The acquisition of the weakstrong distinction

1

• The acquisition of the weak-strong distinction
• and the Dutch quantifier allemaal
• Erik-Jan Smits (E.J.Smits_at_let.rug.nl)
• Bart Hollebrandse

SiN-day November 25, 2004
2
The acquisition of quantificationthe classical
picture and the yes-answer

• Question
• Is every farmer feeding a donkey?
• Possible answers
• (1) No pointing at the donkey
• (2) Yes all the farmers are feeding a donkey
  (adult answer)
• (3) Yes many donkeys are fed by a farmer

Crain et al. (1996)
3
The weak-strong distinction and the acquisition
of quantification

• Weak-strong distinction (Milsark, 1979)
• There are many, few, all, every doctors in
  the room
• Geurts (2003) no experimental data, but
• the grammatical connection between a quantifier
  and its domain of quantification is less rigid in
  children than it is in adults (footnote 3, p.
  10). (cf. Philip (1995), Drozd and many others)

4
Analyzing quantified sentences

• In order to interpret a quantified sentences, one
  should
• Correctly localize the domain of the relevant
  quantifier (or determine its scope)
• Correctly interpret the domain of the relevant
  quantifier (or determine the nature of the
  quantifier)

5
The Dutch quantifier allemaal (1)

• With respect to the correct localization of the
  domain, allemaal is able to quantify over
  subject or object
• Een jongen draagt de koffers allemaal
• A boy is carrying the suitcases all
• A boy is carrying all the suitcases
• 2. De jongens dragen allemaal een koffer
• The boys are carrying all a suitcase
• The boys are all carrying a suitcase

6
The Dutch quantifier allemaal (2)

• With respect to the correct interpretation of the
  domain, allemaal is unique because its
  ambiguity between a strong and a weak quantifier
• 1. Een jongen draagt de koffers allemaal
• A boy is carrying the suitcases all
• A boy is holding all the suitcases
• allemaalstrong (A,B) is true iff A ? B
• 2. Er fietsen allemaal papegaaien
• There are bicycling all parrots
• There are bicycling allemaal (many) parrots
• allemaalweak (A, B) is true iff A ? B ?
  2

7
Experimental design

• Hypothesis Difficulties with understanding
  quantified sentences can not only be found in
  children unable to correctly localize the domain
  of the quantifier, but also in children unable to
  correctly interpret the domain of the quantifier
  (i.e. a consequent strong or weak reading)
• Aim Distinguish children with an adult-like
  quantifier system from children with a weak
  quantifier system.
• Two experiments
• Scope-experiment
• Is a child able to make a distinction between
  allemaal quantifying over the subject or
  object?
• Weak-strong experiment
• Is a child able to make a distinction between a
  weak and strong use of allemaal?

8
General prediction

• A child that is always interpreting a quantifier
  as a weak one in the weak-strong experiment,
  regardless its syntactic position, will judge a
  significantly higher amount of sentences as true
  in the scope-experiment (regardless the fact
  whether the subject or object is within the
  domain of the quantifier) than the child always
  understanding a quantifier as a strong one

9
The scope-experiment

• Predictions
• A child is always quantifying the subject
• A child is always quantifying the object
• A child is a spreader quantifying over both the
  object and the subject
• Method Truth Value Judgment Task also
  questioning the yes-answer.
• 39 kids (aged 4 6)
• 3 items per 2 conditions 3 no-fillers (total 15
  sentences)

10
The scope-experiment test items (1)

• Een paard draagt de meisjes allemaal
• A horse is carrying the girls all
• object Q
• De mannen dragen allemaal een ezel
• The men are carrying all a donkey
• subject Q

11
The scope-experiment test items (2)

• Een robot houdt de ballonnen allemaal vast
• A robot is holding the balloons all
  PART
• object Q
• De mannen tillen allemaal een kist op
• The men are lifting all a box up
• subject Q

12
Results scope-experiment

• Two groups
• Adult answer (no) (26)
• Non-adult answer (yes) (13)

13
Results scope experiment domains of
quantification
14
Results scope experiment domains of
quantification
15
The weak-strong experiment

• Prediction
• Children differ in their interpretation of a
  quantifier as
• A weak one
• A strong one
• A weak or strong one depending on its syntactic
  position the adult analysis
• 39 subjects (aged 4 6)
• Method Truth Value Judgment Task also
  questioning the yes-answer
• Total of test sentences 18 (12 test items, 3
  no-fillers, 3 yes-fillers)

16
The weak-strong experiment test items

• De ezels huilen allemaal
• The donkeys crying all
• (Strong3 items
• 3 items with alle, all)
• Er dansen allemaal meisjes
• There are dancing many girls
• (Weak 6 items)

17
Results weak-strong experiment

• Prediction I children with only a strong reading
  of allemaal
• Prediction II children with only a weak reading
  of allemaal

18
Results weak-strong experiment (2)

• Prediction III children with an adult reading
  (expected yes-answer in the weak-condtion,
  no-answer in the strong-condittion)

19
Scope and the weak-strong distinction

• General prediction
• A child that is always interpreting a quantifier
  as a weak one in the weak-strong experiment,
  regardless its syntactic position, will judge a
  significantly higher amount of sentences as true
  in the scope-experiment than the child always
  understanding a quantifier as a strong one

20
General results scope and the weak-strong-distinc
tion
4
21
Conclusions

• Experiment 1
• The data shows that there are children that have
  a weak reading for a universal strong quantifier
  (13 out of 39).
• Experiment 2
• Children have a preference to analyze allemaal as
  a strong quantifier, in a situation in which not
  all the subjects are participating (35 out of
  39).
• In general
• Children that have a weak quantifier system can
  only be discriminated from children that have an
  adult quantifier system by experiments taking the
  weak-strong distinction into account.
• Problems with quantification are more widespread
  than previously thought.