Foundation literacy acquisition in European orthographies · Foundation literacy acquisition in...

Foundation literacy acquisition inEuropean orthographies

Philip H K Seymour1 Mikko Aro2 and Jane M Erskine1 2

in collaboration with COST Action A8 networksup2

1University of Dundee Scotland UK2University of Jyvaskyla Finland

Several previous studies have suggested that basic decoding skills may develop lesseffectively in English than in some other European orthographies The origins of thiseffect in the early (foundation) phase of reading acquisition are investigated throughassessments of letter knowledge familiar word reading and simple nonword reading inEnglish and 12 other orthographies The results conrm that children from a majority ofEuropean countries become accurate and uent in foundation level reading before theend of the rst school year There are some exceptions notably in French PortugueseDanish and particularly in English The effects appear not to be attributable todifferences in age of starting or letter knowledge It is argued that fundamental linguisticdifferences in syllabic complexity and orthographic depth are responsible Syllabiccomplexity selectively affects decoding whereas orthographic depth affects both wordreading and nonword reading The rate of development in English is more than twice asslow as in the shallow orthographies It is hypothesized that the deeper orthographiesinduce the implementation of a dual (logographic + alphabetic) foundation which takesmore than twice as long to establish as the single foundation required for the learning ofa shallow orthography

There has been much recent attention to the possibility that the ease of readingacquisition may vary between languages because of differences in lsquoorthographicdepthrsquo (Frost Katz amp Bentin 1987) Much of this attention has focused on a contrast

143

British Journal of Psychology (2003) 94 143ndash174copy 2003 The British Psychological Society

wwwbpsorguk

Requests for reprints may be addressed to Philip H K Seymour Department of Psychology University of Dundee DundeeDD1 4HN UK (e-mail phkseymourdundeeacuk)sup2 National representatives of the EC COST Action A8 especially H Wimmer University of Salzburg Austria J LeybaertUniversite Libre de Bruxelles Belgium C Elbro University of Copenhagen Denmark H Lyytinen University of JyvaskylaFinland J-E Gombert University of Brittany and M T le Normand HoAtildepital de la SalpeAtilde triere France W Schneider Universityof Wurzburg Germany C Porpodas University of Patras Greece H Ragnarsdottir University of Rekjavik Iceland P Tressoldiand C Vio University of Padova Italy A de Groot University of Amsterdam and R Licht Free University NetherlandsF-E Tugravennessen Stavanger University College Norway S L Castro University of Porto and L Cary University of LisbonPortugal S De or and F Martos University of Granada Spain and Aordm Olofsson University of Umeaordm Sweden

between English which is regarded as a deep orthography containing manyinconsistencies and complexities and other alphabetic European languages severalof which have shallow orthographies with consistent graphemendashphoneme correspon-dences For example Wimmer and Goswami (1994) compared reading of digits numbernames and nonwords formed by exchanging the onsets and rimes of number names by7- 8- and 9-year-old children in German and English Nonword reading was signicantlyslower and more error prone in English at all three age levels Frith Wimmer andLanderl (1998) used structurally equivalent sets of 1- 2- and 3-syllable nonwords inEnglish and German and again found consistently poorer nonword reading in EnglishSimilar data are reported for comparisons of English with Spanish and French byGoswami Gombert and de Barrera (1998) and with Greek by Goswami Porpodasand Wheelwright (1997) These studies suggest that the decoding process which iscommonly assigned a central role in theoretical accounts of reading acquisition (Ehri1992 Gough amp Hillinger 1980 Share 1995) develops more slowly and less effectivelyin English than in other European languages

The present studyextends this work to a comparison of English with a wider range ofEuropean languages and also seeks to determine the stage in reading acquisition atwhich the orthographic depth effect becomes evident The theoretical context isprovided by the foundation literacy framework developed by Seymour (1990 19971999) This proposes that reading is acquired in phases such that basic foundationalcomponents are established in Phase 1 while the complexities of orthographic andmorphographic structure are internalized in Phases 2 and 3 Asimplied version of themodel is shown in Fig 1 The foundation consists of two processes a logographicprocess involved in the identication and storage of familiar words and an alphabeticprocess which supports sequential decoding Both of these processes are held to bedependent on the availability of letter-sound knowledge (Ehri 1992 1997) In theirturn the foundations underpin the development of an orthographic framework inwhich the full complexity of the spelling system is represented in an abstract general-izable format (Plaut McClelland Seidenberg amp Patterson 1996) Reading acquisition is

144 Philip H K Seymour et al

Figure 1 Schematic representation of the dual foundation model of orthographic development (fromDuncan amp Seymour 2000)

paralleled by developments in linguistic awareness It is supposed that small units(phonemes) are emphasized during Phase 1 and larger units (rimes syllablesmorphemes) during Phases 2 and 3

Seymour and Evans (1999) devised simple procedures for the measurement of thecomponents of foundation literacy These were (1) giving sounds for individual lettersand writing letters in response to their sounds (letter knowledge) (2) reading veryfamiliar words (logographic foundation) and (3) reading aloud and writing to dictationsimple CVC nonwords (alphabetic foundation) The procedures were applied to unse-lected groups of Scottish children in the Primary1 year (age 5 years) and in the Primary2year (age 6 years) Letter-sound knowledge was found to be complete before the end ofthe rst school year Development of the logographic and alphabetic foundations wasclosely linked to reading age and advanced towards an asymptote at about 7 years ofreading age Afurther delay of one or more years was found among samples of childrendrawn from disadvantaged (low SES) areas (Duncan amp Seymour 2000)

The simple model set out in Fig 1 suggests two hypotheses regarding the relation-ship between foundation literacy acquisition and the orthographic complexity of thelanguage which is being learned One possibility is that foundation acquisition precedesthe formation of an orthographic framework If so we could expect to nd that thedevelopment of the foundation components occurred in an approximately equivalentway in the different languages and that the effects of orthographic complexity emergedonly later when the orthographic framework was formed Alternatively it could be thatthe presence of orthographic complexity exerts effects from the very beginning oflearning to read In this case foundation literacy acquisition would be expected toadvance more slowly in deep orthographies than in shallow orthographies

Language differencesThe study aims to test for differences in rates of acquisition of the components offoundation literacy in European languages which vary signicantly in orthographiccomplexity This variation has not yet been submitted to a comprehensive computa-tional linguistic analysis (but see Peereman amp Content 1997 1999 Ziegler Jacobs ampStone 1996 Ziegler Stone amp Jacobs 1997 for steps in this direction) Nonethelessthere is agreement that some European orthographies are relatively shallow (FinnishGreek Italian Spanish German) while others are deeper in the sense of containingmore inconsistent correspondences as well as morphological inuences on spelling(Portuguese French Danish English)

The opportunity to carryout the study was provided by an ECnetwork (COSTActionA8) which brought together researchers from 16 European countries who shared aninterest in reading acquisition and dyslexia (Niessen Frith Reitsma amp Ohngren 2000)One part of this cooperation entailed a review of the characteristics of the Europeanorthographies which were thought likely to affect reading acquisition On this basis itwas proposed that it should be possible to classify the orthographies on the twindimensions of (1) syllabic complexity and (2) orthographic depth The rst dimensionrefers principally to the distinction between the Romance languages which have apredominance of open CV syllables with few initial or nal consonant clusters (egItalian Spanish) and the Germanic languages which have numerous closed CVCsyllables and complex consonant clusters in both onset and coda position (egGerman Danish English) The orthographic depth dimension contrasts alphabetic

European orthographies 145

writing systems which approximate a consistent 11 mapping between letters andphonemes (eg Finnish) with those which contain orthographic inconsistencies andcomplexities including multi-letter graphemes context dependent rules irregularitiesand morphological effects (eg French Danish)

Table 1 identies the languages involved in the research and sets out a hypotheticalclassication of the orthographies in terms of the dimensions of syllabic complexity andorthographic depth1 The expectation is that the difculty of acquiring literacy willincrease as one moves from simple to complex syllabic structures and from shallowtowards deep orthographies Hence if linguistic complexity affects the foundationphase of acquisition it is hypothesized that the initial steps in reading will be traversedmore rapidly in languages with simple syllabic structure than in languages with complexsyllabic structure and that acquisition will be slower in deeper orthographies than inshallow orthographies

General methodThe collaborators coordinated the preparation of test materials and the collectionand coding of the data according to guidelines agreed by the COST A8 group Thegroup aimed to make formally equivalent assessments in each of the participatinglanguages of the components of foundation literacy identied by Seymour and Evans(1999)

ListsThere were three sets of lists designed to assess (1) letter knowledge (2) very familiarword identication and (3) decoding of simple nonwords (see Appendix)


1Frost et al (1987 p 104) state lsquoIn a shallow orthography the phonemic and orthographic codes are isomorphic thephonemes of the spoken word are represented by the graphemes in a direct and unequivocal manner In contrast in a deeporthography the relation of spelling to sound is more opaque The same letter may represent different phonemes in differentcontexts moreover different letters may represent the same phonemersquo The COST A8 representatives completed anlsquoorthographicquestionnairersquo in which they estimated the extent of variability in their own languages These estimates were thendebated and formed the basis of the hypothetical classi cation set out in Table 1

Table 1 Hypothetical classication of participating languages relative to the dimensions of syllabiccomplexity (simple complex) and orthographic depth (shallow to deep)

Orthographic depth

Shallow Deep

Simple Finnish GreekItalianSpanish

Portuguese French

Sylla

bic

stru

ctur

e

Complex GermanNorwegianIcelandic

DutchSwedish

Danish English

Letter knowledgeTwo lists of the letters of the alphabet printed in clear lower case font were assembledfor each language The letters were displayed in vertical columns on A4 sheets

Familiar wordsSets of very familiar high frequency words were sampled from the reading materialsused in the early stage of primary schooling in each language Two sets were compiled(1) two lists of content words (principally imageable nouns) and (2) two lists offunction words (grammatical morphemes) Each set was presented as two separate listsof nine items vertically arranged on A4 sheets The items selected for the deeperorthographies were permitted to contain orthographically complex features (diacriticsmulti-letter graphemes irregularities)

Simple nonwordsTwo sets of nonwords were constructed for each language one consisting of mono-syllables and using the structures CV VC and CVC and the other of bisyllables formedfrom the structures VCV CVCV and VCVC Two separate lists were prepared for bothlengths each containing nine items vertically arranged on A4 sheets The nonwordswere formed by sampling dominant and consistent graphemendashphoneme (1 letter 1sound) correspondences in each language and included no complex structures2

ProcedureEach of the three reading tasks was introduced by presentation of six practice items andan explanation For letters children were asked to pronounce the names or sounds(depending which was emphasized by the local educational system) of each item Forfamiliar words they were asked to read out each item on the list In the case ofnonwords they were told that the items were made-up words which they mightnonetheless be able to pronounce

The lists were presented in a xed sequence (letters familiar words nonwords)Testing was carried out individually Vocal responses in reading the complete sequenceof nine items on each list were tape recorded for subsequent checking and categoriza-tion A stop watch was started when the list was presented to the child and stoppedwhen the last item had been attempted If a child was blocked by a particular item he orshe was encouraged to move on and complete the list

ParticipantsFoundation literacy acquisition in English was tested in two groups of children attendingprimary schools in the city of Dundee on the east coast of Scotland Children in Group 1attended a school located in a predominantly middle class (high SES) area of the city Thesample included Primary 1 children aged 5 years and Primary 2 children aged 6 years


2The construction of nonwords for use in cross-language comparisons is problematic in the light of differences in syllablestructure The COST A8 group debated the relative merits of adopting lsquonaturalrsquo but differing structures for each language asopposed to a common set of structures some of which might appear lsquounwordlikersquo for some subsets of languages Eventuallythe use of a common set of patterns was adopted as consonant with the primitive (non-lexicalized) model of decodingpostulated in the foundation literacy framework Further research to determine how far the use of natural versus unnaturalCVstructures inuences the development of decoding at this early phase would be of interest

Group 2 was drawn from a school located in a socially disadvantaged (low SES) area ofthe city Primary 3 children aged 7 years were tested in addition to P1 and P2groups Reading ages were determined for the high SES sample by applying the wordrecognition subtest of the British Abilities Scale (BAS) (Elliot 1987)

Acomparison of the P1 and P2 results conrmed that there was a large advantage inreading accuracy for the high SES group over the low SES group with F(1 116) = 1814p lt 001 for letters F(1 233) = 4305 p lt 001 for familiar content and function wordsand F(1 233) = 1615 p lt 001 for monosyllabic and bisyllabic nonwords Howeverdifferences between the high SES P1 and P2 groups and the low SES P2 and P3 groupswere not signicant These results agree with previous indications that low SES isassociated with a delay in letter-sound acquisition and a consequent lag of about 1 yearin the development of the logographic and alphabetic foundation processes (Duncan ampSeymour 2000)

Given this effect we will restrict the comparison between English and the otherEuropean orthographies to the high SES Scottish group Table 2 provides details of theP1 and P2 sample sizes and the chronological ages and BAS reading ages at the time oftesting Both groups were reading well ahead of age expectation The comparisonbetween English and other orthographies is based on a sample of children who are notheld back by social disadvantage and who are making excellent progress according toUK norms

Table 2 identies the simple syllable and complex syllable European orthographiesand shows the sample sizes and average ages at the time of testing3 Samples wererecruited from effective schools in non-deprived areas in each country The participantswere enrolled in Grade 1 (P1) classes and were tested in the latter part of their rstschool year In addition Grade 2 (P2) groups participated in France and Denmark as wellas in Scotland

Age and genderThe mean ages of the samples varied in line with national differences in the age ofcommencement of formal schooling The Scottish P1 children entered school at5 years and were younger than any other group The groups learning to readcomplex syllable languages were slightly older than the groups learning simplesyllable languages (89 vs 85 months F(1 531) = 5783 p lt 001) Within the simplesyllable set the Finnish group aged 94 months was differentiated from all othergroups 80ndash84 months F(5 247) = 12670 p lt 001 There were also differencesbetween the complex syllable languages with the main group averaging 82ndash91months and the Norwegian children signicantly older at 95 months (F(6 273) = 3803p lt 001)

The relationship between chronological age and performance on the foundationliteracy tasks was determined in a correlational analysis Table 3 gives the values ofPearson r between the ages of all Grade 1 children at the time of testing and the accuracyand time scores for reading letters familiar words and simple nonwords The correla-tions were calculated with and without inclusion of the Scottish P1 sample With the


3There were some instances of incomplete returnsof data Time scores were missing for two children in the Danish P1 sampleand chronological ages for four children in the Finnish sample The most serious problems arose with the French dataset Timescores were omitted for 31 P1 cases and 8 P2 cases and chronological ages were not given for 18 P1 cases and 21 P2 cases Inthe analyses we have used the maximum available sample (see df values in the ANOVAs)

Scottish data excluded the overall indication is that foundation literacy is not stronglyrelated to age Nonword reading was independent of age and the relationships withletter accuracy and word reading speed although signicant are very weakInclusion of the results for the younger Scottish children increases the strength ofthe correlations with age quite substantially especially for performance on the wordreading task


Table 2 Simple syllable and complex syllable European orthographies sample sizes and average ages atthe time of testing

Country Grade N Age (years)Si

mpl

esy

llabl

est

ruct

ures

Shallow

Deep

FinnishGreekItalianSpanishPortugueseFrenchFrench

1111112

70443840384529

787 (28)675 (27)689 (31)684 (27)701 (32)669 (32)794 (56)

Com

plex

sylla

ble

stru

ctur

es

Shallow

Deep

AustrianGermanNorwegianIcelandicSwedishDutchDanishDanishScottish

Scottish

111111121

2

572840372528654030

30

755 (41)740 (34)794 (28)686 (25)748 (59)697 (36)765 (38)855 (34)

CA 559 (32)RA 619 (63)CA 656 (30)RA 722 (84)

Note Participants were enrolled in Grade 1 (P1) classes and were tested in the latter part of their rstschool year Grade 2 (P2) groups also participated in France Denmark and Scotland

Table 3 Pearson r between the ages of all Grade 1 children at the time of testing and the accuracy andtime scores for reading letters familiar words and simple nonwords

Excluding IncludingScottish P1 Scottish P1

Letter accuracy 110 120Letter timeitem 015 plusmn 189Word accuracy plusmn 018 339Word timeitem plusmn 125 plusmn 375Nonword accuracy plusmn 019 289Nonword timeitem plusmn 020 plusmn 261

p lt 05 p lt 01Note Correlations were calculated with and without inclusion of the Scottish P1 sample

There was an equal balance of male and female participants Tests were made of theeffects of gender on each of the foundation literacy measures but no signicantdifferences were obtained (F usually lt 1)

ConclusionsThis preliminary analysis indicates that foundation literacy acquisition by non-EnglishEuropean groups is not affected by gender and is largely independent of variations in theages at which children commence formal schooling

STUDY 1 LETTER-SOUND KNOWLEDGETable 4 reports the accuracy (per cent correct) and speed (sitem) of letteridentication by each language group A three-step analysis of the data was followed(1) results for simple syllable and complex syllable languages (Scottish excluded)were compared in order to determine whether syllabic complexity inuenced letter-sound learning (2) results for the individual languages within the simple andcomplex series were contrasted in order to test for an effect of orthographicdepth (3) the Scottish data were compared with the other languages as a test of aspecial difculty in English

AccuracyAn analysis of variance was conducted to compare languages with simple syllabicstructure against languages with complex syllabic structure (Grade 1 groups only) Thehypothesis that letter learning might be delayed in complex syllable languages was not


Table 4 Accuracy (per cent correct) and speed (seconds per item) of letter identi cation by eachlanguage group (standard deviations in parentheses)

Letter-sound identi cation

sitem

Sim

ple

sylla

ble

stru

ctur

es

Shallow

Deep

FinnishGreekItalianSpanishPortugueseFrench P1French P2

9444 (471)9640 (509)9513 (767)9555 (499)9248 (784)9122 (1083)9803 (432)

148 (055)105 (027)106 (046)103 (044)140 (050)138 (061)126 (018)

Com

plex

sylla

ble

stru

ctur

es Shallow

Deep

AustrianGermanNorwegianIcelandicSwedishDutchDanish P1Danish P2Scottish P1Scottish P2

9685 (332)9951 (179)9865 (246)9573 (533)9874 (175)8951 (466)9492 (730)9795 (242)9397 (603)9603 (468)

074 (016)075 (019)084 (022)102 (024)136 (037)094 (027)114 (051)063 (017)188 (101)103 (027)

supported Accuracy was slightly higher in complex than in simple syllable languages(96vs 94) F(1 553) = 1384 p lt 001

Further analyses were conducted to test for differences between languages withineach series There was a variation between 91 and 96among the simple syllablelanguages F(5 269) = 34 p lt 01 with French and Portuguese at the lower end andSpanish and Greek at the higher end The variation among complex syllable languageswas slightly wider (89ndash99) F(6 273) = 1585 p lt 001 with Dutch picked out as thelanguage with the lowest accuracy level

The outcome for the Scottish P1 group was then compared with the results for allother languages The language effect was signicant F(13 571) = 765 p lt 001 but thiswas not attributable to an outlying result for English Post hoc tests identied fouroverlapping sets of scores The Scottish result (94) was embedded in the main seriesand was better than Dutch French or Portuguese

SpeedTable 4 gives information about the speed of letter identication in each language Thepreliminary analysis (with English excluded) suggested that letter identication wasslower (124 sitem) in simple than in complex syllable (096 sitem) languagesF(1 522) = 5201 p lt 001 This effect is again in the wrong direction with respect tothe syllabic complexity hypothesis

There were inter-language differences within the simple syllable seriesF(5 238) = 864 p lt 001 and within the complex syllable series F(6 273) = 1724p lt 001 Post hoc (Scheffe ) tests identied French Portuguese and Finnish as theslower subgroup in the rst series and Danish and Swedish in the second seriesNaming time per letter was slower in the Scottish P1 group (188 sitem) than in anyother language An analysis of variance indicated that this effect was signicant inrelation to the complex syllable languages where F(7 302) = 2519 p lt 001 and posthoc tests differentiated the Scottish outcome from Swedish (136 sitem)

ConclusionsThere are variations between languages in the efciency of letter-sound acquisitionduring the rst school year However all groups achieve a mastery level of 90orbetter on average and there is no strong support for the conclusion that learning isaffected by the linguistic factors of syllabic complexity or orthographic depth Inparticular the complexity of the English orthography did not appear to havedelayed the acquisition of letter-sound knowledge in the high SES Scottish P1sample

There were differences in speed of letter identication Most groups demonstrated ahigh level of uency quantiable as a rate of about 1 sitem but speeds were slightlyslower in some languages Again the variations were not obviously attributable tocontrasts in syllabic complexity or orthographic depth There was no general associa-tion between uency and chronological age in European groups aged 6 years and aboveHowever naming speed was slower in the Scottish P1 sample than in any otherlanguage This dysuency could be an effect of immaturity in Scottish children whocommence formal learning at an earlier age (5 years) than any of the other languagegroups


STUDY 2 FAMILIAR WORDSTable 5 summarizes the accuracy and time data for reading lists of very familiar wordsOverall the accuracy level was 90(Scottish data excluded) An analysis of varianceindicated that there was no general difference between content and function wordsF(1 1107) lt 1 The distinction affected reading speed with content words somewhatslower than function words (202 sitem vs 17 sitem) F(1 1043) = 679 p lt 01probably due to length differences between the two sets of items Results for thetwo word sets were combined in the subsequent analyses which followed the samethree-step approach as was adopted for the letters

AccuracyA preliminary analysis was carried out on the European data with English excludedRanges within the simple and complex syllable languages were similar (74ndash98forsimple 71ndash98for complex) and the overall difference between the series was notsignicant F(1 553) lt 1

There was a large variation among the simple syllable languages F(5 269) = 2562p lt 001 This effect occurred because accuracy in French and Portuguese (lt 80) wassignicantly lower according to Scheffe post hoc tests than in any of the other languages(gt 90) The variation was also signicant in the complex syllable languagesF(6 273) = 2194 p lt 001 due to the low accuracy in Danish (71)

Word reading accuracy by the Scottish P1 sample (34) fell far below the range ofthe other languages Language differences with Scottish included were highly signi-cant F(13 571) = 4799 p lt 001 According to post hoc Scheffe tests the Scottish P1mean was signicantly below the results for Danish Portuguese and French whichwere in turn differentiated from the outcomes for all remaining languages TheScottish P2 results formed a subset with Danish Portuguese and French (lt 80accuracy)

These data suggest that competence in familiar word reading approaches masterymuch more slowly in English than in other European orthographies At the end of therst school year accuracy is typicallygt 90in the majorityof languages and gt 70in thedeeper orthographies Danish children require an additional year to exceed the 90threshold English-speaking children remain well below this level even after two years oflearning

An alternative way of expressing this trend is to consider the relationship withreading age Within the Scottish sample overall (P1 and P2 combined) there was astrong association between BAS reading age and accuracy of familiar word readingwith r = + 86 p lt 01 The test for a linear trend in the data was highly signicantF(1 59) = 39722 p lt 001 The function relating the two variables had a slope of 249percentage points accuracy for each month of reading age On this basis it waspossible to calculate that the Scottish group crosses a 90threshold at a reading ageof nearly 8 years However the sample included a few children (N= 5) who had veryhigh reading ages (gt 96 months) The presence of these cases reduces the slope ofthe regression line and increases the reading age at which the threshold is crossed Areanalysis with the advanced readers excluded produced the function graphed inFig 2 This shows the scatter of the data and the regression line for word namingaccuracy against reading age The 90threshold is crossed at a reading age of 7 years4 months



Tab

le5

Acc

urac

y(p

erce

ntco

rrec

t)an

dtim

e(s

econ

dspe

rite

m)d

ata

(sta

ndar

dde

viat

ions

inpa

rent

hese

s)fo

rre

adin

glis

tsof

very

fam

iliar

wor

dsw

ithou

tcom

esfo

rco

nten

tfu

nctio

nan

dco

mbi

ned

cont

ent

and

func

tion

wor

dsfo

rea

chla

ngua

gegr

oup

Fam

iliar

wor

dre

adin

g

Con

tent

wor

dsFu

nctio

nw

ords

Com

bine

dco

nten

tfu

nctio

n

s

item

s

item

s

item

Simplesyllablestructures

Shal

low

D

eep

Finn

ish

Gre

ekIt

alia

nSp

anis

hPo

rtug

uese

Fren

chP1

Fren

chP2

981

7(4

09)

965

9(5

90)

948

8(1

754

)94

17

(10

29)

771

9(1

945

)72

47

(27

57)

984

7(4

67)

138

(09

1)1

60(0

63)

218

(11

8)1

40(1

23)

344

(24

4)7

04(1

087

)1

00(0

29)

984

1(3

43)

986

1(2

97)

957

6(1

661

)95

28

(97

1)69

88

(20

77)

856

8(2

087

)10

000

(00

0)

113

(07

2)1

31(0

67)

123

(05

8)1

31(1

00)

300

(18

1)4

19(6

74)

064

(02

1)

982

9(3

76)

976

0(4

75)

953

2(1

697

)94

72

(99

6)73

54

(20

32)

790

7(2

354

)99

23

(24

5)

125

(08

3)1

45(0

66)

170

(10

4)1

35(1

11)

322

(21

4)5

61(9

00)

084

(02

6)

Complexsyllablestructures

Shal

low

D

eep

Aus

tria

nG

erm

anN

orw

egia

nIc

elan

dic

Swed

ish

Dut

chD

anis

hP1

Dan

ish

P2Sc

ottis

hP1

Scot

tish

P2

970

8(4

94)

974

2(4

90)

915

3(1

655

)93

39

(91

5)93

56

(93

1)92

66

(71

0)70

94

(26

78)

922

2(1

388

)32

59

(24

83)

790

7(1

832

)

134

(06

4)1

35(0

74)

213

(13

9)2

36(1

65)

264

(25

7)2

02(0

84)

115

(10

3)1

18(1

40)

819

(60

8)2

11(1

40)

978

6(4

42)

980

2(3

77)

920

8(1

423

)94

74

(85

8)96

67

(48

1)98

21

(50

3)71

20

(29

93)

929

2(1

045

)35

18

(26

23)

737

0(2

134

)

082

(02

9)0

76(0

39)

171

(11

9)1

88(1

21)

139

(12

2)1

17(0

67)

251

(20

0)0

91(0

82)

736

(64

3)2

25(1

59)

974

7(4

68)

977

2(4

34)

918

1(1

534

)94

07

(88

4)95

11

(75

0)95

44

(67

1)71

07

(27

71)

925

7(1

174

)33

89

(24

97)

763

9(1

927

)

108

(05

6)1

06(0

66)

192

(13

0)2

12(1

46)

201

(20

9)1

60(0

87)

185

(14

2)1

05(1

09)

778

(61

9)2

18(1

41)

SpeedTable 5 reports the reading times for the familiar word lists Reading speeds in simpleand complex syllable languages (193 vs 165 sitem) did not differ F(1 522) = 268 nsThere were variations between languages in both series where F(5 238) = 1138p lt 001 for the simple set and F(6 273) = 541 p lt 001 for the complex set Posthoc (Scheffe ) tests differentiated French (561 sitem) from other simple syllablelanguages and placed Portuguese (322 sitem) at the outer edge of the remaininglanguages There were two overlapping sets among the complex syllable languages andno clear instances of slow reading The Danish data (185 sitem) fell well within theboundaries of the remainder of the series

The reading speed of the Scottish P1 group (778 sitem) was substantially slowerthan that of any other group There was a signicant variation across the full set oflanguages F(13 540) = 2063 p lt 001 which was interpretable according to the posthoc tests in terms of a main subset of languages mostly lt 2 sitem a Portuguese andFrench subgroup and Scottish P1 as an isolated outlier Reading speed for the ScottishP2 group (218 sitem) was close to the range of the main set of languages

According to these results uency may be dened operationally as a speed oflt 225 sitem on the word list reading task This is achieved in most languages beforethe end of the rst school year In the Scottish sample word reading speed wasassociated with BAS reading age r = plusmn 614 p lt 01 by a strongly linear relationshipF(1 59) = 3914 p lt 001 Aplot of this relationship is shown in Fig 3 The data for theve advanced readers have been excluded It can be seen that the Scottish group passesa uency threshold of 225 sitem after a reading age of about 7 years is exceeded

ConclusionsThere are substantial variations among European language groups in facility foridentifying and reading familiar words These differences do not appear to relate to


Figure 2 European mastery thresholds and regression of familiar word reading accuracy (per cent)against BAS reading age (months) for the Scottish P1 and P2 sample

the contrast between languages with simple and complex syllabic structure Howeverthere is good evidence of an effect of orthographic depth Within the simple seriesFrench and Portuguese were associated with large reductions in accuracy and speedAmong the complex syllable languages Danish showed a relative reduction in accuracyalthough this did not affect uency French Portuguese and Danish were identied asthe deeper European orthographies (see Table 1)

The most striking feature is the relative delay in achievement of efcient reading offamiliar words by the English-speaking sample In particular the performance of the P1Scottish group fell far below the levels of rst-year groups in other countries The delaywas quantied in terms of rates of gain in accuracy and speed relative to BAS readingage The results suggest that a reading age in excess of 7 years is needed beforeperformance converges on the levels which are typical for the rst year of learning inthe majority of European orthographies

STUDY 3 SIMPLE NONWORDSThe third study tested capacity to decode simple mono- and bisyllabic nonwords Theaccuracy and time data have been summarized in Table 6 Monosyllables were read moreaccurately than bisyllables 89 versus 81 F(1 1107) = 4335 p lt 001 and at higherspeed 188 versus 294 sitem F(1 1042) = 6603 p lt 001

AccuracyThe preliminary analyses were conducted using the overall scores for nonworddecoding (mono- and bisyllables combined) on the Grade 1 samples Scottish dataexcluded In line with the complexity hypothesis nonword reading was more accurate


Figure 3 European uency threshold and regression of familiar word reading speed (sitem) againstBAS reading age (months) for the Scottish P1 and P2 sample


Tab

le6

Acc

urac

y(p

erce

ntco

rrec

t)an

dtim

e(s

econ

dspe

rite

m)

data

for

read

ing

lists

ofsi

mpl

eno

nwor

ds

with

outc

omes

for

mon

osyl

labi

cbi

sylla

bic

and

com

bine

dsy

llabl

ele

ngth

s(s

tand

ard

devi

atio

nsin

pare

nthe

ses)

for

each

lang

uage

grou

p

Sim

ple

nonw

ord

read

ing

Mon

osyl

labi

cno

nwor

dsBi

sylla

bic

nonw

ords

Com

bine

dno

nwor

ds

s

item

s

item

s

item


Shal

low

D

eep

Finn

ish

Gre

ekIt

alia

nSp

anis

hPo

rtug

uese

Fren

chP1

Fren

chP2

976

2(4

78)

968

4(5

14)

921

1(1

936

)93

06

(10

28)

763

2(2

187

)88

27

(18

43)

992

3(2

45)

108

(06

6)1

32(0

53)

158

(06

8)0

96(0

54)

224

(12

7)3

09(2

48)

086

(02

0)

924

6(8

92)

872

5(1

028

)86

70

(20

85)

845

8(1

448

)77

49

(19

63)

816

0(2

383

)95

59

(11

05)

181

(10

8)1

97(0

69)

241

(09

3)2

03(1

74)

370

(22

9)5

17(5

18)

101

(02

2)

950

4(5

77)

920

5(6

90)

894

0(1

954

)88

82

(11

78)

769

0(1

945

)84

94

(20

61)

974

2(6

38)

144

(08

6)1

64(0

60)

199

(07

7)1

50(1

12)

297

(17

3)4

13(3

79)

094

(01

8)


Shal

low

D

eep

Aus

tria

nG

erm

anN

orw

egia

nIc

elan

dic

Swed

ish

Dut

chD

anis

hP1

Dan

ish

P2Sc

ottis

hP1

Scot

tish

P2

971

7(3

95)

978

2(3

15)

933

3(9

84)

912

9(1

092

)90

89

(89

1)90

48

(10

99)

625

6(3

126

)87

50

(15

14)

405

6(3

665

)74

26

(22

67)

116

(03

9)1

10(0

48)

195

(11

1)1

84(1

14)

237

(19

6)2

65(0

99)

386

(28

4)1

75(1

93)

604

(34

2)2

47(1

43)

866

5(1

115

)90

87

(80

3)88

19

(20

60)

816

8(1

640

)84

44

(11

56)

740

1(2

410

)44

87

(36

09)

750

0(2

544

)17

96

(26

37)

527

8(3

348

)

197

(06

6)1

80(0

67)

286

(15

2)2

79(1

52)

386

(38

4)4

68(2

17)

518

(39

4)2

93(3

13)

735

(46

1)3

87(2

40)

919

1(6

54)

943

5(4

99)

907

6(1

486

)86

49

(11

60)

876

7(8

97)

822

4(1

672

)53

72

(31

82)

812

5(1

914

)29

26

(29

40)

635

2(2

553

)

157

(05

1)1

45(0

56)

240

(13

0)2

31(1

32)

311

(28

8)3

67(1

50)

458

(31

9)2

34(2

51)

669

(38

7)3

17(1

77)

in simple syllable languages than in complex syllable languages 8871 versus 8106F(1 553) = 2018 p lt 001 It seemed possible that the effect might be contingent on thepresence of the deeper orthographies in each series To guard against this the analysiswas repeated with French Portuguese and Danish removed The effect was preserved(9194 for simple languages 8932 for complex) and remained signicantF(1 405) = 518 p lt 05

There were differences among the simple syllable languages F(5 269) = 88p lt 001 Post hoc Scheffe tests differentiated Portuguese and French from the otherlanguages (gt 85correct) Differences among complex syllable languages were alsosignicant F(6 273) = 3319 p lt 001 The post hoc tests separated Danish (5372)from the remainder of the series (82ndash94)

Accuracy in the Scottish P1 sample (2926) was again far below the range for theother orthographies The analysis of variance of all languages indicated highlysignicant differences F(13 571) = 437 p lt 001 which were resolvable into threegroups Scottish P1 Danish and the remaining languages including French andPortuguese The Scottish P2 group at 635 correct was also below all Grade1samples other than Danish and was grouped with Danish in the post hoc tests TheDanish P2 group (8125) was close to the lower boundary of the main set ofEuropean languages

It is again possible to estimate the amount of experience required to achieve masteryof simple decoding in English by examining the relationship between nonword readingaccuracy and BAS reading age In the Scottish P1 and P2 samples accuracy was stronglyassociated with reading age r = + 645 p lt 01 AEuropean Grade 1 mastery level can beapproximated at 85+ for the majority of languages The Scottish group gainedapproximately 218in nonword reading accuracy for each additional month of readingage According to this function the 85 threshold is not passed until reading ageexceeds 8 years 2 months

As in the analysis of familiar word data it seemed possible that the presence of a fewadvanced readers in the Scottish sample might lower the slope of the gain function andraise the estimates of the reading ages required to pass the European threshold Also asalready noted monosyllabic nonwords were read more accurately than bisyllablesExcluding Danish Portuguese and French monosyllables were read with better than90accuracy in all languages and bisyllables with better than 80accuracy Decodingtwo- and three-letter monosyllables represents the most simple and basic level ofdecoding skill imaginable whereas decoding of bisyllables requiring coordination ofsequences of three or four grapheme-phoneme correspondences is more complex anddemanding It seemed worthwhile therefore to examine the gain in competence in theScottish group for these two levels of decoding skill with the advanced readersexcluded

Over the sample as a whole (P1 and P2) correlations with reading age were r = 645p lt 01 for accuracy in reading monosyllables and r = 710 p lt 01 for accuracy inreading bisyllables The gain functions for mono- and bisyllabic nonword readingaccuracy are plotted against BAS reading age in Fig 4 There is a strong linear trendin both sets of data and a slope of approximately 29 percentage points per month ofreading age The data show that Scottish children cross the threshold for simpledecoding (monosyllables) when reading age surpasses 75 years The threshold forcomplex decoding (bisyllables) is passed when reading age exceeds 8 years Acompar-able analysis of the Danish data using chronological age as the index indicated that thethresholds were crossed at just under 9 years of age


SpeedTable 6 gives the reading speeds for the mono- and bisyllabic nonword lists These datawere initially submitted to a general analysis with monosyllables and bisyllablescombined and the Scottish results excluded The main effect of syllabic structure wassignicant F(1 521) = 2438 p lt 001 Average reading speed was faster in the simplesyllable languages than in the complex syllable languages (197 sitem vs 281 sitem)The analysis was repeated without the results for Danish Portuguese and French Asignicant difference F(1 409) = 1797 p lt 001 in favour of simple syllable languagesremained (168 vs 229 sitem)

There were differences between languages within both series (F(5 238) = 1492p lt 001 for the simple series F(6 272) = 1629 p lt 001 for the complex series)Post hoc tests (Scheffe ) differentiated French (413 sitem) from other simple syllablelanguages and placed Portuguese (297 sitem) with Italian in a middle subset Thecomplex syllable languages fell into three overlapping sets with Danish (458 sitem) atthe outer edge of the slowest group together with Dutch and Swedish

The Scottish P1 nonword reading speed (669 sitem) was slower than that of anyother language Comparison with the results for the other complex syllable languagesgave a signicant main effect F(7 301) = 2274 p lt 001 and separated the Scottishmean from all other languages in the series Reading speed in the Scottish P2 group was317 sitem close to the outcomes for Portuguese and French in the simple syllable setand Swedish Dutch and Danish in the complex syllable set Danish P2 children readnonwords at 234 sitem within the range of the faster complex syllable languages

Afurther analysis was undertaken in order to determine the requirement in terms ofreading age for the achievement of uency in nonword reading by the Scottish P1 andP2 sample Simple decoding (monosyllables) and complex decoding (bisyllables) weretreated separately The mean speed of monosyllable reading for languages other thanPortuguese French and Danish was 15 sitem Apreponderance of times were below


Figure 4 European mastery thresholds and regression of accuracy (per cent) in reading simplemonosyllabic (amp ) and bisyllabic (^ ) nonwords against BAS reading age (months) for theScottish P1 and P2 sample

2 sitem the exceptions being Dutch (265 sitem) and Swedish (237 sitem) We willtreat 2 sitem as our best estimate of uent reading of 2ndash3 letter monosyllables Forbisyllables the mean speed without Portuguese French and Danish was 24 sitem Allvalues were below 3 sitem except for Swedish (386 sitem) and Dutch (468 sitem)There is evidence therefore of a general slowness of decoding in these two complexsyllable languages We will adopt 3 sitem as a liberal estimate of the boundary for uentreading of 3ndash4 letter bisyllables

In the Scottish sample nonword reading speed (sitem) correlated with reading ageat r = plusmn 645 p lt 01 for monosyllables and r = plusmn 556 p lt 01 for bisyllables Followingomission of the results for the advanced readers the gain in uency was plusmn 208 sitemper month of reading age for monosyllables and plusmn 177 sitem for bisyllables Bothfunctions were strongly linear F(1 51) = 4846 and 2048 p lt 001 They are plotted inFig 5 against the European uency thresholds for monosyllables and bisyllables It canbe seen that the Scottish data pass the threshold for monosyllables when reading ageexceeds 7 years 4 months and for bisyllables when reading age exceeds 7 years7 months

ConclusionsThe results demonstrate that the capacity to develop simple nonword decoding skillsvaries substantially between orthographies There was evidence that these skills mightbe more difcult to establish in the context of complex syllable languages than in thecontext of simple syllable languages Nonwords were read more accurately and morerapidly by simple syllable than by complex syllable samples There was also evidence oforthographic depth effects paralleling those found for familiar words Reductions in


Figure 5 European uency thresholds and regression of reading speed (sitem) for monosyllabic(amp ) and bisyllabic (^ ) nonwords against BAS reading age (months) for the Scottish P1 andP2 sample

accuracy and uency were apparent in French and Portuguese among the simplesyllable languages and in Danish and to a lesser extent Swedish and Dutch among thecomplex syllable languages

The most striking outcome was the evidence of profound delays in the developmentof simple decoding skills in English The performance of the Scottish P1 sample fell farbelow the range for all other groups Quantication of this effect using the regressionmethod suggested that a reading age of 75 years or above was necessary beforeaccuracy and uency matched the European levels

Lexicality effectsThe lsquolexicality effectrsquo is the processing advantage in terms of error rates and readingspeed for familiar words over unfamiliar nonwords Additional analyses were made ofthe composite word and nonword scores from studies 2 and 3 in order to test for thiseffect in the European orthographies

The rst analysis contrasted word and nonword accuracy in the simple andcomplex syllable languages (Scottish excluded) A summary of the data appears inFig 6 Both groups of languages show a lexicality effect (word advantage) but this issmaller in the simple syllable languages (197 percentage points) than in the complexsyllable languages (863 percentage points) The analysis of variance conrmed anoverall advantage for simple syllable languages F(1 1106) = 146 p lt 001 an effectof lexicality F(1 1106) = 2193 p lt 001 and syllable acute lexicality interactionF(1 1106) = 866 p lt 01 The interaction is interpreted as the consequence of thedifference in magnitude of the lexicality effect in the two sets of languages

The uency scores (sitem) were also analysed Reading was slower incomplex syllable languages F(1 1043) = 537 p lt 05 There was a lexicality effectF(1 1043) = 2410 p lt 001 and a strong interaction between the two factorsF(1 1043) = 2154 p lt 001 Again the interaction is taken to indicate an exaggerationof the lexicality effect in the complex syllable languages (117 sitem) relative to thesimple syllable languages (003 sitem)

Table 7 reports the lexicality effects (wordnonword difference scores) on accuracyfor each language individually In the simple syllable set these differences were allrelatively small and reversed in two cases (French and Portuguese) Analysis of varianceindicated signicant effects of lexicality and language There was a lexicality acute languageinteraction F(5 538) = 284 p lt 05 which was interpreted as reecting the differencesin the direction of the effect Among complex syllable languages the lexicality effectswere all positive (words gt nonwords) but varied in size producing a language acutelexicality interaction F(6 546) = 273 p lt 01 A further analysis of the differencescores conrmed the presence of variations between languages F(6 273) = 1342p lt 001 These were attributable according to post hoc tests to the large effectsobserved in Danish and Dutch The effect was relatively smaller for the Scottish P1sample but substantial for the Scottish P2 and Danish P2 samples

The table also reports the sizes of the lexicality effects on uency (sitem) Thepattern is very similar to the outcomes for accuracy The effect was small and notsignicant in simple syllable languages F(1 476) lt 1 In complex syllable languagesthere was a large effect F(1 551) = 6675 p lt 001 and a strong language acute lexicalityinteraction F(6 545) = 873 p lt 001 which was interpreted following an analysis ofdifference scores in terms of an exaggeration of the effect in Dutch and Danish Theseeffects on uency were not clearly evident in the Scottish P1 or P2 data


ConclusionsThe results show that simple and complex syllable languages may be differentiated withrespect to the size of the lexicality effect (the advantage for words over nonwords) Theeffect for both accuracy and speed was small in simple syllable languages but relativelylarge in complex syllable languages and exaggerated in some cases (Danish and Dutch)

Error patternsThe error responses produced during the attempts on the word and nonword lists wereclassied as refusals word substitutions nonword substitutions or letter soundings (see


Figure 6 Error rates (per cent) for familiar word and simple nonword reading by simplesyllable language groups and complex syllable language groups

Table 8) The incidence of refusals was very low in all languages other than Danish andEnglish The Scottish and Danish P1 samples exhibit a characteristic pattern in whichword substitution errors predominate in responses to word targets while nonwordsubstitutions predominate in responses to nonword targets The Scottish children alsoproduced a number of letter-sounding responses The emphasis on word substitution isrelatively stronger in the results for the French-speaking sample All other groupsPortuguese included show a bias towards alphabetic reading (preponderance ofnonword substitutions) in the attempts at nonword reading

ConclusionsError patterns differed between languages Primitive reactions such as refusals orsoundings of letters were observed in the Scottish samples and to some extent inDanish Word substitution errors occurred in English Danish Portuguese and French inthe attempts on word lists and in English and French in the attempts on nonword listsAll groups Scottish and Danish included gave evidence of alphabetic processingthrough production of nonword error responses

Individual variationAstriking feature of studies 2 and 3 (see Tables 5 and 6) is the differences in the varianceof the accuracy and time scores of each language group Standard deviations average


Table 7 Lexicality effects (wordnonword difference scores) on accuracy (per cent correct) and time(seconds per item) for each language

Lexicality effects

Wordndashnonwordaccuracy

Nonwordndashwordtime (sitem)

Sim

ple

sylla

ble

stru

ctur

es

Shallow

Deep


325 (537)556 (571)592 (820)590 (821)

plusmn 336 (1246)plusmn 586 (1599)

183 (618)

019 (031)019 (039)029 (046)014 (043)

plusmn 026 (078)plusmn 148 (547)

013 (018)

Com

plex

sylla

ble

stru

ctur

es Shallow

Deep


556 (522)337 (333)104 (490)758 (745)744 (593)

1319 (1515)1735 (1743)1132 (1294)463 (2326)

1287 (2343)

049 (022)039 (020)048 (039)019 (053)110 (127)207 (095)277 (235)129 (150)

plusmn 108 (494)099 (170)

Note Lexicality effects re ect word minus nonword differences for accuracy and nonword minus worddifferences for time Standard deviations are shown in parentheses


Tab

le8

Tota

lper

cent

age

ofer

rors

prod

uced

inea

chsa

mpl

efo

rw

ord

and

nonw

ord

targ

ets

and

the

prop

ortio

nsof

erro

rre

spon

ses

whi

chca

nbe

clas

sie

das

refu

sals

wor

dsu

bstit

utio

nsn

onw

ord

subs

titut

ions

or

lett

erso

undi

ngs

Wor

der

ror

patt

erns

N

onw

ord

erro

rpa

tter

ns

Lang

uage

Erro

rR

efus

als

Wor

dsu

bsN

onw

ord

subs

Lett

erso

undi

ngEr

ror

Ref

usal

sW

ord

subs

Non

wor

dsu

bsLe

tter

soun

ding


Shal

low

D

eep

Finn

ish

Gre

ekIt

alia

nSp

anis

hPo

rtug

uese

Fren

chP1

Fren

chP2

171

240

468

528

264

620

93

077

008

006

066

035

175

185

000

099

177

132

243

102

312

35

048

063

051

066

250

120

66

420

29

000

006

197

000

241

025

000

496

795

106

011

18

231

015

06

259

012

025

000

083

066

043

087

076

158

132

222

453

135

80

29

411

612

651

812

146

91

051

05

000

000

263

000

322

000

038


Shal

low

D

eep

Aus

tria

nG

erm

anN

orw

egia

nIc

elan

dic

Swed

ish

Dut

chD

anis

hP1

Dan

ish

P2Sc

ottis

hP1

Scot

tish

P2

253

228

819

456

593

489

288

97

4366

11

236

1

005

000

000

040

060

000

128

60

9724

07

972

127

139

326

337

248

211

109

06

3216

94

100

0

122

089

014

040

143

233

205

007

731

222

000

000

479

040

143

044

077

014

177

81

67

809

565

924

177

613

51

123

346

28

187

570

74

364

8

000

010

000

040

008

089

110

70

9715

00

491

112

060

014

119

270

200

487

264

136

19

91

687

496

444

103

23

009

0024

10

149

319

17

145

4

000

000

465

585

781

056

355

000

230

67

04

8ndash10 percentage points and around 1 sitem for the main set of shallow orthographiesbut are greatly exaggerated in the deeper orthographies To illustrate Fig 7 shows thelsquonormalrsquo range for accuracy of word and nonword reading in shallow orthographiesdened as scores within the boundaries 6 175 standard units from the mean of thecombined distributions for all languages other than Portuguese French Danish andEnglish (gt 80for words gt 70for nonwords) The gure also shows a lsquodisabilityrsquo rangecontaining outlying data points Some of these are quite extreme (beyond 5 standardunits below the mean) but nonetheless exceed 40correct for words and 25correctfor nonwords There was onlyone instance of a lsquonon-readerrsquo a child in the Italian samplewho was unable to score on either the word or the nonword lists

Figure 7 shows comparable data for the Portuguese and French samples and for theDanish and Scottish P2 samples The statistics for the individual languages have beenused to dene a lsquonormalrsquo range 6 175 standard units from the mean In most instancesthe normal range for these groups substantially overlaps the disability range for theshallow orthographies The exceptions are Danish word reading which has a tighterdistribution and Scottish nonword reading where the dispersion is greatlyexaggeratedNon-readers dened as scores of lt 10accuracy in word or nonword reading were rare(two cases in French one in Scottish P2)

The gure also reports the variability of the Danish and Scottish P1 samples TheDanish P1 data for word reading are comparable to the outcomes for Portuguese Frenchand Scottish P2 and overlap the normal and disability ranges for the shallow ortho-graphies A displacement towards the lower end of the scale is apparent in Danishnonword reading and in the Scottish P1 results for words and nonwords The normal


Figure 7 Dispersion of accuracy scores (per cent) for word and nonword reading byshallow orthography groups and by the French Portuguese Danish and Scottish samples The gure indicates the range of scores falling within 6 175 standard deviations of the mean for eachdistribution and the ranges for outlying scores Non-readers (E) scoring less than 10 accuracy are alsoindicated

range for these distributions approaches or extends into the non-reader zone Therewere nine Danish children who were identiable as non-readers The Scottish P1 samplecontained 12 children who were unable to read nonwords and who read only a few(lt 28) words correctly The set included one non-reader (83for words 278fornonwords) and one classic instance of alphabetic dyslexia (83correct for words 6for nonwords) (Seymour amp Evans 1999)

The Danish and Scottish P1 distributions include a lsquoprecocityrsquo zone above the normalrange for these groups Four Danish children were precocious readers according to thisstandard all with 100accuracy on words and 85ndash100on nonwords Another 10scored above the criterion for nonword reading (gt 85) The Scottish sample includedthree children with advanced scores for both words and nonwords (83 and 64 100and 72 94 and 86) as well as three instances of logographic dyslexia (31 vs 69 33vs 67 and 56 vs 89for words and nonwords respectively) (Seymour ampEvans 1999)

ConclusionsThe task of learning to read a deep orthography generates a much wider variation in rateof initial progress than does the task of learning to read in a shallow orthography Scoreswhich reect the normal variation in Portuguese French Danish and English fall withinthe disability range in the shallow orthographies In English and to a lesser extent inDanish the normal range falls close to the non-reader range Appreciable numbers ofEnglish and Danish children remain unable to read after several months in schoolalthough a few make rapid progress and read within the normal range for the shalloworthographies

GENERAL DISCUSSIONThe COST A8 cross-linguistic study demonstrates that the time needed to establishfoundation literacy varies between European languages The acquisition of elementaryword recognition and decoding occurs more slowly in some languages (PortugueseFrench Danish) than in the majority and the delay is greatly exaggerated in EnglishHow are these cross-language effects to be explained

One obvious hypothesis is that the effects are a product of the differences in the agesat which children start to learn to read This could explain the slow progress made bythe English-speaking sample in the P1 year Five-year-old children may lack the maturitynecessary for mastery of an alphabetic orthography However the Scottish P1 grouplearned the letters no less effectively than the other groups (study 1) It could be that thecapacity to develop an explicit awareness of the phonemic structure of speech is thecritical ability which is compromised by immaturity However Duncan Seymour andHill (1997) tested explicit awareness of phonemes and other linguistic units in P1classes in Dundee schools using a common unit identication task Phonemic awarenessdeveloped rapidly and appeared not to be a limiting factor for a majority of children inP1 It is true that the Scottish group gave some evidence of dysuency when reading listsof letters (study 1) Immaturity might curtail the achievement of automaticity in letterprocessing (LaBerge amp Samuels 1974) and this might slow down the formation of thelogographic and alphabetic processes both of which are dependent on letter-soundrepresentations (Fig 1)

In the wider context the study suggests that the age of starting may not be animportant factor Correlations between age and performance on the foundation literacy


tasks were weak or not signicant (see Table 3) Difculty in learning occurred despite awide variation in starting agendash5 years in the UK 6 years in France and Portugal and7 years in Denmark The comparison with Danish is instructive Danish shares withEnglish the features of a deep orthography and a complex syllable structure In Denmarkchildren do not enter primary school until they are 7 years old Despite this 2-year ageadvantage they experience difculties in acquiring the logographic and alphabeticfoundation processes which are comparable to those observed in English although lessextreme Hence although immaturity may contribute to the slow progress observed inEnglish P1 classes it seems unlikely that age is the major cause of the inter-languagedifferences

Another possibility is that differences in teaching methods may be responsible Inshallow orthographies such as German synthetic phonic methods are commonly used(Wimmer 1993) Some commentators argue that rates of progress could be improved byusing these same methods in English However this contention ignores the distinctionbetween shallow and deep orthographies In Scottish schools there is a preference for amixed method which combines the teaching of a vocabulary of sight words withthe teaching of the letters and decoding procedures (Duncan et al 1997) Thesemethods are well adapted for deep orthographies in which commonly occurringwords contain letter structures which are inconsistent with the principles of simplegraphemendashphoneme correspondence

The more interesting hypothesis is that fundamental linguistic differences areresponsible This hypothesis comes in two parts The rst states that languages whosephonology contains a complex syllable structure dened in terms of a predominance ofclosed CVCsyllables and the presence of numerous initial and nal consonant clusterscreate greater difculties for beginning readers than do languages with a simple opensyllabic structure The second holds that acquisition occurs more rapidly in shalloworthographies which are based on consistent one-to-one mappings between graphemesand phonemes than in deep orthographies which include inconsistent bidirectionalone-to-many mappings and lexical and morphological inuences

The syllabic complexity hypothesis was tested in the comparisons betweensimple syllable languages and complex syllable languages The data support theconclusion that syllabic complexity exerts a selective effect on the development ofthe decoding process This conclusion follows from the exaggeration of the lexicalityeffect and the reduced efciency of nonword reading observed in the results of thecomplex syllable languages (study 3) The effect was not attributable to a generaldifference in efciency The complex syllable group was older than the simplesyllable group and more accurate and faster in letter identication and labelling(study 1) The two groups were equivalent in accuracy and speed of familiar wordreading (study 2) It is only in nonword reading that a signicant disadvantage for thecomplex syllable group emerges

It is important to emphasize that the syllabic complexity effect is evident whensimple nonwords are read The items used in study 3 were based on single lettergraphemendashphoneme correspondences No consonant clusters or multi-lettergraphemes were included Hence the conclusion is that straightforward letter-sounddecoding is more difcult to acquire in the context of a language with a complexphonology than in a language with a simple phonology Why should this be Onepossibility is that the embedding of graphemendashphoneme correspondences in consonantclusters impedes acquisition Thus in English the correspondence p p occurs inisolation and as an element in numerous consonant clusters (sp spr spl pl pr mp lp


mple etc) The embedding of simple correspondences in clusters will occur morefrequently in beginning reading materials in a complex syllable language than in a simplesyllable language

We tested the orthographic depth hypothesis by comparing the languages within thesimple and complex series and by relating the outcomes to the intuitive estimates madeby the COST A8 consortium (Table 1) Within the simple series there were twolanguagesmdashPortuguese and Frenchmdashwhich differed from the other Grade 1 samplesThe mean level and variance of both word reading (study 2) and nonword reading(study 3) were affected (Figs 6 and 7) Alexical bias in error production occurred in theFrench sample The effects cannot be attributed to age since the French and Portuguesesamples were the same age as the remaining simple syllable groups other than FinnishHowever letter labelling accuracy and speed were at the lower end of the range for theseries This could have contributed to the effects on word and nonword reading Againstthis the Finnish sample was equally slow at letter naming but without any consequenteffect on word or nonword efciency

The tests for effects of orthographic depth in the complex syllable series identiedDanish and English as the two languages which differed sharply from the others Againboth the logographic process (word reading) and the alphabetic process (nonwordreading) were affected The Danish Grade 1 results displayed enormous variabilityextending from non-readers up to fully competent readers (Fig 7) and included refusaland word substitution errors The Grade 1 and 2 groups both showed enlarged lexicalityeffects indicating a special difculty in developing effective nonword decoding Theseoutcomes were all present in a much more extreme form in the Scottish results Meanaccuracy in the P1 sample fell below 50 Some children were unable to read and othershad dissociated patterns of word and nonword reading analogous to those reported bySeymour and Evans (1999)mdashalphabetic dyslexia in which nonword reading ismassively inferior to word reading and logographic dyslexia in which good nonwordreading is combined with poor word reading

The delayed acquisition of foundation literacy acquisition in Danish and Englishcan be interpreted as a combined effect of syllabic complexity and of orthographicdepth Both languages have a complex syllabic structure and an inconsistent system ofgraphemendashphoneme correspondences The more extreme effects observed in theScottish sample could be a product of the relative immaturity of the children (thedifference in starting at 5 vs 7 years) or of the greater inconsistency of the Englishorthography

The use of the regression method made it possible to estimate the amount of readingexperience readers of English needed to match European mastery and uency levelsFor familiar words (study 2) a BAS reading age in excess of 7 years was necessary Thiswas also true of simple nonwords (study 3) where reading ages above 75 years wereneeded The results closelyparallel the earlier ndings by Seymour and Evans (1999) andDuncan and Seymour (2000) These studies found a strong correlation (gt + 08)between familiar word reading and BAS reading age and a slightly weaker one(gt + 06) between nonword reading and reading age They also pointed to 7 years asthe reading age at which foundation literacy acquisition was normally complete Giventhat the BAS scale starts at 5 years this suggests that readers of English require 2frac12 ormore years of literacy learning to achieve a mastery of familiar word recognition andsimple decoding which is approached within the rst year of learning in a majority ofEuropean languages Thus the rate of foundation literacy acquisition is slower by a ratioof about 251 in English than it is in most European orthographies We were not able to


make a similar estimate for French or Portuguese However the results for Danishsuggested that two years may normally be required to achieve mastery of simpledecoding (study 3)

We can now consider what mechanism might give rise to these effects Katz and Frost(1992) formulated an orthographic depth hypothesis which asserts that lsquo shalloworthographies are more easily able to support a word recognition process that involvesthe languagersquos phonology deep orthographies encourage a reader to process printedwords by referring to their morphology via the printed wordrsquos visual-orthographicstructurersquo (p 71) Translated into developmental terminology (Fig 1) this hypothesisstates that acquisition in a shallow orthography may be based on a single (alphabetic)process whereas acquisition of a deep orthography requires the formation of a dual(alphabetic + logographic) foundation

This theory suggests an abrupt rather than a graded effect of orthographic depth Ifthe orthography satises relevant criteria of simplicity then a single process alphabeticfoundation will be formed as the basis for later reading If these boundaries areexceeded then the cognitive architecture of the reading process changes dramaticallywith the introduction of a dual process system (Fig 1) This type of account correspondswell to our ndings Both the simple and the complex series present a picture of efcientand undifferentiated performance in the majorityof languages and an abrupt reduction ofefciency in the deeper orthographies (see Fig 6) Thus it appears that there is athreshold of orthographic complexity which once exceeded results in a step change inthe way in which foundation literacy is acquired Portuguese French Danish andEnglish are located above this threshold and the remaining languages below it

The contrast between the single and dual process accounts offers an explanation ofthe effect of orthographic depth on rate of learning In shallow orthographies the tasksof familiar word recognition and decoding are based on a common set of principles (aconsistent set of simple graphemendashphoneme correspondences) and may effectively behandled by a single process In deep orthographies the principles underlying wordrecognition and decoding are distinct Beginning readers of English encounter numer-ous common words (house father nice was etc) which contain complex graphemescontextual variations and irregularities which are not consistent with their concurrentlearning of graphemendashphoneme correspondences To accommodate this discrepancyword recognition (the logographic process) follows a distinctive developmental path-way This is based on a partial cue emphasis through which words are initiallydiscriminated by reference to a subset of their component letters usually the initialand nal letters (Ehri 1997 Stuart amp Coltheart 1988) Decoding (the alphabeticprocess) relies on sequential left-to-right identication and blending of individualletter-sounds

Learning under conditions where attention and processing resources are dividedbetween two functions will occur more slowlythan learning under conditions where allresources are focused on a single function Indeed our data suggest that learning may beabout twice as slow under these circumstances In addition dual process learning willdemand the engagement of a wider range of cognitive skills than single process learningThis may explain why the individual variations in rates of progress are so much wider inthe deep than in the shallow orthographies (see Fig 7) If the dual foundation processesare functionally distinct it will make sense that the dissociated patterns of logographicand alphabetic dyslexia should occur in some instances (Seymour amp Evans 1999)

According to this interpretation the slow rate of foundation literacy acquisition bythe Scottish sample can be seen as an inevitable consequence of the complexity of the


orthography and phonology of English It may be that the rate of learning can beinuenced at the margins being further delayed by socioeconomic disadvantage(Duncan amp Seymour 2000) possibly accelerated by modications to the teaching ofphonics and perhaps sensitive to the childrsquos cognitive maturity when the teaching ofreading is introduced However even where these educational aspects are all optimalthere will always be a cost associated with the implementation of a dual foundationprocess and this will create irreducible differences in rates of progress betweenlearning to read in English or other deep orthographies and learning to read in ashallow orthography

AcknowledgementsThis research was facilitated by the EC COSTorganisation (httpwwwbelspobecost) throughsupport for Action A8 Learning Disorders as a Barrier to Human Development 1995ndash99 Wegratefullyacknowledge the participants of COSTA8 and particularlythose who organized the datacollection in their own countries We are also most grateful to the researchers who collected thedata and to the schools and school children for their participation

ReferencesDuncan L G amp Seymour P H K (2000) Socio-economic differences in foundation-level literacy

British Journal of Psychology 91 145ndash166Duncan L G Seymour P H K amp Hill S (1997) How important are rhyme and analogy in

beginning reading Cognition 63 171ndash208Ehri L C (1992) Reconceptualising the development of sight word reading and its relationship to

recoding In P B Gough L C Ehri amp R Treiman (Eds) Reading acquisition (pp 107ndash143)Hillsdale NJ Erlbaum

Ehri L C (1997) Learning to read and learning to spell are one and the same almost In C APerfetti L Rieben amp M Fayol (Eds) Learning to spell Research theory and practice acrosslanguages (pp 237ndash269) Hillsdale NJ Erlbaum

Elliot C D (1987) British Ability Scales Windsor NFER-NELSONFrith U Wimmer H amp Landerl K (1998) Differences in phonological recoding in German- and

English-speaking children Journal of the Society for the Scientic Study of Reading 2 31ndash54Frost R Katz L amp Bentin S (1987) Strategies for visual word recognition and orthographical

depth A multilingual comparison Journal of Experimental Psychology Human Perceptionand Performance 13 104ndash115

Goswami U Gombert J E amp de Barrera L F (1998) Childrenrsquos orthographic representationsand linguistic transparency Nonsense word reading in English French and Spanish AppliedPsycholinguistics 19 19ndash52

Goswami U Porpodas C amp Wheelwright S (1997) Childrenrsquos orthographic representations inEnglish and Greek European Journal of Psychology of Education 12 273ndash292

Gough P B amp Hillinger M L (1980) Learning to read An unnatural act Bulletin of the OrtonSociety 30 180ndash196

Katz L amp Frost R (1992) Reading in different orthographies The orthographic depthhypothesis In R Frost amp L Katz (Eds) Orthography phonology morphology and meaning(pp 67ndash84) Amsterdam North-Holland

LaBerge D ampSamuels J (1974) Toward a theoryof automatic information processing in readingCognitive Psychology 2 293ndash323

Niessen M Frith U Reitsma P amp Ohngren B (2000) Learning disorders as a barrier tohuman development 1995 ndash1999 Evaluation report Technical Committee COST SocialSciences


Peereman R amp Content A (1997) Orthographic and phonological neighborhoods in namingNot all neighbors are equally inuential in orthographic space Journal of Memory andLanguage 37 382ndash410

Peereman R amp Content A (1999) LEXOP A lexical database providing orthographymdashphonology statistics for French monosyllabic words Behaviour Research MethodsInstruments and Computers 31 376ndash379

Plaut D D McClelland J L Seidenberg M S amp Patterson K (1996) Understanding normal andimpaired word reading Computational principles in quasi-regular domains PsychologicalReview 103 56ndash115

Seymour P H K (1990) Developmental dyslexia In M W Eysenck (Ed) Cognitive psychologyAn international review (pp 135ndash196) Chichester Wiley

Seymour P H K (1997) Foundations of orthographic development In C Perfetti L Rieben ampM Fayol (Eds) Learning to spell (pp 319ndash337) Hillsdale NJ Erlbaum

Seymour P H K (1999) Cognitive architecture of earlyreading In I Lundberg F E Toslashnnessen ampI Austad (Eds) Dyslexia Advances in theory and practice (pp 59ndash73) Dordrecht Kluwer

Seymour P H K amp Evans H M (1999) Foundation-level dyslexias Assessment and treatmentJournal of Learning Disabilities 32 394ndash405

Share D L (1995) Phonological recoding and self-teaching Sine qua non of reading acquisitionCognition 55 151ndash218

Stuart M amp Coltheart M (1998) Does reading develop in a sequence of stages Cognition 30139ndash181

Wimmer H (1993) Characteristicsof developmental dyslexia in a regular writing system AppliedPsycholinguistics 14 1ndash33

Wimmer H amp Goswami U (1994) The inuence of orthographic consistency on readingdevelopment Word recognition in English and German children Cognition 51 91ndash103

Ziegler J C Jacobs A M amp Stone G O (1996) Statistical analysis of the bi-directionalinconsistency of spelling and sound in French Behaviour Research Methods Instrumentsand Computers 28 504ndash515

Ziegler J C Stone G O amp Jacobs A M (1997) What is the pronunciation for -ough and thespelling for u A database for computing feedforward and feedback consistency in EnglishBehaviour Research Methods Instruments and Computers 29 600ndash618

Received 16 August 2001 revised version received 14 February 2002

Appendix

Lists of letter items used in letter knowledge tasksFinnish a b d e f g h i j k l m n o p r s t u v y a oGreek a b g d laquo z h v i k l m n y o p r sect t u J x w q

Italian a b c d e f g i l m n o p q r s t u v zSpanish a b c d e f g h i j k l m n o p r s t u v x y z nPortuguese a b c d e f g i j l m n o p r s t u v x zFrench a b d e f i j k l m n o p q r s t u v x zGerm an a b d e f g h i j k l m n o p r s t u v w zNorw egian a b c d e f g h i j k l m n o p r s t u v y aEcirc aelig oslashIce landic a b d e f g h i j k l m n o p r s t u v x y aelig o a e iacute o u y eth THORNSw edish a b c d e f g h i j k l m n o p r s t u v x y z aEcirc a o aDutch a b c d e f g h i j k l m n o p q r s t u v w x y zDanish a b c d e f g h i j k l m n o p r s t u v y aEcirc aelig oslashEnglis h a b c d e f g h i j k l m n o p q r s t u v w x y z


Lists of familiar content and function words used in reading tasks

FinnishContent koti (home) juosta (run) kolme (three) kirja (book) talo (house) pieni

(small) joulu (christmas) takki (coat) isa (father) koulu (school) tehda(make) vene (boat) hyva (good) pallo (ball) kaunis (beautiful) talvi(winter) aiti (mother) kaksi (two)

Functor siella (there) se (it) vain (only) mika (what) alas (down) ei (no) tassa(here) taas (again) etta (that) pian (soon) ylos (up) taalla (here) on (is)joka (that) kun (when) mutta (but) me (we) missa (where)

GreekContent (children) (house) (grandfather) (rain)

(village) (eye) (apple) (door) (night)(grandmother) (rabbit) (day) (bird)

(sun) (hand) (uncle) (line) (time)Functor (and) (for) (how) (here) (up) (now)

(as) (only) (always) (where) (from)(into) (there) (then) (but) (when) (down)

(after)

ItalianContent ape (bee) ago (needle) ora (hour) lago (lake) cima (peak) dama (lady)

prato (meadow) resto (rest) colpo (blow) ala (wing) oca (goose) uva(grapes) pane (bread) mela (apple) toro (bull) libro (book) ponte(bridge) treno (train)

Functor il (art) la (art) su (on) fra (between) per (for) con (with) del (of ) di(of ) tra (between) lo (him) sul (on) un (art) in (in) al (to) da (from) nel(in) le (art) sui (on)

SpanishContent nino (child) alto (tall) padre (father) a rbol (tree) n (end) uno (one) or

(ower) gato (cat) soy (I am) casa (house) algo (something) madre(mother) azul (blue) sol (sun) dos (two) tren (train) goma (rubber)estoy (I am)

Functor de (of ) para (to) con (with) entre (between) desde (from) cual(which) aquiacute (here) poco (few) tan (so) te (you) pero (but) por(by) sobre (on) despue s (after) quien (who) alliacute (there) mucho(much) ma s (more)

PortugueseContent pe (foot) luz (light) ar (air) asa (wing) gato (cat) ovos (eggs) carta

(letter) prado (graze eld) vento (wind) pa (sickle) sol (sun) a s (ace) uva(grape) bola (ball) anel (ring) porta (door) blusa (blouse) mundo(world)

Functor tu (you) por (by) as (art) ali (there) cada (each) logo (soon)umas (some) mesmo (same) nunca (never) se (if ) tal (such) os(art) uma (art) para (for) como (as) apo s (after) desde (since)tanto (much)


FrenchContent heure (hour) voir (see) air (air) femme (woman) vent (wind) roi (king)

homme (man) main (hand) jour (day) aimer (love) e te (summer) ide e(idea) monsieur (mister) diacutener (dinner) souris (mouse) e cole (school)image (picture) e toile (star)

Functor donc (therefore) une (artone) chaque (each) elle (she) pour (for) dans(in) il (he) qui (who) nous (us) enn (at last) alors (then) ici (here)encore (still) assez (enough) avec (with) pendant (during) de ja (already)souvent (often)

Germ anContent Kind (child) Brot (bread) Oma (granny) Maus (mouse) Auto (car) Mutter

(mother) Esel (donkey) Wasser (water) Torte (cake) Schnee (snow) Papa(daddy) Stein (stone) Zahn (tooth) Winter (winter) Spiel (gameplay)Fisch (sh) Teller (plate) Schule (school)

Functor der (artthat) mit (with) bis (until) und (and) was (what) du (you) sind(are) es (it) wer (who) wir (we) ist (is) zu (to) er (he) hat (has) ein(article) wo (where) das (artthat) ich (I)

Norw egianContent bok (book) leke (toy) far (father) jente (girl) bake (bake) tre (three)

skole (school) doslashr (door) hjem (home) mor (mother) baEcirc t (boat) kake(cake) gutt (boy) lekse (lesson) hus (house) y (airplane) saft (juice)ball (ball)

Functor min (my) paEcirc (on) ut (out) og (and) her (here) inn (in) ikke (not) der(there) eller (or) som (like) alle (all) fra (from) en (art) hvor (where)ogsaEcirc (too) til (until) hvem (who) din (your)

Ice landicContent la s (lock) pabbi (daddy) ha r (hairhigh) koma (come) lesa (read) ro la

(swing) mu s (mouse) so l (sun) mamma (mummy) bo k (book) hu s(house) pakki (packagepresent) sko li (school) leikur (game) afmaeligli(birthday) fugl (bird) straeligto (bus) jo l (Christmas)

Functor aeth (toat) og (and) sem (thatwhich) eetha (or) meeth (with) um (about) til(to) en (but) fra (from) he r (here) THORNegar (when) siacute ethan (since) aftur(again) uppi (up) ekki (not) segir (says) hvar (where) sagethi (said)(after)

Sw edishContent liv (life) barn (children) holl (held) hand (hand) del (part) far (father)

mamma (mummy) vag (way) kvinna (woman) dag (day) tid (time) hem(home) folk (people) fall (case) mor (mother) ogon (eyes) dorren (thedoor) brev (letter)

Functor jag (I) paEcirc (on) inte (not) for (for) ett (art) man (you) sin (hisher) ut(out) mot (against) en (one) som (like) var (was) till (to) saEcirc (so) nar(when) fraEcirc n (from) kan (can) nej (no)


DutchContent naam (name) bal (ball) kok (cook) roos (rose) boek (book) geit (goat)

mus (sparrow) huis (house) man (man) school (school) goud (gold)boot (boat) vis (sh) stad (town) pot (pot) boer (farmer) vuur (re) dik(thick)

Functor hij (he) er (there) niet (not) ik (I) het (it) op (on) voor (before) in (in)de (the) een (a) van (of ) ze (they) die (that) met (with) je (you) is (is)maar (but) ze (they)

DanishContent skole (school) bil (car) ser (see) far (father) dag (day) mor (mother) lille

(little) hjem (home) dyr (animal) hus (house) to (two) baEcirc d (boat) elefant(elephant) bor (lives) tur (trip) blaEcirc (blue) traelig (tree) ord (word)

Functor min (my) om (about) dem (them) alle (all) ud (from) man (you) ikke(not) maEcirc (may) er (is) som (like) hvor (where) du (you) ogsaEcirc (also) har(has) men (but) til (to) sin (hisher) her (here)

EnglishContent one high home school boy bird play yellow birthday house boat two

teacher toy elephant tree Tuesday blueFunctor my them about all your when how out many as where you why

every saw here said have

Lists of nonwords used in reading tasks

FinnishMonosyll eh vis li na kup da yt os rom is vor ke tyn ju ep at hal moBisyll varo ame hopa ypot onu ehi iman olus kaju yso vami evot ryhe eda

ukes pilo atu ojun

Greek Monosyll Bisyll

Italian Monosyll af id ot so be fu lat mur pes ul or es pa ti de den pol rBisyll tibe sopa fude ulaf erid otes ari ela ope beti defu pabe idor aful

ides ifa ubi aru

SpanishMonosyll ez fo in su nes ur val fos ta bu sen ar ma ol re dis em farBisyll oti omen te inos abo ledo asur efa fovu ubil ufe sune asu udon ib

pu edos sare

Portuguese Monosyll ba zo gus nir sor ur al ol vu jo ga tor ros cas or ul erBisyll eda iro afo dajo veca buna eris ogal apir ope emo ila tega mipo

fuba uvel adol ifos


FrenchMonosyll ur al oc bo da ja buc nir sof ol ip af ru ko raf vor tupBisyll uba ori afo nita muro silu itol apuf udic ita ulo oba baru vina jotu

otip arol ur

Germ anMonosyll Ol me tis ak su kup it ko dus sim gi af hok he us jal pa etBisyll ebol abi rela awit edu soti orus uki lane eho je uwal oga huko

ibam uki mipu etus

Norw egianMonosyll toslash val mu bik po ir tip an es sut kaEcirc im boslashl na fok ar ve elBisyll ase safo ipa neto alok miva obe edat oslashger toke ate evan bega ode

itaEcirc p mo oslashle efas

Ice landicMonosyll oeth a p u l go fu va jar THORNis mek iacute n em o t ri ba go du l jok THORNa sBisyll era o iacute ku ruli hipo natu a kat imok aneeth abu emi u pa THORNeni fulo

hiacute su a lip utak opis

Sw edishMonosyll ag ef em li ku ci fev daEcirc k nol ot ov op ba hu ky gos bal kesBisyll ata ari osaEcirc kaby nite tila aEcirc rit avad arot una avaEcirc ety vase koge foja

ovan uted alog

DutchMonosyll ak pi lo ruf ol kal ek ro pum ki bog em lir us sa mer du ipBisyll abo kosi seku ores upe aler oki upid dalo pile upa ebor efu abi

bano imon ekal rupa

DanishMonosyll mo git yr paelig tas ne toslashb aeligk oslasht sa kaeligt us gy vul aelign lat ut Bisyll oslashbe aeligkis gima alo imal tasi usoslashk aeligsi sibe utaEcirc alus moslashku ite vamo

oslashsas ilu laelig atib

Englis hMonosyll tas le eb dem vo im ga os p du kib ig je ut vap ca lem opBisyll uli idap umic feno elot mola aco oma suba uba imal sero osi bina

gopa eto afen opud


between English which is regarded as a deep orthography containing manyinconsistencies and complexities and other alphabetic European languages severalof which have shallow orthographies with consistent graphemendashphoneme correspon-dences For example Wimmer and Goswami (1994) compared reading of digits numbernames and nonwords formed by exchanging the onsets and rimes of number names by7- 8- and 9-year-old children in German and English Nonword reading was signicantlyslower and more error prone in English at all three age levels Frith Wimmer andLanderl (1998) used structurally equivalent sets of 1- 2- and 3-syllable nonwords inEnglish and German and again found consistently poorer nonword reading in EnglishSimilar data are reported for comparisons of English with Spanish and French byGoswami Gombert and de Barrera (1998) and with Greek by Goswami Porpodasand Wheelwright (1997) These studies suggest that the decoding process which iscommonly assigned a central role in theoretical accounts of reading acquisition (Ehri1992 Gough amp Hillinger 1980 Share 1995) develops more slowly and less effectivelyin English than in other European languages

The present studyextends this work to a comparison of English with a wider range ofEuropean languages and also seeks to determine the stage in reading acquisition atwhich the orthographic depth effect becomes evident The theoretical context isprovided by the foundation literacy framework developed by Seymour (1990 19971999) This proposes that reading is acquired in phases such that basic foundationalcomponents are established in Phase 1 while the complexities of orthographic andmorphographic structure are internalized in Phases 2 and 3 Asimplied version of themodel is shown in Fig 1 The foundation consists of two processes a logographicprocess involved in the identication and storage of familiar words and an alphabeticprocess which supports sequential decoding Both of these processes are held to bedependent on the availability of letter-sound knowledge (Ehri 1992 1997) In theirturn the foundations underpin the development of an orthographic framework inwhich the full complexity of the spelling system is represented in an abstract general-izable format (Plaut McClelland Seidenberg amp Patterson 1996) Reading acquisition is


Figure 1 Schematic representation of the dual foundation model of orthographic development (fromDuncan amp Seymour 2000)














Orthographic depth

Shallow Deep


Portuguese French

Sylla

bic

stru

ctur

e


DutchSwedish

Danish English





















mpl

esy

llabl

est

ruct

ures

Shallow

Deep


1111112

70443840384529

787 (28)675 (27)689 (31)684 (27)701 (32)669 (32)794 (56)

Com

plex

sylla

ble

stru

ctur

es

Shallow

Deep


Scottish

111111121

2

572840372528654030

30

755 (41)740 (34)794 (28)686 (25)748 (59)697 (36)765 (38)855 (34)

CA 559 (32)RA 619 (63)CA 656 (30)RA 722 (84)













sitem

Sim

ple

sylla

ble

stru

ctur

es

Shallow

Deep


9444 (471)9640 (509)9513 (767)9555 (499)9248 (784)9122 (1083)9803 (432)

148 (055)105 (027)106 (046)103 (044)140 (050)138 (061)126 (018)

Com

plex

sylla

ble

stru

ctur

es Shallow

Deep


9685 (332)9951 (179)9865 (246)9573 (533)9874 (175)8951 (466)9492 (730)9795 (242)9397 (603)9603 (468)

074 (016)075 (019)084 (022)102 (024)136 (037)094 (027)114 (051)063 (017)188 (101)103 (027)

















Tab

le5

Acc

urac

y(p

erce

ntco

rrec

t)an

dtim

e(s

econ

dspe

rite

m)d

ata

(sta

ndar

dde

viat

ions

inpa

rent

hese

s)fo

rre

adin

glis

tsof

very

fam

iliar

wor

dsw

ithou

tcom

esfo

rco

nten

tfu

nctio

nan

dco

mbi

ned

cont

ent

and

func

tion

wor

dsfo

rea

chla

ngua

gegr

oup

Fam

iliar

wor

dre

adin

g

Con

tent

wor

dsFu

nctio

nw

ords

Com

bine

dco

nten

tfu

nctio

n

s

item

s

item

s

item


Shal

low

D

eep

Finn

ish

Gre

ekIt

alia

nSp

anis

hPo

rtug

uese

Fren

chP1

Fren

chP2

981

7(4

09)

965

9(5

90)

948

8(1

754

)94

17

(10

29)

771

9(1

945

)72

47

(27

57)

984

7(4

67)

138

(09

1)1

60(0

63)

218

(11

8)1

40(1

23)

344

(24

4)7

04(1

087

)1

00(0

29)

984

1(3

43)

986

1(2

97)

957

6(1

661

)95

28

(97

1)69

88

(20

77)

856

8(2

087

)10

000

(00

0)

113

(07

2)1

31(0

67)

123

(05

8)1

31(1

00)

300

(18

1)4

19(6

74)

064

(02

1)

982

9(3

76)

976

0(4

75)

953

2(1

697

)94

72

(99

6)73

54

(20

32)

790

7(2

354

)99

23

(24

5)

125

(08

3)1

45(0

66)

170

(10

4)1

35(1

11)

322

(21

4)5

61(9

00)

084

(02

6)


Shal

low

D

eep

Aus

tria

nG

erm

anN

orw

egia

nIc

elan

dic

Swed

ish

Dut

chD

anis

hP1

Dan

ish

P2Sc

ottis

hP1

Scot

tish

P2

970

8(4

94)

974

2(4

90)

915

3(1

655

)93

39

(91

5)93

56

(93

1)92

66

(71

0)70

94

(26

78)

922

2(1

388

)32

59

(24

83)

790

7(1

832

)

134

(06

4)1

35(0

74)

213

(13

9)2

36(1

65)

264

(25

7)2

02(0

84)

115

(10

3)1

18(1

40)

819

(60

8)2

11(1

40)

978

6(4

42)

980

2(3

77)

920

8(1

423

)94

74

(85

8)96

67

(48

1)98

21

(50

3)71

20

(29

93)

929

2(1

045

)35

18

(26

23)

737

0(2

134

)

082

(02

9)0

76(0

39)

171

(11

9)1

88(1

21)

139

(12

2)1

17(0

67)

251

(20

0)0

91(0

82)

736

(64

3)2

25(1

59)

974

7(4

68)

977

2(4

34)

918

1(1

534

)94

07

(88

4)95

11

(75

0)95

44

(67

1)71

07

(27

71)

925

7(1

174

)33

89

(24

97)

763

9(1

927

)

108

(05

6)1

06(0

66)

192

(13

0)2

12(1

46)

201

(20

9)1

60(0

87)

185

(14

2)1

05(1

09)

778

(61

9)2

18(1

41)














Tab

le6

Acc

urac

y(p

erce

ntco

rrec

t)an

dtim

e(s

econ

dspe

rite

m)

data

for

read

ing

lists

ofsi

mpl

eno

nwor

ds

with

outc

omes

for

mon

osyl

labi

cbi

sylla

bic

and

com

bine

dsy

llabl

ele

ngth

s(s

tand

ard

devi

atio

nsin

pare

nthe

ses)

for

each

lang

uage

grou

p

Sim

ple

nonw

ord

read

ing

Mon

osyl

labi

cno

nwor

dsBi

sylla

bic

nonw

ords

Com

bine

dno

nwor

ds

s

item

s

item

s

item


Shal

low

D

eep

Finn

ish

Gre

ekIt

alia

nSp

anis

hPo

rtug

uese

Fren

chP1

Fren

chP2

976

2(4

78)

968

4(5

14)

921

1(1

936

)93

06

(10

28)

763

2(2

187

)88

27

(18

43)

992

3(2

45)

108

(06

6)1

32(0

53)

158

(06

8)0

96(0

54)

224

(12

7)3

09(2

48)

086

(02

0)

924

6(8

92)

872

5(1

028

)86

70

(20

85)

845

8(1

448

)77

49

(19

63)

816

0(2

383

)95

59

(11

05)

181

(10

8)1

97(0

69)

241

(09

3)2

03(1

74)

370

(22

9)5

17(5

18)

101

(02

2)

950

4(5

77)

920

5(6

90)

894

0(1

954

)88

82

(11

78)

769

0(1

945

)84

94

(20

61)

974

2(6

38)

144

(08

6)1

64(0

60)

199

(07

7)1

50(1

12)

297

(17

3)4

13(3

79)

094

(01

8)


Shal

low

D

eep

Aus

tria

nG

erm

anN

orw

egia

nIc

elan

dic

Swed

ish

Dut

chD

anis

hP1

Dan

ish

P2Sc

ottis

hP1

Scot

tish

P2

971

7(3

95)

978

2(3

15)

933

3(9

84)

912

9(1

092

)90

89

(89

1)90

48

(10

99)

625

6(3

126

)87

50

(15

14)

405

6(3

665

)74

26

(22

67)

116

(03

9)1

10(0

48)

195

(11

1)1

84(1

14)

237

(19

6)2

65(0

99)

386

(28

4)1

75(1

93)

604

(34

2)2

47(1

43)

866

5(1

115

)90

87

(80

3)88

19

(20

60)

816

8(1

640

)84

44

(11

56)

740

1(2

410

)44

87

(36

09)

750

0(2

544

)17

96

(26

37)

527

8(3

348

)

197

(06

6)1

80(0

67)

286

(15

2)2

79(1

52)

386

(38

4)4

68(2

17)

518

(39

4)2

93(3

13)

735

(46

1)3

87(2

40)

919

1(6

54)

943

5(4

99)

907

6(1

486

)86

49

(11

60)

876

7(8

97)

822

4(1

672

)53

72

(31

82)

812

5(1

914

)29

26

(29

40)

635

2(2

553

)

157

(05

1)1

45(0

56)

240

(13

0)2

31(1

32)

311

(28

8)3

67(1

50)

458

(31

9)2

34(2

51)

669

(38

7)3

17(1

77)




































Lexicality effects



Sim

ple

sylla

ble

stru

ctur

es

Shallow

Deep


325 (537)556 (571)592 (820)590 (821)

plusmn 336 (1246)plusmn 586 (1599)

183 (618)

019 (031)019 (039)029 (046)014 (043)

plusmn 026 (078)plusmn 148 (547)

013 (018)

Com

plex

sylla

ble

stru

ctur

es Shallow

Deep


556 (522)337 (333)104 (490)758 (745)744 (593)

1319 (1515)1735 (1743)1132 (1294)463 (2326)

1287 (2343)

049 (022)039 (020)048 (039)019 (053)110 (127)207 (095)277 (235)129 (150)

plusmn 108 (494)099 (170)



Tab

le8

Tota

lper

cent

age

ofer

rors

prod

uced

inea

chsa

mpl

efo

rw

ord

and

nonw

ord

targ

ets

and

the

prop

ortio

nsof

erro

rre

spon

ses

whi

chca

nbe

clas

sie

das

refu

sals

wor

dsu

bstit

utio

nsn

onw

ord

subs

titut

ions

or

lett

erso

undi

ngs

Wor

der

ror

patt

erns

N

onw

ord

erro

rpa

tter

ns

Lang

uage

Erro

rR

efus

als

Wor

dsu

bsN

onw

ord

subs

Lett

erso

undi

ngEr

ror

Ref

usal

sW

ord

subs

Non

wor

dsu

bsLe

tter

soun

ding


Shal

low

D

eep

Finn

ish

Gre

ekIt

alia

nSp

anis

hPo

rtug

uese

Fren

chP1

Fren

chP2

171

240

468

528

264

620

93

077

008

006

066

035

175

185

000

099

177

132

243

102

312

35

048

063

051

066

250

120

66

420

29

000

006

197

000

241

025

000

496

795

106

011

18

231

015

06

259

012

025

000

083

066

043

087

076

158

132

222

453

135

80

29

411

612

651

812

146

91

051

05

000

000

263

000

322

000

038


Shal

low

D

eep

Aus

tria

nG

erm

anN

orw

egia

nIc

elan

dic

Swed

ish

Dut

chD

anis

hP1

Dan

ish

P2Sc

ottis

hP1

Scot

tish

P2

253

228

819

456

593

489

288

97

4366

11

236

1

005

000

000

040

060

000

128

60

9724

07

972

127

139

326

337

248

211

109

06

3216

94

100

0

122

089

014

040

143

233

205

007

731

222

000

000

479

040

143

044

077

014

177

81

67

809

565

924

177

613

51

123

346

28

187

570

74

364

8

000

010

000

040

008

089

110

70

9715

00

491

112

060

014

119

270

200

487

264

136

19

91

687

496

444

103

23

009

0024

10

149

319

17

145

4

000

000

465

585

781

056

355

000

230

67

04































































Appendix












(after)






































ukes pilo atu ojun



ides ifa ubi aru


pu edos sare


fuba uvel adol ifos



otip arol ur


ibam uki mipu etus






ovan uted alog


bano imon ekal rupa




gopa eto afen opud















Orthographic depth

Shallow Deep


Portuguese French

Sylla

bic

stru

ctur

e


DutchSwedish

Danish English





















mpl

esy

llabl

est

ruct

ures

Shallow

Deep


1111112

70443840384529

787 (28)675 (27)689 (31)684 (27)701 (32)669 (32)794 (56)

Com

plex

sylla

ble

stru

ctur

es

Shallow

Deep


Scottish

111111121

2

572840372528654030

30

755 (41)740 (34)794 (28)686 (25)748 (59)697 (36)765 (38)855 (34)

CA 559 (32)RA 619 (63)CA 656 (30)RA 722 (84)













sitem

Sim

ple

sylla

ble

stru

ctur

es

Shallow

Deep


9444 (471)9640 (509)9513 (767)9555 (499)9248 (784)9122 (1083)9803 (432)

148 (055)105 (027)106 (046)103 (044)140 (050)138 (061)126 (018)

Com

plex

sylla

ble

stru

ctur

es Shallow

Deep


9685 (332)9951 (179)9865 (246)9573 (533)9874 (175)8951 (466)9492 (730)9795 (242)9397 (603)9603 (468)

074 (016)075 (019)084 (022)102 (024)136 (037)094 (027)114 (051)063 (017)188 (101)103 (027)

















Tab

le5

Acc

urac

y(p

erce

ntco

rrec

t)an

dtim

e(s

econ

dspe

rite

m)d

ata

(sta

ndar

dde

viat

ions

inpa

rent

hese

s)fo

rre

adin

glis

tsof

very

fam

iliar

wor

dsw

ithou

tcom

esfo

rco

nten

tfu

nctio

nan

dco

mbi

ned

cont

ent

and

func

tion

wor

dsfo

rea

chla

ngua

gegr

oup

Fam

iliar

wor

dre

adin

g

Con

tent

wor

dsFu

nctio

nw

ords

Com

bine

dco

nten

tfu

nctio

n

s

item

s

item

s

item


Shal

low

D

eep

Finn

ish

Gre

ekIt

alia

nSp

anis

hPo

rtug

uese

Fren

chP1

Fren

chP2

981

7(4

09)

965

9(5

90)

948

8(1

754

)94

17

(10

29)

771

9(1

945

)72

47

(27

57)

984

7(4

67)

138

(09

1)1

60(0

63)

218

(11

8)1

40(1

23)

344

(24

4)7

04(1

087

)1

00(0

29)

984

1(3

43)

986

1(2

97)

957

6(1

661

)95

28

(97

1)69

88

(20

77)

856

8(2

087

)10

000

(00

0)

113

(07

2)1

31(0

67)

123

(05

8)1

31(1

00)

300

(18

1)4

19(6

74)

064

(02

1)

982

9(3

76)

976

0(4

75)

953

2(1

697

)94

72

(99

6)73

54

(20

32)

790

7(2

354

)99

23

(24

5)

125

(08

3)1

45(0

66)

170

(10

4)1

35(1

11)

322

(21

4)5

61(9

00)

084

(02

6)


Shal

low

D

eep

Aus

tria

nG

erm

anN

orw

egia

nIc

elan

dic

Swed

ish

Dut

chD

anis

hP1

Dan

ish

P2Sc

ottis

hP1

Scot

tish

P2

970

8(4

94)

974

2(4

90)

915

3(1

655

)93

39

(91

5)93

56

(93

1)92

66

(71

0)70

94

(26

78)

922

2(1

388

)32

59

(24

83)

790

7(1

832

)

134

(06

4)1

35(0

74)

213

(13

9)2

36(1

65)

264

(25

7)2

02(0

84)

115

(10

3)1

18(1

40)

819

(60

8)2

11(1

40)

978

6(4

42)

980

2(3

77)

920

8(1

423

)94

74

(85

8)96

67

(48

1)98

21

(50

3)71

20

(29

93)

929

2(1

045

)35

18

(26

23)

737

0(2

134

)

082

(02

9)0

76(0

39)

171

(11

9)1

88(1

21)

139

(12

2)1

17(0

67)

251

(20

0)0

91(0

82)

736

(64

3)2

25(1

59)

974

7(4

68)

977

2(4

34)

918

1(1

534

)94

07

(88

4)95

11

(75

0)95

44

(67

1)71

07

(27

71)

925

7(1

174

)33

89

(24

97)

763

9(1

927

)

108

(05

6)1

06(0

66)

192

(13

0)2

12(1

46)

201

(20

9)1

60(0

87)

185

(14

2)1

05(1

09)

778

(61

9)2

18(1

41)














Tab

le6

Acc

urac

y(p

erce

ntco

rrec

t)an

dtim

e(s

econ

dspe

rite

m)

data

for

read

ing

lists

ofsi

mpl

eno

nwor

ds

with

outc

omes

for

mon

osyl

labi

cbi

sylla

bic

and

com

bine

dsy

llabl

ele

ngth

s(s

tand

ard

devi

atio

nsin

pare

nthe

ses)

for

each

lang

uage

grou

p

Sim

ple

nonw

ord

read

ing

Mon

osyl

labi

cno

nwor

dsBi

sylla

bic

nonw

ords

Com

bine

dno

nwor

ds

s

item

s

item

s

item


Shal

low

D

eep

Finn

ish

Gre

ekIt

alia

nSp

anis

hPo

rtug

uese

Fren

chP1

Fren

chP2

976

2(4

78)

968

4(5

14)

921

1(1

936

)93

06

(10

28)

763

2(2

187

)88

27

(18

43)

992

3(2

45)

108

(06

6)1

32(0

53)

158

(06

8)0

96(0

54)

224

(12

7)3

09(2

48)

086

(02

0)

924

6(8

92)

872

5(1

028

)86

70

(20

85)

845

8(1

448

)77

49

(19

63)

816

0(2

383

)95

59

(11

05)

181

(10

8)1

97(0

69)

241

(09

3)2

03(1

74)

370

(22

9)5

17(5

18)

101

(02

2)

950

4(5

77)

920

5(6

90)

894

0(1

954

)88

82

(11

78)

769

0(1

945

)84

94

(20

61)

974

2(6

38)

144

(08

6)1

64(0

60)

199

(07

7)1

50(1

12)

297

(17

3)4

13(3

79)

094

(01

8)


Shal

low

D

eep

Aus

tria

nG

erm

anN

orw

egia

nIc

elan

dic

Swed

ish

Dut

chD

anis

hP1

Dan

ish

P2Sc

ottis

hP1

Scot

tish

P2

971

7(3

95)

978

2(3

15)

933

3(9

84)

912

9(1

092

)90

89

(89

1)90

48

(10

99)

625

6(3

126

)87

50

(15

14)

405

6(3

665

)74

26

(22

67)

116

(03

9)1

10(0

48)

195

(11

1)1

84(1

14)

237

(19

6)2

65(0

99)

386

(28

4)1

75(1

93)

604

(34

2)2

47(1

43)

866

5(1

115

)90

87

(80

3)88

19

(20

60)

816

8(1

640

)84

44

(11

56)

740

1(2

410

)44

87

(36

09)

750

0(2

544

)17

96

(26

37)

527

8(3

348

)

197

(06

6)1

80(0

67)

286

(15

2)2

79(1

52)

386

(38

4)4

68(2

17)

518

(39

4)2

93(3

13)

735

(46

1)3

87(2

40)

919

1(6

54)

943

5(4

99)

907

6(1

486

)86

49

(11

60)

876

7(8

97)

822

4(1

672

)53

72

(31

82)

812

5(1

914

)29

26

(29

40)

635

2(2

553

)

157

(05

1)1

45(0

56)

240

(13

0)2

31(1

32)

311

(28

8)3

67(1

50)

458

(31

9)2

34(2

51)

669

(38

7)3

17(1

77)




































Lexicality effects



Sim

ple

sylla

ble

stru

ctur

es

Shallow

Deep


325 (537)556 (571)592 (820)590 (821)

plusmn 336 (1246)plusmn 586 (1599)

183 (618)

019 (031)019 (039)029 (046)014 (043)

plusmn 026 (078)plusmn 148 (547)

013 (018)

Com

plex

sylla

ble

stru

ctur

es Shallow

Deep


556 (522)337 (333)104 (490)758 (745)744 (593)

1319 (1515)1735 (1743)1132 (1294)463 (2326)

1287 (2343)

049 (022)039 (020)048 (039)019 (053)110 (127)207 (095)277 (235)129 (150)

plusmn 108 (494)099 (170)



Tab

le8

Tota

lper

cent

age

ofer

rors

prod

uced

inea

chsa

mpl

efo

rw

ord

and

nonw

ord

targ

ets

and

the

prop

ortio

nsof

erro

rre

spon

ses

whi

chca

nbe

clas

sie

das

refu

sals

wor

dsu

bstit

utio

nsn

onw

ord

subs

titut

ions

or

lett

erso

undi

ngs

Wor

der

ror

patt

erns

N

onw

ord

erro

rpa

tter

ns

Lang

uage

Erro

rR

efus

als

Wor

dsu

bsN

onw

ord

subs

Lett

erso

undi

ngEr

ror

Ref

usal

sW

ord

subs

Non

wor

dsu

bsLe

tter

soun

ding


Shal

low

D

eep

Finn

ish

Gre

ekIt

alia

nSp

anis

hPo

rtug

uese

Fren

chP1

Fren

chP2

171

240

468

528

264

620

93

077

008

006

066

035

175

185

000

099

177

132

243

102

312

35

048

063

051

066

250

120

66

420

29

000

006

197

000

241

025

000

496

795

106

011

18

231

015

06

259

012

025

000

083

066

043

087

076

158

132

222

453

135

80

29

411

612

651

812

146

91

051

05

000

000

263

000

322

000

038


Shal

low

D

eep

Aus

tria

nG

erm

anN

orw

egia

nIc

elan

dic

Swed

ish

Dut

chD

anis

hP1

Dan

ish

P2Sc

ottis

hP1

Scot

tish

P2

253

228

819

456

593

489

288

97

4366

11

236

1

005

000

000

040

060

000

128

60

9724

07

972

127

139

326

337

248

211

109

06

3216

94

100

0

122

089

014

040

143

233

205

007

731

222

000

000

479

040

143

044

077

014

177

81

67

809

565

924

177

613

51

123

346

28

187

570

74

364

8

000

010

000

040

008

089

110

70

9715

00

491

112

060

014

119

270

200

487

264

136

19

91

687

496

444

103

23

009

0024

10

149

319

17

145

4

000

000

465

585

781

056

355

000

230

67

04































































Appendix












(after)






































ukes pilo atu ojun



ides ifa ubi aru


pu edos sare


fuba uvel adol ifos



otip arol ur


ibam uki mipu etus






ovan uted alog


bano imon ekal rupa




gopa eto afen opud






















mpl

esy

llabl

est

ruct

ures

Shallow

Deep


1111112

70443840384529

787 (28)675 (27)689 (31)684 (27)701 (32)669 (32)794 (56)

Com

plex

sylla

ble

stru

ctur

es

Shallow

Deep


Scottish

111111121

2

572840372528654030

30

755 (41)740 (34)794 (28)686 (25)748 (59)697 (36)765 (38)855 (34)

CA 559 (32)RA 619 (63)CA 656 (30)RA 722 (84)













sitem

Sim

ple

sylla

ble

stru

ctur

es

Shallow

Deep


9444 (471)9640 (509)9513 (767)9555 (499)9248 (784)9122 (1083)9803 (432)

148 (055)105 (027)106 (046)103 (044)140 (050)138 (061)126 (018)

Com

plex

sylla

ble

stru

ctur

es Shallow

Deep


9685 (332)9951 (179)9865 (246)9573 (533)9874 (175)8951 (466)9492 (730)9795 (242)9397 (603)9603 (468)

074 (016)075 (019)084 (022)102 (024)136 (037)094 (027)114 (051)063 (017)188 (101)103 (027)

















Tab

le5

Acc

urac

y(p

erce

ntco

rrec

t)an

dtim

e(s

econ

dspe

rite

m)d

ata

(sta

ndar

dde

viat

ions

inpa

rent

hese

s)fo

rre

adin

glis

tsof

very

fam

iliar

wor

dsw

ithou

tcom

esfo

rco

nten

tfu

nctio

nan

dco

mbi

ned

cont

ent

and

func

tion

wor

dsfo

rea

chla

ngua

gegr

oup

Fam

iliar

wor

dre

adin

g

Con

tent

wor

dsFu

nctio

nw

ords

Com

bine

dco

nten

tfu

nctio

n

s

item

s

item

s

item


Shal

low

D

eep

Finn

ish

Gre

ekIt

alia

nSp

anis

hPo

rtug

uese

Fren

chP1

Fren

chP2

981

7(4

09)

965

9(5

90)

948

8(1

754

)94

17

(10

29)

771

9(1

945

)72

47

(27

57)

984

7(4

67)

138

(09

1)1

60(0

63)

218

(11

8)1

40(1

23)

344

(24

4)7

04(1

087

)1

00(0

29)

984

1(3

43)

986

1(2

97)

957

6(1

661

)95

28

(97

1)69

88

(20

77)

856

8(2

087

)10

000

(00

0)

113

(07

2)1

31(0

67)

123

(05

8)1

31(1

00)

300

(18

1)4

19(6

74)

064

(02

1)

982

9(3

76)

976

0(4

75)

953

2(1

697

)94

72

(99

6)73

54

(20

32)

790

7(2

354

)99

23

(24

5)

125

(08

3)1

45(0

66)

170

(10

4)1

35(1

11)

322

(21

4)5

61(9

00)

084

(02

6)


Shal

low

D

eep

Aus

tria

nG

erm

anN

orw

egia

nIc

elan

dic

Swed

ish

Dut

chD

anis

hP1

Dan

ish

P2Sc

ottis

hP1

Scot

tish

P2

970

8(4

94)

974

2(4

90)

915

3(1

655

)93

39

(91

5)93

56

(93

1)92

66

(71

0)70

94

(26

78)

922

2(1

388

)32

59

(24

83)

790

7(1

832

)

134

(06

4)1

35(0

74)

213

(13

9)2

36(1

65)

264

(25

7)2

02(0

84)

115

(10

3)1

18(1

40)

819

(60

8)2

11(1

40)

978

6(4

42)

980

2(3

77)

920

8(1

423

)94

74

(85

8)96

67

(48

1)98

21

(50

3)71

20

(29

93)

929

2(1

045

)35

18

(26

23)

737

0(2

134

)

082

(02

9)0

76(0

39)

171

(11

9)1

88(1

21)

139

(12

2)1

17(0

67)

251

(20

0)0

91(0

82)

736

(64

3)2

25(1

59)

974

7(4

68)

977

2(4

34)

918

1(1

534

)94

07

(88

4)95

11

(75

0)95

44

(67

1)71

07

(27

71)

925

7(1

174

)33

89

(24

97)

763

9(1

927

)

108

(05

6)1

06(0

66)

192

(13

0)2

12(1

46)

201

(20

9)1

60(0

87)

185

(14

2)1

05(1

09)

778

(61

9)2

18(1

41)














Tab

le6

Acc

urac

y(p

erce

ntco

rrec

t)an

dtim

e(s

econ

dspe

rite

m)

data

for

read

ing

lists

ofsi

mpl

eno

nwor

ds

with

outc

omes

for

mon

osyl

labi

cbi

sylla

bic

and

com

bine

dsy

llabl

ele

ngth

s(s

tand

ard

devi

atio

nsin

pare

nthe

ses)

for

each

lang

uage

grou

p

Sim

ple

nonw

ord

read

ing

Mon

osyl

labi

cno

nwor

dsBi

sylla

bic

nonw

ords

Com

bine

dno

nwor

ds

s

item

s

item

s

item


Shal

low

D

eep

Finn

ish

Gre

ekIt

alia

nSp

anis

hPo

rtug

uese

Fren

chP1

Fren

chP2

976

2(4

78)

968

4(5

14)

921

1(1

936

)93

06

(10

28)

763

2(2

187

)88

27

(18

43)

992

3(2

45)

108

(06

6)1

32(0

53)

158

(06

8)0

96(0

54)

224

(12

7)3

09(2

48)

086

(02

0)

924

6(8

92)

872

5(1

028

)86

70

(20

85)

845

8(1

448

)77

49

(19

63)

816

0(2

383

)95

59

(11

05)

181

(10

8)1

97(0

69)

241

(09

3)2

03(1

74)

370

(22

9)5

17(5

18)

101

(02

2)

950

4(5

77)

920

5(6

90)

894

0(1

954

)88

82

(11

78)

769

0(1

945

)84

94

(20

61)

974

2(6

38)

144

(08

6)1

64(0

60)

199

(07

7)1

50(1

12)

297

(17

3)4

13(3

79)

094

(01

8)


Shal

low

D

eep

Aus

tria

nG

erm

anN

orw

egia

nIc

elan

dic

Swed

ish

Dut

chD

anis

hP1

Dan

ish

P2Sc

ottis

hP1

Scot

tish

P2

971

7(3

95)

978

2(3

15)

933

3(9

84)

912

9(1

092

)90

89

(89

1)90

48

(10

99)

625

6(3

126

)87

50

(15

14)

405

6(3

665

)74

26

(22

67)

116

(03

9)1

10(0

48)

195

(11

1)1

84(1

14)

237

(19

6)2

65(0

99)

386

(28

4)1

75(1

93)

604

(34

2)2

47(1

43)

866

5(1

115

)90

87

(80

3)88

19

(20

60)

816

8(1

640

)84

44

(11

56)

740

1(2

410

)44

87

(36

09)

750

0(2

544

)17

96

(26

37)

527

8(3

348

)

197

(06

6)1

80(0

67)

286

(15

2)2

79(1

52)

386

(38

4)4

68(2

17)

518

(39

4)2

93(3

13)

735

(46

1)3

87(2

40)

919

1(6

54)

943

5(4

99)

907

6(1

486

)86

49

(11

60)

876

7(8

97)

822

4(1

672

)53

72

(31

82)

812

5(1

914

)29

26

(29

40)

635

2(2

553

)

157

(05

1)1

45(0

56)

240

(13

0)2

31(1

32)

311

(28

8)3

67(1

50)

458

(31

9)2

34(2

51)

669

(38

7)3

17(1

77)




































Lexicality effects



Sim

ple

sylla

ble

stru

ctur

es

Shallow

Deep


325 (537)556 (571)592 (820)590 (821)

plusmn 336 (1246)plusmn 586 (1599)

183 (618)

019 (031)019 (039)029 (046)014 (043)

plusmn 026 (078)plusmn 148 (547)

013 (018)

Com

plex

sylla

ble

stru

ctur

es Shallow

Deep


556 (522)337 (333)104 (490)758 (745)744 (593)

1319 (1515)1735 (1743)1132 (1294)463 (2326)

1287 (2343)

049 (022)039 (020)048 (039)019 (053)110 (127)207 (095)277 (235)129 (150)

plusmn 108 (494)099 (170)



Tab

le8

Tota

lper

cent

age

ofer

rors

prod

uced

inea

chsa

mpl

efo

rw

ord

and

nonw

ord

targ

ets

and

the

prop

ortio

nsof

erro

rre

spon

ses

whi

chca

nbe

clas

sie

das

refu

sals

wor

dsu

bstit

utio

nsn

onw

ord

subs

titut

ions

or

lett

erso

undi

ngs

Wor

der

ror

patt

erns

N

onw

ord

erro

rpa

tter

ns

Lang

uage

Erro

rR

efus

als

Wor

dsu

bsN

onw

ord

subs

Lett

erso

undi

ngEr

ror

Ref

usal

sW

ord

subs

Non

wor

dsu

bsLe

tter

soun

ding


Shal

low

D

eep

Finn

ish

Gre

ekIt

alia

nSp

anis

hPo

rtug

uese

Fren

chP1

Fren

chP2

171

240

468

528

264

620

93

077

008

006

066

035

175

185

000

099

177

132

243

102

312

35

048

063

051

066

250

120

66

420

29

000

006

197

000

241

025

000

496

795

106

011

18

231

015

06

259

012

025

000

083

066

043

087

076

158

132

222

453

135

80

29

411

612

651

812

146

91

051

05

000

000

263

000

322

000

038


Shal

low

D

eep

Aus

tria

nG

erm

anN

orw

egia

nIc

elan

dic

Swed

ish

Dut

chD

anis

hP1

Dan

ish

P2Sc

ottis

hP1

Scot

tish

P2

253

228

819

456

593

489

288

97

4366

11

236

1

005

000

000

040

060

000

128

60

9724

07

972

127

139

326

337

248

211

109

06

3216

94

100

0

122

089

014

040

143

233

205

007

731

222

000

000

479

040

143

044

077

014

177

81

67

809

565

924

177

613

51

123

346

28

187

570

74

364

8

000

010

000

040

008

089

110

70

9715

00

491

112

060

014

119

270

200

487

264

136

19

91

687

496

444

103

23

009

0024

10

149

319

17

145

4

000

000

465

585

781

056

355

000

230

67

04































































Appendix












(after)






































ukes pilo atu ojun



ides ifa ubi aru


pu edos sare


fuba uvel adol ifos



otip arol ur


ibam uki mipu etus






ovan uted alog


bano imon ekal rupa




gopa eto afen opud









sitem

Sim

ple

sylla

ble

stru

ctur

es

Shallow

Deep


9444 (471)9640 (509)9513 (767)9555 (499)9248 (784)9122 (1083)9803 (432)

148 (055)105 (027)106 (046)103 (044)140 (050)138 (061)126 (018)

Com

plex

sylla

ble

stru

ctur

es Shallow

Deep


9685 (332)9951 (179)9865 (246)9573 (533)9874 (175)8951 (466)9492 (730)9795 (242)9397 (603)9603 (468)

074 (016)075 (019)084 (022)102 (024)136 (037)094 (027)114 (051)063 (017)188 (101)103 (027)

















Tab

le5

Acc

urac

y(p

erce

ntco

rrec

t)an

dtim

e(s

econ

dspe

rite

m)d

ata

(sta

ndar

dde

viat

ions

inpa

rent

hese

s)fo

rre

adin

glis

tsof

very

fam

iliar

wor

dsw

ithou

tcom

esfo

rco

nten

tfu

nctio

nan

dco

mbi

ned

cont

ent

and

func

tion

wor

dsfo

rea

chla

ngua

gegr

oup

Fam

iliar

wor

dre

adin

g

Con

tent

wor

dsFu

nctio

nw

ords

Com

bine

dco

nten

tfu

nctio

n

s

item

s

item

s

item


Shal

low

D

eep

Finn

ish

Gre

ekIt

alia

nSp

anis

hPo

rtug

uese

Fren

chP1

Fren

chP2

981

7(4

09)

965

9(5

90)

948

8(1

754

)94

17

(10

29)

771

9(1

945

)72

47

(27

57)

984

7(4

67)

138

(09

1)1

60(0

63)

218

(11

8)1

40(1

23)

344

(24

4)7

04(1

087

)1

00(0

29)

984

1(3

43)

986

1(2

97)

957

6(1

661

)95

28

(97

1)69

88

(20

77)

856

8(2

087

)10

000

(00

0)

113

(07

2)1

31(0

67)

123

(05

8)1

31(1

00)

300

(18

1)4

19(6

74)

064

(02

1)

982

9(3

76)

976

0(4

75)

953

2(1

697

)94

72

(99

6)73

54

(20

32)

790

7(2

354

)99

23

(24

5)

125

(08

3)1

45(0

66)

170

(10

4)1

35(1

11)

322

(21

4)5

61(9

00)

084

(02

6)


Shal

low

D

eep

Aus

tria

nG

erm

anN

orw

egia

nIc

elan

dic

Swed

ish

Dut

chD

anis

hP1

Dan

ish

P2Sc

ottis

hP1

Scot

tish

P2

970

8(4

94)

974

2(4

90)

915

3(1

655

)93

39

(91

5)93

56

(93

1)92

66

(71

0)70

94

(26

78)

922

2(1

388

)32

59

(24

83)

790

7(1

832

)

134

(06

4)1

35(0

74)

213

(13

9)2

36(1

65)

264

(25

7)2

02(0

84)

115

(10

3)1

18(1

40)

819

(60

8)2

11(1

40)

978

6(4

42)

980

2(3

77)

920

8(1

423

)94

74

(85

8)96

67

(48

1)98

21

(50

3)71

20

(29

93)

929

2(1

045

)35

18

(26

23)

737

0(2

134

)

082

(02

9)0

76(0

39)

171

(11

9)1

88(1

21)

139

(12

2)1

17(0

67)

251

(20

0)0

91(0

82)

736

(64

3)2

25(1

59)

974

7(4

68)

977

2(4

34)

918

1(1

534

)94

07

(88

4)95

11

(75

0)95

44

(67

1)71

07

(27

71)

925

7(1

174

)33

89

(24

97)

763

9(1

927

)

108

(05

6)1

06(0

66)

192

(13

0)2

12(1

46)

201

(20

9)1

60(0

87)

185

(14

2)1

05(1

09)

778

(61

9)2

18(1

41)














Tab

le6

Acc

urac

y(p

erce

ntco

rrec

t)an

dtim

e(s

econ

dspe

rite

m)

data

for

read

ing

lists

ofsi

mpl

eno

nwor

ds

with

outc

omes

for

mon

osyl

labi

cbi

sylla

bic

and

com

bine

dsy

llabl

ele

ngth

s(s

tand

ard

devi

atio

nsin

pare

nthe

ses)

for

each

lang

uage

grou

p

Sim

ple

nonw

ord

read

ing

Mon

osyl

labi

cno

nwor

dsBi

sylla

bic

nonw

ords

Com

bine

dno

nwor

ds

s

item

s

item

s

item


Shal

low

D

eep

Finn

ish

Gre

ekIt

alia

nSp

anis

hPo

rtug

uese

Fren

chP1

Fren

chP2

976

2(4

78)

968

4(5

14)

921

1(1

936

)93

06

(10

28)

763

2(2

187

)88

27

(18

43)

992

3(2

45)

108

(06

6)1

32(0

53)

158

(06

8)0

96(0

54)

224

(12

7)3

09(2

48)

086

(02

0)

924

6(8

92)

872

5(1

028

)86

70

(20

85)

845

8(1

448

)77

49

(19

63)

816

0(2

383

)95

59

(11

05)

181

(10

8)1

97(0

69)

241

(09

3)2

03(1

74)

370

(22

9)5

17(5

18)

101

(02

2)

950

4(5

77)

920

5(6

90)

894

0(1

954

)88

82

(11

78)

769

0(1

945

)84

94

(20

61)

974

2(6

38)

144

(08

6)1

64(0

60)

199

(07

7)1

50(1

12)

297

(17

3)4

13(3

79)

094

(01

8)


Shal

low

D

eep

Aus

tria

nG

erm

anN

orw

egia

nIc

elan

dic

Swed

ish

Dut

chD

anis

hP1

Dan

ish

P2Sc

ottis

hP1

Scot

tish

P2

971

7(3

95)

978

2(3

15)

933

3(9

84)

912

9(1

092

)90

89

(89

1)90

48

(10

99)

625

6(3

126

)87

50

(15

14)

405

6(3

665

)74

26

(22

67)

116

(03

9)1

10(0

48)

195

(11

1)1

84(1

14)

237

(19

6)2

65(0

99)

386

(28

4)1

75(1

93)

604

(34

2)2

47(1

43)

866

5(1

115

)90

87

(80

3)88

19

(20

60)

816

8(1

640

)84

44

(11

56)

740

1(2

410

)44

87

(36

09)

750

0(2

544

)17

96

(26

37)

527

8(3

348

)

197

(06

6)1

80(0

67)

286

(15

2)2

79(1

52)

386

(38

4)4

68(2

17)

518

(39

4)2

93(3

13)

735

(46

1)3

87(2

40)

919

1(6

54)

943

5(4

99)

907

6(1

486

)86

49

(11

60)

876

7(8

97)

822

4(1

672

)53

72

(31

82)

812

5(1

914

)29

26

(29

40)

635

2(2

553

)

157

(05

1)1

45(0

56)

240

(13

0)2

31(1

32)

311

(28

8)3

67(1

50)

458

(31

9)2

34(2

51)

669

(38

7)3

17(1

77)




































Lexicality effects



Sim

ple

sylla

ble

stru

ctur

es

Shallow

Deep


325 (537)556 (571)592 (820)590 (821)

plusmn 336 (1246)plusmn 586 (1599)

183 (618)

019 (031)019 (039)029 (046)014 (043)

plusmn 026 (078)plusmn 148 (547)

013 (018)

Com

plex

sylla

ble

stru

ctur

es Shallow

Deep


556 (522)337 (333)104 (490)758 (745)744 (593)

1319 (1515)1735 (1743)1132 (1294)463 (2326)

1287 (2343)

049 (022)039 (020)048 (039)019 (053)110 (127)207 (095)277 (235)129 (150)

plusmn 108 (494)099 (170)



Tab

le8

Tota

lper

cent

age

ofer

rors

prod

uced

inea

chsa

mpl

efo

rw

ord

and

nonw

ord

targ

ets

and

the

prop

ortio

nsof

erro

rre

spon

ses

whi

chca

nbe

clas

sie

das

refu

sals

wor

dsu

bstit

utio

nsn

onw

ord

subs

titut

ions

or

lett

erso

undi

ngs

Wor

der

ror

patt

erns

N

onw

ord

erro

rpa

tter

ns

Lang

uage

Erro

rR

efus

als

Wor

dsu

bsN

onw

ord

subs

Lett

erso

undi

ngEr

ror

Ref

usal

sW

ord

subs

Non

wor

dsu

bsLe

tter

soun

ding


Shal

low

D

eep

Finn

ish

Gre

ekIt

alia

nSp

anis

hPo

rtug

uese

Fren

chP1

Fren

chP2

171

240

468

528

264

620

93

077

008

006

066

035

175

185

000

099

177

132

243

102

312

35

048

063

051

066

250

120

66

420

29

000

006

197

000

241

025

000

496

795

106

011

18

231

015

06

259

012

025

000

083

066

043

087

076

158

132

222

453

135

80

29

411

612

651

812

146

91

051

05

000

000

263

000

322

000

038


Shal

low

D

eep

Aus

tria

nG

erm

anN

orw

egia

nIc

elan

dic

Swed

ish

Dut

chD

anis

hP1

Dan

ish

P2Sc

ottis

hP1

Scot

tish

P2

253

228

819

456

593

489

288

97

4366

11

236

1

005

000

000

040

060

000

128

60

9724

07

972

127

139

326

337

248

211

109

06

3216

94

100

0

122

089

014

040

143

233

205

007

731

222

000

000

479

040

143

044

077

014

177

81

67

809

565

924

177

613

51

123

346

28

187

570

74

364

8

000

010

000

040

008

089

110

70

9715

00

491

112

060

014

119

270

200

487

264

136

19

91

687

496

444

103

23

009

0024

10

149

319

17

145

4

000

000

465

585

781

056

355

000

230

67

04































































Appendix












(after)






































ukes pilo atu ojun



ides ifa ubi aru


pu edos sare


fuba uvel adol ifos



otip arol ur


ibam uki mipu etus






ovan uted alog


bano imon ekal rupa




gopa eto afen opud


















Tab

le5

Acc

urac

y(p

erce

ntco

rrec

t)an

dtim

e(s

econ

dspe

rite

m)d

ata

(sta

ndar

dde

viat

ions

inpa

rent

hese

s)fo

rre

adin

glis

tsof

very

fam

iliar

wor

dsw

ithou

tcom

esfo

rco

nten

tfu

nctio

nan

dco

mbi

ned

cont

ent

and

func

tion

wor

dsfo

rea

chla

ngua

gegr

oup

Fam

iliar

wor

dre

adin

g

Con

tent

wor

dsFu

nctio

nw

ords

Com

bine

dco

nten

tfu

nctio

n

s

item

s

item

s

item


Shal

low

D

eep

Finn

ish

Gre

ekIt

alia

nSp

anis

hPo

rtug

uese

Fren

chP1

Fren

chP2

981

7(4

09)

965

9(5

90)

948

8(1

754

)94

17

(10

29)

771

9(1

945

)72

47

(27

57)

984

7(4

67)

138

(09

1)1

60(0

63)

218

(11

8)1

40(1

23)

344

(24

4)7

04(1

087

)1

00(0

29)

984

1(3

43)

986

1(2

97)

957

6(1

661

)95

28

(97

1)69

88

(20

77)

856

8(2

087

)10

000

(00

0)

113

(07

2)1

31(0

67)

123

(05

8)1

31(1

00)

300

(18

1)4

19(6

74)

064

(02

1)

982

9(3

76)

976

0(4

75)

953

2(1

697

)94

72

(99

6)73

54

(20

32)

790

7(2

354

)99

23

(24

5)

125

(08

3)1

45(0

66)

170

(10

4)1

35(1

11)

322

(21

4)5

61(9

00)

084

(02

6)


Shal

low

D

eep

Aus

tria

nG

erm

anN

orw

egia

nIc

elan

dic

Swed

ish

Dut

chD

anis

hP1

Dan

ish

P2Sc

ottis

hP1

Scot

tish

P2

970

8(4

94)

974

2(4

90)

915

3(1

655

)93

39

(91

5)93

56

(93

1)92

66

(71

0)70

94

(26

78)

922

2(1

388

)32

59

(24

83)

790

7(1

832

)

134

(06

4)1

35(0

74)

213

(13

9)2

36(1

65)

264

(25

7)2

02(0

84)

115

(10

3)1

18(1

40)

819

(60

8)2

11(1

40)

978

6(4

42)

980

2(3

77)

920

8(1

423

)94

74

(85

8)96

67

(48

1)98

21

(50

3)71

20

(29

93)

929

2(1

045

)35

18

(26

23)

737

0(2

134

)

082

(02

9)0

76(0

39)

171

(11

9)1

88(1

21)

139

(12

2)1

17(0

67)

251

(20

0)0

91(0

82)

736

(64

3)2

25(1

59)

974

7(4

68)

977

2(4

34)

918

1(1

534

)94

07

(88

4)95

11

(75

0)95

44

(67

1)71

07

(27

71)

925

7(1

174

)33

89

(24

97)

763

9(1

927

)

108

(05

6)1

06(0

66)

192

(13

0)2

12(1

46)

201

(20

9)1

60(0

87)

185

(14

2)1

05(1

09)

778

(61

9)2

18(1

41)














Tab

le6

Acc

urac

y(p

erce

ntco

rrec

t)an

dtim

e(s

econ

dspe

rite

m)

data

for

read

ing

lists

ofsi

mpl

eno

nwor

ds

with

outc

omes

for

mon

osyl

labi

cbi

sylla

bic

and

com

bine

dsy

llabl

ele

ngth

s(s

tand

ard

devi

atio

nsin

pare

nthe

ses)

for

each

lang

uage

grou

p

Sim

ple

nonw

ord

read

ing

Mon

osyl

labi

cno

nwor

dsBi

sylla

bic

nonw

ords

Com

bine

dno

nwor

ds

s

item

s

item

s

item


Shal

low

D

eep

Finn

ish

Gre

ekIt

alia

nSp

anis

hPo

rtug

uese

Fren

chP1

Fren

chP2

976

2(4

78)

968

4(5

14)

921

1(1

936

)93

06

(10

28)

763

2(2

187

)88

27

(18

43)

992

3(2

45)

108

(06

6)1

32(0

53)

158

(06

8)0

96(0

54)

224

(12

7)3

09(2

48)

086

(02

0)

924

6(8

92)

872

5(1

028

)86

70

(20

85)

845

8(1

448

)77

49

(19

63)

816

0(2

383

)95

59

(11

05)

181

(10

8)1

97(0

69)

241

(09

3)2

03(1

74)

370

(22

9)5

17(5

18)

101

(02

2)

950

4(5

77)

920

5(6

90)

894

0(1

954

)88

82

(11

78)

769

0(1

945

)84

94

(20

61)

974

2(6

38)

144

(08

6)1

64(0

60)

199

(07

7)1

50(1

12)

297

(17

3)4

13(3

79)

094

(01

8)


Shal

low

D

eep

Aus

tria

nG

erm

anN

orw

egia

nIc

elan

dic

Swed

ish

Dut

chD

anis

hP1

Dan

ish

P2Sc

ottis

hP1

Scot

tish

P2

971

7(3

95)

978

2(3

15)

933

3(9

84)

912

9(1

092

)90

89

(89

1)90

48

(10

99)

625

6(3

126

)87

50

(15

14)

405

6(3

665

)74

26

(22

67)

116

(03

9)1

10(0

48)

195

(11

1)1

84(1

14)

237

(19

6)2

65(0

99)

386

(28

4)1

75(1

93)

604

(34

2)2

47(1

43)

866

5(1

115

)90

87

(80

3)88

19

(20

60)

816

8(1

640

)84

44

(11

56)

740

1(2

410

)44

87

(36

09)

750

0(2

544

)17

96

(26

37)

527

8(3

348

)

197

(06

6)1

80(0

67)

286

(15

2)2

79(1

52)

386

(38

4)4

68(2

17)

518

(39

4)2

93(3

13)

735

(46

1)3

87(2

40)

919

1(6

54)

943

5(4

99)

907

6(1

486

)86

49

(11

60)

876

7(8

97)

822

4(1

672

)53

72

(31

82)

812

5(1

914

)29

26

(29

40)

635

2(2

553

)

157

(05

1)1

45(0

56)

240

(13

0)2

31(1

32)

311

(28

8)3

67(1

50)

458

(31

9)2

34(2

51)

669

(38

7)3

17(1

77)




































Lexicality effects



Sim

ple

sylla

ble

stru

ctur

es

Shallow

Deep


325 (537)556 (571)592 (820)590 (821)

plusmn 336 (1246)plusmn 586 (1599)

183 (618)

019 (031)019 (039)029 (046)014 (043)

plusmn 026 (078)plusmn 148 (547)

013 (018)

Com

plex

sylla

ble

stru

ctur

es Shallow

Deep


556 (522)337 (333)104 (490)758 (745)744 (593)

1319 (1515)1735 (1743)1132 (1294)463 (2326)

1287 (2343)

049 (022)039 (020)048 (039)019 (053)110 (127)207 (095)277 (235)129 (150)

plusmn 108 (494)099 (170)



Tab

le8

Tota

lper

cent

age

ofer

rors

prod

uced

inea

chsa

mpl

efo

rw

ord

and

nonw

ord

targ

ets

and

the

prop

ortio

nsof

erro

rre

spon

ses

whi

chca

nbe

clas

sie

das

refu

sals

wor

dsu

bstit

utio

nsn

onw

ord

subs

titut

ions

or

lett

erso

undi

ngs

Wor

der

ror

patt

erns

N

onw

ord

erro

rpa

tter

ns

Lang

uage

Erro

rR

efus

als

Wor

dsu

bsN

onw

ord

subs

Lett

erso

undi

ngEr

ror

Ref

usal

sW

ord

subs

Non

wor

dsu

bsLe

tter

soun

ding


Shal

low

D

eep

Finn

ish

Gre

ekIt

alia

nSp

anis

hPo

rtug

uese

Fren

chP1

Fren

chP2

171

240

468

528

264

620

93

077

008

006

066

035

175

185

000

099

177

132

243

102

312

35

048

063

051

066

250

120

66

420

29

000

006

197

000

241

025

000

496

795

106

011

18

231

015

06

259

012

025

000

083

066

043

087

076

158

132

222

453

135

80

29

411

612

651

812

146

91

051

05

000

000

263

000

322

000

038


Shal

low

D

eep

Aus

tria

nG

erm

anN

orw

egia

nIc

elan

dic

Swed

ish

Dut

chD

anis

hP1

Dan

ish

P2Sc

ottis

hP1

Scot

tish

P2

253

228

819

456

593

489

288

97

4366

11

236

1

005

000

000

040

060

000

128

60

9724

07

972

127

139

326

337

248

211

109

06

3216

94

100

0

122

089

014

040

143

233

205

007

731

222

000

000

479

040

143

044

077

014

177

81

67

809

565

924

177

613

51

123

346

28

187

570

74

364

8

000

010

000

040

008

089

110

70

9715

00

491

112

060

014

119

270

200

487

264

136

19

91

687

496

444

103

23

009

0024

10

149

319

17

145

4

000

000

465

585

781

056

355

000

230

67

04































































Appendix












(after)






































ukes pilo atu ojun



ides ifa ubi aru


pu edos sare


fuba uvel adol ifos



otip arol ur


ibam uki mipu etus






ovan uted alog


bano imon ekal rupa




gopa eto afen opud















Tab

le6

Acc

urac

y(p

erce

ntco

rrec

t)an

dtim

e(s

econ

dspe

rite

m)

data

for

read

ing

lists

ofsi

mpl

eno

nwor

ds

with

outc

omes

for

mon

osyl

labi

cbi

sylla

bic

and

com

bine

dsy

llabl

ele

ngth

s(s

tand

ard

devi

atio

nsin

pare

nthe

ses)

for

each

lang

uage

grou

p

Sim

ple

nonw

ord

read

ing

Mon

osyl

labi

cno

nwor

dsBi

sylla

bic

nonw

ords

Com

bine

dno

nwor

ds

s

item

s

item

s

item


Shal

low

D

eep

Finn

ish

Gre

ekIt

alia

nSp

anis

hPo

rtug

uese

Fren

chP1

Fren

chP2

976

2(4

78)

968

4(5

14)

921

1(1

936

)93

06

(10

28)

763

2(2

187

)88

27

(18

43)

992

3(2

45)

108

(06

6)1

32(0

53)

158

(06

8)0

96(0

54)

224

(12

7)3

09(2

48)

086

(02

0)

924

6(8

92)

872

5(1

028

)86

70

(20

85)

845

8(1

448

)77

49

(19

63)

816

0(2

383

)95

59

(11

05)

181

(10

8)1

97(0

69)

241

(09

3)2

03(1

74)

370

(22

9)5

17(5

18)

101

(02

2)

950

4(5

77)

920

5(6

90)

894

0(1

954

)88

82

(11

78)

769

0(1

945

)84

94

(20

61)

974

2(6

38)

144

(08

6)1

64(0

60)

199

(07

7)1

50(1

12)

297

(17

3)4

13(3

79)

094

(01

8)


Shal

low

D

eep

Aus

tria

nG

erm

anN

orw

egia

nIc

elan

dic

Swed

ish

Dut

chD

anis

hP1

Dan

ish

P2Sc

ottis

hP1

Scot

tish

P2

971

7(3

95)

978

2(3

15)

933

3(9

84)

912

9(1

092

)90

89

(89

1)90

48

(10

99)

625

6(3

126

)87

50

(15

14)

405

6(3

665

)74

26

(22

67)

116

(03

9)1

10(0

48)

195

(11

1)1

84(1

14)

237

(19

6)2

65(0

99)

386

(28

4)1

75(1

93)

604

(34

2)2

47(1

43)

866

5(1

115

)90

87

(80

3)88

19

(20

60)

816

8(1

640

)84

44

(11

56)

740

1(2

410

)44

87

(36

09)

750

0(2

544

)17

96

(26

37)

527

8(3

348

)

197

(06

6)1

80(0

67)

286

(15

2)2

79(1

52)

386

(38

4)4

68(2

17)

518

(39

4)2

93(3

13)

735

(46

1)3

87(2

40)

919

1(6

54)

943

5(4

99)

907

6(1

486

)86

49

(11

60)

876

7(8

97)

822

4(1

672

)53

72

(31

82)

812

5(1

914

)29

26

(29

40)

635

2(2

553

)

157

(05

1)1

45(0

56)

240

(13

0)2

31(1

32)

311

(28

8)3

67(1

50)

458

(31

9)2

34(2

51)

669

(38

7)3

17(1

77)




































Lexicality effects



Sim

ple

sylla

ble

stru

ctur

es

Shallow

Deep


325 (537)556 (571)592 (820)590 (821)

plusmn 336 (1246)plusmn 586 (1599)

183 (618)

019 (031)019 (039)029 (046)014 (043)

plusmn 026 (078)plusmn 148 (547)

013 (018)

Com

plex

sylla

ble

stru

ctur

es Shallow

Deep


556 (522)337 (333)104 (490)758 (745)744 (593)

1319 (1515)1735 (1743)1132 (1294)463 (2326)

1287 (2343)

049 (022)039 (020)048 (039)019 (053)110 (127)207 (095)277 (235)129 (150)

plusmn 108 (494)099 (170)



Tab

le8

Tota

lper

cent

age

ofer

rors

prod

uced

inea

chsa

mpl

efo

rw

ord

and

nonw

ord

targ

ets

and

the

prop

ortio

nsof

erro

rre

spon

ses

whi

chca

nbe

clas

sie

das

refu

sals

wor

dsu

bstit

utio

nsn

onw

ord

subs

titut

ions

or

lett

erso

undi

ngs

Wor

der

ror

patt

erns

N

onw

ord

erro

rpa

tter

ns

Lang

uage

Erro

rR

efus

als

Wor

dsu

bsN

onw

ord

subs

Lett

erso

undi

ngEr

ror

Ref

usal

sW

ord

subs

Non

wor

dsu

bsLe

tter

soun

ding


Shal

low

D

eep

Finn

ish

Gre

ekIt

alia

nSp

anis

hPo

rtug

uese

Fren

chP1

Fren

chP2

171

240

468

528

264

620

93

077

008

006

066

035

175

185

000

099

177

132

243

102

312

35

048

063

051

066

250

120

66

420

29

000

006

197

000

241

025

000

496

795

106

011

18

231

015

06

259

012

025

000

083

066

043

087

076

158

132

222

453

135

80

29

411

612

651

812

146

91

051

05

000

000

263

000

322

000

038


Shal

low

D

eep

Aus

tria

nG

erm

anN

orw

egia

nIc

elan

dic

Swed

ish

Dut

chD

anis

hP1

Dan

ish

P2Sc

ottis

hP1

Scot

tish

P2

253

228

819

456

593

489

288

97

4366

11

236

1

005

000

000

040

060

000

128

60

9724

07

972

127

139

326

337

248

211

109

06

3216

94

100

0

122

089

014

040

143

233

205

007

731

222

000

000

479

040

143

044

077

014

177

81

67

809

565

924

177

613

51

123

346

28

187

570

74

364

8

000

010

000

040

008

089

110

70

9715

00

491

112

060

014

119

270

200

487

264

136

19

91

687

496

444

103

23

009

0024

10

149

319

17

145

4

000

000

465

585

781

056

355

000

230

67

04































































Appendix












(after)






































ukes pilo atu ojun



ides ifa ubi aru


pu edos sare


fuba uvel adol ifos



otip arol ur


ibam uki mipu etus






ovan uted alog


bano imon ekal rupa




gopa eto afen opud





































Lexicality effects



Sim

ple

sylla

ble

stru

ctur

es

Shallow

Deep


325 (537)556 (571)592 (820)590 (821)

plusmn 336 (1246)plusmn 586 (1599)

183 (618)

019 (031)019 (039)029 (046)014 (043)

plusmn 026 (078)plusmn 148 (547)

013 (018)

Com

plex

sylla

ble

stru

ctur

es Shallow

Deep


556 (522)337 (333)104 (490)758 (745)744 (593)

1319 (1515)1735 (1743)1132 (1294)463 (2326)

1287 (2343)

049 (022)039 (020)048 (039)019 (053)110 (127)207 (095)277 (235)129 (150)

plusmn 108 (494)099 (170)



Tab

le8

Tota

lper

cent

age

ofer

rors

prod

uced

inea

chsa

mpl

efo

rw

ord

and

nonw

ord

targ

ets

and

the

prop

ortio

nsof

erro

rre

spon

ses

whi

chca

nbe

clas

sie

das

refu

sals

wor

dsu

bstit

utio

nsn

onw

ord

subs

titut

ions

or

lett

erso

undi

ngs

Wor

der

ror

patt

erns

N

onw

ord

erro

rpa

tter

ns

Lang

uage

Erro

rR

efus

als

Wor

dsu

bsN

onw

ord

subs

Lett

erso

undi

ngEr

ror

Ref

usal

sW

ord

subs

Non

wor

dsu

bsLe

tter

soun

ding


Shal

low

D

eep

Finn

ish

Gre

ekIt

alia

nSp

anis

hPo

rtug

uese

Fren

chP1

Fren

chP2

171

240

468

528

264

620

93

077

008

006

066

035

175

185

000

099

177

132

243

102

312

35

048

063

051

066

250

120

66

420

29

000

006

197

000

241

025

000

496

795

106

011

18

231

015

06

259

012

025

000

083

066

043

087

076

158

132

222

453

135

80

29

411

612

651

812

146

91

051

05

000

000

263

000

322

000

038


Shal

low

D

eep

Aus

tria

nG

erm

anN

orw

egia

nIc

elan

dic

Swed

ish

Dut

chD

anis

hP1

Dan

ish

P2Sc

ottis

hP1

Scot

tish

P2

253

228

819

456

593

489

288

97

4366

11

236

1

005

000

000

040

060

000

128

60

9724

07

972

127

139

326

337

248

211

109

06

3216

94

100

0

122

089

014

040

143

233

205

007

731

222

000

000

479

040

143

044

077

014

177

81

67

809

565

924

177

613

51

123

346

28

187

570

74

364

8

000

010

000

040

008

089

110

70

9715

00

491

112

060

014

119

270

200

487

264

136

19

91

687

496

444

103

23

009

0024

10

149

319

17

145

4

000

000

465

585

781

056

355

000

230

67

04































































Appendix












(after)






































ukes pilo atu ojun



ides ifa ubi aru


pu edos sare


fuba uvel adol ifos



otip arol ur


ibam uki mipu etus






ovan uted alog


bano imon ekal rupa




gopa eto afen opud



Tab

le8

Tota

lper

cent

age

ofer

rors

prod

uced

inea

chsa

mpl

efo

rw

ord

and

nonw

ord

targ

ets

and

the

prop

ortio

nsof

erro

rre

spon

ses

whi

chca

nbe

clas

sie

das

refu

sals

wor

dsu

bstit

utio

nsn

onw

ord

subs

titut

ions

or

lett

erso

undi

ngs

Wor

der

ror

patt

erns

N

onw

ord

erro

rpa

tter

ns

Lang

uage

Erro

rR

efus

als

Wor

dsu

bsN

onw

ord

subs

Lett

erso

undi

ngEr

ror

Ref

usal

sW

ord

subs

Non

wor

dsu

bsLe

tter

soun

ding


Shal

low

D

eep

Finn

ish

Gre

ekIt

alia

nSp

anis

hPo

rtug

uese

Fren

chP1

Fren

chP2

171

240

468

528

264

620

93

077

008

006

066

035

175

185

000

099

177

132

243

102

312

35

048

063

051

066

250

120

66

420

29

000

006

197

000

241

025

000

496

795

106

011

18

231

015

06

259

012

025

000

083

066

043

087

076

158

132

222

453

135

80

29

411

612

651

812

146

91

051

05

000

000

263

000

322

000

038


Shal

low

D

eep

Aus

tria

nG

erm

anN

orw

egia

nIc

elan

dic

Swed

ish

Dut

chD

anis

hP1

Dan

ish

P2Sc

ottis

hP1

Scot

tish

P2

253

228

819

456

593

489

288

97

4366

11

236

1

005

000

000

040

060

000

128

60

9724

07

972

127

139

326

337

248

211

109

06

3216

94

100

0

122

089

014

040

143

233

205

007

731

222

000

000

479

040

143

044

077

014

177

81

67

809

565

924

177

613

51

123

346

28

187

570

74

364

8

000

010

000

040

008

089

110

70

9715

00

491

112

060

014

119

270

200

487

264

136

19

91

687

496

444

103

23

009

0024

10

149

319

17

145

4

000

000

465

585

781

056

355

000

230

67

04































































Appendix












(after)






































ukes pilo atu ojun



ides ifa ubi aru


pu edos sare


fuba uvel adol ifos



otip arol ur


ibam uki mipu etus






ovan uted alog


bano imon ekal rupa




gopa eto afen opud
































































Appendix












(after)






































ukes pilo atu ojun



ides ifa ubi aru


pu edos sare


fuba uvel adol ifos



otip arol ur


ibam uki mipu etus






ovan uted alog


bano imon ekal rupa




gopa eto afen opud










(after)






































ukes pilo atu ojun



ides ifa ubi aru


pu edos sare


fuba uvel adol ifos



otip arol ur


ibam uki mipu etus






ovan uted alog


bano imon ekal rupa




gopa eto afen opud





























ukes pilo atu ojun



ides ifa ubi aru


pu edos sare


fuba uvel adol ifos



otip arol ur


ibam uki mipu etus






ovan uted alog


bano imon ekal rupa




gopa eto afen opud



otip arol ur


ibam uki mipu etus






ovan uted alog


bano imon ekal rupa




gopa eto afen opud


Foundation literacy acquisition in European orthographies · Foundation literacy acquisition in...

Documents

Transcript of Foundation literacy acquisition in European orthographies · Foundation literacy acquisition in...