ED 371 300 CS 011 729 AUTHOR Pellegrini, A. D.; And Others TITLE · 2014-05-07 · ED 371 300. AUTHOR TITLE. INSTITUTION. SPONS AGENCY. PUB DATE CONTRACT NOTE PUB TYPE. EDRS PRICE

ED 371 300

AUTHORTITLE

INSTITUTION

SPONS AGENCY

PUB DATECONTRACTNOTEPUB TYPE

EDRS PRICEDESCRIPTORS

IDENTIFIERS

ABSTRACT

DOCUMENT RESUME

CS 011 729

Pellegrini, A. D.; And OthersJoint Reading between Mothers and Their Head StartChildren. Vocabulary Development in Two Text Formats.Reading Research Report No. 13.National Reading Research Center, Athens, GA.;National Reading Research Center, College Park,MDOffice of Educational Research and Improvement (ED).Washington, DC.94117A2000733p.

Reports Research/Technical (143)

MF01/PCO2 Plus Postage.*Beginning Reading; Mothers; *Parent ChildRelationship; Preschool Education; Reeding Aloud toOthers; Reading Research; Recall (Psychology);Recreational Reading; *Text Structure; *VocabularyDevelopmentCommunication Behavior; Emergent Literacy; ProjectHead Start

A study examined the ways in which Head Startpreschool children's vocabulary developed when they and their mothersengaged in joint reading. Subjects, 19 dyads, were observed as theyinteracted around expository texts presented in both familiar(newspaper toy advertisements) and traditional (trade books) formats.Subjects were observed in their homes for 10 readings each, duringwhich the dyads read a series of presented texts. The children'sability to identify words from the texts read and their comprehensionof a standardized receptive vocabulary list were measured. Motherstalked more than children in all contexts; furthermore, differentforms of talk were observed around the different text formats.Correlational and sequential analyses indicated that children's wordrecall was best predicted by responsive maternal strategies, such asencouraging children to talk about the text, and children's modelingof maternal strategies. (Contains 36 references and 5 tables of data.An appendix of data is attached.) (RS)

***********************************************************************

Reproductions supplied by EDRS are the best that can be madefrom the original document.

***********************************************************************

Joint Reading Between Mothersand Their Head Start ChildrenVocabulary Development in Two Text Formats

A. D. l'ellegriniLee GakiaUniversity of Georgia

Jane PerlmutterWestern Carolina University

It hel Jones

University of Georgia

U DEPARTMENT DE EDUCATION

4.4A, tit ,,f'. f

1 tre el

,e

. ot

NRRC NationalReading ResearchCenter

READING RESEARCH REPORT NO 13

Spring 1994

2 PFST COPY AVAILABLE

NRRCNational Reading Research Center

Joint Reading Between Mothers and Their Head Start ChildrenVocabulary Development in Two Text Formats

A. D. PellegriniLee Galda



Ithel June:,University of Georgia

READING RESEARCH REPORT NO. 13Spring 1994

The work reported herein n. a National Reading Research Project of the University of Georgiaand University of Maryland. It was supported under the Educational Research andDevelopment Centers Program (PR/AWARD NO. 117A20007) as administered by the Officeof Educational Research and Improvement, U.S. Department of Education. The findings andopinions expressed hem do not necessarily reflect the position or policies of the NationalReading Research Center, the Office of Educational Research and Improvement, or the U.S.Department of Education.

NRRC NationalReading ResearchCenter

Executive CommitteeDonna E. Alvermann, Co-DirectorUniversity of Georgia

John T. Guthrie, Co-DirectorUniversity of Maryland College Park

lama F. Baumann, Associate DiredorUniversity of Georgia

Patricia S. Koskinen, Associate DirectorUniversity of Maryland College Park

Linda C. DeGroffUniversity of Georgia

John F. O'FlahavanUniversity of Maryland College Park

James V. HoffmanUniversity of Texas at Austin

Cynthia R. HyndUniversity of Georgia

Robert SerpellUniversity of Maryland Baltimore County

Publications Editors

Research Reports axd PerspectivesDavid Reinking, Receiving EditorUniversity of Georgia

Linda Baker, Tracking EditorUniversity of Maryland Baltimore County

Linda C. DcGroff, Tracking EditorUniversity of Georgia

Mariam Jean DreherUniversity of Maryland

lastractiortal ResowresLee Gelds, University of Georgia

Research HighlightsWilliam G. HollidayUniversity of Maryland College Park

Policy BritfsJames V. HoffmanUniversity of Texas at Austin

VideosShawn M Glynn, University of Georgia

NRRC StaffBarbara F. Howard, Office ManagerMCILIN M. Erwin, Senior SecretaryUniversity of Georgia

Barbara A. Neitzey, Administrative AssistantVakne Tyra, AccountantUruversity if Maryland College Park

National Advisory BoardPhyllis W. AldrichSaratoga Warren Board of Cooperative E.ducationalServices, Saratoga Springs, New York

Arthur N. Appkbee&ate University of New York, Albany

Ronald S. BrandtAssociation for Supervision and CurriculumDevelopmentMarsha T. DcLainDelaware Department of Public Instruction

Carl A. GrantUniversity of Wisconsin-Madison

Walter KintschUniversity of Colorado at Boulder

Robert L. LinnUniversity of Colorado at Boulder

Luis C. MollUniversity of Arizona

Carol N. SantaSchool District No. 5Kalispell, Montana

Anne P. SweetOffice of Educational Research and Improvement,U S. .Department of Education

Louise Cherry WilkinsonRutgers University

Production EditorSusan L. YarboroughUniversity of Georgia

Dissemination CoordinatorJordana E. RichUniversity of Georgia

Text FormatterAnn VanstoneUniversity of Georgia

NRRC University of Georgia318 AderholdUniversity of GeorgiaAthens, Georgia 30602-7125(706) 542-3674 Fax: (706) 542.3678INTERNET: NRRCIauga.cc.uga.cdu

NRRC Univenity of Maryland College Park2102 J. M. Patterson BuildingUniversity of MarylandCollege Part, Maryland 20742(301) 405-8035 Fax (301) 314 9625INTERNET- NRRCgiumail umd.cdu

About the National Reading Research Center

The National Reading Research Center (NRRC) isfinided by the Office of .Educational Research andImprovement of the U.S. Department of EduCation toconduct research on reading and reading instniction.The NRRC is operated by a consortium of the Universi-ty of Georgia and the University of Maryland CollegePark in collaboration with researchers at several institu-

tions nationwide.The NRRC's mission is to discover and document

those conditioni in homes, schools, and communities.that encourage children to become skilled, enthusiastic,lifelong readers. NRRC researchers are committed toadvancing the development of instructional programssensitive to the cognitive, sociocultural, and motiva-tional factors that affect children's success in reading.NRRC researchers from a variety of disciplines conductstudies with teachers and students from widely diversecultural and socioeconomic backgrounds in prekinder-garten through grade 12 classrooms. Research projectsdeal with the influence of family and family-schoolinteractions on the development of literacy; the interac-tion of scciocultural factors and motivation to read; theimpact of literature-based reading programs on readingachievement; the effects of reading strategies instructionon comprehension and critical thinking in literature,science, and history; the influence of innovative groupparticipation structures on motivation and learning; thepotential of computer technology to enhance literacy;and the development of methods and standards foralternative literacy assessments.

The NRRC is further committed to the participationof teachers as full partners in its research. A betterunderstanding of how teachers view the development of

literacy, how they use knowledge from research, andhow they approach change in the classroom is crucial toimproving instruction. To further this understanding,the NRRC conducts school-based research in whichteachers explore their own philosophical and pedagogi-cal orientations and trace their professional growth.

Dissemination is an important feature of NRRC activi-ties. Information on NRRC research appears in severalformats. Research Reports communicate the results oforiginal research or synthesize the findings of severallines of inquiry. They are written primarily for re-searchers studying various areas of reading and reading

instruction. The Perspective Series presents a widerange of publications, from calls for 'research andcommentary on research and practice to first-personaccounts of experiences in schools. InstructionalResources include curriculum materials, instructionalguides, and materials for professional growth, designedprimarily for teachers.

For more information about the NRRC's researchprojects and other activities, or to have your nameadded to the mailing list, please contact:

Donna E. Alvermann, Co-DirectorNational Reading Research Center318 Aderhold HallUniversity of GeorgiaAthens, GA 30602-7125(706) 542-3674

John T. Guthrie, Co-DirectorNational Reading Research Center

2102 J. M. Patterson BuildingUniversity of MarylandCollege Park, MD 20742(301) 405-8035

5

NRRC Editorial Review Board

Patricia AdkinsUniversity of Georgia

Peter ArnerbachUniversity of Maryland College Park

Jelled MksUniversity of Georgia

Patty AsdersUniversity of Arizona

Tom AndertosUniversity of Illinois at Urbana-Champaign

Harriette ArringtonUniversity of Kentucky

ire.. skimPine Springs Elementary SchoolFalls Church, Virginia

Jobs BorkowskiNotre Dane University

Cyathia BowesBaltimore County Public SchoolsTowson, Maryland

Martha CarrUniversity of Georgia

Sozaaae OrwellMontgomery County Public SchoolsRockville, Maryland

Jona CokyWestern Maryland College

lkfichelk CommeyrasUniversity of Georgia

lAsida CooperShaAer Heights City SchoolsShaker Heights, Ohio

Karts CookieConsectkut Department of &locationHarOrd. Comecticut

Karin DahlOhio State University

Lyaae Dias-RkoCalifornia State University-San

Bernardino

Pamela DuastoaClenuos University.

Jim FloodSan Diego State University

Dana FosUniversity of Arizona

Linda GantbrellUniversity of Maryland College Park

Vakrie GarfieldChaualtoochee Elementary SchoolCumming, Georgia

Sherrie Gibaey-ShermanAthens-Clarke County SchoolsAthens, Georgia

Rachel GrantUniversity of Maryland College Park

Barbara GuzzettiArizona State University

Jane HaughCenter for Developing Learning

PotentialsSilver Spring, Maryland

Beth Ann HerrmannUniversity of South Carolina

Kathlees Wilbert'University of Georgia

Slams HillUniversity of Maryland College Park

Sally !Iodine-RomUniversity of Georgia

Cyathia HysdUniversiy of Georgia

Robert JimmiesUniversity of Oregon

Karen JolumoaPennsylvania State University

James KingUniversity of South Florida

Sandra KimbrellWest Hall Middle SchoolOakwood, Georgia

Kate KirbyGwinnett County Public SchoolsLawrencevill6, Georgia

Sophie KowconPrince Georges County SchoolsLandover, Maryland

Linda LabboUniversity- of Georgia

Rosary LalikVirginia PotYtechnk Institute

Michael LawUniversity of Georgia

Sarah McCartheyUniversity of Texas at Austin

Veda McClainUniversity of Georgia

Lisa McFalbUniversity of Georgia

Mike McKennaGeprgia Southern University

Donna MeskyLouisiana State University

Barbara lifschabveFowler Drive Elementary SchoolAthens, Georgia

Akiatunde MorakinyoUniversity of Maryland College Park

Inky MorrowRutgers University

Bruce MurrayUniversity of Georgia

Susan NeumanTemple University

Caroline NoyesUniversity of Georgia

John O'FlahavanUniversity of Maryland College Park

Peony (AntherUniversity of Georgia

Joan PagauccoUniversity of Georgia

Barbara PalmerMount Saint Mary's College

Mike PkkleGeorgia Southern University

Jessie PollackMaryland Department of EducationBaltimore, Maryland

Sally PorterBlair High SchoolSilver Spring. Maryland

Michael PressleyState University of New York

at Albany

John ReadeaceUniversity of Nevada-Las Vegas

Tom ReevesUniversity of Georgia

Lenore RingkrNew York University

Mary RocUniversity of Delaware

Nadia% T. RuizCalifornia State University-Sacramento

Rebecca SammonsUniversity of Maryland College Park

Paula SchwaneollugelUniversity of Georgia

Robert SerpaUniversity of Maryland Baltimore

County

Betty ShockleyFowler Drive Elementary SchoolAthens. Georgia

Swan SonnenscheiaUniversity of Maryland Baltimore

County

Steve StahlUniversity of Georgia

Anne SweetOffice of Educational Research

and Improvement

lAqing TaoUniversity of Georgia

Ruby ThompsonClark Atlanta University

Louise TomrsasoaUniversity of Georgia

Sandy TvasarkinStrawberry knolls Elementary SchoolGaithersburg, Maryland

Sheila VakaciaUniversity of Washington

Bruce VanSiedrightUniversity of Maryland College Park

Chris WakoaNorthern Territory UniversityAustralia

Janet WatkinsUniversity of Georgia

Louise WaynantPrince George's County SchoolsUpper Marlboro, Maryland

PriscWa WaynastRolling Terrace Elementary SchoolTakorna Park, Maryland

Dera WeaverAthens-Clarke County SchoolsAthens, Georgia

Jane WestUniversity of Georgia

Steve WhkeUniversity of Georgia

Allem WigfieldUniversity of Maryland College Park

Shelley Woos'University of Maryland College Park

About the Authors

A. D. Pellegrini is Professor of Early ChildhoodEducation at the University of Georgia and an investiga-tor with the National Reading Research Center. Hereceived his Ph.D. from Ohio State University and hasalso taught at the University of Rhode Island. Hisresearch and teaching interests center around children'sdevelopment in school and family contexts.

Lee Ga Ida is Professor of Language Education at theUniversity of Georgia and an investigator with theNational Reading Research Center. She received herPh.D. from New York University and has writtenwidely in the areas of early literacy and children'sresponse to literature.

Jane Perlmutter received her Ed.D. from tbeUniversity of Georgia and is Assistant Professor ofEducation at Western Carolina University.

Ithel Jones is a graduate student in the Department ofEarly Childhood Education at the University of Geor-gia. He received his B.A. from the University of Walesand his M.A. from the University of Wisconsin. Hisresearch interests include children's writing withcomputers and quantitative research methods.

NO4i001116.diag Reasarch Ciao.lioiveniiiss of Georgia and MarykkadRoadise **search Repoli No. 13SON/ 1994

Joint Reading Between Mothers and Their Head Start Children:Vocabulary Development in Two Text Formats

A.D. PellegriniLee Galda



Ithel JonesUniversity of Georgia

Abstract. This study examined the ways inwhich Head Start preschool children's vocabu-lary developed when they and their motherserataged in joint reading. The authors observednineteen dyads that interacted around expositorytexts presented in both familiar (newspaper toyadvertisements) and traditional (trade books)formats. The authors observed the dyads intheir homes for 10 readings each, during whichthe dyads read a series of presented texts. Thechildren's ability to identify words from the textsread and their comprehension of standardizedreceptive vocabulary list were measured. Moth-ers talked more than children in an contexts;furthermore, different forms of talk were ob-served around the different text formats. Corre-lational and sequential analyses indicated thatchildren's word recall was best predicted byresponsive maternal strategies, such as mcourag-isqg children to talk about the text, and children'smodeling of maternal strategies.

1

Adults reading to children is a very importantevent in children's lives. This event is impor-tant for many aspects of children's develop-ment, such as school-based literacy (e.g., Bus& van ljzendoorn, 1988; Cochran-Smith,1984; Dickinson & Smith, n.d.; Heath, 1982)and early language development (e.g., Ninio &Bruner, 1978). By being read to, children get"hooked on books" (Fader & McNeil, 1966),and they consequently become accustomed toschool-based literacy events (Cochran-Smith,1984; Heath, 1982). In addition, children'svocabulary is developed from tttorials on thewords presented in the books that are beingread to them (Cornell, Senechal, & Broda,1988; Werner & Kaplan, 1952). In thesecontexts, children's langyage is typically ex-panded by mothers in different ways, andmothers often explicitly tnodel and teach labelsfor pictured objects (e.g., Cornell et al., 1988;

2 A.D. Pellegrini, Lee Galda, Jane Perlmutter & Mel Jones

Snow & Goldfield, 1983). In the presentstudy, we extend this body of research byexamining the ways in which a group of HeadStart children recalled word meanings taught tothem by their mothers' presentation of differenttypes of text.

Book reading is a particularly good contextfor developing children's vocabularies. Asnoted by Werner and Kaplan (1952) and Ninio(1983), when adults and children interactaround books, they are jointly focused onpictures and corresponding text. It has beenshown that parents use a variety of strategies toteach and elicit from children the labels forpictures in the text (e.g., Ninio, 1983; Pelle-grini, Brody, & Sigel, 1985; Pellegrini, Perl-mutter, Galda, & Brody, 1990; Sorsby &Martlew, 1991). This rich descriptive researchbase has enabled experimental researchers toestablish causal links between specific maternalteaching strategies and children's word learn-ing status (e.g., Cornell et al., 1988; White-burst et al., 1988).

This research, however, is limited in twoimportant ways. First, with the exception of afew published studies (e.g.. Cornell et al.,1988; Pellegrini et al., 1990; Sorsby & Mart-lew, 1991), researchers have not attended tothe nature of books being read to children. In-deed, in some cases the specific books read arenot even mentioned in Methods sections of re-search reports. Although the nature of thebooks is important in terms of replication byother researchers, the literary genre of thebooks is specifically important to word teach-ing. Books in the narrative mode, traditionalstory books like Peter Rabbit, elicit minimalrates and varidies of mother-child interaction

and, correspondingly, few word teachingstrategies (Cornell et al., 1988; Pellegrini etal., 1990). By comparison, expository texts,like alphabet books, elicit significantly moreinteraction and specific word teaching strate-gies (Pellegrini et al., 1990). Furthermore,maternal word teaching strategies, even whenexperimentally manipulated, are differentiallyeffective for word learning from narrative andexpository texts (Cornell et al., 1988). Inshort, joint book-reading studia have not beenclear about the books they use as stimuli andhave confounded the literary genres of thebooks used in their studies. From the literaturethat does exist, it appears that words can betaught most effectively using expository texts.For these reasons, the children and mothers inthis study were observed interacting with aseries of expository texts.

A second limitation of the existing literatureis that it has generally been conducted withmainstream-culture children and mothers.When nonmainstream-culture children andmothers have been studied (i.e., generally,children and mothers from poor economiccircumstances), researchers observed veryyoung children in the earliest stages of wordlearning (e.g., Ninio, 1983; Ninio & Bruner.1978). In other cases, poor children have beenstudied outside of their home environments,such as in daycare centers (Valdez-Menchaca& Whitehurst, 1992), thereby placing thesefamilies in strange contexts that may inhibit,rather than maximize, their exhibition of com-petence (Bronfenbrenner, 1979).

It is important to study the interactionbetween nonmainstream-culture mothers andchildren in familiar environments. We know

NATIONAL READING RESEARCH CENTER, READING RESEARCH REPORT NO. 13

1 9

Joint Reading Between Mothers and Their Head Start Children

that when we study children and families infamiliar situations we get a more accuratedescription of the way interaction typicallyoccurs than in a strange situation (Bronfen-brenner, 1979), and therefore, parents andchildren often exhibit higher levels of compe-tence. Recent advances in context-specificcognition (e.g., Laboratory of ComparativeHuman Cognition ILCHC), 1983) suggest thatit is theoretically imriortant to describe theways in which different groups learn in localcontexts, as thse descriptions can he used todesign school curricula. Such considerationsare especially important in early educationprograms such as Head Start. Using texts thatare familiar to nonmainstream-culture childrenmay be one instructional strategy that we canuse to provide young children with opportuni-ties to become literate and to maximize theirpossible success in that endeavor.

Based on these assumptions, one wouldexpect these interaction styles between mothersand their children to vary according to thefamiliarity of the expository texts read. Con-tinuing with this reasoning, expository textsthat are familiar to participants, such as adver-tisements from a local store, should elicit highlevels of child participation and child-sensitivematernal teaching strategies (Pellegrini et al.,1990). Traditional expository trade books,such as alphabet and animal-labeling books,however, are less familiar to nonmainstream-culture families (Heath, 1983; Pellegrini et al.,1990). When young children are unfamiliarwith the text format, meaning must be jointlynegotiated with their mothers. In the presentstudy, we observed mothers and childreninteracting around expository texts in familiar

formats, such as newspaper toy advertisements,and in traditional children's book formats.

The techniques chosen to study the mother-child interactions were rather traditional. Wechose to analyze the extent to which maternaland child utterances varied around familiar andtraditional texts. Additionally, we examinedwhich strategies were significant predictors ofchildren's identification of words presented inthe text, as well as how these strategies relatedto a more general measure of children's recep-tive vocabulary, using the Peabody PictureVocabulary Test. A classification system formaternal and child utterances was implementedwith an eye on the function that these utteranc-es might serve in word teaching and recall.Our, implemented category system was influ-enced by the extant joint-book-reading litera-ture, especially the work of Cochran-Smith(1984), Cornell (Cornell et al., 1988), Dickin-son (Dickinson & Smith, n.d.), Heath (1982),and Pellegrini (Pellegrini et al., 1990).

Categories for the types of maternal ques-(ions and the adequacy of children's responseswere generated. Additionally, children's in-correct responses to maternal questions wereexamined. Both sorts of utterances wereanalyzed because they seemed to be explicitlyused by mothers to teach or explicate the wordsprented in these texts. Recognizing thatchildren's contributions to literate discourse areimportant in word comprehension (Cornell etal., 1988; Dickinson & Smith, n.d.; White-hurst et al., 1988), we also considered theirinitiations of and reactions to maternal strate-gies. Specific strategies used by both mothersand children, such as terms explicating linguis-tic process (e.g., read, talk, write), were


4 A.D. Pellegrini, Lee Ga lda, Jane Perlmutter & Mel Jones

examined, as were maternal expansions ofchildren's utterances, which were includedbecause they have been implicated in earlyliteracy (Dickinson & Smith, n.d.; Olson,1983).

Two interaction strategies, text-world andworld-text, seemed particularly important in astudy focusing on teaching and learning wordsin familiar and unfamiliar contexts. Thesestrategies, like Cochran-Smith's (1984) text-to-life and life-to-text interactions, are importantindicators ofyways in which meaning frombooks is extended or applied to extratextualmatters (i.e., text-world and text-to-life), andthe ways in which text is understood and com-prehended in terms of current knowledge (i.e.,world-text and life-to-text). These strategies,which are more explicitly defined with accom-panying examples in the Method section and inthe Appendix, should be important in teachingand recalling words because learning involvesrelating the new information to existing infor-mation, not unlike Piaget's (1983) notion ofassimilation and accommodation.

In an effort to contextually situate thematernal and child utterances, sequential laganalyses were performed. Specifically, thoseutterances that were significantly related tochildren's recall of words from the texts readas well as their scores on a generalized vocabu-lary measure were analyzed in terms of ante-cedent and consequential utterances. In thisstrategy, we identified those utterances thatpreceded and followed the utterances whichwere significantly related to the criterion mea-sured (i.e., recall and vocabulary). For exam-ple, if a child's use of a linguistic term wassignificantly correlated with the recall of words

from a familiar text, we analyzed the maternalutterances that preceded and followed thechild's use of the term. Correspondingly, if amaternal expansion around the traditional textswas related to a child's recall of the wordspresented, we analyzed the child utterances thatpreceded and followed the maternal expansion.By examining target utterances in this way, wecome closer to recognizing the different func-tions served by a specific utterance; that is, thefunction, or consequence, of an utterance canbe determined by examining utterances thatfollow it. Ideally, we would have liked toconsider utterances at the multiple-turn level,that is, beyond lag 1 but unfortunately, sucha strategy requires a sample size (see Bakeman& Gottman, 1986) that is often difficult tomanage with the type of transcription and textanalyses done in the present study.

In this report we examine the ways inwhich mothers taught their Head Start childrenwords depicted in traditional and familiarexpository texts. We hypothesized, followingthe extant joint-reading literature, that thosematernal strategies that maximized children'sverbalizations would be the best predictors ofchildren's identification of the words presentedin the texts, as well as of their more generalvocabulary level.

METHODSubjects

The mothers and children in this study wererecruited as part of a larger study involvingHead Start centers in a small city. Parentswere asked by Head Start teachers if they wereinterested in participating in an early literacy


Joint Reading Between Mothers and Their Head Start Children 5

project. In exchange for participation, recruit-ed parents were paid $40. This resulted in therecruitment of a number of mothers, 19 ofwhom are described in this report. All familieswere classified as lower socioeconomic statusby Head Start standards and identified them-selves as either African-American or Cauca-sian. The mothers' average years of educationwere 10.6.

The children in this study (11 boys and 8girls) had a mean age of 51 months (4.25years). Children's gender was not used as agrouping variable because previous mother-child hook reading research with similar sam-ples (e.g., Pellegrini et al., 1990), as well aswith a middle-class sample (Pellegrini et al.,1985), found no gender-of-child effects.

Procedures

All observations took place in the homes of thetarget families. The first home visits were toestablish rapport with the families, to explainthe procedures of the project, and to completethe HOME inventory (Caldwell & Bradley,1984). Through the HOME interviews, wefound that all families received one or twolocal newspapers. One was a commercial dailynewspaper with national, state, and local news,as well as advertisement sections. The othernewspaper was primarily an advertising paperalthough it did have some local news; it wasdelivered to all homes in the area free ofcharge. Mothers read newspaper comicsandfor toy advertisements with their children.Given the contrived nature of strangers video-taping interactions in the home, four obseiva-tiotg were taken so as to maximize the likeli-

hood of mothers and children exhibiting high-est levels of competence (Tulkin, 1972; Wac-hs, 1985). Individual home ygits were sepa-rated by one week.

The books read by mot s and childrenwere provided by the experimenters. Fourdifferent expository texts were used: Twowere typical children's trade books (Who Livesat the Zoo? and My First Book of Words, bothGolden Books), and two were expository textsconstructed by the experimenters to reflect themore familiar format of the labeled toy adver-tisements from the local newspapers and la-beled pictures from the children's preschool.These texts were constructed and bound asbooks. Mother-child reading sessions werevideotaml and later transcribed.

Measures

Transcription and Coding Scheme Prepa-ration. Transcriptions of all verbalizations andgestures made by both mothers and childrenwere made from the videotapes. The sequen-tial integrity of the verbal and nonverbal be-havior was maintained in the coding schemesand in the data analyses. Coding schemes forchild and mother behavior were developed instages by the first author. First, separately forchildren and mothers, individual utteranceswere labeled according to the specific functionthey seemed to serve in relation to vocabularyteaching. The result was an overwhelming 242maternal categories and 134 child categories.Next, these individual categories were groupedinto categoria that served similar functions.Utterances were included for analyses only ifthey were assigned to the same functional cate-


6 A.D. Pellegrint Lee Galda, Jane Perlmutter & lthel Jones

gory on three out of the four read-throughs.This resulted in 26 molar maternal categoriesand 23 molar child categories (see the Appen-dix for a full list with examples).

The extent to which utterances could bereliably sorted into these molar categories wasdetermined in two ways. First, as noted previ-ously, each utterance was assigned t: the samecategory on three of four independen% i.ortingsby the same coder. Second, a second observersorted 100 randomly chosen utterances intomolar categories to a level of .93 (Spearman'srho). All categories were mutually exclusive.

Mothers' Language. Mothers' languagewas put into 1 of 26 categories listed below(see Appendix for specific descriptions of eachcategory).

1. Answers own questions2. Describes actions3. Clarify4. Conflict/disagree5. Corrects6. Asks for description7. As.s to label8. Asks to read or spell9. Asks about a subordinate

10. Asks about a superordinate11. Describes12. Gives superordinate label13. Gives subordinate label14. Expands child's utterance15. Linguistic question16. Linguistic description17. Narrativizes18. Negative reinforcement19. Orient20. Reads present text21. Reinforce/repeat

22. Slot/frame provided23. Text-world24. WH-question25. World-text26. Yes/no question

Children's Language. There were 23categories as listed below (see Appendix forfull descriptions).

1. Acknowledge2. Ask question3. Clarify4. Conflict/disagree3. Correct answer to subordinate-label ques-

tion6. Correct answer to superordinate-label

question7. Correct answer to other questions8. Inadequate answer9. "Don't know"

10. Describes subordinate11. Describes superordinate12. Expands/extends13. Initiates/labels subordinate14. Initiates/labels superordinate15. Linguistic16. Narrativizes17. Not relevant to preceding utterance or

text18. Reads19. Repeats20. 'Text-world21. "Want"22. World-text23. Other

Children's General Receptive Vocabulary.Children were tested individually on the Pea-body Picture Vocabulary Text (PPVT) (Dunn& Dunn, 1981), a measure of children's recep-


1 4


tive vocabulary. The unit of analysis we usedfor children's receptive vocabulary was thetest's standardized scores (M = 100, SD =15). The average PPVT score for this samplewas 73.40 (SD = 17.82).

Children's Words Identified Front SpecificBooks Read. Immediately after mothers readeach hook to their children, the children wereasked a series of word identification questionstaken from the book just read. More specifi-cally, children were asked by an experimenterto identify (i.e., "What's this called?") a seriesof pictures taken from each text. Childrenwere asked a total of 10 questions from eachtext, and their responses were audiotaped. Thefirst response to the experimenter's questionwas recorded. A response was consideredcorrect if the child used the label given in thetext. The total number of correct responses foreach text format Was the unit of analysis; thatis, the total number correct from each text inthe familiar format was aggregated into onescore. Similarly, the score for the traditionaltexts was the aggregate of the number correctfrom each of the two recall tests.

RESULTS

The initial series of analyses described thevariation in mothers' and childrens' utterancesas a function of text format. To that end, aseries of 2 (person: mother, child) X 2 (textformat: familiar, traditional) repeated-mea-sures analyses of' variance (ANOVA) werecalculated on categories of utterances thatmothers and children used in common. In allcases, except where the total number of utter-ances was the dependent variable, utterances

were expressed in terms of their relative fre-quency to the total number of utterances usedby that individual in that context. The descrip-tive statistics for mothers' and children's utter-ances are displayed in Table I.

The following child and maternal utterancesthat were used in common were analyzed:total utterances, descriptions, questions, clarifi-cations, conflicts, expands, linguistic terms,narrativizes, reads, repeats, text-to-world,world-to-text. When we compared the totalutterances generated by mothers and childrenin a repeated-measurm ANOVA, a significantmain effect for person was detected, F(1, 32)= 25.48, p< .0001, with mothers generatingsignificantly more utterances than children, andfor text format, F(1,32) = 9.29, p< .004, withmore talk being observed around traditionaltexts than around familiar texts. There werealso significant person effects on the followingcategories: clarifications, F(1, 32) = 3.92,p< .05; expands, F(1,32) = IO.2s$,p < .003;and conflicts, F(1,32) = 5.40, p< .02. For allcategories, except conflicts, mothers generatedmore utterances than did children.

For the linguistic terms category, there was

a significant main effect for person, F(1,32) =10.07, p< .003, with more linguistic termsbeing used by mothers than children. Ther-was also a text-format effect for linguisticterms, F(1, 32) = 6.31, p< .01, with morebeing used around familiar format texts thanwith traditional texts. There were significanttext-format effects for both text-world, F(1,32) = 7.07, p< .01, and world-text, F(1, 32)= 5.64, p< .02, categories: More linguisticterms were used with familiar than traditionaltexts.


1 5

A.D. PellegrinI, Lee Ga Ida, Jane Perlmutter & Ithel Jones

Tabie I. Descriptive Statistics for Mothers and Children's Utterances by Text Format'

Traditional Familiar Text (FT)

M SD

Text rnSD

Maternal

Answers own questions 0.02 0.03 0.04 0.04Describes action 0.01 0.02 0.01 0.01

Clarify 0.01 0.01 0.01 0.02Conflict/disagree 0.01 0.01 0.01 0.01

Corrects 0.02 0.02 0.01 0.01

Asks for description 0.02 0.05 0.01 0.01

Asks for a label 0.02 0.04 0.01 0.02Asks to read or spell 0.01 0.01 0.01 0.01

Asks about a subordinate 0.01 0.01 0.01 0.01

Asks about a superordinate 0.01 0.01 0.01 0.01Describes 0.05 0.05 0.07 0.07Gives superordinate label 0.02 0.02 0.04 0.03Gives subordinate label 0.34 0.37 0.12 0.16Expands child's uttrrance 0.01 0.01 0.01 0.01

Linguistic question 0.05 0.05 0.11 0.11

Narrativizes 0.01 0.02 0.01 0.01

Negative reinforc.ement 0.01 0.01 0.01 0.01

Orient 0.01 0.01 0.01 0.02Reads present text 0.12 0.32 0.03 0.06Reinforce/repeat 0.07 0.06 0.09 0.05Slot/frame provided 0.01 0.02 0.01 0.02Text-world 0.01 0.02 0.03 0.03WH-qution 0.14 0.14 0.28 0.20World-text 0.00 0.00. 0.00 0.00Yes/No question 0.01 0.01 0.01 0.01

Total Ur:trances 191.94 183.72 109.70 105.41

*Figures are reported as proportion.s.



Table 1. Descriptive Statistics for Mothers' and Children's Utterances by Text Format (continued)

Traditional Text (T1')

hi SD

Familiar Text (FI)

hi SD

Child

Acknowledge 0.01 0.01 0.02 0.03

Ask question 0.01 0.02 0.02 0.02

Clarify 0.01 0.01 0.00 0.01

Conceptual conflict/disagree 0.02 0.03 0.02 0.02

Correct answer (Subordinate) 0.08 0.10 0.16 0.13

Correct answer (Superordinate) 0.03 0.04 0.05 0.05

Correct enswer (Other) 0.06 0.13 0.07 0.07

Inadequate answer 0.02 0.03 0.02 0.05

'Don't know 0.01 0.01 0.02 0.02

Describes subordinate 0.02 0.03 0.02 0.03

Describes superordinate 0.01 0.01 0.00 0.00

Expands/extends 0.01 0.01 0.00 0.01

Initiates (Subordinate) 0.05 0.06 0.06 0.05

Initiates (Superordinate) 0.01 0.01 0.01 0.01

Linguistic 0.02 0.03 0.03 0.03

Narrativize 0.01 0.01 0.00 0.00

Not relevant 0.50 0.40 0.29 0.25

Reads 0.01 0.01 0.01 0.01

Repeats 0.10 0.11 0.08 0.13

Text:world 0.01 0.01 0.01 0.02

'Want' 0.01 0.01 0.02 0.03

World-Text 0.01 0.01 0.01 0.02

Total Utterances 144.29 157.69 69.70 72.52


1 7

10 A.D. Pellegrini, lee Gia Ida, Jane Perlmutter & Idyl Jones

In the next series of analyses, we examinedthe relation among mothers' utterances, chil-dren's utterances, and children's identificationscores on words presented in the differenttexts. In addition, we examined the ways inwhich these utterances related to children'sgeneral vocabulary (i.e., PPVT) scores. Be-cause of the large number of interaction cate-gories, an a priori decision was made in choos-ing categories to be examined. We includedthose categories that have been shown, in theextant research, to be related to children's vo-cabulary development. Specifically, we exam-ined the following maternal variables: asks forlabels, asks about subordinates, asks aboutsuperordinates, expands children's utterances,linguistic description, and text-world andworld-text statements. The child utterancesconsidered included: correct answer to subor-dinate and superordinate questions, initiatessubordinate and superordinate labels, andworld-text and text-world utterances. Thesecorrelations are displayed in Table 2. In thistable we display the ways in which maternaland child utterances separately relate to theseparate measures of words identified andPPVT scores. As can be seen in this table,maternal and child utterances were correlatedseparately with children's identification ofwords from both familiar and traditional textsas well as with children's PPVT scores.

Regarding maternal utterances, asking forlabels and asking about subordinates bothrelated to children's scores from familiar texts.Maternal expansions of children's utterancescorrelated positively and significantly with allof the child measures. Children's identificationscores from traditional texts were predicted by

maternal expansions, linguistic utterances, andtext-world utterances. Regarding relationswith the PPVT scores, the following maternalutterances around traditional texts were signifi-cant predictors: expands, linguistic, and text-world. Expands, however, was a significantpredictor only for familiar texts.

Children's utterances also were used ascorrelates of their identification scores on thebooks read as well as on the PPVT. Thesecorrelations are also displayed in Table 2.Children's utterances were correlated withword identification and PPVT scores only ontraditional texts. The following child utteranc-es were significant predictors for word identifi-cation: correct answers to subordinates,world-text utterances, and text-world utteranc-es. For PPVT scores, the following childutterances were significant: correct subordi-nate, initiate superordinate, and world-text.

In the next series of analyses, we situatedthe utterances that were significant correlatesof word identification and vocabulary in theirdiscourse contexts by examining antecedentand consequential utterances generated by theother interlocutor. For example, in the case ofa specific target maternal utterance, say expan-sion, which related to children's word identifi-cation from traditional texts, we would exam-ine the child utterances preceding and follow-ing the target maternal utterances. By precedeand follow, we mean utterances generated inthe turn of the other, nontarget interlocutorimmediately before and after the target utter-ances. The analytical procedure used, follow-ing Bakeman and Gottman (1986), constructedtransitional probability matrices and corre-sponding z-scores for a series of two-state

NATIONAL READING RESEARCH CENTER, READING RESEARCH REPORT NO, 13

1 S


Table 2. Correlations Among Maternal and Child Utterances and Word Learning and Vocabulary Measures

Traditional TextPPVT

Familiar TextPPVT

TraditionalText

Words

Familiar TextWords

Maternal

Asks label .17 .18 .37 .40*

Asks abordinate .27 .38 .08 .470*

Asks superordinate -.01 .20 -.20 ..23

Expands .52** .56** .48**

Linguistic .59*** .22 .30

Text-world .42* .36 .46* .38

World-text .00 -.17 .00 -.35

Child

Linguistic .36 .27 .14 .24

Correct suliordinate .47** .27 .46* .33

Correct superordinate .32 .06 -.12 .34

Initiate subordinate .30 .05 .29 -.33

Initiate superordinate .47** .02 .20 -.31

Text-world -.30 .05 -.42* -.04

World-text .63*** .15 .65*** .03

p<.10, 04.p<.05,***p.01

models where the target utterance, such as thetarget of maternal asking for label, was at lag0 and all categories of children's consequentialutterances were at lag 1 . Lag 1 was defined asan utterance in the subsequent turn. A turn isdefined as talk generated by one person, boun-ded by the talk of another person. To deter-mine antecedents of the target utterance, againusing mothers' asking for label as an example,we determined the child utterances, at lag 0,that preceded it at a greater-than-chance rate.Separate two-state models were tested for all of

the previously noted maternal and child vari-

ables. The statistically significant mother-to-child transitions around traditional and familiartexts are displayed in Tables 3 and 4, respec-tively, and the significant child-to-mothertransitions are displayed in Table 5. Z-scoresgreater than 1.96 are indicative of utterancesthat occur at a greater-than-chance probability.

Starting with maternal utteranca generatedaround traditional and familiar texts that relatedto children's identification, we examined thechild utterances that preceded and followed thefollowing maternal categories: asks for labels,asks about subordinates, expands, text-world.


1.9

12 A.D. Pellegrinl, Lee Galda, Jane Perlmutter & Ithel Jones

Table 3. Maternal Utterances As Targets Around Traditional Texts

Child: Antecedent Mother: Target Child: Consequence

Correct SuperordinateGives Label SubordinateNarrativizes

AcknowledgeCorrect SuperordinateInitiates Label SuperordinateNot Relevant (-)

Asks QuestionDescribes SubordinateInitiates/Labels SubordinateInitiate/Labels SuperordinateNot Relevant (-)World-Text

Don't KnowWorld-Text

Ask Abel

Ask About Subordinates

Exiianih

Text-World

Correct Answer SubordinateInitiates/Labels SubordinateNot Relevant (-)Repeat (-)

Text-worldDescribe SubordinateNot Relevant (-)Incorrect Response

AcknowledgeAsks QuestionExpands/ExtendsInitiates/Labels SubordinateNarrativizesNot Relevant (-)

AcknowledgeNot Relevant (-)Text-WorldWorld-Text

It should he noted that one class of maternal

utterances that was significantly correlated withchildren's word identification and vocabularystatus, linguistic terms, was not followed at agreater-than-chance rate by any one class ofchild utterances. Table 3 shows that the childutterances following mothers' asking for labelswere correct answers to subordinates for bothtraditional and familiar texts, and gives subor-dinate label, not relevant responses, and incor-

rect responses for traditional texts. Mothers'asking for labels around traditional texts waspreceded by children's correct superordinate,correct label subordinate, and narrativizes.For familiar texts, children's use of initi-ates/labels subordinates followed mothers' ask-ing for label at a greater-than-chance rate,whereas no one category of child utterancespreceded this maternal category at a greater-than-chance rate.



Table 4. Maternal Utterances As Targets Around Familiar Texts

Child: Antecedent Mother: Target Child: Consequence

Correct Superordinate"Don't Know'Repeat

Correct SuperordinateDescribes SubordinateNot Relevant (-)Text-World

Ask Label

Ask About Subordinate

Expands

Text-World

Correct Answer SubordinateInitiates/Labels Subordinate

Describes SubordinateInitiates/Lthels SubordinateNot Relevant (-)

'Don't Know"Initiates/Lables SubordinateNot Relevant (-)

AcknowledgeCorrect Answer SubordinateWorld-Text

When mothers asked about subordinates,the following child responses occurred at astatistically significant rate, although only fortraditional texts: text-world utterances, de-scribes subordinate, not relevant, and incorrectresponse. This maternal category was preced-ed by children's use of acknowledge, correctsuperordinate, initiates/labels superordinate,and not relevant. When mothers expandedchildren's utterances around traditional texts,children responded in the following ways: asksquestions, acknowledge, expand/extend, initi-ates/labels subordinate, not relevant, and narra-tivize. Mothers' expansions around traditional

texts were preceded by the following children'sresponses: asks questions, describes subordi-nate, initiates/labels subordinate and superor-dinate, not relevant, and world-text. Aroundfamiliar texts, children responded to expan-sions by dscribing subordinates, initi-ating/labeling subordinates, and responding innonrelevant ways. Children's antecedents tomaternal expansions around familiar textsincluded correct superordinate, describessubordinate, not relevant, and text-world.

Because children's utterances were predic-tive of word identification and related to PPVTscores only with traditional texts, maternal


14 A.D. Pellegrini, Lee Ga Ida, Jane Perlmutter & Ithel Janes

Table 5. Child Utterances As Targets Around Traditional Texts

Mother: Antecedent Child: Target Mother: Consequence

WH QuestionAsk Subordinate

Answers Own Question (-)ClarifySlot/Frame

Asks SubordinateGives LabelText-World

LinguisticWorld-Text

Incorrect

Corrtct Atxs-wr Subordinate

Text-World

World-Text

Initiates Superordinate

Answers Own QuestionGives Label SubordinateNegative Reinforce

Givs Label SuperordinateReinforce/RepeatSlot/FrameWH Question

Gives Label SubordinateReinforce/RepeatWH Question

ExpandsReinforce/RepeatSlot/FrameText-World

DescriptionSlot/Frame

antecedents and consequences of children'sutterances will be presented only from thatcontext. These sequential analyses are dis-played in Table 5. Children's correct answersto subordinate questions were followed at a sig-nificant rate by the following maternal utter-ances: gives superordinate label, rein-force/repeat, slot/frame, and WH -question.This child category was preceded by the fol-

lowing maternal categories: answers ownquestions, clarify, and slot/frame. Children'stext-world utterances were followed by thefollowing maternal categories: gives subordi-nate label, reinforce/repeat, and WH-question.The same child category was preceded by thefollowing maternal categories: asks subordi-nate, gives label, and text-world. Children'sworld-text utterances were followed by mater-



nal expansions, reinforce/repeat, slot/frame,and text-world and preceded by maternallinguistic and world-text utterances. Children'sinitiating superordinates were followed bymaternal descriptions and slot/frame utteranc-es. This form of child initiation was not reli-ably preceded by any one form of maternalutterance.

We were also interested in the ways inwhich mothers responded to children's incor-rect responses to questions. Around traditionaltexts, mothers responded with the following:answers own questions, give subordinate label,and negative reinforcement. Around familiartexts, mothers responded with the following:conflict/behavior, gives subordinate and super-ordinate labels, negative reinforcement, rein-force/repeat (negative z), and yes/no question.

DISCUSSION

The general intent of this report is to describethe ways in which mothers and their childreninteracted around familiar and traditional booksand to determine the relations between theseinteraction patterns and children's vocabulary.Our initial analyses were straightforwardcomparisons of types of utterances generatedby mothers and children around traditional-format and familiar-format expository texts.Not surprisingly, mothers did most of thetalking in both contexts. Mothers not onlygenerated more total utterances than children,but they also generated more utteranca in eachcategory, except for the conflict category.Importantly, maternal utterances were respon-sive to or contingent on children's precedingutterances. These results are consistent withmuch research which describes child-adult joint

interaction as reciprocal, with each partici-pant's behavior being contingent on the other's ,

(e.g., Ninio & Bruner, 1978; Vygotsky, 1978).Future research should address joint readingbetween children and other members of theirsocial network, such as siblings. It may be thatthe affective and cognitive qualities of specificrelationships, such as the cooperation betweensiblings during fantasy (Dunn, 1988), haveimportant implications for early literacy andword learning.

The next comparison was in terms of text-format familiarity. We assumed that thesemothers and children would exhibit differentstrategies around different formats. Such de-scriptions are interesting in their own right, tothe extent that they provide important infor-mation on the ways in which these mothers andchildren jointly construct word meaning.These comparisons could also provide a basisfrom which to design early literacy curriculummaterials.

First, more talk was generated around thetraditional-format texts, compared to the famil-iar. This may have been because traditionalexpository texts are typically written in a"decontextualized" style. In such a style,shared knowledge assumptions and contextualclues are minimized; consequently, childrenneed help deriving meaning from these texts.The mother's job here is to help the childnegotiate the meaning of the texts by talkingwith the child about them (Cochran-Smith,1984). The necessity for negotiated under-standing results in more language being gener-ated around traditional than familiar texts.

Specific strategies, which have been impli-cated in early literacy and word learning, weregenerated more frequently around fPaniliar texLs


23

16 A.D. Pellegrini, Lee Galda, Jane Perlmuuer & Ithel Jones

than around the traditional texts. Specifically,mothers and children used more linguisticterms and text-world and world-text utterancesin the familiar, compared to the traditional,context. Use of linguistic terms involvedchildren and mothers talking about the linguis-tic process per se. Examples I and 2 depict theuse of these terms by both mothers (M) andchildren (C).

1. M: OK. Now you try to read.C: I can't [read]M: Yeah, you can [read]. Just try.

2. C: Clown has c.M: Good. What (letter) does toy have?

As can be seen in the preceding examples,when children and mothers use linguisticterms, they are making linguistic processesexplicit by holding them up for examination,discussion, and verbal encoding. Olson (1983),among others (e.g., Pellegrini et al., 1990;Watson, 1990), argues that use of linguisticterms is an indicator of linguistic awareness, orawareness of the processes and rules underly-ing language use. Other work in this areasupports this conclusion to the degree thatchildren's use of linguistic terms, of the sortused in this study, are correlated with measuresof metalinguistic awareness such as phonemicawareness and word-boundary awareness(Galda, Stahl, & Pellegrini, 1993). The use ofthese terms, in turn, has been implicated inschool-based literacy (Galda et al., 1993; Tor-rance & Olson, 1 9a4) and vocabulary develop-ment (Dickinson & Smith, n.d.). Ifmay he, assuggested by Olson and colleagues, that the use

of these terms is an indicator of metalinguisticawareness, which, in turn, is necessary forliteracy. Alternatively, it may be that theterms are important in school-based literacyonly because school literacy lessons are charac-terized by the segmentation and labeling of thelinguistic components of the reading process.

Text-world utterances, illustrated in Exam-ples 3 and 4, and world-text utterances, illus-trated in Examples 5 and 6, were also usedmore frequently around the familiar formatthan around the traditional format.

3. M: That's like ours.C: Yeah, Jackie broke it.

4. M: Trucks. Grandpa has a little truck.C: I wanna big truck.

5. M: Your ball like that.C: Basketball!

6. M: There's your TWdy.C: Yeah! 1,2,3 Ttxidies!

Cochran-Smith (1984) suggests that these twofrequently used joint-reading strategies areindicative of different ways in which meaningis "taken" from text. These ways of takingmeaning from texts are culturally learned, notuniversal (Barthes, 1974; Cochran-Smith,1984; Heath, 1982). Indeed, Heath (1982)suggests that mainstream-culture mothers,while reading to their young children, teachthem specific ways in which to interrelate textand personal experience; Heath's (1982) non-mainstream-culture children and mothers didnot interrelate book knowledge and worldknowledge. This sort of interrelating may not



happen with traditional texts, especially "story"books, but does happen with certain types ofeveryday expository text, like toy advertise-ments. The function of advertisements, bothfor adults and children, is to relate the objectspresented for sale to a personal need or desirefor those objects; thus, the motivational com-ponent of interacting around familiar texts maybe important for the extension from text to lifeand vice versa.

That more of the text-world and world-textstrategies were used in the familiar text formatmay also indicate that participants were able touse a variety of strategies to make sense of thewords presented. Both children and mothersrelated the words presented in the texts to theirpersonal experiences as well as related theirpersonal experiences to the words presented.

That participants took meaning from famil-iar, compared to traditional, texts by bothrelating the text to the world and relating theworld to the texts is methodologically impor-tant as well. This finding supports the long-held criticism of some types of developmentalresearch that study children and their familiesin relatively strange situations for brief periodsof time. Participants will exhibit higher levelsof competence in situations that are familiarand valued, compared to less familiar and lessvalued contexts (Bronfenbrenner, 1979; Ogbu,1981). In familiar tasks, much of the ordinarywork, such as understanding a specific lexicalitem or determining interaction rules, becomesautomatized; consequently, more cognitiveresources can be allocated to reflection. Thatthese familiar tasks are indigenous may alsomean that they are valued by participants; this

value may be what motivated participants toreflect upon the tasks.

The extent to which these strategies areused in different contexts by each participantalso informs us as to the ways in which theytake meaning from traditional and familiartexts. These strategies enable the child tointerrelate the material presented in the textswith what is already known. In this regard thetext-world and world-text strategies are similarto Piaget's (1983) assimilation and accom-modation, respectively. When used in con-junction, these strategies should be importantin word teaching and learning. As illustratedin Examples 6, 7, and 8, children's use ofworld-text strategies was a reliable predictor oftheir word identification and vocabulary status.

Next we examined the interrelation betweeninteraction variables (by mothers and children)within each text format, children's identifica-tion of words presented in those contexts, andthe children's more general vocabulary status.A striking finding in the correlation analyseswas the lack of a single statistically significantcorrelation coefficient (a total of 14 werecalculaux1) between children's utterancesaround familiar-format texts and their wordrecall and vocabulary scores. As reviewers arefond of saying: One or two significant correla-tions would be expected by chance. Therewere, however, numerous significant correla-tions between children's language aroundtraditional texts and word-learning and vocabu-lary status. These findings are certainly note-worthy and in need of explanation. The firstand most obvious explanation might be foundin a ceiling effect On the word scores from the


18 A.D. Pellegrini, Lee Ga Ida, Jane Perlmutter & !the! Jones

familiar texts. No such ceiling was observed:The mean score was 4.95 out of a total of 10.

Our preferred explanation relates to thesimilarity in the design features of the tradi-tional texts and the word-testing contexts,features not shared by the familiar format.Children's books in the traditional expositorygenre are decontextualized to the extent thatmeaning is conveyed by the text itself or byverbal interaction between tutors and tutees,not by shared aFsumptions between interlocu-tors. In joint reading, mothers and childrensocially negotiated the meaning of these textsso that the texts were mutually understood.Similarly, in the specific word texts for eachbook and for the PPVT, children were expect-ed to use language to define words. Here toodecontextualized meaning was conveyed. In

the familiar format, meaning can be conveyedthrough shared knowledge, not explicitly withlanguage, because participants are familiar withthe material being presented.

This argument is further supported whenwe examine the specific child utteranca thatwere predictors of word identification andvocabulary status. Children's correct answersto subordinate questions were significant posi-tive predictors of both measures. in Examples6 and 7, mothers ask a question about a subor-dinate, and children respond with a correctsubordinate label.

6. M: What's that giraffe lookin' at?C: Her baby.

7 M: Where that camel?C: Near the 'potamus.

Similarly, in the testing questions, childrenwere asked to identify specific subordinateclasses. The experimenter showed the child apicture from the book and then asked the child:What's this? The child was expected to re-spond: Giraffe. Thus, in both formal testingand responding with subordinates, children areresponding to demands to provide verballabels. These test-like sequences are reminis-cent of the teacher-child sequences describedby Mehan (1979).

Children's use of world-text utterancesaround traditional texts was also positively andsignificantly related to word learning andvocabulary status_ For example:

8. C: 'lanta zoo like that.

This strategy is used as a way of relatingsomething external to the text to the text itselfand, consequently, making the text comprehen-sible (Cochran-Smith, 1984). In short, it

seems a very effective learning strategy, and itis not surprising that children who used it

identified words from texts and had high vo-cabulary scores.

Children's text-world utterances, on theother hand, were negative predictors of wordidentification (significantly) and vocabulary(not significantly). This was unexpected in thatthis class of utterance can be used as a way inwhich to gain information (Cochran-Smith,1984); information from the text can be appliedto extratextual contexts. Consequently, thisshould have related to vocabulary status. Thisresult becomes more understandable, however,when we examine the matern:41 lerances thatfollowed children's text-v. utterainces. Two



of the three maternal utterances which followeddid not require children to talk further aboutthe reference. One frequently occurring sce-nario had mothers providing labels for chil-dren, whereas another involved mothers repeat-ing children's utterances. Only when mothersasked WH-questions after children's text-worldutterances, were they required to verbalizefurther. As illustrated below, requiring chil-dren to talk about pictures and text, often viamaternal expansions, is a powerful predictor ofword identification.

Maternal utterances that predicted child-ren's word identification and vocabulary in-cluded asks label, asks subordinate, text-worldutterances, linguistic terms, and expansions.Thme findings reinforce extant theory andresearch that highlight the importance of mater-nal strategies that, generally, enable children totalk about the text and, more specifically,enable children to reflect on the text and itsmeaning. Specific to text-world utterances, asnoted previously, this is an important strategyfor understanding text (Cochran-Smith, 1984;Heath, 1982; Stahl & Fairbanks, 1986).

Maternal expansions of children's utteranc-es were important in both text formats. In-

deed, expansions are important for a number ofchild language-learning measures. The use ofexpansions by mothers and other adults pre-dicts many aspects of children's linguisticfacility because they get children to processadult language and then use their Own language(Cazden, 1965; Cornell et al., 1988; Dickinson& Smith, n.d.). In the present study, maternalexpansions were preceded by a number of childutterances, ranging from seemingly irrelevantutterances to providing labels and interrelating

textual and world information. Thus, mothersseemed to use this strategy in a variety ofdiscourse contexts. Expansions were alsofollowed by a variety of child verbalizations,from providing labelS to asking questions. Anexample of a sequence from a traditional textinvolving children's descriptions, maternalexpansion, and children's questions follows.

9. C: It's funny lookin'.M: Yeah, it has a huge mouth.C: How many teeth are there?

Thus, expansions, as well as the other maternalutterances discussed, seem to be importantteaching strategies, to the extent that they elicitchildren's verbalizations and participation injoint reading. When children verbalize aroundlabeling texts, they also identify the labelslater. By extension, mothers' more general useof such strategies is reliably related to moregeneral vocabulary status; thus, mothers whotalk with their children in such a way as toelicit children's talk have children who aremore facile with language.

The sequential analyses also suggest some-thing specific about the ways in which childrentalk after mothers: They frequently repeat thestrategies that mothers used in preceding utter-ancm. Specifically, if one examines the childconsequerces of maternal nonquestioningutterances, it is clear that children repeat moth-ers' utteranc. The reverse, that is, mothers'copying children's preceding utterances, doesnot seem to be occurring. This does not,however, suggest that joint book reading is aunidirectional process of mothers socializingchildren. Mothers' language is also contingent


9

20 A.D. Pellegrini, Lee Ga Ida, Jane Perlmutter & Ithel Jones

to children's. For example, when childrengave incorrect responses to mothers' WH-questions, mothers (at a greater-than-chancelevel) gave the correct answer or told childrenthat the answer was incorrect. Similarly,mothers reinforced children's correct respons-es. In short, mother-child interaction aroundtexts is transactional. Children do not merelycopy maternal strategies; maternal strategiesare responsive to children's performance, andchildren are, in turn, responsive to mothers'strategies.

It is important for future research to exam-ine how these roles vary in other social con-texts, such as with peers at school and withgrandparents and siblings at home. It is proba-bly the case that children take different roles inthese situations than with parents and teachers.For example, there may be more conflict insibling interaction around books than in themother-child situations. Different types ofconflicts may have different implications forchildren's language and literacy development.

Although the present study showed theways in which mothers and their childreninteracted around familiar texts in their homes,much more research in this area is needed.There is urgent need for naturalistic, longitudi-nal research on the language learning andteaching of nonmainstream-culture children.The important work of Heath (1983) andMiller (1982) in this area provides both meth-odological and theoretical guidance for suchventures. Specifically, long-term observationsof children in various literacy events, both athome and in school, are needed. Descriptivedata from such observations should help us

understand the uses of literacy and text indifferent communities. We need to know, forexample, more about the types of texts avail-able and the participation in various literacyevents. The compilation of such descriptivedata is essential to the important work ofeducating all our children.

Author Notes. Additional support for the workreported here was provided by the A. L. MailmanFoundation; the Faculty of Social Science, Universi-ty of Leiden; and the Institute for Behavioral Re-search, University of Georgia. We thank ourcolleagues at NRRC and the University of Leidenfor their comments on earlier versions of this paperpresented at colloquia.

Correspondence concerning this report shouldbe addressed to A. D. Pellegrini, Institute forBehavioral Research, Barrow Hall, University ofGeorgia, Athens, GA 30602.

REFERENCES

Bakeman, R., & Gottman, J. (1986). Observinginteraction. New York: Cambridge UniversityPress.

Barthes, R. (1974). Introduction to S/Z. NewYork: Hill and Wang.

Bronfenbrenner, U. (1979). The ecology of humandevelopment. Cambridge, MA: Harvard Uni-versity Press.

Bus, A., & van Ijr.endoorn, M. (1988). Mother-child interaction, attachment, and emergentliteracy. Child Development, 59, 1262-1272.

Caldwell, B., & Bradley, R. (1984). HOME Ob-servation for Measurement of the Environment.Unpublished manuscript, University of Arkan-sas, Little Rock.


0 S


Carden, C. (1965). Environmental assistance tothe child's acquisition of grammar. Unpub-

lisheddoctoraldissertation,Harvard University,Cambridge, MA.

Cochran-Smith, M. (1984). The making of areader. Norwood, NJ: Abkx.

Cornell, E., Senechal. M., & Broda, L. (1988).

ROCall of picture books by 3-year-old children:Testing and repetition effects in joint readingactivities. Journal of Educational Psychology.80, 537-542.

Dickinson, D., & Smith, M. (n.d.) Long-termeffects of preschool teachers' book readings onlow-income children's vocabulary, story com-prehension, and print skills. Unpublished

manuscript.Dunn. I., & Dunn, L. (1981). Peabody picture

vocabulary test (rev. ed.). Circle Pines, MN:American Guidance Service.

Dunn, J. (1988), Social understanding. Cam-

bridge, MA: Harvard University Press.Fader, D., & McNeil, E. (1966). Hooked on

books. New York: Berkeley.Galda, L., Stahl, S., & Pellegrini, A. D. (1993).

Language and literacy in a first grade-class-room: Annual report to the NRRC. Athens,GA: Universities of Georgia and Maryland,National Reading Research Center.

Heath, S. (1982). What no bedtime story means.Language in Society, I I , 49-76.

Heath, S. (1983). Ways with words. New York:Cambndge University Press.

Laboratory of Comparative Human Cognition.(1983). Culture and cognitive development. InW. Kessen (Ed.), Handbook of child psycholo-

gy: History, theory, and methods (pp. 295-356). New York: Wiley.

Mehan, H. (1979). Learning lessons. Cambridge,

MA: Harvard University Press.

Miller, P. (1982). Amy, Wendy, and Beth: Leant-ing language in south Baltimore. Austin, TX:University of Texu Press.

Ninio, A. (1983). Joint book reading as a multiplevocabulary acquisition device. DevelopmentalPsychology, 19, 445-451.

Ninio, A., & Bruner, J. (1978)., The achievementand antecedents of labeling. Journal of OtildLanguage, 5, 1-15.

Ogbu, J. (1981). Origins of human competence:A cultural-ecological perspective. Child Devel-

opment, 52, 413-429.Olson, D. (1983). "See!. Jumping! Some oral

language antecedents of literacy. In H. Goel-man & F. Smith (Eds.), Awakening to literary(pp. 186-192). Portsmouth, NH: Heinemann.

Pellegrini, A. D., Brody, G., & Sigel, I. (1985).Parents' book-reading habits with their chil-dren. Journal of Educational Psychology, 77,332-340.

Pellegrini, A. D., Perlmutter, J., Galda, L., &Brody, G. (1990). Joint reading between blackHead Start children and their mothers. ChildDevelopment, 61, 443-453.

Piaget, J. (1983). Piaget's theory. In W. Kessen(Ed.), Handbook of child psychology: History,theory, and methods (pp. 103-128). New

York: Wiley.Snow, C., & Goldfield, B. (1983). Turn the page

please: Situation-specific language acquisition.Journal of Child Language, 10, 551-570.

Sorsby, A., & Martlew, M. (1991). Representa-tional demands in mothers' talk to preschoolchildren in two contexts: Picture book readingand a modeling task. Journal of Child Lan-guage, 18, 373-395.

Stahl, S., & Fairbanks, M. (1986). The effects ofvocabulary instruction. Review of EducationalResearch, 56, 72-110.

NATIONAL READING RESEARCH CENTER, READING RESEARCH REPORT NO. 1 3

2 9

22 A.D. Pellegrini, Lee Galda, Jane Perlmutter & Ithel Jones

Torrance, N., & Olson, D. (1984). Oral languagecompetence and the acquisition of literacy. In

A. Pelkgrini & T. Yawkey (Eds.), The dewl-opment of oral and wriuen language Ps socialcontexts (pp. 167-182). Norwood, NJ: Ablex,

Tulkin, S. (1972). An analysis of the conclept ofcultural deprivation. Developmemal Psycholo-gy, 6, 326-339.

Valdez-Meachaca, M., & Whitehurst, G., (1992).Accelerating language development throughpicture book reading. Developmental Psycholo-gy, 28, 1106-1114.

Vygotsky, L. (1978). Mind in society. Cam-bridge, MA: Harvard University Press.

Wachs, T. (1985, April). Measurement of environ-ment in the study of organism-environmentinteraction. Paper presented at the biennialmeeting of the Society for Research in ChildDovelopMent, Toronto.

Watson, R. (1990). Literate discourse and cogni-tive .organization. Applied Psycholinguistics,10, 221-236.

Werner, H., & Kaplan, B. (1952). The acquisitionof word meanings. Monographs of the Societyfor Research in Child Detelopment, 15 (1,Serial No. 51).

Whitehurst, G., Falco, F., Lonigan, C., Fische!.J., DeBaryshe, B., Valdez-Menchaca, M., &Caulfield, M. (1988). Accelerating languagedevelopment through picture-book reading.Developmental Psychology, 24, 552-558.

APPENTIIX

Mothers' Utterance;

I. Answers own questions: Answers ownquestion where child gives no immediately preced-ing response or an inadequate response (e.g., That'san elephant.).

2. Describes actions: Describes an action in thetext (e.g.. Look at that guy run.).

3. Clarify: Any clarification of a precedingutterance (e.g., I asked you to tell me that name.).

4. Conflict/disagree: Where there is disagree-ment about behavior (e.g., I told you to sit.).

S. Corrects: Specific feedback on children'sinadequate response (e.g., That's a zebra [not ahorsel.).

6. Asks for description: Asks to describe partor whole of a text item (e.g., What's the color?).

7. Asks for a label: Asks for a general name orlabel (e.g., What's that called?).

8. Asks to read or spell: Explicitly asks child toread or spell (e.g., Now how do you spell that?).

9. Asks about a subordinate: Subordinates aredefined as particular members of a class; for in-stance, apples are a member of the classfood. Askslocation, an example, or a clarification of a subor-dinate (e.g., Where do zebras live?).

10. Asks about a superordinate: Superordinatesare defined as classifiers that are above the subordi-nate level and below the basic level; for instance,fruit is subordinate to apple because it represents alarger class than apple, but a smaller class thanfood. Asks for the location, an example, or a clarifi-cation of superordinate (e.g. Is that like thehammer we have?).

I 1. Describes: Describes action, attribute,funcron (Look, he's sliding.).

12. Gives superordinate label: Provides superor-dinate label (e.g.. Those are clothes.).


j 9


13. Gives subordinate label: Provides specificlabel (e.g., That's a scarf.).

14. Expands child's utterance: Extends what thechild aaid in the preceding utterance (e.g., And it'sblue too.).

15. Linguistic question: A questioe containing aterm denoting a linguistic process, such as say, talk,read (e.g., Can he really talk on the phone?).

16. Linguistic description: An utterance cootain-ing a term denoting a linguistic process (e.g., Tellme about that.).

17. Narrativizes: A narrative lint is introducedwhen the elements of pretense and temporal motiva-tion are introduced (e.g., The mommy and babycamels are having lunch.).

18. Negative reinforcement: An evaluation ofchildren's performance that contains a negative term(e.g That's wrong.).

19. Orient: Verbal focus and attention statements(e.g., Look here.).

20. Reads present text: Reads text verbatim.21. Reinforce/repeat: Either repeats verbatim the

child's preceding utterance or evaluates it without anegative term (e.g., Right, that's the tape record-er.).

22. Slot/frame provided: Provide an opening ora slot m an utterance for child to fill in (e.g., This

is very tired.).23. Text-world: Relate an item in the text to

something external to the text (e.g., That's likeours.).

24. WH-question: Any interrogative with a what,why, where, when component (e.g., Where do they

25. World-text: Relate something external to thetext to the text itself (e.g., Ours is like that.).

26. Yes/no question: An interrogative that can beanswered adequately with a yes or no (e.g., Is thata ruler?).127. Other: Any utterance that doeszi'l fit in the

preceding categories.]

Children's Utterances

1. Acknowledge: Respoods verbally but withoutpropositional content to mother's preceding utter-ance (e.g., Uh-uh).

2. Ask question: Any interrogative, either director indirect, relating to text (e.g., What's that?).

3. Clarify: Provides clarification of his/ber ownpreceding utterances (e.g., That's where we eatlunch.).

4. Cooceptual conflict/disagree: Disagrees withmother's preceding utterances (e.g., No, that's thezoo thing.).

5. Correct answer to subordinate-label question:In response to mother's preceding question about asubordinate category in the text (e.g., Frog).

6. Correct answer to superordinate-label ques-tion: In response to mother's preceding questionabout a superordinate category (e.g., They're zooanimals.).

7. Correct answer to other questions: Anycorrect response to other questions related to text(e. g. , They're black.).

8. Inadequate answer: A response that doesn'tprovide the information requested in the precedingquestion.

9. -Don't know": Where child responds withsome form of "Don't know.'

10. Describes subordinate: Describes any item intext labeled by mother or child ss a subordinate,such as apple, mouse, watermelon (e.g., Mother:Look at that apple. Child: It's red.).

11. Describes superordinate: Describes any itemin text labeled by mother or child as super-ordinate, such as animal, food (e.g., Mother:There's the animal. Child: It's furry.).

12. Expands/extends: Adds to what mother saidin preceding utterance (e.g., Mother: There's yourplayground; Child: And that's the slide.).

13. Initiates/labels subordinate: Child labels anitem in text as subordinate (e.g., Like my doll.).


3 1

24 A.D. PellegrInt, Lee Galda, Jane Perlmutter & Ithel Jones

14. Initiates/labels superordinate: Child labels aniperordinate (e.g., Look at all the animals.).

15. Linguistic: Any utterance with a term refer-ring to linguistic processes (e.g., I can't read.).

16. Narrativizes: Child intnxluces pretense andtemporal organization to the text (e.g., And the littleboy is scared.).

17. Not relevant to preceding utterance or text:Any non sequitur.

18. Reads: Read words in text verbatim.19. Repeats: Repeats verbatim what mother said

in preceding utterance.20. Text-world: Relates item in text to the

extratextual world (e.g., That's like mine.).21, "Want: Use the word 'want" in any way

(e.g., I want you to.).22. World-text: Relates something from the

external world to the text (e.g., Nick has one ofthose.).[23. Other: Any child utterance not fitting into the

above categories.]


f.t

NRRC NationalReading ResearchCenteril8 Aderhold. l'ttiversity of Georgia4thens. Georgia 30602 '1151102 J. M. Patter-sta, Building. University of Marylatui, Colirge Park MD 2042

Documents

ED 371 300 CS 011 729 AUTHOR Pellegrini, A. D.; And Others TITLE · 2014-05-07 · ED 371 300. AUTHOR TITLE. INSTITUTION. SPONS AGENCY. PUB DATE CONTRACT NOTE PUB TYPE. EDRS PRICE