The interaction between the perfect and the preterite, in English, as well as in a broader typological perspective, has occupied researchers for a very long time, cementing its position as a staple research topic. There has been a long line of seminal studies on the English temporal-aspectual system touching upon this topic, among them Jespersen (Reference Jespersen1931), Mustanoja (Reference Mustanoja1960), Brunner (Reference Brunner1962), Visser (Reference Visser1963–73), Zimmermann (Reference Zimmermann1973), Dahl (Reference Dahl1985), Mitchell (Reference Mitchell1985), Rydén & Brorström (Reference Rydén and Brorström1987), Brinton (Reference Brinton1988), Traugott (Reference Traugott and Hogg1992), Denison (Reference Denison1993), Bybee et al. (Reference Bybee, Perkins and Pagliuca1994), Wischer (Reference Wischer, Jay, Horobin and Smith2004), Kilpiö (Reference Kilpiö, Rissanen, Hintikka, Kahlas-Tarkka and McConchie2007), Łęcki (Reference Łęcki2010), Macleod (Reference Macleod2012, Reference Macleod2014) and, most notably, Elsness (Reference Elsness1997), to name but a few. Indeed, at times the book under review reads like a fitting sequel to Elsness’ (Reference Elsness1997) pioneering volume. While Elsness outlines the big picture, Yao's monograph delves into more detail, in particular genre and register variation. Rather than taking more of a bird's eye view and tracing the developments from Old English onwards, it zooms in on the modern period, more specifically the past three centuries. The main objective is to shed light on the competition between the (simple/synthetic) past tense and the (analytic/periphrastic) present perfect in expressing past-time events. The present perfect seems to have been much less successful in English than in other European languages, including French and German, where the periphrastic forms have more or less ousted the synthetic ones in a process referred to as aoristic shift or drift.
To achieve the aforementioned aims, the author has carried out quantitative statistical analysis and qualitative philological inspection of specific examples and texts. I applaud such a healthy mix of research methods, in particular the fact that the latter, philological, approach has not been ignored here, though this might have been tempting to someone who is so well acquainted with the intricacies of cutting-edge corpus research. Quite the contrary: Yao describes this mix of methods as a combination of ‘the bird's eye view of the corpus linguist’ and ‘the jeweler's eye view of the philologist’ (p. 202). This happy marriage of the two paradigms can bring benefits which cannot be obtained with raw number crunching alone (important though it is), as I argue in some of my own publications (Hristov Reference Hristov2020, Reference Hristov2023).
As Yao points out on p. 5, when Elsness conducted his study back in 1997 the main quantitative tools were restricted to frequency counts and the significance values of chi-square tests. This toolkit has since been greatly expanded to include more advanced techniques which can incorporate multiple explanatory variables at the same time, capturing the complex relationships between those variables, as well as any emerging diachronic trends. The author resorts to binary logistic regression and random forests. In terms of theory, Yao's project likewise aims to bridge theoretical divides, drawing on formal and functional models alike. Unsurprisingly, Yao adopts a usage-based approach, where frequency plays a central part. This is in line with the corpus-based nature of the study. This approach likewise provides an appropriate setting for exploring the relation between usage, synchronic variation and diachronic change, as well as the interconnectedness between linguistic systems with non-linguistic cognitive systems, which is what the monograph sets out to do. Special attention is paid here to regional (British vs American English) and register variation.
Chapter 1 (pp. 1–17) serves as an introduction which states the theoretical affiliations and tackles important preliminaries, for instance methodology, providing a roadmap for the rest of the book. Chapter 2 (pp. 18–49) fleshes out the theoretical framework in which the examination of the past tense and the present perfect will be situated, especially in terms of their characteristic semantics and pragmatics. It draws on earlier descriptive and typological accounts, additionally appealing to more formal Reichenbachian tense–aspect representations, alongside neo-Gricean pragmatics in the form of relevance theory, bringing to the fore the role of pragmatic inference. Equipped with these conceptual tools, the chapter then reviews the major functions and uses of the two constructions, with an emphasis on the role played by pragmatic inference. Yao proposes a discourse-pragmatic principle in order to explain some of the distinctions and the familiar properties of the Modern English present perfect, namely an implicit search for optimal relevance in human interaction. Current relevance is found to be a semantic assertion in the case of the present perfect but a pragmatic inference in the case of the preterite. Firmly grounded in analysing naturally occurring language data, the book offers a sobering assessment of the (in)applicability of discrete theoretical divisions between what ought to be neatly pigeon-holed categories and the actual situation on the ground. After all, ((non-)linguistic) reality is often continuous rather than discrete, so it does not easily lend itself to neat pigeon-holing.
Even though the book focuses on modern and contemporary English, in chapter 3 (pp. 50–66) the English perfect is traced back to its Old English origins, enriching the description with more extensive historical background. The picture is made complete by cursorily charting the relevant changes in the intervening periods too. Yao reviews some of the earlier and more recent research into the history of the past tense and the present perfect in English, with both forms traced back to some of their earliest attestations. Having analysed the functional distinctions between them in Old English, Yao then goes on to examine the shifting patterns in their use, paying attention to different contexts. Some weaknesses besetting previous treatments of the distribution of the two forms since Late Modern English are also highlighted. Chapter 4 (pp. 67–88) adds a cross-linguistic dimension to the diachronic picture presented in chapter 3. It looks into typology and the grammaticalisation paths of perfects and preterites across a number of languages. (A small note is perhaps in order here: it was surprising to find Macedonian subsumed under the Baltic Rim region, p. 86.) An ‘oft-traversed’ path is the resultative > anterior > perfective/past traditionally proposed for have-perfects (as per Bybee et al.'s 1994 seminal work). Yao points out that this path is conditioned by a functional generalisation triggered by pragmatic factors, in particular a new interpretation supported by a new type of context, essentially what is known as pragmatic enrichment. Such paths are claimed to represent universal probabilistic tendencies in the evolution of grammar driven by cognitive and communicative factors operating across diverse speech communities. Quite importantly, divergences from these grammaticalisation paths are also flagged up here, in a reminder that things needn't develop deterministically in favour of the perfect at the expense of the preterite. In view of this, I wonder if the top line on p. 75 should be about the expansion of the preterite (rather than the present perfect), which makes more sense in this context (cf. further Fløgstad's Reference Fløgstad2016 title and Yao's discussion of Argentinian Spanish in chapter 8).
The following three chapters investigate the past tense and the present perfect in British and American English between 1750 and 1999. Chapter 5 (pp. 89–109) justifies the corpus selection procedure, as well as the methods for data extraction and processing. A health warning for readers not well versed in the science of statistics might be necessary at this point: such readers will probably have to do a lot of background reading on statistical methods in order to keep up with the exposition. We see the results of the author's distributional analysis, which is based on variation in terms of register, region and time. Yao compares the distribution of the two competing forms in two fifty-year intervals, 1750 to 1799 and then 1950 to 1999, with the aim of bringing to the fore potentially significant developments in the interim period. The chapter explores patterns of register variation relating them to usage situations.
The distinction between the semantically encoded nature of current relevance in the case of the present perfect, as opposed to it being a pragmatic inference in the case of the preterite, enables Yao to put forward an insightful explanation as to why fewer present perfects are found in actual spoken interaction than in speech-based written genres such as dramatic dialogue or in personal correspondence. Since interlocutors in conversation share their immediate context, they might be more likely to rely on pragmatic inference about relevance, rather than encoding it with a dedicated perfect form. Such a scenario sounds intuitively plausible, but then, as acknowledged on p. 107, why do some letter writers ‘decide’ that they need shorter, more compact forms like the preterite, instead of the present perfect? Without meaning to sound defeatist, I would be tempted to ask: Might this be a sign that language change could proceed in a more random fashion than we as researchers would like it to do (as emerged when I was working on my 2020 book too), a possibility which Yao entertains when discussing the outcome of competition in different languages?
My impression that things can be messier and more arbitrary than expected is confirmed by the statistically significant factors that influence the choice of one form or the other. On pp. 119–22, the author tells us that, compared to the eighteenth-century sample under study, the significant predictors of register, telicity and transitivity are later replaced by temporal specification and negation. The decline of transitivity as a conditioning factor probably has to do with the demise of be as a perfect auxiliary (p. 126). (Yao doesn't examine be-perfects or include them in the same perfect category as those with have.) So, while there are viable explanations for some of these trends, others are harder to account for and remain more recalcitrant, for instance negation. Why should their significance and contribution change over time? The charts on pp. 143–5 also testify to different variables playing different roles in different speech communities across space and time. This reinforces the feeling that change operates blindly and can head in random directions (it might be instructive to compare Keller's Reference Keller1990 ground-breaking ideas about the invisible hand). In the same vein, on p. 149, Yao reports that there has been a tightening connection between perfect and progressive in American, but not in British, English.
Chapter 6 (pp. 110–35) follows the major paths in the development of the past tense and the present perfect by employing multivariate analysis, more specifically mixed-effect logistic regression. The author presents both quantitative results as well as qualitative interpretations. The choice between the two forms is found to be constrained by seven main predictors – namely temporal specification, telicity, transitivity, negation, subject type, clause type and register, including the interactions between them. Yao uncovers pathways of diachronic change and regional variation by comparing the directions and degrees of influence between the two fifty-year periods, as well as between British and American English.
Chapter 7 (pp. 136–61) moves the multivariate analysis forward, applying a series of random forest modelling algorithms, essentially a machine-learning technique performed on the same corpus data. This allows Yao to cross-tabulate the results from the two experiments implementing multivariate methods. Attention is paid here to the influence of the progressive aspect as well as a wide range of temporal adverbials – those would not have easily lent themselves to parametric regression due to limitations in the available corpus material. With this additional data, a more fine-grained picture of variation and change takes shape. Although the results mostly converge, it might be seen as a cause for concern that the two statistical procedures sometimes yield divergent results regarding the same predictor (e.g. register), as conceded on p. 145.
Connecting the empirical findings from the previous chapters with theoretical assumptions about universals of grammatical change, chapter 8 (pp. 162–94) returns to the motivations behind the observed developments from a typological cross-linguistic perspective. In addition to the previously established internal mechanisms, largely due to cognitive and communicative processes, Yao pays attention to external motivations for language change, more specifically language contact (cf. Drinka's Reference Drinka2017 influential contribution). Another factor is brought to the reader's attention too, this time a language-specific one, namely the degree of formal similarity in the inflectional paradigm of English. The author has extracted verbal data from the Corpus of Historical American English (COHA), testing for a potential correlation between formal similarity (i.e. conflation of perfect participial and preterite forms) and any prospective frequency changes. The conclusion is that such language-internal factors cannot have been decisive on their own. Therefore, the chapter further explores the role of language contact as a more viable external motivator for change. This chimes in with arguments in the specialist literature that ambiguous formal marking or syncretism is not enough to trigger change on its own (Hristov Reference Hristov2020, Reference Hristov2023). The author detects parallel diachronic trends in the use of the preterite and present perfect in English and Spanish, more specifically American English and Argentinian Spanish. Yao's story about colonial varieties and the processes at work in them is well-crafted and sounds convincing. This novel comparative perspective is illuminating and is likely to prompt further scholarly inquiry.
Chapter 9 (pp. 195–204) sums up the main findings, underscoring their broader theoretical and methodological implications and pointing to some possible avenues for future research. The shifts between preterite and perfect that Yao has been investigating are situated in the larger context of the cyclical typological oscillation between synthetic and analytic in the history of English.
Overall, the book is characterised by succinctness, clarity of style and perspicuity of the exposition, for instance the fairly detailed user-friendly index, or brief but cogent explanations when it comes to navigating the terminological morass associated with matters such as lexical aspect, Aktionsart or situation type, the term which the author ultimately adopts in order to avoid confusion (p. 21). The following statement on the same page remains somewhat mysterious however: ‘Melanie walked her dog this morning expresses the situation of Melanie walking her dog in a narrow sense and the situation of Melanie walking her dog this morning in a broad sense’ – without any further comment, the uninitiated reader will be none the wiser about the proposed distinction between the broad and the narrow senses. In a slightly misguided passage, Yao asserts that there are two objects in complex transitive clauses like They made him a hero (p. 43). However, the NP a hero is not an object here but an object complement (i.e. an object-related predicative complement), which can be realised by an adjective phrase, as in the examples on p. 42, and which cannot be promoted to subject in the passive, as in He has been made a hero, but not *A hero has been made him. Regarding the discussion of data processing on p. 96, due to the phrasing, I am not sure if ‘coordinate sentences in which the coded verb is preceded by a verb in perfect form’ were included or excluded. Unlike Yao, I find strings like He has had to leave perfectly acceptable.
Traditionally at least, the present perfect in Many people have been taken hostage in the Iraq war (p. 43) would not be deemed acceptable or pragmatically felicitous with an adverbial expression referring to a past event, blocking current relevance (see Palmer's Reference Palmer1974: 49ff. impeccable argumentation, a title that would undoubtedly have enriched Yao's bibliography). Even though there are noteworthy exceptions to this time-honoured generalisation, readers might have expected more discussion of these issues. A cross-reference could have been included here to section 2.2.5 ‘Co-occurrence with temporal adverbials’. Yao does consider these exceptions, making mention of the so-called ‘lifetime effect’ on p. 61, but does not dwell on them very much. Thus, despite the perceptive analyses, I sometimes was not absolutely convinced by Yao's take on certain examples. Another such instance is the following interrogative on p. 44, which is arguably more likely to be construed as a resultative than an experiential: Who has taken them?. The same applies to the other analogous examples there.
Like any human endeavour, this book too is not devoid of small slips or the occasional typo. For instance, there is a typo on p. 54, ex. (8), where Old English clætne should be clæne ‘completely’. There is the odd typo or small linguistic slip throughout the exposition, which is inevitable in a publication of this size, e.g. p. 160: ‘The findings for the progressive aspect has revealed…’ (bold face mine). In the table on p. 168, German bin ‘am’ should be listed as a form of sein ‘be’, and the Canarian Islands are more widely known as the Canary Islands. And yet, all things considered, such lesser issues are only minor shortcomings which are rare in the text, so they do not impede comprehension or detract from the intellectually stimulating reading experience in any substantial way.
In conclusion, this excellent book about the English perfect and preterite is an exemplary study that skilfully combines classical philological techniques with cutting-edge corpus linguistics and statistical methods, so it comes highly recommended.
