LITERACY AND SCHOOLING EFFECTS ON LANGUAGE AND COGNITION

1
LITERACY AND SCHOOLING EFFECTS ON LANGUAGE AND
COGNITION

• José Morais1
• research in collaboration with
• Régine Kolinsky1, 2
• 1. Research Unit in Cognitive Neurosciences
  (UNESCOG), Université Libre de Bruxelles, Belgium
  
• 2. Fonds National de la Recherche Scientifique,
  Belgium

2
Other colleagues and students involved in old
and recent workon this issueLaboratoire de
Psychologie Expérimentale, ULB, BrusselsPaul
Bertelson Alain Content Jesus Alegria
Philippe Mousty Flavia da SilvaUniversity
of LisbonLuz Cary Carlos Brito
Mendes Arlette VerhaegheUniversity
of Tilburg University of Porto
University of MalagaBeatrice de Gelder
Sao Luis Castro José Antonio
AdrianFederal University of Santa
CatarinaLeonor Scliar-Cabral Loni
Grimm-Cabral Rosemeire MonteiroRosilane
dos Passos Elias Mengarda
Celina MacedoC.N.R.S. (France)Jacques Mehler
Juan Segui
3
Literacy is just an additional system of
information processing, changing in no
significant way the properties and the
organization of the universal mind
4
Literacy promotes a new mode of thought,
characterized by context-independent or abstract
thinking, analytic reasoning, taxonomic
classification of knowledge and a new form of
working memory
5
Literate thinking

• The kinds of analysis involved in syllogism and
  in other forms of logic are clearly dependent
  upon writing (J. Goody I. Watt, 1968, p. 68).
• Cultures with symbolic technologies push
  cognitive growth better, earlier and longer than
  others (P. Greenfield J. Bruner, 1969, p. 654).
• As literacy is mastered, and a new stage of
  social and historical practice is reached, major
  shifts occur in human mental activity radically
  affecting the structure of cognitive processes
  (A.R. Luria, 1976, p. 161).
• The psychologist will realize that the language
  he uses to classify the operations of the mind is
  a literate language superimposed upon primary
  modes of thinking that are not conceptual (E.
  Havelock, 1991, p. 26).
• Literacy is necessary for the forms of
  consciousness found in modern Western thought
  (C.F. Feldman, 1991, p. 47).
• Knowledge about linguistic meanings is a
  plausible candidate for a literate mode of
  thought (D. Olson, 1996, p. 99).

6
This history of the human mind is described by
Merlin Donald (1993, p. 737) as involving three
major cognitive transformations, leading to
three human systems of memory representation
The third transition seems to have started ()
with the invention of the first permanent visual
symbols and it is still under way (p. 739). It
introduced external memory storage and retrieval
and a new working memory architecture (p. 739).
Literacy plays an important but not exclusive
role in this process
7
L. Vygotsky (1978) the basic psychological
processes (abstraction, generalization,
inference) are universal and what the symbol
systems affect is their functional organization
8
S. Scribner and M. Cole (1981),  The Psychology
of Literacy 
9
Effects of literacyby comparing illiterate
adults to "ex-illiterate" adults(ex-illiterate
adults are people who never attended school in
childhood but learned to read and write later on
in special classes)

• Effects of schooling and education by
  comparing ex-illiterates, in other words literate
  but unschooled adults, to literate schooled adults

10
Morais, Cary, Alegria Bertelson (1979)

• Deletion Addition
• (mosagtosa) (osagtmosa)
• Illiterates 19 19
• Ex-illiterates 71 73
• Score of correct responses

11
Number of subjects at the different levels of
performance (possible maximum 10)
12
Read, Zhang, Nie Ding (1986)

• Mean scores averaged over deletion (dougtou) and
  addition (ougtdou)
• Alphabetized Ss 83 C.R.
• Non-alphabetized Ss 21 C.R.

13

• gt what entails phoneme awareness is not literacy
  in general but literacy in a writing system that,
  at some level, represents phonemes, i.e.,
  alphabetic literacy

14
whether sensitive periods exist for culturally
transmitted knowledge systems, such as those
responsible for reading (Blakemore Frith,
2000)
15
source Morais, Content, Bertelson, Cary
Kolinsky, 1988 source Content, Kolinsky,
Morais Bertelson, 1986
16
Detection of common rhyme of correct
responsesIlliterates
67Ex-illiterates
92Subjects with performance gt
75Illiterates 10 out of
21Ex-illiterates 18 out of 20Morais,
Bertelson, Cary Alegria (1986)Also Bertelson,
de Gelder, Tfouni Morais (1989)
17
Deletion of V1 (initial syllable) of correct
responsesIlliterates
55Ex-illiterates
85Subjects with performance gt
75Illiterates 12 out of
21Ex-illiterates 19 out of 20Morais,
Bertelson, Cary Alegria (1986)Also Morais,
Content, Cary, Mehler Segui (1989)
18
Judgement of phonological length of correct
choicesConditionNeutral
73.0Incongruent
45.4Subjects with performance gt
75Neutral 6 out of 10Incongruent
2 out of 10Kolinsky, Cary Morais (1987)
19
Comparison of illiterates and ex-illiterates in
metaphonological (non-phonemic) tests of
correct responses
Illiterates Ex-illiteratesDeletion
of initialsyllabic vowel
55 85
Detection of syllable in sentences 62
81Rhyme judgements
67 93Morais, Bertelson, Cary
Alegria (1986)
20
Distinction between  knowing language  and
 believing about language , or between
linguistic and metalinguistic abilities,
respectively it is actually a distinction
between levels of representation and processing

• Illiterate adults are able to discriminate
  minimal pairs (Adrian, Alegria Morais, 1995
  Scliar-Cabral, Morais, Nepomuceno Kolinsky,
  1997) but unable to manipulate phonemes
  intentionally (Morais, Cary, Alegria Bertelson,
  1979)

21
Discrimination of minimal pairs of correct
responses15 Spanish illiterates
96(Adrian, Alegria Morais, 1995)21
Brazilian illiterates 98(Scliar-Cabral,
Morais, Nepomuceno Kolinsky, 1998)
22
Literacy-independent phenomena in speech
perception1. Categorical identification2.
McGurk effect3. Feature blending4. Unit
migration

• For all these four phenomena, illiterates behaved
  like literate participants (Castro, 1993
  Kolinsky Morais, 1996 Morais, Castro,
  Scliar-Cabral, Kolinsky, Content, 1987 Morais
  Kolinsky, 1994 Morais Mousty, 1992)

23
Categorical identification of consonant
soundsWhen listeners have to identify speech
sounds created with different values along the
acoustic continuum that goes, for example, from
an unvoiced to a voiced stop, their perception
changes abruptly at a given boundary
24
(No Transcript)
25
(No Transcript)
26
McGurk effect This effect is due to the
influence that the visual information about the
movements of the speakers mouth has on the
perception of speech, so that, in an incongruent
situation, an auditory /ga/ together with a
visual /ba/ may lead to the perception of /da/
27
The McGurk Effect

• McGurk McDonald (1976)
• Visual silent  ga  auditory  ba   da  is
  perceived by 98 of adults

28
Feature blending error It is observed in
dichotic listening and consists in combining the
place value of the stimulus delivered to one ear
with the voicing value of the stimulus delivered
to the other ear
29
phonetic feature blendings
30
no difference in phonetic feature blending rates
between Portuguese literate and illiterate adults

• Morais, Castro Kolinsky, 1991
• Morais Castro, Scliar-Cabral, Kolinsky
  Content, 1987

31
Speech unit migration errorAlso observed in
dichotic listening, it consists in reporting a
word illusion that, given certain control
conditions, can only result from the fact that a
unit (e.g., a phoneme) of one stimulus takes the
place of the corresponding unit in the other
stimulus. The migration error provides evidence
of perceptual segmentation into the involved
units
32
Generalized blending situation (Kolinsky,
1992Kolinsky, Morais Cluytens, 1995 Kolinsky
Morais, 1996)
33
Morais Kolinsky, 1994 Castro, Vicente, Morais,
Kolinsky Cluytens, 1995 Kolinsky Morais, in
preparation
34
While conscious representations of phonemes are
acquired under the influence of learning
alphabetic literacy, unconscious perceptual
representations of units that correspond to our
concept of phoneme develop prior to the onset of
literacy
35
Lexical restructuring in the absence of
literacyVentura, Kolinsky, Fernandes, Querido,
Morais(soumis)
36
(No Transcript)
37
The phonological fusion effectThe simultaneous
presentation of  back  in one ear and of
 lack  in the other ear leads very often to
the illusory perception of  black 
38
PHONOLOGICAL FUSIONSCongruent with orthography
CARA-LARA -gt CLARA, or PENA-LENA -gt
PLENAIlliterates 60 Literates
60Incongruent FIZ-LIZ -gt F(E)LIZPAR-LAR -gt
P(E)LARIlliterates 55 Literates
17Morais, Castro, Scliar-Cabral, Kolinsky
Content (1987) Morais, Kolinsky Castro (1991)
39
Dichotic word recognitionSegmental versus
global errorsfor example, for the stimuli CAPA
BOTA(fictitious) errors on the underlined
segmentMAPA BOTA segmentalTAMPA BOTA
globalMorais, Castro, Scliar-Cabral, Kolinsky
Content (1987)
40
Dichotic word recognition of errors of a
particular type among total number of errors
Segmental (C1) Global
(syllable)Illiterates 23.8
22.9Semi-literates
29.1 18.5Literates
32.9
14.2Morais, Castro, Scliar-Cabral, Kolinsky
Content (1987)
41
Morais, Castro, Scliar-Cabral, Kolinsky
Content, 1987
42
Castro Morais (unpublished)
43
Alphabetic literacy does not affect early
phonetic processing, but it may have an effect on
later processing, either by allowing
orthographic knowledge to influence the
integration of phonemic sequences or by
contributing to the deployment of a phonemic
attentional strategy
44
Repetition task of
correct repetitions
Word Pseudo-wordIlliterates
94.4 51.1Literates
100
77.8Morais Mousty (1992)
45
Repetition task of
correct repetitions
Word Pseudo-wordIlliterates
92 33Literates
98 93Castro-Caldas
et al. (1998)
46
The illiterate brain Learning to read changes
the functional organization of the brain
Castro-Caldas et al. 98 (PET)
Brain activity when repeating pseudowords in
literate (top) and illiterate (bottom) Portuguese
women
47
Morais Mousty (1992)
48
Coltheart (1999) within-module interlevels of
representation in Fodor s account
Spoken-language module
Lexical entries
Phonetic processor
Written-language module
Abstract letters
Shapes of letters
49
Autonomy and interaction in and between spoken
and written language systems (Kolinsky, 1998
Morais Kolinsky, 1994)
Late processes
Late processes
Early processes
Early processes
Spoken language perception
Written language perception
50
Visual perception and cognition
51
Kolinsky, Morais Verhaeghe, 1994
E match
C mismatch
52
Illusory conjunctions of form and orientation
Unschooled
SchooledDifference in d between
experimental and control displays 0.88
1.28Estimate ofillusory
conjunctionson the basis offalse detections
9.72
10.92Kolinsky, Morais Verhaeghe (1994)
53
Recognition of incomplete figures(test of Pillon
Lhermitte)Mean step for recognitionIlliter
ates 4.42Ex-illiterates
4.56University subjects
3.52Verhaeghe Kolinsky, in preparation
54
Verhaeghe Kolinsky, in preparation
55
Overlapping figures
56
Recognition of overlapping figures(task
yes/no subjects illiterates) of correct
responsesRealistic figures
95.3Geometric figures
96.7Verhaeghe Kolinsky, in preparation
57
Part verification
58
Kolinsky, Morais Brito-Mendes, 1990 Kolinsky,
Morais, Content Cary, 1987
59
Unschooled illiterates and ex-illiterates do not
differ from more educated people at the level of
perceptual, non-explicit processes of feature
extraction But schooling experience
contributes, in some circumstances, to visual
recognition, and, much more dramatically, it
allows the conscious analysis of visual percepts
60
Lhypothèse dite de  neural recycling 
(Stanislas Dehaene) et son corollaire permettant
de prédire de meilleures performances chez les
illettrés dans certaines tâches visuelles (cf.
Gauthier Curby, 2005)

• linvariance dobjet à travers différentes
  orientations?
• rotation mentale?
• perception des visages et de leurs expressions?

61
Memory
62
Donalds (1993, p. 745) assumptionthe role of
biological working memory has been changed by the
heavy use of external memory

• Gathercole Pickerings (2000, p. 378)
  interpretation of dataworking memory measures
  are impervious to environmental indicators such
  as socioeconomic status

63
Illiterates displayed a rhyme effect in the
immediate ordered recall of a series of pictures,
thus showing that they spontaneously use
phonological codes (Morais, Bertelson, Cary
Alegria, 1986)
64
Kolinsky, da Silva, dos Passos, Grimm-Cabral
Morais, in preparation
65
Kolinsky, da Silva, dos Passos, Grimm-Cabral
Morais, in preparation
66
Kolinsky, da Silva, dos Passos, Grimm-Cabral
Morais, in preparation
67
Kolinsky, da Silva, dos Passos, Grimm-Cabral
Morais, in preparation
68
interaction between the nature of the material
used in the span task and schooling

• The superiority of the schooled people over the
  unschooled was larger in the digit span task than
  in the Corsi's blocks task, thus larger with
  verbal than with visuo-spatial material.

69
Kolinsky, da Silva, dos Passos, Grimm-Cabral
Morais, in preparation
70
Kolinsky, da Silva, dos Passos, Grimm-Cabral
Morais, in preparation
71
Kolinsky, da Silva, dos Passos, Grimm-Cabral
Morais, in preparation
72
Kolinsky, da Silva, dos Passos, Grimm-Cabral
Morais, in preparation
73
Kolinsky, da Silva, dos Passos, Grimm-Cabral
Morais, in preparation
74
Strong effect of schooling, but not of literacy,
on short-term and working memory performance,
especially for verbal material
75
Luria (1978, p. 18) illiterate people are
unable to group objects or even to pick out
their abstract features according to abstract
semantic categories. Illiterates resorted to
"concrete, situational thinking" and remained
unconvinced by "every attempt to suggest the
possibility of categorical grouping" (p. 77)

• Scribner and Cole (1981, p. 124), looking for
  literacy effects independent of those of
  schooling, found results that "discourage
  conclusions about a strong influence of literacy
  on categorization and abstraction"

76
classification requiring Ss to match a target
with either a taxonomically- or a
functionally-related item. Ex LEG - arm- pants
77
Morais, Scliar-Cabral, Monteiro Kolinsky, in
preparation
78
Morais, Scliar-Cabral, Monteiro Kolinsky, in
preparation
79
classification requiring Ss to match a target
with a taxonomically related or an unrelated
itemEx DUCK- sheep- sparrow
80
Morais, Scliar-Cabral, Monteiro Kolinsky, in
preparation
81
Morais, Scliar-Cabral, Monteiro Kolinsky, in
preparation
82
Morais, Scliar-Cabral, Monteiro Kolinsky, in
preparation
83
Morais, Scliar-Cabral, Monteiro Kolinsky, in
preparation
84
Morais, Scliar-Cabral, Monteiro Kolinsky, in
preparation
85
Fluency tests
Illiter. 4 years 8 years 11 years
Univ.degreeEstimates of clusteringBousfields
ratio of repetitions 0.55 0.54
0.53 0.48 0.54Troyer et
alsclusters size 2.25 2.06
2.18 2.24 2.11Number of
subcategories 4.00 3.83 5.25
5.56 6.33
86
Taxonomic clustering in fluency task

• In our study, we calculated
• the ratio of subcategory repetitions (Bousfield,
  1953)
• R/(n-1), with r intra-category repetition (ex
  tiger-lion) n total number of items recalled.
  This index is thus independent of the absolute
  performance level of recall
• the mean size of consistent subcategory groupings
  (Troyer, Moscovitch Winocur, 1997), beginning
  with the 2nd word (thus a single word 0 two
  words have a cluster size of 1 etc.)
• These measures, indicating taxonomic clustering,
  were similar in all the literacy groups tested

87
Raisonnement
88
Tfouni (1988) 5 out of the 16 illiterates that
she interviewed could understand and explain
syllogisms after displaying the behavior
described by Luria.

• For example, one illiterate was presented with
  the following syllogism "These people only visit
  their friends on Sundays. Today they are visiting
  their friends. What day is today?"
• He first responded "I like to go and visit on
  Saturday night then, the day after, I can sleep
  until late in the morning."
• At this point, the examiner read the syllogism
  again, and the participant said "Only on
  Sundays? Well, we are Sunday. Given what you
  said, I should answer Sunday"

89
Dias, Roazzi Harris (2005)Scribner (1977)
après 2 ou 3 ans de scolarisation, les
participants tendent à adopter une attitude moins
empirique et plus analytique, encodant les
affirmations faites dans les prémisses
indépendamment de leur familiarité  puisque
vous dites cela... ,  si je suis vos
paroles... 

• Dias Harris (1988, 1990) quand des enfants de
  2 à 6 ans sont invités à prendre les prémisses
  comme se référant à une autre planète, on obtient
  le même changement dattitude

90
Dias, Roazzi Harris (2005)2 groupes non
scolarisés et instruction de base
Contenu inconnu Contenu
contraire Standard
Planète Standard PlanèteR.C.
2.38 3.87 1.33
3.79Just. Anal. 1.00
3.26 0 3.33
Just. Empir. 0.96 2.25
5.33 2.29
91
Lhabileté à raisonner à partir de prémisses
non-familières est souvent masquée par un biais
empiriqueLa référence à une autre planète a
éliminé leffet du contenu

• En conclusion lhabileté à raisonner sur la base
  de prémisses nest pas créée par la
  scolarisation, elle précède celle-ci

92
PM 47 (Raven)Problems correctly
solvedSeries(maximum 12) Illiterates
Ex-illiteratesA
8.4 8.3Ab
6.6
7.0B
4.9 5.6Total
19.9
20.9Verhaeghe Kolinsky (2005)
93
Wason's (1966) selection taskInvolves a
conditional rule that is conventionally expressed
as "if P then Q". Evaluation of the truth or
falsity of the rule requires the subject to
choose both visible P and not-Q.
94
Participants were shown four cards with, on the
visible face, for example, Pelé, Tyson, a soccer
ball and a boxing-glove. They were told that,
when the visible face showed a famous player, on
the invisible face there was a sporting good, and
vice-versa. They were asked to point the cards
to be turned over in order to check, for example,
the statement "If there is Pelé on one side of
the card, then there is a soccer ball on the
other side".

• In the experimental groups, after the first
  problem, if the answer was wrong, the subjects
  were explained why it was wrong and the reasons
  to pick up the correct cards. In the control
  groups, there was no such training.Morais,
  Mengarda, Grimm-Cabral Kolinsky, in preparation

95
Morais, Mengarda, Grimm-Cabral Kolinsky, in
preparation
96
Fonctions exécutives (processus de contrôle)
97
Stroop test types of trials Task tell the
number of elements Ex of expected response
 three 
98
Stroop- results
99
Classification semantic vs. perceptual features
1) mammals X vegetables 2) edibles X edibles 3)
big squares X small squares 4) green on left X
green on right 5) horizontal line X tilted line
100
(No Transcript)
101
semantic vs. perceptual classification results
unschooled quasi-unschooled Ss make (find) less
classifications than schooled Ss but they do not
differ in semantic vs perceptual classifications
and can shift from one type of dimension to the
other
102
Tower of London planning ability ( inhibition)
103
Tower of London- results
Effects of Group and Item type, no interaction
104
Tower of London- results
Item type Group Interaction
105
Tower of London- results
Only effect of Item type
106
Executive processes,more specifically on
inhibition of irrelevant responses, on shifting
between criteria, and on planning ability, show
little if any effect of either literacy or
schooling.
107
Conclusions
108
The acquisition of alphabetic literacy
influences metaphonological development, being
responsible for the establishment of phoneme
awareness. It also stimulates some spoken
language processing strategies, but nothing else.
109
Literacy does significantly change neither the
pre-existent modular systems involved in speech
and visual perception nor the basic non-modular
cognitive processes that are involved in
attending to and categorizing information in
abstract ways, selecting and inhibiting
information, planning actions, etc.