Biberauer Theresa Walkden George Syntax over Time. Lexical, Morphological, and Information – Structural Interactions Oxford Oxford University Press 2015 XVII 418 S.
The book, edited by Theresa Biberauer and George Walkden, is the 15th volume of the series ›Oxford Studies in Diachronic and Historical Linguistics‹, and contains selected papers (altogether 20) originally presented at the 12th Diachronic Generative Syntax Conference (Cambridge 2010). As described by the editors in the introductory chapter, the very object of investigation, syntactic change, is subject to debate: there is no consensus as to whether syntactic change exists as such, or whether it is the result of changes in other domains of language (cf. the Inertial Theory of Longobardi 2001). At any rate, the editors suggest that while syntax may remain an autonomous module, it is one of the core assumptions of the Minimalist Program that syntax is responsible to its interfaces and can be seen as an optimal solution to legibility conditions (cf. the Strong Minimalist Thesis of Chomsky 2000). Seen in this light, the editors name three major areas that have been relevant for syntactic change: changes in the lexicon, morphology, and the marking of information structural properties. The studies presented in the volume are organised into three parts that correspond to these major areas. The first part (›Syntax and the Lexicon‹) contains 6 papers, the second part (›Syntax and Morphology‹) contains 5 papers, while the third part (›Syntax and Information Structure‹) is the largest with 9 papers. The individual papers within each part are organised according to their subtopics, and the interrelatedness of the various studies is highlighted by the editorial introduction. While the individual papers address a wide range of topics, it was a particular pleasure to note that the editors invested some effort into making the volume coherent: thus, the entire book has an additional value beyond the sum of the individual contributions.
The first article in Part I, written by Caitlin Light (pp. 17–35), investigates expletive there in German and English. Light argues that German expletive da in Early New High German was inserted into [Spec,TP] if the subject was located elsewhere; this da is not available in Modern German, as it has been ousted by situational da located in a similar surface position, and facilitated by the availability of a null expletive from Old High German onwards. The behaviour of da in earlier periods of German shows several similarities with English there. The patterns with these elements are amply backed up with corpus data. Light convincingly shows that the EPP was active in German in previous periods, and suggests that it is so in Modern German, too: the lack of overt subjects in [Spec,TP] can be explained by the availability of a null expletive, though it remains unclear how one can exclude the possibility of the EPP being lost.
Joan Maling and Sigríður Sigurjónsdóttir (pp. 36–53) examine the so‑called New Impersonal Construction in Modern Icelandic. It differs from canonical passives in that the verbal object remains in situ (hence postverbal) and is assigned accusative case (unless it bears oblique case). They argue that reanalysis started from impersonal passives, which are syntactically ambiguous between passives and actives: in the former case the subject is empty, while in the latter case the subject is an arbitrary pro. Reanalysis results in impersonal passives appearing with reflexive verbs (in addition to intransitives), and the process is completed by their availability with transitive verbs. The paper is crucially important since the phenomenon refutes Inertial Theory in that syntax is not driven by changes in morphology. The examined process is ongoing in Icelandic at quite a fast rate, as backed up by ample data from two previous studies.
Focussing on be like quotatives, the paper written by William Haddican, Eytan Zweig and Daniel Ezra Johnson (pp. 54–71) is likewise devoted to a change in apparent time. Relying on the results of two acceptability experiments, the authors demonstrate that the acceptability of be like with stative and eventive quotes is primarily dependent on age, younger speakers being more innovative. Still, be like is different from the quotative verb say in that be behaves like an auxiliary and not a lexical verb, refuting the idea previously raised in the literature that eventive be like involves the reanalysis of be into a lower (lexical) verbal head. Instead, it is argued that this follows from the ability of the same be to create both stative and eventive readings. The syntactic innovation lies in the introduction of an underlying null demonstrative (something), which results in an approximate quote meaning.
Veronika Hegedűs (pp. 72–85) describes the grammaticalisation path of postpositional elements (Ps) in Old Hungarian, showing that several present‑day Hungarian postpositions originated as nouns, which were reanalysed into Ps via an intermediate stage when they were AxialParts. While some Ps are already in P in Old Hungarian, others show mixed characteristics: in particular, they may agree with a dative‑marked lexical DP, similarly to possessive constructions. The intermediate status accounts for these mixed properties, and its position between the DP and the PP projections conveniently yields a standard grammaticalisation process whereby lexical elements undergo movement and later reanalysis upwards in the structure, thereby becoming more functional. The argumentation is convincing, though it would have been advantageous to present more of the P‑system rather than just very few examples.
The investigation of Old Hungarian is also the interest of Katalin É. Kiss (pp. 86–101), who very convincingly argues that a negative cycle can be detected in Old Hungarian. Much like other well‑known instances of the Jespersen‑cycle, the introduction of a new negative element was triggered by the weakening of the original negative element. The interesting contribution of Old Hungarian is that it shows a split between morpho‑phonologically opaque and transparent negative elements originally composed of a negative and an indefinite: opaque elements always require the new negative particle in the clause, while it is still optional for transparent ones. The lack of transparency is due to word‑internal phonological processes that bleached the original morphological makeup; hence, syntactic change (ultimately resulting in a left‑peripheral negation head) is driven by morphology.
Ana Maria Martins (pp. 102–122) also examines a topic related to negation: this is the case of the indefinite quantifier algum ›some‹, which entails a positive polarity interpretation pre‑nominally and a negative polarity interpretation post‑nominally in contemporary European Portuguese. In Old Portuguese, however, word order variation was free in the sense that both orders allowed both interpretations. The post‑nominal order was initially derived by the noun moving to a higher specifier position; the first step of reanalysis involves the movement of the noun to the head algum, and this complex head moving subsequently to a higher, DP‑internal NegP. The second step involves the further movement to the highest functional head (D). In this way, the negative polarity item is formed syntactically. Martins also convincingly shows that the cognate elements in Italian and French have already lexicalised.
Part II is opened by a paper written by Chris Reintges (pp. 125–145), investigating the issue of morphological complexity and the relation between syntactic and morphological change. Contrary to the widespread assumption that syntactic change may be induced by morphological change but not vice versa, Reintges claims that the lack of verb movement to higher left‑peripheral positions in the history of Later Egyptian led to the availability of these positions to other elements, such as tam‑markers. In a similar vein, he argues that the activation of the topic field led to the lexicalisation of Top heads by particles. While the overall claim suggests that it is a landmark paper, and the individual claims are certainly valid, there are far too many issues constrained within the frame of a single paper, leaving many questions open for the reader, though the references to Reintges’s extensive work on the subject provide good orientation.
Adam Ledgeway (pp. 146–162) examines the reasons behind diverse patterns regarding complementiser‑drop in the dialects of the Salento. In central‑southern dialects, the absence of an overt cu with irrealis complements is phonological: the complementiser has a phonologically zero variant, which exhibits the same syntactic behaviour as its overt counterpart. Interestingly, in northern dialects the absence of cu has syntactic reflexes: Ledgeway convincingly shows that this is due to the absence of the CP‑projection; hence, the irrealis clause is deficient in this sense. The reinterpretation concerning the absence of cu in these dialects took place in parallel with the reanalysis of the phonosyntactic doubling of the initial consonant of elements following the complementiser (overt or covert): they have been reinterpreted as irrealis markers on T, and as all CP‑related features are marked in a syncretic TP, making CP unnecessary.
Synchronic differences with diachronic roots among dialects are also a key interest of Marit Julien (pp. 163–178), who investigates the reanalysis of the negated perfect into negated past in Sámi and Finnic. The change has been completed in Northern Sámi, her focus, but not in more peripheral dialects, such as Southern Sámi, and there are also transitional varieties. Julien argues that the change was initiated by the introduction of obligatory copulas in the present perfect, which ousted the affirmative pattern involving a single participle, leaving the negative pattern to be interpreted as contrasting with the negative present, a likewise copula‑less structure. Importantly, Julien shows that the participle retains its past tense meaning throughout: it changes from relative into absolute past tense. This induced a change in the functional hierarchy of tense‑related projections, contesting the strict hierarchy of cartographic approaches.
Krzysztof Migdalski (pp. 179–196) examines the history of cliticisation across Slavic, and argues that present‑day patterns are directly related to the presence or absence of the TP. Bulgarian and Macedonian have retained verb‑adjacent pronominal clitics, a pattern going back to Old Church Slavonic, and these languages have a TP: verb‑adjacent clitics adjoin to T. In other Slavic languages, such as Serbo‑Croatian, the relevant clitics move to so‑called second positions which Migdalski shows to be instances of separate specifier positions for each particle. The change in the position of the particles was induced by the loss of TP, which can be observed throughout the history of Slavic. The endpoint of the change is recent, and is in fact ongoing in the Montenegro dialect. The analysis is convincing, though the proposal regarding why clitics cannot move to the second position prior to the loss of TP remains tentative.
The paper written by Dimitris Michelioudakis (pp. 197–216) examines the emergence of the inherent dative case in Greek from Hellenistic Greek to Modern Greek, with particular attention paid to Cypriot Greek. In Hellenistic Greek, direct objects preceded indirect (dative) objects in ditransitive constructions, while Modern Greek is similar to English in that indirect objects precede direct objects, while PPs expressing the indirect object argument follow the direct object. Michelioudakis refutes the idea that the change was triggered by the loss of the morphological dative, as abstract dative case can be detected well after distinctive morphological marking was lost. In this sense, morphological change was merely a facilitator of syntactic change, which in turn was induced by changes in the interpretability of the relevant features. The argumentation is convincing, though at some points a bit too technical.
Part III starts with the investigation of Romanian pe by Virginia Hill (pp. 219–235), who shows that pe involved in Differential Object Marking was reanalysed from a preposition into a topic marker located on the functional left periphery of the DP. Hence, as Hill argues cogently, pe is not a case assigner but a marker related to marking certain information structural properties: contrastive topics in Old Romanian and familiar topics in Modern Romanian. The reanalysis from contrastive into familiar is tied to the appearance of dislocated clitic pronouns, which head the projection that has the pe‑DP in its specifier: the original purpose of pe was thus blurred. Hill’s contribution is crucial for linguistic theory for several reasons, the chief one probably being that it sheds light on the importance of the left periphery of DPs in terms of information structure for topics that remain in situ, hence without movement to the clausal left periphery.
George Walkden (pp. 236–248) examines the V2/V3 alternation in early West Germanic main clauses. He argues that the alternation results from the verb being able to move either to a higher C head (Force) or to a lower one (Fin), and in the latter case there are certain topic projections that may intervene between the leftmost XP and the verb, hence also an information‑structural restriction on the intervening constituent. In Old Saxon, V3 was already a diminishing pattern, while its occurrence was relatively substantial in early Old High German, compared to later Old High German, which ultimately came to be V2 like Old Saxon; in Old English, V3 is well attested throughout. Walkden’s diachronic scenario is that a single type of change spread from Old Saxon to Old High German but not to Old English. His findings are vitally important both for diachronic theory and for the investigation of Old Saxon.
The chapter by Ed Cormany (pp. 249–264) investigates subject clitics in Friulano. The traditional assumption is that syntactic clitics grammaticalise from tonic pronouns via an intermediate step of weak pronouns (phonological clitics). Cormany’s extensive corpus investigation on historical texts shows that this is not the case in Friulano. Instead, he proposes that the high clitic al was probably introduced due to contact with the neighbouring Padovan dialect, where it was already a syntactic clitic. Apart from the gradual appearance of clitic doubling with al, the stepwise extension (from tonic pronoun subjects to DPs and QPs) can be observed, which Cormany attributes to the conflation of left‑peripheral layers originally hosting specific types of subjects. While the analysis for al is plausible, the question of low subject clitics remains unexplained, even though these are also claimed to lack an intermediate weak pronoun stage.
Lieven Danckaert (pp. 265–279) examines left‑peripheral presentational foci in earlier Latin, arguing that both its availability and later decline are tied to whether vP‑movement was possible. He considers embedded clauses only, where certain phrases (of diverse syntactic categories, but invariably presentational foci) could be fronted to FocP above overt complementisers that he locates in Fin. He argues that this was possible as long as the entire vP moved to a functional projection above TP, while the tensed verb itself landed in T; the focussed XP could undergo subsequent movement. Without vP‑movement, the focus may move only as far as a lower FocP in the vP‑periphery, while moving as part of a larger phrase does not violate locality. Due to the loss of vP‑movement via reanalysis processes, high presentational foci were lost; it remains unclear whether such foci are then located in the lower FocP or may remain in situ.
Examining Spanish and Catalan, Montserrat Batllori and Maria‑Lluïsa Hernanz (pp. 280–298) show that weak (non‑contrastive) focus fronting was widely available in both Old Spanish and Old Catalan, while there is an asymmetry in the modern languages: Modern Spanish still allows it, while Modern Catalan does not. However, QP‑fronting is still possible in Catalan. The authors argue that this is so because QP‑fronting was reanalysed as moving to a left‑peripheral polarity phrase, the existence of which they convincingly motivate by other instances of grammaticalisation into polarity markers. Batllori and Hernanz claim that Catalan lost the weak focus phrase projection, though it remains unclear what might have led to this change, and whether there is evidence for QP‑movement reanalysed as movement to the polarity phrase prior to the loss of the weak focus phrase projection.
Roland Hinterhölzl (pp. 299–317) investigates the OV/VO variation in Old High German, arguing that, just like Modern German, Old High German was an OV language, where OV order is derived from a VO base via leftward movement triggered by licensing considerations. Based on the examination of the ›Tatian‹ translation (and especially the differences it exhibits from the Latin original), Hinterhölzl shows that word order variation is closely related to information structural properties. Given elements (background) appear preverbally, while presentational foci occur postverbally; contrastive foci may appear in the immediate preverbal position, that is, in the specifier of a designated Focus phrase. The distinctions follow from restrictions holding in the syntax‑prosody interface, as described by Hinterhölzl’s metrical rules. The theory is convincing, though the diachronic change following Old High German is not fully worked out.
Variation in OV/VO orders in English is examined by Ann Taylor and Susan Pintzuk (pp. 318–335), who show that the appearance of postverbal objects in Old English is more frequent in clauses with Aux‑V than in ones with V‑Aux order, as shown by the results of an extensive corpus study. This is true for referential objects: non‑referential ones appear rarely postverbally. The authors argue that non‑referential objects appear postverbally only if they are base‑generated in that position, that is, when the underlying order is VO. Referential objects, on the other hand, may either be postposed in OV clauses (if they carry new information or count as heavy), or may be base‑generated there in VO clauses. The higher frequency of postverbal objects in Aux‑V clauses is thus due to there being two possible sources of VO, while in V‑Aux clauses it is always the result of movement, as V‑Aux goes only with OV, due to the Final‑Over‑Final Constraint.
Joel C. Wallenberg (pp. 336–349) addresses the issue of Heavy NP Shift (HNPS) across Germanic, focussing on its historical development in English. Since HNPS is generally associated with focus, Wallenberg’s own antisymmetric assumption is that the DP undergoing HNPS moves to a left‑peripheral Focus position, and the remnant of the TP moves to a Topic position above the Focus. Since this is independent from the internal structure of whether the TP and the VP are head‑final, the prediction is that the proportions of HNPS should be constant throughout all periods. This is, however, not the case: Wallenberg calculates the estimated proportion of HNPS for Early Middle English, which is significantly higher than the proportions for earlier and later periods. This is only predicted by the traditional rightward movement approach, where DPs may be right‑adjoined in two positions if the TP is head‑initial and the VP is head‑final.
The last paper, written by Edith Aldridge (pp. 350–370), investigates why certain pronominal objects shifted to a preverbal position in the context of negation in Archaic Chinese. Aldridge argues that a negative head was able to take a nominal complement (containing the verb), which lacked a functional v head that could assign accusative case, hence the object had to move up to the edge of the nominal domain to receive case from the negative head. While full DPs are spelt out in their base position, pronouns showing case distinction are spelt out at their landing site, where the morphologically marked accusative form is valued. As expected, object pronouns do not move upwards when a v head is present for independent reasons, and the pattern ceased to exist once morpho‑phonologically distinctive accusative forms disappeared from the pronominal paradigm. The analysis is clear and convincing, and has far‑reaching theoretical consequences.
In sum, the volume ›Syntax over Time‹ is of exceptionally high quality (both in terms of content and technical editing). While one can naturally identify some minor problems and inconsistencies, this does not affect the overall result. The authors have raised important questions that are relevant even beyond the scope of diachronic investigations, and provided answers (albeit sometimes only partial or tentative ones) that constitute a real and meaningful contribution to the field. The editors must also be praised in this respect. Apart from the editorial introduction, the organisation of the articles into three major parts, as well as the order of the articles within each part, helps the reader identify several – and not necessarily obvious – connections among the various chapters. The rich scientific output that manifests itself in this book will hopefully help the wider community of theoretical syntacticians realise that diachronic investigations, especially when amply supported by empirical data, are in fact central for the development of formal linguistic theory.
Chomsky, Noam 2000: Minimalist Inquiries: The Framework, in: Roger Martin [et al.] (eds.): Step by Step. Essays on Minimalist Syntax in Honor of Howard Lasnik, Cambridge, pp. 89–155.Search in Google Scholar
© 2012 Walter de Gruyter GmbH & Co. KG, Berlin/Boston