Memorization and the Compound-Phrase Distinction

Marcel Schlechtweg

• Publisher: De Gruyter
• Year: 2018
• ISBN:
  9783110568677
• Edition: 1
• Topic: Languages & Linguistics
• Subtopic: Linguistics
• Index: Languages & Linguistics

Over the last decades, it has been hotly debated whether and how compounds, i.e. word-formations, and phrases differ from each other. The book discusses this issue by investigating compounds and phrases from a structural, semantic-functional and, crucially, cognitive perspective. The analysis focuses on compounds and phrases that are composed of either an adjective and a noun or two nouns in German, French and English. Having distinguished compounds from phrases on structural and semantic-functional grounds, the author claims that compounds are by their nature more appropriate to be stored in the mental lexicon than phrases and supports his argument with empirical evidence from new psycholinguistic studies. In sum, the book maintains the separation between compounds and phrases and reflects upon its cognitive consequences.

1Introduction

This work aims at investigating structural, semantic-functional and cognitive aspects of compounds/compound-like constructions, i.e. morphological constructions, and phrases/phrase-like constructions, i.e. syntactic constructions, in German, French and English.1 Generally speaking, I reject approaches that try to unify morphology and syntax and, instead, follow the idea that the two domains of morphology and syntax are distinct. Cognitive aspects of compounds/ compound-like constructions and phrases/phrase-like constructions play the central role in the current contribution. Specifically, I am interested in examining whether compounds/compound-like constructions and phrases/phrase-like constructions differ with respect to how well they are memorized. In the literature, it is suggested that morphological constructions are better candidates for memory storage than syntactic constructions (Olsen 2000a: 899; Wunderlich 1986: 209). I intend to analyze this proposal and answer the question whether compounds/ compound-like constructions show a memorization advantage in comparison to phrases/phrase-like constructions. Doing so might give us new insights into the relationship between morphology and syntax as well as between lexicon and grammar.

A basic assumption of the dissertation is the idea that languages differ with regard to the usage of morphological and syntactic constructions (Spencer 2000: 313–314; Wunderlich 2008: 230). Specifically, it is assumed that the three aforementioned languages do not have the same preferences with respect to the linguistic realization of novel complex lexical concepts. While German can be argued to favor a morphological route in order to name new concepts, i.e. it usually relies on compounding, French mostly takes a syntactic route, i.e. it normally relies on phrases (Bücking 2009, 2010; Dressler 2006: 28; Hüning 2010: 211; Kastovsky 2009: 334; Van Goethem 2009; Zwanenburg 1992: 221). In comparison to the two above-named languages, it is sometimes difficult to keep apart compounds and phrases in English. I argue that we should substitute the terms “compound” and “phrase” by the terms “compound-like construction” (henceforth: CoLiCo) and “phrase-like construction” (henceforth: PhraLiCo) in some cases. Claiming that not only compounds and phrases but also CoLiCos and PhraLiCos exist in English, I assume that this language prefers using compounds or CoLiCos to express novel complex lexical concepts (cf. Downing 1977: 810; Lieber 2005: 378; McCauley, Hestvik & Vogel 2012: 26–27).

So far, in sum, while the first paragraph raises the question given in (1), the second paragraph points to the assumption in (2):

(1)Do morphological constructions, i.e. compounds/CoLiCos, show a memorization advantage in comparison to syntactic constructions, i.e. phrases/ PhraLiCos?

(2)While German prefers using compounds as naming units, French favors phrases for this purpose. In English, compounds or CoLiCos typically serve to name complex lexical concepts.

Combining (1) and (2) finally leads to the central question of the current contribution, which is presented in (3):

(3)Do German compounds and English CoLiCos show a memorization advantage in comparison to French phrases and English PhraLiCos? (cf. Chapters 6 and 7)

However, before I can approach the question in (3), it is indispensable to answer the questions in (4) to (6).

(4)How can one define/characterize compounds/CoLiCos and phrases/ PhraLiCos on structural grounds? (cf. Chapter 3)

(5)Do compounds/CoLiCos and phrases/PhraLiCos differ with respect to semantic-functional aspects? (cf. Chapter 4)

(6)Which implications might the interplay of structural and semantic-fuctional aspects have for the mental representation and processing of compounds/ CoLiCos on the one hand and phrases/PhraLiCos on the other hand? (cf. Chapter 5)

In general, constructions can be compounds/phrases or CoLiCos/PhraLiCos. Assuming that compounds are products of morphology and phrases are syntactic constructions, I both define and characterize the two construction types. That means, having defined compounds and phrases on the basis of one specific factor, namely the primary factor, I characterize these constructions by means of secondary factors. Using one primary factor, which has priority over the secondary factors, I ensure that my argumentation is not circular. Put differently, compounds and phrases are first defined and then characterized. Looking at constructions that are composed of an adjective and a noun, one sees that my approach can be easily implemented in German and French: First, compounds (e.g. Großmutter/grand-mère, big_mother/big-mother, ‘grandmother’) and phrases (e.g. große Mutter/grande mère, ‘big mother’) are defined by applying the primary factor inflectional agreement and, then, they are further characterized by saying, for instance, that German compounds are typically stressed on the first syllable. Concentrating on English adjective-noun (henceforth: AN) constructions, however, one realizes that the primary factor inflectional agreement is not available. Now, two options arise. First, it might be claimed that English has only one type of AN constructions. At first glance, this option seems to be quite attractive because, as stated above, the primary factor does not exist (anymore) in English. However, if one simply assumed that there was just a single construction type, one would take it for granted that our secondary factors were superfluous. Therefore, I prefer an alternative way, which is based on the historic root that German and English share. Although German AN compounds and phrases are defined only on the basis of inflectional agreement, certain secondary factors are regarded as crucial as well. For instance, while the majority of German AN compounds are stressed on the adjective, AN phrases of this language normally bear stress on the noun. By the same token, even if one cannot safely define English AN constructions as either compounds or phrases, they might show certain features that are typical of compounds or phrases (in Germanic languages) nevertheless. Hence, if necessary, the terms “CoLiCos” and “PhraLiCos” are used in English. The fact that inflectional agreement between an adjective and a noun disappeared in the history of English prevents us, if we follow my approach, from defining compounds and phrases of the type AN in this language. Nonetheless, I argue that we can still consider constructions more compound- or phrase-like by investigating secondary factors. Overall, in the case of AN constructions, German functions as a bridge to English: Having defined and characterized a compound or a phrase in the former language, we can hypothesize that the characteristics also apply to the latter because the two languages are closely related.

I am interested in complex constructions that individual language users memorize and that can become lexicalized within a speech community. That means, from a functional perspective, the focus lies on naming units that represent a subkind of the concept expressed in the head noun. Note, however, when defining compounds and phrases on the basis of inflection/inflectional agreement, it does actually not matter whether a construction represents a naming unit or not, whether it is lexicalized or not. Only the primary factor decides whether a construction represents a compound or a phrase. Therefore, the focus on (potential) naming units becomes only relevant if we connect structural to semantic-functional aspects, e.g. in Chapter 4. In this case, exocentric constructions, which do not represent a subkind of the head, and proper names, which do not refer to kinds, are ignored. Moreover, copulative and synthetic compounds are also ignored throughout my entire work. I primarily focus on constructions that are composed of an adjective and a noun in the three languages of interest (e.g. Braunbär, ours brun, brown bear). Looking at AN/noun-adjective (henceforth: NA) constructions, I assume that compounds represent typical naming units in German but phrases are preferred for the naming function in French (Booij 2002: 315; Bücking 2009, 2010; Van Goethem 2009). In English, CoLiCos usually serve as naming constructions (cf. McCauley et al. 2012: 26–27). Apart from AN/NA constructions, German and English constructions that contain two nouns (e.g. Kaffeetasse/coffee cup) as well as French constructions whose two nouns are connected by a preposition (e.g. tasse à café, cup to coffee, ‘coffee cup’) or by the combination of a preposition and a determiner (e.g. café au lait, coffee with.the milk, ‘milky coffee’) are also considered. In this case, German and English prefer compounds to express complex lexical concepts but French favors phrases (Dressler 2006: 28). Focusing on noun-noun (henceforth: NN) compounds in German and English but on noun-preposition-noun (henceforth: NPN)/noun-preposition-determiner-noun (henceforth: NPDN) phrases in French is based on the general observation that Germanic languages often make use of NN compounding where Romance languages opt for NPN or NPDN phrases (Dahl 2004: 222; Dressler 2006: 28; Kastovsky 2009: 334). That means, NN constructions are less frequent than NPN (and NPDN) constructions in French (Nicoladis 2002a: 46; cf. also Clark 2016: 312). ten Hacken’s (2013: 101–102) data support this trend: While only five percent of the English NN compounds investigated in his contribution appeared as NN constructions in French, 61 percent of the English NN compounds were NPN phrases in French.2 Research on first language acquisition of monolinguals and bilinguals also confirms the aforementioned preferences. Referring to data from Seidler (1988), Clark (1998: 521) shows that French-speaking children, who were five or six years of age, almost never relied on compounding to create new items (cf. also Clark 2001: 387). Nicoladis’ (1999) analysis revealed that a French-English bilingual child produced clearly more NN compounds in English than in French from 33 to 39 months of age. Nicoladis (2002a) found in a production task with children, who were between 39 and 59 months old, that French-English bilinguals created a significantly higher number of NN compounds/ constructions in English in comparison to French but a significantly higher number of NPN phrases in French in comparison to English. Overall, while both English monolinguals and bilinguals produced more NN compounds than NPN phrases in English, bilinguals constructed more NPN phrases than NN constructions in French. Note that French monolinguals were not tested in the study (on the study just outlined, cf. also Nicoladis 2002b; for a similar result, cf. Nicoladis 2002c: 646). Therefore, two facts can be kept in mind. First, NPN phrases are more frequent than NN constructions in French and, second, French NPN phrases are often used where Germanic languages favor NN compounds.3 Finally, note that, although I focus on typical naming constructions in the languages under investigation, it is often also necessary to refer to the other kind of construction in a language. For instance, even if I am particularly interested in investigating German AN compounds, I also refer to German AN phrases in order to make clear how the two construction types differ.

The contribution is structured as follows. In Chapter 2, the major assumptions about fundamental terms such as morphology, syntax, lexicon, grammar, memorization as well as lexicalization are specified. I argue that morphology and syntax represent two distinct grammatical domains and that lexicon and grammar have to be kept apart. Further, it is emphasized that memorization, i.e. storage of lexical items in the mental lexicon, differs from lexicalization, i.e. storage of lexical items in the lexicon of a speech community. In Chapter 3, structural factors that have been proposed in order to distinguish between compounds/CoLiCos and phrases/PhraLiCos are investigated. One factor (inflectional agreement/inflection) is treated as the primary factor that is superior to other factors (e.g. stress), so-called secondary factors. Semantic-functional aspects in the context of the compound-phrase distinction are discussed in Chapter 4. Here, it is claimed that compounds/CoLiCos, but not phrases/PhraLi-Cos, are by their nature semantically non-compositional, appropriate to refer to kinds and to function as naming units and, consequently, perfect to enter the lexicon. In Chapter 5, I discuss cognitive aspects of complex constructions and suggest a principle in order to describe the mental representation and processing of compounds/CoLiCos and phrases/PhraLiCos. Due to the structural and semantic-functional peculiarities of compounds/CoLiCos, I believe that these constructions develop their own representation in the mental lexicon earlier than phrases/PhraLiCos. Chapters 6 and 7, where two own experimental studies are described and interpreted, represent the core of the present contribution. In the empirical work to be presented, I focused on AN/NA constructions in German, French and English and examined how well they were memorized. I argue that compounds/CoLiCos show a memorization advantage in comparison to phrases/PhraLiCos and interpret the results of the experiments with respect to the principle introduced in Chapter 5. In Chapter 8, I draw conclusions and connect my analyses to the debate about language complexity. Using AN/ NA constructions in the three languages under investigation as an example, I show that the structural, semantic-functional and cognitive aspects of the compound-phrase distinction nicely fit into the complexity debate.

2Basic assumptions

The objective of the current chapter is twofold: I aim at showing how my work is connected to two “big questions” in linguistics and defining central terms of my contribution. Härtl (2013: 7) recapitulates the following two controversial subjects that have been debated for a long time. On the one hand, researchers discuss whether word-formation belongs to the grammar at all; on the other hand, linguists assume a particular role of word-formation in the grammar and are concerned with its exact nature. In §2.1, I focus on the second issue and contrast several proposals on the question whether one has to distinguish between the two grammatical domains morphology and syntax. In §2.2, I investigate the first controversy and look at suggestions on the nature of lexicon and grammar. The two debates interact and potential answers to one of the issues can have severe implications for the other discussion as well. Having started to reflect on the notion of the lexicon, I introduce and define two essential terms of my work, namely memorization and lexicalization, in §2.3. Fina...

Table of contents

• Cover
• Title Page
• Copyright
• Contents
• Acknowledgments
• Abbreviations
• Some general notes
• List of tables
• List of figures
• 1 Introduction
• 2 Basic assumptions
• 3 Compound-phrase distinction I: Structural aspects
• 4 Compound-phrase distinction II: Semantic-functional aspects
• 5 Compound-phrase distinction III: Cognitive aspects
• 6 Experimental study I: The memorization of compounds/CoLiCos and phrases/PhraLiCos: An investigation on German, French and English
• 7 Experimental study II: The memorization of CoLiCos and PhraLiCos in English
• 8 Conclusion
• Appendix
• References