Human Cognitive -Read Entirely FIRST!!

xzevo103
Chapter9Powerpoint.ppt

Language I:

Basic Issues and Speech Processing

Language Basic Principles

Linguistics and Psycholinguistics

Psycholinguistics

The study of psychological processes involved in producing and comprehending language

Linguistic performance

Linguistics

The study of language in its pure form

Linguistic competence

.

*

Words and Rules

  • Language

A set of symbols (words) and rules (grammar) for the combination of symbols that allow for communication and comprehension

*

Design Features of Language

*

*

*

*

*

*

Animation put asterisks by defining features discussed in book

*

Language in Non-human Animals

  • No doubt animals possess communication systems

Honeybees

Vervet monkeys

...but do they possess the design features of language? of language….

Yes

No

No

No

No

No

Design Feature
Semanticity
Arbitrariness
Displacement
Productivity
Prevarication
Reflectiveness

*

Language Training Projects

Do nonhumans have the capability to learn and use a system of words and rules?

Studies conducted with:

Common chimpanzees

Bonobo chimpanzees

Sea lions

Bottle-nosed dolphins

African gray parrots

Many would argue that performance does not rise to the level of human language

Labeling: Is It Word Learning?

All language training projects involve teaching labels for salient objects in the environment

Non-human animals are able to learn these labels

Does this represent linguistic ability?

Some would say no

They associate labels with objects, but labels are not referential

The label is associated with the object, but does not represent the object in the animal’s mind

*

Language as Learning to Do as You are Told

Some language training projects involve nonhuman animals learning how to follow commands

Learning to follow a command requires an understanding of word order

Example: difference between bring the ball to the trainer vs. bring the trainer to the ball

Does this represent linguistic ability?

Some would say no

Only perform commands to get reward

*

Language as Engaging in Social Routines

Language is social

Intentional and referential communication between at least two people

To assess this ability, cross-fostering studies have been conducted

American Sign Language taught to chimpanzees within context of daily interactions

Eventually, chimps began to produce the symbols

*

Does this represent linguistic ability?

Some would say no

In order to learn symbols, experiments had to shape the hands of the chimps

Once learned, symbols had to used to gain access to desired activities

*

Kanzi and the Bonobos

Most impressive evidence of language in non-humans

This work is noteworthy for several reasons

Language was learned spontaneously

Kanzi learned lexigram symbols by watching researchers use them with his mother

*

Language evolved through his everyday activities; did not attempt to engage in human-like conversation

Led to a great deal of data—13,000 utterances generated over a 4-month period

Analysis revealed he:

could generate novel combinations of symbols

could distinguish between different types of symbols (nouns and verbs)

could place symbols in appropriate slots in an utterance

made up his own grammatical rules and consistently used them

*

Some do not believe this represents true linguistic ability and argue for:

Nativist view of language

Language is based in brain systems that only evolved in humans

Some argue that recursion is a language characteristic that is unique to human

Embedding sentences within sentences

*

Levels of Analysis

Phonology

Morphology

Syntax

Semantics

Pragmatics

*

Phonology: The Sounds of Language

  • Analysis of the basic speech sounds

/ / represents the sound, apart from spelling

Acoustic structure can be viewed with the use of a speech spectrograph, which produces a spectogram

Plots sound waves of differing frequencies that result from speech

*

Phones and Phonemes

  • Phone (phonetic)

An acoustically discriminable language sound ("a" sound in cat, vs. "a" sound in bat)

Phonetic differences do not change the meaning of a word

Phonetic level is acoustic—the physical sound that we produce

*

  • Phoneme (phonemic)

Speech sound represented by a single symbol

letters

Contain no meaning

Phonemic differences do change the meaning of a word

About 200 phonemes across all known languages

44 in English

Practice

Bat

3

Chat

3

Love

3

Find

4

Trick

4

*

Producing Phonemes

  • Consonants

Air flow in the vocal tract is obstructed

Formed by varying

Place of articulation

Where the airflow is obstructed

Manner of articulation

How the airflow is obstructed (stopped, just interrupted, etc.)

Examples:

Stop: complete closure (temporary built up of air pressure) with release (top)

Fricative: narrowing of two articulators, such that a turbulent airflow is produced (see)

Affricates: a stop consonant with a slow release (stop + fricative) (Chicken)

Voice onset time

Temporal relationship between release and vocal cord vibration

Voiced: release and vocal cord vibration occur almost simultaneously (“b”)

Voiceless: release is followed by vocal cord vibration (“p”)

Vowels

Vocal tract is open

Formed by varying placement of the tongue

Vertical: high – mid – low

Horizontal: front – central – back

*

Phonemes also differ by suprasegmental factors

Rate, stress, and intonation

Coarticulation

A given phoneme sounds different depending on neighboring phonemes

Phonemes are articulated simultaneously

*

Perceiving phonemes

Despite coarticulation….

  • Categorical Perception

Tendency to perceive phones within a phoneme category as the same

Phonemes are invariant across different contexts

Cat

/p/

/p/at

/p/et

/p/it

/p/ot

/p/ut

*

Liberman, Harris, Hoffman, and Griffith (1957)

Used speech synthesizer to produce and present speech sounds that differed on a continuum (b-d-g)

Increased voice onset time in 5 millisecond increments

Despite continuous variation in the acoustic signal, participants did not classify the sounds continuously

Classified sounds into 3 distinct categories corresponding to the phonemes /b/, /d/, and /g/

*

Phoneme boundaries

Differ for every language

Development of language specific phoneme boundaries occurs early in infancy (6 months)

“Tune in” relevant phonemic differences

“Tune out” irrelevant phonemic differences

*

The Importance of Context

The role of top-down processing in perceiving phonemes

Phonemic restoration effect 

We tend to perceptually “replace” sounds that are missing from a speech signal

*

Warren (1970)

Presented with one of four sentence:

It was found that the *eel was on the axle

It was found that the *eel was on the shoe

It was found that the *eel was on the orange

It was found that the *eel was on the table

* phoneme replaced by cough

What did subjects report hearing?

Wheel

Heel

Peel

Meal

*

Morphology: From Sounds to Words

Combining phonemes into morphemes

Morpheme

Smallest unit of language that has meaning

Words, prefixes, and suffixes

Types

Free morpheme

Word or part of a word that may stand alone

Bound morpheme

Part of a word that may not stand alone (suffix or prefix)

*

Practice

Loved

Two

1 free (“love”)

1 bound (“-ed”)

Unhappiness

Three

1 free (“happy”)

2 bound (“un-” and “-ness”)

Game

One

1 free (“game”)

*

Producing (or “Morphing” the Spoken Word)

  • How do we change the form of words? (i.e., singular to plural, present to past, etc.)

Two Systems

Regular transformation (askasked; add –ed)

Irregular transformation (know  known)

*

*

*

Perceiving the Spoken Word

Theories

Cohort model

When first phoneme is uttered, a set of possible matches is activated in LTM

With each successive phoneme, the set is reduced until a match found

*

Allopenna, Magnuson, and Tanenhaus (1998)

Procedure

Given spoken instruction to interact with simple objects (e.g., beaker)

Target object was presented with other phonologically similar distractor objects (e.g., speaker and beetle)

Eye movements were tracked as they followed the instructions

Results

Participants glanced at the object that began with same phoneme as the target item (e.g., beetle)

*

Problems with the model

Places importance on perceiving the initial phoneme

Difficult given the continuous acoustic nature of speech

Relies solely on phonology (as do other word recognition models) and ignores prosody

Rhythm, stress, and intonation of speech

*

Nygaard and Queen (2008)

Investigated the role of emotional prosody in word recognition

Pitch and tone associated with emotional speaking

Independent variables

Word valence: Happy, sad, and neutral words

Tone: presented in happy, sad, or neutral tones

Task: repeat the word after hearing it

Dependent variable: RT to say the word

*

Results

Congruency effect

RT to say the word was faster when valence and tone matched than when they mismatched

*

Discerning Word Boundaries

  • Weak relationship between breaks in the acoustic speech signal and breaks between words

Never touch a snake with your bare hands

*

Bottom-up factors

Phonotactic knowledge

Sensitivity to the rules that govern phoneme combinations in a given language

Example: jp never occurs in English

Metrical segmentation

Phonological regularities of a given language

Example: In English, content words tend to start with strong syllables and end with weak syllables

*

Using Statistical Probabilities to Detect Word Boundaries

Statistical regularities present in the speech sequence is used to detect word boundaries

Saffran, Aslin, and Newport (1996)

Phase 1

Presented infants with a two-minute continuous speech stream composed of 4 different 3-syllable nonsense words strung together in random order

Bidaku/padoti/golabu/fotavi

*

Phase 2

Two types of test trials

Previously defined “words” heard by infants

New experimenter defined words containing the same syllables, but in new combinations

Old: bidaku/padoti/golabu/fotavi

New: kupa/dotigola/avibida/bufot

Infants controlled listening time by staring or not staring at a blinking light

Infants prefer “novelty”

*

*

Novel

*

Old

And look away when hearing old words because not novel

*

Prediction

If picked up on words from phase 1

Will look at blinking light when new “words” are heard

If did not pick up on the “words” from phase 1

Will look at light for old and new “words”—both are novel

Results

Animation reveals results

*

Top-down factors

Knowledge of the words in a given language help locate word boundaries

IRANDOWNTHESTAIRS

MYDOGCHASEDTHECAT

THEREDONATEAKETTLEOFTENCHIPS

Visual analog--Just like these examples, speech is a continuous stream of information, and we must impose boundaries

Parse this sentence into words—your knowledge of the words in our language help you to do it—did you get confused at beginning? Yes because Iran is a word in our language. But once you got past that, the rest was easy because down the stairs is the only parsing that works in our language. My dog chased cat—easy as no other interpretation. There donate a kettle of ten chips/the red on a tea kettle often chips/There don ate a kettle of ten chips—multiple interpretations as the boundary between words is difficult.

Consider what would happen if you had little or no familiarity with a language. Have you ever listened to a person engaged in the fluent speaking of a language other than English? It sounds like they’re talking a mile a minute, but really, this is a sort of illusion induced by your lack of familiarity with the language. Because you have only minimal (or no) knowledge of the words in that language, parsing them is nearly impossible.

Think about taking cognitive psychology in Norwegian? Would be hard.

Even in our own language we often we make mistakes

*

Mondegreens

Misheard word boundaries

  • Often experienced in song lyrics

Putting speech to music leads to distortions in the typical acoustic signal for speech

When the data for bottom-up processing is distorted, we apply top-down processing

Extensive experience with artist or type of music

Knowing what the person is singing about

Could additional bottom-up information help?

*

Massaro and Jesse (2009)

Independent variables

Stimuli

Spoken speech

Sung speech

Computerized “talking head”

Present

Provided additional bottom-up information of facial and lip movements

Absent

Identification was better with talking head, but effect was smaller for sung than spoken speech

So the distortion created by setting morphemes to music cannot be completely overcome by additional bottom-up information.

*

The Mental Lexicon

Mental lexicon

General knowledge representation of words in semantic memory

Lexical Access

Process by which a word is activated within the mental lexicon

*

Factors Affecting Lexical Access

Word frequency

How commonly a word occurs in one’s day-to-day linguistic encounters

High frequency words are more easily and quickly accessed than low-frequency words

*

Foss (1969)

Phoneme monitoring task

Listen for a target phoneme (e.g., /g/) while attempting to comprehend a speech passage

Target phoneme occurred after a:

low-frequency word

high-frequency word

Dependent variable: RT to detect the target phoneme

Results

RT was longer after a low than high-frequency word

Additional resources are needed to access a low-frequency word, slowing RT

*

Lexical ambiguity

A word that has more than one meaning

Does encountering an ambiguous word, activate all representations or just the one relevant to the context?

*

Foss (1970)

Phoneme monitoring task

Listen for a target phoneme (e.g., /g/) while attempting to comprehend a speech passage

Target phoneme occurred after:

an ambiguous word

a non-ambiguous word

Dependent variable: RT to detect the target phoneme

Results

RT was longer after ambiguous than non-ambiguous words

Both representations were activated and additional resources are needed to disambiguate, slowing RT

*

The Bilingual Lexicon

Separate stores approach

Two separate mental lexicons; one for each language

The one that is accessed corresponds to the language being used

Support

Priming in lexical decision tasks is stronger for same word than cross-language equivalent prime

“cat-cat” > “gato-cat”

Priming would be the same if gato and cat shared one representation in the mental lexicon

*

Common stores approach

One common mental lexicon that is accessed by all languages

Support

RT to decide if a concept (gato, or cat) is a member of category (animal or sustanivo) is the same for concept-category match as mismatch

Cat-animal = gato-animal

RT would differ if two representations were accessed

*

Preponderance of the evidence supports the common-stores approach, but:

Depends on the nature of the lexical item

Cross-linguistic cognates (flute and flauta) are more likely to share a common representation

Depends on the nature of the bilingual individual

Late bilinguals (2nd language late in development) tend to have separate lexicons for each language

Those who learn both languages early in development tend to have a common lexicon

It is possible that bilinguals can choose which mode to use depending on task demands

*

Entrance Into the Mental Lexicon: Word Learning

Estes, Evans, and Alibali (2007)

Investigated if infants could associate new learned “word” sequences to objects

Procedure

Phase 1

Used Saffran, et al. (1996) procedure to teach infants new “words”

*

Phase 2

Word-labeling habituation task

Object is moved about while hearing a “word” label

Conditions

Word condition

“word” was learned in phase 1

Non-word condition

“word” was not learned in phase 1

“Word” label-object pair shown until habitation

Infant stops looking at moving object

*

Phase 3

Object presented with a word label

Conditions

Same condition

Same label associated with “word” in phase 2

Switch condition

Label was associated with a another “word” in phase 2

*

Prediction

If an infant learned the “word” label in phase 1 and associated label with a specific object in phase 2

Infant should:

be surprised by the new label (switch condition)

not be surprised by the same label (same condition)

Consequence

Infant should stare longer at the object in the switch condition than the same condition

Results

*

Syntax and Semantics: From Words to Sentences

Syntax

Set of rules that specify legal combinations of words within a given language

Semantics

Set of rules that govern the effective transmission of meaning

The boy kissed the girl

The girl was kissed by the boy

Boy the girl the kissed

Colorless green dreams sleep furiously

The cat chased the dog

His insecurities chased him in his dreams

1 and 2 are the same semantically but differ syntactically

3 is syntactically and semantically invalid

4 is syntactically valid but semantically invalid

5 is syntactically and semantically valid but unusual based on our semantic knowledge

6 is syntactically and semantically valid but violates the literal meaning of chase yet we understand the meaning is metaphorical

*

Dissociation that supports distinction between syntactic and semantic processing

Broca’s aphasia

Associated with frontal lobe damage

Loss of syntactic ability, but preserved semantic ability

Wernicke’s aphasia

Associated with temporal lobe damage

Loss of semantic ability, but preserved syntactic ability

*

Transformational Grammar

Noam Chomsky argued that language cannot be learned due to the “poverty of the stimulus”

Children apply rules (represented physiologically from birth) to the symbols of language

Develops in the absence of formal “teaching”

True regardless of the language being spoken, so approach is often referred to as universal grammar

*

Phrase Structure, Surface Structure, and Deep Structure

  • Phrase structure

The rules that define:

the fundamental components of a sentence

the types of words that comprise these components

the ways that words may be arranged

…but phrase structure is insufficient to explain the production and comprehension of a sentences

*

  • Sentences exist at two levels

Surface structure

Superficial external version of sentence

Syntax

Deep Structure

Meaning of the sentence

Semantics

Sentence generation

Phrase structure rules used to create the deep structure

Transformational rules are applied to the deep structure to create the surface structure

*

Challenges to Chomsky

(Non) Universal Grammar?

Evans and Levison (2009) review

Diversity of the world’s languages makes it improbable that a set of features describes all of them

Estimated that current languages only constitute 2% of all languages ever spoken

Current linguists and psycholinguists only study 10% of that 2%!

Making universal language claims based on studies of this limited range of languages seems invalid

*

Pinker and Jackendoff (2009)

Acknowledge linguistic diversity is often ignored

Disagree that no aspect of language can be considered universal

Semanticity  all languages convey meaning

Universal grammar is a set of abilities that allow a person to develop language

These abilities manifest themselves regardless of the specific linguistic context

*

Language: Not So Unlearnable?

Constraint-based approach

Language is filled with probabilistic constraints that can be used during language learning

For example, through linguistic experience we learn that:

Some meanings of words are more probable than others

The structure of a sentence constrains how parts of the sentence can be interpreted

*

Pinker (2007)

Probabilistic constraints used to learn how syllables combine into words don’t generalize to combining words into sentences

Words can be constructed in a finite number of ways

Sentences can be constructed in an infinite number of ways

Sentence and syllable sequencing are fundamentally different

Sentence sequencing isn’t just about sequencing words, but hierarchically organizing them to convey meaning

Syllable sequencing does not require this

*

Child Directed Speech (CDS)

  • Adult speech with infants

Altered prosody

High pitched

Exaggerated ups and downs

Slower cadence

Well-articulated and exaggerated phonemes

Found in all cultures

Infants show preference for this type of speech relative to normal adult speech

*

  • Altered prosody of CDS

Serves two functions

Linguistic function

Helps distinguish non-informative, nonlinguistic sounds from informative, linguistic sounds

Helps segment speech signal into words by isolating them with exaggerated pauses

Conveys syntactic rules

Emotional

Positive affect draws infant’s attention to important aspects of the environment

*

Burnham, Kitamura, and Vollmer-Coona (2002)

We talk to pets in a similar way as we talk to infants

Limited to pitch and affect (emotional prosody), but not vowel hyperarticulation (linguistic prosody)

*

Kempe, Shaeffler, and Thoresen (2010)

Investigated the relationship between the emotional and linguistic prosodic cues to syntax

Varied the emotional relationship between adult and child

High affect: mother talking to own child

Low affect: non-mother talking to child

*

Task

Convey a syntactically ambiguous instruction (touch the cat with the spoon) that represented one of two meanings

Prosodic indication for this interpretation is a pause between cat and spoon

Prosodic indicator for this interpretation is the lack of a pause between cat and spoon

Animation reveals which picture the prosodic indicators represent

*

Results

CDS prosody provided disambiguating information, but only for non-mothers

Mothers lengthened all pauses (emotional prosody) not just the one between cat and spoon

Conclusions

Non-mothers use prosodic cues to syntactic structure to maximize the effectiveness of the communication

Emotional prosodic cues (exaggerated pauses) diminish prosodic cues to syntax

Mothers use CDS to forge a bond with the infant, rather than produce effective communication

Makes evolutionary sense

*

…other evidence for the evolutionary value of CDS

Put the baby down hypothesis

Early in the evolution of language, mothers needed to put their baby down to complete daily tasks

Song-like vocalizations kept the baby content

*

Pragmatics: The Social Aspects of Language

  • Practical knowledge we need to use language effectively

Conversational Structure

Characteristics

Conversations start with one of several standard greetings

Most of the time, standard greetings are met with standard response

Conversations involve turn-taking; conversational overlap is rare

Turn-taking styles vary and tend to be a stable characteristic of a person’s conversational style

  • Three principles that guide turn-taking in conversations

Current speaker may choose the next speaker by directing a comment or question to a particular person

If this does not happen, it is acceptable for anyone to fill the conversational gap

The original speaker may continue talking

  • Four Gricean maxims

Contract between speakers that has the following characteristics

Quantity

Say as much as you need to, but not more

Quality

Don’t say things that you believe to be false or for which you lack evidence

Relation

Be relevant to the topic at hand

Manner

Be clear; avoid obscurity and ambiguity

Gender and Conversation

Conversational differences between men and woman

Woman tend to be more polite than men

Woman tend to use indirect requests

Meaning communicated does match the explicitly stated meaning of the request

Woman tend to employ tag questions and hedges

Tag questions: questions added to the end of a statement like “isn’t it?”

Hedges: qualifiers like “kind of”

Male conversations feature more interruptions

*

What do these gender differences indicate?

Reflect power differences between men and woman

Simply reflect differences in way the gender’s prefer to communicate

Some argue that gender differences are not consistent

*

Mehl, Vazire, Ramírez-Esparza, Slatcher, and Pennebaker (2007)

Investigated the oft-cited 3:1 female to male dominance in conversations

Used an electronically activated recorder (EAR) to record natural conversations

Recorded for 30 seconds every 12.5 minutes  extrapolated likely daily output for the day

400 individuals wore the EAR during waking hours

Results

Conversation output was the same for men and women

*

Thomson, Murachver, and Green (2001)

Speakers change their conversation style so that it matches the style of the person with whom they are conversing

Men talk like women when conversing with women and vice versa

*

Putting It All Together: Speech Perception and Production

Speech Production

Stages of Speech Production

  • Conceptualizing

Deciding what we want to say

  • Planning

Developing a linguistic plan–organizing our thoughts in terms of language

  • Articulating

Executing the linguistic plan

  • Self-monitoring

Tracking what we say to determine if message and tone are as intended

*

Conceptualizing

Difficult to investigate, so little is known about it

Planning and Articulating

Slips of the tongue serve as a rich database for investigating the stages of language production

*

Slips of the tongue

Errors in which sounds or entire words are rearranged between 2 or more different words

8 types

A thee prage paper (three page paper)

  • One segment disappears from its appropriate location and appears somewhere else
  • Shift

R disappears from three and appears in page

*

The cookie is ginormous (gigantic or enormous)

  • Occurs when more than one word is being considered and the two intended items "fuse“ into a single item
  • Blend

Gigantic and enormous fuse into ginormous

*

The new “Winnie the Pooh” Blue Roo disk (Blue Ray disk)

  • Earlier segment replaces a later segment, but does not disappear from its appropriate location
  • Perseveration

Earlier “oo” sound in Pooh or Blue places the later “ay” sound but does not disappear from its appropriate location

*

What’s the feather forecast (weather forecast)

  • Later segment replaces an earlier segment, but does not disappear from its appropriate location
  • Anticipation

Later ‘’f’ sound replaces an earlier “w” sound but does not disappear from its appropriate location

*

Please untie your shoe (tie your shoe)

  • The addition of linguistic materials (added linguistic unit is not from the rest of the utterance)
  • Addition

“un’’ is added to “tie”

Joe Nathan is done—he needs Jimmy John surgery (Tommy John)

*

She is occupied with the problem at work (preoccupied)

  • The removal of linguistic material
  • Deletion

“pre” is removed from occupied

*

I want to bank my thoss (thank my boss)

  • Two segments exchange places; both segments disappear from their appropriate location
  • Exchange

“b” and “th” exchange places, both disappearing from their appropriate location

*

We destroyed the major (revised the major)

  • When one segment is replaced by an intruder (source of the intrusion is not in the rest of the utterance)
  • Substitution

destroyed replaces revised

*

Be Careful !!!!!!

I straw the boy on the seet (saw the boy on the street)

Shift

Remember it is sounds that are moving

Practice:

Shift-- tr is moving NOT s exchanging with str

*

Slips of the tongue can occur at any linguistic level

Phoneme level

Morpheme level

Word level

But within an utterance, slips only occur at one level

*

A thee prage paper (three page paper)

  • Shift
  • Phoneme level

The cookie is ginormous (gigantic or enormous)

  • Blend
  • Word level

The new “Winnie the Pooh” Blue Roo disk (Blue Ray disK)

  • Perseveration
  • Phoneme

What’s the feather forecast (feather forecast)

  • Anticipation
  • Phoneme level

Please untie your shoe (tie your shoe)

  • Addition
  • Morpheme level

She is occupied with problems at work (preoccupied)

  • Deletion
  • Morpheme level

I want to bank my thoss (thank my boss)

  • Exchange
  • Phoneme level

We destroyed the major (revised the major)

  • Substitution
  • Word level

Consistencies in slips of the tongue

Interacting elements tend to

come from similar positions within a word

be similar to one another

receive the same stress they would have if they had not interacted

Interactions tend to be based on phonological, rather than semantic similarity

Slips obey the rules of phonology

.

*

Explanation for slips of the tongue

Cognitive explanations differ in terms of the process involved in planning and articulation

Stages occur in serial or parallel fashion

Stages

Stage 1

Conceptualization and determination of syntactical structure

Stage 2

Insertion of content words and/or free morphemes into the word slots

Stage 3

Insertion of bound morphemes

Stage 4

Insertion of function words

Stage 5

Overt articulation

*

Stage 1

Idea: In the past, an old man was enjoying listening to music

____ ____ ____ ____ ____ ____ ____

Stage 2

____ ____

Stage 3

____ old man enjoy listen ____ music

Stage 4

old man enjoyed listening music

Stage 5

The old man enjoyed listening to music

old

man

enjoy

listen

ed

music

ing

The

to

The old man enjoyed listening to music

Stage 1

  • Conceptualization and determination of syntactical structure

Stage 2

  • Insertion of content words and/or free morphemes into the word slots

Stage 3

  • Insertion of bound morphemes

Stage 4

  • Insertion of function words

Stage5

  • Overt articulation

*

Blend slips of the tongue

Occur at stage 1

Word-level slips

Occur at stage 2

Morpheme level slips

Occur at stage 2 if a free morpheme

Occur at stage 3 if a bound morpheme

Phoneme level slips

Occur at stage 5

A thee prage letter (three page letter)

  • Shift
  • Phoneme Level
  • Stage 5

The cookie is ginormous (gigantic or enormous)

  • Blend
  • Word Level
  • Stage 1

The new “Winnie the Pooh” Blue Roo disk (Blue Ray disk)

  • Perseveration
  • Phoneme Level
  • Stage 5

What is the feather forecast (feather forecast)

  • Anticipation
  • Phoneme Level
  • Stage 5

Please untie your shoe (tie your shoe)

  • Addition
  • Morpheme level

Bound morpheme

  • Stage 3

She is occupied with problems at work (preoccupied)

  • Deletion
  • Morpheme level

Bound morpheme

  • Stage 3

I want to bank my thoss (thank my boss)

  • Exchange
  • Phoneme level
  • Stage 5

We destroyed the major (revised)

  • Substitution
  • Word level
  • Stage 2

Freudian explanation

Emphasized the role of psychodynamic forces in making certain content more available than others

Slips-of-the-tongue arise from the mutually opposing action of two different intentions/thoughts

Conscious intention of the speaker

Unconscious disturbing thought/intention

The disturbing comment is typically censored

When a slip-of-the-tongue occurs, it represents some aspect of the unconscious intention

*

Motley Baars (1979)

Procedure

Presented with pairs of words or nonsense syllables were presented

HAT – RAM

RUF – GAM

At random times, buzzer sounded, and they were to say the previous pair aloud

Design

Experimental groups

Heightened sexual arousal

Heightened fear of shock

Control group

Concerned with neither sex nor shock

Materials

Some nonsense-syllable pairs were designed to promote sex-related slips

GOXY FURL (FOXY GIRL)

LOOD GEGS (GOOD LEGS)

Some nonsense-syllable pairs were designed to promote shock-related slips

SHAD BOCK (BAD SHOCK)

WOT-HIRE (HOT WIRE)

Some pairs did not suggest slips related to either sex or shock

HOT-GAM

Results

Self-Monitoring

Process by which we keep track of what we’re saying and change it as necessary

Unknown if we edit what we say before we say it

But we engage in self-repair after an utterance is made

We interrupt ourselves when we detect an error

We issue an editing expression

We repair what we’ve said

*

Perceptual self-loop theory

Two chances at repair

Internal monitoring before utterance is articulated

External monitoring after utterance is heard and comprehended

*

Slevc and Ferreira (2006)

Investigated if internal monitoring process is more sensitive to phonological or semantic factors

Utilized a stop-signal paradigm

Presented with line drawings (i.e., target) of everyday objects (e.g., couch)

Task: name the object as quickly as possible

Presentation was followed by:

Word label for the object

A different word

This was a stop-signal and participants were to withhold their response

*

Conditions

Varied the relationship between stop-signal and the target

Phonologically related (e.g., pouch)

Semantically related (e.g., sofa)

Prediction

Stopping should be more difficult when stop-signal matches what is going on internally

If internal monitoring process is phonological

Harder to stop when target and stop-signal are phonologically related

If internal monitoring process is semantic

Harder to stop when target and stop-signal are semantically related

Results

Animation reveals the result

*

Speech Perception

  • Modular view

Language processes cannot be explained in terms of other cognitive processes

Speech is special—comprised by a unique set of abilities and capacities

  • Nonmodular view

Language processes can be explained by normal cognitive processes

*

Motor Theory of Speech Perception

  • Modular view of speech perception

Implicit knowledge about how speech sounds are articulated aids perception of those speech sounds

Two important principles:

Only humans possess the ability to produce speech and so only humans can perceive speech

Speech perception is innate—special mechanism present at birth

*

Arguments for a special speech mechanism

Categorical perception

Special mechanism to deal with coarticulation

Different from other perceptual abilities where fine discriminations are possible

The speed of processing required to perceive speech

Phonemes occur at a rate of 10-20 per second

Special mechanism is needed as other perceptual abilities cannot make such fast discriminations

*

Auditory Theory of Speech Perception

Nonmodular view of speech perception

Speech perception is just another exercise in auditory perception and pattern recognition

Other species share similar auditory abilities to humans, so should be able to perceive speech

*

Refuting the speech is special view

Unit of perception may be the syllable, not the phoneme

If so:

Coarticulation is not so problematic

Processing speed could be slower (10-15 units per second)

In line with other perceptual abilities

Categorical perception may not be so categorical

More evident with consonants than vowels

While phenomenologically true, we are capable of finer-grained discriminations

Nonhumans can perceive speech

Quail can perceive phonemic categories

*

Re-Assessment: Kind of Special

A resurgence of interest in the motor theory as it fits with:

Findings from embodied cognition

The processing of mirror neurons

React the same way when perceiving an act as when performing an act

Could be the link between speech perception and production proposed by the motor theory

*

Evidence for speech being special is mixed

Speech would be special if:

It was the only perceptual system directly tied to the motor system

It’s not

It was the only auditory stimulus whose perception required more than just the acoustic signal

It’s not

If it utilized auditory perception mechanisms dedicated exclusively to speech

Different brain areas are devoted to vocal vs. nonvocal sounds

Same brain mechanism used for speech and nonspeech sounds

*

Some think it is special enough in that its link to the motor system warrants further investigation

However this link does not necessitate the need for a “special mechanism”

Others argue that the only thing special about speech is its importance for daily living

The study of speech should be subsumed under a general auditory cognitive science framework

*