Speech Acts in English: From Research to Instruction and Textbook Development

Speech Acts in English: From Research to Instruction and Textbook Development Lorena Pérez-Hernández To my family: my parents Aurora and Luciano, my daughters Lucía and Payal, my sister Mila, and my niece Arsema. You are the engine that fuels this and every project in my life. Table of Contents List of Figures List of Tables Acknowledgements List of Abbreviations Chapter 1. Introduction 1.1. Teaching directive speech acts: is there room for improvement? 1.2. Objectives 1.3. Methodology 1.4. Chapter contents Chapter 2. What Contemporary Research Tells Us about Speech Acts 2.1. Speech Acts: The Player All Linguistic Theories Want in their Team 2.2. Team 1. Codification-Based Theories and the Over-Grammaticalisation of Speech Acts 2.2.1. Weaknesses of the Literal Force Hypothesis and Ross’s Performative Hypothesis 2.2.2. Halliday’s Over-Grammaticalisation of Speech Acts 2.3. Team 2. Convention-Based Theories: Indirect Speech Acts 2.3.1. Searle: Inference and Convention in Speech Acts 2.3.2. Morgan’s Conventions of Usage and Short-Circuiting Implicatures 2.4. Team 3. Inference-Based Theories: Over-Pragmatisation of Speech Acts 2.4.1 Standard Pragmatics Approach: Bach and Harnish’s Speech Act Schemas 2.4.2. Direct Access Approaches I: Leech’s Interpersonal Rhetoric 2.4.3. Direct Access Approaches II: Conversational Approaches 2.5. A Cognitive-Constructional Approach to Directive Speech Acts 2.5.1. What Experimental Linguistics Has Revealed about Speech Act Processing 2.5.2. Redefining the Literal Force Hypothesis in Terms of SentenceType/Speech Act Compatibility 2.5.3. Revisiting the Notions of Direct and Indirect Speech Acts 2.5.4. Constraining Inferences via Cognitive Operations: Conceptual Metonymy 2.5.5. Illocutionary Idealised Cognitive Models and (Multiple Source)-in-Target Metonymies 2.5.6. Assembling the Illocutionary Puzzle: Families of Illocutionary Constructions Chapter 3. Critical Assessment of the Representation of Speech Acts in Advanced EFL Textbooks 3.1. Analytical categories and corpus of textbooks for analysis 3.2. Quantitative Assessment of the Treatment of Directive Illocutionary Acts in Advanced EFL Textbooks 3.3. Qualitative Assessment of the Treatment of Directive Illocutionary Acts in Advanced EFL Textbooks 3.3.1. Assessment of the Inclusion of Semantic/Pragmatic Information about Speech Acts in EFL Textbooks 3.3.2. Assessment of the Treatment of the Constructional Nature of Directive Speech Acts in EFL Textbooks 3.3.3. Assessment of the Treatment of Conversational Aspects of Directive Speech Acts in EFL Textbooks 3.3.4. Assessment of the Treatment of Cross-Cultural and Cross-Linguistic Areas of Discrepancy between L1 and L2 in EFL Textbooks 3.4. Conclusions and Way Forward: Explicit Instruction though a Corpus-Based Cognitive Pedagogical Grammar Chapter 4. A Cognitive Pedagogical Grammar of Directive Speech Acts I: The KnowWhat and Know-How of Directives 4.1. Orders 4.1.1. The Know-What of Orders 4.1.2. The Know-How of Orders 4.2. Requests 4.2.1. The Know-What of Requests 4.2.2. The Know-How of Requests 4.3. Beggings 4.3.1. The Know-What of Beggings 4.3.2. The Know-How of Beggings 4.4. Suggestions 4.4.1. The Know-What of Suggestions 4.4.2. The Know-How of Suggestions 4.5. Advice Acts 4.5.1. The Know-What of Advice Acts 4.5.2. The Know-How of Advice Acts 4.6. Warnings 4.6.1. The Know-What of Warnings 4.6.2. The Know-How of Warnings Chapter 5. A Cognitive Pedagogical Grammar of Directive Speech Acts II: Activities and Practice Materials 5.1. Teaching the Know-What of Directives 5.2. Teaching the Know-How of Directives 5.3. Teaching Cross-Cultural and Cross-Linguistic Issues of Directives Chapter 6. Conclusions List of References Index List of Figures Figure 2.1. Cost-benefit and politeness interactions Figure 2.2. Metonymic activation underlying examples (17) and (18) Figure 2.3. Metonymic activation underlying examples (19)-(22) Figure 2.4. (Multiple-source)-in-target metonymic operation underlying the request act in example (23) Figure 2.5. Conceptual motivation underlying the CAN vs. COULD YOU DO X? request construction in examples (24) and (25). Figure 4.1. Force schema underlying the conceptualisation of orders Figure 4.2. Force schema underlying the conceptualisation of requests Figure 4.3. Force schema underlying the conceptualisation of beggings Figure 4.4. Force schema underlying the conceptualisation of suggestions Figure 4.5. Force schema underlying the conceptualisation of advice acts Figure 4.6. Force schema underlying the conceptualisation of warnings Figure 5.1. Force dynamics of orders. S=speaker; F=force; H=hearer Figure 5.2. Force dynamics of requests. S=speaker; F=force; H=hearer Figure 5.3. Force dynamics of beggings. S=speaker; F=force; HR/S=hearer as rational /social being; HA=hearer as agent Figure 5.4. Force dynamics of suggestions. S=speaker; F=force; HR=hearer as rational being; HA=hearer as agent List of Tables Table 2.1. Deep structure vs. surface structure in questions for information vs. requests interpretation Table 2.2. Attributes configuring the ontology of directive illocutionary ICMs Table 2.3. Ontology for the illocutionary ICM of the act of requesting Table 2.4. Social variables configuring the structure of directive illocutionary ICMs Table 2.5. Structure of the illocutionary IXM if the act of requesting in terms of interactions between attributes, and between attributes and social variables Table 2.6. Realisation procedures for the attributes of the ICM of requesting Table 2.7. Number of occurrences in iWeb for the realisation procedures of the requests in examples (27)-(31) Table 2.8. Base structures for the act of requesting Table 2.9. Summary of contemporary theoretical advancements on speech acts Table 3.1. Number of activities devoted to the teaching/practice of directive speech acts in advanced EFL textbooks Table 3.2. Analytical categories for the assessment of current EFL textbooks as regards their treatment of directive speech acts Table 4.1. The know-what of orders Table 4.2. Base constructions for the act of ordering in English Table 4.3. Imperative base construction + Realisation procedures for orders Table 4.4. Comparison of base constructions for the act of ordering in English and Spanish Table 4.5. The know-what of requests Table 4.6. Base constructions for the act of requesting in English Table 4.7. Imperative base construction + Realisation procedures for requests Table 4.8. Comparison of base constructions for the act of requesting in English and Spanish Table 4.9. The know-what of beggings Table 4.10. Base constructions for the act of begging in English Table 4.11. Imperative base construction + Realisation procedures for beggings Table 4.12. Comparison of base constructions for the act of begging in English and Spanish Table 4.13. The know-what of suggestions Table 4.14. Base constructions for the act of ordering in English Table 4.15. Imperative base construction + Realisation procedures for suggestions Table 4.16. Comparison of base constructions for the act of suggesting in English and Spanish Table 4.17. The know-what of advice acts Table 4.18. Base constructions for the act of advising in English Table 4.19. Realisation procedures for the act of advising Table 4.20. Imperative base construction + Realisation procedures for the act of advising Table 4.21. Solicited and unsolicited bare imperatives advice acts Table 4.22. Comparison of base constructions for the act of advising in English and Spanish Table 4.23. The know-what of warnings Table 4.24. Base constructions for the act of warning in English Table 4.25. Realisation procedures for the act of warning Table 4.26. Comparison of base constructions for the act of warning in English and Spanish Table 5.1. The know-what of suggestions Table 5.2. The know-what of requests (short version) Table 5.3. The know-what of requests (interactions between attributes and variables) Table 5.4. Linguistic realisation procedures for combination with request base constructions Table 5.5. Field work, comparison, and discussion task Table 5.6. Base constructions for requests in English Table 5.7. Base constructions for requests in Spanish Table 5.8. The know-what of advice acts (interactions with socio-contextual variables) Table 5.9. Base constructions for advice acts in English Table 5.10. Discourse completion tasks Table 5.11. Repair and comparison task Acknowledgements 1 Chapter 1. Introduction 1.1 Teaching Directive Speech Acts: Is There Room for Improvement? Every day of our lives we use language to tell stories, to express feelings, to describe and talk about the world, and also, as was lucidly revealed by Austin (1962) in his book How to do things with words, to perform actions. There are some things we do physically with our hands or with our legs, and others we do verbally by means of words. One essential thing we can do, and often do, primarily by means of language is to try to influence other people into a certain path of action. Thus, we can move them to do something in their benefit (by advising or warning them), in our own benefit (by ordering, requesting or begging them), or in mutual benefit (i.e. by suggesting a common course of action). Actions of this kind, which we carry out linguistically and which have as their main objective to prompt our interlocutors to do something, are known in the literature as directive speech acts. Directives are ubiquitous. We start producing them as soon as we learn to speak and, for the rest of our lives, we use them on a daily basis. Directives are useful because they allow us to attain our goals by persuading or pushing other people to act according to our wishes, but they are not free from risks. Using directive acts inappropriately can offend our interlocutors. As a result, the directive act may turn out to be ineffective and our goals may end up unfulfilled. Alternatively, we may succeed in getting the listener to carry out the desired action, but at a high social cost. Human interaction is very sensitive to the use of directives, which are a potential source of interpersonal conflicts. Learning to produce directive speech acts correctly in a second language is a crucial, but difficult task. Speech acts are rich, kaleidoscopic concepts and their use involves juggling with many social, interactional and contextual variables at once: the relative power of the speakers, their social closeness, the formality of the situation, or the cost/benefit of the action being requested, among others. These are only some of the factors that need to be considered by the speaker if she wants to produce a contextually, and socially felicitous speech act that stands a chance to be successful. A request like If you don’t mind, would you please be so kind as to open the window? functions well in a formal context or in a situation in which the speakers are strangers to each other, or one of them is more powerful than the other (e.g. an employee talking to her boss). However, the same request would most probably sound odd, and possibly move the receptor to search for some ulterior hidden meaning (e.g. irony), if used among close friends in an informal setting. 8 conceptual (i.e. meaning) and formal (i.e. linguistic realisations) aspects of speech act production included in objectives (1) and (2) above. Recent contrastive analyses show that languages have different constructions and constructional properties that are likely to cause transfer effects and faulty representations of constructions in a second language (HijazoGascón, Cadierno & Ibarretxe-Antuñano, 2016; Ruiz de Mendoza & Agustin, 2016). In fact, according to Holme (2009: 84), one of the main sources of errors for EFL students is their use of L1 forms and meanings when using an L2. This has prompted many voices within the Cognitive Linguistics paradigm to raise in support of a greater acknowledgement of the pervasiveness and relevance of constructions in the EFL teaching (Römer, O’Donnell & Ellis, 2014). Identifying the semantic, pragmatic and formal mismatches between the two languages can help students to avoid those traps by serving ‘the purpose of preparing teaching materials and exploiting these in L2/FL instruction’ (De Knop & De Rycker, 2008: 2). Despite the relative youth of construction-based pedagogical approaches, experimental evidence on the psychological reality of constructions for EFL learners and the effectiveness of a pedagogical intervention based on the explicit teaching of form-meaning pairings is mounting up in the literature (Holme, 2010; Baicchi, 2016; Sung & Yang, 2016). Altogether, objectives (1)-(3) advocate a cross-linguistic, cognitive-constructional approach to the teaching of directive speech acts. One of the central tenets of Cognitive Linguistics is that language is composed of form-meaning pairings, also known as constructions (Langacker, 2008; Goldberg, 1995, 2006). Linguistic choices are paired with idealised representations of reality that include propositional, pragmatic, contextual and cultural knowledge. Meaning is, therefore, understood in a broad sense, as the idealised collection of encyclopaedic knowledge that language users have about a topic. This view of language allows Cognitive Linguistics to explain the motivation of linguistic forms and makes of it a promising theoretical framework to inform EFL teaching practices and to guide the design of EFL didactic materials. As Jacobsen (2018: 669) puts it: Using select, carefully adapted for unprepared audiences, and contextualised CL concepts in L2 instruction can reveal the perspective of a native speaker and make form-meaning mappings relatively transparent for L2 learners. Accordingly, the underlying conceptual characteristics of CLs make it a good candidate for the role of providing a comprehensive theory that could successfully support L2 instruction (Achard & Niemeier 2004; Tyler & Evans 2004; Tyler 2012). 9 So far, as noted in Jacobsen (2012, 2015), there is mounting experimental, quantitative (Chen & Oller, 2008; Valenzuela & Rojo, 2008; Jacobsen, 2018) and qualitative evidence that CL tenets can successfully be applied to the teaching of diverse aspects of language such as phrasal verbs (Dirven, 2001; Liu, 2010), metaphorical/idiomatic language (Lindstromberg & Boers, 2005; Littlemore & Low, 2006; Boers & Lindstromberg, 2008; Yasuda, 2010), and grammar (prepositions: Lindstromberg, 1996; Tyler & Evans, 2004; Tyler, Mueller & Ho, 2011; tense and aspect: Niemeier & Reif, 2008). There is a notorious lack, however, of applications of Cognitive Linguistics to the teaching of higher levels of language description, among which there is the teaching of speech acts. As shall be explained in the next section, this book incorporates a cross-linguistic, cognitive-constructional approach to directive speech acts into the EFL teaching practice and the design of EFL teaching materials based on explicit instruction. 1.3 Methodology In accordance with the three goals specified in the previous section, our proposal advocates an explicit instruction approach to the teaching of speech acts, which has already been proved to be more beneficial than implicit L2 teaching (Norris & Ortega, 2000; Ellis, 2001, 2002; Spada & Tomita, 2010), with specific studies providing evidence about its effectiveness on the realm of grammar (Sharwood Smith, 1981; Rutherford, 1987; Ellis, 2001), and more recently, also of pragmatics (Taguchi, 2015; Tello, 2016). Explicit instruction is also better adapted to the capacities and needs of advanced students of a second or foreign language, who already have the necessary command of that language to understand the explanations in the L2 explicitinstruction textbook or class. The teachers and the authors of advanced textbooks can profit from their students’ linguistic competence to explicitly teach subtle nuances of the L2, which at lower levels have to be necessarily taught in a more implicit fashion. To facilitate the explicit instruction of directive speech acts, this book takes on a pedagogical grammar approach to their characterisation. Taylor (2008: 38) explains that a pedagogical grammar may be defined ‘as a description of a language [or an aspect of a language] which is aimed at the foreign language learner and/or teacher, and whose purpose is to promote insight into, and thereby to facilitate the acquisition of, the foreign language.’ Pedagogical grammar differs from descriptive or linguistic grammars, which are written by linguists for linguists, and which are not aimed at meeting the needs of the students and/or teachers of a language (Dirven, 1985, 1990, 2001). Thus, pedagogical grammars make use of accessible terminology and represent concepts in a reader-friendly manner to be able to reach their intended audience. In 10 addition, they place the focus of attention on those aspects of a language that may pose a problem or a difficulty for learners, and offer succinct, intuitive, and straightforward explanations of idiosyncratic aspects of the foreign language (Taylor, 2008: 38-39). To comply with the objectives that were set up in section 1.2, the pedagogical grammar offered in this work includes cross-linguistic, cognitive and constructional information about speech acts. A word of caveat is in order. The term ‘grammar’ should not be understood here in a narrow sense as applying exclusively to morphological and syntactic aspects of language. Pedagogical grammar inherits the Cognitive Linguistics conception of grammar as the collection of form-meaning pairings (i.e. constructions) that occur at all levels of linguistic description, including morphology, lexicon, syntax, illocution, and discourse (Goldberg, 1995, 2006; Langacker, 2008; Ruiz de Mendoza & Galera, 2014). Pedagogical grammar is, therefore, an adequate tool for the description and explanation of the directive illocutions under scrutiny. More specifically, the cross-linguistic, cognitive-constructional pedagogical grammar approach to directive speech acts adopted in this book can help teaching professionals (whether teachers, lecturers, and/or textbook writers) to prepare their teaching sessions and design their didactic materials on this topic, while at the same time it may be useful for postgraduate students and novel researchers who are looking to gain an understanding of the workings of directive illocutionary acts. A second relevant methodological decision on which the present proposal hinges is the use of real language data for the analysis of both the semantic and the formal sides of directive illocutionary constructions. At present, corpus-based cognitive pedagogical grammars are scarce. Meunier (2008: 99) points to the relative youth of the fields of corpus linguistics and cognitive pedagogical grammar as the main reason for the lack of availability of this type of materials. One exception is Carter, Hughes and McCarthy’s (2000) Exploring Grammar in Context, which does not, however, include the description of speech acts. As noted in Meunier (2008: 102), whether the use of authentic language in pedagogical grammars and teaching materials is beneficial to the learning process is a matter of controversy. Some authors defend the exclusive use of authentic material (Römer, 2004, 2006) and others argue that learners have more to gain from exposure to prototypical, didactic and learneradapted examples (Widdowson, 2003). The cognitive pedagogical grammar of directive speech acts offered in this book takes sides with Valdman’s (2003: 61) equidistant position to the effect that it reflects ‘the actual speech of target language speakers in authentic communicative situations, and should conform to native speaker’ idealised view of their speech use, should conform to expectations of both native speakers and foreign learners concerning the type of 11 behaviour appropriate for foreign learners, and should take into account processing and learning factors’. Thus, the description and explanation of the meaning and formal sides of directive speech acts in Chapter 4 stems primarily from the analysis of authentic instances of language taken from the British National and the iWeb corpora for English data, and the CREA and CORPES XXI corpora for Spanish data (see below). When possible and as required by the teaching and learning needs, real corpus data has also been used to illustrate the theoretical points offered in Chapter 4, and in the development of activities and practice materials in Chapter 5. In this the present book follows and acknowledges recent fine proposals on how to effectively use corpus data in the development of classroom materials for teaching ESL pragmatics (Bardovi-Harlig, Mossman & Vellenga, 2014a, 2014b, Bardovi-Harlig & Mossman, 2016; Bardovi-Harlig, Mossman & Su, 2017). Nevertheless, both illustrative examples of theoretical aspects of speech acts and those illocutionary act samples included in the activities and practice materials need not always exhaustively copy the rough data in the corpora used for analysis. If needed, they have been adapted to meet teaching and learning requirements. The use of adapted data appears as especially useful when the object of study involves higher level pragmatic concepts such as directive speech acts. As explained in section 1.2, the successful use of directive illocutions often involves managing complex amalgams of pragmatic, semantic and conversational knowledge. For teaching and learning purposes, it is therefore more clarifying to initially present the students with adapted examples that illustrate the workings of just one or two of these variables at a time. Once the different strands of knowledge needed to perform a directive speech act correctly have been mastered, it seems the right moment to present the students with more complex, corpus-based samples and exercises that juggle with several of those variables simultaneously. The present book, therefore, adheres to a mixed approach to the use of corpus data. While the theoretical proposals on the nature of speech acts are based on the analysis of real-language examples, the use of the latter for illustrative purposes or in the design of teaching materials is conditioned to the teaching needs, and both adapted, and corpus-based examples are used as required by instructional purposes. For the English data, the analysis makes use of the British National Corpus (BNC), and the more recent iWeb corpus, which allows specific searches for the different varieties of English. Together, these corpora offer over fourteen billion words, including both written and spoken data, for analysis. For the Spanish data, the Corpus de Referencia del Español Actual (CREA) has been chosen, because it has a good balance of written and spoken sources; together with the Corpus del 12 Español del Siglo XXI (CORPES XXI), which is yearly updated with twenty-five million extra forms. Both the CREA and the CORPES XXI allow specific searches on the Castilian variety of Spanish. Given the kaleidoscopic and often non-conventional nature of speech acts, however, the use of corpora may sometimes run short of providing an exhaustive picture of the semantic and formal characterisation of speech acts. Therefore, a mixed method has been implemented in order to find the most exhaustive collection of linguistic realisations for each speech act type: (1) a corpus-based search has been performed in order to find indirect speech acts which are introduced by reporting performatives (e.g. He requested: ‘Could you help me out with this mess?’) in the BNC, iWeb, CREA, and CORPES XXI corpora described above. In this way, it has been possible to take advantage of the categorisation made by real speakers of the speech acts they are reporting; (2) the initial list has been enriched with other directive configurations taken from a thorough revision of the literature on speech acts and additional research resources, such as the CARLA (Centre for Advance Research on Language AcquisitionUniversity of Minnesota) website; and (3) the final collection of speech acts for analysis has been completed with instances of directive acts that the author has come across in her everyday life interaction with English native speakers or arising from her own native speaker’s intuition in the case of the Spanish language. Before including the linguistic configurations stemming from (2) and (3) in the final list of linguistic realisations for speech acts, an additional search of those configurations in rich computerised corpora (i.e. iWeb, Web Corp Live) has been conducted to assess their productiveness and to confirm that they are in actual use. Given the indirect nature of speech acts, this combined method guarantees a desirable degree of exhaustiveness in the final collection of linguistic configurations. 1.4 Chapter Contents The contents of the book have been organised as follows: Chapter 2. What Contemporary Research Tells Us about Speech Acts Chapter 2 provides an accessible and reader-friendly outlook on contemporary theoretical research on speech acts, explaining and illustrating the latest pragmatic, functional, and cognitive/constructional contributions to the understanding of illocutionary acts. The chapter advocates for the latest contributions carried out within cognitive/constructional theories of speech acts, showing how these approaches are compatible with pragmatic and functional perspectives on illocution, and capable of integrating pragmatic, semantic and 13 syntactic aspects of speech acts into a unified and comprehensive account. This chapter sets the theoretical foundations for Chapters 4 and 5, offering a fully-fledged theoretical proposal on how the description of the semantic and formal features of directive illocutionary acts should be approached in terms of illocutionary constructions. At the end of Chapter 2 and stemming from the previous overview of contemporary theories of illocution, the reader is presented with a relevant list of theoretical aspects about the meaning and form of speech acts, which have been shown to be essential in their production and understanding, and which should, therefore, be included in textbooks and instructional materials in order to guarantee their correct learning. This inventory of relevant theoretical aspects that need to be taken into account in the description of directive speech acts will serve as a guide to the critical assessment of their representation in current EFL textbooks in Chapter 3. Chapter 3. Critical Assessment of the Representation of Speech Acts in EFL Textbooks Chapter 3 reports the results of a study which looks into a collection of textbooks for advanced EFL students and analyses how directives speech acts are currently being approached. The chapter considers aspects related to (1) the quantitative representation of directive speech acts in the textbooks (i.e. determining if there is a balanced portrait of the most frequent categories of directive speech acts), and (2) the qualitative treatment of directive speech acts (i.e. assessing if the depiction of directive speech acts in EFL textbooks has incorporated the main research advancements described in Chapter 2). Chapter 4. A Cognitive Pedagogical Grammar of Directive Speech Acts I: The Know-What and Know-How of Directives Chapter 4 applies the cognitive-constructional approach to speech acts advocated for in Chapter 2 to the task of providing an exhaustive description of the meaning and form of the six directive speech acts under consideration (i.e. orders, requests, beggings, suggestions, advice acts, and warnings), including cross-cultural considerations about the realisation of these illocutionary acts in Spanish (L1) and English (L2). The description of the form-meaning constructional nature of the directive speech acts will be carried out in an accessible, largely jargon-free manner, so that it can be used by teachers and textbook developers for the explicit teaching of the workings of each of the directive illocutions to advanced Spanish EFL students. 14 Chapter 5. A Cognitive Pedagogical Grammar of Directive Speech Acts II: Activities and Practice Materials Chapter 5 draws on the description of the meaning and form of directive speech acts offered in Chapter 4 in order to offer a collection of suggestions and examples of the type of practical activities and exercises that could help in the teaching of the directive speech acts under scrutiny. Activities will be grouped attending to the semantic, formal, or cross-linguistic aspects that need to be taught. Especial emphasis will be placed on those activities aimed at the students’ recognition of those illocutionary constructions and linguistic realisation procedures which allow the communication of the different illocutionary forces. Some of these activities will be specifically devoted to help teachers and textbook developers show students (1) the motivation of the meaning of directive speech acts in their underlying force dynamics, and (2) the role of conceptual metonymy in the production of directive speech acts. Conclusions The final chapter summarises the main contributions of the previous chapters and offers suggestions for future research. 15 Chapter 2. What Contemporary Research Tells Us about Speech Acts 2.1 Speech Acts: The Player All Linguistic Theories Want in their Team The story of theoretical studies of speech acts is one of appropriation: three all-stars teams of linguists, representing largely opposing theoretical positions about the nature of language, wanting so badly to sign speech acts for their club that they do not hesitate in simplifying their true nature in order to have them in their crew. Team 1 represents those theories which support the belief that language allows us, language users, to fully codify what we want to say. Full codification leads to unambiguous, risk-free linguistic interactions, but it denies two of the most central and valuable characteristics of language: its creativity and its flexibility to adapt to diverse communicative needs. Adopting a full codification view of language, however attractive this position may seem in terms of cognitive economy and the prevention of misunderstandings, leads linguists to overlook pivotal aspects of human communication, such as the indirect uses of language that underlie ironic, figurative, or polite expressions, among others. At the opposite end of the spectrum of linguistic theories, supporters of team 2 envision speakers of a language as a sort of intelligent, fast-processing, walking computers that manage to swiftly calculate, on the spot, the meaning of each expression they hear in their daily lives. This theoretical stance is capable of accommodating the creativity and flexibility of so-called indirect forms of language, but it does so at a high cognitive cost. Granted that human brains are powerful devices, still, having to calculate the meaning of each and every single utterance that we come across in our daily interactions appears at first sight as a cognitively costly and time-consuming approach to language. Even computers make use of rules, patterns, and scripts to automatise frequent and recurrent tasks. Why would the human mind function differently? Team 3 holds an intermediate position between the previous two, allowing for the existence of at least some conventional means of expressing our thoughts and wishes and, therefore, for more creative, but also safer, and more cognitively-efficient ways of communicating our thoughts than those proposed by team 1 and 2 respectively. As shall be shown in section 2.4, team 3 also presents some problems in its game: its advocates defend an all-or-nothing view of conventionality. Thus, they claim that a speech act is either conventional or not. CAN YOU DO X? forms, for instance, would be considered conventional expressions of requests by followers of this team. Section 2.5. provides evidence that this is often an oversimplification of how the production of speech acts works. Conventionality is shown to be itself a blurry theoretical category, which can be modulated linguistically to produce speech acts with a higher 16 or lower conventional meaning, which thus gradually approach either the codification or the inferential ends of the illocutionary continuum proposed by teams 1 and 2. For decades, linguists have attempted to confine speech acts into the limits of radical conceptions of language, in which either codification, convention or inferential processes, represented by our three imaginary teams of linguists above, took pride of place in the explanation of the illocutionary component of language. This chapter unveils these attempts and argues in favour of a wide-ranging, comprehensive approach to speech acts that highlights their polychromatic, ductile, and adaptable nature. In so doing, it will be argued that the story of speech acts is also the story of a highly fluid player who can perform equally well for these three teams as required by contextual and interactional needs, and who resists permanent affiliation. What motivates the existence of these three codification-based, convention-based, and inference-based theories of speech acts is their adherence, or lack of it, to the Literal Force Hypothesis (Levinson, 1983). This hypothesis assumes that each sentence has an illocutionary force of its own built into it. In other words, that the speech act conveyed by a sentence derives from its form. This can be done in two different ways: (i) using an explicit performative verb that names the speech act that wants to be communicated, such as beg in No, my lord, I beg you keep this thing secret (BNC). (ii) or, alternatively, using one of the three major sentence-types in English (i.e. imperative, interrogative and declarative), which have the forces traditionally associated with them. (i.e. ordering (or requesting), questioning, and stating respectively). According to this, an interrogative sentence like Can you hear me? (BNC), for example, would have a questioning illocutionary force. Sections 2.2 to 2.4 below analyse the virtues and flaws of the Literal Force Hypothesis and explain how either acceptance or rejection of its premises derives in membership to one of the aforementioned three views of speech acts. Different theories within each ‘team’ are revised, their weaknesses exposed, and their strengths highlighted to be later on integrated into our own theoretical proposal on how speech acts are produced and interpreted (Section 2.5). The resulting proposal advocates a modified, weaker version of the Literal Force Hypothesis (Risselada, 1993) that sets up the possibility of integrating the three existing antagonistic 17 approaches to illocution into a single, all-encompassing framework for the study of speech acts, and its ensuing application to language teaching. 2.2 Team 1. Codification-Based Theories and the Over-Grammaticalisation of Speech Acts 2.2.1 Weaknesses of the Literal Force Hypothesis and Ross’s Performative Hypothesis Accepting the Literal Force Hypothesis leads to a rather limited and rigid view of speech acts. Fully codified, unambiguous, linguistic acts are restricted to those three (i.e. stating, questioning, and ordering/requesting) that can be expressed by means of the corresponding three universal sentence types (i.e. affirmative, interrogative, and imperative), or by means of explicit performatives. A quick search in the British National Corpus (BNC), however, reveals that explicit performatives are rare in real life communicative exchanges. Simple searches of performative sentences like I order you to…, I request you to…, I beg you to…, I advise you to/that…, I warn you to/that…, or I suggest that you… yield less than 20 results each on average in a 100 million word corpus like the BNC. Speakers of English do not often use explicit performatives to order, request, beg, advise, warn, or suggest other people to do something. This is only to be expected since fully explicit, lexicalised acts such as these do not easily adapt themselves to the varied communicative needs of different social settings. By way of illustration, in a situation in which a large social distance or an asymmetrical power relationship between the speakers asks for a polite request, a performative directive like I request you to repeat it does not seem to be the best possible choice to make the act as polite as required by the aforementioned context. An indirect, conventional request such as Can you repeat it, please? (BNC), on the contrary, would come through as more natural. In fact, request constructions of the CAN YOU DO X (please)? form have a higher frequency of occurrence in the same corpus with one hundred fifty-nine occurrences. This latter example reveals yet another weakness of a theory of speech acts based on the acceptance of the Literal Force Hypothesis. The allegedly literal force of the three basic sentence types can be interpreted, and it is in fact, often interpreted with a secondary illocutionary force. Can you repeat it, please? displays an interrogative sentence type. According to the Literal Force Hypothesis, interrogatives are associated with the speech act force of questioning. Nevertheless, if we zoom in into the context in which this specific instance of the CAN YOU DO X (please)? construction was uttered, it is clear, by the addressee’s 24 interpretation for those speech acts which are not linguistically codified (i.e. indirect speech acts). In this he approaches the postulates of team 3 (see Section 2.4). This inferential theory, however, is incapable of accounting for the aforementioned illocutionary phenomenon that he encounters in his study: conventional instances of illocutions. As pointed out by Geis (1995), one of the reasons why Searle’s account cannot deal with conventionality is that it focuses on transactional aspects of speech acts and overlooks their interactional side. In fact, some conventional forms of requests with modal verbs in the past tense (e.g. Could/Would you open the door?) are motivated by the interactional need of acting politely. This weakness is a direct consequence of Searle’s assumption that speech acts are essentially linguistic and his failure to realise their social dimension. As argued in Halliday (1978, 1994), Marcondes (1984) and Geis (1995), among others, warnings, requests, orders, and other communicative acts are social constructs that can also be realised by non-linguistic means (i.e. gestures). The social nature of speech acts and the interactional factors that affect their performance will be extensively dealt with by inference-based theories (see Section 2.4). Searle, however, can be credited for bringing to the front the existence of the especial category of conventional speech acts, which are not fully codified, but are more readily understood than fully inferred/calculated speech acts. The literature shows several attempts to account for them. 2.3.2 Morgan’s Conventions of Usage and Short-Circuiting Implicatures Morgan (1978) was one of the first to take on the challenge of explaining the functioning of conventional speech acts. In order to do so, he distinguished between conventions of language (i.e. those that account for the literal, arbitrary meaning of sentences) and conventions of usage (i.e. those based on cultural conventions that govern the use of a sentence with a fixed literal force for the performance of a different act in a particular culture). Language users have knowledge of conventions of language which account for the fact that an interrogative sentence (e.g. Can you hold the book?) has the literal meaning of a question, as predicted by the Literal Force Hypothesis. But they are also aware of the convention of usage that justifies that the same sentence can also be used as a request. In order to understand why there is a convention of usage by virtue of which a sentence like Can you hold the book? can be interpreted as a request, it would be possible to reconstruct the inferential steps leading from its literal force of a question to its indirect interpretation as a request. This could be done using the same theoretical apparatus proposed by Searle. Thus, CAN YOU DO X? forms question one of the preparatory conditions for the performance of requests. However, it is arguable whether 25 speakers do actually calculate the requestive value of the sentence in such a burdensome fashion. On the contrary, they seem to straightforwardly jump to the request interpretation. Morgan referred to this phenomenon as short-circuiting implicatures: The expression Can you… is not an idiom but has only the obvious literal meaning of a question about the hearer’s abilities. One can readily see how the expression could have, via Grice’s maxims, the implicature of a request. In fact, it has become conventional to use the expression in this way. Thus, speakers know not only that Can you…? has a literal meaning (a convention of language): they know also that using Can you… is a standard way of indirectly making a request (a convention of usage). Both are involved in a full understanding by the hearer of what is intended in the use of the expression (Morgan, 1978: 274). … although Can you pass the salt? is indeed CALCULABLE, it is not in fact calculated: rather, one gets the point more or less directly, without any inferential processing (Morgan 1978: 263). Thus, thanks to the existence of conventions of usage, speakers can avoid tiresome calculations and short-circuit the inferential process that would lead to the interpretation of CAN YOU DO X? forms as requests. Morgan’s proposals allow us to understand the different functioning of near synonym constructions, such as CAN YOU DO X? and ARE YOU ABLE TO DO X?, in terms of degree of conventionalisation. According to his view, only in the case of CAN YOU DO X? interrogative sentences there is a convention of usage that explains their interpretation as requests. No explanation is given, however, about the reasons why these conventions of usage are established for some linguistic forms and not for others that are close synonyms. The explanatory power of Morgan’s conventions of usage and short-circuiting implicatures is not capable either of motivating the existence of a varied range of slightly different linguistic forms of the same speech act. Will/Would you hold the book?, Can/Could you hold the book?, Hold the book, can/will/could/would you?, Would you mind holding the book?, etc. are all possible instances of conventional requests. What prompts speakers to use one or the other is often related to politeness and interactional factors. Just like Searle, Morgan’s view of speech acts overlooks their social side, thus offering a limited portrayal of illocution. 26 2.4 Team 3. Inference-Based Theories: Over-Pragmatisation of Speech Acts In Sections 2.2 and 2.3 it has been shown how the different attempts to explain speech acts as either fully codified or conventional acts, respectively, are not without problems: some of them are not flexible enough to account for the creativity, flexibility, and economy of language (e.g. Halliday’s systemic-functional proposal), some run short of explaining the vast range of linguistic realisations available for the expression of a speech act (e.g. Searle’s and Morgan’s accounts), and most of them cannot provide a full picture of illocutionary performance, since many speech acts, in fact, escape codification and conventionalisation, and require of inferential processes to a larger or lesser extent for their full understanding. This section looks into those proposals that give pride of place to inference in the understanding of speech acts. 2.4.1 Standard Pragmatics Approach: Bach and Harnish’s Speech Act Schemas Bach and Harnish (1979) were among the first linguists to question the validity of the Literal Force Hypothesis, and consequently, of the role of codification in speech acts (Dascal, 1989; Dobrovie-Sorin, 1985; Recanati, 1987, 1994; Bertolet, 1994). Their contribution is relevant, because it paved the way for some mainstream theories of illocutionary performance in the last decades, including conversational approaches (e.g. Levinson, 1983; Geis, 1995), pragmatic accounts (e.g. Leech, 1983, 2014), and relevance-theoretical proposals (e.g. Sperber & Wilson, 1995). Bach and Harnish (1979: 17) approached speech acts from an extreme inferentialist viewpoint, according to which their performance is thought to involve ‘the speaker’s having a special sort of intention (an intention that the hearer make a certain sort of inference) and the hearer’s actually making that inference’. Let us consider example (1) again, reproduced here as (5) for convenience: (5) Albini thinks of Big Black and says, ‘We were a pretty good live band.’ To anyone in a band now, he advised, ‘Develop your aesthetic and execute it, you will find your audience, your audience will find you.’ [BNC] Linguists in team 2 (e.g. Searle, Morgan), who accepted the Literal Force Hypothesis, would claim that the imperative sentence Develop your aesthetic and execute it has both a literal force of order (i.e. the one associated with its imperative sentence form), and an indirect force of advising, which is reached through several inferential steps (see Section 2.3). In their account, the literal force is codified in the linguistic form, the indirect force is inferentially calculated. 27 In addition, it is argued that speakers only bother to calculate the indirect force when a trigger makes them realise that the literal force does not fit the context. In contrast to this, the extreme inferentialists in team 3 reject the Literal Force Hypothesis, and widen the scope of the inferential processes involved in speech act interpretation to the effect that both literal and indirect speech acts need to be calculated (Bach & Harnish, 1979: 132). The inferential speech act schemas put forward by Bach and Harnish do not assume that the imperative in example (5) is directly link to the illocutionary force of order. On the contrary, these schemas make use of mutual contextual knowledge and general principles about the nature, stage, and direction of the talk exchange (i.e. Conversational Presumption) to infer whether the imperative sentence is in fact intended as an order. If the contextual information and the conversational presumptions validate this interpretation, the inferential process would come to an end. However, since this is not the case in the example under consideration, the listener needs to continue with her inferential calculations until she reaches the advising interpretation intended by the speaker. 4 Surprisingly enough, when faced with conventional instances of speech acts, like examples (2)-(4) above, Bach and Harnish adopt a position close to that of Morgan. Bach and Harnish (1979: 198) refer to conventional speech acts, such as If I were you, I would do X pieces of advice or Can you do X? request forms as ‘standardised’. They further argue that in these cases the inferential route is short-circuited, and the hearer identifies the speaker’s advising or requestive intent without having to grasp the literal intent of questioning associated with the interrogative sentence type (see also Bach, 1998). 2.4.2 Direct Access Approaches I: Leech’s Interpersonal Rhetoric Following the inference-based path to the analysis of speech acts opened by Bach and Harnish, other linguists have attempted to investigate the factors that guide and motivate the inferential interpretation of the intended force of speech acts. Among them, special attention should be paid to pragmatic proposals which not only provide a principle-based account of the inferential 4 For a detailed description and critical analysis of Bach and Harnish’s (1979) Speech Act Schemas, see PérezHernández (2001: 41-45). Sperber and Wilson’s (1995) Relevance Theoretic account follows Bach and Harnish in their assumption that even the literal meaning of an utterance needs to be inferred. They propose their own schema of interpretation based on the Principle of Relevance: first assessing the literal interpretation, then assessing its relevance, and, if it does not fit the contextual needs, deriving the non-literal interpretation. Bach and Harnish’s and Sperber and Wilson’s approaches are known as Standard Pragmatics models, which differ from the so-called Direct Access models (represented by Leech’s contributions, and those of the conversational theories of illocution) in that the latter make use of contextual information to derive the intended meaning of a speech act directly, without needing to assess the validity of the literal force of the utterance in the first place. 28 mechanisms involved but do so by acknowledging the social and conversational nature of speech acts. Undoubtedly, a prominent representative of this trend is Leech’s (1983, 2014) account of speech acts in terms of Interpersonal Rhetoric. In relation to the codification/conventionalisation vs. inference dichotomy, Leech states the following: There are important issues which cannot be pursued here, particularly the question of how far the relation of questions and mands to the illocutions they typically perform (viz. askings and impositives) is conventional, rather than determined by Interpersonal Rhetoric. Kempson (1975: 147) opts for a conventional mapping of one on to the other set of categories, whereas I prefer to go the whole pragmatic hog, and attempt an explanation entirely in terms of Interpersonal Rhetoric (Leech, 1983: 17). In other words, Leech takes on an extreme inferentialist approach to the analysis of speech acts, according to which the distinction between literal and indirect speech acts becomes irrelevant and it is necessary to calculate the meaning of all speech acts. As opposed to Bach and Harnish, however, Leech does not consider it necessary to calculate the literal meaning first, and then, if shown that it does not fit the context, to continue with the calculation of the indirect force. In Leech’s account, speakers make use of contextual and social information to directly calculate the intended meaning. 5 In fact, what makes his proposal stand out from previous inferential accounts is its focus on the social dimension of illocution. More specifically, Leech states that the need to be polite in our social interactions is one of the essential motivations underlying and guiding the production and interpretation of illocutionary acts. 6 For different reasons, on many occasions in our daily lives we are forced to carry out linguistic acts that may put our social relations at risk: acts that involve asking our interlocutors to do something that is costly to them in terms of time, resources, etc. (e.g. requests, beggings), acts that may be felt as impositions on their freedom (e.g. orders, threats), acts that may clash with our listeners’ plans 5 This Direct Access model of speech act interpretation has been followed by other authors such as Gibbs (1994, 2002). 6 Considerations of politeness as motivating factors of speech act performance were also explored by other contemporary authors like Brown and Levinson (1987 (1978)). Brown and Levinson’s proposal in terms of the notion of face (i.e. social image of a person) leads to a binary conception of politeness, whereby speech acts are either polite (when they maintained people’s face) or impolite (when they do not). Leech’s pragmatic scales, on the contrary, allow for a scaled gradation in the assessment of the politeness requirements of particular instances of speech acts, which reflects the needs of actual social interaction more closely. Leech’s proposals have been chosen for this study due to their encompassing and detailed nature. For a comparison of Brown and Levison’s and Leech’s politeness theories, see Pérez-Hernández (1999, 2001). 29 (e.g. suggestions), etc. To minimise or prevent the social conflicts that may arise from acts like these, speakers make use of their knowledge of Interpersonal Rhetoric, a keystone of which is the Politeness Principle: ‘minimise the expression of impolite beliefs, and maximise the expression of polite beliefs’ (Leech, 1983: 81). In the case of directive speech acts, being polite involves, among other things, trying to minimise the cost of the action that is requested from the hearer (Tact Maxim). In order to calculate the cost of the act and the corresponding degree of politeness needed to compensate for it and to keep interaction running smoothly, Leech (1983, 2014) puts forward a scale of cost-benefit or weightiness of the transaction. Figure 2.1 illustrates how different directive speech acts can be ordered along a scale depending on how costly they are, and how the need for politeness increases with the cost. Figure 2.1. Cost-benefit and politeness interactions (Leech, 1983: 107) A costly action such as peeling the potatoes would require a more polite linguistic realisation than the offer Have another sandwich, which is beneficial to the hearer. Directive speech acts are very sensitive to social factors and their cost is not the only variable that affects their potentiality for causing social disruption. As pointed out by Leech (2014: 11), power asymmetries between speakers (i.e. Scale of vertical distance) or the existence of a large social distance between them (i.e. Scale of horizontal distance) may also call for higher doses of politeness. In asking someone to do something, we may get by using a simple imperative like Peel the potatoes when talking to a sibling or a friend, because they are close to us and hold a similar power status. But if we wanted a stranger to comply with that costly request, we would preferably make use of a more polite realisation (e.g. If you don’t mind, will you peel the potatoes, please?). Additionally, the degree of politeness of a directive speech act can be modulated by varying the optionality offered to the hearer to comply with the action (Optionality Scale) or the 30 indirectness of the act (i.e. the length of the inferential path linking the illocutionary act to its communicative goal, Indirectness Scale). Requests expressed by means of interrogative sentences display a higher degree of optionality by default. By way of illustration, an utterance like Can you peel the potatoes, please? leaves a door open for the hearer to express noncompliance. On the contrary, imperatives present the action as unavoidable: Peel the potatoes! Thus, leaving the hearer without a chance to say ‘no’ is felt as impositive, and therefore, less polite. Likewise, the more indirect a request is, the more its politeness increases, since it is easier for the hearer to deny understanding of the intended meaning, and hence avoid compliance without losing facing. A highly explicit request expressed by a performative (e.g. I request you to peel the potatoes) leaves little option to the hearer but to comply with it. Explicitness reduces optionality, and hence, the politeness of the act. A highly implicit request like Has someone peeled the potatoes yet?, on the contrary, allows the hearer to overlook the requestive interpretation and answer to the question force of the utterance with a simple ‘yes’, ‘no’, or ‘I don’t know’. Scales, like those of cost-benefit, optionality, indirectness, social power, social distance, etc. allow speakers to assess the politeness needed in a particular interaction and to act accordingly, thus keeping their social interactions largely free from the conflict inherent to costly speech acts like directives. In his more recent 2014 revision of the politeness of speech acts, Leech (2014: 14) further argues that pragmalinguistics should additionally be concerned with the task of revealing the ‘lexico-grammatical resources of language, their meanings […], and how they are deployed as linguistic strategies of politeness’. In this, Leech paves the way for a constructional account of illocution such as the one we shall propose in section 2.5. His model, however, runs short of explaining the conceptual operations that work on illocutionary constructions and which are responsible for their flexibility and adaptability to different contexts and interactional needs (i.e. conceptual metonymy). These issues will be addressed in our proposal. Leech’s scales are also useful in distinguishing between different subtypes of directives. Thus, orders, requests, and beggings involve a cost to the hearer, while pieces of advice and warnings are beneficial to them. Requests and suggestions are characterised by their high optionality, as opposed to orders or threats. Orders are typically associated with powerful speakers, while requests and beggings are not. In Chapter 4, we shall make extensive use of Leech’s scales in our description of the semantic side of illocutionary constructions. 31 2.4.3 Direct Access Approaches II: Conversational Approaches Conversational approaches represent a mainstream group of speech act accounts that share with others in team 3 their rejection of the Literal Force Hypothesis and their placing inference in the centre of speech act performance (e.g. Schegloff, 1979, 2007; Levinson, 1983; Mey, 1993; Geis, 1995; Kasper, 2006). These accounts focus on suprasentential aspects of illocutionary acts in terms of pre-sequences, adjacency pairs and preferred/dispreferred responses. Levinson’s (1983: 361) well-known example of a request for cigarettes illustrates these notions: (6) A: Do you have Marlboros? (T1) B: Yeah, hard or soft? (T2) A: Soft, please. (T3) B: Okay. (T4) For a request to be successful it is necessary that the addressee possesses the object that the speakers wants. Thus, it could be argued that, in example (6), T1 is a simple question about possession, and the actual request is formulated in T3 (i.e. Soft, please). Levinson points out that the recurrent occurrence of conversational exchanges like the one in example (6) eventually leads speakers to recognise pre-sequences like T1 as full requests by dropping the second and third turns of the conversational sequences. Example (6) also illustrates the notions of adjacency pairs and preferred/dispreferred response. The notion of adjacency pair constitutes a relevant contribution of conversational approaches to the study of illocution. Locastro (2012) explains that it is usually the case that speech acts require an answer by the addressee in the form of a confirmation, ratification or rejection. Questions, for instance, form adjacency pairs with positive or negative answers. Requests can be followed by the acts of compliance, or declination to comply, the first one being the preferred option. Example (6) is formed by two adjacency pairs (i.e. question-answer; request-answer). In addition, preference organisation (Glaser, 2009; Carroll, 2011) explains that some of those answers are conversationally preferred (i.e. those that agree with the position of the speaker taking the first action), and others are dispreferred (i.e. those that show disagreement, refusal, or rejection). Preferred turns are usually unmarked; and therefore, they are not preceded by silence or produced with delays or mitigation, as is the case with dispreferred turns. In the request adjacency pair in example (6), speaker B has chosen the 32 preferred response (i.e. compliance), which fits the social expectation that one should help others if possible. As illustrated by the analysis of example (6), conversational linguists believe that illocutionary meaning is not static and dependent on the speaker’s intention, as proposed by pragmatists and speech act theorists, but that it is rather built through interaction. A consequence of this is the emphasis they place on the analysis of large corpora of real data and their attention to suprasentential aspects of speech act performance. In this, the conversational analysis approach to illocution also differs from previous studies carried out within traditional speech act theory, which focus their attention on invented examples, most of them restricted to the limits of the sentence level. Conversational analysis transcends these boundaries and offers a more realistic account of speech acts. The advantages of considering conversational issues in the correct performance of speech acts have been amply dealt with in the literature, and this is in fact one of the few research advances that has already made its way into current EFL textbooks (see Pérez-Hernández, 2019). Leaving aside the ground-breaking contributions to the study of the interactional, social, and conversational aspects of illocution reviewed in sections 2.4.2 and 2.4.3, inference-based accounts of illocution in team 3, as a whole, are characterised by their rejection of the Literal Force Hypothesis. This allows extreme inferentialists to provide a unified account of how speech acts are understood in terms of inferential calculations. The account, however, is inconsistent with typological evidence supporting the fact that most languages display three universal sentence types (e.g. imperative, affirmative, interrogative). It seems counter-intuitive that these three linguistic forms would not be somehow linked to three salient meanings or functions that are equally pivotal in all cultures. Section 2.5 reveals how a weaker version of the Literal Force Hypothesis may reconcile the typological evidence about the existence of three universal sentence types with their use in the expression of different illocutionary acts beyond the three basic ones of ordering, asserting, and questioning. An extreme inferentialist stance on speech acts also fails to acknowledge the fact that some linguistic expressions are more frequently and extensively used than others for communicating a speech act force (e.g. CAN YOU DO X? forms are more widely used for expressing a request than the synonymous ARE YOU ABLE TO DO X? linguistic realisations). The conventional status of some speech acts, as revealed by those linguists in team 2, is a reality that cannot be overlooked. In fact, some linguists in team 3 have restated their views on this issue. Leech (2014: 14), for instance, has recently acknowledged the fact that some expressions may become highly conventionalised for the performance of some speech acts. He refers to the process as 33 pragmaticalisation, and further argues that pragmalinguistics should be concerned with the lexico-grammatical resources of language and with how their frequency of use leads to their conventionalisation/pragmaticalisation. In section 2.5, we shall argue that a sound and exhaustive account of speech acts should also aim at explaining the motivations underlying the wider use of conventional forms. 2.5 A Cognitive-Constructional Approach to Directive Speech Acts The revision of previous theoretical accounts of speech acts in sections 2.2 to 2.4 reveals that the portrayals of illocutionary performance are diverse and often even contradictory. If we thought of those portraits as jigsaw puzzles, some of them would have relevant gaps. Thus, social considerations are not taken into account in Searle and Morgan’s accounts, the role of inference is overlooked in Halliday’s functional-systemic approach, the possibility of full codification is not considered by Bach and Harnish or Leech, and the motivations of conventionality are not clear in most of them. These are just some of the missing pieces that render traditional accounts of illocution somehow incomplete. While the need to understand speech acts as social constructs (Leech, 2014) rather than as exclusively linguistic in nature (Searle, 1975), or the advantages of considering the conversational dimension of speech acts (Levinson, 1983; Geis, 1995; Schegloff, 2007) have gradually been accepted in the literature; other issues, like the compatibility of the existing typological evidence supporting the Literal Force Hypothesis, on the one hand, with the role played by codification, conventionalisation, and inference in the performance of speech acts, on the other, are still open to debate. In fact, these three pivotal pieces concerning the relative weight of linguistic form, convention, and inference in the performance of speech acts have not yet found their exact place in the puzzle. So far, as shown in Sections 2.2 to 2.4, theoretical efforts have in most cases been devoted to justifying the predominance of one of them to the exclusion of the others. 2.5.1 What Experimental Linguistics Has Revealed about Speech Act Processing In this connection, experimental studies aimed at assessing the psychological reality of those competing models have been scarce and often contradictory. Back in the 1970s, Clark and Lucy’s (1975) experiments suggested that the interpretation of indirect speech acts took longer than that of literal speech acts, which was taken as proof that the Literal Force Hypothesis is valid, and that it is necessary to recognise the direct force of an utterance prior to the understanding of the indirect meaning intended by the speaker. More recently, eye-tracking 40 usually the case that this information can be directly retrieved from the context. Let us see each mechanism in turn. On the one hand, the exact illocutionary force could be further specified through linguistics means, as illustrated by the following elaborations of example (7): (8) Turn down the volume, will you?/please/if you don’t mind (9) Turn down the volume if you don’t want to become deaf (10) Turn down the volume or I’ll punish you (11) Please, please, please, turn down the volume Each of the examples above activates a key attribute of a specific directive speech act, thus cutting down or straightforwardly restricting the interpretation of the illocutionary force of the initial imperative to that of a particular speech act. In example (8), the adverb please, the question tag, or the conditional phrase explicitly convey the fact that the addressee has freedom to decide whether to do the action or not. This element of optionality is incompatible with directives like orders and threats, but central to requests, which turn out to be the most likely interpretation of this example. The utterance in (9) makes use of a conditional phrase that states the negative consequences that the action of hearing the radio at a high volume may have for the listener. Helping others avoid costs or harms is central to the nature of the act of warning, and the imperative is thus very likely to be interpreted as such. In example (10) the juxtaposed phrase communicates the action that the speaker will carry out against the listener if he does not comply. Stating negative actions against the addressee in case of non-compliance is a pivotal aspect of threats, which is the most predictable reading of this imperative. Finally, the iterative use of politeness markers (e.g. please, please, please) in example (11) reflects an insistence which stems from the lack of power of the speaker, and the intensity of his wanting, both of which define the act of begging. In all examples above the use of additional linguistic resources has explicitly activated items of the semantics of different directive categories, thus guiding the listener to the most likely interpretations. The clues leading to a final interpretation of the initial imperative in (7) need not be linguistic, they can also stem from the context in which the utterance has been produced. Thus, imagine that example (7) is uttered by a boy sharing his room with one of his friends. In this context, it is already manifest to both speaker and listener that the latter has freedom to choose whether or not to do the action. The optionality feature that characterises the act of requesting is already active and its knowledge mutually shared by both interactants. This makes it unnecessary to use additional linguistic resources to make it explicit (e.g. please, if you don’t 41 mind, etc.). This mutually shared knowledge of the listener’s optionality makes the order, threat, begging readings easily discarded, and leads to a request interpretation. Relying on contextual information is clearly riskier. Speakers may assume that there is mutual knowledge of the necessary pieces of information for the interpretation of an illocutionary act, and this may not be the case. However, all things being equal, it can be entertained, as a theoretical hypothesis, that the interpretation of sentence (7) as a request can be reached as fast when the optionality variable is activated by the context as when it is activated by linguistic resources. In fact, this theoretical proposal fits the experimental data available to date. As shown in Section 2.5.1, those experiments which were rich in providing contextual information retrieved fast response times for so-called indirect speech acts (see section 2.5.1 on Gibbs’, Ervin-Tripp et al.’s and Gibbs & Gerrig’s experiments). And it was also the case that an increase in the linguistic specification of the force of the act also yielded lower response times for conventional speech acts (see section 2.5.1 on Clark’s, Holtgraves’, and Abbeduto, Furman & Davies’ experimental findings). Those seemingly contradictory results stemming for experimental studies of speech acts find an explanation in the present theoretical proposal that advocates for a weaker form of the Literal Force Hypothesis in terms of sentence type-speech act compatibility. 2.5.3 Revisiting the Notions of Direct and Indirect Speech Acts Accepting the revised weaker version of the Literal Force Hypothesis put forward in Section 2.5.2 naturally leads to a redefinition of the concepts of direct and indirect speech acts. It can be concluded that a speech act is direct when the addressee has enough available information to be able to recognise its illocutionary force. This information can be provided linguistically, or contextually, as shown in the previous section. Either way, if there is enough information at the addressee’s disposal, the interpretation of the illocutionary force of an utterance will be reached easily and unproblematically. On the contrary, a speech act is indirect if the information available is not so rich. Lack of information, either contextual or linguistic, necessary implies a higher cognitive cost, as well as a higher risk of misinterpreting the intended force of the utterance. It is important, in this respect, to distinguish clearly between the notions of indirectness and implicitness. In the present account, whether a speech act is explicit or implicit is a linguistic issue dependent on the degree to which his conceptual nature has been linguistically revealed. Whether a speech act is direct or indirect, on the contrary, depends on how much information the speaker has about the conceptual nature of the act, regardless of whether that information 42 is provided by contextual or linguistic means. Thus, as argued in Section 2.5.2, a bare imperative, like the one in example (7), is implicit (i.e. the lack of explicitness of its linguistic form does not provide enough information to determine its illocutionary force), but not necessarily indirect. If the context provides the addressee with the relevant information, the illocutionary interpretation of the imperative will be straightforward (i.e. direct). It should be noticed that, since our revised version of the Literal Force Hypothesis does not assign specific speech act forces to the sentence types, ‘indirectness’ does not involve a two-step process in which the literal force needs to be reinterpreted according to the context, but rather just a higher amount of inferential calculations to determine which of the illocutionary forces compatible with that sentence form is the one intended by the speaker. The notions of indirectness and implicitness have often been merged in traditional theories of speech acts. The weaker version of the Literal Force Hypothesis assumed in this piece of research allows us to draw a clear distinction between them. In addition, it may offer a plausible solution to the experimental puzzle, reported in Section 2.5.1, regarding the fact that the process of interpretation of indirect speech acts offered different response times in different experiments depending on the degree of contextual cues available to the addressee. Under the present redefinition of the notions of indirectness and implicitness, it is to be expected that those so called ‘indirect’ speech acts that are not only implicit (i.e. linguistically underdetermined), but also indirect (i.e. lacking relevant information, either linguistic or contextual about their semantics) will show higher response times than those which are implicit, but not indirect (i.e. those for which the context provides relevant cues, thus making up for the underdetermination of the linguistic form). Interestingly enough, the newly re-defined notions of direct and indirect speech acts are capable of showing prototypicality effects. The indirectness of a speech act may vary depending on the amount of information provided by the context, through linguistic means, or through a combination of both. Consider the following examples of requests: (12) Open the door (13) Can you open the door? (14) If you don’t mind, can you open the door? (15) If you don’t mind, can you open the door, please? (16) If you don’t mind, could you open the door, please? 43 Let’s assume a situation between two strangers in which the context does not provide any relevant information that may help the listener identify the illocutionary forces of these utterances. In this situation, example (12) would come through as a rather indirect speech act, which would probably leave the listener wondering if he is being ordered, suggested, or requested to carry out the action. The lack of contextual information, together with the linguistic under specification of the utterance (i.e. a simple imperative sentence presenting a proposition for its realisation) does not ease the interpretation process. Example (13), which makes use of an interrogative sentence type presents the proposition as partially open. This is compatible with the illocutionary force of questioning, but also with that of requests (which are characterised by their optionality, see Chapter 4). The optionality feature, implicitly activated by the use of the interrogative sentence type, also makes the utterance incompatible with other illocutionary forces that do not present their content as open, but rather as compulsory (e.g. orders, commands, threats, etc.). Consequently, the range of potential interpretations of example (13) has been reduced and its indirectness has decreased. Examples (14) to (15), activate the attribute of optionality more explicitly, not only by means of the openness of the interrogative sentence, but also through the use of the phrase if you don’t mind, which clearly leaves the decision of carrying out the action in the hands of the listener. In addition, examples (15) and (16) also activate linguistically the attribute of politeness (please, use of past modal could), which is characteristic of requests, but not compatible with questions for information. As the explicitness of the linguistic forms increases, the indirectness of the utterances gradually decreases to the point that examples (15) and (16) could hardly be interpreted as anything other than requests. The previous explanation of examples (12) to (16) assumes a neutral context, which does not provide speakers with relevant information that can guide them in the interpretation process. The latter, therefore, relies exclusively on linguistic cues. Under those circumstances, example (12) is fully indirect, while examples (13) to (16) gradually increase their degree of directness. Does this mean that simple imperatives like (12) are always indirect requests simply because they are linguistically underspecified? Under the present proposal, the answer is no. Let’s imagine a rich context in which the speakers have a similar social status, in which it is clear that neither of them can impose his will on the other (e.g. siblings). In this context, it is mutually known by both speakers that politeness requirements are low or inexistent, and that optionality is taken for granted (i.e. one does not feel under an obligation to act as told by their equals if one does not want to). The context itself activates the optionality and politeness attributes 44 characteristic of requests and fills in the necessary information for the listener to interpret (12) as a request and not as an order. In this context, the simple imperative is as direct and straightforward an example of request as example (16) was shown to be in the neutral, uninformative context above. Indirectness would only arise in those cases in which both the context and the linguistic expression fail to provide the necessary information to identify the intended illocutionary force of an utterance. This model also fits well with one piece of experimental evidence about speech act processing that has puzzled linguists up to date: the fact that examples like (12), which have traditionally been considered indirect speech acts, seem to involve higher memory retrieval (Coulson & Lovett, 2010). Going back to our explanation of examples (12) to (16) in the presence vs. the absence of contextual and/or linguistic information, it is easy to see in which way the interpretation processes differ and the memory retrieval requirements increase significantly in relation to example (12). In the presence of linguistic information, the listener needs only retrieve his knowledge of the semantic attributes characterizing the different speech acts compatible with the sentence type, and then see which one matches those attributes that are activated linguistically. Thus, example (16) activates the attributes of politeness and optionality that prototypically define requests. The listener only needs to confirm this matching, by retrieving from his memory the semantics of this particular speech act type. When the information derives from the context and not from linguistic cues, the process of interpretation is not as straightforward. In the case of the simple imperative in (12), the context tells us that the speakers, being siblings, are equals, as regards their power status, and close to each other in terms of social distance, but it does not activate explicitly the attributes of optionality and politeness as the linguistic cues do in example (16). These semantic attributes of requests are reached inferentially and through a process that involves memory retrieval not only of the speech act categories compatible with the sentence type, but also of other relevant principles and maxims of social interaction. Thus, the politeness principle and the social distance scale, for instance, inform the listener that politeness is not as necessary in interactions between people who are socially close to each other, thus making unnecessary the use of a polite request; the optionality scale tells him that the closer two people are in the social distance dimension, the less one can impose on the other, thus ruling out the potential imposing flavour of the bare imperative. Since optionality is presumed in this context, many speech acts types are discarded from the interpretation (i.e. orders, threats, commands). Additionally, since politeness is not necessary, other directives are ruled out (i.e. beggings, which require of 45 politeness even when carried out between socially closed speakers). As a result of these context-based inferences, the request interpretation is reached. As spelled out in the previous paragraphs, the present theoretical model is capable of offering an explanation for the higher memory retrieval requirements of those speech acts whose interpretation relies exclusively on contextual information. By way of summarizing, linguistic information can directly activate semantic attributes of speech act categories. Contextual information can only activate social variables (i.e. social power, social distance, formality), and then the speaker needs to further retrieve his knowledge (1) of the semantics of the different speech act categories that are compatible with the sentence form of the utterance, and (2) of the principles and scales of social interaction that act on the semantics of speech act categories, helping listeners to infer which speech act categories are compatible both with the sentence type of the utterance and with those social scenarios. Thus, the higher memory retrieval requirements of non-conventional indirect speech acts, as attested in experimental studies, find a plausible explanation within this model of illocution. 2.5.4 Constraining Inferences via Cognitive Operations: Conceptual Metonymy As illustrated in 2.5.3, the interpretation of speech acts involves a process in which the listener identifies the illocutionary force intended by the speaker, among those that are compatible with the sentence type used for its expression. This process is guided and enabled by cues provided either contextually or linguistically. Such cues have been shown to reflect one or more of the characteristic attributes of the speech act type involved (e.g. optionality, politeness, etc.). If all the attributes of an illocutionary category were made explicit, the listener would not have any difficulty in interpreting such a direct instance of speech act. For different reasons, however, the fact is that fully explicit, direct illocutions are scarce. One obvious reason is that speech acts are polyhedral, kaleidoscopic phenomena whose semantics comprises varied information regarding interactional, transactional and social features. To give just one example, the meaning of requests includes the knowledge of who is expected to perform the requested action (i.e. addressee), who is the beneficiary of the action (i.e. speaker), the cost of the action and the person upon whom it will fall (i.e. addressee), the degree of optionality that the addressee has to decide as to whether to perform the requested action or not (i.e. optionality), the politeness requirements of the act of requesting (i.e. request are expected to be polite), and the knowledge of how the cost of the requested action, the social distance and power relationships between the speaker and the addressee, and the formality of the context may affect the politeness requirements of the act, etc. Making all these semantic ingredients of requests explicit in their 46 linguistic expression would turn communication into a cumbersome exchange of long-winded utterances. Often this is also unnecessary, because some of the semantic attributes are provided by the context. And many times, it is not even desirable, because by not making all the semantic items explicit, the speaker can regulate the indirectness of her illocutionary acts, thus increasing the optionality granted to the addressee, decreasing the imposing feeling of her speech acts, and eventually increasing their politeness. In other words, lack of explicitness can be exploited as a means of producing speech acts that do not threaten social conviviality. Granted that full explicit rendering of all the features defining an illocutionary category is often not an option, the question that arises is how it is possible for the addressee to identify the speech act category intended by the speaker on the basis of the linguistic expression of just a few of the semantic attributes that define it. Cognitive linguists have provided an answer to this question based on the interaction of two cognitive models: a non-dynamic model of propositional knowledge organisation (whose nature may vary slightly in different approaches and may be referred to as illocutionary frame, propositional cognitive model or scenario) and a dynamic cognitive model that operates on the former (e.g. conceptual metonymy). 8 The first attempt to explain the interpretation of speech acts in these terms was put forward by Panther and Thornburg (1998: 759). These authors coined the term illocutionary scenario to refer to the collection of semantic attributes defining each illocutionary category that speakers have stored in their long-term memories. The request scenario was said to consist of the following elements: (a) Before component: The hearer (H) can do the action (A). The speaker (S) wants H to do A. (b) Core component: S puts H under a (more or less) obligation to do A. H is under an obligation to do A (H must/should/ought to do A). (c) After component: H will do A. S has emotional response. 8 The distinction between dynamic and non-dynamic cognitive models stems from Ruiz de Mendoza and Galera Masegosa (2014) work on cognitive modelling. 47 Speakers do not make explicit all the knowledge they have about requests when they want to perform this speech act. Panther and Thornburg’s proposal elaborates on Gibbs’ (1994: 352) insight that ‘speaking and understanding indirect speech acts involves a kind of metonymic reasoning, where people infer wholes (a series of actions) from a part’. Thus, Panther and Thornburg argue that, in order to produce a request, speakers explicitly activate one of the components of the scenario (e.g. I would like/want a coke, Can you hold on a sec? - Before component) to evoke the whole illocutionary act. In this way, metonymic projections on illocutionary scenarios provide natural inferential schemata for the performance of speech acts. Panther and Thornburg’s model of illocutionary scenarios reflects the transactional and dynamic nature of speech acts by focusing on the temporal sequence of events characterizing them (i.e. before, core, after components). Nevertheless, it largely overlooks the social and interactional side of illocutions. Together with the illocutionary scenario components, knowledge about the act of requesting also comprises facts about the politeness needed to make it successful, and about how social aspects like the degree of intimacy or power between the speakers, the cost of the requested action, or the formality of the context affect the politeness requirements of the act (Pérez-Hernández, 2001; Huddleston & Pullum, 2002; PérezHernández & Ruiz de Mendoza, 2002, Mauri & Sansò, 2011; Takahashi, 2012). Formalizing all this rich transactional and interactional information about speech act categories exceeds the potentiality of illocutionary scenarios and asks for a more comprehensive model of knowledge organisation, as well as for more specific types of metonymic projections. Section 2.5.5. deals with these issues in detail and lays out a constructional model of illocution that attempts to explain how speech act performance is operationalised. 2.5.5 Illocutionary Idealised Cognitive Models and (Multiple Source)-in-Target Metonymies In the light of the above observations, a more granular conceptual characterisation of illocutionary categories is required to account for all the possible metonymic manifestations of a particular speech act such as the request instances analysed in the previous examples. In this connection, different authors such as Pérez-Hernández (2001, 2013), Huddleston and Pullum (2002), Mauri and Sansò (2011), and Takahashi (2012) have argued that additional parameters like those of power, cost, benefit, optionality, and politeness, among others, would be essential for a comprehensive, solid conceptual characterisation of directive speech acts. As shown above, this type of attributes exceeds the limits of a scenario approach and require of a more comprehensive model of knowledge organisation for their formalisation. 48 Pérez-Hernández (2001, 2013) has offered corpus-based evidence about the potentiality of propositional idealised cognitive models (henceforth illocutionary ICMs) for this task. Illocutionary ICMs consist of an ontology and a structure (Lakoff, 1987: 285). In relation to illocutionary acts, the ontology would comprise the values taken by the attributes conforming the semantics of each speech act category (e.g. addressee’s capability, speaker’s need, costbenefit, optionality, politeness, etc.). In turn, the structure would capture the interplays between the different attributes (e.g. how the cost of the requested action affects the politeness requirements of the act), as well as between those attributes and certain extralinguistic variables that may affect their assessment (e.g. how the power relationship between the speakers determines the optionality of the addressee to decide about his course of action). Since our description of directive illocutions in Chapter 4 will be based on this theoretical construct, let us look in more detail into the nature of the ontology and structure of directive speech acts in turn. The ontology of directive illocutionary ICMs comprises the attributes listed in Table 2.2 (elaboration of the initial proposal in Pérez-Hernández (2013: 133-134) based on the new corpus data for the present study): Table 2.2. Attributes configuring the ONTOLOGY of directive illocutionary ICMs Agent: person who is expected to perform the proposed action (i.e. speaker, addressee, both of them). Beneficiary of the action: person who benefits from the action (i.e. speaker, addressee, or both). Agent’s capability: assessment of the agent’s capability to perform the proposed action. Speaker’s willingness: assessment of the speaker’s desire that the proposed action is carried out. Addressee’s willingness: assessment of the addressee’s desire to perform the proposed action. Possession of the requested object: assessment of possession of the requested object by the speaker or the addressee. Speaker’s need: assessment of the speaker’s need that the proposed action is carried out. Cost-benefit: assessment of the cost/benefit of the proposed action. Optionality: assessment of the degree of freedom granted to the agent to carry out the proposed action or to opt out. Mitigation: assessment of the degree to which the cost of the proposed action is mitigated. Politeness: assessment of the politeness requirements of the illocutionary act. As revealed in previous sections, the description of the semantic side of illocutionary acts should include considerations about the participants (e.g. agent’s capability, speaker’s needs, etc.), dimensions of social interaction (e.g. politeness, optionality, etc.) and aspects of the transactions involved (e.g. cost/benefit, possession of the requested object). By way of illustration, Table 2.3 displays the specific ontology of the illocutionary ICM of requesting, 49 which captures the prototypical parametrisation of the attributes conforming this illocutionary act. Table 2.3. Ontology for the illocutionary ICM of the act of requesting Agent: addressee Beneficiary: speaker Agent’s capability: the agent needs to be capable of performing the requested action. Speaker’s willingness: the speaker wants the requested action to be carried out. Addressee’s willingness: the addressee’s willingness to carry out the action is required. Requested object: the requested object should be in the possession of the agent or accessible to him. Speaker’s need: the speaker needs the requested action to be carried out. Cost-benefit: the requested action is typically costly. Optionality: the addressee is typically free to decide upon his course of action in relation to the request. Mitigation: the cost of the requested action is mitigated. Politeness: requests are polite speech acts. As shall be made apparent in the remainder of this section, the exhaustive description of the semantic attributes that define the act of requesting in the form of an illocutionary ICM enhances the scope of its metonymic exploitation. Additionally, it increases the explanatory power of a metonymic account of illocution, overcoming some of the limitations of the illocutionary scenario approach, at least in four different respects: 1. Illocutionary ICMs allow speakers to distinguish different speech act categories. According to the illocutionary scenario approach, in sentences (17) and (18) the core component of the request scenario is activated by means of the imperative sentence, which puts the hearer under the obligation to hold the book for the speaker. If they both activate the same component, it would logically follow that they realise the same act of requesting. However, while the adverb of immediateness now prompts an order interpretation, the conditional clause if you don’t mind favours a request reading. The explanatory power of the scenario approach is not capable of distinguishing specific illocutionary acts (e.g. orders, requests, warnings, etc.) within the broader category of directive speech acts. (17) Hold this book now! [Web Corp Live, accessed August 5th, 2012)] (18) Hold this book, if you don’t mind. [Web Corp Live, accessed August 5th, 2012)] 56 As shown so far, the ontology of each illocutionary ICM captures the attributes that define the prototypical instances of a particular speech act category and serves as basis for the metonymic activation of the particular illocutionary constructions. As captured in Table X, some of the attributes are scalar in nature, and may take different values in different contexts (e.g. the cost of the requested action may vary, the speaker’s need or desire that the action is carried out may also change from one situation to another, etc.). More importantly, the values taken by one attribute may affect those of others. To give just one example of the potential interactions between attributes, if the cost of the requested action increases (i.e. if the speaker is asking for something that has a high cost in time or effort for the addressee), then the need to mitigate the cost of the act and the politeness requirements to secure compliance with the request will also increase. Additionally, the values taken by the attributes conforming the ontology of the illocutionary ICM may also be affected by some social variables, such as the power relationship between the participants, the social distance between them, and the formality of the context in which the interaction is taking place. Thus, if the social power of the addressee is markedly higher than that of the speaker, the latter will feel a need to increase the optionality, the politeness and/or the mitigation of his act to guarantee its success. This is so because speakers are aware of the workings of social interaction and have learned through experience that imposing their wishes on someone who is socially more powerful requires some tact. These interplays between the attributes of an illocutionary ICM or between one of these attributes and one or more social variables (social power, social distance, formality) need to be formalised to reach a proper understanding of the functioning of speech acts. In the present proposal, such interplays are included in the so-called structure of each illocutionary ICM. The social variables conforming such structure are defined in Table 2.4. Table 2.4. Social variables configuring the STRUCTURE of directive illocutionary ICMs Power: power relationship holding between the speakers. The vertical asymmetries of power between speakers can stem from their social status as members of different social classes, professional categories, institutional ranks, or knowledge groups. Social Distance: social distance existing between the speakers. This is a horizontal dimension pertaining to familiarity or friendship relationships (i.e. friends, strangers, relatives). Formality: the formality of the context in terms of its politeness requirements. 57 These social, extralinguistic variables interact with four of the ontological attributes (i.e. costbenefit, optionality, mitigation, and politeness), forcing their reassessment. The specific structure for each illocutionary ICM will comprise these interactions together with those holding between the ontological attributes themselves. Thus, by way of illustration, in relation to the illocutionary act of requesting, some of the interactions conforming the structure of its illocutionary ICM are captured in Table 2.5. Table 2.5. STRUCTURE of the illocutionary ICM of the act of requesting in terms of interactions between attributes, and between attributes and social variables Power-Optionality: the relative power of the speakers affects the optionality of the addressee to perform the action. Thus, the more powerful the speaker, the lower the addressee’s freedom to decide upon his course of action and vice versa. Social distance-Optionality: the closer the speakers are in the social distance axis (i.e. the higher their degree of intimacy), the higher the optionality that the addressee has to choose upon his course of action, and vice versa. Optionality-Politeness: the lower the amount of optionality granted to the speaker to decide upon his course of action, the lower the politeness of the speech act, and vice versa. Cost/benefit-politeness: the higher the cost of the requested action, the higher the politeness needs will be in order to secure the addressee’s compliance, and vice versa. These interplays are specific of each illocutionary act category and they guide and constrain the speakers’ choice of realisation procedures for the performance of their speech acts. Thus, even though prototypical requests are characterised by their politeness, speakers know that in a context of social familiarity (e.g. mum-son relationship), optionality increases, and that this, in turn, makes politeness requirements decrease. Thus, in such a context, an apparently nonpolite formula (e.g. a simple imperative like Mum, pass the salt) will be unproblematically understood as a request. Before ending this section devoted to the semantics of illocutionary categories, it is necessary to bring to the front a relevant theoretical postulate offered as far back as the 80s by Givon (1989). As this cognitive-functional linguist remarked, it is essential to approach the study of the semantic fabric of speech acts from a typological perspective. Cross-cultural studies of illocution will reveal that societies often differ as to the politeness requirements needed and/or expected for the performance of a particular speech act, as well as to the degree to which social variables (power, social distance, formality) affect the realisation of illocutionary acts. Explicit instruction of these subtle cultural differences in the semantics of speech acts is pivotal for the correct learning of the illocutionary component of a second language. 58 2.5.6 Assembling the Illocutionary Puzzle: Families of Illocutionary Constructions The constructional treatment of speech acts is an old longing of linguists from all traditions. The Literal Force Hypothesis itself was an attempt to pair linguistic forms (sentence types) with particular illocutionary forces. Since the very beginning of speech act theory, linguists have tried to identify fixed form-meaning pairings for the expression of speech acts. Communication would be all the more straightforward and unproblematic if speakers had at their disposal clear linguistic formulae with which to communicate their wishes unequivocally. Following this path of thinking, Searle (1969) put forward the notion of illocutionary force indicating devices (henceforth IFIDs), which were expressions that were thought to be activators of a certain illocutionary category. He observed that I’LL DO X formulae, such as I’ll buy you a bike are conventionally understood as promises, and CAN YOU DO X? expressions activate the illocutionary force of requesting. Dik (1989, 1997) introduced the notion of illocutionary converter. These are certain linguistic items that can change the illocutionary force of an utterance, such as the tag question will you? or the adverb please, which can turn an imperative, usually associated with orders, into a request (e.g. Turn off the radio, will you?/please). These and similar attempts have not withheld the test of use, since all the examples corresponding to Searle’s IFIDs, as well as those stemming from the application of Dik’s (1989) illocutionary converters, can ultimately express other illocutionary forces in alternative contexts (e.g. I’ll buy you a bike tomorrow, for example, need not be a promise, but just a statement of intention). As explained in Pérez-Hernández (2013: 140), more recently, Stefanowitsch (2003) has made a strong case in support of the constructional nature of some speech act expressions. He argues that some request constructions display a number of semantic and formal properties that cannot be predicted from their component parts or from previous constructions. Thus, the CAN YOU DO X? request construction, according to Stefanowitsch (2003: 16), displays an agentive subject and allows preverbal request markers (e.g. please, kindly), alternation of present and conditional forms (e.g. Can/could you do X?), and co-occurrence with a preposed subordinate clause explaining the speaker’s reason for making the request (e.g. Since you are stronger than me, can/could you please hold this?). This view is largely shared by most contemporary constructional accounts of illocution (Panther & Thornburg, 2005; Ruiz de Mendoza & Baicchi, 2007; Brdar-Szabó, 2009; Del Campo, 2013). However, this proposal does not account for the fact that an expression like Can you hold this for me?, which contains no request marker and no preposed subordinate clause, is still 59 unequivocally understood as a request. Contemporary constructional accounts of illocutions also fail at explaining the different degrees of explicitation of request expressions. As shown in Section 2.5.5, illocutionary constructions can be made more explicit as a higher number of attributes of an illocutionary ICM are linguistically realised: Can you hold this? > Can you hold this for me?> Can you please hold this for me?>Could you please hold this for me?> If you don’t mind/If it’s not too much trouble, could you please hold this for me for just a sec?, etc. Pairing full linguistic forms with illocutionary meanings has long failed as an attempt to describe illocutionary constructions. As was made apparent in Section 2.5.5 in our description of illocutionary ICMs, the conceptual nature of speech acts is too complex and dynamic to be confined into a fixed form. As argued before, it is often the case that speakers do not want or need to make all the conceptual material of a speech act explicit, either because there is no need (the context fills in the missing information), or because it is interactionally desirable to highlight some aspects of the semantic fabric of the act over others: if a request is costly, speakers may appeal to this issue and try to minimise it (e.g. Can you step aside for a sec?), if the addressee’s cooperation is not clear, we may target his willingness as an strategy to comply with the request (e.g. If you will, could you step aside?), if there is a large power asymmetry between speakers, a more polite realisation may be preferred (e.g. Student to headmaster: Could you step aside please?), while that same politeness will become unnecessary if the social distance between them is small (e.g. Son to his mother: Can you step aside?), etc. Does this intrinsic variability make a constructional account of speech acts unfeasible? The present proposal argues that it is in fact possible to offer a constructional treatment of illocution, as long as the notion of construction is redefined and adapted to the dynamic, multidimensional nature of speech acts. Goldberg (1995, 2006) defined constructions as non-compositional learned pairings of forms and semantic and/or discourse functions in which one or more of their properties are not predictable from knowledge of other constructions in the language. Constructions vary in size, complexity and generality, and include words and idioms, as well as phrasal linguistic patterns (Goldberg & Suttle, 2010: 469). The orthodox notion of construction, as defined by Goldberg (1995, 2006) within Cognitive Linguistics, was originally designed to account for argument-structure configurations. This linguistic phenomenon, however, does not need to accommodate the rich social, transactional and interactional variability that characterises higher levels of linguistic description (pragmatics, discourse/conversation), and that affects so straightforwardly the performance of speech acts. Hence the need to adapt the original notion of construction for this purpose. 60 As pointed out by Goldberg and Suttle (2010: 469-470), ‘it is clear that we need to posit a construction when there is something non-compositional that needs to be learned [ … as well as when usage yields a] highly frequent form-meaning pairing. […] both the items and the generalisations are represented within the interrelated network of constructions: the ‘constructi-con’’. As the description of directive speech acts in Chapter 4 will show in detail, this is certainly the case with speech acts. On some occasions illocutionary constructions can be highly idiomatic, while in most cases they are non-compositional form-meaning pairings motivated by a variable number of the semantic attributes of illocutionary ICMs. Thus, expressions like WOULD YOU MIND DOING X? are idiomatically used as requests as attested by the over 7000 instances of this construction found in the iWeb corpus. In turn, CAN YOU DO X? expressions constitute highly-frequent, non-compositional constructions with varying degrees of conventionality and explicitation. This latter type of illocutionary constructions, to which we shall refer as fluid illocutionary constructions, do not exhibit a fully fixed idiomatic formulae, but rather they are capable of adapting themselves to the variable needs of interpersonal communication, modulating their degree of explicitness and their semantic focus as needed in a particular situation. Fluid illocutionary constructions consist of a fixed base structure, which metonymically activates one attribute of the corresponding speech act category, plus a variable number of optional realisation procedures, which instantiate other attributes of the same speech act as needed in each concrete interaction. Consider the following instances of requests: (26) Can you add a mildewcide to paint at the time of purchase? [iWeb] (27) Can you give me a sample of a recommendation letter? [iWeb] (28) Can you please take a look at the one linked below? [iWeb] (29) Can you describe a bit more the problem with combat controls? [iWeb] (30) If possible, can you post the solution in the comments? [iWeb] (31) Horace if it’s not too much trouble can you post the locations of the events here? [iWeb] All examples (26)-(31) share the same base structure which consists in an interrogative sentence of the CAN YOU DO X? type. This base structure metonymically activates two attributes of the ICM of requests: the addressee’s capability to perform an action, by means of the use of the modal verb can, and the optionality attribute through the use of the interrogative sentence type, which presents an event for its potential realisation, but also leaves open the possibility for the event not taking place at all. The constructional nature of this base structure 61 for the expression of requests is attested by its high frequency of occurrence in the iWeb corpus, in contrast to similar expressions, such as the ARE YOU ABLE TO DO X? construction, which activates the same attributes, but which has specialised as a question for information. Thus, a random search of 5,000 instances of the CAN YOU DO X? construction shows that 4,304 instances correspond to requests and only 696 to questions for information. Those CAN YOU DO X? instances that function as questions for information display perception or understanding verbs (e.g. hear, see, understand, etc.), which unlike action verbs (e.g. give, read, sing, etc.), cannot have the speaker as their beneficiary, and therefore, do not comply with one of the central attributes of the ICM of requesting. On the contrary, a similar search for the ARE YOU ABLE TO DO X? structure reveals that this construction is used mainly as a question for information in 4987 of the cases (e.g. Are you able to state positively whether or not that indentation was present before the shooting?), and as a request in just 13 occurrences, all of which include the politeness marker please (e.g. Are you able to please provide us some more details regarding your hire?). Experimental studies have argued (Ruytenbeek, Ostashchenko, & Kissine, 2017) that ARE YOU ABLE TO DO X? expressions are understood as requests as easily as CAN YOU DO X? constructions. This is only to be expected, since both expressions activate the same attributes of the ICM of requesting. However, the fact that both of them could be understood as requests does not mean that speakers use them interchangeably. The aforementioned corpus-based evidence shows a high degree of conventionalisation of the CAN YOU DO X? base structure for the expression of requests. In fact, its specialisation for the expression of this illocutionary force becomes manifest in examples like If you are able, can you check what his current overall ranking is now? (iWeb), where the division of work between these two constructions becomes transparent. It has also been argued in the literature that CAN YOU DO X? forms may also be used to perform other directive acts, like orders or begs. For example, in a context in which the speaker is socially more powerful than the addressee, CAN YOU DO X? may be used to express a polite order. Again, although possible in very specific contexts, corpus data indicates that this is not the prototypical use of this construction. The random sample of 5,000 CAN YOU DO X? instances yields no occurrences in which this construction is used to express an order. These findings show that previous studies were mistaking the theoretical potential of the construction with its actual use. The CAN YOU DO X? formula is rarely found instantiating any other speech act type than a request in actual usage, and when it does, it corresponds to non-prototypical instances of other illocutionary categories such as the aforementioned polite order. 62 In addition, the CAN YOU DO X? base structure can be further specified linguistically for the instantiation of requests, so that its directive flavour is highlighted and its interpretation as a question for information is totally ruled out. Let us consider once more examples (27)-(31), reproduced here for convenience: (27) Can you give me a sample of a recommendation letter? – Beneficiary(speaker) (28) Can you please take a look at the one linked below? – Politeness (29) Can you describe a bit more the problem with combat controls? – Mitigation (30) If possible, can you post the solution in the comments? – Optionality (31) Horace if it’s not too much trouble can you post the locations of the events here? – CostBenefit The words in italics activate additional specific attributes of the requesting ICM. In (27), the personal pronoun me instantiates the speaker-as-beneficiary feature. In (28), the adverb please instantiates the politeness attribute of requests. In (29), a bit functions as a mitigator of the cost of the proposed action (i.e. mitigation attribute). In (30), if possible presents the request as tentative thus increasing the freedom of the addressee to comply with it or to refuse to do it (i.e. optionality attribute). Finally, in (31) the conditional phrase if it is not too much trouble directly alludes to the cost of the requested action, thus instantiating this specific attribute. It should be noted that the different attributes could be linguistically instantiated by other expressions with similar semantics. Although the inventory of realisation procedures for each attribute of the ICM of requesting will be dealt with in detail in Chapter 4, Table 2.6 includes some illustrative examples: Table 2.6. Realisation procedures for the attributes of the ICM of requesting Beneficiary Me/us, for me/for us E.g. Can you give me a book? Can you find out the answer for me? Cost-Benefit If it is not too much trouble, if it doesn’t bother you, if it’s not too much to ask E.g. If it’s not too much to ask, can you find out the answer for me? Optionality If possible, if you will, if you are able, maybe, perhaps E.g. If possible, can you reach me the book? Mitigation A bit, a little, a sec, for a second, just E.g. Can you wait for a second? Politeness Please, kindly, use of conditional modals (could), if you please E.g. Could you please stay still for a sec? As shown in Section 2.5.5 in relation to examples (19) to (23), a single base structure can also be combined with a varying number of realisation procedures, thus metonymically activating 63 a smaller or larger number of illocutionary attributes and yielding instances of requests that differ as to their level of explicitness. The different combinations of the base structure (CAN YOU DO X?) with specific realisation procedures for the attributes the request ICM conform a family of constructions whose members display different degrees of conventionality. As shown in Table 2.7, different combinations have been found to have different degrees of productivity in the iWeb corpus, which suggests that some of them are more conventional than others: Table 2.7. Number of occurrences in iWeb for the realisation procedures of the requests in examples (27)-(31) Base structure + Realisation procedures for specific attributes Nº of occurrences (iWeb) Can you do X? + Beneficiary (me) 73.852 Can you do X? + Politeness (please) 32.690 Can you do X? + Mitigation (a bit) 4.577 Can you do X? + Optionality (if possible) 197 Can you do X? + Cost-Benefit (if it’s not too much trouble) 139 The above data indicates that illocutionary constructions, just like constructions at other levels of linguistic description (Goldberg & Suttle, 2010), capture generalisations within a given language. The members of the CAN YOU DO X? family of request constructions are related via an inheritance hierarchy, with more abstract, productive constructions (i.e. base structure) being directly related to their more conventional instantiations. The description of the family of constructions for each speech act category should include its specific set of base structures plus the inventory of realisation procedures that help instantiate further attributes of the corresponding illocutionary ICM. Table X captures an illustrative inventory of realisation procedures for the act of requesting. It is important to draw attention to the fact that the realisation procedures activate a very specific set of attributes (i.e. beneficiary, optionality, cost-benefit, mitigation and politeness) among those that conform the illocutionary ICM of requesting. Other attributes of the same requesting ICM, which do not function as realisation procedures, seem to have specialised as base structures. These include the following: addressee-as-agent, addressee’s capability, addressee’s willingness, speaker’s need, speaker’s willingness, and possession of the requested object. An illustrative inventory of base structures for requests is included in Table 2.8: 64 Table 2.8. Base structures for the act of requesting Attribute of the requesting ICM Base structures Addressee-as-agent Imperative sentence, Vocative + Imperative sentence, You will do X + question tag -construction Addressee’s capability Can you do X? -construction, You can do X + question tag-construction Addressee’s willingness Will you do X? -construction, Would you be willing to do X? -construction, If you will do X-construction Speaker’s willingness I want X/I want you to do X-constructions, I’d like X/I’d like you to do Xconstructions Speaker’s need I need X /I need you to do X-constructions, I could do with X-construction Possession of requested object Have you got X?-construction, You’ve got X + Question tag-construction The attributes that activate base structures correspond to essential aspects for the realisation of the requested action (i.e. capability, willingness, and possession). Thus, if the addressee cannot physically perform the requested action or is not in possession of the requested object, there is no point in addressing the request to him. Similarly, if the speaker has no desire or need for the requested object or action, there is no need for the performance of the request at all. On the contrary, those attributes that correspond to additional realisation procedures that combine with base structures are connected to dimensions of social behaviour and social expectations that could turn the request more or less successful. Thus, modulating the cost, optionality, mitigation or politeness of the request may influence the addressee’s decision as to whether or not to perform the request. As was the case with the semantics of speech acts (see previous section), their constructional realisation should also be considered from a cross-linguistic perspective. Different languages offer diverse realisation procedures and favour different base constructions. Again, as with the attributes and variables that conform the illocutionary ICMs, succeeding in teaching speech acts constructional families in an L2 requires a previous comparison with those constructions available in L1, as well as explicit instruction aimed at making students aware of the similarities, differences and mismatches between the constructions in each language. Chapter 4 will offer detailed corpus-based inventories of the most prevalent base constructions and realisation procedures that English speakers use in their everyday life interactions for the expression of the directives under investigation. Illocutionary constructions will also be shown to be motivated by the semantics of speech act as captured by the attributes conforming its corresponding illocutionary ICMs. The constructional approach to speech acts that is advocated in this book attempts to accommodate the most relevant theoretical findings about the meaning and form of illocutionary acts that have been shown to stem from the contemporary cognitive, pragmatic and functional works revised in this chapter. By way of summary, Table 2.9 lists those theoretical advancements that shall guide our revision of the 65 way in which speech acts are presently being taught to advanced Spanish students of English as an L2 (Chapter 3), as well as our pedagogical description of directive speech acts and the subsequent design of related teaching activities (Chapters 4 and 5, respectively): Table 2.9. Summary of contemporary theoretical advancements on speech acts Functional Paradigms -The notion of speech acts as social, as well as linguistic constructs -The idea that the language system is capable of offering enough realisation procedures for the full codification of speech acts Pragmatic-Conversational Paradigms -The realisation that speech acts show different degrees of conventionalisation and inference -The focus on social/interactional aspects of illocution (i.e. pragmatic scales and variables) -The relevance of conversational aspects of speech act performance (i.e. adjacency pairs, preference organisation) - The relevance of the use of real data and sufficient contextual information Cognitive-Constructional Paradigms -The realisation that speech act categories show prototype effects -The realisation that codification, convention, and inferential processes are not discrete categories -The realisation that the formalisation of the semantics of speech acts in terms of illocutionary ICMs is capable of capturing their conceptual fabric and to distinguish prototypical cases of each illocutionary category -The realisation that the formal side of illocutions is motivated by (multiple source)-in-target metonymical operations on illocutionary ICMs. -The description of the idiosyncratic nature of illocutionary constructions in terms of base constructions + linguistic realisation procedures, both motivated by the semantics of each illocutionary ICM. -Focus on the need of cross-cultural and cross-linguistic comparisons of the illocutionary ICMs and illocutionary constructional families in L1 and L2 for its successful teaching. 72 earlier stages: requests are included both in the starters and again in the advanced level textbooks, while other directives are not. Another piece of research carried out by Pérez-Hernández (2019) looked in detail into the types of directives taught at three different levels of proficiency (starters, intermediate and advanced) in a larger collection of textbooks, including eight course book series by mainstream editorial houses like Cambridge, Oxford and Pearson. Her findings revealed that the representation of directives was poor at all three levels of instruction. Starters textbooks only dealt with requests and suggestions, leaving aside basic and highly frequent directives like ordering, warning and advising. The intermediate level textbooks added some subtypes of orders (i.e. commands), and other directives like giving instructions, while still excluding some of the most common types of directive acts: advising, warning, begging, among others. The advanced series were incapable of fully compensating for these gaps in previous levels. Most advanced textbooks included requests, suggestions and advising, and only a few of them enlarged the inventory of directives to incorporate the acts of recommending or proposing future actions. Nguyen’s (2011) and Ren and Han’s (2016) papers provide similar data stemming from an analysis of lower-intermediate textbooks, which again focus on the acts of requesting, suggesting and advising to the exclusion of other directives. These studies are largely compatible with the findings of the present study, which also shows that requesting, suggesting and advising are the directive acts that receive the highest amount of attention in all advanced textbooks under scrutiny. From the above considerations, it can be concluded that the limited representation of directive speech acts in advanced textbooks series is not due to the fact that their teaching has already been undertaken in previous academic stages. Acts like requests, suggestions, and to a lesser extent, advising, are taught at all levels of proficiency (i.e. starters, intermediate and advanced), while others are simply and randomly left out in all three levels of instructions. An extreme case is that of the act of begging which has only been found in one advanced textbook (TB1). Warning and ordering are also largely underrepresented, with only half the number of advanced textbooks in our corpus offering some kind of instructional material about them. In addition, the amount of exercises made available for the learning of the six directive speech acts under inquiry was generally scarce with an average of just 1 or 2 exercises per speech act. In most cases, the space allowed for the teaching of a particular speech act within an exercise was shared with other linguistic issues. For example, those exercises that presented the use of modals like must and have to for the expression of the act of ordering, also included other uses of the modals, such as the deductive must (TB1). Some of the activities including 73 directives did not even have the teaching of the speech act as their main focus. The few instances of warnings that have been found in the textbooks, for example, appear in exercises which are mainly concerned with reported speech: the student is given a direct form of warning (e.g. Don’t touch the stove or you’ll get burnt) and is asked to use a performative verb (i.e. warn) to turn it into a reported speech expression (e.g. He warned me not to touch the stove…). Since the focus of the exercises is the teaching of reported speech, no information or instruction is provided about the directive speech act itself. Zooming in onto other quantitative dimensions of the data, it was observed that the amount of attention, understood in terms of the number of exercises/activities devoted to the teaching of a speech act, was randomly assigned and not justified from a theoretical or pedagogical point of view. One may wonder why some textbooks (TB4) offer up to five activities for the teaching of advising acts and only one for requests, while others (TB7) devote five exercises to latter and only one to the former. In any case, there is nothing intrinsically more complex in the conceptual nature of either of these speech acts that would make one of them more challenging than the other and require deeper practice for its learning. Although not the focus of this investigation, the planification of the teaching of directives speech acts throughout the different levels of proficiency also seems to be largely unsatisfactory. As already hinted at in previous works (Moradi, Karbalaei & Afrad, 2013), a comparison of the request and advising formulae used in the starters, intermediate and advanced instructional material for each of the course book series under inquiry reveals that the same linguistic expressions are being taught at all levels, following no specific pattern and with no significant increase in their difficulty or complexity. CAN/COULD YOU DO X? and I’D LIKE X requests are the linguistic formulae found in all textbooks. Whether both are used, or which one is used at either level of instruction varies greatly among editorial houses, which again suggests lack of planification based on pedagogical and/or theoretical criteria. 3.3 Qualitative Assessment of the Treatment of Directive Illocutionary Acts in Advanced EFL Textbooks The following subsections offer the results of a qualitative assessment of the EFL textbooks under scrutiny in relation to the analytical categories described in section 3.1, which are summarised in Table 3.2 for the readers’ convenience. 74 Table 3.2. Analytical categories for the assessment of current EFL textbooks as regards their treatment of directive speech acts Semantic/Pragmatic aspects AC1. Transactional attributes (e.g. agent, beneficiary, cost) AC2. Interactional attributes (e.g. optionality, mitigation, politeness, etc.) AC3. Social variables (e.g. social power, social distance, formality) Constructional aspects AC4. Base constructions AC5. Realisation procedures AC6. Cognitive operations (metonymy, metaphor) Conversational aspects AC7. Adjacency pairs AC8. Preference organisation AC9. Use of real data and contextual information Typological aspects AC10. Cross-cultural mismatches in the semantics of illocutionary categories AC11. Cross-linguistic mismatches in the form of speech acts 3.3.1 Assessment of the Inclusion of Semantic/Pragmatic Information about Speech Acts in EFL Textbooks In Chapter 2, the description of the semantic and pragmatic knowledge characterizing each directive category in terms of illocutionary ICMs was proved to be a keystone of a sound theory of speech acts. In fact, the information comprised within each illocutionary ICM provides the grounding and motivation for the choice of speech act constructions and linguistic realisation procedures to be used in conversational interactions. Current functional and pragmatic research has evinced the roles and relevance of several types of attributes and variables in building the conceptual fabric of illocutionary ICMs (see Sections 2.2.2., 2.4.2 and 2.5.5. for a detailed description). By way of summary, transactional attributes inform about the agent, beneficiary and cost/benefit of the transaction involved in the directive speech act; interactional attributes capture social and pragmatic conventions that regulate social exchanges and verbal interactions (e.g. politeness, optionality, mitigation), and social variables, like social power, social distance and formality, have been shown to interact with the two previous sets of attributes and to be pivotal in determining their values (e.g. a large social distance increases the politeness, optionality and mitigation requirements of the speech act). Altogether they conform the ontology and structure of illocutionary ICMs (Section 2.5.5). The extent to which these semantic and pragmatic items have been implemented in the textbooks under inquiry is disturbingly low, especially given the importance of mastering this type of information for the performance of socially and contextually felicitous speech acts, as has been extensively argued in the literature. None of the textbooks under analysis offers information about the transactional attributes involved in the semantics of directive speech acts or their interconnections and synergies with interactional and social variables (e.g. the fact that as the cost of the requested action increases, 75 the politeness of the act will need to be modulated accordingly). Interactional attributes and social variables are also largely overlooked. The only interactional attribute that is given any consideration is that of politeness. Only three of the textbooks in our corpus offer polite formulae for requests and suggestions, but they do not offer any instruction as to when such polite forms should be used (e.g. high cost of the requested action, large social distance, power asymmetries, etc.). It has also been observed that most of the illocutionary constructions included in the textbooks are highly polite (TB3, TB4, TB5, TB7, TB10). This is so especially in connection to the act of requesting (e.g. Could/would you do X? forms). This type of polite requests, however, may not fit many of the everyday life interactional contexts in which a more relaxed register is expected and in which the use of such polite formulae may come through as odd, ironic, or straightforwardly clumsy. No information is offered about which contexts and which social variables license the use of those polite illocutionary constructions. Other interactional attributes, such as those of mitigation and optionality, are not given any consideration. Regarding social variables only two textbooks (TB and TB7) deal with the formality of the context, and just one introduces the variable of social distance in relation to requests (TB7). The variable of social power, an essential ingredient in the semantics of some directives (e.g. orders) does not receive any attention. Most textbooks teach linguistic formulae for the production of directive speech acts without instructing students about the fact that the power relations between the speakers, their social distance or the formality of the context often have a bearing on their needs for politeness, mitigation and/or optionality. The only exception to the lack of semantic/pragmatic information about the conceptual nature of directive illocutions is textbook TB7, where a rich number of illocutionary constructions are offered with some explicit instruction about the contexts (formal/informal), social variables (distance), and some of the interactional attributes (politeness) that guide the speakers’ choices in different situations. Unfortunately, this wealth of information is offered only in relation to the act of requesting. The rest of the directives included in this textbook are deprived of analogous explanations. There doesn’t seem to be any sound reason why requests require explicit teaching about social and interactional aspects of their use, while other directives like advising, warning, begging, ordering or suggesting do not. As pointed out at the beginning of this chapter, the act of requesting has received a significantly higher amount of attention in current theoretical and pedagogical research than other directives. The poor coverage of the semantic/pragmatic nature of directives different from requests in the textbooks under analysis may simply be a reflection of this situation. This brings to the front the need for 76 detailed theoretical descriptions of speech acts like the one provided in Chapter 4 for their subsequent implementation in EFL textbooks and teaching materials. The reported findings on the lack of adequacy of the semantic/pragmatic descriptions of directive speech acts in the collection of textbooks under analysis are consistent with previous literature on this issue. Crandall and Basturkmen (2004), Vellenga (2004), Usó-Juan (2008), Nourdad and Khiabani (2015), Ren and Han (2016), and Pérez-Hernández (2019) inform about the lack of meta-pragmatic information in textbooks for the teaching of English as an L2. Diepenbroek and Derwing (2013: 1) highlight the fact that ‘the quality of the pragmatic information included in textbooks does not allow learners to develop their pragmatic competence in the target language’. More recently, Borer (2018) and Ton Nu (2018) not only confirms the paucity of explicit information on pragmatic aspect of illocutionary performance, but also reveals that the few semantic/pragmatic aspects that are dealt with in the textbooks are inadequate according to current theories of L2 pragmatics teaching. Farashaiyan, Tan and Shahragard (2018) also report a lack of cross-cultural perspective in the teaching of the pragmatic dimension of illocutionary acts. All in all, the semantic and pragmatic knowledge of directive speech acts has been shown to be under-represented in EFL textbooks: students are generally offered a limited number of linguistic forms for each directive, and related meta-pragmatic information is insufficient both in terms of quantity and quality. The amount of practice material is also scarce (Shimizu, Fukasawa & Yonekura, 2007). 3.3.2 Assessment of the Treatment of the Constructional Nature of Directive Speech Acts in EFL Textbooks Learning to form grammatically correct sentences is just the first step towards being able to communicate in a foreign language. As was made manifest in Chapter 2, however, learning to pair different linguistic formulae with specific communicative purposes, in particular contexts, and bearing in mind the interactional and social variables that are at work in each particular communicative exchange is what is actually needed in order to master the use of a second language. Examining how the formal, linguistic realisations of directive speech acts are presented in the textbooks and whether they are taught in relation to their specific meanings is essential in order to assess the adequacy and effectiveness of current teaching materials. Previous research on this issue offers a bleak panorama that is confirmed by the data in our corpus. As already reported in the literature, the main focus is on the teaching of lists of expressions for the production of each directive speech act, which are randomly included in the 77 grammar sections, or as part of dialogues and/or exercises (Farashaiyan & Muthusamy, 2017). The number and representativity of the linguistic expressions included in those lists does not follow any systematic pattern. We have found textbooks (e.g. TB7) offering up to 20 different formulae for the expression of requests, four for the act of advising, and two for ordering. Alternatively, TB4 offers 12 expressions for the act of advising, and only one for requests. Yet other textbooks (e.g. TB1) provide just one linguistic formula for each directive category. The theoretical description of directive speech acts that will be presented in Chapter 4 reveals that the English language does not discriminate among different directive acts as to the wealth of linguistic formulae that it offers for their realisations. The unbalanced portrayal found in textbooks will inevitably have a negative impact in the learning process of L2 students of English. Beyond the above quantitative issues, Bouton (1996) points to another weakness of the treatment of formal and constructional aspects of speech acts in EFL textbooks: the lack of representativity of the formulae chosen as teaching targets. In his study on the act of inviting in EFL textbooks, the invitation formulae presented to the students were found to rarely occur in published native speaker’s corpora. In this same vein, Nguyen (2011) reported that sometimes textbooks stress one formula over others which are equally or more frequent in real life use, thus providing teachers and students with misleading information. This point has also been corroborated by the data in our corpus. By way of illustration, in relation to the act of ordering, all textbooks that include this directive in their syllabus offer the same two linguistic formulae for its expression, namely, the modal verbs must and have to (e.g. You must study, You have to study). The corpus search for the act of ordering (see Section 4.1.2 in the next chapter), however, reveals that these are neither the only, nor even the most frequent linguistic formulae for the communicative rendering of this directive. The IMPERATIVE construction and declarative-based constructions like YOU ARE TO DO X, for instance, which are also amply used by speakers, are not considered in the textbooks. None of the textbooks under scrutiny devotes specific sections to speech acts. The latter are introduced as part of the teaching of other grammatical issues. For instance, orders are dealt with when teaching modal verbs. Requests are introduced when teaching the uses of modal verbs in the past tense (e.g. could/would). Warnings have exclusively been found in connection to reported speech, when asking students to use the performative verb warn to report the corresponding indirect act. The fact that speech acts are always approached in relation to other linguistic issues, instead of being treated as independent teaching topics, may explain their shallow and partial representation in the textbooks under scrutiny. 78 The analysis of the textbooks chosen for this study also reveals that the linguistic formulae included in the teaching materials do not distinguish between illocutionary base constructions and realisation procedures. This distinction was shown to be relevant in Section 2.5.6, since realisation procedures act on base constructions to modulate their illocutionary force, so that they can fit different interactional/social contexts. Thus, the same base construction (e.g. CAN YOU DO X?) may be combined with different realisation procedures to modulate its levels of politeness (e.g. Can you do X, please?, Could you do X?) and even to change its illocutionary force from one of request to that of an order or a begging (Can you do X, please? vs. Can you do X at once!! vs. Can you please, please, please do X?). Finally, the description of directive illocutionary constructions in the textbooks under scrutiny also fails to inform students about the cognitive operations that motivate them (see Sections 2.5.5 and 2.5.6). By way of illustration, students are unaware of the fact that questions about ability, willingness or possession can metonymically activate a directive category simply because those are three basic components of the semantics of directive acts. Teaching the motivation of illocutionary expressions on the semantics of the illocutionary acts is likely to facilitate the memorisation of those formulae by L2 students, as has already been shown to be the case in other areas of L2 learning (e.g. prepositions, metaphors, idioms, etc.; Boers & Demecheleer, 1998; Shaffer, 2004; Boers & Lindstromberg, 2008; Tyler, 2012). 3.3.3 Assessment of the Treatment of Conversational Aspects of Directive Speech Acts in EFL Textbooks As explained in detail in Section 2.4.3, conversational approaches to illocution have highlighted the relevance of considering real language data in their analysis of speech acts. This theoretical postulate has also been assessed in relation to the portrayal of illocutionary acts in EFL textbooks, under the premise that the lack of real language examples can hinder the correct learning of a second language. As pointed out by Nguyen (2011), the literature provides ample evidence that English textbooks do not offer classroom learners adequate opportunity for learning authentic language (Bardovi-Harlig, 2001; Grant & Starks, 2001; Wong, 2002; Vellenga, 2004). In much the same vein, Jiang (2006: 36) points out that ‘although the new generation textbooks introduce more linguistic structures […] than the old generation textbooks, the discrepancies between real language use and EFL textbooks are still apparent’. This has also been made manifest in our assessment of the linguistic constructions and realisation procedures for directive speech acts included in our collection of textbooks (Section 3.3.2): the limited list of linguistic formulae that are currently being taught have been 79 found to be highly formulaic, and hardly representative of the wealth and variety of illocutionary constructions that the English language has to offer. In fact, out of the ten textbooks chosen for analysis, only five make it clear in their methodological descriptions that they make use of real language data in the texts and conversations included in the books. Authors like Bardovi-Harlig, Mossman and Vellenga (2014a, 2014b), and Bardovi-Harlig and Mossman (2016), however, have provided ample evidence about the feasibility of using corpus-based authentic language for developing teaching materials. These authors have shown how to use language samples to help students identify pragmatic routines for speech acts and notice how they are used in context. The lack of a widespread corpus-based approach to the design of instructional material for the teaching of speech acts finds little justification in the time of digital humanities. The use of real language examples as teaching material becomes pivotal when assessing the degree to which textbooks succeed in helping students to master the conversational/discursive dimension of speech acts. In this connection, the use of naturally occurring conversations is essential for students to understand how interactions are locally managed by speakers by means of adjacency pairs, turn-taking strategies, pre-sequences, and preferred/dispreferred responses. Pérez-Hernández (2019) reported on the assessment of eight textbooks of English as an L2 in relation to the attention devoted to three conversational aspects (i.e. adjacency pairs, presequences, and preference organisation). Her analysis reveals that their coverage of these conversational aspects varied with the proficiency level of the textbooks. In fact, the space devoted to these issues was found to be inversely proportional to the increase in proficiency, with advanced EFL textbooks devoting the lowest degree of attention to these matters. As Pérez-Hernández (2019: 269) points out, ‘the focus of intermediate and advanced textbooks also shifts towards writing and reading comprehension tasks, thus focusing on assertive speech acts like those of describing, speculating, comparing, argumenting, etc., which are less prone to a conversational handling’. These findings are compatible with the results stemming from the analysis of the teaching material assessed in the present study. As already pointed out in the two previous sections, the textbooks under scrutiny often introduce directive speech acts in connection to other linguistic aspects, such as modal verbs, hedging, or polite/tactful expression. This approach generally limits the scope of the explanations to the phrase or sentence level. Suprasentential aspects of the performance of speech acts, such as adjacency pairs and preference organisation, are consequently left out. 80 In those few exceptions in which students are led to consider directive speech acts in conversations, the scarcity or lack of explicit instruction about their conversational and/or discursive functioning is recurrent. In TB2, students are asked to role play a conversation in which one of them is meant to advise the other. In order to do so, students are given just one example of a linguistic construction for the act of advising (e.g. I’d do X, if I were you forms). No information is added about the preceding or following conversational turns, and no explicit instruction is provided about the contexts in which offering a piece of advice is a preferred or dispreferred conversational move. In some extreme cases, students are asked to role play a particular directive in the absence of any information about its semantic, pragmatic, linguistic or conversational workings. Thus, in one exercise in TB4, students are asked to formulate suggestions with no prior instruction about how to produce this speech act correctly and felicitously in the second language. Shockingly enough, this much need instruction is partially and implicitly offered in the following page of the textbook as part of an exercise about rephrasing and grammatical inversion (e.g. You should not do X > Under no circumstances should you do X). As will be made apparent in Chapter 4, offering students corpus-based real language examples of the conversational workings of directive speech acts is possible. More importantly, in the case of some directives, looking beyond the sentence towards higher levels of language description is in fact necessary in order to achieve a realistic description of their formal layout. As observed in the corpus study reported in Chapter 4, the act of offering advice, for instance, is hardly ever performed within the scope of a single sentence in the data extracted from our corpora of real language. Pieces of advice are generally scattered in several subordinate and/or juxtaposed sentences, and often realised through one or more conversational turns. 3.3.4 Assessment of the Treatment of Cross-Cultural and Cross-Linguistic Areas of Discrepancy between L1 and L2 in EFL Textbooks Interest on cross-cultural and cross-linguistic aspects of the teaching of speech acts started as far back as the 1990s. In Bouton’s (1996) classical volume on pragmatics and language learning, Cenoz and Valencia (1996) provide a case for taking into account cultural conventions on politeness and on the use of linguistic mitigation when teaching EFL students how to perform speech acts in a second language. According to their findings, Spanish students of English as a second language make use of less mitigating devices and more direct formulae for the expression of requests both in their native language and in their conversations in English. The higher need for politeness and indirectness that characterises the use of requests in British 81 English had not been properly acquired by the students taking part in this study, thus pointing to a failure in the teaching of pragmatic and cultural competence. As this type of sociopragmatic, cultural competence is concerned, the comparison of native and non-native speakers of English and Spanish reveals that both groups are aware of the different situations and use different degrees of directness according to the context. These observations in the field of EFL teaching come to confirm the findings of an extensive body of theoretical research on crosscultural variations in the performance of speech acts known as Interlanguage Pragmatics (Blum-Kulka, 1982, 1991; Blum-Kulka & House, 1989; Blum-Kulka, House & Kasper, 1989; Kasper, 1989; and Rintell & Mitchell, 1989; among others). Teachers and researchers on language learning and teaching have reached ample agreement on the relevance of including intercultural dimensions of language in the EFL instructional materials (Edelhoff, 1993; Byram & Risager, 1999; Sercu, 2002; Castro, Sercu and Garcia, 2004; Atay et al., 2009). In this connection, Castro (1999: 92) points out that teachers are now expected not only to teach the foreign linguistic code but also to ‘contextualise that code against the socio-cultural background associated with the foreign language and to promote the acquisitions of intercultural communicative competence’. Nevertheless, as Neddar (2012) states, it is arguable to what extent the findings in Interlanguage Pragmatics have actually been given consideration in EFL textbooks, thus contributing to raising learners’ awareness of crosscultural aspects of speech acts. Already in the 1990s, authors like Kachru (1994) pointed to the need of increasing the classroom practice on cross-cultural aspects of speech acts. The situation does not seem to have changed much since that decade. More recently, in her Masters’ Thesis on the pragmatics of EFL course books for Finnish students, Luomala (2010) points to the lack of explicit instructional and teaching materials devoted to making students aware of cross-cultural asymmetries in the conceptual and formal layout between speech acts in English and those in their native language. Loumala (2010: 7) remarks that ‘although some of the necessary pragmatic knowledge can be transferred directly from the learner’s first language (L1), more often than not the pragmatic rules and conventions in L1 and a foreign language differ significantly, and therefore need to be consciously learned and practiced’. Neddar’s (2010) study on cross-cultural pragmatic information in Algerian EFL course book series corroborates these findings. In much the same vein, our own assessment of advanced EFL textbooks shows that cross-cultural and cross-linguistic information is not included neither through explicit, nor through implicit instruction. Given the general consensus on the relevance of mastering this type of knowledge for the successful performance of speech acts in a foreign language, Neddar 88 Whether the speaker and the hearer want the ordered action to take place or not correlates with who benefits from the action. As shown above, in most cases the speaker is the beneficiary, either on his own or in conjunction with others. In these cases, the speaker clearly wants the action to be carried out, as is sometimes even manifested in the expression of the order itself: (6) ‘Your father has finally found a bride for your brother, Itachi is going to be married soon and when his betrothed comes I want you to be on your best behaviour, you're going to have to share your brother and I don't want you clinging to him when his bride arrives.’ The matriarch ordered. [iWeb] When the beneficiary of the order is the hearer or a third party, the speaker’s wanting the action to take place becomes less relevant. Still, when ordering someone to do something, the speaker expects the hearer to comply, even if the proposed action is not in his own benefit. As regards the hearer, his wanting to carry out the ordered action is prototypically low, since it generally involves a cost to him. Only in those situations in which the order is in his benefit (see example [5] above), will his desire to perform the action increase. What is important to highlight, however, is that whether or not the hearer wants to do the action is irrelevant for the speaker. When ordering, the latter does not take into consideration the hearer’s disposition and only expects compliance. This explains that there is not one single instance of order construction in our corpus based on the hearer’s willingness to perform the action. Expressions like WILL YOU DO X? (e.g. Will you vacuum the house for me?) have not been found in the corpus in relation to the act of ordering. It could be argued that, when used by a powerful speaker, these expressions could be understood as orders. Our corpus study shows that, even if the notion of order is a prototypical category and WILL YOU DO X? instances of orders could be accommodated as peripheral (polite) instances of the category, the fact is that native speakers do not prototypically use these expressions when they want to order someone to do something. The attested lack of occurrences of order constructions exploiting the hearer’s willingness or lack of it to carry out the action, correlates with the results of the analysis of the conceptual layout of orders, according to which the hearer’s willingness to perform the action is irrelevant to the speaker, who just expects compliance with his order. Is the action expressed in the order necessary? In most cases, the speaker needs the action to be carried out because he is to benefit from it: 89 (7) ‘Shut up. We need to keep going!’ Hou Qing ordered as he took the lead once more [iWeb] Still, even in those cases in which he does not have a strict necessity that the action is carried out, he still expects compliance with the order. How much freedom does an order allow the hearer? Whether or not the hearer follows an order is beyond the speaker’s control, but when a speaker gives an order, he presents it as non-negotiable. Orders communicate the idea that the hearer has no freedom to choose a path of action different from the one expressed in the order: (8) ‘…I want you to be on your best behaviour, you're going to have to share your brother and I don't want you clinging to him when his bride arrives.’ The matriarch ordered. She was sweet and caring like every other mother, but she was also an Uchiha and she didn't ask for things, she's demanded for them. Sasuke wasn't being given an option; his mother's words were law. [iWeb] As shall be made apparent in Section 4.1.2, this attribute of the conceptual layout of orders largely influences the formal traits of order constructions. Does the speaker attempt to minimise the cost of the action? When giving an order, the speaker does not make any explicit attempt to bring down the cost of the proposed action. In fact, it is usually the case that such cost is increased by adding a requirement of immediateness: (9) ‘Get your hands up and against that wall,’ the woman ordered, stepping away from her slightly. ‘Now!’ [iWeb] Does the speaker attempt to be polite? Politeness can be found in any type of speech act. Even threats can be polite (i.e. Honey, please, if you don’t finish up your soup, I will have to punish you without playtime). However, as the data in our corpus shows, orders are not prototypically polite. Over 99 percent of the orders in our data show no overt attempt at being polite. The opposite is usually the case. Most order constructions include strategies that highlight their impositive nature and, therefore, render 90 them fairly impolite (e.g. expressions of immediateness, forceful intonation, vocatives stressing the hearer’s inferiority, etc.): (10) ‘Girls, into your room, now.’ Her father ordered sternly as he rose to the door. [iWeb] (11) ‘Finish it, NOW!!!’, Alice ordered grimly. [iWeb] (12) ‘Sit down, you moron!’ Herrera ordered. [iWeb] In fact, in connection with orders the use of politeness strategies is anecdotal and when they are used, they often yield peripheral examples of ironic, sarcastic or even humiliating orders, such as the following example in which politeness is used in an order addressed to a hearer who is clearly being denigrated by the speaker: (13) ‘Shelby walked over to Sara with a dog leash in her hand. She clipped the leash to Sara's collar […] she went to her knees and crawled behind Shelby. The house was so big that her knees were very sore by the time they got to the main dining room […] ‘Under the table please,’ Shelby ordered. [iWeb] Is social power relevant to the act of ordering? The power relationship that holds between the speakers is central to the act of ordering. Orders are about power. An unstoppable force can move an object if it is strong enough to counteract the weight of the object. Likewise, a directive speech act is an order if the speaker has enough power over the addressee to impose his willing. Other directive acts also exercise power onto the addressee. Threats, for instance, are also impositives. In this case the power of the speaker comes from the fact that he can somehow harm or wrong the hearer if he doesn’t comply with his wishes. The power involved in the act of ordering does not imply this type of retaliation. All instances of orders in our corpus are uttered by speakers who either are socially more powerful than the hearers or adopt a position of superiority in relation to the latter. The notion of social power has been found to be quite heterogenous. The power of the speaker may be physical, granted by social institutions (a company, the military, the family hierarchical structure), or linked to social conventions (age, social class, knowledge). (14) ‘Itee! Make some tea for your grandfather. Its five o’clock.’, ordered her father. [iWeb] (15) ‘Go, find out where he is,’ the king ordered [iWeb] (16) ‘Sign to them to keep off,’ Colonel Forsyth ordered [iWeb] 91 (17) ‘So, now tell me in full detail everything you thought and felt while you sat within the chamber,’ the Doctor ordered [iWeb] The power of the speaker may even arise from the moral superiority stemming from the knowledge that the action expressed in the order is in the benefit of the hearer or a third party. Example (18) illustrates this: a brother orders his sibling (an equal) to stop seeing her girlfriend under the premise that it is not good for him to do so. What seems essential is that the speaker makes use of a real or perceived superiority to present the action as an imposition that leaves the hearer little freedom to refuse to comply: (18) ‘You must stop seeing that Russian girl,’ I ordered my brother […] Echoing the prejudiced, ignorant sentiment that I had grown up with, I believed it was wrong to become seriously involved with a person who does not follow the Hindu religion and is not a member of the Indian race. [iWeb] Is the social distance between the speakers relevant to the act of ordering? Our data shows that orders can be performed regardless of the social distance that exists between the speakers: as long as the speaker has the necessary power, she can direct her order to people who are socially close to her (relatives, friends, colleagues) or far away in the social scale (strangers). Learners of English as a second language need to be aware, however, that the closer the speakers are socially, the less impositive and forceful the order tends to be. As example (19) shows it is not socially acceptable to impose on our people who are socially close to us, even if we have the power to do so: (19) ‘Ok sweetie let me go to sleep on this cloud you've giving me...’ she softly ordered. Turning her head to the side she continued, ‘... but first give me a kiss Goodnight.’ [iWeb] Is the formality of the context relevant to the act of ordering? Orders are not tied to a specific type of context. Around fifty percent of the orders in our corpus take place in formal contexts and the other half in informal situations. 92 Table 4.1. The Know-What of Orders Orders are directive speech acts in which a speaker imposes a course of action on the hearer(s). When someone orders someone else to do something, they communicate the following: ‘I have the necessary power/authority to impose on you a particular course of action. I want you to do it because it is beneficial to me or to someone that I want to benefit from the action. I know it is costly to you, but I still expect you to comply with my order.’ What you need to bear in mind in order to perform a successful order: -The proposed action may benefit the speaker, the hearer, a third party or a combination of the former. -Orders stem from a need or a desire on the part of the speaker, who takes for granted that the hearer is capable of performing the action. -Whether or not the hearer wants to do as told is irrelevant for the speaker, who assumes that she has the power (either real or perceived) to impose her wishes on the former. -Prototypical orders are not polite, they do not attempt to minimise the cost of the proposed action, and they offer little or no freedom to the hearer to refuse to do as told. -Orders can be performed in all contexts regardless of their formality, and they can be directed to hearers regardless of the social distance that separates them from the speaker. BE AWARE!!! -Orders are prototypically not polite and unmitigated. The overt use of politeness or mitigation gives way to peripheral instances of orders that may be felt as ironical or may be mistaken for requests. -You can perform an order regardless the social distance that separates you from the hearer, but you should bear in mind that the smaller the social distance between the participants, the less impositive and forceful your order will be perceived. It is not socially acceptable to impose on people to which we feel emotionally attached. 4.1.2 The Know-How of Orders How are orders expressed in English? The analysis of the first 500 instances of orders returned by the iWeb corpus offers the following portrait of how English speakers realise this speech act linguistically (see Table 4.2). Table 4.2. Base constructions for the act of ordering in English ORDER BASE CONSTRUCTIONS OCCURRENCES PERCENT IMPERATIVE 436 87.2% YOU DO X 6 1.2% X SHALL BE/DO 6 1.2% YOU ARE (NOT) TO DO X 5 1% YOU MUST DO X 4 0.8% YOU HAVE TO DO X 4 0.8% X MUST DO Y 4 0.8% LET X DO Y 4 0.8% GO DO X 4 0.8% I WANT YOU TO DO X 4 0.8% I WANT YOU + PREPOSITION 3 0.6% SEE THAT X IS DONE 2 0.4% I’D LIKE X 2 0.4% 93 Looking at the numbers, it is no surprise that the imperative sentence type has traditionally been associated with the speech act of ordering. Over eighty seven percent of the instances of orders in our corpus make use of imperative sentences as their base constructions: (20) [BARE IMPERATIVE] ‘Release the lady,’ the soldier ordered. [iWeb] (21) [SEE THAT X IS DONE]: ‘These men will not be required to work for two weeks,’ he ordered sternly. ‘And see that good provisions are handed out to them,’ he added. [iWeb] (22) [GO DO X]: ‘Right. Go lock the door then so mother won’t be disturbing us,’ he ordered. [iWeb] The rest of the orders in our corpus make use of a varied array of base linguistic constructions which activate different elements of the meaning of orders, therefore metonymically activating their interpretation as such. Thus, there are some constructions that focus on the speaker’s desire that the action is carried out by the hearer: (23) [I WANT YOU TO DO X] When she ordered: ‘I want you out of here now...’ [iWeb] (24) [I’D LIKE X/I’D LIKE YOU TO DO X] ‘I’d like a rowboat ready to cast off, mister Quidd,’ the Admiral ordered. [iWeb] There are also some base constructions focusing on the lack of freedom of the hearer to refuse to comply with the order: (25) [YOU DO X] ‘You stay here and find one,’ he ordered. [iWeb] (26) [YOU HAVE TO DO X] ‘You have to taste it before complaining,’ she ordered him [iWeb] (27) [YOU ARE ORDERED TO DO X] ‘You are ordered to co-operate…’ [iWeb] X MUST BE DONE 2 0.4% X WILL BE DONE 2 0.4% YOU HAD BETTER DO X 2 0.4% YOU ARE ORDERED TO DO X 2 0.4% X IS PROHIBITED TO DO X 2 0.4% X IS DIRECTED TO DO Y 2 0.4% X WILL (NOT) BE REQUIRED TO DO Y 2 0.4% THERE BETTER BE X 2 0.4% TOTAL 500 100% 94 Others exploit the impositive flavour of orders and present the realisation of the action as unavoidable: (28) [X SHALL DO Y] ‘Defendant School District shall pay a fine per infraction to the Plaintiff…’ the judge ordered [iWeb] (29) [X WILL BE DONE] ‘…whatever we find in the shape of an upper class in Poland will be liquidated,’ Hitler had ordered. [iWeb] With the exception of those constructions which have a marked impositive flavour (e.g. YOU ARE TO DO X, YOU HAVE TO DO X, YOU DO X, X SHALL DO Y, X WILL BE DONE, etc.), the rest of the base constructions for ordering found in our corpus are also used in the expression of other speech acts (see sections in relation to requests, beggings, suggestions, pieces of advice and warnings). Imperatives, for instance, simply instruct the hearer to do something, and this is a goal that is shared by all directive speech acts. However, as shown in Section 2.5.6, base constructions rarely appear on their own. They combine with other linguistic strategies that further activate other aspects of the meaning of illocutions, thus clarifying the nature of the intended speech act. How is the imperative used when the aim of the speaker is to perform an order? Table 4.3 summarises the linguistic strategies that accompany the IMPERATIVE base construction in the expression of orders. Table 4.3. Imperative base construction + Realisation procedures for orders A central characteristic of the linguistic strategies that combine with the base constructions used in the expression of orders is that they generally aim at increasing the imposition and the force of the act. Thus, the three main realisation procedures that combine with IMPERATIVE base structures in the production of orders are (1) a forceful/impositive intonation, (2) IMPERATIVE CONSTRUCTIONS OCCURRENCES PERCENT BARE IMPERATIVE 254 58.2% IMPERATIVE + IMPOSITIVE INTONATION 116 26.7% IMPERATIVE + EXPRESSION OF IMMEDIATENESS 29 6.7% IMPERATIVE + EXPRESSION DOWNGRADING THE HEARER 17 3.9% IMPERATIVE + EXPRESSION OF SOCIAL CLOSENESS 17 3.9% IMPERATIVE + EXPRESSION OF POLITENESS 3 0.6% TOTAL 436 100% 95 expressions of immediateness, and (3) vocatives downgrading the hearer, as illustrated by examples (30)-(32) respectively. (30) ‘Come on!’ he ordered roughly. [iWeb] (31) ‘Get your hands up and against that wall,’ the woman ordered, stepping away from her slightly. ‘Now!’ [iWeb] (32) ‘Open your mouth whore!’ he ordered [iWeb] Still, around half of the IMPERATIVE base constructions used in the expression of orders in our corpus are bare imperatives. As will be made apparent in Section 4.2.2, this is also the case with requests. The use of bare imperatives, deprived of any other lexico-grammatical forms that help to specify the type of directive involved, can be an economic strategy for performing directives, as long as the semantic attributes of the speech act that is to be communicated are contextually activated. In the case of orders, EFL students should be aware that the use of a bare imperative would be successfully interpreted as an order in contexts in which the superiority of the speaker is known to all participants. The superiority of the speaker, in the case of orders, leads him to not even entertain the possibility of a refusal to comply by the hearer. Hence, when this is clear from the context, he does not feel the need to make use of any of the strategies included in Table 4.3 to force the hearer to obey his order. If speakers are socially equal, however, the interpretation of the IMPERATIVE base structure as an order will be ruled out in favour of alternative directive readings like requests, beggings, suggestions, etc. The examples in which politeness strategies (i.e. use of the adverb please) are used in connection to the imperative for the expression of an order are either sarcastic polite orders (see example 13 above), camouflaged orders, or cases in which the social distance between the speakers is small (i.e. they are friends, relatives, long-time colleagues). With respect to the latter, students need to be aware of the fact that the impositive force of an order is sensitive to the social distance that exists between speakers. As explained in Section 4.1.1, when the speakers are socially close (friends, relatives), the imperative is perceived as less forceful because it is taken for granted that it is not socially acceptable to impose on people who are close to us. In these cases, if the speaker wants to maintain the impositive flavour of an order, he should make use of one of the strengthening strategies described above (i.e. forceful intonation, expression of immediateness, downgrading vocative). If, on the contrary, he wishes to soften the strength of the order to adapt it to the requirements of a short social distance, he 96 can either make use of politeness strategies (example 33), or simply use expressions that signal the social closeness with the hearer (example 19, reproduced below as 34 for convenience): (33) ‘I wrote it up myself,’ Harper’s secretary replied. […] ‘Dig it out for me and bring it in here, please,’ Harper ordered. [iWeb] (34) ‘Ok sweetie let me go to sleep on this cloud you've giving me...’ she softly ordered. Turning her head to the side she continued, ‘... but first give me a kiss Goodnight.’ [iWeb] The data reported in this section suggests that teaching Spanish EFL students to perform orders in English should necessarily include the most frequent constructions in Table 4.2. In addition, it would be useful to explain in detail the use of the IMPERATIVE base construction in the production of orders, since this has turned out to be the most frequent linguistic strategy for their expression in English. In this respect, two issues should be highlighted: (1) If the features included in the know-what of an order are evident from the context, a bare imperative is one of the preferred, most prototypical linguistic strategies for ordering someone to do something in English; and (2) in a less informative context, the order interpretation of an IMPERATIVE construction can be secured by making use of additional strategies aimed at increasing its impositive character (i.e. expressions of immediateness, forceful intonation, downgraders, etc.) or by avoiding those that soften its force (i.e. politeness strategies, expressions of social closeness). Finally, the strategies offered by the English language for the expression of orders have been compared with those that are most frequent in the native language of the prospective Spanish EFL students. Such comparison allows us to offer learners some red flags about possible mismatches between both languages. Table 4.4. Comparison of base constructions for the act of ordering in English and Spanish ORDERS BASE CONSTRUCTIONS (ENGLISH) Nº % Nº % ORDERS BASE CONSTRUCTIONS (SPANISH) IMPERATIVE 436 87.2% 380 76% IMPERATIVE YOU (WILL) DO X 6 1.2% 29 5.8% (TÚ/VOSOTROS/USTEDES) HARÁS/HARÉIS X YOU ARE TO DO X 5 1% 13 2.6% VAS A HACER X YOU MUST DO X 4 0.8% 4 0.8% DEBES HACER X YOU HAVE TO DO X 4 0.8% 5 1% TIENES QUE HACER X 97 Table 4.4 shows the number of occurrences and percentages of the different order constructions both in English and Spanish. The data reveals that there are some parallel constructions for the expression of orders that have a similar frequency of occurrence in both languages. Among them, the most frequent are the IMPERATIVE, YOU (WILL) DO X, YOU MUST DO X, YOU HAVE TO DO X and YOU ARE TO DO X constructions (in green in Table 4.4). Many others are found in the corpora for both languages, but they are more frequently used in one of them (i.e. those in orange in Table 4.4). Thus, order constructions that refer to the speaker’s wanting the action to take place (e.g. I WANT YOU TO DO X, I WANT + PREP) are found in both languages, but slightly more frequent in English. Spanish, on the contrary, displays a preference towards the use of direct performatives (e.g. TE ORDENO QUE (YOU ARE ORDERED TO DO X)). Finally, there are constructions that only arise in the corpus for one of the languages. In some cases, this is so because there exists no direct counterpart in the other language. Spanish, for example, makes use of empathic order constructions like QUE HAGA(N) X (THAT THEY X WILL/SHALL BE/DO X 6 1.2% 3 0.6% (ÉL/ELLA/ELLOS/AS) HARÁ(N) X X MUST DO Y 4 0.8% 2 0.4% DEBE HACER X GO DO X 4 0.8% 1 0.2% VE Y HAZ X I WANT YOU TO DO X 4 0.8% 1 0.2% QUIERO QUE HAGAS X I WANT YOU + PREPOSITION 3 0.6% 1 0.2% TE QUIERO + PREPOSITION X MUST BE DONE 2 0.4% 9 1.8% (SE) DEBE HACER(SE) X YOU ARE ORDERED TO DO X 2 0.4% 13 2.6% TE ORDENO QUE… LET X DO Y 4 0.8% 0 0% Possible, but not productive SEE THAT X IS DONE 2 0.4% 0 0% Possible, but not productive I’D LIKE X 2 0.4% 0 0% Possible, but not productive X WILL BE DONE 2 0.4% 0 0% Possible, but not productive YOU HAD BETTER DO X 2 0.4% 0 0% Possible, but not productive X IS PROHIBITED TO DO X 2 0.4% 0 0% Possible, but not productive X IS DIRECTED TO DO Y 2 0.4% 0 0% Possible, but not productive X WILL BE REQUIRED TO DO Y 2 0.4% 0 0% Possible, but not productive THERE BETTER BE X 2 0.4% 0 0% Possible, but not productive Not possible 0 0% 30 6% QUE HAGA(N) X Not possible 0 0% 9 1.8% A + HACER X TOTAL 500 100% 500 100% 104 something along the line of You might want to collect those documents before the bank close. In spite of the freedom conveyed by the linguistic form, the hearer is well aware of the superiority of the speaker and of his obligation to comply. These are, in fact, orders camouflaged as requests or polite orders. Students should be aware of interactions of this type between the semantic attribute of optionality and the semantic variable of social power, to the effect that asymmetrical power relationships may override the effect of the optionality conveyed linguistically. Does the speaker attempt to minimise the cost of the action? Among the obstacles that may lead to an unsuccessful request there is also the cost of the action. Since the speaker is not powerful enough to impose his will on the hearer, she needs to carefully consider the cost of the requested action and attempt to minimise it as much as possible to secure the hearer’s compliance. The minimisation of cost serves as the grounding for some request constructions (i.e. If only you could do X/If only you did X). It can also be achieved by the use of a particular set of realisation procedures, mainly adverbs like a little, a bit, just, simply, a minute, only, etc. By means of illustration, see examples (18) and (19). (18) ‘When they see a plow truck, just give us the opportunity to do what we need to do to help them get where they're going,’ requested Maroon. [iWeb] (19) Simba requested: ‘Taraja, could you stay here a bit while Akili and I climb out.’ [iWeb] Does the speaker attempt to be polite? Politeness has traditionally been considered a defining trait of requests. However, as was the case with optionality, politeness per se may not lead to a request interpretation, especially in those cases in which the speaker is clearly more powerful than the hearer and it is obvious from the context that he has reasons to use such power to coerce the hearer’s course of action. In asymmetrical relationships, politeness, without optionality, leads to fake requests or polite orders such as the one in example (20): (20) [Team captain to players] ‘Would players three and eleven please step forward?’ Caius requested. (iWeb) Being aware of this overriding effect of the social power variable, students will be able to understand that the use of politeness is only a straightforward signal of requests when the 105 speaker is either socially equal or inferior than the hearer. When this situation holds, politeness is an effective way of overcoming some of the obstacles that may prevent the request from succeeding (e.g. the hearer’s unwillingness to do the action or the high cost of the action). This is the reason for the pervasiveness of politeness in the realisation of requests. However, as shown by the data in our corpus, the politeness of requests may decrease in several circumstances, such as when the speaker is more powerful than the hearer (21), and when the beneficiary of the requested action is the hearer himself (example 22), or both speaker and hearer (example 23). Politeness is also taken for granted when the speaker and the hearer are socially close (example 24). (21) ‘STEWARDESS SNIPPEE!’ Captain boomed. ‘YOU ARE REQUESTED ON FLIGHT DECK!’ [iWeb] (22) Then Jill took off her pants and handed them to Jack, as she said, ‘Here, you try on mine!’ As she requested, he tried them. [iWeb] (23) ‘Keep our studio clean,’ another sign requested. [iWeb] (24) ‘Go down to Thomas and get me some more,’ his father requested. [iWeb] Is social power relevant to the act of requesting? Requests can be uttered regardless of the power relationship that holds between the speakers. Nevertheless, as explained above, students should be aware of the fact that as the power of the speaker increases, (1) the need for politeness to secure compliance from the hearer decreases and requests gradually fade into more impositive speech acts like orders or commands, and (2) the optionality of the hearer to decide whether to act as requested decreases in spite of the politeness used in the expression of the request. Mutual knowledge of the speaker’s superiority overrules the effects of politeness and limits the freedom of the hearer to decide on his course of action, thus turning those instances of requests into more peripheral members of the category. Is the social distance between the speakers relevant to the act of requesting? Requests can be addressed to people who are both close and distant to the speaker on the axis of social relationships. Still, students should be aware of the interaction of this social variable with the semantic attribute of politeness. As shown above, the closer the speakers are, the lower the need for overt politeness in the expression of requests will be. In conversations among socially close people, politeness is taken for granted, since the unwritten rules of social 106 interaction already apply, and people are not expected to impose on others who are socially or emotionally close to them (see example 24). Is the formality of the context relevant to the act of requesting? Requests are performed in all contexts regardless of their degree of formality. The use of politeness tends to increase with the formality of the context. Formal contexts often also imply a larger social distance between speakers, which also favours the aforementioned increase in the use of politeness. Table 4.5. The Know-What of Requests Requests are directive speech acts in which a speaker asks the hearer(s) to carry out an action. When someone requests someone else to do something, they communicate the following: ‘I want you to do something because it is beneficial to me or to someone that I also want to benefit from the action. I know it is costly to you and I don’t have the necessary power to impose the action on you. I acknowledge your freedom to comply or to opt out.’ What you need to bear in mind in order to perform a successful request: -Prototypically, the requested action benefits the speaker, but more peripheral instances of requests may be intended to benefit the hearer, a third party or a combination of the former. -Requests stem from a need or a desire on the part of the speaker, who does not have the necessary social power to impose his will on the hearer and who is, therefore, aware of the fact that he may need to overcome some obstacles in order to get the hearer to comply with his wish. -Among the obstacles to overcome, there is the willingness and ability of the hearer to carry out the proposed action, the cost of the action, and the freedom of the hearer to carry out the action or not. -To overcome the aforementioned obstacles, prototypical requests are polite, they attempt to minimise the cost of the proposed action, and they offer freedom to the hearer to refuse to do as told. BE AWARE!!! -Prototypical requests are uttered by speakers who do not have the necessary power to impose their will on the hearers. If you are a socially powerful speaker, be aware that your social power may overrule the effect of politeness and the intrinsic optionality of a requestive act, thus giving way to cases of fake requests or camouflaged orders. -You can perform requests in all contexts regardless of their formality, and they can be directed to hearers regardless of the social distance that separates them from the speaker. But you should be aware that the higher the formality of the context, or the larger the social distance between the participants, the more polite the request will need to be to secure compliance. Remember that politeness can also be achieved indirectly by means of minimizing the cost of the action and/or increasing the optionality of the hearer to do as told. -You should be aware of the fact that the higher the cost of the action, the higher the need to be polite, to acknowledge the freedom of the hearer to refuse, and/or to minimise the cost of the action. 4.2.2 The Know-How of Requests Studies on the linguistic formulae used by speakers of different languages to request someone to do something have flourished since the beginning of speech act theory. It is possible to distinguish four main lines of research on this matter, including pragmatic, interlinguistic, cognitive and conversational approaches. Within the pragmatic strand, Leech (2014) has provided one of the most comprehensive works to day, devoting a full chapter of his book to a 107 qualitative corpus-based taxonomy of the different linguistic realisation procedures for requests, or in his own words, the pragmalinguistic strategies and lexico-grammatical forms available for the expression of requests in English. His main classification criterion is based on the degree of directness/conventionality of the different formulae. It does consider some motivating factors, such as cost-benefit, social distance, and power, but ignores others like the capability and willingness/need of the participants, the formality of the context, and the potential interactions between semantic variables and attributes, thus falling short of offering a comprehensive picture of the semantic motivation that underlies the diverse range of linguistic constructions at hand for the realisation of requests. The work by Blum-Kulka, House and Kasper (1989) has delved on interlinguistic issues affecting the production of requests, with a view to highlighting cultural idiosyncrasies that need to be taken into account when attempting to perform a request in a second language. Cognitive approaches (Pérez-Hernández, 2001, 2013, 2019; Del Campo, 2013; Vassilaki, 2017) offer a more detailed account of the experiential grounding (i.e. removal of restraint force schema), cognitive operations (i.e. illocutionary metonymy), and cultural conventions (ICMs) that motivate the strategies for requesting. Finally, conversational approaches (Blum-Kulka, House & Kasper, 1989; Faerch & Kasper 1989), have shown that the expression of requests often exceeds the limitations of a single utterance, and have also looked into the supportive moves, and preferred-dispreferred conversational turns in which requests are involved. In the framework adopted for this book (see Chapter 2), the portrayal of request constructions brings together these four essential aspects of the expression of requests. There are not many studies on frequency data for different request constructions. Not even Leech’s (2014) comprehensive taxonomy of requesting strategies provides information about their frequency of use. This type of information, however, is essential for teaching purposes. Due to space limitations, EFL course books cannot possibly cover every single request strategy, which makes it important to know which ones are the most frequently used by native speakers in order to include them in the teaching materials. This information is also useful when planning the teaching of speech acts throughout the different levels of instruction. Frequency data can help teachers and editors to choose the most common constructions for the beginners’ level and gradually increase the difficulty with less frequently used constructions at the intermediate and advanced levels. Based on the London-Lund corpus of spoken English, Aijmer (1996: 157) reports two constructions (i.e. YOU COULD DO X and COULD YOU DO X? as those showing a higher frequency of occurrence among British ‘educated speakers’ from the 1950s- 108 70s generation which populated the data of that particular corpus. Along this line, as shown in Chapter 3, textbooks of English as a second language offer a very limited number of canonical formulae for requesting such as CAN/COULD YOU DO X?, IMPERATIVE + PLEASE, OR WOULD YOU MIND DOING X? The iWeb corpus, containing fourteen billion words, offers a slightly different picture and paves the way for a richer approach to the programming of the teaching of requests in EFL textbooks. As can be seen in Table 4.6, the data in this corpus yields a large number of linguistic constructions with varied rates of occurrence for the expression of requests in English. Table 4.6. Base constructions for the act of requesting in English REQUEST BASE CONSTRUCTIONS OCCURRENCES PERCENT IMPERATIVE 280 56% CAN YOU DO X? 45 9% COULD YOU DO X? 25 5% I WOULD LIKE X I WOULD LIKE YOU TO DO X 20 4% I WISH X I WISH YOU TO DO X 15 3% IF YOU WILL DO X 15 3% WILL YOU DO X? 10 2% WOULD YOU DO X? 10 2% DO YOU HAVE X? HAVE YOU GOT X? 10 2% CAN WE NOT DO X? 5 1% COULD WE NOT DO X? 5 1% MAY YOU DO X? 5 1% I WANT YOU TO DO X 5 1% I NEED YOU TO DO X 5 1% IT’S TIME FOR YOU TO DO X 5 1% IT WOULD BE GOOD IF YOU COULD DO X 5 1% YOU THINK YOU COULD DO X 5 1% WOULD YOU MIND DOING X? 5 1% ANY CHANCE YOU COULD DO X? 5 1% I DIRECT THAT YOU SHALL DO X 5 1% I ASK YOU TO DO X I ASK THAT YOU DO X 10 2% YOU ARE REQUESTED TO DO X 5 1% TOTAL 500 100% 109 The request strategies taught in EFL textbooks are often restricted to CAN/COULD YOU DO X?, WILL/WOULD YOU DO X?, WOULD YOU MIND DOING X? constructions. The use of the IMPERATIVE construction in association with the politeness adverb please is also a typical request construction taught to EFL students. The data in our corpus, however, should lead to reconsider present-day approaches to the teaching of request strategies. Imperative constructions As illustrated in Table 4.7, and as it was the case with the act of order, IMPERATIVES are by far the most common construction for the expression of requests in English. The canonical CAN/COULD YOU DO X?, WILL/WOULD YOU DO X?, or WOULD YOU MIND DOING X? constructions, among others, occupy secondary distant positions. Table 4.7. Imperative base construction + Realisation procedures for requests The high frequency of use of the IMPERATIVE construction, however, calls for a closer examination of its particular use in the realisation of requests. As can be observed in Table 4.7, and as was the case with orders, bare IMPERATIVES constitute nearly half the occurrences of IMPERATIVE-BASED request constructions in English. Using a bare imperative for requesting is an effective strategy, but also a complex and risky one, since imperatives can realise any type of directive (see the Know-How sections for other directives in this chapter). A successful use of bare imperatives as requests will depend on whether the semantic attributes of the ICM of requesting (see Section 4.2.1) are activated contextually. If it is clear that the action is in the speaker’s benefit, that participants are equals and that the speaker does not have enough power to impose his will on the hearer, the bare IMPERATIVE CONSTRUCTIONS FOR REQUESTS OCCURRENCES PERCENT BARE IMPERATIVE 115 23% IMPERATIVE + EXPRESSION OF POLITENESS (PLEASE) 65 13% IMPERATIVE + EXPRESSION OF OPTIONALITY 30 2% IMPERATIVE + EXPRESSION OF SOCIAL CLOSENESS 30 6% IMPERATIVE + EXPRESSION OF MINIMISATION OF COST 20 4% IMPERATIVE + EXPRESSION OF SUPERIORITY OF HEARER 10 2% IMPERATIVE + REASON CLAUSE 5 3% IMPERATIVE + EXPRESSION PRAISING THE HEARER 5 1% TOTAL 280 100% 110 imperative will be effective as a request. This is illustrated by example (25), where a lover asks his partner to tell her that he loves her: (25) ‘Sirius’, meekly she requested, ‘Tell me you love me’. [iWeb] Otherwise, other directive interpretations may be activated. The student should also be made aware of the fact that in some contexts, especially those that are highly formal (e.g. legal documents), the use of a bare imperative may be a dispreferred construction for requesting even if all semantic attributes of requests are clearly activated by the context. Compare the formal request in (26) with a fabricated alternative that makes use of a bare imperative in (27). (26) ‘I direct that my executors shall arrange for my remains to be taken to the country of Bali and to be cremated there in accordance with the Buddhist rituals of Bali’, Bowie allegedly requested in his last will testament. [iWeb] (27) ‘Take my remains to the country of Bali and cremate them there in accordance with the Buddhist rituals of Bali’, Bowie requested in his last will testament. This is so because formal legal contexts often require a higher use of explicitness. Mastering the use of bare imperatives for the performance of the different types of directives is a highly complex task. Fortunately, the language system offers a myriad of lexico-grammatical realisation procedures that can be used to secure a requestive use of the imperative. These are different from the lexico-grammatical realisation procedures that lead to an order interpretation of the IMPERATIVE construction (see Section 4.1.2), and they are semantically grounded in the semantics of requesting (see ICM of requesting in Section 4.2.1). Thus, the lack of social power of a speaker issuing a request prevents her from being able to impose her will on the hearer as is the case with other directives like orders and threats. For this reason, when requesting something, speakers attempt to achieve the hearer’s compliance through other means. The most frequent lexico-grammatical realisation procedures used in connection with an IMPERATIVE construction to secure a request interpretation are those that express (1) politeness, (2) mitigation of cost, (3) optionality and (4) social closeness. The most widespread one is the use of politeness through the use of the adverb please, or alternative expressions (e.g. if you please). (28) ‘Please, make way,’ I requested. [iWeb] 111 Also, highly frequent is the use the IMPERATIVE construction together with expressions of optionality, aimed at enhancing the freedom of the hearer to refuse to carry out the requested action (e.g. if you can, if you don’t mind, question tags like can/could/will/would you?): (29) ‘While you’re out, if you can, take some pictures of the birds and send them to me.’ She requested. [iWeb] IMPERATIVE constructions with vocatives signalling social closeness and affection also lead towards a request interpretation, since imposing on people who are close to us is generally not socially acceptable: (30) ‘July, wait,’ he requested politely. [iWeb] Lexico-grammatical forms that mitigate the cost of the proposed action also help to guarantee a request interpretation of IMPERATIVE constructions. If the speaker had the power to impose his will (as happens with orders), he would not need to consider the cost of what he asks for or attempt to mitigate it. In fact, as shown in Section 4.1.2, IMPERATIVE constructions for orders were not mitigated in the data from our corpus. Requests realised through IMPERATIVE constructions, on the contrary, often shown mitigation of the cost of the required action through the use of adverbs like just and only, diminutives, expressions of diminishing like a little, a bit, a little bit, etc., or a combination of one or more of these linguistic resources: (31) ‘So, Bella, tell us a little bit about yourself.’ He requested politely. [iWeb] (32) ‘Just a little off the bottom please,’ requested the unsuspecting lady. [iWeb] Signalling the superiority of the hearer by means of vocatives expressing the asymmetric power relationship between participants is also an indirect way of acknowledging that the hearer is free to comply (i.e. optionality), thus tilting the interpretation towards that of a request: (33) ‘Sir, come over to our vigil,’ she requested in vain. [iWeb] Another less frequent means of acknowledging lack of power to impose an action is by alluding to rational arguments in order to persuade the hearer to comply. Thus, some request 112 IMPERATIVE constructions are either preceded or followed by reason clauses stating the motives that lead the speaker to ask the hearer to do something: (34) ‘Ok, look, I’m not good with words so bear with me here,’ I requested. [iWeb] Finally, IMPERATIVE constructions with vocatives praising the hearers are also prototypically interpreted as requests because they come through as polite: (35) ‘Tell me more, you big handsome sailor,’ she requested. [iWeb] All in all, if compared to the IMPERATIVE constructions for orders (Section 4.1.2) and other directive types (see following sections), the realisation procedures accompanying IMPERATIVE constructions for requests are specific enough for hearers to identify them as such and for speakers to grant the intended interpretation. While over eighty percent of order constructions in our corpus are based on the use of the imperative, the act of requesting offers a more varied range of construction types for its expression. Nearly fifty percent of requests in our corpus are realised by means of declarative or interrogative-based constructions. The most frequent of them are those that question the ability of the hearer to perform the requested action (i.e. CAN/COULD YOU DO X?) (36) ‘Can you tell me where the bathroom is on this ship?’ she nervously requested. [iWeb] (37) ‘Could you teach me how to dance, Misty?’ He requested. [iWeb] This type of interrogative-based construction, grounded on an essential semantic attribute for directives (i.e. the fact that the hearer has to be able to perform the action), gives the latter freedom to refuse compliance without losing face. In this way, one of the key semantic attributes of requests (i.e. the fact that they allow optionality to the hearer) is activated. If the past modal is used, the politeness of the act increases, thus further activating another essential feature of requesting (i.e. its polite nature). 13 Through this activation of three nuclear semantic variables of requesting (i.e. ability of the hearer to perform the action, optionality and politeness) CAN/COULD YOU DO X? constructions constitute a very effective means of 13 For explanations on the politeness of past modals, see Pérez-Hernández (1996) and Leech (2014: 153-155), among others. 113 conveying a request, which explains their productivity in actual use as shown by the data (see Table X). In his taxonomy of request strategies, Leech (2014: 148-149) includes declarativebased ability constructions (i.e. YOU CAN/COULD DO X) as strategies for requesting. These could in fact be used as requests in a proper context, however, our corpus yields no occurrences for constructions of this type, except for two variants in which the politeness and optionality of the act is highlighted through other means. Thus, the constructions IT WOULD BE GOOD IF YOU COULD DO X and I THINK YOU COULD DO X activate the politeness and optionality variables by means of past modals (i.e. it would be good if…) and expressions of tentativeness (i.e. I think…) respectively. Affirmative ability-based constructions with no overt politeness or optionality markers are hardly used as requests, probably because they feel too impositive. Other ability-based interrogative constructions, like CAN/COULD WE/YOU NOT DO X? (as in Can we not use that phrase, please?) are far from polite, and as explained in detail in Leech (2014: 156), are accompanied by a tone of annoyance. Another group of fairly productive constructions for requesting is formed by declarative constructions expressing the willingness of the speaker that the action is carried out (i.e. I WOULD LIKE YOU TO DO X/I WISH YOU TO DO X): (38) ‘And now, if you will examine the floor,’ the magician requested. ‘All of it. I wish you to assure the audience that there are no hidden trapdoors.’ [iWeb] This type of construction also activates the politeness and optionality variables through the use of past modals or lexical means (i.e. wish is more tentative than want). The present tense variant (i.e. I WANT YOU TO DO X), which does not activate the key request attributes of optionality or politeness, is much less productive in the corpus. Interrogative constructions that activate the attribute concerning the willingness of the hearer to perform the requested action are also fairly frequent (i.e. IF YOU WILL DO X, WILL YOU DO X?, WOULD YOU DO X?): (39) 'Would you pass the grated cheese?’ I requested. [iWeb] (40) ‘If you will examine the floor,’ the magician requested. [iWeb] Both the interrogative sentence type and the conditional phrase activate the optionality of the speaker to comply, and in turn, the politeness of the request. In (39) politeness if overtly expressed by means of the use of a past modal. 120 are more prone to resort to blackmailing and coercive means (see Pérez-Hernández, 2001: Chapter 10). Speakers uttering beggings also work under the assumption that the hearer’s willingness to comply with their proposal is low (Pérez-Hernández 2001: 211). The speaker’s strong desire that the action is carried out, his lack of power to impose his will on the hearer, and his assumption that the hearer is probably not willing to comply explains the insistence, and higher use of politeness that characterises the act of begging. Example (5) clearly illustrates these semantic traits: (5) Given Jacob’s history with hobbies, it was no surprise that Jacob’s father was reluctant to buy him a magician’s kit for his birthday. ‘Geez, Jacob… You sure you wouldn’t rather I got you more guitar lessons?’ He suggested. Jacob was insistent. ‘Dad, you’ve got to get me the magician’s kit. This time I’ll stick with it for real. I promise! Come on, Dad.’ Jacob begged. [iWeb] In fact, in those cases in which the hearer makes his willingness and/or disposition to comply explicit in advance, politeness and insistence are not deem so necessary and the begging tends to resemble an average request for action. This is the case with example (6), which is much less insistent (i.e. there are no repetitions) and less polite (i.e. no overt use of politeness markers like please) than prototypical pleas like (1). (6) A cyclist lying on his deathbed asked his best friend to do him a favour when he’d gone. ‘Anything,’ replied his friend. ‘Just don’t let my wife sell my bikes for what I told her I paid for them,’ he begged. [iWeb] Is the action expressed in the act of begging necessary? Like orders and requests, begs may also stem from the speaker’s needing an object or from his needing an action to be carried out. This semantic attribute motivates some begging constructions based on the expression of necessity: (7) ‘Please sir, I need medical help,’ I begged, barely able to speak. [iWeb] How much freedom does an act of begging allow the hearer? The speaker uttering a begging does not have the necessary power to impose his will on the hearer. The freedom of the latter to do as asked is therefore high by default. This was also the 121 case with requests, which respected that optionality and even attempted to increase it as an indirect means of enhancing the politeness of the act (see Section 4.2.1). Beggings work differently. Motivated by his strong desire that the proposed action is carried out, the speaker attempts to somehow limit the possibilities of the hearer to refuse to comply. Limiting the optionality of the hearer cannot be done through imposition (as in the case of orders) or coercion (as in threats), due to the lack of power that characterises speakers uttering beggings. Consequently, the powerless, but eager, speaker favours other strategies to secure the hearer’s compliance. Among them, he will make extensive use of a combination of insistence (to make his degree of desire clear to the hearer) and a marked use of politeness (to gain the hearer’s compliance while avoiding imposition at the same time). Increasing the expressiveness of the act (by means of interjections, forceful intonation, lengthening of sounds, etc.) may also help increase the hearer’s compliance. These realisation procedures allow the speaker to convey his strong desire that the action is carried out. The overt expression of the speaker’s wishes puts the hearer in the position of having to comply if he does not want to confront the conventions of social interaction included in the Politeness Principle (Leech 1983). To a lesser extent, other strategies, like negotiating, minimizing of cost, pampering, or appealing to justifying reasons, have also been found to be used in order to restrict the hearer’s optionality and to move him to comply with the speaker’s begging. Example (8) illustrates several of the strategies by means of which the speaker attempts to force the hearer to grant his wishes (i.e. politeness, insistence, negotiation, mitigation, expressiveness): (8) ‘Oh, please don’t Aunt Lucy! I promise I won’t do it again. I’ll give you everything I have. Just please don’t tell Mum,’ he begged. [iWeb] Does the speaker attempt to minimise the cost of the action? Previous studies based on more limited data have suggested that beggings, like requests, minimise the cost of the requested action as a way of persuading the hearer to comply (PérezHernández, 2001: 214-215). The results of the present corpus analysis, based on 500 instances of beggings, contradict previous findings. Minimisation of cost, although a possible strategy compatible with the aim of beggings (see example 6), is not extensively used by English speakers. When begging, they do not rely on this strategy to achieve their goal, favouring others like the use of politeness, negotiation, or offering justifications. It is only natural that minimisation of cost should not be central to the conceptualisation of beggings. Minimizing the cost involves asking for less, and this contradicts one of the essential semantic attributes of 122 begging: the fact that the speaker has a strong desire to achieve his goal. It makes no sense, therefore, to ask for less than what is actually wanted in order to minimise the cost of the action. As a result, the compliance of the hearer is searched through other means (e.g. politeness, negotiation, etc.). Does the speaker attempt to be polite? Politeness, in combination with directness and insistence, has been shown to be a characterizing semantic feature of the act of begging. The lack of social power of the speaker, his high desire to get the hearer to carry out the proposed action, and his assumption that the hearer may not be willing to comply motivate the use of politeness as a means to persuade the hearer into action. Politeness is also a way of humbly acknowledging the inherent freedom of the hearer to decide for or against carrying out the required action. Example (9) illustrates the extensive use of politeness found in prototypical beggings: (9) ‘Please, please’ she begged, ‘Rape me, take my money but please don’t kill me please.’ [iWeb] The need for politeness decreases when the beneficiary of the action includes the hearer as well as the speaker (see example (3) above), as well as when the power of the hearer increases or the social distance between speakers is smaller, as in examples (10) and (11) respectively: (10) ‘Tell me who you are,’ the King begged, for he was amazed by her beauty and magic. [iWeb] (11) ‘Mama, get me out of here,’ the girl begged. [iWeb] Is social power relevant to the act of begging? As illustrated by examples (1) to (11), beggings can be performed regardless of the power relationship that holds between speakers. Even kings, who represent a clear type of social power, may find themselves in the need of begging in certain contexts (see example 10 above). Nevertheless, we have argued that speakers uttering beggings typically lack the power to impose their will. As explained in Pérez-Hernández (2001: 216), the hearer’s power need not arise from an institution. In the case of beggings, the hearer is the person who has the ability to grant the speaker’s wishes and this endows him with an ad hoc superiority over the latter. Thus, in example (10), in spite of his institutional power, the king, who is in love with the hearer, 123 finds himself in a position of inferiority because of this, being aware of the fact that his power as king does not grant him control over the woman’s emotions). This inferiority of the speaker, whose wishes depend on the hearer’s disposition, is one of the motivating factors that explain the insistence and higher levels of politeness that characterise beggings, two strategies used to overcome this inferiority and to persuade the hearer into compliance. Is the social distance between the speakers relevant to the act of begging? Beggings can be performed regardless of the social distance that exits between speakers. One can beg to a stranger and to an intimate alike. Students of EFL, however, should be aware of the interactions of the variable of social distance with other semantic attributes of the act of begging. Thus, the data in our corpus suggests that, when begging, the larger the social distance between the speakers, the higher the need to increase the politeness of the act will be: (12) The old beggar woman saw Calista and began speaking. ‘Oh m’lady, would you be so kind as to…’. ‘Get out,’ Calista demanded. ‘Please, I am not asking for much, just food scraps, maybe an old blanket…’ the old woman begged. [iWeb] On the contrary, example (11) above showed how short social distances are linked to smaller needs of politeness. Pérez-Hernández (2001: 217) has explained that small social distances correlate with a higher predisposition of the hearer to grant the speaker’s wishes. In these contexts, the familiarity between the participants gives the speaker reasons to expect collaboration on the part of the hearer. This makes politeness and insistence less necessary, unless the speaker asks for something too costly or too important for the hearer. Is the formality of the context relevant to the act of begging? The corpus under scrutiny yields instances of begging in all types of contexts, regardless of their formality, though they tend to be more common in informal contexts. Formal contexts are generally more structured and less prone to expressions of subjectivity or emotions. In our corpus, those beggings found in formal contexts tend to be less insistent, and to rely on negotiation strategies to secure the hearer’s compliance. Example (13), set in the formal context of a royal audience, illustrates this use of negotiation, in which the speaker promises something back in return for the king’s compliance. In addition, this instance of 124 begging does not display the insistence (i.e. repetitions, intensifiers) or expressiveness (i.e. exclamations, interjections) that characterise its informal counterparts. (13) Shaqiq went to the Prince. ‘Give me three days, and I will bring your dog back to you. Set my friends free,’ he begged. [iWeb] Table 4.9. The Know-What of Beggings Beggings are directive speech acts by means of which a speaker asks the hearer(s) insistently to carry out an action. When someone begs someone else to do something, they communicate the following: ‘I really want you to do this. I know you have the power to decide for or against doing it. I acknowledge your power, but I have to keep trying hard to get you to do it because I really want to achieve my goal.’ What you need to bear in mind in order to perform a successful act of begging: -The speaker, who is the prototypical beneficiary of the action, has a strong desire to achieve his goal (i.e. get the hearer to do as told), which leads him to be direct, unambiguous, and insistent in the expression of his begging. -The hearer is perceived as being more powerful than the speaker. This power need not be institutional. Simply the fact that he is the one who can make the speaker’s wishes come true endows him with an ad hoc superiority. -The speaker feels that he lacks the necessary power to impose his will on the hearer. -The hearer is largely free to decide for or against doing what the speaker says. -The speaker acknowledges and respects that the hearer’s optionality by being overtly polite, but he also keeps trying to achieve his goal by insisting, negotiating with him, pampering him, and to a lesser extent, by minimizing the cost of the required action. BE AWARE!!! -Politeness is not so necessary when the benefit is both for the speaker and the hearer, and/or the action is not very costly. -You can beg someone to do something regardless of how powerful you may be, but less (social, institutional) power you have, the more polite you need to be. -You can beg someone to do something regardless of the social distance that separates you from the hearer, but the larger the social distance, the more polite you need to be. -You can beg someone to do something in all contexts regardless of their formality, but the more formal the context, the less insistent you have to be and the more extensive your use of negotiation strategies should be. 4.3.2 The Know-How of Beggings The lack of research on linguistic constructions and realisation procedures for the expression of beggings parallels the virtually inexistent attention devoted to the teaching of this directive speech act in the ten textbooks analysed for this study. 14 As reported in Chapter 3, only one of them takes into consideration this speech act. This is done in passing, as part of a reported 14 The only previous in-depth study of beggings dates from the beginning of the century (Pérez-Hernández, 2001: Ch. 8) and it is based on a smaller collection of examples. The present study comes to confirm Pérez Hernández’s (2001) initial conclusions regarding the realisation procedures that activate each attribute of the begging ICM (i.e. each piece of knowledge included in the Know-What of beggings). In addition, it offers a collection of base constructions for the expression of beggings in English, and a comparison with those of the Spanish language. 125 speech activity in which students are asked to rephrase the direct act of begging (i.e. Please, please, don’t wear those old jeans) using the speech verb to beg. There is no explicit explanation about the semantics of this directive act, much less about the characteristics of the linguistic construction that is used as illustration. As shall be made apparent in this section, the English language offers a rich array of base constructions and realisation procedures to allow the expression of the act of begging, all of which find a motivation in the semantic description of the act provided in Section 4.3.1. EFL students at the advanced level of instruction would surely benefit from the teaching of at least the most frequent among these constructions. Table 4.10 summarises the quantitative data on begging constructions resulting from the analysis of the first random 500 occurrences of beggings in the iWeb corpus: Table 4.10. Base constructions for the act of begging in English With over seventy four percent of the total number of occurrences, IMPERATIVE base constructions clearly outnumber other linguistic constructions in the expression of beggings. This is only to be expected since the speaker uttering a begging has a strong desire to achieve her goal. In turn, this motivates the use of the imperative as a direct, unambiguous way for the speaker to communicate the hearer the action that she expects him to carry out. BEGGING BASE CONSTRUCTIONS OCCURRENCES PERCENT IMPERATIVE 374 74.8% I BEG YOU TO DO X 46 9.2% CAN I/YOU/WE DO X? 28 5.7% CAN’T I/WE/YOU DO X? 16 3.2% I WANT/NEED X I WANT/NEED YOU TO DO X 10 2% YOU HAVE/MUST DO X 6 1.2% COULD YOU DO X? 4 0.8% WON’T YOU DO X? 4 0.8% WILL YOU DO X? 2 0.4% WOULD YOU DO X ? 2 0.4% DO YOU THINK YOU COULD DO X? 2 0.4% YOU’LL DO X (WON’T YOU?) 2 0.4% YOU CAN DO X 2 0.4% TOTAL 500 100% 126 As has previously been shown to be the case with orders and requests, the use of IMPERATIVE constructions for the expression of beggings displays some specific characteristics of its own which constitute an essential teaching target for EFL students. The realisation procedures that combine with the IMPERATIVE base construction in the expression of beggings (see Table 4.11) can be divided in two main groups: (1) those that, assuming the lack of power of the speaker to impose his will, attempt to persuade the hearer to do as asked, and (2) those that communicate the speaker’s strong desire that the action is carried out. Let us see each of them in turn. Table 4.11. Imperative base construction + Realisation procedures for beggings How can a powerless, but eager to achieve his will, speaker attempt to persuade the hearer to do as told? The data in our corpus shows that in many cases this is done by making use of politeness. The use of the IMPERATIVE base structure in combination with the adverb please amounts to over fifteen percent of the total instances of beggings in our corpus: (14) The brother then begged him: ‘Please, father, intercede for me to God that I may be allowed a little more time in which to amend my life’ The speaker in example (14) also attempts to minimise the cost of the required action (i.e. a little more time). Mitigation of cost was shown to be a pervasive strategy in the case of requests. This realisation procedure is also used in some instances of beggings in our corpus, but to a lesser extent (around two percent). As explained in relation to the Know-What of Beggings, since the speaker has a strong desire that the action is carried out, it does not make much sense IMPERATIVE CONSTRUCTIONS FOR BEGGING OCCURRENCES PERCENT BARE IMPERATIVE 18 4.3% IMPERATIVE + EXCLAMATIONS/INTONATION 78 18.6% IMPERATIVE + EXPRESSION OF POLITENESS (PLEASE) 66 15.7% IMPERATIVE + REPETITION 42 10% IMPERATIVE + NEGOTIATION (PROMISES) 22 5.2% IMPERATIVE + REASON CLAUSE 12 2.8% IMPERATIVE + EXPRESSION OF MINIMISATION OF COST 10 2.4% IMPERATIVE + A COMBINATION OF 2 OR MORE OF THE REALISATION PROCEDURES BELOW 172 41% TOTAL 420 100% 127 to ask for less than what is actually wanted. The higher degree of speaker’s will in the case of beggings, as compared to that of requests, provides an explanation for the more limited use of this realisation procedure. In begging, speakers seem to prefer other persuasive strategies that do not imply minimising the desired objective. Thus, around ten percent of the IMPERATIVE constructions in our data make use of negotiation and reasoning as a way to get the hearer to comply. Consider examples (15) and (16) respectively: (15) ‘Turn yourself in,’ Phil Vetrano begged his daughter’s killer two weeks after her death. ‘I will make sure the reward money goes to the person of your choice. Your sister, your brother, your mother. It’s a life changer.’ (16) ‘Simon, put him down. You’ll hurt him!’ Sean begged. In example (15) the speaker attempts to persuade the hearer to comply with his wish by promising to reward one of his relatives. The use of the IMPERATIVE base construction followed or preceded by a promise that benefits the hearer is a fairly frequent realisation procedure for begging. In (16), the speaker’s persuasion strategy is based on the provision of rational arguments for him to do as asked. Thus, the speaker points out that someone is to be hurt if the hearer does not carry out the required action. Appealing to the hearer’s rationality by means of including a reason clause, either before or after the IMPERATIVE base construction, is another common mechanism for begging. A second group of realisation procedures that combine with imperatives to render a begging interpretation is grounded on the strong desire of the speaker to achieve his goal, which is a central aspect of the meaning of beggings. Examples (17) to (18) below illustrate those realisation procedures that enable its linguistic instantiation: (17) ‘Oh God,’ the grave keeper begged, praying. ‘Save me.’ (18) ‘Don’t do that! Don’t do that! I’ll die.’ He begged hard, but no use… (19) ‘NOOOOOOO!!!! PLEASE, Aaron, STO-HAHAHAHA-P!!! Stop tickling me, please!’ Mike begged with tears starting to come out of his eyes. (20) ‘Please, don’t murder me? Pretty-pretty-please?’ he begged. The above examples illustrate some recurrent linguistic resources that accompany IMPERATIVE base constructions when used as beggings. There are exclamations and interjections (e.g. Oh God, NOOOOO!!!), the use of a particular type of intonation that may 128 range from persuasive to insistent (see example 20), and above all the use of repetitions (examples 18, 19 and 20). These realisation procedures help the speaker to convey his strong desire that the required action is carried out. In addition, as explained in detail in PérezHernández (2001: 224-226), this type of linguistic strategies exploits the Tact Maxim of the general Principle of Politeness (Leech, 1983), according to which participants are expected to maximise benefit to others. By stating his wishes clearly through an imperative sentence and by communicating how strong those wishes are by means of exclamations, interjections, repetitions, etc., the speaker is pragmatically constraining the hearer’s freedom to refuse to comply with them, since this would involve a breach of the Tact Maxim and would make him lose face. Some of the realisation procedures described above have already been shown to co-occur with the IMPERATIVE base construction in the expression of other directives. For example, the use of politeness (e.g. please) or minimisation of cost (e.g. just, only) have previously been shown to direct hearers to a request interpretation of the imperative (see Section 4.2.2). How can we help EFL students to produce clear-cut examples of beggings that differ from those of requests that use similar realisation procedures? As shown in Table X, forty-one percent of the instances of begging in our sample make use of a combination of an IMPERATIVE base construction and two or more realisation procedures among those of the two main groups of resources describe above. Native speakers favour this mechanism, which is only natural given that the use of a combination of two or more realisation procedures together with the IMPERATIVE base construction manages to activate all the essential key attributes of the act of begging: the speaker’s strong desire to achieve his goal and the acknowledgement of his inferiority and lack of ability to impose his will on the hearer. Example (21) illustrates this type of clear-cut beggings that make use of a combination of several linguistic resources to activate the meaning essentials of the directive under consideration: (21) ‘Oh, please don’t Aunt Lucy! I promise I won’t do it again. I’ll give you everything I have on you. Just please don’t tell Mum’, he begged. Thus, in (21) the speaker’s high desire that the hearer complies with his wishes is conveyed by means of interjections (i.e. Oh), exclamations, and repetitions (i.e. please…please). In addition, example (21) also exhibits attempts at persuading the hearer to do as told by means of negotiations (i.e. promises in return of his compliance) and the minimisation of the cost of the required action (i.e. just). 129 All in all, this combination of an IMPERATIVE base construction plus several realisation procedures of the types described above manages to activate a clear begging interpretation and makes a good teaching target for EFL advanced students. The rest of the base constructions for begging that have been found in the data under study are much less frequent (less than nine percent of the total number of occurrences in our corpus). Advanced students could also be taught that it is possible to beg by means of base constructions that make use of explicit performatives (i.e. I BEG YOU TO DO X), or which activate the ability variable (i.e. CAN/CAN’T/COULD YOU DO X?), the speaker’s willingness or need for the action to be carried out (i.e. I/WE WANT/NEED YOU TO DO X) or the hearer’s volition (i.e. YOU’LL DO X and WILL/WON’T/WOULD YOU DO X?). As can be observed in examples (22) to (24), these base constructions also combine with the realisation procedures described in relation to IMPERATIVE base constructions to secure a clearer begging interpretation: (22) CAN YOU DO X? base construction + expression of politeness (please), exclamation/intonation, and negotiation (promise): ‘Mom, can you please stop buying store brand sausage? I will pay for Jimmy Dean!’ she begged. (23) I/WE NEED X base construction + repetition, expression of politeness (please): ‘We need to go to the club now. Please. Please. Right now,’ she begged and pleaded. (24) YOU’LL DO X base construction + repetition (tag question): ‘You’ll come out after supper, won’t you?’ he begged. Comparison with the constructions for beggings in the Spanish corpus yields relevant information for EFL professionals. The first four base constructions in Table 4.12 are used in both languages. Nevertheless, the first two (in green) have a similar low frequency of occurrence both in English and Spanish. Therefore, the most productive base constructions for EFL teaching purposes are those that make use of the imperative (+ one or more realisation procedures; see above) or of an explicit performative verb (i.e. I BEG YOU TO DO X). Both languages display a more extensive use of the imperative than of the performative, although in Spanish the use of the performative almost triples that in English. The explicit performative is a very direct, unambiguous way of performing a begging, and the Spanish language has often been characterised as having a higher tendency to directness than that of English. 136 (10) I turned and faced him [a stranger] ‘Why don’t we talk a little, then?’ I suggested. ‘You can tell me a bit about yourself and how you manage to defeat formidable foes in gym battles.’ Does the speaker attempt to be polite? Suggestions are characterised by their lack of overt politeness. None of the suggestions in our corpus displays the use of the adverb please or is realised by means of formulaic expressions of politeness (e.g. WOULD YOU MIND DOING X?). If at all, politeness is conveyed through indirection (i.e. interrogative sentence types, use of modals in the past tense). These resources are primarily aimed at increasing the optionality of the agent. Only secondarily, by decreasing the imposition of the act, they increase its politeness. The lack of deference is a feature of suggestions that has been identified in different languages as noted in cross-cultural studies like those of Rintell (1979), who observed that suggestions were less polite than requests both in English and Spanish. Politeness is a persuasive strategy aimed at gaining the hearer’s compliance. Since the speaker uttering a suggestion is not as concerned or interested in obtaining such compliance as is the case with requests, politeness is not felt as necessary. Is social power relevant to the act of suggesting? Suggestions can be performed regardless of the power relation that holds between participants. Students should be aware, however, that if there is an asymmetrical power relationship, the speaker may want to attempt to minimise the cost, and/or increase the politeness of the act. If the speaker is more powerful than the hearer, this will help to avoid imposition and to keep the optionality of the hearer within the limits expected in the act of suggesting. (11) Why don’t you go ahead and take a little bit of a swim anyway? he suggested. ‘Maybe it’ll cool you off.’ In example (11) a father avoids the imposition that may characterise a paternal command by increasing the tentativeness (i.e. Why don’t you…?) and by minimising the cost (i.e. a little bit) of the action he suggests to his son. If the speaker is less powerful, the minimisation of cost and/or an increase in politeness will help keep the imposition of the act low and avoid the negative consequences of a facethreatening act (see House & Kasper (1981: 143) for similar conclusions). This is the case in 137 example (12) in which the hearer has a higher social status than the speaker. The latter’s suggestion relies on tentativeness (i.e. Why don’t you…?) and minimisation of cost (i.e. a bit) to avoid imposition. (12) Somehow Amanda had a hard time thinking of this poor unhappy girl as a recording star, even though she was clutching a guitar case to her chest. ‘Why don't you sit down and rest a bit,’ suggested Amanda, adopting a solicitous attitude. Is the social distance between the speakers relevant to the act of suggesting? Most suggestions in our corpus display participants who are socially close. It is also possible to perform the act of suggesting when the social distance between participants is large. Students should be reminded that these situations require higher dose of minimisation of cost and politeness to avoid imposing on the hearer. See examples (10) and (12) above. Is the formality of the context relevant to the act of suggesting? Most suggestions in our corpus take place in informal contexts. Those uttered in more formal settings have been found to display higher levels of minimisation of cost and politeness; or alternatively to make use of unambiguous performative verbs that have lexicalised the nonimpositive nature of suggestions (i.e. I SUGGEST THAT YOU DO X). Formal contexts, by default, tend to be more polite. In this regard, suggestions do not differ from other directives. Table 4.13. The Know-What of Suggestions Suggestions are weak directive speech acts by means of which a speaker asks the hearer(s) to consider the merits or benefits of a potential course of action. When someone suggests someone else to do something, he communicates the following: ‘I believe that a particular course of action may be positive for you, for both of us or for someone else. You could consider doing it. I have no particular interest in you carrying out that action, neither do I have the social, institutional, or experiential power to impose the action on you, but I share it with you for your consideration.’ Therefore, the main semantic attributes of the act of suggesting are the following: -The beneficiary of the action may be the speaker, the hearer, both of them or a third person. -Regardless of who is to benefit from the action, the speaker shows no special desire or interest in its materialisation. He merely presents it for the consideration of the hearer. -Whether or not the hearer wants to do as told is unknown or irrelevant for the speaker, who respects the hearer’s freedom to comply or refuse to do as suggested. -In order to respect the optionality of the hearer, the speaker makes no overt attempt to minimise the cost of the action or to persuade the hearer to comply through insistence, imposition or the use of politeness. BE AWARE!!! You can make suggestions in all contexts, regardless of their formality, and of the power relationship or social distance that separates you from the hearer(s). However, if there are power asymmetries between the speakers, 138 if the social distance between them is large, or if the context is formal, your suggestions will need to include attempts to minimise their cost and/or exhibit a higher use of tentativeness and politeness. 4.4.2 The Know-How of Suggestions Research on linguistic constructions available for the expression of suggestions is scarce. One of the few existing taxonomies in English is based exclusively on a compilation of those formulae found in theoretical works (Martínez-Flor, 2005) and offers no information as to which constructions are preferred by native speakers. Jiang’s (2006) more extensive corpusbased classification of suggestions provides quantitative data about the frequency of occurrence of each construction, as well as a comparison between those found in real language and those that appear in EFL textbooks. He concludes that there is a relevant gap between the way native speakers perform this speech act and the linguistic strategies included in EFL teaching materials. Jiang’s (2006) taxonomy, however, is restricted to suggestions produced in a very specific setting (i.e. student study groups and student-teacher office hour interactions), in which speakers have a fixed social and power status. In this particular setting, the LET’S DO X construction and performative structures like I SUGGEST YOU DO X, which do not appear in the textbooks under his analysis, are found to be some of the most frequently used structures for suggestions. Jiang (2006) also notices that all textbooks emphasise the use of the modal should (YOU/WE SHOULD DO X), while his corpus shows that constructions with have to and need to are used much more often in the registers under examination. Finally, Jiang (2006) points out that the formulaic expressions HOW ABOUT DOING X?/WHY DON’T YOU DO X?, etc., which are ubiquitous in EFL textbooks, are not common at all as expressions for suggestions. As shall be shown below, the analysis of a larger and more varied corpus, which is not limited to a single situational type, yields somehow different results. Table 4.14. Base constructions for the act of suggesting in English SUGGESTING BASE CONSTRUCTIONS OCCURRENCES PERCENT IMPERATIVE 88 17.6% LET’S DO X 78 15.6% WHY DON’T YOU/WE DO X? 70 14% YOU/WE SHOULD DO X YOU/WE OUGHT TO DO X 52 10 10.4% 2% YOU/WE CAN DO X 46 9.2% HOW ABOUT DOING X? 45 9% 139 The IMPERATIVE and the LET’S DO X base constructions stand out from the rest as regards their frequency of occurrence in our corpus. This could be shocking under the traditional speech act theory view of imperatives as impositive strategies. In this line of reasoning, Koike (1994), Hinkel (1997) and Martínez-Flor (2005), among others, state that imperatives are regarded as the most direct and impolite forms of making a suggestion. If IMPERATIVE base structures were actually as impolite as traditional pragmatists deem them to be, we may wonder why so many native speakers chose this linguistic strategy for the performance of an intrinsically nonimpositive speech act like suggesting. As was explained in detail in Chapter 2, however, it is possible to adopt a weaker version of the literal force hypothesis, according to which imperative sentences are defined simply as those which present the content of a proposition for realisation (Risselada 1993: 71). Under the light of a weaker literal force hypothesis, imperatives are not impositive per se. In previous sections it has also been shown how the combined used of the IMPERATIVE base construction with specific realisation procedures for the semantic variables of different directive categories leads to a metonymic activation of particular speech acts. IMPERATIVE base constructions are combined with impositive linguistic strategies (e.g. expressions of immediateness, forceful intonation, etc) to activate orders; with strategies of mitigation and politeness (e.g. adverb please, expressions of optionality, mitigation, and social closeness) to yield requests; and with a combination of strategies of persuasion and insistence (e.g. repetitions, adverb please, negotiation and reasoning) to prompt a begging interpretation. The use of the imperative for the expression of suggestions also displays some peculiarities of its own as captured in Table 4.15. YOU/WE COULD DO X 34 6.8% WHY NOT JUST DO X? 15 3% YOU/WE NEED TO DO X 12 2.4% YOU MIGHT DO X YOU MIGHT WANT TO DO X YOU MIGHT WANT TO TRY TO DO X 10 2% I SUGGEST YOU/WE DO X 9 1.8% YOU/WE CAN/COULD ALWAYS DO X 7 1.4% YOU/WE’D BETTER DO X 7 1.4% WHAT IF YOU/WE DO X? 7 1.4% WHAT ABOUT DOING X? 5 1% BETTER DO X 4 0.8% WHAT DO YOU SAY TO DOING X? 1 0.2% TOTAL 500 100% 140 Table 4.15. Imperative base constructions + Realisation procedures for suggestions As the data in our corpus reveal, what characterises IMPERATIVE base constructions in relation to the act of suggesting is their use in isolation. As defined by Risselada (1993: 71) the imperative simply presents the content of a proposition for realisation, and that is precisely the aim of a suggestion: to present an action for the consideration of the hearer. Bare imperatives used with a neutral, non-impositive intonation suit well the nature of suggestions and this is in fact the way in which native speakers use them. In those rare occasions in which the IMPERATIVE base construction is found in combination with other linguistic strategies, these are aimed at increasing the tentativeness of the act (e.g. use of interrogative intonation or adverbs of probability like maybe, perhaps, etc.). The six occurrences in which the speaker provides the reasons that motivate him to ask the hearer to perform the proposed action imply benefits either to a third person (as in example 13), or both to speaker and hearer (as in example 14), and constitute peripheral instances of suggestions which could be also interpreted as weak requests or pieces of advice, respectively: (13) ‘So, get him another,’ Niklas suggested. ‘He could make a few credits up there.’ (14) ‘Then don't bother,’ Ginger suggested. ‘Use the museum as an excuse to raise the front gate price, to maybe five-fifty, or even six bucks. A buck isn't going to cut the traffic down much, especially since you can advertise the museum as part of the gate price. That would also help cover the cost of the back-pit improvements.’ Declarative sentences with modal verbs of ability, necessity, probability and obligation are also extensively used as strategies for the expression of suggestions. Altogether they amount to thirty-two percent of the suggestions in our corpus. Native speakers show a preference for constructions including the modal verbs should (i.e. YOU/WE SHOULD DO X) and can (i.e. YOU/WE CAN DO X), which amount to roughly ten percent each of the total number of suggestions: IMPERATIVE CONSTRUCTIONS FOR SUGGESTIONS OCCURRENCES PERCENT BARE IMPERATIVE 78 88.6% IMPERATIVE + REASON CLAUSE 6 6.8% IMPERATIVE + INTERROGATIVE INTONATION 2 2.7% IMPERATIVE + EXPRESSION OF PROBABILITY 2 2.7% TOTAL 88 100% 141 (15) ‘Perhaps you should go to the military museum and see it,’ I suggested. (16) ‘You can tell us about the time you fell into Clearedge Lake,’ Pip suggested. This contrasts with Jiang’s (2006) findings. According to this author, declarative base constructions with the modals have to and need to outnumber those with should. However, it should be taken into account that his corpus included a limited set of situations (i.e. study groups and student-teacher interactions). In addition, student-teacher interactions involve an asymmetrical relationship in which the teacher has a higher institutional and knowledge power. This could explain the higher frequency of occurrence of more impositive modals like have to in this particular setting. In our corpus, however, there are no instances of suggestions with the modal verb have to and those with need to are just one point six percent of the total number of examples. These results, which are in line with those of Edmonson and House (1981), Banerjee and Carrell (1988), Koike (1994), Bardovi-Harlig and Hartford (1996), and Alcón and Safont (2001), seem more compatible with the non-impositive nature of suggestions. Ability modals and the tentative modal should respect the freedom of the hearer to choose whether to do the action or not, which characterises suggestions. In this connection, it has been attested that constructions for suggestions based on modal verbs tend to combine with an additional linguistic strategy: the use of expressions of tentativeness and probability. Thus, sixty-two out of the 160 constructions based on the use of modal verbs includes adverbs of probability like maybe or perhaps, expressions of tentativeness (e.g. YOU MIGHT TRY AND DO X), or an interrogative intonation. These linguistic strategies are used to further increase the optionality granted to the hearer. The third group of constructions for the expression of suggestions that stands out in our data comprises a rich collection of interrogative-based formulae like WHY DON’T YOU/WE DO X?, WHY NOT (JUST) DO X?, HOW ABOUT DOING X?, WHAT ABOUT DOING X?, WHAT IF YOU/WE DO X?, etc. Altogether they represent nearly one third of the examples in the corpus. Most authors agree that the interrogative forms employed by using these formulae are typical of suggestions (Wardhaugh, 1985; Wierzbicka, 1987; Koike, 1994). The vast majority of them are highly tentative since they do not even make explicit the identity of the agent of the suggested action (e.g. HOW ABOUT DOING X?, WHAT ABOUT DOING X?, WHY NOT JUST DO X?), which again suits well the fact that suggestions are characterised by allowing high doses of optionality to the hearer. These constructions were not common in Jiang’s (2006) corpus of teacher-student interactions. The asymmetrical power relationship between the participants in Jiang’s corpus, where teachers have knowledge 142 authority over their students, explains that their suggestions were on average less tentative and made a lesser use of interrogative formulae like the ones under consideration. In situations in which the speaker has a higher status than the hearer, the dividing line between a suggestion and a piece of advice or an order becomes blurry. The last set of strategies found in our corpus represents an heterogeneous group that includes constructions based on the use of performative verbs (e.g. I SUGGEST YOU DO X), and declarative sentences expressing the benefits of a certain action (e.g. IT MIGHT BE WORTHWHILE TO DO X, YOU’D BETTER DO X, BETTER DO X, etc.). None of these constructions is very productive. Koike (1997) and Koester (2002) have already argued that performatives are not widely used as suggestions, since they are felt as too direct and impositive. Martínez-Flor (2005: 175) found these formulas mainly in formal contexts. Table 4.16 compares the constructions used in Spanish and English for the expression of suggestions. In green there are those constructions that have a similar number of occurrences in the corpora for both languages. As can be observed, IMPERATIVE base constructions, including LET’S DO X performatives, which in Spanish correspond to both the first-person plural of the present subjunctive and the VAMOS A HACER X formula, take pride of place in both languages with a similar number of tokens. The conventional formulae for suggestions (i.e. WHY DON’T YOU/WE DO X?, WHAT ABOUT DOING X?) also have counterparts in Spanish (i.e. ¿POR QUÉ NO HACES/HACEMOS X?, ¿QUÉ TE PARECE HACER X?). These constructions show a similar number of occurrences in both languages, the first one being much more productive in both of them. Likewise, the use of performative verbs for the expression of suggestions is possible in both languages, displaying a similar low degree of occurrence. Given their existence and similar degree of occurrence in both languages, these constructions are expected to be easy for Spanish speakers to learn. This should be taken into account by EFL teachers and professionals in order to schedule the teaching of suggestion constructions throughout the different levels of instruction. The easiest constructions could be programmed for the initial levels. Those constructions that show differences between the two languages (in orange and red in Table 4.16; see explanation below) should be dealt with in subsequent years and would most likely require a larger amount of time and resources. Suggestion constructions based on the use of modal verbs (i.e. YOU SHOULD DO X, YOU CAN/COULD DO X, YOU NEED TO DO) are found in both languages. The English language shows a preference for the use of modal verbs within declarative sentences, while Spanish displays a more varied range of formulas based on the use of modal verbs, including declarative, but also impersonal forms (i.e. LO QUE DEBERÍA HACERSE ES X (WHAT 143 SHOULD BE DONE IS X), LO QUE SE DEBE HACER ES X (WHAT HAS TO BE DONE IS X), SE PUEDE/PODRÍA HACER X (X CAN/COUND BE DONE), and interrogative-based constructions (i.e. ¿NO PODEMOS HACER X?, ¿NO TENDRÍAS QUE HACER X?). The latter two do not yield any occurrences in the English corpus, thus suggesting a low productivity in this language. SPANISH also makes quite a significant use of the modal verb TENER QUE (HAVE TO) in constructions like TIENES/TENEMOS QUE HACER X (YOU/WE HAVE TO DO X). At first sight, it may strike as odd that this impositive modal is used for suggesting. However, a closer look at the examples shows that their impositive force is generally softened by means of a reason clause justifying the command and/or pointing out the benefits that the hearer may obtain if he decides to comply (e.g. ¡Está casi lloviendo! Tenemos que llegar cuanto antes a un lugar habitable (It’s almost raining! We have to get to a shelter as soon as possible!)). Interrogative-based constructions that are possible in both languages include WHY NOT DO X? and WHAT IF YOU/WE DO X? formulae, both displaying slightly different amounts of productivity in each language. Likewise, some declarative constructions based on the statement of what is considered to be the best course of action according to the speaker (i.e. YOU’D BETTER DO X, BETTER DO X) are found in both languages with a slightly higher number of occurrences in the Spanish corpus. Finally, there is a number of constructions in red that either are not possible in one of the languages or do not yield any occurrences in one of the corpora. Among these, the HOW ABOUT DOING X? construction stands out from the rest. This formula has a significant presence in the English language (nine percent of the suggestions in the corpus) but shows no counterpart in Spanish. This is an English construction to which EFL materials should, therefore, devote special consideration. EFL textbooks should also make Spanish EFL students aware of the fact that some constructions for suggesting that are common in their language (i.e. HAY QUE HACER X, ¿HACEMOS X?) do not have a direct counterpart in English. Thus, the impersonal HAY QUE HACER X construction corresponds to personal constructions like YOU HAVE TO/MUST DO X in English. The latter do not show productivity in the English corpus, which is only to be expected since this type of personal constructions are highly impositive. The impersonal formula in Spanish, on the contrary, suits well the non-impositive nature of suggestions. Not making explicit the identity of the agent increases the freedom of the hearer to choose whether to comply with the suggestion or not. Posing a simple question in the present indicative tense (i.e. ¿HACEMOS X?) also respects the hearer’s optionality and, therefore, is useful in the expression of suggestions. This type of constructions, however, are 144 not possible in English (*Do we play cards after dinner?). Spanish EFL students should be made aware of this to avoid faulty realisations. Table 4.16. Comparison of base constructions for the act of suggesting in English and Spanish SUGGESTIONS BASE CONSTRUCTIONS (ENGLISH) Nº % Nº % SUGGESTIONS BASE CONSTRUCTIONS (SPANISH) IMPERATIVE 88 17.6% 136 27,2% IMPERATIVE (2nd person) LET’S DO X 78 15.6% 41 28 8.2% 5.6% VAMOS A HACER X HAGAMOS X (1st person plural Present Subjunctive) WHY DON’T YOU/WE DO X? 70 14% 50 10% ¿POR QUÉ NO HACES/HACEMOS X? WHAT ABOUT DOING X? 5 1% 4 0.8% ¿QUÉ TAL SI HACES/HACEMOS X? I SUGGEST YOU/WE DO X 9 1.8% 6 1.2% SUGIERO QUE HAGAS/HAGAMOS X YOU/WE SHOULD DO X YOU/WE OUGHT TO DO X 52 10 10.4% 2% 34 6.8% (LO QUE) DEBERÍAS/ DEBERÍAMOS HACER ES X YOU/WE CAN DO X 46 9.2% 26 5.2% PUEDES/PODEMOS HACER X YOU/WE COULD DO X 34 6.8% 20 4% PODRÍAS/PODRÍAMOS HACER X WHY NOT JUST DO X? 15 3% 8 1.6% ¿POR QUÉ NO HACER X? YOU/WE NEED TO DO X 12 2.4% 2 0.4% NECESITAS/NECESITAMOS HACER X WHAT IF YOU/WE DO X? 7 1.4% 20 4% ¿Y SI HACEMOS/HICIESEMOS X? Verb in 1st person pl. Present Simple Indicative or Past Simple Subjunctive YOU/WE’D BETTER DO X 7 1.4% 16 3.2% MEJOR HAZ/HAGAMOS X LO MEJOR ES QUE HAGAS/HAGAMOS X BETTER DO X 4 0.8% 10 2% LO MEJOR ES HACER X ES MEJOR HACER X QUIZÁS SEA MEJOR HACER X WHAT DO YOU SAY TO DOING X? 1 0.2% 0 Possible, but not productive HOW ABOUT DOING X? 45 9% 0 Not possible YOU/WE CAN/COULD ALWAYS DO X 7 1.4% 0 Not possible YOU MIGHT DO X YOU MIGHT WANT TO DO X YOU MIGHT WANT TO TRY TO DO X 10 2% 0 Possible, but not productive Not possible 0 29 5.8% HAY QUE HACER X Possible, but not productive 0 17 3.4% TIENES/TENEMOS QUE HACER X Possible, but not productive 0 12 2.4% (LO QUE) DEBES/DEBEMOS HACER ES X Possible, but not productive 0 10 2% (LO QUE) DEBERÍA HACERSE ES X Possible, but not productive 0 10 2% (LO QUE) SE DEBE HACER ES X Not possible 0 10 2% ¿HACEMOS X? ¿ [Present Indicative 1st person pl.] X? Not possible 0 8 1.6% SE PUEDE/PODRÍA HACER X Possible, but not productive 0 2 0.4% ¿NO PODEMOS HACER X? Possible, but not productive 0 1 0.2% ¿NO TENDRÍAS QUE HACER X? TOTAL 500 100% 500 100% 145 4.5 Advice Acts Searle (1969) defined the speech act of advice as ‘a weak directive whose illocutionary force is to suggest a future action to the hearer that the adviser believes will benefit the hearer’. Searle used the speech act verb suggest in the definition of advice acts. In fact, the defining line between these two illocutionary acts is very thin and some authors include both within the group of advice acts (Wierzbicka, 1987: 181-190). Suggesting and advising also share similar force dynamics. In both cases the speaker directs a force towards the hearer with the aim of getting him to consider a beneficial future course of action. If this force succeeds in persuading the hearer that the action is in his benefit, then it can eventually move him into action. The initial force (F), which needs to go through the filter of the hearer’s consideration and rationality (i.e. HR), could be blocked by the latter if deemed unnecessary or not beneficial. In Section 4.4.1, it was argued that the speaker uttering a suggestion does not make any overt attempt to persuade the hearer about the benefits of the suggested action. As will be shown below, this is main difference with advice acts. When advising, the speaker does attempt to persuade the hearer about the benefits of the action, and he does so by appealing to his rationality (i.e. by presenting the reasons, advantages and benefits of the proposed action). 15 As a result, the initial force of an advice act is stronger than that of suggestions and more likely to achieve its goal of turning the hearer into an agent that moves into action (i.e. HA). Since the force still needs to go through the filter of the hearer’s consideration, however, it is not felt as impositive as that of orders or threats. Figure 4.5. Force schema underlying the conceptualisation of advice acts 4.5.1 The Know-What of Advice Acts Advising is a complex category of speech acts which includes both those pieces of advice which have been asked for (i.e. solicited advice) and those which have been uttered spontaneously by the speaker (i.e. unsolicited advice). As shall be explained below, it is 15 In this respect, advice acts differ from requests, which are aimed at the social dimension of the hearer rather than to his rationality, thus exploiting well-established conventions of politeness and correct social behavior, as shown in Section 4.2.1.