DETERMINING TELICITY: Entailments Test

1
Are We Done Yet? Childrens Interpretation of
Verb Phrase Telicity Diane A. Ogiela1, Michael
W. Casby2 and Cristina Schmitt3 1Department of
Communication Sciences and Disorders, TWU
2Department of Communicative Sciences and
Disorders, MSU 3Department of Linguistics, MSU

Results
Introduction and Linguistic Background
Adult Behavior
LINGUISTIC VARIABLES
Build-type vs. Eat-type Verbs by Determiner-Type
for Children

• RESEARCH QUESTION
• What do typically developing (TD) children know
  about the linguistic means of indicating
  telicity?
• TELICITY
• Refers to whether or not a linguistic
  description indicates that an event culminates in
  a logical end point.
• Not morphologically marked Determined by the
  entire verb phrase (VP)
• Is compositional, i.e., determined by the
  interaction between the properties of the verb
  and the properties of the direct object
  determiner phrase (DP) or prepositional phrase.
• EXAMPLES OF VP TELICITY
• Draw ? atelic (- logical end point) Push a cart
  ? atelic (- logical end point)Draw a circle ?
  telic ( logical end point) Push a cart to the
  park ? telic ( logical end point)
• Draw circles ? atelic (- logical end point) Push
  over a cart ? telic ( logical end point)

• Statistical Summary
• Between subjects effect
• 3-yr-olds sig. diff. than 5- 6-yr-olds.
• No main effect for verb-type.
• Sig. verb X determiner interaction.
• More telic responses with eat- type verb two
  than the (for 3-, 4-,
• 6-yr-olds).

Example stimuli Did the man build the/two
houses? Did the man eat up the /two brownies?
ADULTS INTERPRETATIONS OF TELICITY

• Summary for Adults
• Distinguish between partitive non- partitive
  QS verbs regarding telicity.
• VPs w/ partitive QS verbs are interpreted as
  telic more often with two than w/ the.
• The resultative particle up disambiguates
  between partitive and non-partitive
  interpretations of eat-type verbs.

Eat-type vs. Eat up-type Verbs by
Determiner-Type for Children

• Statistical Summary
• Between subjects effect
• 3-yr-olds had fewer telic responses.
• Main effect for verb-type.
• Eat up-type verbs were interpreted as telic more
  often than eat-type verbs.
• Main effect for determiner-type.
• More telic responses w/ two than the.
• Planned comparisons
• No sig. diff. between eat up two and eat two
  for any age group.

• DETERMINING TELICITY Entailments Test
• VERB PROPERTIES
• Quantity-sensitive (QS) verbs amount of object
  affected matters. e.g., build, eat.
• Quantity-insensitive (QI) verbs amount of object
  affected does not matter e.g., push, carry.
• DP PROPERTIES
• Non-quantized DP no explicit quantity
  information (e.g., bare plurals such as toys,
  apples).

If I was drawing a circle and suddenly stopped
drawing a circle, did I draw a circle? NO.
Therefore, the VP is telic.
If the man was pushing the cart and suddenly
stopped pushing it, did the man push the cart?
Yes. Therefore, the VP is atelic.
Responses to Yes/No questions following videos
that depict non-culminating events.
Hypotheses Predictions
Push-type vs. Push over-type Verbs by
Determine-type for Children

• Verb-type and determiner-type will interact to
  produce different response patterns to questions
  about non-culminating events. The predicted
  hierarchy from most to least telic is
• Build-type verbs with a cardinal number in the
  object DP
• Build-type verbs with a definite determiner in
  the object DP
• Eat-type verbs with a cardinal number in the
  object DP
• Eat-type verbs with a definite determiner in the
  object DP
• Older children will demonstrate greater
  sensitivity to verb-type and determiner-type with
  regard to telicity than younger children.
• Resultative particles specify an end point.
  Therefore, when presented with eat-type VPs,
  children will be more sensitive to the presence
  of resultative particles than the presence of a
  definite determiner or cardinal number for the
  interpretation of VPs as telic.

• Statistical Summary
• Main effect for verb-type.
• Push over-type verbs had more telic responses
  than push-type verbs.
• Main effect for determiner-type.
• More telic responses for two than the.
• Interaction of verb-type X determiner- type.
• More telic responses w/ push-type two than
  push-type the.

Method
Discussion

• 80 typically developing English-speaking
  children ages 3 (n15), 4 (n24), 5 (n18), and 6
  (n23).
• Passed speech, language and hearing screenings.
  
• The visual stimuli were videos of events that
  either culminated at a logical end point (e.g.,
  actor builds two toy houses) or that were
  discontinued before the end point was reached
  (e.g., actor builds one house completely and the
  second partially). The linguistic stimuli were
  as listed in the table above.
• Participants watched movies with a puppet,
  who routinely fell asleep.
• They were instructed to watch each movie.
  After each video, the puppet was woken up and he
  asked the child a Yes/No question.
• For experimental items, YES ? atelic
  interpretation and NO ? telic interpretation.
• 30 target videos and 6 filler videos (to check
  for attention).
• The data (for non-culminating events) were
  analyzed using repeated measures ANOVAs.

• Children make some, but not all, of the
  verb-type distinctions as adults re QS vs. QI,
  but by age 6 they do not yet distinguish between
  partitive and non-partitive verbs.
• The cardinal number serves as a strong
  indicator of telic VPs for children, for some,
  even in VPs w/ QI verbs.
• Consistent with finding that telic VPs help
  children count whole events (Wagner, 2006).
• The cardinal number may be modifying the whole
  event and not just the objects.
• Resultative particles may not be unambiguous
  indicators of telic VPs for children as old as 6.
  
• Although many children interpret resultative
  particles as cues for telic VPs, the particles
  were not better indicators than the cardinal
  number two. This was an unexpected result.
• 3-year-olds have fewer telic interpretations
  than older children overall.
• Overall patterns of responses are similar
  across the 4-, 5-, and 6-year-old groups.
• With age, childrens patterns increasingly
  approximate the adult patterns.

EXPANDED LINGUISTIC VARIABLES
This study was completed as part of the first
authors dissertation work at Michigan State
University. For correspondence, please contact
Diane Ogiela at ogiela_at_twu.edu. Partial funding
was provided by the American Speech-Language-Heari
ng Foundation New Century Scholars Doctoral
Scholarship and an MSU Dissertation Completion
Fellowship.
ASHA, 2007