One remarkable fact about language is that with a finite number of words and a fairly small number of ways of combining them, it allows us to refer to indefinitely many things, to express indefinitely many thoughts, to formulate indefinitely many requests, and so on. There are two core mechanisms that explain this remarkable fact: context-dependence and compositionality. When it comes to context-dependence, natural languages contain a set of designated expressions whose function is to exploit contextual features in order to pick out different things in different contexts. In English, indexical and demonstrative pronouns (I, here, this, etc.) are a case in point. The context exploited may be a concrete, real-world setting in which the utterance takes place, but it may also be the linguistic material that surrounds a given expression. Ellipsis and anaphora are phenomena in which linguistic context gets exploited. Yet a third kind of context-dependence that we find in language is a purely pragmatic one, in which the beliefs and intentions of the participants of a conversation allow them to endow their utterances with contents that go well beyond the literal meanings carried by the words that they use. Implicatures belong in this category, and so does the controversial mechanism of ‘free enrichment’ (see Recanati Reference Recanati2004: 10). Last but not least, we use language smoothly and efficiently because the context also allows us to take certain things for granted; this is the phenomenon of presupposition.
How does brevity fit into this picture of language and its use? We could say that ‘brevity’ is an encompassing term that covers the various mechanisms by means of which language users manage to do a lot with only a little. Indexicality, ellipsis, anaphora, implicature, free enrichment and presupposition would then all be seen as different facets of an overarching phenomenon of brevity. But ‘brevity’ is also intimately linked with Grice's maxim of Manner, one of whose submaxims is ‘Be brief’ (Grice Reference Grice1989: 27). One issue that not only underlies the present volume, but remains one of the most important open issues in the study of language is whether or not the above-mentioned phenomena all derive from some single, more basic phenomenon, as would be, for instance, some general rational cooperative principle that governed conversation, as Grice held. While this fundamental issue provides an excellent rationale for collecting articles on context-dependence–related phenomena in a volume about brevity, Brevity as a whole does not leave the reader with any clear idea as to what the answer might be.
The volume is composed of seventeen essays, distributed over four parts: ‘Brevity in language and thought’ (eight essays); ‘The philosophy of brevity’ (four essays); ‘Experimenting with brevity’ (four essays); and ‘Prolixity’ (one essay). While the essays in the first two parts are very diverse, those in Part III achieve internal cohesion by focusing on empirical data gathered through experimental studies.
The first three essays, by Jason Merchant, Lyn Frazier, Thomas Weskott & Charles Clifton, Anne Bezuidenhout, and Eleni Gregoromichelaki, Ronnie Cann & Ruth Kempson, offer different perspectives on the topic of fragment answers and ellipsis. The first essay, in line with Merchant's previous work, presents two studies whose aim is to show that a fragment answer comes with some syntactic material that has been elided. The first study shows that a question like ‘What did John say about Sue?’ is preferably answered with a ‘that’-clause (e.g. ‘That her husband abandoned her’ rather than a plain sentence such as ‘Her husband abandoned her’). The second study shows that in German, to a sentence like ‘With whom did Ana speak?’, subjects prefer an answer that includes the preposition ‘with’ to one that does not. One intriguing observation about the two studies is that, while there is this preference, a plain sentence answer is still judged as perfectly acceptable (with a 3.73 means on a five-point scale); and so is a plain noun phrase answer (e.g. ‘John’ rather than ‘With John’) (4.76 on a seven-point scale). Although Merchant et al. hold that ‘it is the relation between the means for the NP and the PP fragments that matters’ (29), it is hard to see how a mere preference for a pattern that is arguably best seen as involving syntactic ellipsis could be used to argue that in general, fragment answers come with such inaudible syntax. In the best case, what this shows is that between two competing mechanisms, one that involves ellipsis and another that does not, there may be some preference for the first one. What is more, it is not even clear that in fragment answers that do articulate a complementizer (‘that’), a preposition (‘with’), or, language permitting, a case-marker (e.g. accusative), syntactic ellipsis must be involved. As Bezuidenhout rightly observes, ‘on the assumption that syntactic structures are stored in the lexicon along with argument structures, lexical knowledge of the verb “see” would determine the correct marking for the fragment answer’ (41). Indeed, Bezuidenhout's essay aims at discrediting the idea that in fragment answers, additional syntactic structure first gets generated and then deleted. The third essay, by Gregoromichelaki et al. discusses fragment answers and ellipsis while paying close attention to coordination in dialogue. Their proposal brings together the linguistic framework of Dynamic Syntax with philosophical insights from pragmatists like Brandom. However, unlike Brandom, the authors do not think that high-level inferential reasoning should be essential to utterance interpretation. Rather, they see linguistic coordination as ultimately grounded in low-level pre-conceptual processing mechanisms.
Shifting gears, Christopher Gauker's contribution, ‘Inexplicit thoughts’, analyzes brevity as it is reflected not in language but in thought. He argues that even fully conceptual thought may still remain inexplicit. Reinaldo Elugardo's essay takes us back to language, and to the phenomenon of subsentential speech acts: like fragment answers, subsentential utterances do not correspond to full sentences, but unlike them, they do not follow any question or any other overt linguistic material; they may stand on their own, or even initiate a discourse. Elugardo argues that in order to understand what a speaker asserts by a subsentential utterance, the hearer must enrich the semantic content. In doing so, he spends some time criticizing John Perry's idea of reflexive content. Michael Glanzberg's essay, ‘A new puzzle about discourse-initial context’, presents a puzzle that, ironically, is not about discourse initials, but, rather, about anaphora and ellipsis. What worries Glanzberg is a possible scenario in which a machine randomly generates some sentence and then a speaker hooks up on it with an anaphoric device. The worry seems to be that even a randomly generated piece of discourse achieves almost automatically something that otherwise requires a lot of effort, shared background and intention-recognition, namely, introducing a discourse referent and making it an available anchor for anaphora. The essay by Anouch Bourmayan & François Recanati discusses the possibility of using certain verbs that have both transitive and intransitive uses (such as ‘eat’) in its intransitive form but with a transitive meaning. They hold that a sentence like ‘Sally ate’ can, in a suitable context, and via free enrichment, express e.g. the proposition that Sally ate the mushrooms that John cooked. They try to respond to a series of objections by giving a new twist to Recanati's older proposal (Recanti Reference Recanati2004) using resources from situation semantics. The last essay of Part I, by Matthew Stone, brings us back to the issue of brevity proper. The author draws on ideas from computational linguistics to offer a novel perspective on the phenomenon. It can be observed about human action in general that agents achieve greater efficiency by overloading their intentions, that is, by using an intended action to contribute to other goals, in addition to the original goal. Stone suggests that the same phenomenon affects discourse: by overloading one's communicative intentions, one conveys information in a more economical way. For instance, if one is given the instruction ‘Cut the square in half to create to triangles’, one easily infers that one should cut it on the diagonal. The speaker's intention to instruct that the square be cut on the diagonal is loaded into her original communicative intention that the square be cut so as to create two triangles (see Di Eugenio & Weber Reference Di Eugenio and Webber1996 for discussion). Communicative economy, then, could account for brevity (at least in some of its guises).
Part II, ‘The philosophy of brevity’, contains four articles that are philosophical in nature, but that do not address the deeper philosophical question as to whether the different phenomena related to brevity are all grounded in some single principle or process. Laurence Goldstein's ‘Some consequences of “speaking loosely”’ explores, albeit not in any conclusive way, the relationship between brevity and three sets of problems that have long been of interest to philosophers: Frege's puzzle (that is, the failure of substitutivity of coreferential names in attitude ascriptions), the Sorites paradox and vagueness, and Liar's paradox and self-reference. Jeff Pelletier's ‘Context, compositionality and brevity’ is concerned with the question of how compositionality may be preserved in the face of pragmatic phenomena of the sort discussed by contextualists. The other two essays in this part of the book are devoted to specific questions regarding presupposition. Andreas Stokke defends the Context Change Theory, developed by Irene Heim (Heim Reference Heim, Portner and Partee1983/2002), in face of charges originally raised by Mats Rooth (personal communication to Heim, 1987) and Soames (Reference Soames, Gabbay and Guenther1989) and more recently revived in Schlenker's (Reference Schlenker2012) critique of dynamic approaches to presupposition. By rebutting those criticisms, Stokke undermines one important motivation for Schlenker's Transparency Theory, which heavily relies on Grice's maxim of Manner. Manuel García-Carpintero is also critical of the pragmatic approach to presupposition advocated by Stalnaker (Reference Stalnaker, Munitz and Unger1974) and endorsed by Schlenker. Focusing on presupposition accommodation, García-Carpintero outlines a semantic account and compares it to the one recently put forward by Mandy Simons (Simons Reference Simons2006).
The four essays that constitute Part III all discuss brevity from an empirical point of view. In ‘“Be brief”: From necessity to choice’, Eve V. Clark & Chigusa Kurumada report evidence from language acquisition that shows that children gradually acquire a skill of being brief, i.e. being concise and providing necessary information in an optimal way. Their study shows that children's utterances become more compact between the ages of four and five, which they take to be correlated with their ability to make pragmatic inferences. Julie Sedivy surveys experimental evidence from studies with adults which suggests that interpreting an utterance may depend on the information that the listener has about specific speakers. She then turns to children's sensitivity to the particular characteristics of speakers, and presents two case studies that show that already early on, children discern between reliable and unreliable speakers, and their linguistic abilities, such as the contrastive interpretation of adjectives, get modulated when it comes to interpreting speakers with anomalous behavior. The results from psycholinguistics thus appear to show that one's mastery of cooperative principles develops early on, can be extremely computationally efficient, and does not follow the pattern of articulate, conscious inferential reasoning that one would normally associate with Gricean cooperative principles. Dan Grodner & Rachel Adler address the issue of how a speaker selects some referential expression over another and, in particular, how the divergence between the speaker's and the hearer's perspectives affects this choice. They report experiments whose results show that referential descriptions are tailored to respond to the addressee's communicative needs as well as being constrained by the speaker's communicative goals. Ira Noveck & Nicola Spotorno propose that brevity in language derives from a cognitive process that they call narrowing. They distinguish between voluntary narrowing, in which the hearer derives a content that is more informative than the one literally expressed out of her own initiative, and imposed narrowing, in which deriving such a content is necessary to make sense of the speaker's speech act. Crucially, they believe that in order to provide an account for pragmatic processes, we cannot merely rely on theoretical motivations and our own intuitions, but must have recourse to experimental methodology, which is what they do by reporting a case study on scalar implicatures as a case of voluntary narrowing, and by citing psycholinguistic evidence on metaphor as a case of imposed narrowing.
In the last, stand-alone essay of the volume, Friedrich Christoph Doerge is concerned with the notion of prolixity and aims to understand why people sometimes add linguistic material that seems unmotivated and unnecessary. He defends a broadly Gricean account against the explanation offered by Relevance Theory.
As can be seen from this survey, the essays that compose Brevity cover a wide range of issues at the interface of linguistics, philosophy and psychology, and exemplify a variety of approaches and methodologies. While each essay is worth reading on its own, as a whole, they form a fairly heterogeneous collection that does not tell us whether the family of phenomena that are somehow related to brevity all stem from some single underlying principle or process. What this volume does, though, is make us aware that there are still many unresolved issues that can only be accounted for through an interdisciplinary enterprise of the kind intended in this project.
