This book "supplies a vocabulary of English words and idiomatic phrases 'arranged … according to the ideas which they express'. The thesaurus, continually expanded and updated, has always remained in print, but this reissued first edition shows the impressive breadth of Roget's own knowledge and interests."
AUTHOR: Geert Booij TITLE: Construction Morphology PUBLISHER: Oxford University Press YEAR: 2010
Saskia Schuster, Institute of German and Dutch Philology, Free University Berlin
SUMMARY ''Construction Morphology'' (CM) presents a synthesis of Geert Booij's writings on constructionist approaches to word formation and lexical units over the past years. A number of ideas and case studies have been published before, but are now gathered for the first time in a monograph to contribute to a detailed picture of the interaction of morphological and syntactic constructions.
In chapters 1 - 3, Booij develops his idea of how a lexicon should be conceived of from a constructionist point of view. According to him, the lexicon is not a ''prison for the lawless'' (Di Sciullo and Williams 1987), but a hierarchically structured entity that contains pairings of form and meaning, i.e. constructions in the sense of Goldberg (2006). These constructions may be idiomatic and/or conventionalized (complex) words and phrases on the one hand and partially underspecified constructional schemas on the other hand. Assuming constructions as a basic entity in the lexicon is motivated by the fact that complex linguistic items may have holistic properties that cannot be derived from one of their constituents (36ff.). For instance, exocentric compounds in French such as ''chauffe-eau'' (heatV-waterN 'water heater') may have an agentive or instrumental meaning although this cannot be assigned to any of its constituents, but is a property of the whole.
The lexicon is the place where all constructions -- may they be individual words or constructional schemas -- are related to each other in several ways. They may be related via instantiation: for instance concrete words such as ''skyper'' or ''baker'' instantiate the abstract constructional schema [[x]V er]N 'one who Vs'. On a more general level, AN compounds (with an adjective as first member) also instantiate the more general schema [[a]Xk [b]Ni]N for nominal compounds where the word class of the first member is not further specified. Moreover, there are part-of relationships which can hold between a base word and a word derived from it (as in the case of ''skype'' and ''skyper'') or between items with the same degree of morphological complexity: these can share the same idiosyncratic semantics although there is no common base in the lexicon, see Dutch ''padvind-er'' 'boy scout' -- ''padvind-ster'' 'girl scout' -- *''padvind-en'' 'be a scout'. The essential mechanism for linking specific constructions and general schemas is default inheritance: words inherit the specification of a particular property from the dominating node, unless their actual lexical entry has another specification concerning this property (27ff.) Through this, regularities and exceptionalities among related, hierarchically structured constructions can be explained.
According to Booij, morphological schemas can be considered to be basically word-based rules in the sense of Aronoff (1976). Moreover, he follows Jackendoff (2008) assuming that there is not a principled difference between words and rule (schemas actually in his terminology): A lexical entry is said to be more word-like if it is fully specified, and more rule-like if it contains some variables (15). Morphemes do not play a role in this constructional account as units on their own, but only exist as bound forms inside larger constructional schemas. CM thus is to be considered a kind of word-based morphology.
The essential mechanism for coining new items or schemas is unification: e.g. the verb ''bake'' is unified with the schema [[x]V er]N to coin ''baker''. Schemas might be unified among each other to coin new (and more elaborate) schemas, e.g. the unification of [de[X]V]V and [[x]N ate] V which makes it possible to derive ''decaffeinate'' immediately from ''caffeine'' without assuming a possible intermediate verb ''caffeinate''. Productivity, an essential issue in morphology, is accounted for by assuming subschemas, i.e. low-level schemas that are more specific in contrast to higher ranked, abstract schemas. They indicate productive subtypes of a general schema (51f.), e.g. nominal compounding in Germanic languages can be accounted for by a more specific subschema [[a]Nk [b]Ni]Nj with special properties (e.g. recursivity), while isolated examples of non-productive patterns are related directly to the general schema [[a]Xk [b]Yi ]Yj in the lexicon (e.g. N+V-compounds such as ''babysit'').
Chapters 4 - 8 provide case studies on phrasal structures that are functionally similar to complex words: quasi-incorporation, that is, noun-verb-combinations that display word-like properties though not being one single morphological item (chapter 4); particle verbs, a very common phenomenon in Germanic languages (chapter 5); the grammaticalization of progressive constructions with special focus on the ''aan het +INF''-construction in Dutch (chapter 6); adjective-noun combinations that function as phrasal names (chapter 7) and numerals as a mixed category of phrasal and morphological structure in Dutch (chapter 8). Booij mainly focuses on Dutch data, though he also relies on data from other languages from time to time.
Chapter 9 deals with what Booij terms construction-dependent morphology: morphological markers may get tied to specific constructions in the course of time and are reanalyzed as constructional markers (e.g. the former morphological marker for genitive -s which is now used in phrasal constructions such as ''The king of England's crown'' (219)). Chapter 10 is about the way in which phonological information is stored in the lexicon and deals with the question if abstract underlying phonological forms are necessary (238). Chapter 11 finally presents some issues that are rarely touched upon in the book and hints at future fields of research.
EVALUATION Construction Morphology aims at a ''better understanding of the relation between syntax, morphology, and the lexicon, and at providing a framework in which both the differences and the commonalities of word level constructs and phrase level constructs can be accounted for'' (p. 1). There are thus two distinct formal mechanisms, morphology and syntax, which differ in their domain of application: syntax deals with linguistic items above the word limit, morphology deals with linguistic items beneath it. However, they may serve the same purposes in language from which follows that lexicon and grammar (comprising both morphology and syntax) are not as neatly separated as it is assumed in strictly modular approaches (see Goldberg 2006).
In my view, the great strength of a constructionist approach to language lies in the fact that it allows us to capture morphological and syntactic structures that are functionally equivalent with one theoretical notion (that of ''construction''). Furthermore, meaning and form are both part of the analysis and are not considered separately, which is important because there is evidence that both interact in various ways (see Hay and Baayen 2005). The case studies presented in CM show exemplary phenomena where morphological and syntactic procedures complement each other as in the case of numerals or where there are systematic syntactic alternatives to non-productive morphological patterns. For example, in chapter 4 Booij presents phrasal N-V constructions like ''piano spelen'' 'to play the piano' as having special syntactic and semantic properties that allow these constructions to fill the (semantic) gap of verbal compounding which is unproductive in Dutch. According to the author, constructions that instantiate these patterns display a greater tightness than normal regular syntactic phrases and might be considered to be ''syntactic compounds'' (see 101f., 181f.): although not being single morphological items, they manifest both syntactic and lexical properties. Another example that makes clear that meaning should be included in the analysis are phrasal combinations of adjective and noun used as names, which are discussed in chapter 7. They also display properties that do not comply with their phrasal nature. For instance, they may not be modified without losing their kind-referring function: ''bruine beren'' 'brown bears' (ursus arctos) - *''heel bruine beren'' 'very brown bears'. As this class of expressions can be readily extended and is not a closed set of phrases, Booij proposes to capture these phrasal patterns by assuming a constructional schema with open slots for adjectives and nouns. This schema, then, can be unified with appropriate adjectives and nouns to coin a new phrasal name that displays the same properties (that is, among other things the impossibility to modify the adjective).
An interesting issue that deserves some more discussion in the book is the question of whether constructional schemas, which are thought of as being symbolic representations in CM, give a proper characterization of the speakers' linguistic knowledge. Booij discusses one possible alternative to this view, namely linguistic models that negate symbolic representations and rely on the concept of analogy. In a range of cases, this has proven to be very useful in order to explain linguistic behavior, see for instance the choice of linking elements in Dutch compounds (see Krott 2001). Rather than simply stating that there are several (partially) abstract schemas equal in terms of meaning and function, analogy might help explain why one special option is selected out of a pool of competing options (as in the case of the compound linking elements -s-, -en-, -Ø- in Dutch). Booij addresses the issue of schemas vs. analogy briefly in chapter 3 and at the end of the book, but leaves it finally open to further research, especially with regard to its psychological dimensions.
As is clear from my comments so far, Construction Morphology is more and something other than what the title suggests: Although morphological structures are dealt with at length in the first chapters, the middle part of CM (chapter 4 - 8) discusses multi-word units. Booij gives two reasons for this: first of all, they are functionally similar to complex words. Second, he stresses that syntactic structures might also perform morphological tasks, as can be seen from the passive in English (20 f.). Thus in fact, CM is a book about the interaction of morphological and syntactic constructions in the domain of word and word-like entities, and it presents a conceptualization of the nature and structure of the lexicon from a constructionist point of view. Also it is not about inflectional phenomena, as Booij indicates right at the beginning. A theoretical account of constructional morphology with respect to inflection and morphosyntax can be found in Gurevich (2006). In contrast to Riehemann (2001) who gives an account of derivation of adjectives in German, CM deals with both derivation and compounding.
All in all, this very interesting book fills a gap in construction grammar in regard to word formation. Booij succeeds in linking syntax, morphology and the lexicon in a coherent way and giving fascinating insights into the nature of the lexicon as a network of syntactic and morphological constructions with varying degrees of abstractness. Concerning its formal qualities, I really enjoyed the abundant use of examples, its minimally formalized presentation as well as its clear writing style, which is very characteristic for Booij's writings.
REFERENCES Aronoff, Mark. 1976. Word formation in generative grammar. Cambridge, MA: MIT Press.
Di Sciullo, Anna Maria & Edwin Williams. 1987. On the definition of word. Cambridge, MA: MIT Press.
Goldberg, Adele E. 2006. Constructions at work: the nature of generalization in language. Oxford: Oxford University Press.
Gurevich, Olga. 2006. Constructional morphology: the Georgian version. Stanford: Ph.D. dissertation, Stanford University.
Hay, Jennifer B. & R. Harald Baayen. 2005. Shifting paradigms: gradient structure in morphology. Trends in Cognitive Sciences 9, 342-348.
Jackendoff, Ray. 2008. Construction after construction and its theoretical challenge. Language 84, 8-28.
Krott, Andrea. 2001. Analogy in morphology: the selection of linking elements in Dutch compounds. Nijmegen: Max Planck Institut für Psycholinguistik.
Riehemann, Susanne Z. 2001. A constructional approach to idioms and word formation. Stanford: Ph.D. dissertation, Stanford University.
ABOUT THE REVIEWER
ABOUT THE REVIEWER:
Saskia Schuster is a research assistant in the department of German and
Dutch Philology at the Freie Universität Berlin and is currently working on
her thesis as part of the project “Words and phrases” (for more
Her research interests concern historical linguistics, construction grammar
and contrastive linguistics (Dutch, German, French).