1 / 1

What “Mice Trap” tells us about the mental lexicon Carolyn J. Buck-Gengler 1,3 , Lise Menn 2,3 , and Alice F. Healy 1,3

What “Mice Trap” tells us about the mental lexicon Carolyn J. Buck-Gengler 1,3 , Lise Menn 2,3 , and Alice F. Healy 1,3 University of Colorado at Boulder 1 Department of Psychology 2 Department of Linguistics 3 Institute of Cognitive Science. Abstract

roshaun
Download Presentation

What “Mice Trap” tells us about the mental lexicon Carolyn J. Buck-Gengler 1,3 , Lise Menn 2,3 , and Alice F. Healy 1,3

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. What “Mice Trap” tells us about the mental lexicon Carolyn J. Buck-Gengler1,3, Lise Menn2,3, and Alice F. Healy1,3 University of Colorado at Boulder 1Department of Psychology 2Department of Linguistics 3Institute of Cognitive Science Abstract Two experiments focused on the effect of noun regularity on the task of producing a singular form from a plural or the reverse. Experiment 1 used English; Experiment 2 used a miniature artificial language in which word frequency and form were controlled. Subjects learned the words in the artificial language before proceeding to the test phase of Experiment 2. In the test phase of both experiments, subjects were given one of the two forms (singular or plural) and required to produce either the same form or the opposite form. In half the trials the grammatical number of the picture and the required response matched, and in half the number did not match. In Experiment 1, response time per letter was faster when the number of cue and required response matched than when they differed, especially for the irregulars. In Experiment 2, there were also significant effects of matching, but these effects were not influenced by either word frequency or regularity. From these results we conclude that the preference for irregular plural as first element of noun-noun compounds can be explained by processing factors which hold for both children and adults in both natural and artificial languages. No appeal to innate grammar is required to explain the similarity between this aspect of child and adult linguistic behavior. Background Regular and irregular nouns behave differently Example: noun-noun compounding in English regularirregular singular rat catcher mouse catcher plural *rats catcher mice catcher General rule: The first noun must be singular (e.g., rat catcher, toy box, but not *rats catcher, *toys box, even when talking about multiple rats or toys), but the plural of irregular nouns is also acceptable. Pinker (e.g., 1994: 146-7) argues that the rule is likely to be innate because children obey it (Gordon, 1985), even without being exposed to adult examples of irregular plurals as first elements of compounds. Gordon (1985): When prompted with phrases like“What do you call someone who eats <X>?”, where X was either a regular or irregular plural noun, children frequently used plural when primed with an irregular plural, but almost never used plural when primed with a regular plural. Findings were explained in terms of an innate grammar, specifically Level Ordering (e.g., Kiparsky, 1982). Alternate explanation: Processing difficulty Buck-Gengler, Menn, and Healy (2001): Elicited noun-noun compounds (similar to Gordon, 1985) from adults Subjects saw fill-in-the-blank sentence, responded with compound a TUB holding X is a _____ X was a singular or plural regular or irregular English noun Example: duck/ducks/goose/geese Subjects produced far more plurals for the first noun of the compound when the cue was an irregular plural (geese) than when it was a regular plural (ducks) or either type of singular (goose/duck). When subjects did produce singular first forms (goose tub/duck tub) RTs were significantly longer for irregular plural cues than for the other three kinds of cue. Preference for producing irregular plurals as first elements of compounds in such elicitation tasks can therefore be explained by processing difficulty: Goose is harder to access from geese than duck is from ducks (as might be predicted by, e.g., Allen & Badecker, 2002, or Levelt, Roelofs, & Meyer, 1999). • The Current Experiments • Main Task: Given either a singular or a plural of a noun, produce either same or opposite form • Experiment 1: English – similar set of words as in Buck-Gengler et al. (2001) • Stimuli: Pictures representing a matched set of imageable English nouns • IrregularRegular • child children car cars • foot feet fork forks • goose geese gun guns • louse lice letter letters • man men match matches • mouse mice moon moons • ox oxen owl owls • tooth teeth tree trees • Matched on frequency, first letter, length • Experiment 2: Miniature artificial language • Why an artificial language? To control for factors in real languages (e.g., word frequency, type of irregularity, aspects of form including onset, length, etc., and other idiosyncrasies) that could be contributing to the response. • Description of artificial language: • Structure -- controlled for form and frequency • • Form • singular plural • CVCV C1 ≠ C2, V1 ≠ V2 Plural method • C = {k,g,f,v,m,n,s,z,t,d,p,b} Noun type Prefix Infix • V = {a,i,o,u} Regular e-CVCV C-e-VCV • Irregular C-#-VCV #-CVCV • # is a nonpredictable vowel (not e) • • Frequency • 12 low frequency regular 4 low frequency irregular • 4 high frequency regular 4 high frequency irregular • Stimuli: Pictures representing the English concept for each artificial language word Method Both experiments: 2 phases Phase 1 Exp. 1: familiarization with pictures see picture, type word; go through list twice Exp. 2: Learn miniature artificial language words Phase 2 Main task Participants and Apparatus: • 16 (E1), 24 (E2) native speakers of English • Presented on computer • Responses and RTs recorded by computer Main task: In both experiments, participants saw a picture and a fill-in-the-blank phrase which was either the number “1” or the number “4” followed by a blank (in the English experiment the numbers were spelled out). Example fill-in-the-blank phrases: “one _____” or “four ____” (Exp. 1) “1_____” or “4 ______” (Exp. 2) Participants typed the appropriate form of the English or artificial language word representing the picture to go with the number in the fill in the blank phrase. For example: If they saw a picture of four trees and “one ____” (Exp. 1) or “1 ____” (Exp. 2) they responded by typing “tree” (Exp. 1) or “bidu” (Exp. 2) Design: Between subjects factor: (Exp. 2) Plural method of regular (prefix; infix) Within subjects factors: Noun type (regular, irregular) Grammatical number of response (singular, plural) Match (Response grammatical number match/not match cue picture) Frequency (high, low) (Exp. 2 only) Measure: Time/letter (Total Response Time/# letters in the word) Scoring: Only correct typed trials (no backspacing/correction) included Results Exp. 1 (English): Main effects for Response Time/Letter: • Regulars faster than irregulars • Additional time is required when the picture number and response number don’t match Interactions: Exp. 2 (Miniature Artificial Language): Main effects for Response Time/Letter: • Regulars faster than irregulars • Additional time is required when the picture number and response number don’t match Conclusion From these and previous results we conclude that the frequently reported preference for irregular plural as first element of noun-noun compounds can be explained by processing factors (i.e., the accessibility of the singular from the plural) that hold for both children and adults, and in both natural and artificial languages. No appeal to innate grammar is required to explain the similarity between child and adult performance on this aspect of linguistic behavior. More generally, the closer two forms of the same word (lemma) are to each other, the faster one will be accessed from the other. Unlike English, the artificial language ensured that the difference in form between singular and plural was equivalent for both regulars and irregulars. With this control, we’ve teased apart the issues of regularity and similarity, and have shown them to have separate, non-interacting effects. In addition, infixation and affixation appear to be equivalent when transparency is controlled. Letters in irregular singulars are typed faster than those in irregular plurals, but letters in regular plurals are typed faster than those in regular singulars. Shows the two main effects; moreover, the interaction approaches significance: Extra response time is needed when cue and response numbers don’t match for irregular nouns. Shows the two main effects (interaction is not significant). __________________________________ References Allen, M. & Badecker, W. (2002). Inflectional regularity: Probing the nature of lexical representation in a cross-modal priming task. Journal of Memory and Language, 46, 705–722. Buck-Gengler, C. J., Menn, L., & Healy, A. F. (2001). Mice Trap: A new explanation for irregular plurals in noun-noun compounds. Proceedings of the Twenty-Third Annual Conference of the Cognitive Science Society (pp. 140-145). Mahwah, NJ: Erlbaum. Gordon, P. (1985). Level-ordering in lexical development. Cognition, 21(2), 73-93. Kiparsky, P. (1982). From cyclic phonology to lexical phonology. In H. v. d. Hulst & N. Smith (Eds.), The Structure of Phonological Representations (pp. 131-175). Dordrecht, The Netherlands: Foris Publications. Levelt, W. J. M., Roelofs A., & Meyer, A. S. (1999). A theory of lexical access in speech production. Behavioral and Brain Sciences, 22, 1–75. Pinker, S. (1994). The Language Instinct. New York: W. Morrow and Co. Learning the Miniature Artificial Language • Subjects saw a picture and associated word, and typed the word • 12 words (mixture of high and low frequency, singular and plural, regular and irregular) formed a subset • After each subset, quiz to reinforce learning, with feedback • 8 subsets comprised one round; within the 8 subsets each word was seen either once (low frequency) or 4 times (high frequency) • After one round of learning and quizzes, test of all words, with feedback • When criterion (90%) reached on test, proceeded to main task • If criterion not reached, repeated learn/quiz, with different order and grouping of words FAll Ss learned to criterion within 7 rounds; over half learned to criterion in 2 or 3 rounds. __________________________________ Acknowledgments This research was supported in part by Army Research Institute Contract DASW01-99-K-0022 and Army Research Office Grant DAAG55-98-1-0214to the University of Colorado (Alice Healy, Principal Investigator; Lyle Bourne, Co-Principal Investigator). This research was also supported in part by a Student Research Award from the Institute of Cognitive Science to the first author.

More Related