Our project

1
Our project

• Extension of Sanders et al (2002).
• In Sanders et al (2002) all words were
  3-syllabic, hence knowing one word would allow
  the subject to deduce all other word onsets
• We will use 6-8 bisyllabic and trisyllabic
  words
• 3-syllabic batuki, gopeda, pudeto, tebogu
• 2-syllabic kopi, beka, gido, tadi

2
Our project

• Phase 1 record the subjects brainwaves in
  response to words in an unknown language
  presented as one continuous streambatukipudetote
  bogutadibekapudetogopedabekagidokopitadibatukigope
  dakopigidotebogu
• Phase 2
• test whether the subject learnt these words from
  exposure
• Teach the words to the subject
• A subset of words show the object
• Phase 3 record the subjects brainwaves in
  response to words in a known language

3
Hypothesis

• If there is an increased N100 to word onsets in
  Phase 3 vs. Phase 1, then this component is
  sensitive to word segmentation
• Finer analysis of the N400 component

4
Our project

• Q Why no pauses in the stream?
• A Because pauses are robust markers of
  word-boundaries, i.e. subjects can infer which
  are words in our nonsense language from the
  location of pauses without knowing words per se