2 Part I: Delineating the ROC

As mentioned, Bouso anchors her study of the English ROC in the context of English having undergone a process of ‘transitivisation’, viz. having moved towards a greater reliance on transitive structure over time. More specifically phrased in Construction Grammar terms, such movements constitute valency-increasing argument structure changes (or changes of argument augmentation), adding an extra argument to originally less transitive/intransitive patterns (p. 16). While this and other constructionist notions are taken up more explicitly in later sections, chapter 2 first traces the research history on transitivisation and gives an overview of (in)transitivity as discussed in diachronic and synchronic sources (section 2.1). Precisely, Bouso here compares Visser’s (Reference Visser1963–73) classification of intransitive patterns in Old English (OE) to Liu’s (Reference Liu2008) taxonomy for Present-day English (PDE). This comparison indicates that the number of verb types only used intransitively (e.g. arrive) has decreased considerably between OE and PDE, while verbs which can be used in either an intransitive or a transitive structure (referred to as ‘amphibious’, ‘double-functioned’ or ‘labile’) have increased (e.g. The wood burns vs They burnt the wood). Section 2.2 further corroborates the trend by using more detailed figures provided in a range of earlier studies on amphibious verbs. The causes for this development are explored in section 2.3, which briefly reviews the loss of prefixation and case marking, changes in prepositional patterns, certain ambiguities surrounding past participles and a growing popularity in sound and gesture verbs in LModE as potential factors in the shift.

Chapter 3 similarly starts by consulting (descriptive) reference grammars on historical and contemporary English, in particular Jespersen and Visser for the former, and Quirk et al. (Reference Quirk, Greenbaum, Leech and Svartvik1985) as well as Huddleston & Pullum et al. (Reference Huddleston and Pullum2002) for the latter (section 3.1). This time, the sources are compared in terms of their treatment and classification of ‘objects’, with the main intention of establishing prototypical and non-prototypical features of objects across periods, and positioning the ROCs (and related patterns) within these categories. For example, both reaction objects and way-objects feature as special cases in Jespersen’s (Reference Jespersen1909–49) group of ‘objects of results’ (contrasted with ‘ordinary’ objects), whereas they are strikingly absent from Visser’s taxonomy. Likewise, ROCs are not explicitly mentioned in all PDE grammars, which is taken as an indication that the phenomenon may be rather recent and over all relatively rare. Section 3.2 then introduces ROCs in more detail, and approaches the pattern as well as cognate objects and way-objects in the context of (non-)prototypical object features discussed in earlier literature, such as topicalisation or the ability to be questioned (see pp. 90–1 for specific constraints on the ROC). This analysis ultimately substantiates that all three object types ‘exhibit a series of morphosyntactic features that clearly set them apart from the prototypical objects’ (p. 97). Furthermore, the reviewed historical evidence suggests that the ROC – like the other patterns – may have expanded from transitive uses to intransitive verbs in a gradual analogical extension process.

The last chapter of this part finally sees an introduction to (usage-based) Construction Grammar, laying out the main principles of the framework (sections 4.1 and 4.3), as well as giving a short survey of the history of the field (section 4.2) – the strong focus on idiosyncratic, idiomatic phenomena in the earliest constructionist accounts matches well with Bouso’s interest in the ROC. Perhaps most importantly for the present study, section 4.4 outlines constructionist views on language change, including (lexical vs grammatical) constructionalisation, which refers to the ‘creation of a form_new meaning_new pairing’, as well as (pre- or post-constructionalisation) constructional changes (Traugott & Trousdale Reference Traugott and Trousdale2013). In addition, Bouso here briefly reports on changes in constructional networks and changes in the links between constructions. This section concludes the setting-the-scene part of the book, providing the backdrop for the empirical study in part II.

3 Part II: Investigating the ROC

Part II is again divided into three main chapters, the first of which consolidates the background knowledge and especially the constructionist notions from part I with further meta-evidence from previous sources (chapter 5). This chapter, titled ‘The formation of ROCs’, starts by characterising the construction in PDE, zooming in on formal and semantic aspects of the pattern, and discussing the verbs and object types occurring in it (section 5.1). For instance, it is stated that the ROC most typically appears with manner-of-speaking verbs (e.g. mutter, murmur), or verbs of gestures and signs (e.g. wink, smile, nod). The objects, denoting ‘expressive speech acts resulting from a gesture or sound performed by the subject of the construction, who expresses a mental state towards an antecedent event’ (p. 130, quoting Martínez-Vázquez Reference Martínez-Vázquez2015: 152), can be distinguished into delocutive nouns (welcome, goodbye), deverbal illocutionary nouns (assent, acquiescence) and predicative expressive nouns, either positive (delight, joy) or negative (disdain, discontent). Other frequent properties of ROCs include premodification of the object by a possessive (smiled their approval) and the presence of an indirect object (waved them goodbye). These features, among other constraints, are finally linked to other English constructions such as the ditransitive/communicative (gave them a book, told them a story), the experiencer construction (They like us) and the resultative construction (kissed them unconscious), modelling the ROC as a hybrid pattern characterised by multiple inheritance in a complex network (see the full sketch on p. 148). Section 5.2 moves from the synchronic to the diachronic account. Again using the previous reference works of Visser (Reference Visser1963–73), Jespersen (Reference Jespersen1909–49) as well as Levin (Reference Levin1993), among others, as sources of data, Bouso establishes a list of verbs and examples qualifying as instances of ROCs in historical texts together with their dates of first attestation. In addition to providing a wealth of textual evidence, this allows her to locate the construction’s peak in LModE, which strikingly overlaps with the dates of emergence and spread of the related transitivising cognate object and way-constructions, as also nicely visualised on pp. 201–2.

Chapters 6 and 7 report on the corpus studies conducted as part of the project, with the former concentrating on Late Modern British English and the latter on (later) Late Modern American English. For British English, Bouso uses data from 1710–1920 in the multi-genre Corpus of Late Modern English Texts, version 3.0 (CLMET3.0, De Smet et al. Reference De Smet, Diller and Tyrkkö2011), extracting instances of forty potential ROC verbs and manually pruning the hits for relevant instances. This procedure must have involved massive effort, as from the initial number of almost 45,000 tokens of forty verbs, only about 300 instances of twenty-six verbs remained after pruning, again also highlighting the overall rare status of this phenomenon. The final dataset was then analysed in terms of frequency distributions across time, genres and text types, as well as object types and verb types/classes (corroborated by means of collostructional analysis, Stefanowitsch & Gries Reference Stefanowitsch and Gries2003). In a nutshell, the results show that the ROC increased from the mid-eighteenth century onwards and had its heyday around the turn of the nineteenth century (up to 1850); in terms of genre, it is closely associated with narrative fiction, drama and poetry, and can be linked to the rise of the novel as a literary text type around this time, as well as to an increase in the popularity of sound verbs in such texts. As for verb types, the study identifies five prototypical ROC verbs, of which two instantiate manner-of-speaking verbs (murmur, mutter) and three verbs of nonverbal communication (smile, nod, wave). An in-depth look into the cooccurrence patterns of these verbs and particular syntactic features of the objects, such as absence/presence of an indefinite article, leads Bouso to conclude that the subconstruction expressing manner of speaking is represented on a more abstract, more schematic level, while nonverbal communication involves more lexical constraints on a lower level (viz. specific verbs being associated with specific object types). The development of the ROC is furthermore interpreted in terms of constructionalisation and constructional changes, arguing that while precursors of the ROC existed in Old and Middle English, the pattern constructionalised only from the sixteenth century onwards (Early Modern English), and saw various post-constructionalisation constructional changes in LModE (see the elaborate overview on p. 267).

Chapter 7 gives a similar analysis of ROCs in American English, making use of the Corpus of Historical American English (COHA, Davies Reference Davies2010–), which covers the time frame from 1810 to the 2000s. In this case, the verb slot is left open but the noun element is restricted to delocutive nouns (hello, goodbye and thank you) in the retrieval, ultimately resulting in a total of eighty verb types across almost 1,000 tokens. The findings suggest that, as expected, the ROC in this variety spreads later than in British English, in line with the American novel tradition also developing later. Here, the way-construction seems to differ from the ROC, as the former proliferates more in American English than in the British variety. Despite the overall low frequency of the ROC-pattern in American English and only a slight increase over time, the chapter shows that the construction experiences a rise in productivity, with the most productive class being verbs of sound emission. A very useful and detailed graphic outline of the development of the ROC in constructionist terms, extending to Present-day American English, and including productivity changes as further post-constructionalisation constructional changes, is given on p. 306, compared to a similar sketch for the way-construction on p. 305.

Chapter 8 sums up the book and its main findings, traces the trajectory of the ROC and points out desiderata for future research, including some shortcomings of the dataset and historical data in general.

4 Wrap-up: objectness, transitivisation and the ROC

In sum, the reviewed book presents an excellent read for anyone who has ever wondered about odd structures like smiling hello, waving goodbye or nodding agreement (and also for those who never have, but will surely encounter them everywhere now). It is furthermore a must-read for researchers interested in careful implementations of (diachronic) Construction Grammar for specific case studies, and a prime example of how to conduct corpus studies on somewhat elusive patterns in an as bottom-up, exploratory way as possible. The study clearly achieves its goal of providing a meticulous and convincing application of by now rather standard concepts in Construction Grammar to a particular, to-date unexplored phenomenon. On a more methodological level, Bouso nicely emphasises and also demonstrates the value of older descriptive sources (above all on historical English), which, despite being compiled in pre-corpus times, can still provide important and useful insights about a construction’s development if used as diligently as done here. At the same time, Bouso is also acutely aware of what insights her study can or cannot provide, which is showcased by her reflections on the representativeness of the data sources used, specifically the Oxford English Dictionary (OED) and previous studies (e.g. p. 195).

Last, Bouso’s book also inspires a number of interesting follow-up points especially in light of more recent (post-publication) research. On the one hand this relates to definitions of ‘transitivity’ (Næss Reference Næss2007) and ways of measuring such (e.g. Guajardo Reference Guajardo2023), as well as to testing the validity of the presumed transitivisation shift as a whole. For example, while the trade-off between purely intransitive and labile verbs is assessed in detail in Bouso’s study, possible de-transitivisation trends may be given as well. Here, some examples given in section 2.3 triggered my curiosity – verbs such as complain or marvel, which seem to have developed transitive uses in the timeframe investigated in Bouso (p. 54), in fact seem to have lost this ability again in later times (OED, s.v. complain, marvel). This raises the question of whether at least some transitivisation subtrends may only have been temporary, or whether the entire shift is perhaps not entirely linear (cf. Szmrecsanyi Reference Szmrecsanyi and van Gelderen2016 for a similar point about the move from synthetic to analytic). On the other hand, revisiting Bouso’s proposed scenario in view of current debates on the distinction between constructionalisation and constructional changes, included as one of the major open questions in diachronic Construction Grammar in Hilpert (Reference Hilpert, Coussé, Andersson and Olofsson2018), among others, would likely yield further insights into the phenomenon and also the further refinement and development of these concepts (cf. also later publications by the author such as Bouso Reference Bouso2022a, Reference Bouso2022b, where this is done more explicitly).

Overall, then, this monograph constitutes a great contribution to the growing body of (diachronic) Construction Grammar, and a model illustration of how constructionist thinking can be used for the characterisation and tracking of peculiar but thought-provoking linguistic patterns.