This monograph introduces three types of discursive connector relationships and an approach to Construction Grammar that may not be very familiar to researchers in English linguistics in the UK and US, but deserves attention. Various models of Construction Grammar have been proposed in the last forty years or so, among them the model used in this volume: the Lexical Constructional Model outlined in Mairal Usón & Ruiz de Mendoza Ibáñez (Reference Mairal Usón, Ruiz de Mendoza Ibáñez, Butler and Martín Arista2009). This model seeks to integrate aspects of several types of grammar, among them Functional Grammar (e.g. Dik Reference Dik1997; Dik & Hengeveld Reference Dik and Hengeveld1997), natural semantic metalanguage (e.g. Goddard & Wierzbicka Reference Goddard and Wierzbicka2002), cognitive semantics (e.g. Lakoff & Johnson Reference Lakoff and Johnson1999) and Cognitive Construction Grammar (e.g. Goldberg Reference Goldberg1995, Reference Goldberg2006). Four levels of linguistic organization are distinguished in the Lexical Constructional Model: argumental, pragmatic-implicational, illocutive and discursive. As Iza Erviti says, the Lexical Constructional Model is ‘highly ambitious’ (p. 25). She seeks to enrich the fourth discursive component by developing a systematic account of constructions at the discourse level, with emphasis on discourse relationships that convey an element of contrast like X, however Y, X never mind Y and the ‘family’ resemblances among them. A ‘family’ of constructions is unified by a ‘functional (or conceptual) space onto which given constructions can be mapped (Kiss & Alexiadou Reference Kiss and Alexiadou2015)’.
Citing Ruiz de Mendoza Ibáñez (Reference Ruiz de Mendoza Ibáñez, Diedrichsen and Nolan2013: 238), ‘construction’ is defined as ‘a conceptually entrenched (i.e. frequent and straightforwardly replicable) form–meaning association where form is realizational of conceptual representation in any degree of complexity’ (p. 2). Not much is said in the book, however, about form. The discourse relationships include what are often called Discourse Markers (DMs) (e.g. and, but, or), but these are not the center of attention. It is mostly coordinate relationships such as X but Y and an inventory of their realizations that are the topic of the book. So the crucial difference from approaches such as Fraser's (e.g. Reference Fraser1996) is that the focus is not on the marker itself and its function but on the function of the co-text. While Fraser's ‘segment 1’ and ‘segment 2’ are not distinguished for content, Iza Erviti's X and Y are characterized in different ways, depending on the relationship they realize.
Chapter 1 is a brief introduction to the goals of the study and the perspective adopted.
Chapter 2 is rather abstract, with few examples, leaving some questions unanswered, e.g. what are examples of ‘internal and external constraints’ (p. 14)? It is in two main parts. The first part is a survey of approaches to connectivity, including Rhetorical Structure Theory (e.g. Mann & Thompson Reference Mann and Thompson1988) and Relevance Theory (Blakemore Reference Blakemore1987), and provides extensive bibliographic references, especially to Ruiz de Mendoza Ibáñez’ work. Iza Erviti proposes (p. 9) that DMs should be understood as discourse connectives that have content of a ‘high-level, relational nature’ (p. 9). DMs are said to refer, in their broadest sense, ‘to any linguistic mechanism, independently of its syntactic status, used to link two or more predications into a more complex conceptual package expressing conceptual relations’ (p. 9). They are analyzed qualitatively, with focus on ‘potential replicability’. The analyst's aim ‘is to ascertain the licensing factors for such a construction in its context of production’ (p. 10).
The second part of chapter 2 provides an account of the most significant features of the Lexical Constructional Model adopted. It is a functionalist usage-based ‘comprehensive model of meaning construction through language in context’ (p. 23). It explores ‘how communicative activity impinges on the linguistic form by investigating its cognitive grounding’ (p. 25). And it offers an inventory of constructions at the discourse level along with a detailed description of their role in discourse structure. A preliminary list of discourse relations developed by Ruiz de Mendoza Ibáñez & Gómez-González (Reference Ruiz de Mendoza Ibáñez, Ángeles Gómez-González, González, Ruiz de Mendoza Ibáñez and Gonzálvez-García2014) is provided on p. 29. This list is said to be ‘the first attempt to clarify discourse constructions according to the meaning relation they generate’ (p. 28).
Chapter 3 is on methodology and data. The corpus used was specially compiled largely from the British National Corpus and the Corpus of Contemporary American English (via the formerly BYU corpora interface), dictionaries like the Collins COBUILD Dictionary and the Merriam-Webster Dictionary Online, and selected Google searches. This means the data used were both UK and US data. It would have been interesting to know whether any differences between British and American usage emerged with respect to any of the connective relationships discussed. Searches were conducted for three types of construction: (i) ‘complementary alternations’ such as neither X nor Y, X let alone Y, in which the X and Y alternates are not mutually exclusive; (ii) ‘complementary contrastives’, which are opposites, but not exclusive of each other (more dead than alive); and (iii) ‘contrasts’, in which the alternates are exclusive (either you win or you lose). Meanings were characterized according to Langacker's (Reference Langacker1987, Reference Langacker1991) concepts of base, profile, and active zones.
Chapter 4 details the essential features of a family of nineteen ‘complementary alternation’ constructions. In the configurations in this chapter, X and Y ‘are two different states of affairs such that Y adds to X on the basis of a subjective speaker's judgment’ (p. 44). Four subtypes are discussed: neutral constructions (Neither X nor Y), understatement constructions (X not to say Y), expanding constructions ((Not) X not to mention Y) and condensing constructions (X in fact Y). These are not rigidly separate subtypes; indeed, X never mind Y is classified as both neutral and expanding (highlighting and expressing what is obvious from the speaker's point of view). Among expanding constructions is X let alone Y, which was analyzed at length in Fillmore et al. (Reference Fillmore, Kay and O'Connor1988). Iza Erviti proposes, without much argument, that both parts of the construction are equally informative, but one is presented as less likely to occur than the other. She concludes that Fillmore et al.'s analysis, which focuses on relative prominence of the focused elements and redirection of the addressee to a new, more informative proposition Y, ‘is not entirely correct’ (p. 65). The chapter ends with an illuminating discussion of the distinction between X even less Y and X much less Y. In X even less Y, ‘the speaker presupposes that the hearer thinks that Y is the case’, but not in X much less Y. The examples (p. 71), reproduced here as (1) and (2), illustrate this well. In (1) the speaker presupposes that Augier's daughter knew at least something about her business. This is not the case in (2).
(1) Augier's daughter knew nothing about art, even less about her business. (COCA)
(2) Augier's daughter knew nothing about art, much less about her business. (constructed)
Chapter 5 investigates a family of ‘complementary contrastive discourse constructions’. These convey a contrastive meaning between X and Y and ‘also contain an additive value’ (p. 75). This additive value appears to be strengthening of the profile of Y. A list of thirty-nine discourse constructions is introduced and then discussed in terms of seven subgroups: neutral (X on the other hand Y), concessive (X although Y), correcting (X anyhow Y), topic changing (X be that as it may Y), topic avoiding (X never mind Y), refusal-apology (X all the same Y) and evaluative (not so much X as Y). X but Y is cited in each of these subgroups. It is said to have a very generic meaning and to be schematic in the sense that it sketches out a non-content relation between two constructional variables. Making use of a high-level metonymy generic for specific, and contextual cues, ‘but is adjusted to each of the seven specific meaning dimensions complementary contrastive constructions can profile’ (p. 79, original italics). Whether Iza Erviti hypothesizes that it is the speaker, the addressee, the context or the marker that does the adjusting is not specified.
The concessive relation exemplified by X although Y, X all the same Y, X still Y, X while Y, is far more restrictive than the X but Y relation. It is one of ‘partial opposition’ on a contrast continuum. Two main types are noted, both indicating that, from the speaker's perspective, what matters is Y (p. 86): (i) those that convey the idea that a state of affairs holds despite opposition (against her will, she glanced down), and (ii) X and Y hold, but Y is more relevant (I hate eating fish. However, I really enjoyed your mum's cod).
This brings us to chapter 6, on a family of thirty-eight contrast constructions. Contrast is defined as ‘the act of distinguishing or of being distinguished by comparison of unlike or opposite qualities’ (p. 121). ‘The act’ suggests a speech act orientation, but Iza Erviti appears to envisage constructions, not speakers, as acting (see p. 75). Contrast constructions are divided into four subfamilies: contraposition (X whereas Y), used to express at least one relevant opposition; exception (There's no X like Y), used to express an exception to the characteristics of another state of affairs; alternative contrastive constructions (X or Y); and disagreement constructions (disagreeing with X, Y).
Chapter 7 provides a useful summary of the main findings of the book. These are to identify constructional families that are operational at the discourse level of the Lexical Constructional Model (LCM), and to make ‘the first contribution to the creation of a fully-fledged Constructicon consistent with the descriptive and explanatory mechanisms of the LCM’ (p. 155, original italics). The main characteristics of each of the three families of constructions discussed in chapters 4–6 are outlined. Iza Erviti sees identification of the complementary contrastive constructions family as an especially important contribution to the study of language because these constructions combine contrastive and additive operations in one form. They also allow for operations such as highlighting, reinstatement and broadening. The chapter ends with a call for research into the relationship between the constructional families she identifies and others such as addition and comparison.
The inventory of relational constructions and their subtle meaning differences in this study is highly informative about almost 100 discourse relations. It will be of great interest to lexicographers, not only compilers of dictionaries like the Collins COBUILD English Dictionary, the Merriam-Webster Dictionary Online and Thesaurus.com that Iza Erviti says sometimes do not list particular subtypes of relationships she identifies, e.g. ‘expanding scenarios’ that add something obvious, as in (3):
(3) But will finding her tomb, not to say her body itself, deepen our portrait …? (COCA, p. 55)
The book will most especially be useful to compilers of thesauruses, given the focus on families of constructions. The theoretical linguistic value lies in the range of factors that the Lexical Constructional Model encompasses, especially its similarity in aim to Frame Semantics, which is alluded to several times. There will also be special interest in Iza Erviti's contribution to the constructional lexicon, or ‘constructicon’. If the constructicon is conceived not as an unstructured list, but as a ‘network’ (Goldberg Reference Goldberg2003: 219), the connections with other families that Iza Erviti calls for will be essential.
The usefulness of the book to both lexicographers and constructionalist linguists would be greatly enhanced by clearer explanation of why a particular interpretation is adopted, most especially whether the construction or the content of X or Y is driving the analysis (or both). Sometimes there is inconsistency. For example, Iza Erviti explicitly excludes such expressions as in conflict with from the contrast family because they introduce prepositions, not propositions (p. 121). It is not clear why X distinct/different from Y and other members of the subfamilies such as opposing X, Y, which ‘adopts the formal aspect of a prepositional phrase’ (p. 126), are not excluded on the same grounds. There are several valuable relatively abstract figures, e.g. figure 4.4 illustrating domain reduction, highlighting, abstraction and strengthening activated by X never mind Y. But X leave alone Y, which is said to be based on the metaphor ‘emotional distance is physical distance (i.e. we leave behind what is not dear to us)’ (p. 66) is illustrated with a somewhat representational image of a couple standing back to back, with a split heart between them. It is not clear exactly how the metaphor and its representation fit examples cited like:
(4) this man took her back to her family without any mention, leave alone discussion, of the dilemma which confronted her (BNC-BYU) (p. 67, italics added)
What is dear to whom in (4)? A possible alternative analysis might be that users of X leave alone Y analogize it to X let alone Y using a kind of ‘folk etymology’ that renders component parts of a fixed expression more comprehensible.
Sometimes there is insufficient discussion of which interpretation of a term is being adopted. For example, Iza Erviti endorses the Lexical Constructional Model definition of discourse constructions as ‘idiomatic form–meaning pairings’ (p. 28) without discussing the ambiguity of the term ‘idiomatic’. She appears to understand the term to mean ‘conventional’ when she comments that Excuse me/I'm sorry but Y is ‘a highly idiomatic construction with a very wide active zone’ (p. 115). This is a usage often found in the constructionalist literature. For example, Goldberg (Reference Goldberg2006: 13) says ‘[i]t is much more idiomatic to say I like lima beans than it would be to say Lima beans please me’. But Goldberg also uses ‘idiomatic’ to mean ‘non-compositional’, for example when she cites kick the bucket ‘die’ or give someone the eye ‘to look seductively at someone’ (Goldberg Reference Goldberg2006: 137, fn. 4). Such non-compositional idioms often need to be translated. Some of Iza Erviti's interpretations of connective relationships raise the question whether a non-compositional connective is being interpreted as a compositional one with an older meaning. A case in point is concessive X all the same Y, which is said to mean ‘all things being equal at this time’ (p. 97). Although historically all the same originates in a literal comparative expression of identity or similarity (Traugott Reference Traugott2022: 132–3), there is no evidence that it is currently compositional when it is used as a concessive, or that equality or sameness are understood to be part of the active zone of this use. The lexical approach to meaning deserves to be nuanced with a more pragmatic approach.
Reservations of this kind aside, Discourse Constructions in English enriches our understanding of Lexical Constructional Model and discourse connectivity. It will inspire much further work on nuances among licensing factors for connective constructions.
