490 likes | 569 Views
Last week. Characteristics of language production: Coordinating multiple sources of information in real time Methodological issues. This week. Overview of the production architecture Lexical representation: Tip of the tongue states. Levelt’s (1989) model. Formulation.
E N D
Last week • Characteristics of language production: • Coordinating multiple sources of information in real time • Methodological issues
This week • Overview of the production architecture • Lexical representation: • Tip of the tongue states
Formulation • Linguistic encoding of conceptual message • Speaker must choose appropriate lexical items to convey intended message • Lexical representations incorporate semantic, syntactic, morphological and phonological content • How are these retrieved?
The “tip-of-the-tongue” experience • “an instrument used for measuring angular distances, used especially in navigation to observe the altitude of celestial bodies” • The tip-of-the-tongue experience occurs when people are sure a word is in memory but they are unable to access it
“The state of our consciousness is peculiar. There is a gap therein; but no mere gap. It is a gap that is intensely active. A sort of wraith of the name is in it, beckoning us in a given direction, making us at moments tingle with the sense of our closeness and then letting us sink back without the longed-for term. If wrong names are proposed to us, this singularly definite gap acts immediately so as to negate them. They do not fit into its mould. And the gap of one word does not feel like the gap of another, all empty of content as both might seem necessarily to be when described as gaps.” (William James, 1890, pp. 251-252)
How to study • Naturalistic studies (e.g., using a diary) • suggest fairly frequent (a few times a week) • other instances may be forgotten • Experimental procedures • usually, recall to a definition • Though also sometimes pictures • state “things you know” about the word • E.g., first sound, number of syllables • often test if the word is later recalled • sometimes deliberately induce TOT states
Brown & McNeill (1966) • Studied 56 American undergraduates. • 49 low-frequency words (e.g., apse, nepotism, sampan), prompted by brief definitions. • On c. 8.5% of trials, tip-of-the-tongue state ensued: • Had to guess word's first or last letters, the number of syllables it contained, and which syllable was stressed • Total of 360 TOT states: • 233 ="positive TOTs" (subject was thinking of target word, and produced scorable data • 127 = "negative TOTs" (subject was thinking of other word, but could not recall it) • 224 similar-sound TOTs (e.g., Saipan for sampan) • 48% had the same number of syllables as the target • 95 similar-meaning TOTs (e.g., houseboat for sampan).Of the SS items • 20% had same number of syllables as target.
Some findings • Experimental procedures often generate TOTs on about 10-15% of trials • Seems to occur for most people • Seems to occur for most items tested • probably less with very frequent or very infrequent words
What information is recalled? • “The rhythm of the lost word may be there without the sound to clothe it; or the evanescent sense of something which is the initial vowel or consonant may mock us fitfully, without growing more distinct.” (James, 1890, p. 251)
Partial access? • Similar words come to mind about half the time • but how much is just guessing? • First letter: correct 50-71% of time (vs. 10% by chance) • first sound: 36% of time (vs. 6% by chance) • Additional letters: less but above chance • more for last than middle letters • hence, a U-shaped curve • so first letter is privileged (a “pointer”?)
Other aspects • Knowledge of number of syllables seems above chance • Knowledge of other aspects of the word they are looking for
Patients with anomia • Patients show more extreme patterns of word-finding difficulty • e.g. EST (Kay & Ellis, 1987) • may be like an exaggeration of normal TOT states
Is the experience universal? • Schwartz (1999) found that 45 out of 51 languages use the “tongue” metaphor, suggesting that it probably is universal • not in Icelandic, Amharic, Indonesian, Kiswahili, Kalenjin • no “tip of the finger” expression in American Sign Language
Translations of expressions • “on the tongue” • “on the tip/point/head of the tongue” • most common • “on the top of the tongue” • “on the front of the tongue” • “sparkling at the end of the tongue” • “in the mouth and throat”
Word production & TOTs • TOTs and Anomias suggest a basic split between semantics/syntax and phonology: • people can access meaning and grammar but not pronunciation
What might be known in a TOT? • Semantics • Syntax • grammatical category (“part of speech”) • e.g. noun, verb, adjective • Gender • e.g. le chien, la vache; le camion, la voiture • Number • e.g. dog vs. dogs; trousers vs. shirt • Count/mass status • e.g. oats vs. flour
Vigliocco et al. (1997) • Subjects presented with word definitions • Gender was always arbitrary • If unable to retrieve word, they answered • How well do you think you know the word? • Guess the gender • Guess the number of syllables • Guess as many letters and positions as possible • Report any word that comes to mind • Then presented with target word • Do you know this word? • Is this the word you were thinking of?
Vigliocco et al (1997) • + TOT • Both reported some correct information in questionnaire • And said yes to recognition question • - TOT • Otherwise • Overall, 9% + TOT; 19% - TOT
Vigliocco et al (1997) • + TOT: 84% correct gender guess • - TOT: 53% correct gender guess • chance level • Conclusion • Subjects often know grammatical gender information even when they have no phonological information • Supports split between syntax and phonology in production
Vigliocco, Vinson, Martin & Garrett (1999) • English: Count vs mass nouns (e.g., egg vs sugar) • Definitions to elicit ToT: • Final part that typically serves to round out or complete the design of a literary work; conclusion (EPILOGUE) • White, fatty cartilage found in meat (GRISTLE) • Asked whether participants could tell them which was appropriate: • There is_/There is a _ • There won’t be much_/There won’t be many_ • There is some_/There are a few_ • Better at this when in +TOT than –TOT state
Separate components in language production • Language production appears effortless and therefore “one stage” • TOT data suggest this is wrong – production can be difficult • It seems to break down “in the middle” • Semantic processing can be OK • “I know what I’m trying to say” • But phonological processing can be impaired • “but I just can’t think of the word” • Vigliocco and colleagues’ studies suggest syntax can be accessed without phonology
Other evidence • Much other evidence for the same conclusion • TOT data showing other syntactic knowledge (Vigliocco et al., 1999, and Caramazza and Miozzo, 1997, on grammatical gender) • Other TOT data (e.g., priming data - Rastle & Burke, 1996) • other aphasia data (Badecker et al., 1995; Henaff Gonon et al., 1989) • slips of the tongue, experiments on time-course of retrieval etc. (see later)
Separable components • Hence, a “barrier” between phonological and semantic processing • Evidence emerging for a “barrier” between phonological and syntactic processing
Speech error data • Substitution errors: • All I want is something for my shoulders [intended: elbows] • I’ve got whipped cream on my mushroom[intended: moustache] • Exchange errors: • I left the briefcase in my cigar • Do you reel feally bad?
Speech errors • Two types of errors (semantic and phonological) • errors at different stages in the production process? • Semantic exchange errors • exchanges tend to be same grammatical class • exchanges tend to occur between phrases • don’t need to occur in similar phonological environments
Speech errors • Phonological exchange errors • need not be words • tend to be adjacent words, or at least within phrase • tend to be different grammatical classes • Hence, errors support the notion of two production stages, in accord with TOT data.
Lemma and Word-Form • Word form (Lexeme): phonologically specified representation • Lemma: representation specifying syntactic features • in some accounts, also specifies semantic information • but the recent evidence supports a purely syntactic lemma
Lemmas and word-forms • Normally, accessing a word involves lemma access followed rapidly by word-form access • But sometimes only the lemma is accessed • during TOT state • regularly in anomic patients
Lemmas vs word-forms • In models of sentence production, syntactic processing occurs before phonological processing • syntactic processing uses lemma • phonological processing uses word-form
Levelt et al (1999) model • Lemma may consist of a “node” containing information about the “base form” of a word (e.g. give) and “pointers” to syntactic features, e.g.: • grammatical number (singular or plural, normally) • grammatical gender (masculine, feminine, neuter – depending on language) • category (noun, verb, etc.) • subcategory (transitive verb, intransitive verb, etc.) • count or mass noun • tense of verb (present, past, etc.)
Two accounts of TOTs • Blocking – target is suppressed by a stronger competitor • rememberers retrieve “blockers” or “interlopers” related to the correct target • these interlopers inhibit the correct target • but rememberers recognise the interlopers as incorrect as well • Partial activation – target is only weakly represented
Partial activation • Meyer and Bock (1992) provide indirect evidence for partial activation (contra earlier findings): • Participants produced responses to definitions • Definitions accompanied by cues • related in sound to target (phonological cue) • related in meaning to target (semantic cue) • we’ll ignore • unrelated to target (unrelated cue)
Meyer & Bock (1992) • Blocking hypothesis • phonological cue should interfere with target access • i.e., decrease correct responses relative to unrelated cue • Partial activation hypothesis • phonological cue should assist target access • i.e., increase correct responses relative to unrelated cue
Meyer and Bock (1992) • Critically, more correct responses following phonological cue than unrelated cue • phonological information provides activation in lexical selection • and TOT states due to partial activation rather than blocking • note, hard to interpret the TOT states themselves
Meyer & Bock (1992) • no differences between conditions before cue (as expected) • But differences after cue • Again, more correct responses following phonological cue than unrelated cue • supports partial activation account again • facilitation works with a late presentation of the cue
Partial activation • See also Harley and Brown (1998) • TOTs are more common for words with few phonological neighbours (i.e. few words that differ by only one phoneme from the target) • Hence suggests that neighbours facilitate rather than inhibit target retrieval
Summary • 3 stages to production: • Conceptualisation • Formulation • Articulation • Formulation involves lexical retrieval: • Semantic/syntactic content (lemma) • Phonological content (word-form) • Lemma can be retrieved without word-form necessarily being retrieved • When this occurs, ToT state ensues
Phenomenology and metacognition • Doctrine of concordance: very high correlation between cognitive processes, behaviour and phenomenological experience
Tulving (1989) criticised this doctrine – for instance, retrieval is not the same as the experience of recollection • e.g., words are often retrieved during speech without any experience of recollection • “mental time travel” only occurs in certain situations involving episodic memory • other non-retrieval factors affect our judgement of “pastness”
For instance, metacognition literature demonstrates clear differences between “feeling of knowing” and actual knowledge • Two separate components: • one carries out the process (“object-level”) • the other monitors the process (“meta-level”) • TOT states could reflect meta-level processes
Cues and cue familiarity • TOTs are not based on sensitivity to inaccessible targets • Instead, rememberers infer the target’s existence from clues • the cue (e.g., the definition) • retrieved partial information • any other generated information
Thus, Reder (1987) and Reder and Ritter (1992) • rapid judgements of feeling of knowing (e.g., What is the capital of Peru?) depend on familiarity with the terms in the question • these judgements are improved by priming the terms in the questions, not by priming the answers
Koriat and Lieblich (1977) – more repetitive definitions led to more TOTs • suggests that cue factors play a role in TOT states • Schwartz – TOTs are a “lens” to bring phenomenology into focus, not a “window” on the nature of retrieval • a note of caution • but lots of parallels between TOT data and other tests of lexical retrieval mechanisms
And more recently still, Harley and Bown (1998) varied the frequency and phonological distinctiveness of the target words and found "that TOTs are more likely to arise on low-frequency words that have few close phonological neighbours" (p151). They use their data to reflect upon the broader process of "lexicalisation"[glossary], which they define as "the process of phonological retrieval in speech production given a semantic input" (p152), and they opt for a "two-stage" explanatory model of lexical access, that is to say, a model which strictly separates each word's semantic and phonological representations. TOTs can therefore be seen as arising "when the first stage of lexical access is completed successfully, but not the second" (pp152-153). However, the critical point as far as Harley and Bown are concerned is as follows ..... • "Our central result is that phonological neighbours contribute to, rather than hinder, phonological retrieval in speech production. [.....] A TOT occurs when the semantic specification successfully accesses the abstract lemma [glossary]. This causes the 'feeling of knowing' the word. Nevertheless, the lemma is then unable to pass sufficient activation onto and thereby access the corresponding phonological word form. [.....] There are two possible reasons for failure at this stage. Either the connections between the lemma and the phonological forms might be weakened, or the phonological forms might themselves be weakly represented for these items." (Harley and Bown, 1998, p162)