The syntactic licensing of ellipsis

The syntactic licensing of ellipsis. By Lobke Aelbrecht. (Linguistik aktuell/Linguistics today 149.) Amsterdam: John Benjamins, 2010. Pp. xii, 230. ISBN 9789027255327. $149 (Hb).

Reviewed by Dimitrios Ntelitheos, United Arab Emirates University

This book attempts a detailed study of ellipsis, i.e. the non-pronunciation of a linguistic string whose semantic contribution can be recoverable from the context. The book further attempts to provide an explanation on why certain types of ellipsis exist in some languages but not in others and what exactly is the syntactic mechanism that licenses ellipsis.

Ch. 1, ‘What is ellipsis?’ (1–16), provides a general discussion of ellipsis as a linguistic phenomenon, including a detailed literature review of both structural and non-structural approaches to ellipsis. In particular, two structural approaches—null proform and deletion at phonological form (PF)—that dominate the field are discussed. The chapter ends with a discussion of different restrictions on elided material related to how this material is recoverable and how the ellipsis is licensed.

Ch. 2, ‘Dutch modal complement ellipsis’ (17–86), introduces the different syntactic properties of Dutch modals. A number of diagnostic tests show that Dutch modals behave like raising verbs and not like control verbs. Complements are shown to be reduced clausal strings that contain a vP layer and tense projection. Finally, the categorial status of the modals is shown to be Mod/V head rather than auxiliary or inflectional heads. A examines more closely the properties of Dutch modal complement ellipsis, demonstrating that this type of ellipsis is only allowed with root modals and affects a complete constituent. A number of extraction tests reveal that extraction out of the ellipsis site of Dutch modal complements is possible with subjects but restricted with objects and impossible with adjuncts. Additionally, ellipsis is possible when the subject is an expletive if the associate is not deleted. Mismatches in form between the elided material and its antecedent are allowed to some extent.

Ch. 3, ‘Ellipsis licensing’ (87–156), introduces a theory of ellipsis based on the notion of Agree. A proposes that ‘ellipsis is licensed via an Agree relation between an ellipsis feature and the ellipsis licensing head’ (87). When this head enters the derivation, ellipsis takes place instantaneously and blocks additional computations from occurring within the ellipsis site. The feature checking mechanism results in not allowing the lexical insertion of material at the level of PF. The analysis explains previously observed extraction contrasts between subjects and objects in Dutch ellipsis sites.

Ch. 4, ‘Extending the analysis to other ellipses’ (157–206), applies the author’s analysis to other cases of ellipsis, including sluicing, English VP-ellipsis, pseudogapping, and British English do.  In particular, the analysis predicts that all material that escapes the ellipsis site before the ellipsis licensing head merges is available for further computations. On the other hand, material trapped within the ellipsis site is deleted and thus, is not available to further operations in the narrow syntax.

Finally, Ch. 5 ‘Conclusion and issues for further research’ (207–14), presents A’s concluding remarks and lists a number of issues that require additional study. Most notably, issues are left open concerning null complement anaphora that are traditionally assumed to involve null pronominal forms, as well as the cases of stripping, gapping, fragments, and nominal ellipsis that also involve elided material of some sort.

The book is a valuable addition to the discussion of syntactic licensing mechanisms in different cases of material that is not phonetically realized.