EXPERIMENTAL PARADIGMS IN DEVELOPMENTAL PSYCHOLINGUISTICS by chenboying

VIEWS: 33 PAGES: 24

									     EXPERIMENTAL PARADIGMS IN
          DEVELOPMENTAL
         PSYCHOLINGUISTICS

Zsuzsanna Schnell
Ph. D. student
Institute of Psychology
University of Pécs
           Language acquisition
   journey that begins in the womb and continues
    throughout childhood, early adulthood and even
    beyond.
   past: acquisition literature situated the onset of lg at
    about 12 months – when 1st recognizable words are
    produced - hence focus on production
   BUT new methods: LA starts with PROCESSING
   paves the way, fetus eavesdropping -> prosody, intonation of
    mother tongue -> segmentation of speech stream, familiarizing
    with mother’s voice and mother tongue (Karmiloff-Smith –
    Karmiloff 2002).
         Findings of modern
     psycholinguistic experiments
   what the child understands is much more indicative of
    their stage in LA
   they actually have a knowledge that is not displayed in
    their utterances (espec. late starters age 3)
   same is true for L2 learners.
   bias for production: behaviorist views
   1960’s Chomsky’s views revolutionarized psycholing.
    research
   focus shifted from production to comprehension
                          Typology

Distinction based on:

   1. Sample: Longitudinal vs. Cross sectional
   2. Method: Observation based (diaries,
                   Experimental
    questionnaires) vs.

   3. Temporal aspect: Pre-natal vs. Post natal
   4. Linguistic behavior: Pre-verbal vs. Verbal
    Research methods in the study of
        language development
  Diaries and parental reports:
as early as 18th century – own children, diaries.
   remain a valuable source,
!: misleading: tempted to write down unusual,
or interesting, and not the ordinary but relevant.
 new version: McArthur Communicative
    Development Inventory (CDI) by Larry Fenson et al.
 checklist prepared by psycholinguists distributed to
    parents
 has been validated in laboratory settings.
      Advantages vs. Disadvantages

   ADVANTAGES: natural setting, family
    context
   fast, no funding needed
   no training required
   DISADVANTAGES: not suitable for older
    children, whose lg use is more complex
   may be misleading
   ALL IN ALL: If used correctly, the home-
    based parental questionnaire is a reliable
    indicator
           Observational data
  first motivation: reconcile allegations
of revolutionary nativism: UG, LAD
 first: tape recordings  transcripts.
 Roger Brown 1960-ies, Harvard:

longitudinal study, studying the lg dev.
of Adam, Eve and Sarah.

 transcripts fostered our understading of child
 grammar, acq. of basic sentence structures
   still a major research design
 Advantages vs. disadvantages
ADVANTAGES: Observation allows indirect scrutiny,
  metalinguistic competence is not a must.
 rich source of data.
 focus on environmental factors and on influence of
  socialization (effect of parental input).

DISADVANTAGES: longitudinal studies are
- confined to a small n. of children (luckily results are
    representative of all children)
!:transcripts may not include all situational factors, social and
    contextual effects
- Difficulties arise from individual encoding – devising their own
    system, rendering it unusable for others.
     CHILDES – The Child Language
        Data Exchange System
      http://childes.psy.cmu.edu
   unified approach
   standardization in research methods,
   of recording and analyzing data on child language production.
   1980’s: Brian MacWhinney and Catherine Snow proposed
    setting up a computerised data-base that would be available
    for the entire academic community.
                   CHILDES
           http://childes.psy.cmu.edu

   the database can be accessed worldwide, to which
    relevant material can be contributed in a standard
    format.
   today contains data from 22 lgs
   on both first LA and second lg learning
   from normally and from atipically developing
    children
   from early mother-child dialogues to lengthy
    narratives, monologues
   prosody, stress, pause patterns, metalinguistic
    fators.
   Allows initial assessment of hypotheses.
        Longitudinal vs. Cross-
          sectional studies
   Cross sectional studies: gather groups of children
    to assess their abilities
   compare performance of groups.
   ADVANTAGES: allows to obtain a great deal of
    data in a short time,
   large samples yield results that can be
    generalized.

   Cross Sectional or Longtiud. studies can be either
    Observational or Experimental.
       Observational studies

The researcher tries not to interfere with the
  subject’s natural use of lg.
a.) Naturalistic O.S. - focus on real-life
  situations  linguistic socialization, env.
  factors shaping LA
b.) Controlled O. S. – laboratory setting and
  props are the same for all subjects
  (puppets).
            Experimental studies

Do involve interference from the part of the researcher.
   WUG test - Jean Berko Gleason ’50-ies – Elicited
    Production experiment Using Nonce Words.
   m.i.: when prompted by the linguistic context in
    which the term appears, children will transform
    these so as to obey the grammatical rules.
   picture shown and is told it is called a „wug”. Then
    subject is shown another „wug” and told: „So now
    there are 2 of them.” And asked: „Now I have two…?”
    Expected answer: „wugS”.
   2. Every day I „gorp”. Just like every day, yesterday
    I……….? „GORPED”.
           Observational vs.
          Experimental studies
   OS: we do not know for sure if plural marker „s” is
    consciously put there, or blind imitation
   ES: semi-controlled setting, eliciting the plural
    marker  RULE application  true status of
    grammatical markers

DISADVANTAGE of ES:
 Rigidly controlled settings may mean difficulties,
  such demands of the experimental task may distort
  validity and reliability of results.
 It is therefore important to verify that such patterns
  are present in natural speech.
                 PREVERBAL I.
         Pre-natal Studies of speech
                 perception
Prenatal (fetal):
from 6th month of gestation
method: microphone placed outside the womb,


measure modifications in:
a.) Cardiac acceleration/deceleration:
festus shows sensitivity to changes
in lgic stimuli (habituation)
b.) Fetal movement (kicking rate)
corresponds to changes in stimuli (US)
                  PREVERBAL II.
        Post-Natal Studies of speech
                perception

HASP: High Amplitude Sucking Paradigm
?: categorical perception of infants
capitalizes on natural sucking instinct

                      using sucking rate
                    as measure of interest
                  habituation / dehabituation
                          paradigm




If baby starts sucking vigorously  can
         /continue to decline  can/’t discriminate sounds / m’s voice
          …with older children
  Conditioned head-turn paradigm
(Polka, Jusczyk, Rvachew 1995)
 - with infants 6-18 months of age
-a stimulus sound is presented
repeatedly
   - Whenever there is a change in a stimulus, an
    electric toy (to the side of the child’s line of vision. )
    is lit up and activated.
   Eventually the child learns to anticipate that
    sound changes provoke a visual display, and
    they will turn their head to look for it when
    hearing a sound change.
methods for studying comprehension
…beyond speech (sound) perception
   preferential looking paradigm
(Hirsh-Pasek – Golinkoff 1996).
capitalizes on
differential looking behavior
when given a choice from 2 displays

   child grammar: passive voice
   lexicon: pictures of objects, animals (nouns)

   m.i.: child will look longer in the direction of the named object
    if they understand the word / construction.
   if neither object corresponds to the experimenter’s label, then
    they will look at both stimuli for eqivalent periods of time.
   the same holds if the baby does not know the word.
          …in the verbal stages
   elicited response, interviews (monologue, retelling an
    adventure>5 yrs).
 Off-line techniques
measure comprehension by getting children
a.) perform an action (act out)
b.) retell the event after hearing a linguistic stimulus
c.) picture pointing – subject is asked to point to the pic.
   that depicts the action in the experimenter’s utterance.
The dog pushes the cow                vs.
The cow is pushed by the dog          (active vs. passive)

DISADVANTAGE:
experimenter influences answers (multiple choice)
- involves conscious decisions and reflections by the child
    (performance limitations)    (metalinguistic competence.)
           On-line techniques I.
-   measure language processing as it occurs in real time.
-   e.g.: distinction between transitive vs. intransitive verbs.

-   I. Task: to listen for a particular word, e.g. „DOG”.
-   The subject is asked to push a button as soon as they hear
    the target word.
-   a.) The boy was out on a walk and he watched a DOG passing by.
-   b.) The boy was out on a walk and he watched to*a DOG passing
    by.
-   If they are sensitive to in-/transitive distinction, it will take
    longer to react in the erroneous case (watched to* a DOG…)
-   because the processing of ungrammatical structures will take
    longer if they have already acquired the correct grammatical
    rule.
         On-line techniques II.
-   Non-invasive brain imaging studies: HD Event
    Related Potential (ERP) Studies measuring electric
    activity in brain areas.

-   Eye-tracking experiments (head-mounted and
    remote eye-tracking systems) (syntactic parsing,
    topicality, idiom research looking time –
    segmentation correlation).

-   Reaction-time techniques: computers measure
    reaction time in milliseconds. Time correlates with
    parsing difficulties (gaden-path sentences).
    Advantages vs. Disadvantages
        of on-line methods
-   ADVANTAGES:
-   fast, accurate,
-   reveals fine-grained differences in performance,
-   memory is not overburdened, does not require reflection,
    since it is unconscious.
-   reveal competence off-line don’t (focus on comprehension)
-   online: investigates the interaction between competence
    and performance factors.

-   DISADVANTAGES:
-   difficult to design,
-   fallible, detail sensitive.
               Conclusions

   Innovative Experimental techniques
    counterbalance the limitations of a purely
    observational research
   Passive is more
   Processing paves the way
   Experimentally gained data needs to be
    complemented by observations of natural
    linguistic performance (social factors).
                                             Thank you
                                                for your
                                              attention!
Literature:
Gleason, J. B. - Ratner, N. B. (1998). Psycholinguistics. Harcourt Brace
    College Publishers. Orlando, FL.
Karmiloff-Smith, A. – Karmiloff, K. (2002). Pathways to Language. Harvard
    University Press.
Sowden, H. (2008). Developmental Psycholinguistics. On-line methods in
    children’s language processing. Language acquisition and language
    disorders series. John Benjamins, Amsterdam.

								
To top