Date: Sun, 21 Apr 2002 12:56:58 -0400
From: David Golumbia
Subject: Alexiadou (2001), Functional Structure in Nominals
David Golumbia, New York, NY
Alexiadou provides a dense and powerful argument about a central issue throughout the history of generative grammar, the role played by nominalized verbs (for example, "destruction" in "the destruction of the city"), most famously as the subject of Chomsky (1970). The distribution of such phenomena seem to reveal a great deal about the structure of nominal phrases (NPs) and more generally determiner phrases (DPs), especially in languages like English that display several nominalization patterns. Alexiadou's study is emphatically cross-linguistic, using Modern Greek as its empirical base and providing examples from a range of sample languages including English, French, Spanish, Italian, German, Russian, Hungarian, Hebrew and Arabic, with briefer glances at several other modern and indigenous languages. This especially multivariate empirical base provides strong support for the book's argument, which is carefully integrated into recent generative theory.
Alexiadou argues that the distribution of nominals reflects abstract functional projections, especially DPs/NPs. This abstract level of representation determines the structure of apparent phrases, and not vice versa. Following Halle and Marantz (1993), Alexiadou suggests that the "syntactic categories N, V and A are actually morphological categories created by the syntax, i.e. post-syntactically realized" (211). Chapters 1 and 2 provide an overview of this argument; Chapters 3 and 4 provide a deep and impressive survey across languages to support the argument; Chapter 5 extends this survey to argue that the reliance on abstract functional projections reveals "certain similarities between nominals in nominative-accusative languages and certain patterns found in ergative languages" (167), ergativity taken here in the spirit of the best-known study of the pattern, Dixon (1994).
1. Introduction (1-26)
The Introduction provides a brief summary of the argument to come. The goal is to understand "certain aspects of the internal structure of DPs" (1) beyond simply the nominalizations indicated in the title, and in common with other recent transformationalist studies. Alexiadou surveys the types of nominals in English, breaking them down into three categories:
(1) Gerunds (e.g., "John's criticizing the book")
(2) Derived Nominals (e.g., "The barbarian's destruction of the city")
(3) Mixed Nominalizations (e.g., "Belushi's mixing of drugs led to his
Situating her work within a Distributed Morphology (DM) framework, following Halle and Marantz (1993) and Marantz (1999), she argues that these nominalization patterns can best be understood via "an approach to word formation according to which lexical elements, unspecified for syntactic category are introduced into variable syntactic environments" (7); "on such an approach, functional heads fully determine the category of a lexical head." This allows a satisfyingly abstract approach, in that similar mechanisms are thought to determine the pieces of linguistic phenomena regardless of the functional level at which they appear.
Alexiadou follows Grimshaw (1990) among others in analyzing the event structure of nominalizations to distinguish between process and result nominalizations. She surveys eleven diagnostics for distinguishing between the types, of which the first three are: "process nominals denote an event; result nominals denote the output of an event"; "process nouns take internal arguments obligatorily, while result nominals never do" and "process nominals can take agent-oriented modifiers, while this is not possible with result nouns" (9-10).
Alexiadou contrasts her account with those of Grimshaw and Chomsky (1970), but in concert with more recent generative approaches. She argues that "there is a VP node within event/process] nominals" (14) which is not true of result nominals. She then presents her model for "the fine structure of process nominals" (16) via a functional projection tree structure (19) with around seven nodes. The DP head of this projection is said to be common to both verbs and process nominals, thus collapsing the apparent noun/verb distinction for certain kinds of surface forms; "the difference between result and process nouns relates to the presence vs. absence of certain verbal-like functional projections, much like the VP analysis of process nominals. However, it differs from this approach in that it allows for result nominals to take complements as well, since both types are derived from unspecified roots that can take internal arguments" (20). While this discussion forms the main part of the book, as Alexiadou notes, it leads fairly naturally toward discussions of ergativity that advance nominalization as a possible source.
2. The Functional Architecture of Nominalizations (27-76)
The first part of the main discussion returns to the structure of event nominals, specifically examining the nodes of the tree presented at the end of the introduction. In particular she surveys various language data to emphasize the presence of an Aspect Phase (AspP) and a light v/Voice Phase (vP). Recent work on the status of NP vs DP as the dominating element of the overt NP. Elements thought to be captured by this approach include the presence of agreement, number and gender. The status of Spec,DP as A vs A' is explored, and Alexiadou's polyglot approach is especially welcome here, since it has been argued that "a crucial difference between English and Greek is that in English Spec,DP corresponds to Spec,IP, thus being an A-position, while in Greek it corresponds to Spec,CP, thus being an A'-position" (30). Exploring the Greek DP, she demonstrates that "Greek shows a preliminary distinction between event vs. result nominals identified by the criteria of manner and aspectual modification" (46). She goes on to explore the properties of AspP and vP in process nominals, again pressing on the process vs. result distinction. "The difference between result and process nominals is not one of argument structure, since both nominal types can have complements ... but rather it relates to the presence vs. absence of functional layers that bring about process/event readings" (58).
While vP and AP layers are found in process nominalizations, there is an absence of phenomena related to Tense (as well as other inflectional categories such as Mood). More generally, "both sentences and NPs denote events, but the manner with which they denote is quite different [as] distinguished by their types. A proposition is seen as the result of applying tense to an event description. To this end, tense acts as a generalized quantifier over event descriptions. D takes this function in the nominal context" (62). This allows Alexiadou to account for the existence of languages (like some in the Salish group) with weak or no noun/verb distinctions, by "predicting that languages that lack Tense as a separate category should neutralize the distinction between nouns and verbs" (62). She reviews and rejects some arguments about the presence vs. absence of Tense in DPs and an objection to the question of licensing in arguments, and then reviews a manuscript proposal from Embick (1998) according to which "the presence of agreement is not statable as a property of the roots, but as a property of roots in syntactic environments" (68).
3. Intransitivity in Nominalization (77-126)
Chapters 3 and 4 present a detailed, step-by-step account of the architecture of process nominals. Chapter 3 focuses on the vP found in process nominals. She situates the argument by claiming that "event nominals are ergative constructions" (78), providing an extensive survey of examples from a wide variety of languages, showing that they appear to be uniformly either "passive" or unaccusative (88), according to a relatively theory-specific definition of passive (accounting for her consistent use of scare quotes around the term). Her demonstration shows that nominals in general fall into a passive/unaccusative inflectional category, so that "(a) process nominals of the destruction type are intransitive; (b) English 'passive' nominals are event nominals and thus have argument structure, and (c) intransitive nominals are not derived by a process of passivization. This is consistent with the assumption that transitivity is a derived property" (111). Returning to process nominals, then, she again invokes DM to propose that the "type of v found in process nominalizations is the one found in unaccusative structures, i.e. the one that does not project an external argument/agent" (112). She then glances at by-phrases (e.g., "the destruction of the city by the barbarians") noting certain patterns in the availability of this construction with nominals and arguing that "the by-phrase is strictly parallel to ergative case, if this is analyzed as a prepositional/lexical case" (119).
4. Variation in Functional Structure (127-166)
The architecture of nominalizations is explored in this chapter via an especially wide typological survey of nominalization patterns. Alexiadou accounts for variation in terms of the kinds of functional projections encountered in across languages, so that core features of nominalization can vary widely. She surveys differences in patterns of presence vs. absence of a Complementizer (C) level, of AspP, vP, number, D, and the presence of agents in Spec,DP. The survey is meant to demonstrate that "the presence of Aspect contributes to event specification, and moreover, it is linked with the availability of adverbial modification"; that "while the presence of v contributes to eventive interpretation, the transitivity of v regulates the agentive vs. non-agentive character of the nominal and the availability of the accusative case;" and "the A vs. A' status of D regulates the availability of transitive nominalizations" (157). In a brief Appendix (158-162) to this chapter, Alexiadou surveys nominal infinitives in Italian, Dutch and German, finding more evidence for the claim that "nominalized infinitives include AspP and vP" (158).
5. Nominalization and Ergativity (167-210)
In a sense, the preceding chapters have been building a base for the argument put forward in the book's fifth and final chapter. Here, Alexiadou turns to what is her main theoretical target, which is to craft a connection between the structure of nominalization and the development of ergative patterning in some languages. This follows a distinguished tradition in the generative literature, due in no small part to the obvious typological splits in world languages (especially between ergative-absolutive and nominative-accusative clause alignments), which present a challenge to models of universal functional structure. Alexiadou considers construals of ergative patterns in terms of Case and Agreement and the formation of the perfect, and then examines in detail genitive constructions and constructions with auxiliaries. She finds significant similarities in patterns of nominalization and ergative patterns in both Case and the formation of perfects. In an Appendix to this chapter (198-204), she follows a complex recent line of discussion about participle formation, finding that "it must be the case that adjectives and participles differ in that the former are bare roots, while the latter include certain layers of functional structure" (204).
6. Conclusions (211-213)
The brief conclusion largely reiterates material from the Introduction; again drawing attention to her DM foundation according to which "there are no differences between word formation and syntax" (213).
This is a fascinating, dense and important book that both provokes and to a lesser degree frustrates. It continues the important work of Alexiadou and her colleagues in various (especially European) groups of generative grammarians, and the polyglot base of her analysis is fascinating. The book speaks to what are some of the most central and challenging issues in the generative tradition, and knits arguments about the issues together to show how they have persisted throughout the literature. Her presentation of the evidence and arguments is succinct, interesting, and often extremely pointed with regard to the question at issue.
At the same time, this book suffers from issues one is familiar with in volumes from this publisher. There are a number of proofreading and copyediting errors, occasionally making the interpretation of phrases and clauses ambiguous. Alexiadou's English is very strong, but one nevertheless suspects it is not her first language, and certain key argumentative passages suffer from a lack of editing that might have made the passages clearer. This is usually a minor problem, but because this book runs up against some of the deepest issues in generative grammar, the reader cannot help wanting to understand the author's points as precisely as possible.
The book is persuasive, although there remain open questions even at its core; for example, and it is hard to say whether this is due to editing or not, it remains hard to keep Alexiadou's base observations about nominals clearly in sight at all moments. Does the process vs. event reading rest on the presence of functional layers, or are the layers simply options that can be present but are not coextensive with a lexical phenomenon? That is, there are hints of circularity around some of her main contentions, so that one worries that the theory is verging on description as it becomes complex enough to handle so much typological data. These are metatheoretical worries that Alexiadou, like many of the best generative grammarians, attempts to address as her argument goes forward, and one looks forward to more work along these lines with anticipation.
Chomsky, N. (1970). Remarks on Nominalization. In Chomsky (1972),
Chomsky, N. (1972). Studies on Semantics in Generative Grammar.
The Hague: Mouton.
Dixon, R. M. W. (1994). Ergativity. New York: Cambridge University Press.
Embick, D. (1998). Syntax and Categories: Verbs and Participles in the Latin Perfect. Ms., MIT.
Grimshaw, J. (1990). Argument Structure. Cambridge, MA: MIT Press.
Halle, M. and Marantz, A. (1993). Distributed Morphology and the Pieces of Inflection. In K. Hale and S. J. Keyser (eds.), The View from Building 20. Cambridge, MA: MIT Press.
Marantz, A. (1999). Creating Words Above and Below Little v. Ms., MIT.