinitive adverbial clauses, instead of deriving them from the same deep structure. These differences were related to the notion of mental spaces–the finite structures were dislocated from a parent mental space of origin to mental spaces with a higher degree of conceptual complexity. On the other hand, the infinitive structures are more restricted to the mental space of origin. In sum, the analysis was based on notions like grounding, subjectification, dominion and mental spaces–notions that are anchored in a conceptualist vision of language and linguistic analysis. To sum up, every analysis is guided by the perspective in which it is conducted. If the aim is to prove that human beings possess a Universal Grammar, and that all languages are instantiations of Universal Grammar, the analyst will have to create underlying linguistic structures in order to explain language diversity. Generative Grammar is committed to this approach. Consequently, focus is put on the analysis of grammar as a formal device to the detriment of semantic analysis. On the other hand, we may conceive of language as a fundamental device for creating meaning. In this sense, linguistic analysis should be devoted to the mechanisms behind this apparatus, and related to them. As Langacker points out: The most fundamental issue in linguistic theory is the nature of meaning. I take it as self-evident that meaning is a cognitive phenomenon and must eventually be analyzed as such. Cognitive Grammar therefore equates meaning with conceptualization…explicated as cognitive processing… (Langacker 1987: 5)

The aim of the present analysis has been to follow these guidelines, offering a conceptual approach to the Portuguese infinitive, to the clitic pronoun se and to finite verb forms in adverbial clauses of European Portuguese. In doing so, focus has been put on the relation between the conceptualizer, the context and the meaning and function of different adverbial structures.

REFERENCES

ACHARD, MICHEL. 1998. Representation of Cognitive Structures – Syntax and Semantics of French Sentential Complements, Berlin/New York: Mouton de Gruyter. ARIEL, MIRA. 1991. The function of accessibility in a theory of grammar. Journal of Pragmatics 16, 443-463. —. 1994. Interpreting anaphoric expressions: a cognitive versus a pragmatic approach. Journal of Linguistics 30, 3-42. —. 1996. Referring Expressions and the +/- Coreference Distinction”. In: Thorstein Fretheim and Jeanette K. Gundel (eds.) Reference and Referent Accessibility. Amsterdam/Philadelphia: John Benhamins Publishing Company, 13-36. ATHANASIADOU, ANGELIKI, COSTAS CANAKIS and BERT CORNILLIE .2006. Introduction. In: Angeliki Athanasiadou, Costas Canakis and Bert Cornillie (eds.) Subjectification Various Paths to Subjectivity. Berlin/New York: Mouton de Gruyter, 1-13. BACELAR DO NASCIMENTO, MARIA FERNANDA, MARIA LÚCIA GARCIA MARQUES and MARIA LUÍSA SEGURA DA CRUZ. 1987. Português Fundamental. Vol. II, Métodos e Documentos, tomo 1, Inquérito de Frequência. Lisbon: instituto Nacional de Investigação Científica / Centro de Linguística da Universidade de Lisboa. BECHARA, EVANILDO. 1999 Moderna Gramática Portuguesa. Rio de Janeiro: Lucerna. BERMÚDEZ, FERNANDO. 2003. Llegando a la conclusión: la escena del camino en los conectores consecutivos”. Revue Romane 38-2, 239-271. BLACKWELL, SARAH E. 2000. Anaphora interpretations in Spanish utterances and the neo-Gricean pragmatic theory. Journal of Pragmatics 32, 389-424. BRITO, ANA MARIA. 1995. Algumas considerações sintácticas do Português no quadro das línguas românicas: Sujeito nulo, Infinitivo Flexionado e Clíticos Nominativos. Lusorama 27, 17-27. CAETANO SILVEIRA, JANE R, LUCIENE SIMÕES, SABRINA ABREU GISELA COLLISHONN and DELZIMAR LIMA. 1994. O infinitivo flexionado em Português: um reestudo de Raposo (87). Letras de Hoje 96, 135-146. CARLSON, GREG N. 1977. Generics and the atemporal when. Linguistics and philosophy 3, 49-98.

178

References

CARECHO, JUDITE. 1996. Sobre a semântica das construções com «quando». Lisbon, Faculdade de Letras da Universidade de Lisboa, MA thesis. CASTELEIRO, JOÃO M. and MARIA FERNANDA BACELAR DO NASCIMENTO. 2001. Português Falado, Documentos Autênticos. Lisbon: CLUL and Instituto Camões. CHAFE, WALLACE. 1987. Cognitive constraints on information flow. In: Russel S. Tomlin, (ed.) Coherence and Grounding in Discourse. Amsterdam/ Philadelphia: John Benjamins Publishing Company, 21-52. —. 1994. Discourse, Consciousness and Time – The flow and displacemente of conscious experience in speaking and writing. Chicago/London: University of Chicago Press. —. 1996. Inferring identifiability and accessibility. In: Thorstein Fretheim and Jeanette K. Gundel (eds.) Reference and Referent Accessibility. Amsterdam/ Philadelphia: John Benjamins Publishing Company, 37-46. CHOMSKY, NOAM. 1957. Syntactic Structures. The Hague/Paris: Mouton. —. 1967. Current Issues in Linguistic Theory. The Hague: Mouton. —. 1981. Lectures on Government and Binding, Dordrecht, Foris. —. 1995. The Minimalist Program. Cambridge, Massachusetts, London: The Mit Press. —. 2000. New Horizons in the Study of Language and Mind. Cambridge: Cambridge University Press. —. 2002. On Nature and Language. Cambridge: Cambridge University Press COWAN, NELSON. 1988. Evolving Conceptions of Memory Storage, Selective Attention, and Their Mutual Constraints Within the Human Information-Processing System. Psychological Bulletin 104, 163-191. CUNHA, CELSO and LINDLEY CINTRA. 1984. Nova Gramática do Português Contemporâneo, Lisbon: Edições de João Sá da Costa. DANCYGIER, BARBARA and EVE SWEETSER. 2000. Constructions with if, since and because: Causuality, epistemic stance, and clause order. In: Bernd Kortmann and Elizabeth Couper-Kuhlen (eds.) Cause, Condition, Concession, Contrast: Cognitive and Discourse Perspectives. Berlin: Mouton de Gruyter, 111-142. DECLERCK, RENAAT. 1996a. Tense choice in adverbial when clauses. Linguistics 34, 225-261. DECLERCK, RENAAT. 1996b. A functional typology of English whenclauses. Functions of Language 3-2, 185-234. DE MELLO, GEORGE. 1995. Preposición + Sujeito + Infinitivo: ‘Para yo hacerlo’. Hispania 78, 825-836.

A Cognitive Approach to Adverbial Subordination in European Portuguese 179

DE SMET, HENRIK and JEAN-CHRISTOPHE VERSTRAETE. 2006. Coming to terms with subjectivity. Cognitive Linguistics 17-3, 365-392. FANT, LARS. 1985. Procesos anafóricos y valor enfático en el español hablado. Español Actual 43, 5-26. FAUCONNIER, GILLES. 1990. Domains and connections”. Cognitive Linguistics 1-1, 151-174. —. 1994. Mental Spaces – Meaning Construction in Natural Language. Cambridge: Cambridge University Press. FAUCONNIER, GILLES and EVE SWEETSER. 1996. Spaces, Worlds and Grammar. Chicago: University of Chicago Press. FERREIRA, PAOLO M. 1984. Algumas considerações sobre o conjuntivo nas línguas Românicas”. In: José G. Herculano de Carvalho and Jürgen Schmidt-Radefeldt (eds.) Estudos de linguística portuguesa (vol 1). Coimbra: Coimbra Editora, 257-392. FOX, BARBARA. 1982. Figure-ground in language: a study of several topic continuity devices in Chamorro. Glossa 16-2, 149-180. GARCÍA, ERICA. 1975. The Role of Theory in Linguistic Analysis: The Spanish Pronoun System. Amsterdam/Oxford/New York: North-Holland Publishing Company. GEERAERTS, DIRK. 2006. Introduction: A rough guide to Cognitive Linguistics. In: Dirk Geeraerts (ed.) Cognitive Linguistics: Basic Readings. Berlin/New York: Mouton de Gruyter, 1-28. GIVÓN, TALMY. 1983. Topic Continuity in Discourse: a Quantitative CrossLanguage Study. Amsterdam/Philadelphia: John Benjamins Publishing Company. —. 1992. The grammar of referential coherence as mental processing instructions. Linguistics 30-1, 5-55. —. 1993. English Grammar: A function-Based Introduction, Vol. II. Amsterdam/ Philadelphia: John Benjamins Publishing Company. —. 1995 Coherence in text vs. coherence in mind. In: Morton A. Gernsbacher and Talmy Givón (eds.) Coherence in spontaneous text. Amsterdam/Philadelphia: John Benjamin Publishing Company, 59-116. —. 2001. Syntax, Vol. 2. Amsterdam/Philadelphia: John Benjamins Publishing Company. GUNDEL, JEANETTE K., NANCY HEDBERG and RON ZACHARSKY. 1993. Cognitive status and the form of referring expressions in discourse. Language 69-2, 274-307. HAIMAN, JOHN. 1980. The Iconicity of Grammar: Isomorphism and Motivation. Language 56-3, 515-540. —. 1985. Iconicity in Syntax, Amsterdam/Philadelphia: John Benjamins Publishing Company.

180

References

HAVU, JUKKA. 1997. Acerca del uso de los tiempos verbales en las proposiciones subordinadas de tiempo. Neuphilologische Mitteilungen 4, 365-378. HOPPER, PAUL J. and SANDRA A. THOMPSON. 1980. Transitivity in Grammar and Discourse. Language 56-2, 251-299. HOPPER, PAUL J. and SANDRA A. THOMPSON. 1985. The iconicity of universal categories nouns and verbs”. In: John Haiman (ed.) Iconicity in Syntax. Amsterdam/Philadelphia: John Benjamins Publishing Company, 151-186. HUANG, YAN. 2000. Discourse anaphora: Four theoretical models. Journal of Pragmatics 32, 151-176. IKEDA, SUMIKO. 1977. A função do pronome se. PUC-SP, MA thesis. JORNAL-ELAN. 2003. Endereço electrónico: . KORTMANN, BERND. 1997. Adverbial Subordination. A Typology and History of Adverbial Subordinators Based on European Languages Berlin/New York: Mouton de Gruyter. KORTMANN, BERND and ELIZABETH COUPER-KUHLEN (2000) Cause, Condition, Concession, Contrast: Cognitive and Discourse Perspectives, Berlin/New York, Mouton de Gruyter. KÖNIG, EKKEHARD and PETER SIEMUND. 2000. Causal and concessive clauses: Formal and semantic relations. In: Elizabeth Couper-Kuhlen and Berndt Kortmann (eds.) Cause, Condition, Concession, Contrast, 341-360. Berlin/New York: Mouton de Gruyter. LAKOFF, GEORGE. (1987) Women, Fire, and Dangerous Things. Chicago: University of Chicago Press. LAKOFF, GEORGE and MARK JOHNSON (1980) Metaphors we live by. Chicago: University of Chicago Press. LANGACKER, RONALD W. 1985. Observations and speculations on subjectivity. In: John Haiman, J. (ed.) Iconicity in Syntax. Amsterdam/Philadelphia: John Benjamins Publishing Company, 109150. —. 1987. Foundations of Cognitive Grammar, Vol. 1 – Theoretical Prerequisites. Stanford, California: Stanford University Press. —. 1988a. An overview of Cognitive Grammar. In: Brygida Rudzka-Ostyn (ed.) Topics in Cognitive Linguistics. Amsterdam/Philadelphia: John Benjamins Publishing Company, 3-48. —. 1988b. A view of linguistic semantics. In: Brygida Rudzka-Ostyn (ed.) Topics in Cognitive Linguistics. Amsterdam/Philadelphia: John Benjamins Publishing Company, 49-90.

A Cognitive Approach to Adverbial Subordination in European Portuguese 181

—. 1988c. The nature of grammatical valence. In: Brygida Rudzka-Ostyn (ed.) Topics in Cognitive Linguistics. Amsterdam/Philadelphia: John Benjamins Publishing Company, 91-126. —. 1988d. A Usage-Based model. In: Brygida Rudzka-Ostyn (ed.) Topics in Cognitive Linguistics. Amsterdam/Philadelphia: John Benjamins Publishing Company 127-161. —. 1990. Subjectification. Cognitive Linguistics 1-5, 5-38. —. 1991. Foundations of Cognitive Grammar, Vol. 2 – Descriptive application. Stanford, California: Stanford University Press. —. 1993. Reference point constructions. Cognitive Linguistics 4-1, 1-38. —. 1999. Loosing control: grammaticization, subjectification and transparency. In: Andreas Blank and Peter Koch (eds.) Historical Semantics and Cognition. Berlin/New York: Mouton de Gruyter, 147175. —. 2000. Grammar and Conceptualization, Berlin/ NewYork, Mouton de Gruyter. —. 2001. Discourse in Cognitive Grammar. Cognitive Linguistics 12-2, 143-188. —. 2003. Extreme subjectification: English tense and modals. In: Hubert Cuyckens, Thomas Berg, René Dirven and Klaus-Uve Panther (eds.), Motivation in language. Studies in honor of Günter Radden. Amsterdam/Philadelphia: John Benjamins Publishing Company, 3-26. —. 2006. Subjectification, grammaticization and conceptual archetypes. In: Angeliki Athanasiadou, Costas Canakis and Bert Cornillie (eds.), Subjectification. Various Paths to Subjectivity. Berlin/New York: Mouton de Gruyter, 17-40. —. 2008. Cognitive Grammar: A Basic Introduction. New York: Oxford University press. LEVINSON, STEPHEN C. 1983. Pragmatics. Cambridge: Cambridge University Press. —. 1987. Pragmatics an the grammar of anaphora: a partial pragmatic reduction of Binding and Control phenomena. Journal of Linguistics 23, 379-434. —. 1991. Pragmatic reduction of the Binding Conditions revisited. Journal of Linguistics 27, 107-161. LEVINSON, STEPHEN AND NICHOLAS EVENS. 2009. The myth of language universals: Language diversity and its importance for cognitive science. Behavioral and Brain Sciences 32-5, 429-492. LINGUATECA. 2005. . LUNN, PATRICIA. 1987. The Semantics of Por and Para. Bloomington: Indiana University Linguistics Club.

182

References

—. 1989. Spanish mood and the prototipe of assertability”. Linguistics 27, 687-702. MALDONADO, RICARDO. 1988. Energetic reflexives in spanish”. Proceedings of the anual Meeting of the Berkley Linguistics Society, 153-165. —. 1992. Middle Voice: The case of Spanish se. San Diego: University of California dissertation. —. 1995. Middle-Subjunctive links. In: Peggy Hashemipour, Ricardo Maldonado and Margaret van Naerssen (eds.) Studies in Language Learning and Spanish Linguistics in honor of Tracy D. Terrel. New York: McGraw Hill, 399-418. MANSFIELD, TRACY C. 1997. Prominence: from Sensation to Language. San Diego: University of California dissertation. MARQUES, RUI. 1995. Sobre o Valor dos Modos Conjuntivo e Indicativo em Português. Lisbon: Faculdade de Letras da Universidade de Lisboa, MA thesis. MAURER JR, THEODORO. H. 1951. Dois problemas da língua portuguesa O infinitivo pessoal e o pronome se. São Paulo: Boletim da faculdade de letras. —. 1968. O infinitivo flexionado Português (estudo histórico-descritivo). São Paulo: Companhia Editora Nacional. MEJÍAS-BIKANDI, ERRAPEL. 1993. Syntax, discourse and acts of mind: a study of the indicative/subjunctive in Spanish. San Diego: University of California dissertation. —. 1995. Presupposition inheritance and mood in Spanish. In: Peggy Hashemipour, Ricardo Maldonado and Margaret van Naerssen (eds.) Studies in Language Learning and Spanish Linguistics in honor of Tracy D. Terrel. New York: McGraw Hill, 375-384. —. 1996. Space Accessibility and Mood in Spanish. In: Gilles Fauconnier and Eve Sweetser (eds.) Spaces, worlds and grammar. Chicago: University of Chicago Press, 157-178. MILANEZ, WÂNIA. 1982 Recursos de indeterminação do Sujeito. Campinas: UNICAMP, MA thesis. NUYTS, JAN. 2001. Subjectivity as an evidential dimension in epistemic modal expressions. Journal of Pragmatics, 383-400. NUNES, JAIRO. M. 1990. O famigerado SE: uma análise sincrônica e diacrônica das construções com SE apassivador e indeterminador. Campinas: Universidade Estadual de Campinas, MA thesis. OLIVEIRA, TERESA. 2001. Valores do conjuntivo em construções sintacticamente dependentes. In: Maria Helena Mira Mateus and Clara

A Cognitive Approach to Adverbial Subordination in European Portuguese 183

Nunes Correia (eds.) Saberes no tempo: homenagem a Maria Henriqueta Costa Campos. Lisbon: Colibri, 417-425. OVERSTEEGEN, ELEONORE. 1997. On the Pragmatic Nature of Causal and Contrastive Connectives. Discourse Processes 24, 51-85. PANDER MAAT, HENK and LIESBETH DEGAND. 2001. Scaling causal relations and connectives in terms of speaker involvement. Cognitive Linguistics 12-3, 211-245. PANDER MAAT, HENK and TED SANDERS. 2000. Domains of use or subjectivity? The distribution of three Dutch causal connectives explained”. In: Bernd Kortmann and Elizabeth Couper-Kuhlen (eds.) Cause, Condition, Concession, Contrast: Cognitive and Discourse Perspectives. Berlin/New York: Mouton de Gruyter, 57-72. PANDER MAAT, HENK and TED SANDERS. 2001. Subjectivity in causal connectives: An empirical study of language in use. Cognitive Linguistics 12-3, 247-273. PERINI, MÁRIO A. 1977. Gramática do infinitivo Português. Petrópolis: Editora Vozes Ltda. PIT, MIRNA. 1997. Subjective and objective construal in coherence relations. In: Jan Don and Ted Sanders (eds.) Yearbook 1997. Utrecht: Utrecht institute of Linguistics, 45-70 —. 1998. Gramática descritiva do Português. São Paulo: Editora Ática. QUESADA, DIEGO J. 1997. Obituary: Adios to Passive in Spanish”. La linguistique – revue de la société internationale de linguistique fonctionelle, 41-62. —. 1998. Transitivity, Voice, And The Middle – Spanish Se Revisited. Romanishe Forshungen: Vierteljahrsshrift für romanishe Sprachen und Literaturen, 1-36. RAPOSO, EDUARDO. 1975. Uma restrição derivacional global sobre o infinitivo em Português. Lisbon: Universidade de Lisboa. —. 1987. Case theory and Infl-to-Comp: The inflected infinitive in European Portuguese. Linguistic Inquiry 18-1, 85-109. RUDZKA-OSTYN, BRYGIDA. 1988. Semantic extension into the domain of verbal communication”. In: Brygida Rudzka-Ostyn (ed.) Topics in Cognitive Linguistics. Amsterdam/Philadelphia: John Benjamins Publishing Company, 507-554. SCIDA, EMILY. 2004. The Inflected Infinitive in Romance Languages. New York/London: Routledge. SILVA, SOARES DA, AUGUSTO. 1997. A linguística cognitiva: uma breve introdução a um novo paradigma em linguística. Revista Portuguesa de Humanidades. Estudos Linguísticos 1, 59-101.

184

References

—. 2004. Imagery in Portuguese causation/perception constructions. In: Barbara Lewandowska-Tomaszczyk and Alina Kwiatkowska (eds.) Imagery in Language. Festschrift in Honour of Professor Ronald W. Langacker. Frankfurt/Main: Peter Lang, 2004, 297-319. —. 2005. Between finite clauses and infinitival clauses: The Portuguese inflected infinitive and its conceptual basis. Communication presented at the conference From Grammar to Mind: Grammar and Cognition, Bordeaux. —. 2008. The Portuguese inflected infinitive and its conceptual basis. In: Barbara Lewandowska-Tomaszczyk (ed.) Asymmetric events. Converging evidence in language and communication research 11. Amsterdam/Philadelphia: John Benjamins Publishing Company, 227243. SMITH, MICHAEL B. 2002. The polisemy of German es, iconicity, and the notion of conceptual distance. Cognitive Linguistics 13-1, 67-112. SPERBER, DEIRDRE and DAN WILSON. 1995. Relevance: Communcication and Cognition. Oxford: Blackwell Publishing. SWEETSER, EVE. 1996. Mental Spaces and the Grammar of Conditional Constructions. In: Gilles Fauconnier and Eve Sweetser (eds.) Spaces, worlds and grammar. Chicago: University of Chicago Press, 318-333. SLOBIN, DAN I. 1996. From “thought and language” to “thinking for speaking”. In: John J. Gumpertz and Stephen C. Levinson (eds.) Rethinking linguistic relativism. Cambridge: Cambridge University Press, 70-96. TALMY, LEONARD. 1978. Figure and Ground in Complex Sentences. In: Joseph Greenberg (ed.) Universals of human language. Stanford, California: Stanford University Press. —. 1988. Force dynamics in language and cognition. Cognitive science 12, 49-100. —. 2000. Toward a Cognitive Semantics, Vol. 1 Concept Structuring Systems. Cambridge/Massachusetts: The MIT press. TLÂSKAL, JAROMIR. 1984. Observações sobre tempos e modos em Português. In: José G. Herculano de Carvalho and Jürgen SchmidtRadefeldt (eds.) Estudos de linguística portuguesa (vol 1). Coimbra: Coimbra Editora, 257-392. TOMLIN, RUSSEL S. 1987. Linguistic reflection of cognitive events. In: Tomlin, R.S. (ed.) Coherence and Grounding in Discourse. Amsterdam/Philadelphia: John Benjamins Publishing Company, 455480. TOMLIN, RUSSEL S. and MING-MING PU. 1991. The management of reference in Mandarin”. Cognitive Linguistics 2-1, 65-95.

A Cognitive Approach to Adverbial Subordination in European Portuguese 185

TRAUGOTT, ELIZABETH. 1989. On the rise of epistemic meaning in English: An example of subjectification in semantic change. Language 65-1, 3155. —. 1995. Subjectification in grammaticalisation. In: Dieter Stein and Susan Wright (eds.), Subjectivity and subjectivisation. Cambridge: Cambridge University Press, 31-55. —. 1996. Subjectification and the development of epistemic meaning: The case of promise and threaten. In: Toril Swan and Olaf Jansen Westvik (eds.), Modality in Germanic Languages. Historical and Comparative Perspectives. Berlin/New York: Mouton de Gruyter, 185-210. —. 1999. The rhetoric of counter-expectation in semantic change: a study in subjectification. In: Andreas Blank and Peter Koch (eds.) Historical Semantics and Cognition. Berlin/New York: Mouton de Gruyter, 177196. —. 2003. From subjectification to intersubjectification. In: Raymond Hickey (ed.), Motives for Language Change. Cambridge: Cambridge University Press, 124-139. TRAUGOTT, ELISABETH and RICHARD B. DASHER. 2001. Regularity in Semantic Change. Cambridge: Cambridge University Press. TRAVIS, CATHERINE E. 2003. The semantics of the Spanish subjunctive: Its use in the natural semantic metalanguage. Cognitive Linguistics 14-1, 47-69. WALLACE, STEPHEN. 1982. Figure and Ground: The interrelationships of linguistic categories. In: Paul J. Hopper (ed.) Tense-Aspect: Between semantics and pragmatics. Amsterdam/Philadelphia: John Benjamins Publishing Company, 201-223. VERHAGEN, ARIE. 1995. Subjectification, syntax and communication. In: Dieter Stein and Susan Wright (eds.), Subjectivity and subjectivisation. Cambridge: Cambridge University Press, 103-128. —. 2000. Concession implies causality, though in some other space. In: Elizabeth Couper-Kuhlen and Berndt Kortmann (eds.) Cause, Condition, Concession, Contrast, 361-380. Berlin/New York: Mouton de Gruyter. —. 2005. Constructions of intersubjectivity, discourse, syntax and cognition. New York: Cambridge University Press. VESTERINEN, RAINER. 2007. Complementos finitos e infinitivos dos verbos perceptivos ver, ouvir e sentir: iconicidade linguística e subjectificação. Revista Portuguesa de Humanidades 11-1, 251-283. —. 2008. Direct, indirect and inferred causation: Finite and Infinitive complements of Deixar and Fazer. Journal of Portuguese Linguistics 71, 23-50.

186

References

—. 2010a. The relation between Iconicity and Subjectification in Portuguese complementation: Complements of perception and causation verbs. Cognitive Linguistics 21-3, 573-600. —. 2010b. Uma aproximação cognitiva ao modo conjuntivo. Revista Portuguesa de Humanidades. Estudos Linguísticos 14-1, 151-174. —. Forthcoming. Iconicity, Subjectification and Dominion in Portuguese concessive clauses: conceptual differences between concessive clauses introduced by Apesar de and Embora. International Journal of Cognitive Linguistics. WILCOX, SHERMAN. 2004. Cognitive iconicity: conceptual spaces, meaning, and gesture in signed languages. Cognitive Linguistics 15-2, 119-142. WILSON, DAN and DEIRDRE SPERBER (1990) Outline of Relevance Theory. Hermes 5, 35-56. VOGT, CARLOS. 1976-77. Indicações para uma análise semântica argumentativa das conjunções porque, pois e já que. Alfa 22-23, 139155.

INDEX

Name Index Achard, 121, 143 Ariel, 54, 62, 64, 65, 66, 164 Athanasiadou et al., 122 Bechara, 11, 12, 16, 17, 18, 20, 173 Bermúdez, 138 Blackwell, 57 Brito, 11 Caetano Silveira et al., 11, 15 Carecho, 129 Carlson, 128 Casteleiro, 36 Chafe, 37, 54, 62, 63, 64, 164 Chomsky, 24, 54, 115, 169, 170, 171 Cowan, 63 Cunha, 11, 12, 15, 16, 18, 20, 174 Dancygier, 132 De Mello, 4 Declerck, 128, 129 Fant, 52 Fauconnier, 122, 123, 124 Ferreira, 141 Fox, 52 García, 83 Geeraerts, 173 Givón, 52, 54, 85, 95, 98, 121 Gundel et al., 57, 58 Haiman, 34, 97, 116, 166 Havu, 129 Hopper, 88, 89, 109, 116, 165 Huang, 57 Ikeda, 17 Johnson, 117 Kortmann, 116 König, 155 Lakoff, 117, 172

Langacker, 1, 11, 21, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 35, 37, 43, 46, 48, 52, 54, 68, 76, 93, 97, 118, 119, 121, 122, 143, 149, 154, 159, 162, 163, 166, 170, 172, 176 Levinson, 40, 54, 56, 57, 58, 59, 60, 62, 76, 163, 171, 173 Lunn, 138, 141 Maldonado, 85, 86, 93, 94, 99, 142, 143, 165, 167, 175 Mansfield, 41 Marques, 141 Maurer, 7, 11, 13, 14, 20, 34, 173 Mejías-Bikandi, 126, 127 Milanez, 18, 174 Nunes, 18 Nuyts, 121 Oliveira, 141 Oversteegen, 134 Pander Maat, 121, 134 Perini, 11, 14, 15, 17, 18, 19, 20, 34, 173, 175 Pit, 121 Quesada, 85, 91, 92, 94, 109 Raposo, 11, 14, 15, 18, 19, 20, 34, 175 Rudzka-Ostyn, 43 Scida, 2 Silva, 32, 113, 121 Slobin, 170 Smet, 121 Smith, 116 Sperber, 40, 60, 61, 76, 163, 173 Sweetser, 126, 132 Talmy, 28, 48

188

Index

Tlâskal, 141 Tomlin, 54, 62, 63, 164 Traugott, 118, 119, 122, 159, 166 Travis, 141 Wallace, 51, 52

Verhagen, 119, 121, 155 Vesterinen, 121, 136, 140, 155 Wilcox, 116 Wilson, 40, 60, 61, 76, 163, 173 Vogt, 36, 135

Subject Index accessibility, 62, 64, 65, 66, 67, 73, 76, 126, 127, 164, 174 activation, 62, 63, 77, 98, 164, 174 anaphoric expression, 54, 67, 76, 163, 164 attention, 12, 27, 28, 32, 43, 46, 48, 50, 51, 53, 60, 62, 66, 67, 76, 77, 99, 101, 106, 174 base, 30, 31, 33 basic reality, 143, 149 clitic pronoun, 7, 8, 12, 15, 16, 17, 18, 20, 34, 42, 53, 77, 79, 80, 81, 93, 104, 108, 109, 161, 166, 168, 174, 175 Cognitive Grammar, 1, 11, 21, 22, 23, 24, 25, 26, 29, 30, 31, 32, 33, 37, 42, 43, 44, 45, 46, 47, 48, 62, 68, 77, 81, 87, 93, 99, 110, 161, 162, 170, 172, 173, 176 conceptual complexity, 34, 115, 118, 128, 139, 146, 158, 166, 176 conceptual distance, 34, 117, 118, 139, 146, 158 conceptualization, 23, 24, 26, 29, 47, 68, 69, 71, 79, 82, 83, 93, 97, 99, 104, 117, 118, 119, 165, 171, 172, 173 conceptualizer, 24, 26, 32, 48, 56, 60, 85, 86, 90, 95, 103, 115, 119, 120, 121, 122, 126, 127, 128, 134, 135, 136, 137, 138, 140, 142, 143, 144, 149, 150, 151, 153, 155, 156, 157, 158, 159, 167, 171, 172, 175, 176

context, 4, 8, 13, 14, 26, 29, 34, 37, 41, 45, 51, 54, 58, 59, 61, 62, 64, 68, 69, 70, 72, 73, 76, 77, 80, 106, 109, 110, 124, 126, 131, 138, 146, 151, 152, 162, 164, 168, 170, 171, 173, 174, 176 control, 92, 109, 122, 127, 141, 143, 144, 145, 146, 147, 149, 151, 152, 154, 156, 157, 158, 159, 167, 168 deep structure, 14, 18, 19, 115, 170, 175, 176 domain, 29, 30, 31, 33, 34, 119, 122, 123, 154 dominion, 140, 142, 143, 144, 149, 150, 151, 156, 158, 159, 167, 176 elaborated reality, 143, 151, 153, 167 encyclopedic, 29, 32, 33, 87 figure, 28, 31, 33, 43, 47, 48, 54, 76 focal attention, 63, 64, 164 focus, 28, 30, 32, 33, 41, 43, 46, 47, 48, 50, 51, 53, 62, 63, 68, 69, 71, 77, 81, 83, 84, 85, 87, 96, 99, 101, 103, 104, 106, 108, 109, 110, 164, 165, 168 formal complexity, 34, 35, 118, 139, 146, 158, 159 formal distance, 14, 34, 66, 74, 75, 117, 118, 139, 146, 158, 166 Generative Grammar, 11, 169, 170, 176

A Cognitive Approach to Adverbial Subordination in European Portuguese 189 generic trajector, 34, 53, 79, 80, 81, 82, 83, 85, 86, 87, 90, 92, 93, 94, 95, 97, 98, 101, 102, 103, 104, 106, 107, 108, 110, 164, 165, 168, 174, 175 grammatical structure, 27, 47 ground, 28, 31, 33, 43, 48, 54, 76, 119, 120, 121, 122, 138, 140, 143, 149, 151, 158, 159, 166, 175 grounding, 113, 120, 121, 122, 130, 139, 151, 158, 159, 175, 176 iconicity, 111, 115, 116, 117, 118, 128, 159, 166 indicative, 114, 126, 127, 128, 140, 141, 142, 149, 151, 153, 158 inflected infinitive, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 19, 20, 33, 39, 53, 58, 59, 60, 106, 113, 161, 162, 163, 173, 174, 175 landmark, 31, 32, 33, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 54, 76, 82, 101, 103, 139, 163 Levinson’s principles, 56, 58, 59, 60, 62, 76 Logical Form, 115, 170 mental space, 115, 122, 123, 124, 125, 126, 128, 134, 139, 140, 144, 149, 157, 158, 159, 166, 167, 168, 176 mental spaces, 111, 115, 122, 123, 124, 126, 128, 132, 133, 134, 144, 158, 166, 168, 176 Phonetic Form, 115, 170 phonological pole, 25, 33, 172 pragmatic, 40, 41, 42, 54, 56, 57, 58, 60, 73, 76, 118, 123, 134, 135, 163, 173 pragmatic function, 123 profile, 27, 30, 31, 33, 48, 51, 121, 129, 130, 133, 139 projected reality, 143, 147 prominence, 33, 39, 40, 41, 42, 46, 52, 53, 54, 64, 66, 76, 77, 79,

85, 98, 99, 101, 102, 103, 107, 108, 110, 163, 174, 175 reference-point, 47, 49, 50, 51, 54 referential activation, 174 referential expression, 54, 57, 58, 60, 62, 63, 64, 65, 66, 67, 68, 72, 77 relative prominence, 101 Relevance Theory, 60, 61, 62, 76, 163, 173 semantic meaning, 9, 21, 22, 25, 26, 29, 30, 32, 33, 34, 35, 47, 81, 115, 128, 140, 141, 161, 164, 169, 172, 173 semantic pole, 25, 33, 172 semantic structure, 21, 23, 27, 29, 32 semantics, 1, 21, 24, 29, 32, 33, 42, 68, 127, 131, 141, 142, 149, 150, 151, 152, 158, 161, 167, 168, 173, 174 space builder, 124, 125, 134, 139 subjectification, 35, 111, 115, 118, 119, 120, 121, 122, 128, 134, 140, 158, 159, 166, 168, 176 subjunctive, 6, 7, 10, 114, 126, 127, 128, 140, 141, 142, 143, 144, 146, 147, 149, 151, 152, 153, 154, 157, 159, 167, 168 surface structure, 14, 18, 19, 115, 170, 175 symbolization, 21, 22, 24, 29, 172 trajector, 31, 32, 33, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 61, 62, 68, 69, 70, 72, 73, 74, 75, 76, 83, 84, 85, 92, 93, 94, 95, 98, 99, 100, 101, 102, 103, 104, 107, 108, 109, 110, 111, 112, 120, 130, 139, 143, 145, 149, 152, 153, 154, 156, 158, 159, 163, 164, 165, 167, 168, 174 Universal Grammar, 169, 170, 171, 176 usage based model, 21, 25, 29, 37