Adults’ Gestures During Object Naming and Preverbal Infants’ Word Mapping: Does Type of Motion Matter? Lakshmi J. Gogate & Dalit J. Matatyaho State University of New York Health Science Center at Brooklyn, NY
Prior Studies of Maternal Naming to Infants What is the Relevance of Object Motion and Synchrony to Infants in Word Learning Contexts? Prior Studies of Maternal Naming to Infants During semi-structured play episodes, when naming novel objects or actions, mothers of preverbal infants provide object motion in synchrony with the novel name more often than they provide names with static objects (Figure 1; Gogate, Bahrick & Watson, 2000; Gogate, et al., 2006).
Figure 1. The mean proportion of total target word occurrences (PTTW and SD) in mothers’ bimodal naming in the context of objects moving in temporal synchrony, asynchrony, remaining still, or being held by the infant. a b c Scheffe’s two-tailed p < .05 Gogate, Bahrick & Watson, 2000
The Relevance of Object Motion and Synchrony to Infants in Word Learning Contexts (contd.) Infant-Controlled Habituation Experiments Following habituation to two syllable–object pairs, infants look longer on switch test trials relative to control trials if the objects are moved while the syllable is spoken but not if stationary objects are presented with the syllables (Figure 2; Gogate & Bahrick, 1998).
Figure 2. Seven-month-old infants’ mean looking (and SD) to syllable-object displays as a function of trial type and condition. Syllables - /a/-/i/ Mean Looking Time (secs.) ** p < .01 Gogate & Bahrick, 1998 1 Visual recovery is a difference score between visual fixation during the switch trials versus visual fixation during the control trials.
How Do Object Motion and Synchrony Facilitate Infants’ Learning of Word- Object Mappings? Synchrony, an amodal property available across auditory-visual sensory modalities, foregrounds a novel word-object relation. In addition, the trajectory of maternal motion might bring the object into infants’ line of sight and assist them in attending to the very object the mother intended to name, likely reducing referential ambiguity.
What We Know and Don’t Know In general, we know that object motion, a unimodal property, in synchrony with an utterance facilitates learning of word-object relations by preverbal infants. We do not know if type of object motion or hand-held caregiver gesture with objects plays a role in infants’ learning of word-object relations. Specifically, we do not know if some types of object motion more than others, in synchrony with spoken words, facilitate learning of word-object relations.
Does Type of Motion Matter during Word Mapping? A Prior Field Study (Zukow-Goldring & Ferko, 1994) Under naturalistic conditions, Mexican and Caucasian-American mothers predominantly use showing gestures such as waggling or looming an object during object naming to their young infants. OUR MAIN HYPOTHESIS Within an embodied system (Smith, 2005), where there are bound to be tightly coupled interactions between infant and environment, some maternal object motions might be more salient than others for infants in word mapping contexts.
Two Studies of Preverbal Infants Study 1 - maternal use of object motion to their infants during object naming, infant attention, and word learning. Study 2 (in progress) - 8-month-old infants’ word mapping under different motion conditions in infant-controlled habituation experiments.
Method (Same data set as in Gogate, et al., 2006): Study 1: Type of Maternal Object Motion During Naming Predicts Infants’ Word-Mapping Method (Same data set as in Gogate, et al., 2006): 3-min play episode - Mothers (N = 24) taught their 6- to 8-month-old infants the names, Gow and Chi, for two target objects (Fig. 3) one at a time while other toys were present on the scene (video clip). A two-choice intermodal word-mapping test of infants’ learning of word-object pairings - Main Measure - proportion of first looks to the word matched object [PFLs across 12 trials].
Figure 3: The objects mothers used to teach infants the two word-object pairs Chi/Gow Gow/Chi
Figure 4: Video clip of a mother teaching her 7-month-old the name “Gow” for an object.
Coding of play episodes Prior Coding - Maternal use of synchronous naming (.63) vs. other bimodal naming categories (e.g., asynchrony (.07), or static object naming (.14) Prior Coding - Infants’ attention during maternal bimodal naming [e.g., looks to mother (.05) , looks to object (.59), gaze switch from mother to object (.08), or object to mother (.05) or elsewhere (.23)]. Novel Coding - Type of object motion during maternal bimodal naming (e.g., forward, backward, upward, shaking motions (Matatyaho & Gogate, under review).
Results (Table 1) Shaking and looming object motions occurred more often in synchronous maternal naming than other motions. Insignificant motions such as twitches or jerks of the arm occurred more often during asynchronous naming.
Table 1. Maternal Motion Types Table 1. Maternal Motion Types* as a Function of Bimodal Naming Category. Motion Types Naming Category Shaking Forward/ Downward “looming” Upward Backward Other Insignificant (twitch or jerk) Synchronous Raw total (M freq) M Prop. (SD) 347 (14.46) .34 (.27) 295 (12.29) .37 (.23) 26 (1.08) .03 (.04) 29 (1.21) .03 (.05) 65 (2.71) .07 (.07) 106 (4.42) .16 (.16) Asynchronous 19 (.79) .16 (.20) 12 (.50) .12 (.16) 3 (.13) .08 (.22) 2 (.08) .04 (.09) 9 (.38) .13 (.18) 35 (1.46) .47 (.36) *Out of 986 occurrences of maternal naming that were coded as synchronous or asynchronous, 38 (4%) were not coded for motion type due to disagreement between motion versus bimodal naming category coders.
Results (Contd.) Analyses of covariance of infants’ proportion of first looks (PFL) to the word matched object - Covariate - the proportion of each maternal motion type in the presence of synchronous or asynch. naming Independent variable - the proportion of infants’ gaze-switches from mother to object or other attention types. Present Finding: Maternal shaking motions during synchronous naming co-varied with infants’ word mapping ability (Figure 5, p = .037). Prior Finding: Infants’ eye-gaze switches from mother-to-object influenced their mapping of the words onto objects (p = .01).
High Gaze Switchers (Mother to Object) Figure 5. Maternal Shaking Motions during Synchronous Naming and infants’ Proportion of First looks to the Word Matched Object (PFL) as a Function of Gaze-Switching Ability High Gaze Switchers (Mother to Object) Low Gaze Switchers (Mother to Object) Proportion of first looks Individual Dyads
Study 2: Do 8-month-olds Map Words in Some Motion Conditions Better than in Others? Method: Participants: 2 groups of 8-month-olds (n = 12 in each) in one of two motion conditions – shaking and sideways (Figures 6, 7a and 7b). Infant-Controlled Habituation Procedure: Habituation Phase - 2 Synchronous Word-Object pairs, e.g., wem -lamb chop and baf - squiggly until habituation criterion. Test Phase - Following two post hab. trials - Switch trials e.g., wem - squiggly and baf - lamb chop Control trials – e.g., wem - lamb chop and baf - squiggly
Figure 6: The lamb chop, the squiggly toy (above), Figure 6: The lamb chop, the squiggly toy (above), the fish and the dragonfly (below).
Figure 7A- A word-object pairing where the object is moved using a shaking motion
Figure 7B- A word-object pairing where the object is moved sideways
Study 2: Results – Habituation Phase During infant controlled habituation, infants showed greater attention to word-object pairings right from the start on baseline trials in the shaking motion condition relative to the sideways motion condition.
Table 2. 8-month-old infants' looking time (sd) during habituation to two word-object pairs under the shaking and sideways motion conditions. Habituation Trial Type Shaking Motion (N = 12) Sideways Baseline (mean of first two habituation trials) 49.13a (10.70) 34.92b (19.24) Mean looking on the last two habituation trials 11.0 (4.74) 10.74 (5.87) Mean seconds to reach habituation criterion 268.38 (123.62) 164.18 (91.11) a b Independent samples t-test p < .01
Study 2: Results - Test Phase Infants learned the word-object relations in the shaking motion condition but not in the sideways motion condition. Visual recovery to switch trials relative to control trials was greater in the shaking motion condition than in the sideways motion condition (Figure 8).
Figure 8. Eight-month-old infants’ mean looking (SD) to word-object displays as a function of trial type and motion condition. Words - wem / baf 37.81 (18.11) Mean Looking Time (secs.) *p = .026 1 Visual recovery is a difference score between visual fixation during the switch trials versus visual fixation during the control trials.
Recap. of Findings Study 1 : Mothers of 6 to 8-month-old infants predominantly use shaking or looming motions in synchrony with objects when teaching their infants the names for objects. The infants, in turn, attend to shaking object motions during maternal synchronous naming and learn novel word mappings by switching gaze from the mother to an object. Study 2: 8-month-olds learn two word-object relations when shaking object motions are presented in synchrony with spoken words, but not when sideways motions are presented in synchrony with the words.
Conclusion THUS, TYPE OF MATERNAL GESTURE MATTERS DURING INFANT WORD LEARNING!!! Shaking but not sideways motions of objects likely add a dynamic property to the objects that facilitates preverbal infants’ attention to objects in word learning contexts.
Acknowledgements The Thrasher Research Fund (#02819-1), Dean’s Research Initiative Award, College of Medicine, SUNY. Staff: Zachary Mason, Christine Coward, Nory Jun Cabanilla, Samantha Berkule, Divya Awal. The mothers and infants who participated. Special thanks to Christopher Prince for his insights on object motion and word learning. Thank You!