Language Evolution: The Windows Approach addresses the question: "How can we unravel the evolution of language, given that there is no direct evidence about it?"
The LINGUIST List is dedicated to providing information on language and language analysis, and to providing the discipline of linguistics with the infrastructure necessary to function in the digital world. LINGUIST is a free resource, run by linguistics students and faculty, and supported primarily by your donations. Please support LINGUIST List during the 2016 Fund Drive.
AUTHOR: De Belder, Marijke TITLE: Roots and Affixes SUBTITLE: Eliminating Lexical Categories from Syntax SERIES: LOT dissertation series YEAR: 2011 PUBLISHER: Netherlands Graduate School of Linguistics / Landelijke -- LOT
Mercedes Tubino, Department of English, University of Seville (Spain)
This monograph develops a new syntactic/morphological model that stems from the revision of the two main current models, Distributed Morphology (DM, Halle & Marantz 1993) and the Exo Skeletal Model (Borer 2003). Like its two predecessors, this is a Late Insertion Morphological model that assumes that lexical material enters any derivation only after the narrow syntax part of the derivation is concluded. De Belder’s doctoral thesis goes a step further in separating the functional from the lexical part of any derivation in that it: (i) eliminates categorial heads (n, v, a) from syntax under the assumption that lexical categories are not language primitives; (ii) defends a view in which what corresponds with lexical material (known as Root Terminal Nodes in DM (Harley 2009) and l-morphemes (Harley & Noyer 1999, 2000)) is empty of any content in syntax.
Chapter 1 introduces the book’s research questions as well as the theoretical and empirical basis assumed throughout the monograph. De Belder first raises the double question of whether features of lexical vocabulary items (VIs) and categorial heads -- two elements generally assumed in DM -- are really active in narrow syntax. The aim of her thesis is to answer this question negatively. She argues that lexical vocabulary items cannot contain features linking them with particular syntactic contexts. For instance, she gives the example of the VI 'dog', which she claims cannot contain such features due to its count and non-count uses: (2) Three dogs (count use, p. 20) vs. (3) There is dog in the soup (non-count use, p. 20). She claims that, while both examples involve the same VI, ‘dog’, the syntactic context is what determines its meaning as count or non-count. On the other hand, she argues that categorial functional heads do not exist in syntax, contra other approaches such as DM. She claims that derivational affixes, many of which are traditionally seen as the phonetic realization of categorial positions, are in fact instantiations of Root Terminal Nodes. For De Belder, the categorial interpretation of words as nouns, verbs or adjectives is a by-product of functional structure dominating roots, not the result of categorial heads merged above such roots. For this author, there is a fundamental difference between lexical and functional material; only the former is malleable and subject to coercion. She will base many of her contrasts on this fact.
In Chapter 2, De Belder develops a theory of the syntax of roots, which is a revision of previous late insertion accounts. This chapter is the result of her joint work with Jeroen van Craenenbroeck (2011). She adopts Borer's view that roots are void of syntactic features, so they cannot select for particular categories (contra DM's approaches such as Harley & Noyer 2000). De Belder defends the view that roots are defined structurally and that all VIs, including those corresponding to root positions, are subject to competition (e.g. they are all subject to the Subset Principle, as in Halle 1997), although in the case of lexical VIs, their insertion finally occurs by free choice. This differs from both traditional and revised DM approaches. For instance, Halle & Marantz (1993) claim that only VIs that are inserted in functional positions are subject to competition, whereas the insertion in root positions is done by free choice. Recent DM approaches (e.g. Siddiqi 2009) argue in favor of competition of VIs to occupy root positions, but they also assume that lexical VIs contain features. De Belder rejects these accounts on these grounds, although she adopts a late (as opposed to early) insertion approach. In the first part of this chapter, she argues that only a late insertion approach to morphology can explain why we may find functional VIs occupying root positions, e.g., (7) Martha is mijn tweede ik, 'Martha is my best friend, lit. Martha is my second I' (p. 42). These examples also show that roots are defined structurally and not because of their inherent properties (e.g. the first person pronoun, ik 'I', does not carry person features with it when used as a root). In the second part of the chapter, De Belder puts forward the syntactic model that will be used throughout the book, which involves a revision of previous syntactic models on Merge (Chomsky 2000 and subsequent work). She follows Zwart (2009) in noticing that previous accounts on Merge involve an asymmetry; whereas primary Merge involves selecting two elements from numeration at once, subsequent Merge operations involve selecting one object at once. She proposes a system based on Unary Merge in which Primary Merge consists of one object selected from Numeration that is combined with an empty derivation (i.e. the empty set). She argues that this empty syntactic position is what corresponds with root positions. This implies that each root in a syntactic structure implies a new derivation, although the specific details of this mechanism are not discussed until Chapter 5. In the last part of the chapter, De Belder proposes a modification of the Subset Principle to ensure the insertion of functional VIs in root positions through competition. This happens if we assume that the phonological exponent of a VI results from multiplying the featural content of the syntactic position it will occupy by the features the VI contains (any multiplication by zero will void the featural content of the VI, so any features originally contained in the functional VI will not block their insertion in root nodes). She finishes the chapter by addressing potential problems and other accounts (e.g. Harley 2009; Pfau 2009) she terms as 'late insertion approaches with a limited type of early insertion'.
In Chapter 3 De Belder shows that functional structure is sufficient to define categories. She first presents a case study on the countability of nouns to show that fine-grained distinctions in the nominal domain are determined syntactically. That is, she shows that the fact that the same VI (e.g. water) may be used as count and non-count is the result of syntactic features and their corresponding functional projections rather than inherent properties of the VIs. The relevant syntactic features accounting for the different readings of nouns are [Num] and [Size]. Thus, a VI receives a 'mass' reading if it is inserted in a syntactic position void of either feature; a VI receives a 'kind' reading if [Num] is contained in its insertion site, and if both [Num] and [Size] are contained in its insertion site, then a VI receives a 'unit' reading. She also argues that a syntactic position containing [Size] only is not possible. Possible restrictions associated with a count or non-count interpretation of some VIs are placed on extralinguistic (e.g. conceptual) knowledge. In the second part of the chapter, De Belder discusses the case of semi-lexical VIs that may have both a functional and a lexical use, such as the Dutch word ‘heel’ ('whole'), which may function as both an adjective and a universal quantifier. She argues that their different denotation is derived from the terminal nodes this VI may realize as the complement of an adjectival projection, if it is an adjective. As a quantifier, it originates in Sizeº, moves up to Numº and ends up in Dº. As a lexical VI, it gets its interpretation entirely from the Encyclopedia (their denotation is learned and arbitrary), and as a functional VI, it gets its interpretation from the functional features it realizes in syntax.
In Chapter 4, De Belder discusses why categorial heads do not exist as primitives of language. Derivational affixes have been long assumed to realize categorial heads. De Belder, however, discusses the possibility that they realize ‘something else’ and argues that her approach is more advantageous, especially when it comes to affixes associated with more than one category. In the first part of the chapter, De Belder puts forward a theory of homonyms, which she identifies as affixes with the same phonetic realization that are listed as separate vocabulary items. She presents three empirical tests to distinguish between homonyms and different instances of the same affix based on allomorphy, based on whether the different instances of the affix have the same synonyms and co-occurrence. She illustrates the section with the Dutch suffix -er, which she concludes is the same affix when it is a deverbal and a denominal agentive suffix, but not in the case of pluralctional -er, which is a case of homonymy. In the second part of the chapter, De Belder argues that multicategorial affixes cannot be the realization of categories, hence categories are not primitives of grammar. She argues that the fact that her model does not rule out examples such as ‘*to ugliness’ is not a problem for the claim that categories do not exist in grammar because approaches that assume categorial heads have the same problem, since the conversion mechanism that explains cases such as ‘to proposition’, a derived noun used as a verb, also fails to rule out ‘*to ugliness’. She concludes that her model is advantageous because even though both proposals are equally inadequate in explaining the ungrammaticality of ‘*to ugliness’, hers is simpler by doing away with mechanisms such as conversion.
In Chapter 5, De Belder provides an alternative approach to derivational word formation. She argues that since the meaning of derivational affixes is lexical, they realize lexical positions (i.e. they realize roots), but like other VIs that realize root nodes, derivational affixes may be semi-lexical too (i.e. lexical VIs that realize functional nodes). She presents the case study of Dutch collective mass nouns (e.g. ondergoed ‘underwear’) to illustrate this point. She argues that these nouns are derivational word forms that resist plural (a sign that they realize functional positions), whereas their collective reading is derived from the lexical meaning of the affixes. In this chapter, De Belder also proposes that derived words are concatenated root nodes, where the base VI realizes the lowest root node and the affixes realize the highest ones. According to De Belder, root nodes are defined by the absence of any syntactic features (they are empty derivations that are merged with objects from Numeration in Unary Merge fashion). Since two concatenated root nodes cannot result from this mechanism, De Belder proposes that words are the result of (i) Merge of an object (F, according to De Belder) from Numeration with the empty set/derivation; (ii) The newly merged complex object is returned to Numeration; (iii) Merge of the complex object from Numeration with the empty set/derivation, and so on. Each instantiation of the empty set in a derivation will correspond to a ‘lexical’ VI. The remainder of the chapter consists of different instantiations of the model.
In Chapter 6, the author discusses the theoretical repercussions of her proposed model and briefly addresses further issues associated with her proposal. For instance, she suggests that what we know as lexical categories are in fact defined by groups of features such as [Size] or [Num], in the case of nouns. She briefly suggests that this can be extended to verbs, which would be the result of cognitively recognizing information about time, modality, and aspect in conjunction with argument introducing projections. Another issue briefly treated in this chapter is, for instance, the elimination of the distinction between inflection, derivation, and compounding, which should now be viewed as the distinction between functional and root terminal nodes.
De Belder’s monograph makes an outstanding contribution to syntactic and morphological theory, as well as to the analysis of the Dutch nominal domain, with plenty of examples illustrating all the issues treated, especially regarding the derivation of Dutch nominal word-forms. The author raises important questions regarding many assumptions existent in present morphological theory and effectively challenges them as she searches for a simplified apparatus that would more effectively account for morphological processes in language, such as derivation and the existence of derivational affixes.
I did have some concerns regarding certain issues, both in content and presentation. First, the monograph would benefit from a revised organization, since the reader is constantly referred to later sections for information that may as well have been discussed all at once. The presentation of the theoretical model proposed is especially problematic, as De Belder covers some parts in Chapter 2, and leaves the rest for Chapter 5. Although the reader is referred to Chapter 5 for further information on how the model works, a full description of the model right at the beginning would make the proposal clearer and save the reader some unnecessary questions.
Second, the proposal is supported throughout with interesting and new data as well as illustrations of different phenomena affecting the nominal domain. However, this discussion would be stronger if examples were provided from languages other than Dutch (it just includes a few additional examples from English). Also, while the proposal defended in this thesis is intended as a theoretical model for grammar, the discussion is limited to the nominal domain, while its application to the verbal domain is only suggested, but not addressed, at the end of the book.
The model proposes as one of its main points the elimination of categories as part of grammar. However, traditional cases of nominalization or verbalization are not specifically, but just vaguely, addressed in the present model, which may leave the reader wondering how they would be resolved.
The main problem I see with this model is the fact that the insertion of Vocabulary Items in root nodes (corresponding with elements void of any phonological, syntactic or semantic information in this model) is completely unrestricted, as the author herself admits. She suggests that the restriction on VIs is to be placed on extralinguistic/conceptual knowledge and convention, an explanation that seems unsatisfactory; while the concept associated with Vocabulary Items is clearly extralinguistic, Vocabulary Items themselves are linguistic representations. Their association with particular syntactic positions needs to be restricted somehow.
Finally, the monograph would have benefited from more careful proof-reading, as the text contains a considerable number of misspellings that may be distracting to the reader.
In sum, De Belder makes an excellent contribution to syntactic and morphological theory, challenging many ideas widely assumed (i.e. the existence of syntactic categories as a primitive of grammar). She effectively challenges these assumptions by showing the reader that a simplified model may be equally explanatory, hence more advantageous than other models that need to resort to more complicated apparatus to explain the same phenomena.
Borer, Hagit. 2003. Exo-Skeletal vs. endo-skeletal explanations. In John Moore and Maria Polinsky (eds.), The Nature of Explanation in Linguistic Theory. Chicago: CSLI and University of Chicago Press.
Chomsky, Noam. 2000. Minimalist inquiries: the framework. In Roger Martin, David Michaels and Juan Uriagereka (eds.), Step by Step: Essays on Minimalist Syntax in Honor of Howard Lasnik. Cambridge: The MIT Press.
De Belder, Marijke, & Jeroen van Craenenbroeck. 2011. How to merge a root. Ms., HUBrussel & Utrecht University.
Halle, Morris, & Alec Marantz. 1993. Distributed Morphology and the pieces of inflection. In Kenneth Hale and Samuel Jay Keyser (eds.), The view from building 20. Cambridge: MIT Press, 111-176.
Harley, Heidi. 2009. Roots: Identity, Insertion, Idiosyncracies. Talk presented at the Root Bound workshop, USC, February 21, 2009.
Harley, Heidi and Rolf Noyer. 1998. State-of-the-Article: Distributed Morphology. Glot International 4.4: 3-9.
Harley, Heidi and Rolf Noyer. 2000. Formal versus encyclopedic properties of vocabulary. Evidence from nominalisations. In B. Peters (ed.), The Lexicon-Encyclopedia Interface. Amsterdam: Elsevier, 349-374.
Pfau, Roland. 2009. Grammar as processor. A Distributed Morphology account of spontaneous speech errors. Amsterdam: John Benjamins.
Siddiqi, Daniel. 2009. Syntax within the word: economy, allomorphy and argument selection in Distributed Morphology. Amsterdam: John Benjamins.
ABOUT THE REVIEWER
ABOUT THE REVIEWER:
Mercedes Tubino has a PhD in Linguistics from the University of Arizona.
She presently teaches English at the University of Seville (Spain). Her
primary research interests include syntactic theory, morphology,
lexico-semantics, Second Language Acquisition, and Hispanic and Amerindian