Emerald | Current Research in the Semantics/Pragmatics Interface | Table of Contents

1 The Conceptual-Procedural Distinction: Past, Present and Future

literatinetwork@emeraldinsight.com (Deirdre Wilson) — Mon, 25 Jul 2011 00:00:00 +0100

My aim in this chapter is to reassess the conceptual-procedural distinction in the light of the last twenty years of research, and to consider some possible revisions or extensions. The first section is a brief introduction. In the second section, I outline the rationale for drawing a conceptual-procedural distinction of the type proposed in relevance theory (Blakemore, 1987, 2002; Wilson and Sperber, 1993). In the third section, I discuss some current issues and objections, looking in particular at whether procedural meaning is properly regarded as semantic, and at whether a single item can encode both conceptual and procedural meaning. In the fourth section, I suggest some possible revisions and extensions in the light of recent research on lexical pragmatics (Wilson and Carston, 2007; Sperber and Wilson, 2008a). In the fifth section, I discuss the relation between the conceptual-procedural distinction and the ‘massive modularity’ hypothesis (Sperber, 2005; Carruthers, 2006), and suggest some further revisions or extensions. I conclude that the conceptual-procedural distinction is well founded, and may have much more general application than has previously been thought.

2 On the Status of Procedural Meaning in Natural Language

literatinetwork@emeraldinsight.com (Carmen Curcó) — Mon, 25 Jul 2011 00:00:00 +0100

Procedural meaning has standardly been treated in the literature as semantic and representational. Recently, arguments have been put forward to suggest that in fact procedural meaning is dispositional and pragmatic in nature. In this chapter I revise these arguments in detail, and conclude that there are important reasons to assume the existence of a procedural non-dispositional semantics. I argue that procedural lexical items behave rather differently from typical dispositions, so that a dispositional view of them makes the wrong predictions. I sketch a possible alternative view that overcomes the problems envisaged by the dispositional view, but which maintains a representational and semantic stand on procedural meaning. I suggest a role for two possible genuine dispositions inbuilt in the semantic cognitive sub-system. Besides, I look at a number of communicative phenomena that have a procedural character, but which seem to be better understood as causal dispositions of the pragmatic system. The upshot is that not everything procedural in language and communication is of a kind. While it seems that procedural lexical items are better treated as encoding explicitly instructions on how to manipulate representations as part of their semantics, there are other aspects that affect the construction of meaning in interaction that are indeed better understood in dispositional terms.

3 On Some Methodological Issues in the Conceptual/Procedural Distinction

literatinetwork@emeraldinsight.com (Louis de Saussure) — Mon, 25 Jul 2011 00:00:00 +0100

The assumption developed in this chapter is that the available criteria for classifying an expression as conceptual or procedural are not specific enough, since they may lead one to think that some procedural expressions encode some relevant conceptual load of information (and vice versa). A methodological criterion is proposed: if all possible meanings of an expression across contexts can be predicted on the basis of a conceptual core and general pragmatic cognitive principles of inference seeking for relevance, then the considered expression has to be considered representational, hence conceptual. An expression should be considered procedural only when its meanings cannot be accounted for as conceptual. This assumption entails that the conceptual-procedural distinction does not have to map onto the distinction between grams and lexemes. Along these lines, it is also suggested that no expression should be expected to bear the two possible types of information, a claim which is illustrated by the case of modal verbs. Other examples discussed in this chapter are French connectives parce que (‘because’), ensuite and puis (‘then’) and some effects of the French imperfective past tense that do not seem predictable on any conceptual basis.

4 On the Rigidity of Procedural Meaning

literatinetwork@emeraldinsight.com (Victoria Escandell-Vidal, Manuel Leonetti) — Mon, 25 Jul 2011 00:00:00 +0100

This chapter puts forward the claim that rigidity is one of the central, characterising properties of procedural meaning, which plays a crucial role in accounting for the inferential resolution of a number of linguistic mismatches. Rigidity implies that linguistically encoded instructions have to be obligatorily satisfied in the interpretive process; contrary to conceptual information, they cannot be adjusted to comply with the requirements of other elements, nor can they be cancelled and modified by any pragmatic process. They systematically prevail over conceptual and contextual information whenever a mismatch or a contradiction arises between the meanings of two linguistic expressions or between a linguistic expression and the available contextual information. Three different kinds of mismatch involving procedural items are revised in order to show that the pragmatic processes triggered in the resolution of mismatches are to a large extent predictable. Conflicts between procedural meaning and contextual assumptions give rise to cases of accommodation; conflicts between procedural elements and conceptual content typically generate coercion phenomena; finally, a clash between two procedural items can only be solved, in the cases where this option is available, by means of a special ‘splitting’ mechanism and a reportive or quotative reading. Thus, significant generalisations about reinterpretation processes can be obtained if rigidity is taken as the most outstanding feature of procedural meaning.

5 Exploring the Borderline between Procedural Encoding and Pragmatic Inference

literatinetwork@emeraldinsight.com (Christoph Unger) — Mon, 25 Jul 2011 00:00:00 +0100

The insight that verbal communication involves a coding process that feeds into inferential ones naturally led to the conclusion that natural languages may encode not only conceptual content but also instructions for the inferential processing of conceptual content (Blakemore, 1987, 2002). But the inferential nature of verbal communication also suggests that linguistic expressions may be used for purposes other than the meaning they linguistically encode. Work on lexical pragmatics has shown this to be the case for words encoding conceptual meaning. In this chapter I argue that the same is true for lexical items encoding procedural information as well: tense, aspect, modality, and evidentiality markers may be used in redundant ways to achieve various cognitive effects, and noun phrase morphology can be exploited to trigger inferences as side effects. I account for this use of procedural markers for purposes other than their linguistically encoded meaning in terms of Vega Moreno's (2007) notion of pragmatic routines, that is, the process of whole inference patterns – premises and conclusions – becoming easily accessible in one go through repeated access.

6 Description as Indication: The Use of Conceptual Meaning for a Procedural Purpose

literatinetwork@emeraldinsight.com (Thorstein Fretheim) — Mon, 25 Jul 2011 00:00:00 +0100

A singular term in pronominal form does not survive the hearer's pragmatic processing but undergoes saturation and delivers what is understood to be its referent to the propositional content of the utterance. This chapter argues that the concepts encoded by the descriptive material of a referentially used definite description or a demonstrative description are also not entered into the content. The communicative function of such conceptual material is oftentimes to constrain the hearer's inference needed to identify the referent but descriptive meaning in referential terms does not necessarily facilitate reference resolution, it may instead direct the hearer to information that goes beyond the truth-conditional content. Linguistic data to be discussed and offered as support for my arguments include descriptive material used for metacommunicative pointing to antecedent information, description that may be relevant even though it does not facilitate reference resolution, and description intended to direct the hearer's attention to an extralinguistic stimulus. Descriptive terms used referentially are obviously not specialized for the expression of procedural meaning, but procedural information that constrains the hearer's computation is sometimes more effectively conveyed by means of a description than by means of an indexical or other context-sensitive lexical item that encodes procedural meaning.

7 Definiteness, Procedural Encoding and the Limits of Accommodation

literatinetwork@emeraldinsight.com (Christopher Lucas) — Mon, 25 Jul 2011 00:00:00 +0100

This chapter has two main parts. The first offers a defence and revision of Hawkins' (1978, 1991) theory of definiteness. I argue that Hawkins' pragmatic approach, centred on the notion of uniqueness within a mutually manifest ‘P-set’, is fundamentally on the right track, but that it is best formulated in consistently relevance-theoretic and, in particular, procedural-encoding terms, and that relevance theory provides the appropriate tools for a better understanding of some of the theory's key concepts. The second part of the chapter defends this procedural version of Hawkins' theory from a commonly cited class of apparent counterexamples whereby speakers routinely utter sentences containing definite NPs for which hearers cannot locate a mutually manifest P-set, but which are nevertheless judged acceptable. I argue that this is not a special property of definites, but a manifestation of the wider phenomenon of presupposition failure with accommodation, such that acceptability in this case is determined by the hearer's ability to adjust the context so that an appropriate P-set becomes mutually manifest. Such utterances do not, therefore, constitute counterexamples to the procedural theory of definiteness; in fact it is only terms of this kind of theory that they can be properly understood.

8 Beyond Reference: Concepts, Procedures and Referring Expressions

literatinetwork@emeraldinsight.com (Kate Scott) — Mon, 25 Jul 2011 00:00:00 +0100

This chapter focuses on the selection and interpretation of referring expressions. Working within the relevance theory pragmatic framework, I make two main claims. Firstly, by taking seriously the idea that referring expressions carry information which interacts with relevance theory principles, we can formulate an account of referring expression use that does not rely on a system of linguistic marking. I argue that whilst formulations of Accessibility, Givenness and so on may be descriptively useful, they are not necessary in an analysis which incorporates the notion of procedural meaning. Secondly, I argue that a relevance-theoretic analysis allows us to look beyond the mere act of referring and consider how referring expressions might contribute to what is implicated as well as what is explicitly expressed. This idea is missing from most accounts and it goes a long way towards explaining the choices made by speakers when selecting one referential expression rather than another. I focus on data comparing the acceptability of definite descriptions with complex demonstratives in various discourse contexts to explore this approach.

9 Child Language, Theory of Mind, and the Role of Procedural Markers in Identifying Referents of Nominal Expressions

literatinetwork@emeraldinsight.com (Jeanette K. Gundel) — Mon, 25 Jul 2011 00:00:00 +0100

The absence of a simple one-to-one mapping between linguistic form and a speaker's intended interpretation is perhaps most evident in the case of referring expressions. This is true for pronouns (e.g., she, that), which encode little if any conceptual content, as well as for full nominal expressions such as the papers in this volume, the woman in the red dress, all of which can refer to different entities in different contexts of use. This chapter examines the use of referring expressions in spontaneous conversation by children aged 3 and younger, with specific focus on determiners/pronouns which constrain possible interpretations by encoding procedural information about how the referent is to be mentally accessed (i.e., whether it is in focus, activated, familiar, uniquely identifiable, referential, or type identifiable). I argue that the procedural nature of the meaning of these forms explains why children are able to use them appropriately before they exhibit the representational ability to correctly assess the mental states of others that is measured by standard theory of mind tests.

10 Cross-linguistic Variation in Procedural Expressions: Semantics and Pragmatics

literatinetwork@emeraldinsight.com (José Amenós-Pons) — Mon, 25 Jul 2011 00:00:00 +0100

The verbal systems of Spanish and French share a compound and a simple past form (CP and SP), but their uses differ cross-linguistically. In French, the CP is often used as a narrative tense, whereas the Spanish CP has very limited narrative properties; conversely, the French SP only appears in very specific discourse functions, whereas its Spanish counterpart is used in all types of discourse. Tenses are procedural devices with a rigid and unique meaning, while eventualities have a more adaptable conceptual meaning. However, procedural instructions are general enough to enter different processes of pragmatic enrichment, giving way to diverse results according to specific contextual needs and following the principle of relevance. Tenses determine the way in which an eventuality is represented, and they contribute decisively to temporal reference. Despite their varied contextual effects, the linguistically underdetermined procedural meaning of CP and SP is (respectively) one and the same in Spanish and French. The differences between both languages correspond to dissimilar ways of materializing the procedural instruction of each tense; those differences can be described as an interaction of pragmatically enriched procedural and conceptual meanings taking tendencially divergent inferential paths in each language. Conventionalization of frequent interpretive paths (understood as privileged access to specific ways of carrying out a procedural instruction) comes as an effort-saving process. In the cases analysed here, such process is linguistically mandated: not internally (by the semantics of a single tense) but externally (by the structure and needs of the entire verbal paradigm).

11 Assertion, Relevance and the Declarative Mood

literatinetwork@emeraldinsight.com (Mark Jary) — Mon, 25 Jul 2011 00:00:00 +0100

This chapter puts forward an account of the linguistic semantics of the declarative mood that is able to explain its assertoric potential without thereby predicting that all uses of the declarative are assertoric. This is done by treating mood as providing non-conceptual input to the pragmatic process of utterance interpretation, as postulated by Relevance Theory. The claim is that the declarative delivers up a propositional representation in a format that, other things being equal, makes it likely that this representation will be employed as a premise in the derivation of contextual effects, thereby making a direct contribution to the relevance of the utterance in its own right. The nature of the most accessible context then determines, in part, whether the utterance has assertoric force. This approach has a number of advantages. First, it explains why only the declarative can be used to assert the proposition explicitly expressed by the utterance, a fact that is not explained by the standard Relevance-Theory account of mood. Second, it explains the relationship between assertion and main-point status without identifying the two. Third, it accounts straightforwardly for both main-clause and embedded occurrences of the declarative, even explaining cases where the embedded clause achieves main-point status. Both English and Spanish data are discussed.

12 The Procedure of Marking Contrast with Alternatives: A Constraint in the Derivation of Higher Level Explicatures

literatinetwork@emeraldinsight.com (Susana Olmos, Laura Innocenti, John Saeed) — Mon, 25 Jul 2011 00:00:00 +0100

Relevance Theory (RT; Sperber and Wilson, 1986) adopts an interactional model to describe the reliance on inferential behaviour in language comprehension. Blakemore (1987) was the first to provide an RT procedural account of but as a procedural unit which guides the hearer towards the cognitive effect of contradiction and elimination of a manifested assumption. The contribution of but was described as a constraint on the derivation of the implicit content. What we propose is a different view of the procedural meaning encoded by the Spanish counterpart pero. We propose that pero instructs the hearer to access alternatives, which are constructed online and constrained by the Principle of Relevance. By marking this contrast with the explicit representation that the connective introduces, the speaker is then constraining the propositional attitude in which the explicit representation should be embedded. We argue that our analysis of pero gives a more convincing account of the function of this connective than previous accounts based on the elimination and suspension of implicatures. We also want to propose that this approach can be further applied to other procedural expressions, such as Spanish connectives sino and aunque (Olmos and Ahern 2009) and temporal markers such as Italian già and ancora.

13 A Procedural Analysis of kadhalik

literatinetwork@emeraldinsight.com (Mai Zaki) — Mon, 25 Jul 2011 00:00:00 +0100

The objective of this chapter is to apply a procedural analysis to the demonstrative form kadhalik in Modern Standard Arabic. It is argued that the form kadhalik can function either as a demonstrative or as a discourse marker. In the first use it consists of kaaf ’al tašbiih (kaaf for simile)+distal demonstrative dhalik, while in the second use it has grammaticalised into a single semantic unit. Using corpus examples, these two uses will be differentiated as the chapter further argues that the semantic contribution of both kaaf ’al tašbiih and kadhalik as a discourse marker can be systematically accounted for in procedural terms. This explains how both terms are used to make a discourse relation explicit by encoding procedural constraints on the interpretation process. The distinction of the two uses of kadhalik in procedural terms also explains distribution patterns in corpus data.

14 Sentence Stress and the Procedures of Comprehension

literatinetwork@emeraldinsight.com (Daniel J. Sax) — Mon, 25 Jul 2011 00:00:00 +0100

This chapter develops a relevance-theoretic model – building on and refining the ideas of Wilson and Sperber (1979), Sperber and Wilson (1995), and Breheny (1998) – of how one aspect of intonation, the position of sentence stress, affects the incremental processing of simple utterances in English. After proposing a notation (adapted from Wedgwood, 2005) for modeling how cognitive effects may be derived from not-yet-fully-decoded utterances, the chapter advances the central tenet that the position of sentence stress should be seen as marking a particular stage in the hearer's incremental processing, encouraging the investment of effort in deriving cognitive effects at that stage. It sketches how this approach compares against other work within Relevance Theory, including efforts to analyze intonational contours or tones themselves in terms of procedural meaning, and it concludes, contra Breheny (1998), that although the position of sentence stress has an impact on the procedures of utterance comprehension, it is not best viewed as constituting a type of “procedural meaning” per se.

15 Procedural Encoding and Tone Choice in Buenos Aires Spanish

literatinetwork@emeraldinsight.com (Leopoldo O. Labastía) — Mon, 25 Jul 2011 00:00:00 +0100

This chapter explores some tone choices in the spontaneous speech of Buenos Aires (Argentina) Spanish and attempts to account for them in procedural terms along the lines suggested by Relevance Theory (Wilson and Wharton, 2006). In particular, it analyses stretches of discourse beyond individual tone units, and tries to show how nuclear tone and boundary tone choices and pitch range management affect the interpretation of those tone units in terms of their function in spoken discourse, and how participants in a conversation organize information to indicate dependence, continuity and discontinuity (House, 2006). More generally, it aims to show that these tone choices encode specific instructions to guide the hearer to the most relevant interpretation of discourse by reducing the processing effort necessary to achieve the desired cognitive effects, and therefore argues for a procedural account of intonation.

The prosodic analysis is carried out in the Autosegmental-Metrical framework (Pierrehumbert, 1980; Ladd, 1996) using its application to the study of Spanish intonation known as the Tone and Break Indices (Sp-ToBI) transcription system (Beckman et al., 2002; Hualde, 2003; Sosa, 2003; Estebas-Villaplana and Prieto, 2008), also used specifically in the study of Buenos Aires Spanish (Gabriel et al., 2010a).

The results of the analysis suggest that relevance is pursued both at a local and at a global level, and they lend support to a compositional approach to intonational meaning, in which different prosodic choices and their combinations guide pragmatic interpretation at different levels. Overall, the chapter aims to show that Relevance Theory in general, and procedural encoding in particular, offer an insightful way to deal with prosodic phenomena and their meaning.

Introduction: Procedural Meaning