1
From context to text
LESLLA learners between situated learning and logic reasoning
Jeanne Kurvers
13th Nordic Conference
Helsingør, 6-4-2017
Overview
• Literaturereview– Literacyandmetalinguis6cawareness– Literacyandprocessingoflinguis6cinforma6on
• Literacyandcogni6on– Taxonomicclassifica6on– Syllogis6creasoning
Afewobserva6ons• Reading
– CouldyoureadthisleCertome?– Hassanstoothache– Fatmagoestoschool
• Wri6ng:– Abookofmylife– Emergentwriters
• Outofschoollearning– Somaliinterpreter
Emergent writers
aosevtok Fa6ma
Jahmila
Kwaku
Literaturereview
• Longlas6ngresearch– Doesliteracyimpactmetalinguis6cawareness?– Doesliteracyimpactlogic/deduc6vereasoning?àToday’stopic
• Morerecentresearch– Doesliteracyimpacttheprocessingof(linguis6c)informa6on
Literacyandmetalinguis6cawareness
• Phonologicalawareness– Sounds:(howmanysoundsincat?)– syllables
• Wordawareness– Whatisaword?– Whatisthelastwordyouheard?– HowmanywordsinJohntakesthetrain?
• Printawareness– Streetsigns,leCers,register
2
Resultsmetalinguis6ctasks
• Onnearlyalltasks:• non-literatesdifferstronglyfromreaders
– Excep6on:rhymerecogni6onandsegmenta6oninsyllables
Segmenta6onsentencesandwords
• Couldyousegmentintopieces(orally)
– IcomefromthesouthofSomalia
– Theoldman
– Intheshop– Tomatoes
Examplesnon-literates• IcomefromthesouthofSomalia
– Icome/fromsouthSomalia
– Youhavethesouthandthenorth,isthatit?
• Theoldman
– Noyoucan't
– Doyoumeanoldmenandyoungmen?
• Intheshop
– No,thatisoneplace
• Tomatoes
– Everyoneatomato
– Intofourparts
– To/ma/toes
Could someone write this?
• I live in Holland
• Outside
• I was raining yesterday
• Ten
• A baby is very old
Examplesnon-literates
• Yes,becauseIdoliveinHolland.• Youcouldwrite‘tree’butnot‘outside’• Ten,yes,thatcanbewriCen• No,becauseitwasnotrainingyesterday• Ifitwasrainingyesterday,youcouldwritethatdown.
• No,ofcoursenot,ababyisnotold.• Youcouldwriteitdown,butitiss6llnonsense
Impactonlanguageprocessing
• Repeattable,repeathable(wordandpseudo-word)n• Verbalfluency
– Men6onasmuchwordsasyoucanwithap– Men6onasmuchanimals/foodasyoucan
• Workingmemory:repeatstringsofdigitsorwords
• Results:– no(big)differencesbetweennon-literatesandliteratesinusing
seman6cinforma6on,– butbigdifferencesinprocessingphonologicalinforma6on
3
LanguageProcessing:wordrepe66on(Reis&Castro-Caldas,1997)
% correct Words Pseudo words
Non-literate
92% 33%
Literate 98% 99%
Literacyandcogni6veopera6ons
• ClaimVygotskyandLuria:literacychangeswaysof(deduc6ve)reasoning
• Studiesrevealdifferentoutcomes• However:simplesyllogis6creasoningtasksrevealintriguingconsistentresults.
(Luria,1975;Scribner,1977;Scribner&Cole,1981;Kurvers,2002;Diasetal.,2005;Haan,2007;Counihan,2007)
Examplesresearchonreasoning(Luria,1975)
Taxonomicclassifica6onTheoddoneout- glass,pan,glasses,boCle- rifle,bowandarrow,gun,bird- Saw,hammer,log,axe
Syllogisms:AllbearsonNovaZembla,farupintheNorth,arewhite.Lastyear,mycousinsawabearonNovaZemblaWhatwasthecolourofthebear?
Syllogis6creasoning:AllXareY
Premise-based Non-literates Literates
Luria(1930) 22.5% 100%
Scribner(1997) 22.3% 75%
Scribner&Cole(1981)* 27% 29%/50%
Kurvers(2002) 20% 68%
Haan(2007)Counihan(2007)
14%30%
-66%
* Firstexperiment,percentagesliterates:resp.Vai-literatesandschooledliterates
StudyKurvers(2002)• Comparisonofthreegroups:
– preliteratechildren,non-literateadults,low-educatedliterateadults
• Tasks:– metalinguis6ctasksandcogni6vetasks
• Ques6on:Impactofliteracyorsomethingelse?– Ifchildrendifferfromadults(irrespec6veofliteracyexperience)ànoimpactofliteracy
– Ifreadersdifferfromnon-readers(irrespec6veofage)àimpactofliteracy
Respondents
• Pre-literatechildren,lasttermKindergarten(N=23)
• Non-literateadults(N=25)• Low-educatedliterateadults,4yearsprimaryschool(N=24)
• All:Berber,Somali,Turks,samebackgrounds;adultsecondlanguagelearnersofDutch
4
Cogni6vetasks
• RavenSPM:nonverbalIQtest,culturefree(adultsonly)(k=42)
• Taxonomicclassifica6on(k=8)• Simplesyllogisms(k=5)
• TasksconductedinL1(Berber,Turkish,Somali)unlessrespondentpreferredL2.
Cogniton: Raven SPM
ResultsRavenSPM(max=42) Classification
Examplesclassifica6on1
• MostLiterates– Picture,becausetheotherthreeareforreading
• Non-Literates– Picture,becausethatisonthewall– Newspaper,becauseyoucanthrowitawaywhenyouhavefinishedreading.
– LeCer,becausethatcomesthroughthepostbox– Photo,becauseyouneedasonfortheotherthree
Exampleclassifica6on2
• Mostliterates– Fish
• Non-literates– Dog,becausewedonoteatdogs– Rabbit,becausenotusefulforpeople– Rabbit,doesnotliveintheNetherlands– Dog,becauseadogisallowedinthelivingroom– Fish,becausetheothersdonotliveinthewater
5
Analysisclassifica6on
• Example:Saw, hammer, log, axe• Taxonomic
– (wood)log,becauseotherthreearetools• Situa6onal/func6onal
– Youalsoneedthewood,becauseotherwisethereisnothingtosaworhammer
• Idiosyncra6c– Thesaw,becauseyoucannotsawwiththeotherthree
Resultsclassifica6on
Child Non-literate Literate
taxonomic 38% 55% 77%
situa6onal 19% 26% 16%
idiosyncra6c 43% 19% 7%
Boxplotsclassifica6ontaxonomic Examplessyllogism
• AllwomeninMarkeyaremarried• Fatmaisnotmarried• DoesFatmaliveinMarkey?
• Allstonesonthemoonareblue• Amanwenttothemoonandfoundastone.
• Whatwasthecolourofthatstone?
ExamplesreasoningSyllogismtask
• Does Fatma live in Markey? • Most literates:
– No, because all women are married there
• Non-Literates: – No, because I know Fatma. She lives here. – How should I know, I have never been there. – We have to ask Fatma. – It can not be that there is a country where all
women are married. – Should I give my opinion, or react on your words?
ExamplesreasoningSyllogismtask
• What was the colour of that stone? • Most literates:
– Blue, because all stones are blue there
• Non-Literates: – Black, because it is very hot there – How should I know, I have never been there – There are no stones on the moon – Brown, just look outside. – I think blue, because the sky is blue. – Black or white, that depends
6
Typesofarguments• Premisebased
– Becauseallstonesonthemoonareblue– Becauseotherwiseheshouldhavehadthreeheads– Becauseyoutoldmeallstonesareblue– Ifshelivedthere,shewasmarried
• Experiencebased– BecauseIhavebeeninAmsterdam– IknowFatma,sheismarried– Becauseofthecolorofthewater– Lookoutside,allstonesarebrown– WehavetoaskFatma– Peopletoldmeitisanicecity
TypesofargumentsCtn.• Discussionpremise(alsoexperiencebased)
– Itcannotbethatthereisacountrywhereallwomenaremarried
– Therearenostonesonthemoon– Ahumanpersoncannothavethreeheads
• Don’tknow/noargument– HowwouldIknow?– Youdidn’ttellme.
Frequencies arguments by group
Child Non-literate Literate
Premise based 33% 19% 67%
Experience based
39% 75% 27%
No argument 28% 6% 6%
PearsonCorrela6ons
Classifi-ca>on
Raven Printawareness
Meta-linguis>c
L1reading
syllogisms .26 .39* .57** .77** .59**
**p<.000,*p<.05
Differencesbetween5items?
1. AllCi6esinHollandarenice2. AllwomeninMarkeyaremarried3. Allstonesonthemoonareblue4. AllpeopleonMarshavethreeheads5. Achmedwentforawalk(allstonesinthe
riverwereyellow)– Storyembeddedsyllogism
MeansbygroupItem1-5
7
Conclusionssyllogism1-5• Differencesbetweenadultnon-literatesandliteratesstrongestforthemoon(syllogism3)andtheriversyllogism/story(syllogism5)
• Childrenseemtoprofitfromstory-embedding(syllogism5),non-literateadultsdonot
• Reasoningnon-literatesinstoryembeddedsyllogismsimilar– White,becauseGodmadethestoneswhite– Idon’tknow,Ihaveneverbeenthere– Blue,becausethewatermadethemblue– Itmustbeabeau6fulcolour.Wedon’tknowthecolour
Acloserlookatreasoning
• Isittheques6on?• Isittheverbalaspect?
– Boxtask(Haan2007)• Isitthemeaningoftheconcept“all”?
– Brothers(Haan2007)• Somein-betweenanswers
Scribner&Cole(1981)
• E:AllKpellemenarericefarmers.MrSmithisnotaricefarmer.IsheaKpellemen?
• S:Idon’tknowthemaninperson.Ihavenotlaideyesonthemanhimself.
• E:Justthinkaboutthestatement.• S:IfIknowhiminperson,Icananswerthatques6on,butsinceIdonotknowhiminpersonIcannotanswerthatques6on.
Kurvers,2002
• E:Listen(repeatssyllogisme).DoesFatmaliveinMarkey?
• Arkem:FatmalivesinMarkey,orinTurkey(laughs).Fatmaisnotmarried,hè?Allwomenaremarried,sheisnot.Butwhyisshenotmarried?
• E:DoessheliveinMarkey,youthink?• Arkem:Idon’tknow.Shemightlivethere,orhere.
Verbalaspect:Boxtask(Haan,2007)
• Threeredboxesinatray.• Eachboxcontainsaping-pongball(show).• Closeallthreeboxes,hidethethreeboxes,showoneoftheredboxesagain.
• “Whatisinthisbox?”
• Correct:69%• Respondentscandeduceinforma6onfromthe‘premises’iftheinforma6onispresentedvisually.
Conceptall?(Haan,2007)• Simplifiedsyllogism
– Ihavethreebrothers.AllthreeofmybrothersliveinRoCerdam.Janisoneofmybrothers.InwhichcitydoesJanlive?
• Correct:25%• Whatisthedifferencewiththebox-task?
8
Examplebrothers’task• Exp:WheredoesJanlive?• Lahcen:[longpause]YoudidnottellmewhereJanlives.YoutoldmethatyourbrothersliveinRoCerdam,butnotwhereJanlives.
• Exp:AllthreeofmybrothersliveinRoCerdam,allthree.Janisoneofmybrothers.WheredoesJanlive?
• Lahcen:ThosethreebrothersofyoursliveinRoCerdam,hemaybeoneofthem.
• Exp:Janisoneofmybrothers.• Lahcen:ThentheyallliveinRoCerdam
In-betweens
• Highscoringnon-literates– Khadizja(1experiencebased,4deduc6ve)– Habiba(1experiencebased,4deduc6ve)– Lionel(2experiencebased,3deduc6ve)
• Onnotaskallin-betweensdodifferfromtheaverageofthewholegroup,exceptformetalinguis6cawarenessandprintawareness
Waysofreasoningin-betweens
• Blue, you told me all stones are blue there • If she lived in Markey, she would have been
married • I think yes, although I have never been there. • No, she is not married. That is not allowed. • Yes blue, all stones are blue there, isn’t it. • Shall I give my opinion, or react on your
words?
Compare• Lahcen:YoudidnottellmewhereJanlives
– Implicitques6on:“DoyourememberwhatItoldyouaboutJan?”
• Khadizja:Blue,becauseyoutoldme– Answertoadifferentques6on:“WheredoesJanlivewhenAandBaretrue?”
• Compareexperienceswithreadingcomprehensioninliteracyclasses
Conclusions• Literacyopensnewwaysofhandlingverbalinforma6on• Defaulthandling:rela6ngverbalstatements(separate
facts;exemplars)successivelyandonebyonetotheimmediate,outsidecontext,thedirectworld;situatedcogni6on,combiningandintegra6ngac6ngandspeaking:contextualverbalreasoning
• Literate(metalinguis6c)handling:rela6ngverbalstatementsfirstofalltoeachother,withinthetext:textualverbalreasoning.
• Theliterate(metalinguis6c)pointofview:integra6ngverbal(textual)informa6onbeforecontextualchecking
• Nextstep,‘symbolic’cogni6on,withwithin(inside,text-bound)trueandfalsevalues:symbolicreasoning,formallogic
References Counihan,M.(2007),Lookingforlogicinallthewrongplaces.aninvesGgaGonoflanguage,literacyandlogicinreasoning.Amsterdam:ILLCCastro-Caldas,A.,&Reis,A.(2003).Theknowledgeoforthographyisarevolu6oninthebrain.ReadingandWriGng:AnInterdisciplinaryJournal,16,81–97.Diasetal.,(2005).ReasoningFromUnfamiliarPremisesAStudyWithUnschooledAdults.Psychologicalscience,16,7,550-554.Haan,G.(2007);HowilliteratesinterpretsyllogisGcproblems.MScthesisinLogic.Amsterdam:UVAKurvers,J.,(2002).MetongeleCerdeogen.Kennisvantaalenschri|vananalfabeten.Amserdam:AksantAcademicPublishers.J.Kurvers,I.vandeCraats&R.vanHout(2015)Footprintsforthefuture:Cogni6on,literacyandsecondlanguagelearningbyadults.In:IvandeCraats,J.Kurvers&RvanHout(Eds.)Adultliteracy,secondlanguageandcogniGon.EditorsNijmegen:CenterforLanguageStudies(CLS)Luria,A.(1975),CogniGvedevelopment:itsculturalandsocialfoundaGons.Cambridge:CambridgeUniversityPress.Ong,W.(1982).OralityandLiteracy:TheTechnologizingoftheWord.NewYork:Methuen.Reis,A.,&Castro-Caldas,A.(1997).Illiteracy:Acauseforbiasedcogni6vedevelopment.JournaloftheInternaGonalNeuropsychologicalSociety,3,444–450.Scribner,S.(1977).”Modesofthinkingandwaysofspeaking”in:JohnsonLaird&Wason(eds),Thinking.ReadingsinCogniGveScience.CambridgeUniversityPress,CambridgeMassachuseCsScribner,S.,&Cole,M.(1981).Thepsychologyofliteracy.Cambridge,MA:HarvardUniversityPress.Vygotsky,L.(1978).Mindinsociety:Thedevelopmentofhigherpsychologicalprocesses.Cambridge,MA:HarvardUniversityPress
9
Thankyou!
Doinbetweensdiffer
• Background(age,schoolingetc)?• Non-verbalintelligence?• Memory?• Theconceptall?• Waysofreasoning?• Whatelse?
Correlations
syllogisms literacy classification metalinguistic raven syllogisms Pearson Correlation 1 ,572** ,263* ,772** ,385*
N 62 62 62 62 34 literacy Pearson Correlation ,572** 1 ,431** ,663** ,583**
N 62 62 62 62 34 classification Pearson Correlation ,263* ,431** 1 ,202 ,202
N 62 62 62 62 34 metalinguistic Pearson Correlation ,772** ,663** ,202 1 ,595**
N 62 62 62 62 34 raven Pearson Correlation ,385* ,583** ,202 ,595** 1
N 34 34 34 34 34
Relevant tasks / tests Literacy and raven, adults only
Syllogisms highest corelation with metalinguistic abilities
Coefficientsa
Model
Unstandardized Coefficients
Standardized Coefficients
t Sig. B Std. Error Beta 1 (Constant) -1,750 1,006 -1,739 ,093
raven -,029 ,031 -,136 -,913 ,369 literacy ,004 ,015 ,052 ,277 ,784 classification -,012 ,237 -,007 -,052 ,959 metalinguistic ,078 ,017 ,826 4,614 ,000
2 (Constant) -1,789 ,668 -2,679 ,012 raven -,028 ,031 -,135 -,929 ,360 literacy ,004 ,014 ,049 ,280 ,781 metalinguistic ,078 ,017 ,825 4,701 ,000
3 (Constant) -1,730 ,625 -2,769 ,009 raven -,026 ,029 -,124 -,900 ,375 metalinguistic ,081 ,013 ,855 6,207 ,000
4 (Constant) -1,903 ,593 -3,210 ,003 metalinguistic ,074 ,010 ,781 7,077 ,000
Predicting syllogisms
Continuum, correlation
Background NL in-betweens Age L2 school textL1
74 25 5 mth 0 yes
76 47 2 mth 0 no
131 37 10 mth 2 Some decoding
Average group 38.8 2.3 0.4 no
10
Background in-betweens classification
Raven Sentence repetition
objectivation
Picture story: coherence
Picture story questions
74 4 14 8 2 3 3
76 4 20 8 2 4 2
131 3 31 2 0 3 3
Average group
4.3 14.9 6.2 0.5 2.2 1.4
Examplebrothers’task
• Exp:CanyourememberwhatItoldyouaboutmybrothers?
• Zina:Youtoldmeyouhavethreebrothersandonesister,andtheplaceyoumen6onedIdonnotknow/Icannotremember.[..]
• Exp:AllthreeofmybrothersliveinRoCerdam.Janisoneofmybrothers.InwhichcitydoesJanlive?
• Zina:Idon’tknow.
Boxplotssyllogisms,correctpremise-basedanswers Predic6ngsyllogis6creasoning
Model
Standardized Coefficients
t Sig. Beta 1 (Constant) -1,739 ,093
Raven -,136 -,913 ,369 Print awareness ,052 ,277 ,784 Classification -,007 -,052 ,959 Metalinguistic ,826 4,614 ,000
2 (Constant) -2,679 ,012 Raven -,135 -,929 ,360 Print awareness ,049 ,280 ,781 Metalinguistic ,825 4,701 ,000
3 (Constant) -2,769 ,009 Raven -,124 -,900 ,375 Metalinguistic ,855 6,207 ,000
4 (Constant) -3,210 ,003 Metalinguistic ,781 7,077 ,000