Maximality and Domain Restriction: evidence from adjectival - - PowerPoint PPT Presentation

. .

Maximality and Domain Restriction: evidence from adjectival modification

Alan Munn1 Karen Miller2 Cristina Schmitt1

1Michigan State University 2Calvin College

BUCLD 34 .

.

The Definite Determiner Semantic properties

Semantics of Definiteness

.

Definites have a uniqueness presupposition

. . A definite description is felicitous if there is a unique/maximal (singular or plural) object in the domain of discourse that satisfies the descriptive content of the NP. (1)

• a. A boy and a girl came in. The girl sat down.
• b. Two women came in. #The woman sat down.
• c. Three women came in. The women sat down.
• d. John climbed the highest mountain in Tibet.
• e. John bought a bicycle. The saddle was missing.
• f. John bought a bicycle. #The tire was flat.

Munn, Miller, Schmitt (MSU and Calvin College) Maximality Restrictions BUCLD 34 2 / 33

.

The Definite Determiner Semantic properties

Domain restriction

.

Picking the right domain

. . like other determiners, definites require a contextually determined domain relative to which maximality is defined. (2)

• a. Please pick up the kids from school.
• b. The students are waiting outside.
• c. The Americans cheered for the Americans.

Munn, Miller, Schmitt (MSU and Calvin College) Maximality Restrictions BUCLD 34 3 / 33

.

Previous Work: Maximality and Plurality Children’s misuse of definites

Children’s (mis)use of definites

. . .

. . .

.

Child

. . Give me the red ball. Give me the yellow ball. .

Adult

. . Give me the red ball. Give me one of the yellow balls. (Karmiloff-Smith 1979; Maratsos 1972)

Munn, Miller, Schmitt (MSU and Calvin College) Maximality Restrictions BUCLD 34 4 / 33

.

Previous Work: Maximality and Plurality Children’s misuse of definites

Children’s (mis)use of definites

.

.

Egocentrism is the problem

. . Children have an ‘egocentric’ interpretation of the definite. (Karmiloff-Smith and Maratsos) .

Maximality is the problem

. . Definite errors persist, even when no element is egocentrically ‘in focus.’ Children lack the Maximality Presupposition of the definite

• determiner. (Wexler in press)

.

Domain Restriction is the problem

. . Children have problems with setting domain restrictions on

• determiners. (Drozd 2001; Drozd e Loosbroek 2004; Guerts

2003; Krämer 2002; Miller e Schmitt 2004; Munn et al. 2006)

Munn, Miller, Schmitt (MSU and Calvin College) Maximality Restrictions BUCLD 34 5 / 33

.

Previous Work: Maximality and Plurality Munn et al. 2006

Testing Maximality vs. Domain Restriction

.

Experimental sentences

. . “Give me the frog next to the barn.” Singular “Give me the frogs next to the barn.” Plural .

Controls

. . “Give me a frog next to the barn.” Singular “Give me all the frogs next to the barn.” Plural

Munn, Miller, Schmitt (MSU and Calvin College) Maximality Restrictions BUCLD 34 6 / 33

.

Previous Work: Maximality and Plurality Munn et al. 2006

Testing Maximality vs. Domain Restriction

.

Experimental sentences

. . “Give me the frog next to the barn.” Singular “Give me the frogs next to the barn.” Plural .

Controls

. . “Give me a frog next to the barn.” Singular “Give me all the frogs next to the barn.” Plural

Munn, Miller, Schmitt (MSU and Calvin College) Maximality Restrictions BUCLD 34 6 / 33

.

Previous Work: Maximality and Plurality Munn et al. 2006

Testing Maximality vs. Domain Restriction

.

Experimental sentences

. . “Give me the frog next to the barn.” Singular “Give me the frogs next to the barn.” Plural .

Controls

. . “Give me a frog next to the barn.” Singular “Give me all the frogs next to the barn.” Plural

Munn, Miller, Schmitt (MSU and Calvin College) Maximality Restrictions BUCLD 34 6 / 33

.

Previous Work: Maximality and Plurality Munn et al. 2006

Experimental setup

.

Toy house and barn

. . .

6 toy figures of each type:

. . English: frogs, rabbits, cats, monkeys Spanish: gatas (‘cats’), vacas (‘cows’), muñecas (‘dolls’), arañas (‘spiders’)

Munn, Miller, Schmitt (MSU and Calvin College) Maximality Restrictions BUCLD 34 7 / 33

.

Previous Work: Maximality and Plurality Munn et al. 2006

Results

.

Controls

. .

Children had adult-like performance on all controls

0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 Singular Maximal Plural Indef Singular All

% Correct: Control Sentences

English Spanish Adults

.

Definites

. .

Children differed in the singular but not the plural

94.9 67.3 100.0 98.2 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 Maximal Maximal Def Plural Def Singular

% Correct Maximality (English and Spanish combined) Children Adults

Munn, Miller, Schmitt (MSU and Calvin College) Maximality Restrictions BUCLD 34 8 / 33

.

Previous Work: Maximality and Plurality Munn et al. 2006

Explanation of the results

.

Explicit Restriction

. . Give me the frogs next to the barn. The maximal set of frogs next to the barn Explicit restriction on “the”: next to the barn .

Implicit Restriction

. . Give me the frog next to the barn. The maximal singleton frog next to the barn Implicit restriction on “the”: the frog closest to the barn

Munn, Miller, Schmitt (MSU and Calvin College) Maximality Restrictions BUCLD 34 9 / 33

.

Previous Work: Maximality and Plurality Maximality vs. Domain Restriction

Conclusions

.

Maximality

. . The “no Maximality Hypothesis” is not supported by this experiment. Children only made maximality errors in the singular condition, and not in the plural condition. .

Domain Restriction

. . The Domain Restriction Hypothesis is supported. Children had problems with implicit but not explicit restrictors on the definite.

Munn, Miller, Schmitt (MSU and Calvin College) Maximality Restrictions BUCLD 34 10 / 33

.

Manipulating restrictors Research Questions

Understanding domain restrictions

.

Research questions

. . What underlies children’s different choices in the singular condition in Munn et al. 2006? Are children sensitive to different sorts of restrictors? Can manipulating restrictors provide evidence for children’s knowledge of the Maximality presupposition?

Munn, Miller, Schmitt (MSU and Calvin College) Maximality Restrictions BUCLD 34 11 / 33

.

Manipulating restrictors Research Questions

Understanding domain restrictions

.

Scalar and colour adjectives

. . 1 2 3 4 (3)

• a. Point to the green penguin next to the igloo.
• b. Point to the tall penguin next to the igloo.

Munn, Miller, Schmitt (MSU and Calvin College) Maximality Restrictions BUCLD 34 12 / 33

.

Manipulating restrictors Maximality and scale structure

Maximality and scale structure

.

Size adjectives

. . Size adjectives like tall are open scale adjectives (Kennedy 2001) In a given context, objects can be ordered on this scale Size adjectives can be used to satisfy the Maximality presupposition of the definite

Munn, Miller, Schmitt (MSU and Calvin College) Maximality Restrictions BUCLD 34 13 / 33

.

Manipulating restrictors Maximality and scale structure

Maximality and scale structure

.

Colour adjectives

. . Colour adjectives are ambiguous between a ‘classificatory’ and a scalar interpretation (Kennedy e McNally 2010) (see also Rothschild e Segal 2009; Szabó 2001) When interpreted as individual level properties, they are non-gradable Non-gradable adjectives cannot provide a scale with a contextually determined endpoint Classificatory colour adjectives cannot be used to satisfy the Maximality presupposition

Munn, Miller, Schmitt (MSU and Calvin College) Maximality Restrictions BUCLD 34 14 / 33

.

Manipulating restrictors Maximality and scale structure

Hypotheses

. .

Point to the tall/green penguin next to the igloo.

1 2 3 4 Size adjectives provide an extra restrictor which can be used to satisfy Maximality Colour adjectives do not provide an extra restrictor

Munn, Miller, Schmitt (MSU and Calvin College) Maximality Restrictions BUCLD 34 15 / 33

.

Testing domain restriction Experimental setup

Experimental Setup

.

Picture

. .

1 2 3 4

LIGHT GREEN DARK GREEN RED LIGHT GREEN

5 6 7 8

LIGHT GREEN RED DARK GREEN LIGHT GREEN

.

Sentences

. .

(4)

• a. Point to the penguin

that is next to the igloo.

• b. Point to the tall

penguin that is next to the igloo.

• c. Point to the green

penguin that is next to the igloo.

• d. Point to the red/small

penguin next to the igloo.

Munn, Miller, Schmitt (MSU and Calvin College) Maximality Restrictions BUCLD 34 16 / 33

.

Testing domain restriction Experimental setup

Predictions

.

1 2 3 4

. .

Both adults and children will behave differently on implicit restrictions with scalar adjectives compared to colour adjectives Subjects should choose either the tallest penguin or the closest penguin for (4b) but

• nly the closest penguin for

(4c)

.

Sentences

. .

(4)

• a. Point to the penguin

that is next to the igloo.

• b. Point to the tall

penguin that is next to the igloo.

• c. Point to the green

penguin that is next to the igloo.

• d. Point to the red/small

penguin next to the igloo.

Munn, Miller, Schmitt (MSU and Calvin College) Maximality Restrictions BUCLD 34 17 / 33

.

Methods and Materials Subjects

Subjects

.

Adults

. . Adults Exp. 1 N=19 Adults Exp. 2 N=20 .

Children

. . Size Exp. 3 N=10; mean age 4;7 Colour Exp. 4 N=11; mean age 4;4

Munn, Miller, Schmitt (MSU and Calvin College) Maximality Restrictions BUCLD 34 18 / 33

.

Methods and Materials Materials

Materials

.

Adults

. .

• Exp. 1 & 2: 64 experimental sentence/picture pairs + 64

distractors. .

Children

. .

• Exp. 3 (size) 12 experimental sentence/picture pairs + 4

controls.

• Exp. 4 (colour) 12 experimental sentence/picture pairs +

4 controls.

Munn, Miller, Schmitt (MSU and Calvin College) Maximality Restrictions BUCLD 34 19 / 33

.

Adult Experiments Experiments 1 and 2

Experimental Pictures

.

Design

. . Filler pictures were identical to the experimental pictures, but had distractor sentences. Half of the objects were animate and half inanimate. Pictures used:

▶ butterflies/flagpoles next to a school and the Whitehouse ▶ dinosaurs/clouds next to a lake and a volcano ▶ trees/tractors next to a house and a barn ▶ bears/penguins next to an igloo and a snowman

Stimulii were counterbalanced so that half the choices were to the top row and half were to the bottom row. Sentences were presented in a random order (same order for all subjects)

Munn, Miller, Schmitt (MSU and Calvin College) Maximality Restrictions BUCLD 34 20 / 33

.

Adult Experiments Experiments 1 and 2

Experiments 1 and 2

.

Experiment 1: Picture Selection Task

. . Participants viewed pictures on a video monitor, and mark

• n a coding sheet the items that corresponded to the

sentence. Answers were coded as matching or not matching the theoretical predictions. .

Experiment 2: Truth Value Judgement Task

. . Pictures were modified so that a particular set of objects were circled in red. Sentences were modified to the following form:

▶ “We circled the tall tree next to the house.” Munn, Miller, Schmitt (MSU and Calvin College) Maximality Restrictions BUCLD 34 21 / 33

.

Adult Experiments Experiment 1

Results: Experiment 1

.

Size vs. Proximity

. .

0.461 0.475 0.5 1 Closest Biggest/Tallest

Closest Biggest/Tallest .461 .475

.

Colour vs. Proximity

. .

0.949 0.005 0.5 1 Closest Reddest/Greenest

Closest “Greenest/Reddest” .949 .005

Munn, Miller, Schmitt (MSU and Calvin College) Maximality Restrictions BUCLD 34 22 / 33

.

Adult Experiments Experiment 1

Results: Experiment 1

1 2 3 4

.

Size

. . Adults chose equally between the tallest object or the closest object when the restrictor was a scalar adjective .

Colour

. . Adults always chose the closest object when the restrictor was a colour adjective

Munn, Miller, Schmitt (MSU and Calvin College) Maximality Restrictions BUCLD 34 23 / 33

.

Adult Experiments Experiment 2

Results: Experiment 2

.

Size

. .

0.73 0.30 0.00 0.20 0.40 0.60 0.80 1.00 Expected T Expected F

.

Colour

. .

0.95 0.675 0.2 0.4 0.6 0.8 1 Expected T Expected F Munn, Miller, Schmitt (MSU and Calvin College) Maximality Restrictions BUCLD 34 24 / 33

.

Adult Experiments Experiment 2

Results: Experiment 2

.

Size

. . True items were treated differently from False items In both T and F items, levels of matching were relatively low .

Colour

. . True items were treated differently from False items In both T and F items, levels of matching were relatively high Adults treated size and colour adjectives differently

Munn, Miller, Schmitt (MSU and Calvin College) Maximality Restrictions BUCLD 34 25 / 33

.

Adult Experiments Experiment 2

Discussion

.

Size

. . In the size condition, the scalar adjective provided a second scale which could satisfy the maximality requirement. This accounts for the even split in the selection task and the lower rates of matching in the TVJT. .

Colour

. . In the colour condition, the colour adjective did not provide a second scale which could be used to satisfy maximality. This accounts for the categorical behaviour in the selection task, and the increased matching rates in the TVJT.

Munn, Miller, Schmitt (MSU and Calvin College) Maximality Restrictions BUCLD 34 26 / 33

.

Child Experiments Experiment 3

Results: Size vs. Proximity

.

Children

. .

0.475 0.525 0.5 1 Closest Biggest/Tallest

Closest Biggest/Tallest .475 .525

.

Adults

. .

0.461 0.475 0.5 1 Closest Biggest/Tallest

Closest Biggest/Tallest .461 .475

Munn, Miller, Schmitt (MSU and Calvin College) Maximality Restrictions BUCLD 34 27 / 33

.

Child Experiments Experiment 4

Results: Colour vs. Proximity

.

Children

. .

0.944 0.059 0.5 1 Closest Reddest/Greenest

Closest “Reddest/Greenest”/Tallest .944 .059

.

Adults

. .

0.949 0.005 0.5 1 Closest Reddest/Greenest

Closest “Reddest/Greenest” .949 .005

Munn, Miller, Schmitt (MSU and Calvin College) Maximality Restrictions BUCLD 34 28 / 33

.

Child Experiments Experiment 4

Results

1 2 3 4

.

Size

. . Children chose equally between the tallest object or the closest object when the restrictor was a scalar adjective, just like the adults did. .

Colour

. . Children always chose the closest object when the restrictor was a colour adjective, just like the adults did.

Munn, Miller, Schmitt (MSU and Calvin College) Maximality Restrictions BUCLD 34 29 / 33

.

Conclusions

Conclusions

Both children and adults behave similarly in distinguishing size from colour adjectives. For size adjectives, subjects chose either closest or biggest For colour adjectives, subjects overwhelmingly chose closest Children understand Maximality, since they use whichever restrictor provides them with a scale that has a contextually determined maximum.

Munn, Miller, Schmitt (MSU and Calvin College) Maximality Restrictions BUCLD 34 30 / 33

.

Conclusions

References

Drozd, Ken F. (2001). «The effect of context in children’s interpretations of universally quantified sentences.» Em: Language Acquisition and Conceptual Development. Ed. por Melissa Boweman e Stephen C. Levinson. Cambridge: Cambridge University Press,

• pp. 340–376.

Drozd, Ken F. e Erik van Loosbroek (2004). «The effect of context in children’s interpretations

• f universally quantified sentences.» Em: Semantics Meets Acquisition. Ed. por

Geerle van Geenhoven. Dordrecht: Kluwer. Guerts, Bart (2003). «Quantifying Kids». Em: Language Acquisition 11, pp. 197–218. Karmiloff-Smith, Annette (1979). A Functional Approach to Child Language: A study of determiners and reference. Cambridge: Cambridge University Press. Kennedy, Chris (2001). «Polar Opposition and the Ontology of ’Degrees’». Em: Linguistics and Philosophy 24, pp. 33–70. Kennedy, Christopher e Louise McNally (2010). «Color, Context and Compositionality». Em: Synthese 174.1, pp. 79–98.

Munn, Miller, Schmitt (MSU and Calvin College) Maximality Restrictions BUCLD 34 31 / 33

.

Conclusions

References (cont.)

Krämer, Irene (2002). «Reference of indefinite and pronominal noun phrases in a story context: English children’s comprehension.» Em: BUCLD 27: Proceedings of the 27th annual Boston University Conference on Language Development. Ed. por Barbara Beachley et al. Cascadilla Press, pp. 449–460. Maratsos, Michael (1972). The Use of Definite and Indefinite Reference in Young Children. Cambridge: Cambridge University Press. Miller, Karen e Cristina Schmitt (2004). «Wide-scope ndefinites in English child language». Em: Proceedings of GALA 2003. Ed. por Jacqueline van Kampen e Sergio Baauw. Utrecht: LOT,

• pp. 317–328.

Munn, Alan et al. (2006). «Maximality and plurality in children’s interpretation of definites». Em: Proceedings of BUCLD 30. Ed. por David Bamman et al. Cascadilla Press, pp. 377–387. Rothschild, Daniel e Gabriel Segal (2009). «Indexical Predicates». Em: Mind and Language 27,

• pp. 467–493. DOI: doi:10.1111/j.1468-0017.2009.01371.x.

Szabó, Zoltan (2001). «Adjectives in context». Em: Perspectives on Semantics, Pragmatics and Discourse: A Festschrift for Ferenc Kiefer. Ed. por Istvan Kenesei e Robert M. Harnish. Amsterdam: John Benjamins, pp. 119–146.

Munn, Miller, Schmitt (MSU and Calvin College) Maximality Restrictions BUCLD 34 32 / 33

.

Conclusions

References (cont.)

Wexler, Kenneth (in press). «Maximal trouble in the determiner system». Em: The Processing and Acquisition of Reference. Ed. por Edward Gibson e Neal Pearlmutter. Cambridge, Mass.: MIT Press.

Munn, Miller, Schmitt (MSU and Calvin College) Maximality Restrictions BUCLD 34 33 / 33

.

Acknowledgments

Thanks

.

.

Graduate Students

. .

Joe Jalbert

.

Undergraduates

. .

Brittany Baker, Lindsey Bird, Lyn Blanco, Erik Butterfield, Amy Cronin, Karl DeVries, Thao Dinh, Erin Dixon, Matthew Karram, Braden Leinbach, Lydia Oehring, Megan Sutton, Alanna Thiede, Jessica Wallace, Jessica Zwirner

.

Daycares

. .

Eastminster Child Development Center Stepping Stones Montessori School, East Lansing Okemos Kids Club at Chippewa Middle School

• St. Martha School, Okemos

.

Funding

. .

MSU Honors College MSU College of Arts and Letters

Munn, Miller, Schmitt (MSU and Calvin College) Maximality Restrictions BUCLD 34 34 / 33