This book is a revised version of Marc Felfe's (henceforth, MF) PhD dissertation. The main subject of the study is German Particle Verb Constructions (PVCs) with the particle ‘an’. The core meaning of the homophonous (and related) preposition ‘an’ is a contact relation between a Figure and the external region of a Ground (Stiebels 1996:89). The author claims that standard compositional analyses - starting out from the assumption of a minimal argument-structure for verb and particle - can predict approximately fifty percent of the overall transparent form-meaning pairs exhibited by PVCs with ‘an’. Such analyses would be then forced to assume further machinery and operations affecting argument structures. MF argues instead for a holistic approach, modeled in the framework of Construction Grammar (CG): the PVCs under scrutiny are claimed to result from the fusion of abstract argument-structure constructions and verb-classes.
Chapter 1 (“Einführung” [Introduction]) starts out with the observation that often one and the same verb can be used to build syntactically and semantically different PVCs with ‘an’. An example is given for the verb ‘tanzen’ [to dance], which appears in (at least) five PVCs. Two additional remarks follow: for one, although the presented constructions do not correspond to the primary, canonical uses of ‘tanzen’, speakers of German have no problems in interpreting them correctly. For another, if other verbs are used with the same constructions, the resulting event-types do not change. These premises enable the author to list the main research questions of the study: what licenses the combination of verb (classes) and ‘an’? How can the meaning contribution of each component be captured? What should a compositional analysis of the PVCs at stake look like? After a sketch of the analysis to be proposed, the structure of the book is presented to the reader, and the analysis of different corpora is described as the main empirical basis of the work (in line with the usage-based approach of CG).
Chapter 2 (“Eigenschaften von PVK” [Properties of PVCs]) illustrates the behavior of German PVCs with respect to word order (adjacency vs. separation of verb and particle; fronting of the particle), syntactic and semantic autonomy of its components ((im)possibility for the particle to appear in isolation and/or to be modified), and lexicalization and argument-structure changes. The author provides a brief survey of morphological and syntactic generative analyses of German PVCs, and shows that neither can do justice of all the properties listed above. MF claims that a crucial error of both approaches lies in the assumption of a strict modularity of the grammar, and thus prepares the ground for an analysis where no sharp boundaries are put between lexicon and syntax.
In Chapter 3 (“Kompositionalität” [Compositionality]), MF brings arguments against approaches that assume argument-structure-altering processes (functional composition, argument extension) as a central machinery for explaining the licensing of spatial particles and additional argument slots. For example, instead of deriving one argument structure from the other (e.g. by applicativization), the author argues for the existence of two separate argument-structure constructions, encoding different event-types. The analysis is then presented to the reader: PVCs with ‘an’ result from the fusion of argument constructions (form-meaning pairings consisting of argumental XP-slots and underspecified event structures) and verb classes. The fusion occurs via conventionalized semantic and pragmatic implicatures.
Chapter 4 (“Was bedeutet Konstruktionsgrammatik?” [What does Construction Grammar mean?]) introduces basic assumptions of CG. Different theoretical approaches situated within CG are compared with respect to issues like the nature of grammar, formalization, and predictive power. MF then presents the definition of 'construction' he adopts, and discusses varying parameters like formal complexity, generality, compositionality, and productivity.
Chapter 5 (“Verblexeme” [Verbal lexemes]) focuses on frame semantics, presented as an alternative to a sharp separation between linguistic and conceptual information associated with lexemes. Both cognitive premises and a detailed introduction to frame semantics are offered to the reader. As a further clarification, a frame is proposed for the German verb ‘tanzen’ [to dance]. A sample of the most frequent syntactic contexts where this verb occurs is then brought as an example of the syntactic filling of different frame slots.
In Chapter 6 (“Argumentkonstruktionen mit ‘an’” [Argument Constructions with ‘an’]), MF explains how argument constructions work. Syntactically identical PVCs with ‘an’ can be either homonymous instances of (semantically) different argument constructions, or polysemous instantiations of the same construction (via metonymy or metaphor). The author claims that many 'minimal pairs' of PVCs - traditionally analyzed as alternative realizations of one and the same argument structure - do not actually stand in relation to each other. In a second part, possible representation formats for form and meaning of argument constructions are critically discussed. MF opts for a formal side with case-marked XP-slots and a semantic side containing variables and a schematic characterization of the event-type.
Chapter 7 (“Einzelne Argumentkonstruktionen mit ‘an’” [Single Argument Constructions with ‘an’]) is, together with the following two chapters, the core of MF's analysis. Eight argument construction groups are presented to the reader (names, examples and translations follow):
(1) (a) Change of Place and State: ‘NPnom NPacc an-kleben’ [to glue, to paste NP] (b) Startup: ‘NPnom NPacc an-schalten’ [to switch NP on] (c) State of contact: ‘NPnom an-grenzen’ [to abut, to border] (d) Intensification: ‘NPnom an-steigen’ [to increase] (e) Directionality: ‘NPnom NPacc an-lächeln’ [to smile at NP] (f) Force vs. Counterforce: ‘NPnom gegen NPacc an-kämpfen’ [to fight against NP] (g) Partiality: ‘NPnom NPacc an-lesen’ [to read NP not until the end] (h) Perception: ‘NPnom NPdat NPacc an-merken’ [to notice NP on NP]
Each group comprises different but related argument constructions, which show varying degrees of productivity. Besides a detailed description and discussion of every construction, four digressions focus on issues pertaining to single groups (medial PVCs, lexicalization of the particle, problems in individuating instances of a given construction, and the function of ‘an’ as construction-specific marker).
Chapter 8 (“Das System der Partikelverben mit ‘an’” [The system of Particle Verbs with ‘an’]) provides a deeper scrutiny of how the eight main groups, the argument constructions belonging to each one, and finally the instances of these (i.e., the actual PVCs) stay in a relation to each other, from both a quantitative and a qualitative point of view. In the first part, quantitative results regarding the type-frequency are presented (by means of figures and charts) and discussed; in the second part, the author introduces the concept of 'inheritance link', which offers a way of formalizing hierarchical relations among constructions, between constructions and their instances, and even between more and less general instances.
Chapter 9 (“Fusion und Gebrauch” [Fusion and Usage]) addresses the issue of the integration of argument constructions and verb classes. After a digression on the great productivity of PVCs in German(ic), the author considers single argument constructions with respect to the verbs they can appear with. The fusion of constructions and verb classes obtains via usage-driven 'fusion-routines' (regulated by conventionalized implicatures), paradigmatic forms and contrast with other constructions. Argument constructions can either profile slots already contained in a verb frame, or can contribute arguments on their own.
In Chapter 10 (“Ein- und Ausblick” [Overview and Outlook]), MF recapitulates both issues and findings. He then points out additional desiderata for future research and proposes two practical applications of the study. An appendix follows, where all the PVCs taken into consideration are divided according to the argument constructions they belong to.
MF's investigation can be of interest to both the theoretically and the empirically oriented linguist. On the one hand, it addresses problems regarding the argument structure of particle verbs that have been (and still are) matters of theoretical debate (cf. McIntyre (2007), Roßdeutscher (2012), Svenonius (2003), Stiebels (1996)). On the other hand, the detailed classification of the data presents challenges for the implementation of computational resources (cf. also Springorum et al. (2012)). Moreover, attention is paid to old vocabularies and lexicons, which completes the synchronic corpus-based investigation with insights on the historical development of single PVCs. This makes the study profitable reading also for language historians. Since the book is written in German, however, it is accessible first and foremost to Germanists. This is an undesirable limitation, for the author's constructionist analysis features non-standard solutions that could stimulate an interesting discussion within the world-wide CG community. Therefore, the reviewer thinks that an additional, smaller publication in English would be useful for a better circulation of MF's work.
Besides giving a detailed classification of the PVCs at stake, the author undertakes an accurate scrutiny of the data in terms of degrees of productivity. The reader can thus acquire information about which argument constructions are fully productive, which ones are semi-productive, which ones build a niche, and which ones have been fully lexicalized. This is often a neglected topic in works (especially theoretical ones) dealing with German(ic) particles. Moreover, the constructionist analysis defended by MF fits the data nicely. As regards argument constructions encoding Directionality, for example (‘NPacc an-lächeln’ [to smile at NP]), both a typical pragmatic value (actions connoted as 'negative') and contrast formation are claimed to be crucial factors motivating productivity. The first factor explains occasional use of nouns and adjectives as bases for 'an-verbs' (e.g. ‘NPacc an-kröten’ [to call NP a toad] < N ‘Kröte’ [toad; contemptible person]): as MF points out, there seem to exist no denominal or de-adjectival formations exempt from this negative value. On the other hand, the entailment of 'partial affectedness' common to many 'an-verbs' of this class emerges as a result of a systematic contrast to verbs with the prefix ‘be-’, which denote 'totally affecting' actions (e.g. ‘NPacc an-grabschen’ [to grab NP] vs. ‘NPacc begrabschen’ [to paw NP]). In sum, this study confronts scholars of different theoretical persuasions with one important issue: how to incorporate a sustainable explanation of varying degrees of productivity in a given analysis of particle verbs (a contribution in this direction is Roßdeutscher (2012)).
A quite strong claim of this work is that there are hardly any semantic or structural restrictions on the usage of PVCs with ‘an’. The combination of verbs and argument constructions is governed by semantic and pragmatic implicatures, but these are not interpreted as constraints on possible PVCs. Rather, they allow for 'prototypical predictions' on expectable vs. unexpectable ones. This conclusion is supported by a case-study on the verb ‘schlafen’ [to sleep]. Although this verb does not belong to any verb class that normally combines with ‘an’, it can occasionally occur in some of the argument constructions listed in (1), yielding fully transparent structures. It is up to future research to falsify or confirm the author's thesis. If it turned out to be tenable, then the question to raise would be what exactly licenses this absence of constraints.
The book offers a different view of PVC-pairs like ‘Farbe (an die Wand) an-streichen’ [to spread varnish (on a wall)] vs. ‘die Wand (mit Farbe) an-streichen’ [to cover a wall (with varnish)]. Such pairs are usually analyzed as alternants where the particle ‘an’ licenses either a Figure or a Ground. From both a semantic and a syntactic point of view, the phenomenon bears similarity to the familiar 'spray/load'-alternation (cf. Stiebels (1996:105-6)). What MF argues for, on the contrary, is that the two structures appearing in the alleged ‘an-streichen’-alternation instantiate different argument constructions. The first one is an instance of Change of Place/State in (1a), the second one corresponds to the construction encoding Directionality in (1e). According to the author, then, no 'Ground-Promotion' (cf. McIntyre (2007)) obtains here. Real 'promotion'-phenomena do exist, but they are always licensed by a metonymy relation between the referents of two arguments. An example is ‘Erde an-schütten’ [to bank up earth] vs. ‘einen Damm an-schütten’ [to build up a bank], where there is a causal contiguity relation. The hypothesis that metonymy is what really triggers 'promotion' is interesting, and it would be worth testing with respect to other German PVCs where similar phenomena can be found (e.g. particle verbs with ‘aus’ [out], cf. McIntyre (2001: 275-9)).
In addition to the previous issues, German PVCs with ‘an’ also present difficulties in identifying the precise aspectual contribution supplied by the particle. In particular, this is the case for the argument constructions in (1e) (Directionality) and in (1g) (Partiality) (cf. Stiebels (1996: 162-5; 78-82)). With respect to PVCs encoding Directionality, the author convincingly argues that - at least in some cases - they can be characterized as unspecified for perfectivity (attained vs. unattained result). The construction encoding Partiality is discussed in detail, too (e.g. pp. 30-31; 156-9; 218-223). MF challenges the common assumption that ‘an’ here does not affect the argument structure of the verb, bringing examples where a (qualitative or quantitative) change is indeed the case. The aspectual import of this construction is extensively explained to the reader, but a precise formalization is not provided. This would have been a welcome contribution, since a formal description of this use of ‘an’ is hardly given in the literature (Springorum (2011) being an exception).
This shortcoming aside, MF's study succeeds in providing insightful observations that the literature on the topic still lacked. All central issues about PVCs with ‘an’ are tackled, and the proposed analysis covers the whole spectrum of meanings exhibited by the data. Although single aspects of the account may be criticized by scholars working in different frameworks, the overall picture appears to be complete and coherent. This makes this monograph a model for future case-studies on single particles. In sum, I recommend this book to all scholars interested in the investigation of German(ic) particle verbs.
McIntyre, Andrew. 2001. German Double Particles as Preverbs: Morphology and Conceptual Semantics. Tübingen: Stauffenburg. (Studien zur deutschen Grammatik 61).
McIntyre, Andrew. 2007. “Particle verbs and argument structure”, In Language and Linguistics Compass 1(4):350-397.
Roßdeutscher, Antje. 2012. “Hidden quantification in prefix and particle verbs”, in: Proceedings of Sinn und Bedeutung 16, MIT, (in press).
Springorum, Sylvia, Sabine Schulte im Walde and Antje Roßdeutscher. 2012. “Automatic Classification of German an Particle Verbs” In: Proceedings of the 8th International Conference on Language Resources and Evaluation. Istanbul, Turkey, May 2012.
Springorum, Sylvia. 2011. “Drt-based analysis of the German verb particle an”. Leuvense Bijdragen 97.
Svenonius, Peter. 2003. “Limits on P: Filling in holes vs. falling in holes” in Nordlyd 31.2: 431-445.
Stiebels, Barbara. 1996. Lexikalische Argumente und Adjunkte: zum semantischen Beitrag verbaler Präfixe und Partikeln. Studia Grammatica 39. Berlin: Akademie Verlag.
ABOUT THE REVIEWER:
Stefano Quaglia is a Ph.D. student at the Department of Linguistics, University of Konstanz. His research interests include Italian and Germanic particle verbs, the syntax and semantics of adpositional elements, and their interaction with the argument structure of verbs.