Везде
Рец. на: J. Coon, D. Massam, L. deMena Travis (eds.). The Oxford Handbook of Ergativity. Oxford: Oxford University Press, 2017. 1296 p. ISBN: 978-0-19-873937-1
Оглавление
Аннотация Оценить Содержание публикации
Библиография Комментарии
Поделиться
QR
Метрика
Рец. на: J. Coon, D. Massam, L. deMena Travis (eds.). The Oxford Handbook of Ergativity. Oxford: Oxford University Press, 2017. 1296 p. ISBN: 978-0-19-873937-1
Аннотация
Код статьи
S0373658X0004306-5-1
DOI
10.31857/S0373658X0004306-5
Тип публикации
Рецензия
Статус публикации
Опубликовано
Авторы
ден Диккен Марсель  
Аффилиация:
Венгерская Академия наук
Университет имени Лоранда Этвёша
Выпуск
Номер 2
Страницы
151-159
Аннотация

  

Классификатор
Получено
28.04.2019
Дата публикации
05.05.2019
Всего подписок
89
Всего просмотров
598
Оценка читателей
0.0 (0 голосов)
Цитировать   Скачать pdf
1 The 35 chapters of this volume, to which 53 scholars (including the three editors) contributed their expertise, in conjunction with the extensive bibliography (covering fully 100 pages), offer outstanding coverage of the problems presented by ergative alignment, in synchronic morphosyntax (across an impressive range of languages from all of the world’s populated continents), in diachrony, and also in acquisition and sentence processing. The last section of the volume (Part IV) contains sixteen chapters presenting a treasure trove of case studies of individual languages or language families, written by seasoned experts;1 but the thirty chapters that precede Part IV also feature a wealth of data from a smorgasbord of languages with ergative traits. Reading the whole volume cover to cover is a tour de force that probably very few users of the handbook (apart from its reviewers) will subject themselves to, and which is not in every way entirely satisfying (for reasons mentioned towards the end of this review). But for all linguists and other interested parties who want or need to familiarise themselves with or remind themselves of the hallmarks, explananda, explanations, open questions, and controversies associated with ergativity, this volume will from now on be their first port of call. The handbook is doing linguistics a tremendous service — and it also shows how far linguistics has come as a field, in its analytical depth and sophistication as well as the precision of the debates about matters on which consensus remains elusive.
1. A very laudable ingredient of Part IV is the chapter (written by Christa König) on ergativity in the languages of Africa — a continent that does not usually figure prominently (if at all) in the context of ergativity studies. (In truth, this is not without reason: König herself identifies only one ‘full-fledged ergative language’ on the continent (Shilluk), and mentions a handful of split-ergative languages.)
2 One thing that emerges perhaps most saliently from a perusal of the handbook is that it is highly unlikely that ergativity is a parameter that distinguishes between languages. It is a rare thing indeed for a LANGUAGE to have consistent ergative alignment in every aspect of its morphosyntax and information structure. Ergativity manifests itself in the case system (with the subject of transitive clauses getting a different case, the ‘ergative’, from the subject of intransitive clauses and the object of transitive clauses), in the agreement system (the head-marking counterpart to dependent marking in terms of case), in the syntax (accessibility hierarchy effects in the realm of Ā-dependencies; control; perhaps even word order), and in discourse — but rarely if ever in ALL these ways in one and the same language. Therefore, the locution ‘ergative language’ is probably best avoided in linguistic parlance.2 Yet, stating (as Léa Nash does in the opening paragraphs of her chapter 8, which presents a study of split ergativity in Georgian) that ‘ergative languages are never fully ergative’ (p. 175) may be too categorical — for instance, Laz (Kartvelian) is featured in the handbook as a language that is fully ergative (see p. 7 of the introduction, and section 9.6.3.2 of Woolford’s chapter).
2. Geoffrey Haig’s chapter (p. 468) quotes Bickel [2011: 442] saying that ‘once popular expressions like “ergative language” are simply senseless’. 
3 Be that as it may, the discussion of ergativity splits at many points throughout the handbook certainly indicates that there are lots of split personalities in the ergative universe. Ellen Woolford’s chapter gives a very handy overview of the range of definitions given for ‘split ergativity’ in the literature (pp. 206–207), and answers the question of ‘whether the various types of ergative splits are present in syntax or are purely morphological’ by saying that ‘with the notable exception of person / animacy / NP splits’, most ergative splits are syntactic (p. 224).3 This dampens the prospects of the ‘TotalErg’ hypothesis (advanced in Itziar Laka’s chapter), which has it that ergativity does not split. It may very well be, however, that split ergativity is epiphenomenal, derivable from syntactic factors that are not just the privilege of ergative systems, as Jessica Coon and Omer Preminger argue in their chapter. Differential case marking is the key phrase here: so-called split ergativity is just a different name for ‘differential subject marking’, the companion to the ‘differential object marking’ patterns familiar from many languages that do not have ergative alignment patterns. Andrej Malchukov confirms in his chapter that the differential case-marking patterns exhibited by languages show a correlation with the role-indexing pattern that sets ergative–absolutive and nominative–accusative systems apart. But differential subject marking is not impossible in languages that usually mark all subjects (of finite clauses) as nominative, nor is differential object marking absent from languages with ergative alignment: the passive and antipassive diathesis alternations, respectively, instantiate these patterns.
3. Especially interesting in connection with syntactic ergativity splits is Woolford’s discussion of predicate-type-based split ergativity in Nepali (based on [Butt, Poudel 2007]), where individual-level (IL) transitive predicates have an ergative-marked subject whereas the subject of stage-level (SL) transtives is inative (p. 213). (On p. 214 Woolford notes that Nepali also has two different copulas for the IL and SL predicates — without drawing attention in this context to Spanish and Portuguese, which famously do, too.)
4 While passive and antipassive give each alignment system (nominative–accusative and ergative–absolutive) the differential marking pattern which is more typical of the other system, what makes the differential marking differential even more complicated is the fact that these voice alternations are both found across alignment types: as Maria Polinsky points out in her chapter on antipassive (see pp. 328–329), there are ergative systems with a passive; and there may be a basis for thinking that the antipassive is compatible with nominative–accusative alignment. I used more caution in my formulation of the second conjunct of the previous sentence than in the first, because I remain less than convinced, at least for the more familiar languages that Polinsky includes in her list (German, Romance, Slavic), that the case for them having a genuine antipassive is airtight. For instance, Postal’s [1977] argument for antipassivisation in French, based on unspecified object deletion in ‘faire-infinitive’ causative constructions, is far from conclusive: much depends on one’s outlook on the syntax of the case pattern of the Romance causative.
5 Now that the Romance causative construction has entered the discussion, this may be a good moment to draw attention to the occurrence of ‘ergative’ case patterns in otherwise nominative–accusative languages. The fact that in the French equivalent of I will make him eat his soup we see oblique (here, dative) case on the causee (Je lui ferai manger sa soupe ‘I him.DAT will. make eat his soup’) whereas in faire-causatives with an intransitive infinitive (whether it be unergative or unaccusative) the full-nominal causee bears accusative case (Je le ferai rire / tomber ‘I him.ACC will.make laugh / fall’) is a differential case-marking pattern very much like what we find in so-called ergative languages.4 Remarkably, the ‘faire-infinitive’ causative of the Romance languages is not mentioned anywhere in the handbook, even though Woolford, in particular, is describing a pattern directly from the playbook of the Romance causative. On p. 215, Woolford lays her finger on the role of control in the split-ergative case system of Folopa, where ‘[i]f a person uses the ergative [he] thus states his control’ over the situation (in the words of Anderson & Wade [1988: 11]). The role of control in split ergativity rears its head again on pp. 259 (Sa- moan) and 267 (Lezgian), in Malchukov’s chapter. The parallel with the French ‘faire-infinitive’ causative is hard to miss here: both in the functional literature (see especially [Huffmann 1997]) and in generative work [Authier, Reed 1991], it has been pointed out that there is an explicit connection, particularly when the causee is a clitic, between the case of the causee and the degree of control exerted by it over the causativised event. We are plainly looking at the same phenomenon in both French and Folopa, Samoan and Lezgian.
4. It may be relevant to note in this context that the 19th century missionary George Taplin, in his analysis of Ngarrindjeri, used the term ‘causative’ for what we would now call ‘ergative’ (see William McGre- gor’s chapter, p. 453).
6 This leads me to a more general point. In their introduction to the handbook (p. 3), the editors point out correctly that ‘not only do we find non-ergative patterns throughout languages traditionally labeled “ergative,” we also find ergative patterns in a number of language [sic 5] and domains normally considered “nominative–accusative.” But although at various points throughout the handbook there is discussion of the occurrence of ergative-like patterns in languages that do not generally show ergative–absolutive alignment, it seems to me that more could have been made of potentially profound parallels between familiar nominative–accusative patterns and things we find in systems with ergative alignment. The classic hypothesis that the source of ergative alignment is the passive construction is brought up in several chapters — and it is roundly debunked by all authors who mention it. But the fact that the passive of, say, English singles out the notional subject of transitive clauses and marks it with a special case or preposition is not the only thing reminding one of ergativity in nominative–accusative systems. There are quite a few alternations in so-called nominative–accusative languages (in addition to the passive or the faire-causative already mentioned) in which a particular argument can be marked either with structural case or in some oblique way. The conative alternation (He ate (at) the meat), which confines itself to the object of transitive clauses, is mentioned explicitly by Polinsky in the context of a discussion of the antipassive (see pp. 325, 329) and is also brought up in passing by Coon & Preminger (p. 234) and in Mary Laughren’s chapter on Warlpiri (the Warlpiri conative strikingly has an ergative-marked agent and a dative-marked object: p. 953). But Warlpiri’s rendition of The man shot the kangaroo with a rifle (Laughren, p. 953), which has two ergatives, one on ‘man’ and the other on ‘rifle’, seems to me much more interesting. Laughren’s observation that ‘[i]n Warlpiri, the ergative-marking of instrument DPs is only grammatical if the subject is also marked ergative’ (p. 957), hence not in intransitive clauses, reminds one of the case alternation on the instrument argument in English He opened the door with the key and The key opened the door. Perhaps even more to the point is the case alternation on the cause argument seen in English alternations such as The door closed from the wind and The wind closed the door, which finds an exact match in two Samoan examples mentioned in Malchukov’s chapter (p. 259) — though unfortunately the parallel is missed due to the fact that Malchukov translates the locative case variant in the same way as the ergative one. A translation with from would have been a useful segue to the relationship between ergative case and directional P-elements such as English from, which is addressed elsewhere in the handbook — prominently in William McGregor’s chapter on the grammaticalisation of ergative case marking (see p. 449), in passing in Michelle Sheehan’s contribution (see p. 61), and also in an interesting way in Eva Schultze-Berndt’s discussion of the ablative in Jaminjung, which can mark the A-argument of a transitive clause with either the ergative or the ablative, depending on animacy and information structure (more on the latter below). The fact that English has a case alternation involving precisely the preposition from and finding a close parallel in some ergative case systems is surely indicative of the fact that there is no deep distinction between ergative and non-ergative languages.
5. This typo will serve as my cue for a few small remarks about some presentational imperfections of the handbook. In several places, names of linguists are misspelled (often with dogged persistence) — thus, Sheehan’s name is consistently misreproduced as ‘Sheenan’ in Laka’s chapter (which suffers from a variety of other typos as well), Altshuler’s name appears as ‘Artshuler’ and Depraetere’s as ‘Depreatere’ in Nash’s chapter (the former but not the latter also shows up with the wrong spelling in the general bibliography), and Pylkkänen’s name lacks its umlaut (‘Pylkkanen’) every time Kaufman’s chapter makes mention of it. Another recurrent problem is the variable spelling of certain language names throughout the handbook (e. g., Ch’ol / Chol, Yup’ik / Yupik). On p. 249, in (39a), the gloss is mixed up (so’ should have been glossed as ‘am’ and magnate as ‘eaten’, rather than the other way round); but thankfully this is an example from Italian, a relatively familiar language, so probably most readers will be able to recover from this error. The occasional missing word hampers parsing of some sentences in the handbook (see e. g. the last sentence of the first paragraph of the conclusion of Legate’s chapter). Missing italicisation on some in-text citations of linguistic material (e. g. on p. 65) and omission of first-line indentation at the start of several new paragraphs in chapter 22 will not affect readability but should also have been fixed at the typesetting stage. Overall, however, considering the sheer size of this volume and the large number of contributors, the handbook looks excellent. 
7 Also worth exploring is the possible connection between Malchukov’s Lezgian example (p. 267) in which the ‘involuntary agent’ of the event is marked with adelative case (as opposed to the ergative case that the deliberate agent of the same event would bear) and English experiential constructions such as The pot broke on Zamira. These English cases are striking not only for introducing with a preposition an argument higher on the Silverstein hierarchy than the nominative argument, but also for the fact that they alternate with have-sentences in which the theme is the accusative object and the ‘involuntary agent’ is the nominative subject, obligatorily ‘doubled’ by a pronoun inside a PP headed by on, as in Zamira had the pot break on her (see [Belvin 1996] for detailed discussion). Highly relevant in this context is the intriguing difference between the Shipibo rendition of My monkey died on me (where ‘monkey’ is marked ergative; p. 117) and the Yup’ik version of A bearded seal sank on that guy (with ‘guy’ marked ergative; p. 119), both taken from Mark Baker and Jonathan Bobaljik’s chapter. Though much will depend on how the language does ‘have’, what comes to mind is that the Yup’ik example is in fact the equivalent of the English have-sentence That guy had a bearded seal sink on him. The more closely one looks, the more parallels between English and ‘exotic’ ergative languages one is likely to find.
8 If the suggestion made in the previous paragraph is on target, we are dealing in this Yup’ik example with the ergative–absolutive case pattern familiar from transitive sentences — but the noun phrase in the subject position of the have-sentence is probably a DERIVED subject (see already [Benveniste 1966], and more recently [Freeze 1992] and [Kayne 1993], on the syntax of have-constructions; Itziar Laka’s discussion of the Basque equivalent of need is also relevant here). This is an important point because it bears directly on what is perhaps THE major bone of contention in the literature on ergativity: the status of ergative case as an inherent case (most extensively defended in the handbook by Julie Legate) or as a dependent case in the sense of [Marantz 1991] (elucidated and advocated by Baker & Bobaljik, and seconded in Léa Nash’s chapter on Georgian). Tarald Taraldsen mentions (p. 341) a few languages that allow ergative derived subjects, singling out Hawrami [Holmberg, Odden 2004] as an interesting case in point: Hawrami has ergative–absolutive alignment in passivised double-object constructions, though not in passivised monotransitives. This is perfectly in line with Baker & Bobaljik’s dependent case approach: in passive double-object constructions there is case competition, forcing the deployment of dependent (ergative) case, whereas in the passive of a monotransitive verb a single absolutive is sufficient. Baker & Bobaljik devote careful attention, in connection with the inherent vs. dependent case distinction, to the occurrence of ergative on internal arguments, presenting Shipibo as the primary case in point (but mentioning Kalaallisut, Chukchi and Yup’ik as well). Laughren observes that ‘[a]s in many other Australian languages, the ergative suffix marks DPs in several non-subject roles in Warlpiri’ (p. 957) — indeed, Baker & Bobaljik assert that cross-linguistically ‘there is no type of theta-role that is totally immune to ergative case’ (p. 121). This all points squarely in the direction of the dependent case approach.
9 Yet, at the other end of the spectrum, Legate’s chapter draws attention to the important fact that ‘an absolutive object is not required for ergative case assignment’ (p. 143), although ‘[t]he presence of a complement IS RELEVANT’ (p. 148, my emphasis). The availability of ergative case for the subject in the absence of an absolutive object is of course naturally expected on an inherent approach to ergative case. The dependent case approach can accommodate this, for example, by postulating a silent object ‘taking’ absolutive case, but this raises non-trivial questions. However, the inherent case approach to the ergative faces an even larger question (to which [Polinsky 2016] offers a possible answer from an acquisition point of view): if ergative is an inherent case assigned by v to its specifier, why don’t all languages have inherently case-marked external arguments of verbal predicates? Perhaps subscribers to an analysis of faire-causatives in which the dative causee has an inherent case assigned by v, or followers of Collins’ [2005] approach to the external argument of passives, would counter that in fact they do. But that would be a mistake, in part for the same reasons as the ones brought up by Baker & Bobaljik in connection with ergative case assignment: the dative causee of a faire-causative and the by-phrase of a passive can have a wide range of θ-roles.
10 As I see it, what ought to settle the debate between proponents of the inherent case approach to the ergative and the adherents of the dependent case perspective is that there is arguably no sense in which ergative could truly be an inherent case assigned by v and tied to a thematic role (or a cluster of θ-roles) assigned by v. At the heart of the matter is the fact that the source of the external argument’s θ-role is not v but the VP in its complement. We have known since Marantz’s [1984] cogent arguments to this effect that the nature of the θ-role borne by the subject of predication is determined compositionally by the verb and its complement together: in throw a ball and throw a fit we are dealing with different θ-roles because the predicates are different (even though the verbs are the same). A major step forward in the v–VP approach to the structure of transitive sentences was precisely the ease with which the compositionality of external θ-role assignment is structurally encoded: v is the mediator of a predication relationship between VP (the predicate) and its external argument (in SpecvP), not a θ-role assigner. But if v is not a θ-role assigner, it cannot be an inherent case assigner either. This strikes me as a happy out- come because I know of no convincing cases elsewhere in the grammar in which an EXTERNAL argument (i. e., a subject of predication) is assigned inherent case.
11 Particularly pertinent here is a brief examination of the genitive possessor of noun phrases. Though the literature on genitive case has erred frequently on this point (see e. g. [Chomsky 1986]), it cannot be that the genitive is an inherent case — cross-linguistically it can be assigned to a wide variety of argument types, including internal arguments (Mary’s portrait); it can even be assigned to non-arguments (yesterday’s newspaper). There are frequent parallels across ergative languages between ergative and genitive case (a case in point is the so-called ‘relative case’ of the Eskimo languages, which covers both). Besides the ergative case, the genitive is really the only POTENTIAL candidate for being an inherent case assigned to an external argument (of n in this case; see Artemis Alexiadou’s chapter for discussion of the nP structure of complex noun phrases, and also Edith Aldridge’s and Daniel Kaufman’s chapters for discussion of the relationship between nominality / nominalisation and ergativity). So the fact that the genitive case is NOT an ACTUAL candidate for such a treatment casts a dark shadow on the prospect of an analysis of ergative case as inherent. It seems to me likely that the only assigners of inherent case are prepositions — which may leave an opening for ergative case if there are languages for which the ergative is plausibly analysed as a P-assigned case. Judging from Nash’s assessment (see p. 190), such an approach is not viable for Georgian; but it may be right for other languages (see [Polinsky 2016]).
12 One last remark that I would like to make in the context of case assignment in languages with ergative–absolutive alignment concerns the structural height of the case assigners.6 There is debate on this point in the literature for both ergative and absolutive case. Legate’s chapter addresses ergative case and structural height, arguing that the ergative argument is typically low in the structure (SpecvP). She leaves open the question of ‘whether true “high ergative” languages may be found’ (p. 157) — a tantalising question because, as Legate is right to stress on p. 135, there is a nomenclatural issue here: how is a “high ergative” language in which ergative is assigned by T and absolutive by v different from a nominative–accusative language? A language in which ergative and absolutive are BOTH assigned by T would certainly be different from a nominative–accusative one, for which v is indubitably the source of accusative case. In Michelle Sheehan’s chapter, the question of whether the source of absolutive case is parametrised (i. e., whether there are ‘low’ as well as ‘high’ absolutive languages) is reviewed briefly, with reference to the relevant literature.
6. A question which I have no space to discuss at any length here is whether default case might play a role in ergative systems, as it does in nomative–accusative ones. In Longenbaugh & Polinsky’s chapter the possibility of there being two different ways of dealing with the absolutive case (as a default case or as the equivalent of nominative case) is brought up explicitly (p. 732). See also Taraldsen’s chapter for a brief and highly relevant remark (p. 350), and Austin’s chapter on the role of defaults in the acquisition of Basque ergative and dative morphology.
13 Related to the question of the structural height of the ergative and absolutive arguments is the relationship between ergativity and word order (especially the claim that ergative alignment does not go together with a basic SVO word order — ‘Mahajan’s generalisation’), as well as the important question of the information-structural profile of sentences with ergative–absolutive alignment. On the former point, the reader will find important discussion in Tarald Taraldsen’s contribution, and also in Ritsuko Kikusawa’s chapter, which argues (among other things) that one of the key factors resulting in a diachronic change in the case-alignment system is word-order change. In her chapter on ergativity in the languages of Africa, Christa König observes (p. 922) not only that in these languages ergativity is associated with ‘an unusual constituent order’ (viz., ‘the odd OVA / SV order’), but also that its occurrence is confined to ‘pragmatically marked con- structions’ — in particular, ‘the pragmatic marking of topicalized or focused objects’, or, perhaps more accurately, the marking of contrastiveness, which topics and foci can both be associated with (see also the discussion in Schultze-Berndt’s chapter of the role of contrastiveness or un- expectedness in the case-marking of agents in Jaminjung; pp. 1106–1109).
14 The connection between ergativity and information structure plays a significant role in Alana Johns and Ivona Kučerová’s chapter, and especially in John Du Bois’s contribution, which takes a functionalist and highly comparative approach that I found a real eye-opener to a very complex and timely topic. Starting off with a close reading of a story fragment from Sakapultek (Mayan), Du Bois observes that ‘the discourse distribution of lexical [as opposed to pro-dropped — MdD] arguments (and of new information) corresponds to the absolutive category in the grammar of ergative languages, while topically continuous elements are found in what would be the subject in accusative languages’ (p. 29) — a conclusion that he subsequently confirms on the basis of an examination of a wide range of other languages from several families. The picture that emerges is that, whereas in nominative–accusative alignment patterns the bearer of the unmarked case (nominative) is usually topical, in ergative systems new information is paired with un- marked case (absolutive). Malchukov’s observations on p. 265 confirm this for some languages not mentioned by Du Bois; but the generalisation that ergative-marked transitive subjects are overwhelmingly definite and given appears to be counterexemplified by Ika, Gooniyandi and Newari (pp. 263–264). The complete picture is no doubt more complex than broad brushstrokes can paint — just as it is with nominative–accusative alignment. The message here is that discussions of the information-structural profile of ergative alignment systems are strongly encouraged, and should invite an integrated investigation that also encompasses differential object marking (see Coon & Preminger’s, Malchukov’s and Butt & Deo’s chapters), object shift (see especially Woolford’s pp. 221–222), and ‘scrambling’ phenomena more generally (including extraposition). All of these are known to have a cocktail of morphological, word-order and information-structural effects. For a small subset of the world’s languages, the interactions between these effects are beginning to be understood to quite an advanced degree. It is a very welcome development that they are garnering more and more careful attention in studies of ergative systems as well.
15 In a short review of a book totalling well over a thousand pages of densely packed text and a huge wealth of information, I obviously cannot do justice to everything that is being discussed. In the present paragraph and the next, I will merely highlight, by way of invitations for further research, two other themes that stand out in the handbook. The first is diachronic development and its directionality. There is ongoing debate in the literature on Oceanic languages (dating back to the 1970s; see [Chung 1978]) about the direction of alignment change. According to Kikusawa, in Austronesian and Oceanic languages there is a ‘general flow of ergative to accusative change’ (p. 587), although AFTER the completion of such change some Oceanic languages underwent a more recent change from accusative back to ergative. In the Austronesian context, the developmental relationship between ergativity and the complex voice systems that many of these languages are known for is also an important matter. Mitcho Erlewine, Ted Levin & Coppe van Urk’s chapter concludes on this point that ‘ergativity cannot be the only route to a voice system’ (p. 395). More work on this is called for.
16 A second invitation for research emerges from Martina Wiltschko’s contribution, which is so programmatic (in the author’s own words) that it rather stands out in this handbook. Wiltschko’s chapter argues that ‘ergative constellations are detectable in the syntax of speech acts’ (p. 441), helping to mark relations between speaker and hearer and their commitment to the proposition. The interesting Nez Perce data that William McGregor presents on pp. 457–458 of his chapter fit into Wiltschko’s programme like a hand in a glove. I mention this connection here because readers who look in isolation at either Wiltschko’s or McGregor’s chapter (which are consecutive but separated by the split between Parts II and III of the handbook) may not realise that the other chapter contains highly relevant information: unfortunately these chapters contain no cross-references to one another.
17 The relative dearth of cross-references between chapters is one of the minor flaws of the handbook as a whole. Perhaps a little bit more editing could have been done to enhance the cohesion of the volume. One also notices that there is quite a bit of overlap between individual chapters, which becomes mildly annoying after a while to readers (such as your reviewer) going through the handbook linearly.7 But few readers are likely to consume the handbook linearly from cover to cover, so this redundancy may be a blessing rather than a curse. On a related note, a welcome exercise in cross-chapter editing would have placed the various individual disprovals of the passive as the source of ergative alignment (figuring prominently in three successive contributions to Part III (chapters 20–22) but at other points in the volume as well) in one single location in the handbook.8
7. For instance, in chapter 23, halfway into the handbook, we are suddenly presented again with a text- book-style, paragraph-long exposé of what ergativity is (p. 553), complete with a footnote defining the by now very familiar labels ‘A’ (for ‘agent of transitive verbs’), ‘P’ (‘patient of transitive verbs’), and ‘S’ (‘subject of intransitive verbs’).

8. Another small point of criticism is that neither the editors, in their introduction, nor any of the other contributors discusses the origin of the term ‘ergative’. We encounter the term ‘ergator’ (not included in the index) once, as an alternate name for the ‘A’ argument in a quote from England [1991:484] presented in Du Bois’s chapter (p. 40).
18 As the editors point out in their introduction, ‘the approaches of the authors and the scope of the studies vary considerably’ (p. 14). There is some variability in the quality of the hand- book’s individual contributions as well. Most are superb, but a few chapters strike me as sub-par, and some contributions seem to me out of place, being more like research papers than handbook pieces.9 And although it is a very good thing that there are sections dedicated to language acquisition and experimental approaches in the handbook, the three chapters in the acquisition section are, unfortunately, not particularly robust. On the other hand, Adam Zawiszewski’s short digest of sentence-processing studies of ergative systems is a useful prelude to Nicholas Longenbaugh and Maria Polinsky’s chapter on the processing of subject-gap and object-gap relative clauses in split-ergative systems, showing how the two central hypotheses regarding the processing of relative clauses (the subject processing advantage hypothesis, and the case-cueing hypothesis), while aligned in nominative–accusative patterns, make different predictions in split-ergative systems. These two chapters should serve as a helpful guide to anyone interested in adding to the (modestly) burgeoning volume of experimental studies performed on languages with ergative alignment.10
9. Chapter 3 is a somewhat superficial exercise in parameter hierarchies, not very helpful in a broader context. Chapter 4 makes a simple (though very valid) point, but is really more like a research paper than a handbook chapter (and it could perhaps have been an appendix to the other chapter on Hindi/Urdu). Chapter 7 is extremely narrow in scope. Chapter 12 is rather too technical and idiosyncratic for a handbook piece. Chapter 15 is a recital of (mostly the author’s own) work on nominalisation, which chapters 21 and 24 do a much better job at relating directly to ergativity.

10. Longenbaugh & Polinsky’s chapter is also noteworthy for containing new studies of Niuean and Avar. In this context, I would also like to highlight the novel discussion of Dinka Bor in Erlewine et al.’s chapter, and the fact that Kaufman’s contribution features original material on Mamuju, a language not previously described at any length.
19

The editors ‘have done [their] best to include chapters representing a range of different theoretical and methodological traditions, though, as with any volume, imbalances reflecting the orientation of the editors nonetheless exist’ (p. 2). These imbalances did not strike me as serious. The handbook presents a veritable who’s who of the field of ergativity studies enlivened with occasional contributions from relative outsiders whose out-of-the-box thinking about ergativity places the central questions in this realm in a broader perspective (as John Du Bois aptly puts it, ‘[e]rgativity is too important to be left to the specialists of ergative languages’; p. 57). Here I am thinking, for instance, of Tarald Taraldsen’s exercise in questions of the general type ‘if x is true of ergative systems, what are x’s implications and roots?’. As an illustration of Taraldsen’s dialectic approach, let me point to p. 340, where we find the following passage (somewhat condensed here): ‘If the generalization does hold for tripartite languages, we conclude that the generalization really is to be understood as *SErgVO, but if it doesn’t, the generalization must be *SVOAbs. If the generalization is valid for neutral languages with Erg / Abs agreement patterns, we conclude that the generalization is really about syntactic case <...> If the active agreement pattern reflects underlying Active / Inactive <...> case alignment <...>, one may conjecture that *SErgVOAbs doesn’t hold for active languages <...> If a suitably precise version of the *SErgVOAbs generalization holds, one will ask why <...> If SVO order is merely incompatible with assignment of syntactic absolutive case to the object <...>, there should be tripartite SVO languages <...> But if SVO order is incompatible with the subject surfacing with ergative case, there should be no tripartite SVO languages <...> At present, however, no conclusion can be drawn from this, since the data available to me seems insufficient.’ 11

11. Readers interested in tripartite (or three-way) systems may find it useful to know that Gereon Müller and Daniela Thomas argue in their chapter that tripartite systems do not exist.
20 All things considered, The Oxford Handbook of Ergativity is a splendid resource and a proud testament to the achievements of the many linguists who have made empirical and analytical contributions to ergativity in all its many hues. Logically organised and jam-packed with data and analyses, this is the place to go for specialists, ‘accidental tourists’ (in the words of Longenbaugh and Polinsky; p. 710), and anyone who has ever had a burning question about ergativity but was afraid to ask it: here are the answers, and plenty of food for thought and avenues for future research into the bargain.
21 Acknowledgement
22 I would like to express my deep gratitude to Masha Polinsky for offering me a copy of the handbook, suggesting my name to the review editor of this journal (whom I thank for the invitation), and providing excellent comments on a first draft of this piece. The pointed remarks of an anonymous reviewer also led to improvements. The responsibility for any misrepresentations of the facts presented and views expressed in the handbook is entirely mine.

Библиография

1. Anderson, Wade 1988 — Anderson N., Wade M. Ergativity and control in Folopa. Language and Linguistics in Melanesia, 1988, 19: 1–16.

2. Authier, Reed 1991 — Authier J. M., Reed L. Ergative predicates and dative cliticization in French causatives. Linguistic Inquiry, 1991, 22: 197–204.

3. Belvin 1996 — Belvin R. Inside events: The non-possessive meanings of possession predicates and the semantic conceptualization of events. Ph.D. diss. Los Angeles: Univ. of Southern California, 1996.

4. Benveniste 1966 — Benveniste E. Problèmes de linguistique générale. Paris: Gallimard, 1966.

5. Bickel 2011 — Bickel B. Grammatical relations typology. The Oxford handbook of linguistic typology. Song J. J. (ed.). Oxford: Oxford Univ. Press, 2011, 399–444.

6. Butt, Poudel 2007 — Butt M., Poudel T. Distribution of the ergative in Nepali. 2007. URL: http://ling.uni-konstanz.de/pages/home/tafseer/leipzig07-hnd.pdf

7. Chomsky 1986 — Chomsky N. Knowledge of language: Its nature, origin, and use. New York: Praeger, 1986.

8. Chung 1978 — Chung S. Case marking and grammatical relations in Polynesian. Austin: Univ. of Texas Press, 1978.

9. Collins 2005 — Collins C. A smuggling approach to the passive in English. Syntax, 2005, 8: 81–120.

10. England 1991 — England N. Changes in basic word order in Mayan languages. International Journal of American Linguistics, 1991, 57: 446–486.

11. Freeze 1992 — Freeze R. Existentials and other locatives. Language, 1992, 68: 553–595.

12. Holmberg, Odden 2004 — Holmberg A., Odden D. Ergativity and role-marking in Hawrami. Conference handout. 2004.

13. Huffmann 1997 — Huffmann A. The categories of grammar: French lui and le. Amsterdam: John Benjamins, 1997.

14. Kayne 1993 — Kayne R. Toward a modular theory of auxiliary selection. Studia Linguistica, 1993, 47: 3–31.

15. Marantz 1984 — Marantz A. On the nature of grammatical relations. Cambridge (MA): MIT Press, 1984.

16. Marantz 1991 — Marantz A. Case and licensing. ESCOL 91: Proc. of the Eighth Eastern States Conference on Linguistics. Westphal G. F., Ao B., Chae H.-R. (eds.). 1991, 234–253.

17. Polinsky 2016 — Polinsky M. Deconstructing ergativity: Two types of ergative languages and their features. Oxford; New York: Oxford Univ. Press, 2016.

18. Postal 1977 — Postal P. Antipassive in French. Linguisticae Investigationes, 1977, 1: 333–374.

Комментарии

Сообщения не найдены

Написать отзыв
Перевести