Psychological Mechanisms and the Perception of Language

1
Psychological Mechanisms and the Perception of
Language
2
Part 1 Information Processing Models

• Currently most models of language comprehension
  and/or production fall within the category of
  Information Processing Models.

3
Key Concepts for IPMs

• Memory
• Working memory
• Permanent memory
• Sensory memory (Some theories refer instead to
  receptive and productive gestalts for processing
  within each sensory input system.)
• Receptive channel
• Productive channel
• Capacity

4
Diagnosis Corner - Dc

• Assume that memory is a central IP process. I am
  faced with the following scenario.
• Wandering around a conference hotel, I step into
  a session in progress. At first it sounds very
  much like a talk about new directions in the
  recovery of language functions for brain-injured
  children. I start to have trouble understanding
  the material within a very short time.

5
Dc continued

• The seminar is really about decoding the
  language of dolphins. The presenter is using
  spectrographs to provide physical evidence of
  prosodic variation that may have semantic
  importance. In other words, the researcher is
  trying to understand better how the creatures
  communicate under water.

6
Dc continued

• IP suggests that I am depending mostly on memory,
  both to follow the presentation, and then to
  decide that I had better get up and leave,
  because I am out of my league.

7
Dc continued

• Memory I have a set of concepts that I have
  stored. They are based on past learning that
  overlaps somewhat, but not enough, with the new
  material.
• Working memory I compare new concepts with old
  ones constantly to make sense of what I am
  hearing.
• Sensory memory (mostly visual and auditory)
  allows me to double check new information - from
  slides and from researcher speech - against what
  I think I already know.
• Permanent memory Some of this new information
  is going to become part of my constant store of
  information. As I process, I store in memory. Who
  knows? I may draw on these new ideas at a later
  time.

8
Dc continued

• The new material is entering through the
  receptive channel. There are as many of these
  as there are senses used to receive information.
  Sensory material is then passed onto Memory to be
  coded and organized.
• Fortunately I wont be asked to repeat or retell
  what I just heard. But if that did happen, I
  would be using the productive channel to
  recreate the information. Organized information
  would be fed through by working memory. The
  details of course would be retrieved from
  permanent memory.

9
Dc continued

• At a very profound level, the inability to
  understand and recreate experience using
  receptive and productive language speaks to
  impairments of IP. This is one area where
  diagnosis of language difficulties can start.
  There are other ways of understanding where to
  begin diagnosis that are layered on top of these
  very basic processes.

10
Capacity

• The whole or entirety of ones psychic potential.
• Capacity is like space on your hard drive. The
  more you tax it with serial and/or parallel
  processing tasks, the less there is available for
  new work.
• Lets spend some time thinking about how that
  would work in a variety of language contexts.

11
Why Do We Need to Think About Capacity?

• A professor is constructing a series of power
  point slides about a technically complicated
  topic. As she types, she also modifies her
  language to move it away from the more technical
  jargon (which is in fact how she understands the
  topic herself). She is interrupted by an
  administrator who asks her a question about
  Roberts Rules of Order. Can we construct a model
  of all the psychological and linguistic acts that
  must be occurring and taxing her capacity?

12
Quick Break for Vocabulary Review

• Serial processing A particular mental strategy
  is dependent on one that occurs before it.
• Parallel processing Two or more mental
  strategies that can or must be engaged at the
  same time.
• Top-down versus bottom-up processing A global
  or holistic understanding is constructed and is
  later analyzed in terms of part/whole
  relationships. Versus, specific features of a
  problem that are processed first and later
  connected to form an integrated pattern.
• Automatic versus controlled processes Some
  processing must occur with automaticity in order
  to be functional. Other processes require
  conscious monitoring in order to be functional.
• Modularity A range of variation in behavior is
  possible within normative limits. Beyond those
  limits the behavior is unintelligible.
• .

13
Serial processing A particular mental
strategy is dependent on one that occurs before
it.

• Twas brillig
• I decode this in a particular order of
  sound/letter relationships.

14
Parallel processing Two or more mental
strategies can or must engage at the same time.

• And the slithey tove
• While I am combining sound/letter relationships,
  I am also creating gestalts, organized
  information.
• That is, I am trying to group these patterns to
  find their meaning.
• Slithey is sounding like something I heard of
  before. So is tove.

15
Top-down versus bottom-up processing

• Is it global first, and then I begin to work in
  the particulars?
• Or do I deal with the particulars and see how
  they fit together?
• Lets see.

16

• did gire and gimble in the wabe
• Top-down, I start trying to form a gestalt that
  two events or acts are occurring in the same
  location (as in toss and tumble in the waves).
• Bottom-up, and I return to the statement several
  times, as I try to decide which elements support
  my overall interpretation.
• It is important to keep in mind that this happens
  reciprocally. It is not a matter of one or the
  other.

17
A language event.How do IP models work in
theory?

• Twas brillig and the slithey tove did gire and
  gimble in the wabe
• HuH!
• By using organized nonsense it is possible to see
  how quickly the mind springs into gear and tries
  to create meaning from nonsense. A number of
  psychological processes are brought into play.
  Can you name a few as you monitor your own
  awareness of thought?

18
The Diagnosis Corner

• By making a list of all the problems you are
  encountering when trying to translate Lewis
  Carrolls line of poetry, you might end up with a
  fairly complete set of problems that anyone faces
  when trying to process language that is new,
  difficult, or never heard it before. Try it out
  and see what happens.

19
Part 2 Perception of Language

• Mostly it is all about phonics and prosody.

20
Modularity

• We need to contrast two opposing concepts
• Invariance
• Individual variation
• This discussion becomes extremely important when
  thinking about dialect variation.

21
Invariance

• To be understood, we have to utter sounds in an
  intonational range that is perceptible.
• To be understood, the sounds themselves have to
  fall within a set of parameters that set a
  bandwidth for variation in sound.

22
Perceptibility and Invariance

• The sound heard for e has a set of parameters
  for phonic variation outside of which it is no
  longer recognizable as long e.
• It can vary in the English language from what
  would roughly be eeee to ay depending upon
  regional factors.

23

• Invariance is a requirement for intelligibility.
• Consider the implications of this simple
  principle for victims of brain trauma to language
  centers, individuals with motor impairments of
  the speech apparatus, children who are exposed to
  adult speakers who vary in the extreme from what
  is modular.

24
Individual variation

• Within the bandwidth that establishes modularity
  much variation actually occurs. That is what
  makes my voice distinguishable and gives me a
  speech signature that is almost as good as my
  fingerprints. There are also ranges of variation
  that establish the bandwidth for dialect
  variations in phonological contour. Some dialects
  vary so greatly in terms of sound patterning that
  they are not intelligible to speakers of the same
  language but of different dialect. The sound
  contour of Italian spoken in Turin is virtually
  unintelligible to a Neapolitan speaker (excluding
  semantic and syntactic variation in dialects).
  When differences are that great, the principle of
  modularity has been breached. Confusion reigns.