David Peterson's typological study Applicative constructions is divided into seven chapters, all of which, except chapter 7 with concluding remarks, are summarized below. The book's three appendices are not included in the present evaluation either.
The first chapter of Applicative constructions is a short introduction to the basic nature of applicative constructions. The author also explains there the purpose and the organization of his study. According to Peterson's definition, applicative constructions can be conceived of as involving morphosyntactic means to encode a thematically peripheral participant as one of the core arguments of the verb. However, such ‘augmentation’ of a peripheral entity is a valence-changing mechanism and, accordingly, applicative constructions are usually seen as involving transitivization. Crucially, applicativization affects intransitive verbs by making them transitive and transitive verbs by making them ditransitive.
Peterson provides the reader with two examples of applicative constructions: one example comes from Ainu and involves a Locative applicative construction (‘He lives in a big house’); the second example comes from San Lucas Quiaviní Zapotec and illustrates a Comitative applicative construction (‘Mike sang with John’)
In concluding the chapter, Peterson makes some methodological observations concerning the nature of previous research on applicative construction. In particular, Baker's (Reference Baker1988) contribution, couched in the Government-Binding (GB) framework, is acknowledged, as well as other studies, particularly within the framework of Lexical Functional Grammar (LFG). Peterson points out that the non-syntactic and non-structural properties of applicative constructions such as their pragmatic function, historical development and typological structural correlates have mostly been overlooked and that the present study aims at filling this gap.
In chapter 2, ‘Two case studies of applicative morphosyntax’, Peterson narrows the discussion on applicatives to two genetically-unrelated languages, Bukusu (Bantu) and Hakha Lai (Tibeto-Burman). The purpose of these case studies is to further illustrate certain pervasive characteristics of applicative constructions.
Beginning with Bukusu, the author claims that the nature of applicative constructions in this language depends on the semantic interpretation of the arguments of the verb. In particular, it is argued for ditransitive environments that when one of the two objects is thematically interpreted as Beneficiary, both objects (which Peterson calls ‘base’ object and ‘applicative’ object, respectively) are treated similarly vis-à-vis object properties. If, on the other hand, one of the objects is Locative or Instrument, then the objects exhibit asymmetric behaviour with regard to Bantu objecthood heuristics. A common trait in applicative constructions is the topicality of the applicative argument: applicative objects are apparently usually associated with a high degree of topicality in this language. In fact, in Hakha Lai, the applicative object can indeed co-occur with a specialized topicality marker.
Peterson notes various differences between Benefactive and Instrumental applicatives. In particular, in Benefactive applicative constructions the two objects tend to exhibit more symmetric behaviour than in Instrumentals. For instance, both objects may be encoded with an object pronominal or undergo passivization as well as relativization. The author concludes that both objects largely receive ‘symmetric’ treatment, modulo the linear order of the objects where it is invariably the Beneficiary that is adjacent to the verb. In the case of Benefactive applicatives, the object properties may also be distributed in the sense that one of the objects is eligible for e.g. passivization while the other is encoded with pronominal marking on the verb. The symmetric treatment of the objects is not, however, unconditional, but apparently regulated by animacy: Peterson argues that a considerable amount of asymmetry in the treatment of the objects arises when both of them are animate; in such cases only the Beneficiary is eligible for e.g. passivization.
Instrumental applicatives differ in many respects from Benefactive applicatives. First of all, the Patient (not the Instrument) has exclusive access to object pronominals. Secondly, only the Patient (not the Instrument) may undergo passivization. In relativization things change, however, in the sense that it is in fact the Instrument (not the Patient) that displays object properties (i.e. undergoes relativization).
In chapter 3, ‘Morphosyntactic and semantic variation in applicative constructions’, Peterson describes the syntactico-semantic variation in applicative constructions. The central questions concern the obligatoriness of applicative constructions in some languages (and in some semantic flavours) and, again, in asymmetry in the treatment of the applicative objects. Peterson points out that applicative constructions exhibit cross-linguistic variation first of all regarding the semantic role of the applicative object. He argues that the most common applicative construction is the Recipient/Benefactive/Malefactive applicative construction, but it is by no means the only one attested. Languages also differ with regard to applicative marking itself. Some languages use the same morpheme across the board, while others employ distinctive morphology in each applicative construction type.
Peterson then embarks on a discussion on the obligatoriness (or non-obligatoriness) of applicative constructions. The questions related to this topic concern the definition of the applicative construction itself, namely, whether or not the applicative construction is the only one available (i.e. whether or not there exists any oblique variant to it), and whether or not it is legitimate to call the non-alternating constructions applicative construction. The obligatoriness of applicative constructions seems to be contingent upon semantic roles and animacy in the sense that in many languages animate objects require the use of an applicative construction. The author concludes that the obligatoriness of applicative constructions might have a diachronic dimension.
Peterson continues the discussion of symmetry between the base and the applicative object. He points out that Benefactive applied objects are particularly prone to acquire object properties at the cost of relinquishing the object properties on the base object at the same time. In general, it is the applicative objects that control object agreement (with the exception of Instrumental applicatives) and topicality. On the other hand, it seems to be the case that while applicative objects can indeed acquire object properties, it is not invariably the case that the base object loses these characteristics in the presence of an applicative object. In fact, sometimes the base object only loses some of its original object properties. In Kalkatungu, for instance, the base object may lose its property to control verbal object agreement, but it still retains the ability to interact with cross-clausal reference and antipassivization.
In chapter 3, Peterson also discusses how the presence of applicative morphology alters the semantics of the verb and the construction as a whole. The semantic changes seem to split into two types: (super)transitivization and re-arrangement. In (super)transitivization, the applicative morpheme transitivizes an intransitive predicate and supertransitivizes a transitive predicate (giving rise to a double object verb). Languages like Kinyarwanda follow the (super)transitivization pattern. On the other hand, languages like Chichewa seem to retain a monotransitive pattern in the applicativized verb, dispersing the object properties between the two objects. According to the author, Chichewa belongs to the second semantic type, the re-arrangement type. In re-arrangement languages it makes less functional sense to applicativize intransitive roots.
Concluding chapter 3, the author refers to previous, mainly syntactico-centric, research on applicative constructions. The contributions of Relational Grammar, LFG, GB theory and minimalism are acknowledged. Pylkkänen's (2002) high/low applicative typology within the minimalist framework receives some discussion but the author does not adhere to any of the central assumptions advocated in this research.
In chapter 4, ‘The discourse functions of applicative constructions’, Peterson argues, in the spirit of functional grammar, that the discourse functions of applicative constructions are orthogonal in understanding the basic nature of applicative constructions in human language. The discussion in this chapter is largely comparative, the object of comparison being the discourse properties of applicative objects vis-à-vis oblique objects. It is shown that applicative objects usually have greater tendency to display characteristics of animacy, pronominality, properness and activeness. Applicative objects are also lexically ‘shorter’ than oblique arguments, typically encoded as animate pronominals.
The author launches a theory according to which the primary function of applicative objects is to encode an argument with greater discourse salience and topic continuity. The hypothesis is consistent with the fact that one of the core discourse functions of applicative constructions is to allow peripheral arguments to access specific syntactico-semantic operations such as relativization, topicalization and, in some cases, passivization. In addition, it is not uncommon for applicatives to be used in interrogative environments such as wh-questions. Whether or not this indicates the interaction of applicatives with (contrastive) focus is cursorily mentioned but left without a more detailed discussion.
Many parallels between applicative constructions and dative shift constructions are drawn: dative-shifted arguments are assumed to display greater topic persistence than non-dative-shifted arguments, just like applicative objects, especially Recipients. In fact, Peterson points out that the data from Hakha Lai and Wolof reveal that there is almost no difference in topic-worthiness between applicative objects and Recipients.
The discussion in chapter 5, ‘The evolution of applicative constructions’ seems to have a two-fold purpose. First of all the author is concerned with the grammaticalization source of applicative morphemes. Secondly, the author also embarks on a discussion concerning the evolution of applicative morphemes themselves, the use of which is sometimes extended to new grammatical territories.
According to Peterson, there are two primary sources for applicative morphemes: adpositional and de-verbal. In fact, the author points out that the etymology of the applicative morpheme sometimes mirrors the grammatical relation of the applicative object, which, as we recall, usually display ‘object-like’ behaviour in various contexts. Peterson argues that the object-like behaviour of applicative objects is to be expected given that applicative objects sporadically originate as objects/complements to adpositions or verbs such as the typical ditransitive ‘to give’. Applicative constructions in languages like Nadëb and Kinyarwanda seem to have evolved from adpositional sources while applicative constructions in Sapathian-Klamath, the Papuan language Yimas as well as the by-now-familiar Hakha Lai, have de-verbal sources, particularly the verb ‘to give’ mentioned above.
Applicative morphemes may also have de-nominal source, although this type of grammaticalization is apparently not as extensively attested as grammaticalization from adpositions or verbs. Peterson also points out that grammaticalization from nominal sources presumably involves an intermediate adpositional stage. That said, Peterson shows that nominal sources of applicatives indeed do exist, and that they often materialize as incorporated body parts, such as ‘eyes’ or ‘face’.
Concerning the ways in which applicative morphemes themselves are subject to further evolution and grammaticalization, Peterson claims that applicative morphemes apparently have the tendency to grammaticalize e.g. into topicality markers; Locative and Instrumental applicative markers seem to be particularly prone to such developments. Instrumental applicative markers, on the other hand, also tend to develop into relativization markers encoding oblique objects. Furthermore, Peterson observes that the verb marked with an applicative morpheme may lexicalize and develop into a novel verb with a distinctive meaning.
The topic of discussion in chapter 6, ‘Structural correlates of applicative constructions’, is whether or not the presence of applicative constructions correlates with the presence of other constructions or more abstract structural features in a given language. The research is corpus-based, drawing on a ‘convenience sample’ consisting of 100 languages with worldwide distribution; 50 of the languages have applicative constructions and 50 of the languages lack applicative constructions.
Peterson returns to the issue of identifying applicative constructions. He also addresses apparent dilemmas related to typological linguistics that, at least in its most traditional forms, is inherently dependent on second-hand information in the form of vague descriptive grammars with inconsistent terminology, in addition often written by linguists with non-native proficiency of the particular language.
The parameters Peterson tests as possible correlates for applicative constructions are related to (i) the nature of applicative construction itself, i.e. the variety of applicatives that can be encoded in a language, (ii) head vs. dependent marking, (iii) morphological complexity, (iv) word order, (v) alignment (i.e. the way of expression of core thematic relations), (vi) relative clause structure, (vii) means for encoding spatial relations, (viii) the presence of other valence affecting constructions, and (ix) Instrumental noun incorporation. On the basis of the discussion concerning these parameters, Peterson reiterates the fact already noted in Chapter 3 that Recipient/Benefactive/Malefactive applicatives are typologically the most common variety of applicative constructions, followed by Comitatives/Instrumentals and Locatives. A negative correlation between Benefactives and Instrumentals was detected. On the other hand, the existence of Locative applicatives usually entails the presence of some other type of applicative. Instrumental and Benefactive applicatives do not require the presence of other types of applicatives.
Whether or not a language exhibits head or dependent marking does not have any significant correlation with the presence of applicative constructions; similarly, there is no correlation between morphological complexity and applicative constructions. There is apparently no correlation either with word order, relativization or alignment. Peterson does, however, point out that applicatives might be disfavoured in an accusative language due to an alternative method of topicalization available in such a language, namely passivization. Applicative constructions might therefore be more frequent in ergative alignment languages, a hypothesis that, however, remains unconfirmed although it seems to be the case that passives exhibit positive correlation with Benefactive applicatives. It also seems to be the case that the existence of causative morphology correlates with the presence of applicatives. Recall, however, that in some languages, such as Hulapai and Caquinte, there exists applicative/causative isomorphism in the sense that the applicative morpheme is also used as a causative marker.
Following these summaries of the individual chapters, I turn to a critical evaluation.
As I already pointed out in the beginning of this review, Peterson provides the reader with two examples of applicatives from Ainu and San Lucas Quiaviní Zapotec. Unfortunately, however, the two examples given in (1) at the beginning of chapter 1 are almost the only clear and illustrative examples of the applicative construction in the whole book. In addition, Peterson does not discuss whether or not his examples in (1) are ‘typical’ instantiations of the construction. Discussion of this sort would be very useful, as it in fact turns out that the applicative examples in (1) – Comitatives and Locatives – may in fact not be prototypical applicatives in a cross-linguistic perspective. Peterson points out throughout the book that it is Recipient/Benefactive/Malefactive applicatives that constitute a typical applicative construction. It therefore strikes me as surprising that the author has chosen Comitatives and Locatives as the introductory pair of examples.
When discussing his views on terminology, Peterson claims that ‘applicative constructions are constructions . . . which involve a participant that normally would not be instantiated in a core object relation but rather as an oblique of one or another sort, in a core (usually direct object) instantiation’ (p. 39). I found it confusing that although the author makes reference to the oblique counterpart of applicative constructions, he virtually never gives an example of the corresponding oblique. In fact, I very often gathered the impression that the applicative variant of a construction was actually the only possible one, given the systematic lack of comparison with obliques. However, it indeed seems to be the case that some applicative variants lack an oblique counterpart altogether. Peterson claims that the reason for this could be that such a variant has not arisen yet or that it (i.e. an existing alternative) has faded away in the course of the diachronic development of the language. I find the discussion in this section interesting, but highly stipulative as the author fails to substantiate this claim with suitable examples of diachronic nature. These claims therefore qualify more as a footnote at this point and require more careful study in the future.
Some additional terminological remarks are in order. Peterson seems to take for granted the notions of ‘core’ and ‘peripheral’ arguments, as there is no relevant metatheoretical discussion on these notions in his book. It is also unfortunate that when discussing object properties in Bantu, the author does not make it clear what he means by ‘typical object properties’ or ‘symmetric behaviour’. Nor does he give an explicit account of the diagnostics for objecthood in Bantu. Object properties are, however, discussed in some detail in chapter 3. As a matter of fact I find the author's discussion of the retention and relinquishment of object properties (of base objects in applicatives) extremely enlightening also in a more general perspective, as the discussion gives a good overview of the complex spectrum of ‘objecthood’, specifically the fact that objecthood does not depend solely on one specific property, e.g. object agreement or passivization, but includes characteristics such as cross-clausal reference and antipassivization.
The two case studies conducted on the basis of two genetically-unrelated languages (Bukusu & Hakha Lai) in chapter 2 is admittedly an excellent way to introduce the reader to the multifaceted nature of applicative constructions. The exposition of the similarities and the differences between various types of applicative constructions is clear and informative. I believe, however, that the discussion would have greatly benefited from more examples with NP-arguments. It seems to me, especially concerning Hakha Lai, that an overwhelming majority of the examples cited involve two pronominal arguments rather than full NPs. This, on the other hand, becomes understandable when in the subsequent sections the reader of the book is informed that in fact applicative constructions in this language habitually involve pronominal arguments. However, until the reader receives this crucial piece of information, he is left wondering about the absence of full NPs in Hakha Lai applicatives. An additional noteworthy expository problem related to the examples with only pronominal arguments is that that the pronominal arguments are often encoded as object agreement markers in the verb. This means that the lexical material in the examples becomes reduced to essentially a single word – the (morphologically complex) verb. I do not find such examples very illustrative or pedagogically well-selected given that it is likely that a prospective reader does not have – even rudimentary – knowledge of the languages under discussion in Applicative constructions.
Some further minor remarks are in order at this point. First of all, Peterson's claim that there is ‘no Case marking’ in Bukusu is very surprising. I understand the author's affiliation with a more or less conservative branch of ‘basic linguistic theory’, but he should nevertheless be aware of the fact that claims on ‘non-existence’ of Case-marking in Bukusu might sound naïve to readers with background in theoretical linguistics. Another awkward point – consistent, however, with Peterson's conservative approach to language study – is that the author does not seem to acknowledge the existence of zero instantiations of applicative morphemes in human language. This line of reasoning is conspicuously inconsistent with the author's frequent references to ZERO ANAPHORA (p. 55), ‘zero-marking of objects in Wolof’ (p. 108) or that ‘3rd person patients are zero-marked in Hakha Lai’ (p. 16) (emphasis added). The author is thus obviously aware of the existence of zero-marking in human language in various forms, but he still eschews the possibility that an applicative construction might be zero-marked as well.
The neglect of null applicatives precludes interesting discussion of English-style double object constructions as potential applicative constructions (cf. Marantz Reference Marantz and Mchombo1993). Peterson disqualifies the English double object construction from being an applicative construction simply by not acknowledging covert applicatives as ‘genuine’ applicatives. I find it a serious shortcoming of the discussion in the whole book that the author has not referred to (nor included in the references) the work by Alec Marantz, especially Marantz (Reference Marantz and Mchombo1993) mentioned above, where zero applicatives and the potential affiliation to English-style double object constructions are discussed in depth.
On the other hand, I agree with Peterson in that ‘overt morphological marking’ undoubtedly facilitates linguistic inquiry, but ‘overtness’ should certainly not be regarded as a pre-requisite of such an inquiry in contemporary linguistics. In addition, the author is surely wrong in claiming that dative shift is not especially productive in English and that ‘dative shift constructions may not be adequately described in the languages that have them’ (p. 39. Such claims sound especially bizarre given that the author actually ignores a bulk of prestigious literature where the subject has been studied in depth (Oehrle Reference Oehrle1976, Stowell Reference Stowell1981, Pinker Reference Pinker1989, Levin Reference Levin1993, Marantz Reference Marantz and Mchombo1993 and many others).
Concerning the morphosyntactic and semantic variation in applicative constructions (the essence of chapter 3) the author seems to take it for granted that Benefactives and Malefactives are two sides of the same coin. However, while Benefactives might be associated with ditransitive predication, Malefactives might not, although Malefactives could arise from similar morphosyntactic resources in the form of low applicatives (cf. Pylkkänen Reference Pylkkänen2002). Some discussion of ethical datives would also be in order, given that it has been noted (see König & Haspelmath (Reference König, Haspelmath and Feuillet1998) for extensive discussion of e.g. Chichewa) that ethical datives are in fact encoded with applicative morphology in various languages.
I would like to add that the author questions the empirical correctness of Pylkkänen's distinction of high/low applicatives and McGinnis's (Reference McGinnis2001) phase theory related to high and low applicatives by citing Hakha Lai as a potential counterexample to these theories. According to Peterson, Hakha Lai applicatives qualify as ‘low’ in being asymmetric, yet applicative objects can apparently control depictive predicates in this language and applicativization can also target unergatives as well, a fact that suggests that Hakha Lai applicatives are ‘high’ after all. While the discussion might be on the right track, it seems unfortunate that Peterson does not corroborate his tentative counter-arguments with a pair of practical examples from Hakha Lai. Some examples of this sort are provided in Peterson (Reference Peterson2006), but the reader should not be obliged to recourse to the author's previous work in order to find the necessary examples. However, even in Peterson (Reference Peterson2006) the author does not provide the reader with a (much desired) example on applicativized unergatives in Hakha Lai.
Another type of applicatives overlooked in Applicative constructions is the possessive applicative. I find the lack of discussion of possessive applicative constructions a significant shortcoming, as possessive applicatives are widely attested as Possessor Raising constructions (König & Haspelmath Reference König, Haspelmath and Feuillet1998, Haspelmath Reference Haspelmath, Payne and Barshi1999, Pylkkänen Reference Pylkkänen2002) and applicative morphemes might, in fact, occur even internal to possessive noun phrases (see Lecarme Reference Lecarme, Lecarme and Guéron2004 for an interesting discussion of Somali).
As for the remainder of the book, I would first like to draw attention to chapter 5 where Peterson discusses the evolution of applicative construction with reference to the grammaticalization theory. This chapter is undoubtedly the pièce de résistance of the book; the discussion is not only extremely interesting for any linguist regardless of their theoretical affiliation due to its unique nature, but it is also clear, well-written, concise and credible in a general typological perspective due to the many genetically-unrelated languages discussed. This kind of ‘general applicability’ would be much called for also in the preceding chapter where Peterson discusses the discourse functions of applicatives. The author only concentrates on two arbitrarily selected languages (Hakha Lai and Wolof) and the reader is left questioning whether or not the conclusions drawn on the basis of these two languages would be applicable more widely.
Applicative constructions constitutes a valuable contribution to the study of applicative constructions in a cross-linguistic perspective. In fact, regarding the cross-linguistic scope of the content, the book is undoubtedly one of a kind. In addition, due to the descriptive and typological nature of Peterson's book as well as the wide range of topics discussed in this work, Applicative constructions has somewhat encyclopaedic character and will be of great interest to any linguist interested in applicative constructions (or verbal valence alternations in general), irrespective of methodological preferences. To a large extent I found the book well-written and informative, despite the fact that the topics discussed in various places do not always thematically correspond to their chapter-affiliations. As a whole, the reading experience can be characterized as insightful and very educational.
That said, I would probably not recommend Applicative constructions to someone without any previous knowledge of the syntactic and semantic properties of applicative constructions, despite the fact that the virtues of the book are far greater than its shortcomings. In fact, it feels at times that the reader must be familiar with the author's previous work on the subject in order to fully follow the discussion and the line of argumentation. In fact, concerning the low/high applicative typology introduced in Pylkkänen (Reference Pylkkänen2002) and its potential shortcomings in Hakha Lai, Peterson (Reference Peterson2006) is worth reading along with the book for a more detailed account of the subject.
