Nominal composition and the demarcation between morphology and syntax

Holden Härtl, Katja Böer, Sven Kotowski & Marco Benincasa Universität Kassel

holden.haertl@uni-kassel.de www.uni-kassel.de/go/haertl

Introduction

2

Why do we need compounds?

Classical answers refer to the naming function of compounds, which somehow represent a category / a lexicalized concept.

Why does a language have morphology?

In contrast, syntactic phrases are often claimed to provide descriptions.

Problem:

Many compounds are not established names: Freitagsentscheidung, Nacktprotest, Terroropa

Many phrases are established names: Grüner Tee, Kleiner Tümmler, Deutsch als Fremdsprache

Can we perhaps say that novel compounds are “suggestions for lexicalizations”, cf. Lipka (1977); Motsch (2004)?

Question / Contents

3

What is the cognitive status of morphological products like compounds?

Are they lexicalized differently? Are they more salient in discourse?

1 Linguistic differences between compounds and phrases

2 Learning study: adjective-noun compounds

4 Conclusion

We will argue that compounds indeed deserve a special status as names and have as such, possibly because they are morphological products, a distinct cognitive status.

Grammatical differences

Semantic-pragmatic differences

3 Questionnaire study: discourse salience

Grammatical differences

4

Compounds display lexical integrity:

(1) a. Mia ist Fahrerin eines Audis. Der hat nun einen Motorschaden.

b. Mia ist Audifahrerin. *Der hat nun einen Motorschaden.

The internal structure of (synthetic) compounds is not accessible to syntactic operations:

(2) a. a drunk driver and a melancholic one

b. *a truck driver and an Audi one

cf. Booij (2009); Giegerich (2006)

Semantic-pragmatic differences

5

Compounds often denote kinds:

(1) a. ??Eine Flasche mit Bier hat einen Kronkorken.

b. Eine Bierflasche hat einen Kronkorken.

Novel compounds are linguistically marked expressions. As such, they are compatible with sogenannte-contexts:

(2) a. ??Das ist ein sogenanntes langes Messer.

b. Das ist ein sogenanntes Langmesser.

(3) a. ??Das ist ein sogenannter Henkel der Vase.

b. Das ist ein sogenannter Vasenhenkel.

cf. Bücking (2009); Carlson (1977); Krifka et al. (1995); Schlücker & Hünning (2009)

Semantic-pragmatic differences

6

Compounds often have a specialized meaning.

(1) Max ist ein schöner Raucher.

Max is a smoker and beautiful (intersective)

Max is somebody who smokes beautifully (non-intersective)

(2) Max ist ein Schönraucher.

Max is somebody who smokes beautifully (non-intersective)

cf. Bücking (2009); Egg (2006); Schlücker & Hünning (2009); Schäfer (2010)

For example, an intersective reading is dispreferred with A-N compounds:

Are compounds cognitively different?

7

Compounds are different from phrases.

Does this give us reason to believe that they are treated differently from a cognitive point of view?

Aphasic data: selective impairment for syntactic phrases like strange fever with compound retrieval left intact, see Mondini et al. (2002)

Morphology as an economic way of producing and memorizing complex linguistic structures, see Wunderlich (2008)

Are novel compounds memorized differently than the corresponding phrases?

Williams syndrome: selective impairment for lexical computation/access with grammatical computation left intact, see Clahsen & Almazan (2000)

Learning study

8

Learning phase: subjects were asked to memorize unkonwn picture labels

eine Kurzsäge ein breiter Kamm

Compound Phrase

N = 6 N = 6

see Schöpperle & Härtl (2011)

Learning study

9

Recall phase: subjects were asked to decide on correct / incorrect labels

eine Kurzsäge eine Flachsäge

Compound: learned Compound: not learned

N = 6 N = 6

Response variable: reaction times to decide

Learning study

10

ein breiter Kamm ein tiefer Kamm

Phrase: learned Phrase: not learned

N = 6 N = 6

Response variable: reaction times to decide

Recall phase: subjects were asked to decide on correct / incorrect labels

Learning study

11

Entire procedure was repeated over three days:

H1: Compounds are memorized differently than phrases over time.

Day 1

Day 4

Day 8

Learning study

12

Results: main effects

Learned items are decided faster (p < .001)

Phrases are decided faster (p < .01)

You get better over time (p < .001)

Learning study

13

ITEM TYPE × DAY interaction (not significant)

neither type is memorized better over time (p < .26)

860

910

960

1010

1060

1110

Session 1 Session 2 Session 3

Compounds

PhrasesRT

Learning study

14

LEARNED × ITEM TYPE interaction (p < .09)

stronger effect of memorization for compounds (p < .001)

900

950

1000

1050

1100

1150

Phrases Compounds

Not learned

Learned

not learned compounds take longer to decide than phrases (p < .001)

this difference disappears when the compounds are learned (p < .67)

RT

Learning study

15

Error rates: LEARNED × ITEM TYPE interaction (p < .001)

compounds profit from learning, phrases don’t (p < .75)

4,5

4,6

4,7

4,8

4,9

5,0

5,1

5,2

5,3

5,4

Not learned Learned

CompoundsPhrases

compounds are decided as correctly as phrases when learned (p < .99)

CORRECT

ANSWERS

Learning study

16

Main findings

> =

> >>

(1)

(2)

Testing discourse salience

17

Are novel compounds more salient in discourse?

Environment: Verbs of implicit causality create a strong bias for pronouns to be resolved as STIMULUS.

(1) STIM-EXP: Das Lexikon begeistert den Studenten, weil es / er …

weil es viele Abbildungen enthält.

weil er gerne neue Fakten lernt.

(2) EXP-STIM: Der Student schätzt das Lexikon, weil es / er …

weil es viele Abbildungen enthält.

weil er gerne neue Fakten lernt.

cf. Brown & Fish (1983); Grosz et al. (1995); Härtl (2001)

Testing discourse salience

19

Questionnaire study

(1) a. Die flache Säge begeistert Christoph, weil [sie | er] …

Sentence completion task

Factors: STIM as compound vs. phrase STIM-EXP vs. EXP-STIM verb

b. Die Flachsäge begeistert Christoph, weil [sie | er] …

(2) a. Johanna schätzt das schmale Messer [sie | es] …

b. Johanna schätzt das Schmalmesser, weil [sie | es] …

H1: Novel compounds produce a stronger bias towards STIMULUS explication

N = 6 x 2

N = 6 x 2

Testing discourse salience

20

Results:

Highly significant main effect for implicit verb causality (p < .0001)

STIMULUS is explicated more often than the EXPERIENCER.

Testing discourse salience

21

Results:

Tendency for verb type (not significant)

S-E verbs attrack sightly more STIMULUS explications.

Testing discourse salience

22

Results:

Significant compound effect (p < .02)

Compound STIMULI are explicated more often than phrase STIMULI.

Conclusion

The data are compatible with a lexicalist view, which implies a separation between syntax and morphology.

Novel compounds of the A-N type are linguistically marked and have a pronounced cognitive status.

This is reflected in the memorization procedures employed for the two types of expressions.

Unlearned novel compounds are harder to process than phrases.

Novel compounds raise the discouse salience of the corresponding referents.

23

For compounds a stronger memorization effect was detected.

In the context of implicit verb causality, pronouns are resolved more often as STIMULUS if it is a novel compound.

….

Thank you.

The authors would like to thank Dirk Koester, Susan Olsen, Mathias Schlesewsky, Peter Schöpperle, Katharina Spalek, Tobias Richter and Sascha Schröder.

24

References

Booij, Geert (2009) Lexical integrity as a formal universal: a constructionist view. In: Scalise, Sergio; Elisabetta Magni, and Antonietta Bisetto (eds.) Universals of Language Today. Berlin: Springer Science & Business Media, 83-100.

Brown, Roger and Deidre Fish (1983) The psychological causality implicit in language. Cognition, 14, 237-273.

Bücking, Sebastian (2009) German nominal compounds as underspecified names for kinds. Linguistische Berichte Sonderheft, 17, 253-281.

Carlson, Greg (1977) A unified analysis of the English bare plural. Linguistics and Philosophy, 1-3, 413-458.

Clahsen, Harald & Mayella Almazan (2001) Compounding and inflection in language impairment: evidence from Williams Syndrome (and SLI), Lingua, 111, 729-757.

Egg, Markus (2006) Anti-Ikonizität an der Syntax-Semantik-Schnittstelle. Zeitschrift für Sprachwissenschaft, 25-1, 1-38.

Giegerich, Heinz (2006) Attribution in English and the distinction between phrases and compounds. In: Rösel, Peter (ed.) Englisch in Zeit und Raum - English in Time and Space: Forschungsbericht für Klaus Faiss, Trier: Wissenschaftlicher Verlag.

Grosz, Barbara, Aravind K. Joshi, and Scott Weinstein (1995) Centering: A framework for modelling the local coherence of discourse. Computational Linguistics, 21(2), 203–225

Härtl, Holden (2001) CAUSE und CHANGE: Thematische Relationen und Ereignisstrukturen in Konzeptualisierung und Grammatikalisierung (= studia grammatica 50). Berlin: Akademie-Verlag.

Jespersen, Otto (1942) A Modern English Grammar. Part VI, Morphology. Copenhagen: Ejnar Munksgaard.

Krifka, Manfred; Francis. J. Pelletier; Greg. N. Carlson; Alice ter Meulen; Gennaro Chierchia & Godehard Link (1995) Genericity: an introduction. In Greg N. Carlson & Francis J. Pelletier (eds.), The Generic Book. Chicago/London: University of Chicago Press, 1-124.

Lipka, Leonhard (1977) Lexikalisierung, Idiomatisierung und Hypostasierung als Probleme einer synchronen Wortbildungslehre. In: Brekle, Herbert & Dieter Kastovsky (eds.) Perspektiven der Wortbildungsforschung, Bonn: Bouvier, 155-164.

Mondini, Sara; Gonia Jarema; Claudio Luzzatti; Cristina Burani, and Carlo Semenzai (2002) Why is red cross different from yellow cross? A neuropsychological study of noun-adjective agreement within Italian compounds. Brain and Language, 81, 621-634.

Motsch, Wolfgang (2004) Deutsche Wortbildung in Grundzügen. Berlin/New York: Walter de Gruyter.

Olsen, Susan (2000) Composition. In: Booij, G. et al. (eds.) Morphologie / Morphology, 897-916, Berlin: Mouton de Gruyter.

Plag, Ingo; Gero Kunter; Sabine Lappe, and Maria Braun (2008). The role of semantics, argument structure, and lexicalization in compound stress assignment in English. Language, 84-4, 760-794.

Wunderlich, Dieter (2008) Spekulationen zum Anfang von Sprache. Zeitschrift für Sprachwissenschaft, 27-2, 229-266.

Schäfer, Martin (2010) Prä- und postnominale Modifikation im Englischen und das Sitationsargument. Talk the 10th workshop Ereignissemantik, Universität Tübingen.

Schlücker, Barbara & Matthias Hüning (2009) Compounds and phrases. A functional comparison between German A+N compounds and corresponding phtrases, Rivista di Linguistica, 21-1, 209-234.

Schöpperle, Peter & Holden Härtl (2011) Morphological vs. syntactic composition. The case of nominal compounds. Talk at the workshop „Wort oder Phrase“, DGfS conference, Universität Göttingen.

25