Chapter 4 Flashcards
The aim of language production is
To describe the linguistic operations (functional arquitecture) applied to a given prelinguistic input to the generation of a motor output
Questions of interest:
How many/which are the processors
How dos they work (serial vs. parallel)
How they connect to each other (autonomous vs. interactive)
Speech error: Anticipation
= occur when a later segment takes the place of an earlier one. They differ from shifts in that the segment that intrudes on another also remains in its correct location and thus is used twice
Back my bike (Take my bike)
It’s a meal mystery (It’s a real mystery)
Speech error: Persevation
= occur when an earlier segment replaces a later item.
He pulled a pantrum (tanrum)
Speech error: Shift
= one speech segment disappears from its appropriate location and
appears somewhere else
That’s so she’ll be ready in case she decide to hits it (decides to hit it).
Speech error: Exchange
= double shifts, in which two linguistic units exchange places
Fancy getting your model renosed (getting your nose remodeled).
Speech error: Addition
= add linguistic material
I didn’t explain this clarefully enough (carefully enough).
Speech error: Deletion
= leave something out
I’ll just get up and mutter intelligibly (unintelligibly)
Speech error: Substitution
= occur when one segment is replaced by an intruder. These differ from
previously described slips in that the source of the intrusion may not be in the
sentence
At low speeds it’s too light (heavy).
Speech error: Blend
= apparently occur when more than one word is being considered and the two intended items ‘‘fuse’’ or ‘‘blend’’ into a single item.
That child is looking to be spaddled (spanked/paddled)
Syntactic priming phenomenon
Subjects trend to use preactivated syntactic structures
Tip of the Tongue phenomenon TOT (Brown y McNeill, 1966)
The “tip of the tongue” (TOT) phenomenon is a state in which one cannot quite recall a familiar word but can recall words of similar form and meaning.
Utterance generator Fromkin (1971)
It derives from the analysis of the errors produced by speakers in natural contexts
Each type of error refers us to a type of mental representation and each of them refers to a processing phase
Autonomous generator, serial descending with no feedback loops
6 phases in which different representations of the message take place
Utterance generator Fromkin: 6 fasen
Fase 1 = Meaning to be conveyed
Syntactic structure genetaror + semantic structure genetator
Fase 2 =Syntactic-semantic structures
Intonation controur generator
Fase 3 = Structures with primary stress and intonation specified
Fase 4 = Words are selected from the lexicon
Fase 4 = Strings of segments devided in syllables - Syntactic phonological F’s speficied
Morphophonemic rules
Fase 5 = Strings of phonetic segment
Phonetic rules
Fase 6 = Fully specified phonetic segments in syllables
Motor commands to muscles
Spreading-activation theory of retrieval in sentence production of Dell (1986)
Connectionist model based on the concept of spreading activation
Derived from the analysis of errors
Functional architecture: representations and tactical frameworks at 4 levels (semantic, syntactic, morphological and phonological)
Stages of language production in the model of Ellis & Young
Autonomic serial model from neuropsychological dissociations
Processing units for oral and written production
Widely used in neuropsychological clinic and speech therapy (Cuetos 1998)
Limitations:
- Lack of description of the mechanisms of the modules
- Lack of syntax specification
Theory of lexical access in speech production of Levelt, Roelofs, & Meyer (1999)
1: Lexical selection: Conceptualization
Producing a word begins by activating a lexical concept and selecting it for expression (determining what to say). Speakers contruct a preverbal message through conceiving an intention and selecting relevant information from memory or the environment in order to prepare for the contruction of an utterance.
Involve: macroplanning + microplanning
Image naming and categorization tasks: They involve recognizing the object and selecting the appropriate concept
- Word production starts activating an appropriate lexical concept and selecting it for production
- Naming and categorization tasks:
Involve object recognition and selection of the appropriate concept
N2 in Go/No Go categorization task: Animal (No Go) vs. Non-animal (Go)
At 150 ms after presenting visual stimulus subjects have enough information to initiate response inhibition
Access to the lexical concept operates at this temporal window
Accessing to the lexical concept of an object takes place been 150-200 ms
Theory of lexical access in speech production of Levelt, Roelofs, & Meyer (1999)
2: Lexical selection: Lemma activation
It consists of accessing the syntactic characteristics of the target word (syntactic word)
Lexicalization + syntactic planning
Each node of the “conceptual stratum” is linked to a single node of the “stratum of the lemmas”
- To access to the syntactic features of a target word (syntactic word)
- Each node in the conceptual stratum is linked to a single node of the lemma stratum
- At which time do we get access to the lemmas?
- N2 effect in the Conceptual task (peak latency 477) started earlier than in the Syntax task (peak latency 550)
- The 73 ms of difference will reflect mean time taken to activate gender feature of the lemmas
Lemma selection takes 73 ms and happens between150-350 ms
Theory of lexical access in speech production of Levelt, Roelofs, & Meyer (1999)
3: Morphological encoding:
The operations involved in coding the form begin by accessing the phonological form and end with the articulation of the word.
Three stages involve turning words into sounds:
Morphological codes retrieval: accessing the phonological form of words
Phonological encoding: syllabification or spelling of morphemes in ordered segments
Phonetic encoding: transformation of syllables into motor instructions
Form encoding starts getting access to the phonological shape and ends with the articulation of words
Morphological encoding: Recover of morphological codes
Hyp: If syntax precedes phology during production, syntactic information (prepared hand) will be sent to the motor system before phological information (taking the Go/No Go decision), thus generating some transient preparation (LRP) in No Go trials
- At 370 ms go and no go trials difered from baseline
- At 410 ms the accessibility to phonological information inhibits motor preparation in no go trials
- 40 ms after getting access to syntactic information, phonological information is activated thus inhibiting motor preparation
Morphological encoding last between 40-113 ms after lemma selection
Morphological encoding: Syllabification
Hyp: Target phoneme variations will not change preparation start (LRP), but preparation inhibition (the time point where go and no go responses diverge due to the access of phonological information)
370-400 ms: go and no-go trials from both conditions diverge from baseline
Preparatory responses are inhibit earlier in start phoneme trials
Phonological encoding operates in a left to right manner at a rate of 25 ms per segment
Syllabification: 25 ms/segment
Dialog
complex and dynamic interaction between speakers directed by rules or conventions determined by the social an cultural environment.
The interactive alignment account of dialogue (Pickering y Garrod, 2004):
Four basic cognitive mechanisms to achieved alignment:
- Priming
- Inferences
- Production of Routines
- Monitoring mechanisms.
Cognitive mechanisms during conversation
Priming mechanisms:
People do not explicitly negotiate expressions and meanings that use along a conversation.
Instead it seems that there exist automatic mechanisms (priming) that activates common representations as conversation takes place
The advance in the interaction between speakers will progressively align representations increasing the probability of repetition of the lexical, syntactical structures, intonation…
This will happen in an interactive manner at the different representational levels (semantic, pragmatic, etc.) until alignment is achieved.
Cognitive mechanisms during conversation
2.- Inferences:
Inferences are performed by participants regarding the mutual knowledge or “background of implicit knowledge” (believes, attitudes…)
Interactive restoration mechanisms: top-down mechanisms when alignment is no sufficient.
Consist in two recursive mechanisms:
a. - To check whether a given information can be directly interpreted according to my own representations
b. - To reformulate the expression in order to allow the establishment of a “background of implicit knowledge” (in response to the failure of the first mechanism)
Cognitive mechanisms during conversation
3.- Routine expressions:
Contribute to the fluency of dialoged and involves automatic activation of routines or frequently used expressions
For example:” I know what you said…”
Cognitive mechanisms during conversation
4.- Monitoring:
It refers to alignment processes operating within a subject = auto-alignment between different levels of representation
