Assessing event perception in adults and prelinguistic children: A prelude to syntactic bootstrapping Angela Xiaoxue He, Alexis Wellwood, Jeffrey Lidz & Alexander Williams Department of Linguistics, University of Maryland, College Park Event Representation in Verb Learning Problem: one event, many mutually entailed concepts Participant-Argument-Match (PAM) PAM Sentence: N-Arguments “Anne is swimming.” Presupposition of PAM: Swimming is viewed under SWIM but not SWIMP, SWIMT, etc. SWIMP: swimmer, path Event Concept: N-Participants SWIMT: swimmer, time “Anne is swimming.” SWIM: swimmer FLAIL: flailer MOVE: mover BREATHE: breather Question: Is this presupposition true? If not, PAM won’t work. SWIMM: swimmer, means of propulsion SWIM(e) ⟺#SWIMP(e)#⟺#SWIMT(e)#⟺#SWIMM(e)# Identifying Participants Question For a scene that can be described by 2-argument sentences, and is perhaps viewed under 3-participant concepts: 1) Do adults view the scene under a 3-participant concept? 2) Do infants view the scene under a 3-participant concept? Sentence: 2-Arguments PAM?? Event Concept: 3-Participants “Mike stole a purse.” STEAL: thief, loot, victim “Mike jimmied the door.” JIMMY: agent, patient, lever Experiment 1 & 2 giving g2# Contrast A g1: give [+teddybear] g2: give [-teddybear] h1: hug [+teddybear] h2: hug [-teddybear] h2# h1# Experiment 1: Adults N=12# ns# p=0.07# We compare such contrasts with mere perceptual differences between scenes. Adults: Similarity rating (1-7), explicit judgment & RT measures Children: Habituation/Switch (Gordon 2003), dishabituation measure Experiment 3 & 4 hugging g1# Do people care about the involvement of a third entity in a scene? If yes: Consistent with participant-hood for that entity If no: Inconsistent with participant-hood for that entity Experiment 3: Adults Experiment 2: Infants N=12# p=0.04# ns# o1: open [-instrument, from right] o2: open [+instrument, from right] o3: open [-instrument, from left] o3# o1# N=12# N=49# p=0.06# Experiment 4: Infants N=12# Preliminary#Results# N=14# t1# Figure#3:#Look#Time# Implications: Adults: “giving” = 3 participants “hugging” = 2 participants Infants: “giving” = 3 participants “hugging” = 2 participants Habit GIVE HUG g1 h1 Test Same Switch g1 g2 h1 h2 Figure#4:#Ra.ng# t3# N=12# p=0.04# Figure#6:#Look#Time# Habit Adults: “open-w/o-instrument” = 2 participants “open-w-instrument” = 3 participants “open-from-left” = 2 participants “open-from-right” = 2 participants Scenes described by N-argument sentences are not always viewed under N-participant concepts. In such cases, PAM won’t work as a learning heuristic, predicting difficulty in acquisition of these verbs. Future direction: Is there such a difficulty? Figure#5:#Reac.on#Time# Implications: Implications t1# Experiment 5: Adults p=0.02# Figure#2:#Reac.on#Time# t2# t1: take [+owner, from left] t2: take [-owner, from left] t3: take [+owner, from right] ns# Figure#1:#Ra.ng# Contrast B Contrast A Contrast B o2# o1# Experiment 5 o1 Test Same o3 Switch o2 Selected References [1] Gleitman, L. (1990). The structural sources of verb meanings. Language acquisition, 1(1), 3-55. [2] Fisher, C., Gertner, Y., Scott, R. M., & Yuan, S. (2010). Syntactic bootstrapping. Wiley Interdisciplinary Reviews: Cognitive Science, 1(2), 143-149. Gordon, P. (2003). [3] The origin of argument structure in infant event representations. In A. Brugos, L. Micciulla & C. E. Smith (Eds.), Proceedings of the 28th annual Boston University Conference on Language Development. Boston, MA: Cascadilla Press. Implications: Adults: “take-from-owner” = 3 participants “take-from-table” = 2 participants “take-from-left” = 2 participants “take-from-right” = 2 participants Thanks Xuan Wang, Dept, of Philosophy, Univ. of Maryland Rachel Dudley, Dept. of Linguistics, Univ. of Maryland Project on Children’s Language Learning UMD Language Science fellowships to He & Wellwood N=12#
© Copyright 2024 ExpyDoc
