Zenzi M. Griffin

Structural Priming as Implicit Learning: A Comparison of Models of Sentence Production

Structural priming reflects a tendency to generalize recently spoken or heard syntactic structures to different utterances. We propose that it is a form of implicit learning. To explore this hypothesis, we developed and tested a connectionist model of language production that incorporated mechanisms previously used to simulate implicit learning. In the model, the mechanism that learned to produce structured sequences of phrases from messages also exhibited structural priming. The ability of the model to account for structural priming depended on representational assumptions about the nature of messages and the relationship between comprehension and production. Modeling experiments showed that comprehension-based representations were important for the models generalizations in production and that nonatomic message representations allowed a better fit to existing data on structural priming than traditional thematic-role representations.

Phonological Influences on Lexical (Mis)Selection

Speakers produce words to convey meaning, but does meaning alone determine which words they say? We report three experiments that show independent semantic and phonological influences converging to determine word selection. Speakers named pictures (e.g., of a priest) following visually presented cloze sentences that primed either semantic competitors of the target object name (“The woman went to the convent to become a …”), homophones of the competitors (“I thought that there would still be some cookies left, but there were …”), or matched unrelated control object names. Primed semantic competitors (nun) were produced instead of picture names more often than primed unrelated control object names, showing the well-documented influence of semantic similarity on lexical selection. Surprisingly, primed homophone competitors (none) also substituted for picture names more often than control object names even though they only sounded like competitors. Thus, independent semantic and phonological influences can converge to affect word selection.

A reversed word length effect in coordinating the preparation and articulation of words in speaking.

Speakers tend to prepare their nouns immediately before saying them, rather than preparing them further in advance. To test the limits of this last-second preparation, speakers were asked to name object pairs without pausing between names. There was not enough time to prepare the second name while articulating the first, so the speakers’ delay in starting to say the first name was based on the amount of time available to prepare the second name during speech. Before speaking, they spent more time preparing a second name (e.g., carrot) when the first name was monosyllabic (e.g., wig) rather than multisyllabic (e.g., windmill ). When additional words intervened between names, the length of the first name became less important and speech began earlier. Preparation differences were reflected in speech latencies, durations, and eye movements. The results suggest that speakers are sensitive to the length of prepared words and the time needed for preparing subsequent words. They can use this information to increase fluency while minimizing word buffering.

The influence of age on the time course of word preparation in multiword utterances

Sentence production requires speakers to co-ordinate the preparation of words so that they are ready for articulation when they are needed. Ageing appears to influence both the speed and likelihood of successful word retrieval. We examine how age differences in word production might influence the production of larger units of speech such as sentences. Speakers described displays containing three objects of systematically varied naming difficulty. The latency, duration, content, and fluency of speech in addition to its co-ordination with eye movements indicated that both young and older adults prepared their words immediately before uttering them. As a consequence, older adults were also significantly less fluent in their utterances than were younger adults.

Properties of Spoken Language Production

Publisher Summary Language production is logically divided into three major steps: deciding what to express (conceptualization), determining how to express it (formulation), and expressing it (articulation). Although achieving goals in conversation, structuring narratives, and modulating the ebb and flow of dialogue are inherently important to understanding how people speak, psycholinguistic studies of language production have primarily focused on the formulation of single, isolated utterances. An utterance consists of one or more words, spoken together under a single intonational contour or expressing a single idea. The simplest meaningful utterance consists of a single word. Generating a word begins with specifying its semantic and pragmatic properties—that is, a speaker decides upon an intention or some content to express (e.g., a desired outcome or an observation) and encodes the situational constraints on how the content may be expressed. The next major stage is formulation, which in turn is divided into a word selection stage and a sound processing stage. Sound processing, in contrast, involves constructing the phonological form of a selected word by retrieving its individual sounds and organizing them into stressed and unstressed syllables and then specifying the motor programs to realize those syllables. The final process is articulation—that is, the execution of motor programs to pronounce the sounds of a word.

The Eyes Are Right When the Mouth Is Wrong

When describing visual scenes, speakers typically gaze at objects while preparing their names. In a study of the relation between eye movements and speech, a corpus of self-corrected speech errors was analyzed. If errors result from rushed word preparation, insufficient visual information, or failure to check prepared names against objects, speakers should spend less time gazing at referents before uttering errors than before uttering correct names. Counter to predictions, gazes to referents before errors (e.g., gazes to an axe before saying “ham-” [hammer]) highly resembled gazes to referents before correct names (e.g., gazes to an axe before saying “axe”). However, speakers gazed at referents for more time after initiating erroneous compared with correct names, apparently while they prepared corrections. Assuming that gaze nonetheless reflects word preparation, errors were not associated with insufficient preparation. Nor were errors systematically associated with decreased inspection of objects. Like gesture, gaze may accurately reflect a speakers intentions even when the accompanying speech does not.

Speakers gaze at objects while preparing intentionally inaccurate labels for them.

When describing scenes, speakers gaze at objects while preparing their names (Z. M. Griffin & K. Bock, 2000). In this study, the authors investigated whether gazes to referents occurred in the absence of a correspondence between visual features and word meaning. Speakers gazed significantly longer at objects before intentionally labeling them inaccurately with the names of similar things (e.g., calling a horse a dog) than when labeling them accurately. This held for grammatical subjects and objects as well as agents and patients. Moreover, the time spent gazing at a referent before labeling it with a novel word or accurate name was similar and decreased as speakers gained experience using the novel word. These results suggest that visual attention in speaking may be directed toward referents in the absence of any association between their visual forms and the words used to talk about them.

Retrieving Personal Names, Referring Expressions, and Terms of Address

Abstract Why is it more difficult to recall the names of celebrities and old acquaintances than other words that one seldom uses? Several factors related to the way information about people are structured and how words are produced conspire to make personal names particularly difficult to retrieve. In contrast, expressions such as descriptive nicknames, kinship terms, and titles appear easier to retrieve. A review of how people are named, referred to, and addressed across cultures and situations suggests that there is broad range in the relative difficulty of producing terms and that several social variables must be considered in a full account of name retrieval.

Recency Effects for Meaning and Form in Word Selection

Three experiments examined the contribution of phonological availability in selecting words as predicted by interactive activation models of word production. Homophonous words such as week and weak permitted a words phonological form to be activated on priming trials without selection of its meaning or lemma. Recent production of a homophone failed to significantly increase production of its twin as a sentence completion. However, speakers were significantly more likely to complete a sentence with a recently read or generated unambiguous word. This increase in response probability was unaffected by word frequency. The results constrain the degree to which experience and phonological availability may affect word selection in spoken language production.

Effects of Age of English Exposure, Current Input/Output, and Grade on Bilingual Language Performance.

This study evaluates the effects of Age of Exposure to English (AoEE) and Current Input/Output on language performance in a cross-sectional sample of Spanish-English bilingual children. First- (N = 586) and third-graders (N = 298) who spanned a wide range of bilingual language experience participated. Parents and teachers provided information about English and Spanish language use. Short tests of semantic and morphosyntactic development in Spanish and English were used to quantify childrens knowledge of each language. There were significant interactions between AoEE and Current Input/Output for children at third grade in English and in both grades for Spanish. In English, the relationship between AoEE and language scores were linear for first- and third-graders. In Spanish a nonlinear relationship was observed. We discuss how much of the variance was accounted for by AoEE and Current Input/Output.