Langston, Psychology of Language, Notes 4 -- Speech Production
 
I.  Goals.
A.  Production errors.
B.  A production model.
 
II.  Production errors.  We need to rethink the way we’re studying language.  Instead of thinking about how you understand language, we’re going to look at how you produce language.  The main way to tell what’s happening in production is to look at the kinds of errors that people make.
A.  Some of the more common types of errors:
1.  Shift:  One speech segment moves from its appropriate location “That’s so she’ll be ready in case she decide to hits it.”
2.  Exchange:  Two units change place “Fancy getting your model renosed.”
3.  Anticipation:  A segment is produced too early “Go ahead, bake my bike.”
4.  Perseveration:  A segment gets repeated “He pulled a pantrum.”
5.  Addition:  Add a segment from somewhere “I didn’t explain this clarefully enough.”
6.  Deletion:  A segment gets left out “I’ll just get up and mutter _intelligibly.”
7.  Substitution:  Substitute a related segment for the intended segment “At low speeds it’s too light.”
8.  Blends:  Two possible segments become blended into one output “That child is looking to be spaddled (spanked/paddled).”
B.  Where do errors occur?  All levels of production (phonological, syntactic, lexical insertion...).  But, you don’t see mixes of levels.  So, for example, phonological errors usually occur in syntactically correct sentences.
C.  Common properties of errors:
1.  Elements that interact come from similar environments.  Beginnings change with beginnings, ends with ends, etc.
2.  Elements that interact tend to be similar.  Consonants don’t replace vowels.
3.  Even when errors produce novel linguistic items (spaddled) they’re consistent with the rules of the language (in this case spelling rules).
4.  Errors will have the same stress pattern as the thing they’re replacing.
D.  Where do errors come from?
1.  Freud:  Errors are Freudian slips.  It’s a way to look at what someone is really thinking.  You have a lot of stuff going on in your unconscious, and some of it’s bad and being repressed.  But, repressed stuff sneaks out in production errors.  For example, there’s a cartoon with a king and a queen, and the king is saying “Good morning beheaded... er, beloved.”  The slip is that she’s about to be beheaded...
Experimental evidence:  Take male subjects and put them in a room with a scantily clad female experimenter.  Errors like saying “fine body” instead of “bine foddy” when reading out loud under time pressure indicate that errors are based on impulses from the Id.
2.  ReverseSpeech:  To fit with the unconscious message being uttered in reverse, the forward message will sometimes need to be off a bit.  More on this later.
3.  Psycholinguistics:  The errors come from the complexity of producing language.  A lot of stuff is happening, it’s happening fast, and it’s happening in a system with limited capacity.  All of that makes errors likely.  What is the production system?
 
Top
 
III.  A production model.  The basic model has four parts:
A.  Conceptualize:  This buys into the language of thought idea (that thoughts are in mentalese, not language).  You have to have a thought before you can begin to produce an utterance.
B.  Form a linguistic plan:
1.  Identify meaning:  It’s kind of like looking in a dictionary backwards.  You have a thought you want to express, and you look for definitions that match the thought.
2.  Select syntactic structure.  We will have a rule in our phrase structure grammar that says S -> NP + VP (a sentence is a noun phrase plus a verb phrase).  The real grammar has many possible sentence types.  Choose one here and build the tree (like Chomsky’s toy grammar).
3.  Generate an intonation contour.  Fans of Seinfeld will know that if Jerry’s not sure if he’s invited to a party and the host says to Elaine “Why would Jerry bring something?” it’s not the same as “Why would Jerry bring something?”  In the one case, Jerry’s probably not invited.  In the other case, he’s invited, just not expected to bring anything.  This step is where you lay out stress patterns.
4.  Insert content words:  Put in the words.
5.  Form affixes and function words.
6.  Specify phonetic segments:  Figure out how it’s pronounced.
7.  Edit:  Look at the “darn bore” task.  People are much more likely to say “barn door” after the biasing context than “bart doard.”  That’s because “barn door” makes words (even if it’s not the intended utterance), but “bart doard” doesn’t.  So, people probably have an editing function at the end of the plan to check for basic errors.
Two things to say about this model:
1.  Errors support dividing the plan into these particular steps.  For example, if you say “Stop beating your brick against a head wall” it’s a problem at only stage four (insertion).  Everything else seems to have gone off OK.
2.  Errors support this particular sequence.  For example, if you say “It certainly run outs fast” and pronounce the ‘s’ in “outs” as /s/, it indicates that the sound part came after the affix part (because if the sound came first, the ‘s’ would sound like /z/, since that’s how it would sound in “runs”).
C.  Implement the plan:  Once you make a plan, you have to say what you came to say.  What do I want to say about this?  That planning and production seem to go in cycles.  If you graph time on the x-axis, and amount produced on the y-axis, it looks like people plan a while, say that much, plan some more, say some more, etc.  Why?  Because it’s a hard task, which probably maxes out working memory all the time.  You have to do it in cycles.
D.  Self-monitoring:  Obviously, some errors get by the editing step.  What happens when errors get out?  Stop and make a self-repair.  Three steps:
1.  Self-interruption:  Signal that you’ve spotted the error and stop.  A procedure to get a lot of this is to have people describe colored diagrams (like the one below).  They have to describe them so someone else can draw them.  People make a lot of mistakes, and have to correct themselves a lot.
 
Colored diagram
 
a.  Some statistics about when people interrupt themselves:  Interrupt within a word:  18% of the time “We can go straight on to the ye-, er pink.”  Interrupt after the word:  51% “Straight on to green- to red.”  Interrupt at some later point:  31% “and from green left to pink, er from blue left to pink.”
b.  How do they interrupt themselves?  Utter an editing expression to indicate trouble (generally “uh”).
c.  Self-repair:  Fix the problem.
 
Top
 

Psychology of Language Notes 4
Will Langston

 Back to Langston's Psychology of Language Page