The present perfect is often described as the most challenging of the English verb forms, from both a theoretical and a pedagogical viewpoint. Part of the challenge is to draw up a satisfactory functional line of division between this verb form and its chief rival, the preterite. It is generally acknowledged that the preterite is numerically predominant in most types of text, but it is also frequently pointed out that there is considerable geographical variation, not least between the two major varieties, British and American English (BrE and AmE), the present perfect being somewhat more frequent in BrE. Today the English present perfect may seem to be declining further. This sets English apart from many other languages: especially in speech the present perfect appears to be in the process of ousting the preterite in languages such as German and French.

The continuing interest in the English present perfect as an academic topic is confirmed by the fact that when the 35th ICAME conference was held at the University of Nottingham in 2014, a pre-conference workshop was devoted to this topic. The book under review is the result of that workshop. It is organised into three parts, after an initial survey chapter by the three editors: part I, ‘Diachronic and synchronic perspectives on the perfect in native varieties of English’; part II, ‘Perfects across varieties of English’; and part III, ‘Building bridges’.

Part I begins with a chapter by Berit Johannsen, entitled ‘From possessive-resultative to perfect? Re-assessing the meaning of [hæbb- + past participle] constructions in Old English prose’. Johannsen's research is based on the York–Toronto–Helsinki Parsed Corpus of Old English Prose. In her search for the likely origin of the perfect construction she lists four alternative constructions with post-object past participles: Adnominal (including postposed participle), as in Mary has a shirt made in China; Attained State, as in Mary has her opponent cornered; Affectee, as in Mary had a rock thrown at her; and Causative, as in Mary had the papers graded by an assistant. The fifth possible use of [hæbb- + past participle] is recognised as the Perfect, which Johanssen divides into the four uses distinguished by Comrie (Reference Comrie1976) and adopted by quite a few subsequent writers, sometimes under different names: result, experiential, persistent situation and recent past. A total of 92 per cent of her constructions are classified as perfects and a further 5 per cent as ambiguous between perfect and attained state. Johannsen argues that it is therefore unlikely that the perfect evolved from the adnominal construction, with hæbb- acting as a lexical verb meaning ‘possess’, ‘hold’, as has often been assumed. Because of the much larger number of constructions she recognised as expressing the attained state meaning, or more often as being ambiguous between that and the perfect, she concludes that a semantic shift from attained state to perfect is a more likely source. However, Johanssen herself sounds a necessary note of caution: co-existence does not prove semantic shift from one pattern to the other. A strict division between the different constructions may in any case be dubious.

The second chapter, ‘The to-infinitival perfect: A study of decline’ by Jill Bowie and Sean Wallis, is by far the longest chapter. Since it does not focus on the present perfect, it renders the title of the book slightly misleading, but it is definitely worthy of inclusion. It reports an investigation into Mark Davies's Corpus of Historical American English, which covers the last 200 years. Noting that the to-infinitival perfect mirrors both the preterite and the present perfect – broadly, it may express both specified past time and past time which is vague or extends up to the present moment – the authors choose as their baseline for comparison the combined frequencies of those two verb forms (thus disregarding all the other ways in which reference to past time may be expressed). An overall decline of some 80 per cent in the frequency of the to-infinitival perfect is recorded. Examination of the British component of the International Corpus of English, ICE-GB, revealed that a clear majority of to-infinitival perfects functioned as complements of verbs, mostly with the infinitival perfect immediately after the verb. The major part of their investigation is then focused on such constructions. The authors find that the 30 most frequent governing verbs account for 95 per cent of occurrences. All but 5 of these verbs are classified into four groups based on meaning and grammatical properties: ‘seeming’, ‘cognition and saying’, ‘modality’ and ‘prospective meaning’. The first group is the largest, the verb seem itself accounting for as much as 39 per cent. Throughout they provide detailed statistical data to underpin their analyses but they have rather less to say about likely causes of the results recorded and possible alternant constructions. They point out that ‘there is no clear limit to the set of possibilities’ (p. 70), adding that ‘investigation of them all is beyond the scope of the current paper’.

Chapter 3 is ‘Expression of the perfect in two contact varieties of English’ by Markku Filppula, where the two contact varieties addressed are Irish (IrE) and Hebridean English (HebE). Both are considered high-contact varieties, having emerged as the result of a language shift from the indigenous Celtic languages. Filppula reports that in both varieties meanings which in Standard English are expressed by the present perfect often take other forms. In particular, the preterite is used to express what he calls the indefinite anterior perfect, as in That never happened in this world yet. Now, the use of the preterite for the perfect is a common phenomenon even in other varieties of English, not least in AmE, as Filppula himself points out, but this usage appears to be especially frequent in IrE and HebE. A more unique feature is the use of the present tense (simple or progressive), accompanied by appropriate adverbial specification, to express what in Standard English would be the extended-now perfect. An example is They're fourteen or fifteen years married now. Even more special is the ‘after-perfect’, in which case what is taken as the typical perfect meaning of ‘hot news’ is expressed as in You're after ruining me. All these expressions are ascribed to influence from the Celtic substrate languages. Filppula could have made some of his definitions clearer. For example, he might have made it plain from the outset that for him the terms indefinite anterior perfect and extended-now perfect do not refer to different meanings of the present perfect verb form but rather (somewhat counterintuitively) to cases where similar meanings are expressed by other verb forms.

In chapter 4 Sophie Richard and Celeste Rodríguez Louro look at ‘Narrative-embedded variation and change: The sociolinguistics of the Australian English narrative present perfect’. They take as their starting-point the observation that the English present perfect does not generally serve as a narrative tense. Other researchers had already found that certain variants of Australian English may constitute an exception. The chapter reports findings in a corpus of ‘original talk-in-interaction data’ made up of 220 ‘strictly defined Labovian narratives’ (p. 128), analysed according to the classical Labovian model. The authors set out to study tense alternation in what they take to be the most typical narrative cases. In a variable rule analysis, priming, in the form of the tense of the preceding complicating action clause, turns out to be the weightiest linguistic conditioning factor. The present perfect is particularly common with the quotative verb go (as in I ’ve just gone, ‘Ah, ok’). Various social factors also play a significant part: users of the narrative present perfect tend to be male, less well educated and belong to the older generation. Hence the authors conclude that the narrative present perfect hardly represents a far-reaching change in progress but is instead socially as well as grammatically constrained.

The second major part of this volume begins with chapter 5, which is ‘Present perfect and past tense in Black South African English’ by Bertus van Rooy. The author compares three registers of Black South African English, defined as a non-native variety of English used by speakers of African languages. The three registers are face-to-face conversation, student writing and news reportage, which are taken to represent different kinds of text production circumstances: conversation is unmonitored, spontaneous language production, student writing represents a typical form of text production by non-native learners with texts that are not edited by another party, while news reportage usually goes through an independent editing phase. His corpora for the three registers were all sampled in the period 2000–7. On top of that, van Rooy brings in newspaper texts from two earlier periods, 1884–1918 and 1944–52. The diachronic investigation shows a clear development away from the present perfect in favour of the preterite in news reportage. The overall difference between the registers of his contemporary data is also striking, student writing returning a remarkably high present perfect/preterite ratio of 0.42, well above those for conversation and, especially, contemporary news reportage. He sees a normative constraint at work here, in that the most edited register leads the development away from the present perfect. And yet whether there really is any very strong normative constraint in this direction remains an open question. In the case of the much higher ratio recorded for student writing, he thinks a processing constraint is a better explanation, assuming that these writers, who have not yet reached the level of proficiency achieved by journalists, prefer to use an overt analytical form more frequently. Van Rooy does not address the question as to whether possible differences in general temporal orientation between registers might help to explain the very marked differences recorded.

In chapter 6 Julia Davydova looks at ‘The present perfect in New Englishes: Common patterns in situations of language contact’. Three English varieties are treated: Indian (IndE), Singapore and East African English, based on material from the respective spoken sections of ICE. In addition the London–Lund Corpus is used as a BrE control corpus. The author decides on a function-based approach and sets out to study the ‘variation of verb forms in semantic-pragmatic environments reported to be fundamental to the category of the present perfect’ (p. 174). These environments are defined in terms of Comrie's categories. A major problem is that in lots of cases, especially among those labelled ‘experiential’, the preterite, and sometimes even other verb forms, might also be selected. Davydova presents a detailed list of criteria she applied in her manual selection of tokens. Even so, many troublesome borderline cases must have remained. What impact they had on her results is difficult to assess. The overall distribution of verb forms reveals a marked difference between the three new Englishes, on the one hand, and the BrE control corpus, on the other: in the latter, 90 per cent of the verb forms selected are have perfects, while this percentage varies from 56 to 59 among the three indigenised English varieties. The lower percentages recorded in the L2 varieties are seen as a reflection of the functional complexity of the present perfect, and possibly also of L2 speakers being influenced by the world-wide trend towards a lower frequency of the English present perfect. Davydova takes the speakers in ICE to represent educated upper-mesolectal English and compares her IndE results with figures for basilectal IndE which she had published earlier. There the result for the have perfect is a very low 5 per cent. Interpreting her results in the light of Schneider's distinction between diffusion and selection, she argues that the kind of English used by the speakers represented in ICE, who must be assumed to have learnt their English largely through classroom instruction, is best understood in terms of the diffusion mechanism, ‘a process whereby linguistic features are transmitted from the parent variety to the daughter variety with minimal modifications and language-internal restructuring’ (p. 173). The selection mechanism is seen as a better model for describing the linguistic transfer taking place in more naturalistic settings, as evidenced by the greater diversity of verb forms she found in basilectal IndE.

Chapter 7 is ‘The perfect space in creole-related varieties of English: The case of Jamaican English’ by Elena Seoane. Previous comparison of parts of the Jamaican component of ICE (ICE-JA) with several of the other components had shown Jamaican English (JamE) to be the only variety where non-perfects outnumber have perfects in what are taken to be perfect contexts. A major problem is again that the author remains vague about the definition of such contexts, beyond saying that they are ‘the contexts where standard varieties would be most likely to select a have + past participle periphrasis’ (p. 195). This ignores the uncomfortable fact that there are lots of cases where the selection of either the present perfect or the preterite would be straightforward, and also that ‘the standard varieties’ (presumably including AmE and BrE at the very least) do not always agree amongst themselves. The perfect meaning is divided into the usual four categories. Seoane admits that ‘the semantic categorization was not always straightforward and [that] some borderline cases were found’ (p. 201), in which case ‘expert colleagues’ were consulted. Again it is difficult to assess what impact such difficulties may have had on the results. The backdrop to the research reported is the complex language situation in Jamaica, described as a creole continuum, with basilectal speakers of Jamaican Creole (JamC) at one end and acrolectal speakers of standard JamE at the other. Seoane interprets her results for JamE in the light of possible influence from JamC, where there is no distinction between the simple past and the present perfect, temporal meanings being expressed by the combination of preverbal particles and bare verb forms. In her ICE-JA material the have perfect accounts for just over half the occurrences of verbs taken to express a perfect meaning, with the preterite in second place. Quite a few bare forms and participles with or without be also occur. These appear mostly in the spoken material and are seen as the result of JamC influence.

In chapter 8 Robert Fuchs looks at ‘The frequency of the present perfect in varieties of English around the world’. He uses Mark Davies's 1.9-billion-word Corpus of Global Web-based English (GloWbE), from which he takes in all the twenty national varieties. Fuchs settles on a form-based definition of the present perfect, which saves him from some of the problems that several of the other contributors to this volume are up against. Quite a lot of variation in the relative frequency of the present perfect is uncovered: Kenyan English comes out on top, followed by BrE. At the other end of the scale Philippine English has the lowest frequency. The difference is substantial: the top variety has 44 per cent more present perfects than the one with the fewest. Fuchs looks at a number of factors which might help to explain the recorded variation, particularly geographical region (roughly, continent), position in Kachru's Circle Model, development phase in Schneider's Dynamic Model, formality, and reported speech. Only one of these returns a result of statistical significance: geographical region; varieties from the same region tend to display similar frequencies of the present perfect. Fuchs explains this tendency as the result of population movement and cultural and dialect contact, and also refers to possible influence from similar substrate languages. A further explanatory factor may be different influences from the superstrate language depending on when a region was colonised, since the use of the present perfect in BrE is generally assumed to have changed during the colonising period.

Chapter 9 is ‘Rise of the undead? be-perfects in World Englishes’ by Valentin Werner. He uses GloWbE and ICE, and makes frequent references to eWAVE (cf. Kortmann & Lunkenheimer Reference Kortmann and Lunkenheimer2013), where the be perfect (henceforth BEP) obtains high pervasiveness ratings in just 9 of the 76 varieties represented. The results Werner was left with after a comprehensive manual disambiguation process show some very distinct differences among the twenty national varieties making up GloWbE. The highest frequency was recorded for Bangladesh English, the lowest for South African English. Even the former frequency is no more than 21.2 per 10 million words, the latter is as low as 3.1 per 10 million words. The distribution shows a fairly clear divide between L1 and L2 varieties, in that all the varieties with the highest frequencies represent L2. A semantic analysis into action, result and state reveals that the result meaning predominates in both GloWbE and ICE, and even some of the verbs with a clear action meaning are taken to express resultative connotations, as in Shri Suvarna is informed the house that . . . (from ICE-IND). In the prototypical English BEP construction, and also in similar constructions in a number of both Germanic and Romance languages, the verb is intransitive. Werner's BEP material, however, shows an almost even distribution between transitive and intransitive clause patterns, but with some very marked differences among the national varieties. The author considers several possible explanatory factors. One of them is phonological similarity with forms of the have auxiliary, and it turns out that BEP is most frequent with auxiliaries is and was, i.e. the two forms that are most similar to has/’s. He concludes in favour of a multifactorial explanation. It has often been assumed that BEP is no longer productive and only survives as a relic construction with come and, especially, go. Werner's results show a much wider spread of verbs, most notably in some of the L2 varieties but also in the L1 ones. The fact remains, however, that its frequencies are extremely low even in the kind of largely unedited, electronically mediated language making up GloWbE. As for ICE, the vast majority of examples come from the spoken sections. Werner convinces the reader that BEP is more than just a historical relic in present-day English world-wide. In standard L1 English it still seems to be no more than a marginal phenomenon.

Chapter 10, ‘The present perfect in learner Englishes: A corpus-based case study on L1 German intermediate and advanced speech and writing’ by Robert Fuchs, Sandra Götz and Valentin Werner, is the first of the two chapters in part III. Extensive corpus material is employed, covering the English writings of German learners at school and university levels plus the English speech of German university students, accompanied by appropriate native-English control corpora. The expected influence from German on the German learners is not confirmed. On the contrary, they use the present perfect much less frequently than the English control groups overall, although the frequency increases as one moves up the age (and education) brackets, university students approaching the native level; and the present perfect frequencies are lower in their spoken English. The clearest distinguishing factor is how early students started English instruction at school: university students with more than nine years of English instruction come significantly closer to the native control groups than the others. As the authors are well aware, however, the research reported here does not give all the answers. In particular, numerical equivalence does not necessarily equal native-like competence. For a full answer a more detailed error analysis is needed, as they point out.

The volume concludes with a short chapter by Björn Rothstein entitled ‘Afterthought: Some brief remarks on autonomous and speaker-centered linguistic approaches to the present perfect’. The author draws a very sharp distinction between two major traditions in linguistics. Many of the most important treatments of the present perfect are said to belong in the autonomous tradition, also referred to as ‘speaker-free’. The other, speaker-centred approach is characterised by bringing in sociolinguistic and text-linguistic factors. Somewhat surprisingly, Rothstein places all the papers of the current volume firmly in this category. And yet references to perfect meanings such as current relevance and extended now, which Rothstein mentions as typical of the autonomous approach, abound in this volume. On the whole one may feel that Rothstein's distinction between work that belongs on one or the other side of the dividing line is too rigid.

Werner, Seoane and Suárez-Gómez are to be commended for making these papers from their 2014 workshop available to a wider audience, in a nicely produced volume. They manage to break new ground by focusing on aspects of the distinction between the present perfect and the preterite that had so far largely been ignored, in particular the treatment of the two verb forms in other than the major national varieties. All the chapters are well written and provide valuable insights, although in some cases the present reviewer would have liked to see clearer definitions, especially of what the pivotal ‘perfect meaning’ involves. Interested researchers can take comfort from the fact that the last word has not yet been said, if it ever will be.