Scrapbooking

Morphological Structure and Hemispheric Functioning: The Contribution of the Right Hemisphere to Reading in Different Languages

Description
This study examined the relationship between morphological structure of languages and performance asymmetries of native speakers in lateralized tasks. In 2 experiments, native speakers of English (concatenative morphology stem plus affix) and of
Categories
Published
of 15
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
  Morphological Structure and Hemispheric Functioning: The Contribution of the Right Hemisphere to Reading in Different Languages Zohar Eviatar University of Haifa Raphiq Ibrahim University of Haifa and Rambam Medical Center This study examined the relationship between morphological structure of languages and performanceasymmetries of native speakers in lateralized tasks. In 2 experiments, native speakers of English(concatenative morphology stem plus affix) and of Hebrew and Arabic (nonconcatenative root plusword-form morphology) were presented with lateralized lexical decision tasks, in which the morpho-logical structure of both words and nonwords was manipulated. In the 1st study, stimuli were presentedunilaterally. In the 2nd study, 2 stimuli were presented bilaterally, and participants were cued to respondto 1 of them. Three different indexes of hemispheric integration were tested: processing dissociation,effects of distractor status, and the bilateral effect. Lateralization patterns in the 3 languages revealed bothcommon and language-specific patterns. For English speakers, only the left hemisphere (LH) wassensitive to morphological structure, consistent with the hypothesis that the LH processes right visualfield stimuli independently but that the right hemisphere uses LH abilities to process words in the leftvisual field. In Hebrew and Arabic, both hemispheres are sensitive to morphological structure, andinterhemispheric transfer of information may be more symmetrical than in English. The relationshipbetween universal and experience-specific effects on brain organization is discussed. Keywords:  language morphology, Hebrew, Arabic, lexical decision, interhemispheric integration Half a century of intensive research has revealed that althoughthe left hemisphere (LH) is dominant for most language functionsin 99% of right-handers and in the majority (65%) of left-handers,there is widespread variability in the linguistic abilities of the righthemisphere (RH). This variability has been noted in studies of thedisconnected RH in split-brain patients (see Zaidel, 1998, for areview); in case reports of the reversal of laterality patterns inaphasia; and in cases of recovery of function believed to reflect RHcompensation for a severely damaged LH (Kinsbourne, 1998). Ithas been suggested (e.g., Kinsbourne, 1998) that the languagepotential of the RH may be genetically underdetermined and thatit may be subject to factors in embryogenesis or postnatal experi-ence. Much research has been done looking at the effects of biological variables, such as sex and handedness (e.g., Eviatar,Hellige, & Zaidel, 1997), on individual differences in patterns of hemispheric specialization. The research reported here focuses onthe effects of a specific kind of postnatal experience, specifically,on the structure of the language that the individual uses.The last few years have seen a large number of studies exploringRH involvement in the processing of discourse aspects of lan-guage. Much of this interest is based on the findings from meta-bolic imaging studies that have revealed extensive RH activationswhile people are performing linguistic tasks (see Bookheimer,2002, for a review). Faust et al. have examined RH involvement insyntactic processes (e.g., Faust, Bar-Lev, & Chiarello, 2003). Thegeneral findings have been that although the RH is sensitive tomessage-level processes, all the effects of syntactic manipulationsare stronger in the LH. The role of syntactic processes is tocompute the structural and thematic relationships between thewords in the sentence, so it makes sense to assume that the morehighly a language is inflected, the closer the interaction of thesesyntactic processes is with lexical representation. Thus, the mannerin which words are represented in the mental lexicon has impli-cations for the mechanisms involved in comprehension of themessage level of sentences. When these underlying representationsdiffer among languages, is it the case that the relative involvementof the cerebral hemispheres differs as well?Brain imaging studies examining the processing of morpholog-ical aspects of language performance have tended to focus onregions of interest in the LH (e.g., Lehtonen, Vorobyev, Hugdahl,Tuokkola, & Laine, 2006). Studies examining Indo-European lan-guages have focused primarily on the representation of morpho-logical knowledge, testing dual-route models versus connectionist,single-route representations in English (e.g., Devlin, Jamison, Mat-thews, & Gonnerman, 2004; Joanisse & Seidenberg, 2005; Tyler,Stamatakis, Post, Randall, & Marslen-Wilson, 2005; Vannest,Polk, & Lewis, 2005), or on the separability of morphologicalprocessing from other aspects of word processing in morphologi-cally richer languages, such as German (e.g., Longoni, Grande,Hendrich, Kastrai, & Huber, 2005) and Italian (Marangolo, Piras,Galati, & Burani, 2004). Studies examining morphological pro-cessing in non-Indo-European languages with rich morphologicalstructures have shown RH involvement in word perception, thoughthe RH was not sensitive to inflectional morphological manipula- Zohar Eviatar, Psychology Department and Institute for InformationProcessing and Decision Making, University of Haifa, Haifa, Israel; Ra-phiq Ibrahim, Learning Disabilities Department, University of Haifa, andCognitive Neurology Unit, Rambam Medical Center, Haifa, Israel.This research was supported by Grant 53/2000 from the Israel Founda-tion Trustees to Zohar Eviatar and Raphiq Ibrahim.Correspondence concerning this article should be addressed to ZoharEviatar, Psychology Department and Institute of Information Processingand Decision Making, Haifa University, Haifa 31905, Israel. E-mail:zohare@research.haifa.ac.il Neuropsychology Copyright 2007 by the American Psychological Association2007, Vol. 21, No. 4, 470–484 0894-4105/07/$12.00 DOI: 10.1037/0894-4105.21.4.470 470  tion, in Hebrew (Palti, Ben-Shachar, Hendler, & Hadar, 2007), andwas less sensitive than the LH, in Finnish (Laine, Rinne, Krause,Teras, & Sipila, 1999).Examinations of the effects of morphology in lateralized para-digms with native speakers of English (Burgess & Skodis, 1993)and French (Koenig, Wetzel, & Caramazza, 1992) have suggestedthat only the LH is sensitive to the morphological structure of words. However, in Finnish, which is not an Indo-European lan-guage and is highly inflected, Laine and Koivisto (1998) found nointeraction of performance asymmetries with word morphology.They concluded that both hemispheres of Finnish readers arecapable of morpheme-based lexical access, but that this mecha-nism is more accurate in the LH. A number of studies haveexplored the underlying representation of words in the lexicon of Semitic languages and have proposed that this representation isdifferent from that of words in Indo-European languages (Feld-man, Frost, & Pnini, 1995; Prunet, Beland, & Idrissi, 2000). Thepresent experiments were part of a research plan that takes advan-tage of some principled differences between Hebrew, Arabic, andEnglish to explore a componential analysis of the reading processin different languages and the effects of different components(visual, orthographic, and morphological demands) on the hemi-spheric division of labor in reading. Language-Specific Hemispheric Effects Arabic, Hebrew, and English use alphabetic orthographies butdiffer in interesting ways. English is read from left to right,whereas both Arabic and Hebrew are read from right to left.Previously, we have shown that this difference in directionality,affecting lifelong reading habits, has implications for the efficiencywith which skilled readers can ignore information on the side fromwhich reading usually begins (Eviatar, 1995) and for performanceasymmetries in tasks thought to reflect RH dominance (Eviatar,1997). Examination of the differences in orthography and phonol-ogy relations among the languages, together with the languageexperience of the participants, revealed (a) that strategies of pho-nological encoding that are specific to an orthography seem also tobe used while reading a second language (Eviatar, 1999) and (b)that the processing of Arabic orthography seems to make differentdemands on the cognitive system both in beginning (Ibrahim,Eviatar, & Aharon-Peretz, in press) and in skilled readers (Ibrahim,Eviatar, & Aharon-Peretz, 2002). We have suggested that this isbecause Arabic orthography specifically disallows the involvementof the RH in letter identification, even though the RH of the sameparticipants does contribute to this process in English and inHebrew (Eviatar & Ibrahim, 2004; Eviatar, Ibrahim & Ganayim,2004). The grapheme–phoneme relationship in Arabic is verycomplex, with each phoneme represented by three or four differentgraphemes, depending on where it appears in the word (beginning,middle, or end) and whether it follows a connecting letter or not.Very different phonemes (such as /t/ and /b/) are represented by thesame basic shape, differing only in the number and placement of dots (e.g., /t/ and /b/ ).We have shown that native Arabic readers respond equivalentlyto Hebrew and to Arabic when letters are presented in the rightvisual field (RVF) but make selectively more errors in Arabic thanin Hebrew when letters are presented in the left visual field (LVF).In fact, they make as many errors as participants who cannot readArabic.The focus of the present research is differences between thelanguages in morphological structure. In English, which has aconcatenative morphology, multimorphemic words are usuallycreated by affixation, a process in which the stem is usually a worditself, and its orthographic integrity is largely preserved. Arabicand Hebrew are characterized by a nonconcatenative, highly pro-ductive derivational morphology (Berman, 1978). Most words arederived by embedding a root into a morphophonological wordpattern. In both languages, most words are based on a trilateral rootand various derivatives that are formed by the addition of affixesand vowels. The roots and phonological patterns are abstractentities, and only their joint combination forms specific words. Thecore meaning is conveyed by the root, whereas the phonologicalpattern conveys word-class information. For example, in Arabic,the word  takreem  consists of the root  krm  (whose semantic spaceincludes things having to do with respect) and the phonologicalpattern ta_ _ ee_. The combination results in the word  honor  . InHebrew, the word  sifra  consists of the root  sfr   (whose semanticspace includes things having to do with counting) and the phono-logical pattern _i_ _a, which tends to occur in words denotingsingular feminine nouns, resulting in the word  numeral.  The lettersthat make up the root may be dispersed across the word, interdig-itated with letters that can double as vowels and other consonantsthat belong to the morphological pattern.A number of psycholinguistic studies (Berent, 2002; Deutsch,Frost, & Forster, 1998; Feldman et al., 1995; Frost & Bentin, 1992;Frost, Forster, & Deutsch, 1997) have explored the effects of themorphology and orthography of Hebrew on lexical access and thestructure of the mental lexicon. Two conclusions from these stud-ies are especially relevant to the present study. The first is that thenonconcatenative and agglutinative morphological structure of He-brew, together with the distributional properties of abstract wordforms, results in the inclusion of subword morphological units inthe mental lexicon of Hebrew speakers. The second is that mor-phologically complex Hebrew words cannot be read via incremen-tal parsing (sequentially, letter by letter). This last claim convergeswith the conclusions of Eviatar (1999, Experiment 4), who showedthat nonwords are processed sequentially in both visual fields byEnglish speakers but in neither visual field by Hebrew speakersand hypothesized that this is because Hebrew nonwords cannot beread sequentially. Farid and Grainger (1996) suggested the samefor the reading of Arabic. They showed that initial fixation positionin a word results in somewhat different response patterns in French(which is similar to English in morphological structure and inreading direction) and in Arabic (which is similar to Hebrew inmorphological structure and in reading direction). In French, fix-ation slightly to the left of the word’s center results in bestrecognition for both prefixed and suffixed words, whereas inArabic, prefixed words result in best recognition from leftwardfixations, and suffixed words result in best recognition from right-ward initial fixations. They suggested that this pattern is due to thegreater importance of morphological structure in Arabic, because“much of the phonological representation of the word can berecovered only after successfully matching the consonant cluster toa lexical representation” (p. 364), that is, after extraction of theroot. Berent (2002) has also concluded that in Hebrew, “speakers 471 LANGUAGE STRUCTURE AND HEMISPHERIC ABILITY  decompose the root from the word pattern in on-line word identi-fication” (p. 335). Prunet et al. (2000) reported a case study of anArabic–French agrammatic patient who showed identical deficitsin the two languages, except for a specific type of error, metathesis,in which he modified the order of the root consonants, with thevowel patterns remaining intact, in Arabic and not in French. Theyinterpreted this finding as reflecting the manner in which words arestored in the mental lexicon in the two languages: whole wordsplus affixes in French and roots plus word patterns in Arabic. The Present Study The hypotheses tested here revolved around two related foci.The first focus is the relationship between the morphologicalstructure of a language and performance asymmetries in a lateral-ized lexical decision task. The second focus is three indexes of hemispheric functioning that are informative about interhemi-spheric integration. In the two experiments described below, nativespeakers of English, Hebrew, and Arabic were presented with alateralized lexical decision task, in which the morphological struc-ture of both words and nonwords was systematically manipulated.In the first study, stimuli were presented unilaterally (in one visualfield on each trial). In the second study, two stimuli were presentedbilaterally, and participants were cued to respond to one of them.This design allowed us to test three different indexes of hemi-spheric integration: the processing dissociation, the effects of distractor status, and the bilateral effect. These are described indetail below. The Processing Dissociation The processing dissociation is an interaction between a stimulusvariable (in our case, lexicality or morphological complexity) andthe visual field to which it was presented. The interpretation of such a pattern rests on the reasoning that if the stimulus variableaffects responses in one visual field and not the other, we haveevidence for different and independent processes in the two hemi-spheres. Thus, effects of stimulus lexicality (word vs. nonword) ormorphological complexity, demonstrated in one visual field but notthe other, will support the hypothesis that each stimulus is pro-cessed by the hemisphere contralateral to the stimulated visualfield. For example, if  Koenig et al. (1992) were correct in suggest-ing that only the LH is sensitive to morphological structure, weshould see the effects of this variable in the RVF but not in theLVF.  Distractor Status In the second study, we used bilateral versions of the lexicaldecision task. Here, participants are presented with two stimuli oneach trial and are required to ignore one and to make a lexicaldecision on the other. We can examine the effects of the lexical ormorphological effect of the distractor (the stimulus to be ignored)on the response to the target. The logic is the following: If LVFstimuli are processed independently by the RH, then the lexicalstatus or morphological complexity of the distractor presented tothe LH should not affect performance. However, if the RH drawsupon LH resources to perform the task, we will see an effect of thedistractors. Iacoboni and Zaidel (1996) found such an effect inEnglish for words but not for nonwords. They concluded that thisis evidence that the RH can reject nonwords independently but thatit draws upon LH processes to accept words. The Bilateral Effect  The third measure of interhemispheric integration results fromcomparison of the unilateral and the bilateral conditions in equiv-alent language groups. Boles (1990) reported that performanceasymmetries are larger when stimuli are presented bilaterally (witha cue marking the side to which to respond) than when they arepresented unilaterally. He proposed that this “bilateral effect”occurs because bilateral presentation of different stimuli to homol-ogous areas of the two hemispheres disrupts communication be-tween them. Iacoboni and Zaidel (1996) have suggested that thedegree of the bilateral effect allows assessment of interhemisphericinteractions for different types of stimuli. That is, processes thatare performed independently by each hemisphere should not resultin different performance asymmetries with either unilateral orbilateral presentations, whereas processes that require interhemi-spheric cooperation should result in larger performance asymme-tries with bilateral presentation.This comparison will also allow a test of the hypothesis thatreading scanning habits have a larger effect on performance asym-metries with bilateral presentation than with unilateral presentation(Eviatar, 1995). Two hypotheses resulted in contradictory predic-tions. Iacoboni and Zaidel’s (1996) hypothesis predicted that thesize of the bilateral effect will vary with the other indexes of hemispheric involvement, when conditions in which there is hemi-spheric independence will not show a difference, whereas condi-tions suggesting interhemispheric cooperation will show this ef-fect. Eviatar’s (1995) hypothesis predicted that Hebrew and Arabictests will result in larger performance asymmetries with bilateralthan with unilateral presentations across all of the conditions. Thishypothesis was based on findings that when two different stimuliare presented bilaterally and participants are asked to ignore one of them, Hebrew readers find it harder to ignore a stimulus in theRVF than one in the LVF, and English readers show the oppositepattern. This finding was interpreted as reflecting an automatictransfer of attention to the side at which reading usually begins (theright for Hebrew readers and left for English readers) and thenecessity to disengage from the distractor before processing thetarget in the other visual field. Thus, Hebrew readers and Arabicreaders should show a larger RVF advantage with bilateral pre-sentations, because targets in their LVF will only be processedafter disengagement from the RVF. Study 1: Unilateral Presentations of the Lexical DecisionTask   Method Participants The participants were 60 students at Haifa University, 20 in eachnative language group. The native English speakers were recruitedfrom the summer overseas program. All were American and werepaid for their participation. The native Hebrew and Arabic speak-ers were all students at Haifa University. Most of them completed 472  EVIATAR AND IBRAHIM  the experiments for course credit, with some receiving paymentinstead. All were right-handed, were neurologically normal, andhad normal or corrected vision. Stimuli We compiled three lists of 80 words and 80 nonwords, one ineach language. Of the words, 40 were morphologically simple,and 40 were morphologically complex. In Hebrew and Arabic,morphological complexity was operationalized as the transparencyof the root, such that words derived from generative roots wereconsidered complex, and words in which the roots are not gener-ative (they appear only in that form) were considered simple. 1 InEnglish, morphologically complex words were derivations.The lists were equated on the average frequency of the words,number of letters, and initial letters. For the English list, we usedKucera and Francis’s (1967) frequency counts; the mean frequen-cies of the complex words and the simple words were 37.02and 38.6, respectively. There are no frequency norms in Hebrewand Arabic, and we did not pretest our particular lists. However,none of the stimuli are infrequent or uncommon words. Nonwordswere also morphologically manipulated. Morphologically simplenonwords in English were derived from the simple words bychanging one or two letters, and complex nonwords were illegalcombinations of real morphemes (e.g., gapty). In Hebrew andArabic, morphologically simple nonwords were derived from thesimple words by changing one or two letters, and complex non-words were constructed from nonexistent roots embedded in legalword forms. The stimuli are listed in the Appendix. Procedure The participants were tested individually. The stimuli werepresented on a Silicon Graphics Workstation. On each trial, thesequence of events was the following: A 1,000-Hz tone soundedfor 100 ms to alert the participant that the trial was beginning.Then the fixation cross was presented for 100 ms. The stimuli werepresented for 180 ms horizontally, with their inner edge 2 o of visual angle offset from fixation, either in the LVF or the RVF.English stimuli were in the Times New Roman font, the Hebrewstimuli were in Guttman Miryam, and the Arabic stimuli were inMCS Madinah S_U normal, all point size 22. The longest wordssubtended 2.5 o of visual angle. All appeared as white letters on agray screen. The stimuli were followed by a pattern mask thatremained onscreen until the participant had responded or 3 s hadpassed. The screen was blank for 2 s, and then the next trial began.The participant responded on the keyboard by pressing the uparrow if the stimulus was a real word and pressing the down arrowif it was not.  Results We measured median reaction time (RT) and percentage errors.Trials in which RT was shorter than 100 ms or longer than 3,000ms were excluded from analysis. None of the groups showedspeed–accuracy trade-offs. For Hebrew and Arabic speakers, therewas no correlation between response time and error (  p  .05). ForEnglish speakers, the correlation was  r  (19)  .23,  p  .05. This isa positive relationship, showing that fast responders made fewererrors.Because error rates in some conditions were high, we performeda signal-detection analysis to make sure that participants weresensitive to the difference between words and nonwords in all of the conditions. We computed the sensitivity measure  d    as thedifference between the  z  scores for the probability of hits (forwords) and for false alarms (for nonwords) and computed thecriterion as  c    z [  p (hits)]    z [  p ( FA )]/2. We used correctioncomputations for probability values of 1 (later changed to 1/[2  N  ])and 0 (later changed to 1  1/[2  N  ]) on the basis of suggestions of Macmillan and Creelman (1990). These data are presented inFigure 1. It can be seen that all of the participants showed highersensitivity to the lexicality status of the stimuli in the RVF than inthe LVF: Arabic speakers,  F  (1, 19)    8.04,  p    .05,    p 2   .29;Hebrew speakers,  F  (1, 19)    8.76,  p    .01,    p 2   .32; Englishspeakers,  F  (1, 19)  23.58,  p  .001,   p2  .55. All of the mean d    scores were significantly larger than 0. Thus, all of the partic-ipants showed a pattern indicating LH specialization for this lin-guistic task and better-than-chance performance in all of the con-ditions. The criterion measure shows that Arabic speakers had asignificant “yes” bias in the RVF, whereas their responses in theLVF were unbiased. Hebrew speakers showed a “no” bias in theRVF for all stimuli and in the LVF for morphologically simplestimuli. English speakers were unbiased in all of the conditions.  Native Language Effects Separate three-way analyses of variance were performed on themedian RTs and percentage errors of each language group. The 1 This is based on the finding of Feldman, Frost, and Pnini (1995) thatskilled Hebrew readers can detach a word form more quickly from agenerative root than from a nongenerative root. That is, they can divide theword into its root and word-form morphemes more easily when the root isgenerative than when it is not. d' measures 00.511.522.53complex simple complex simple complex simple Arabic Hebrew English        d       ' LVFRVF response criterion -0.5-0.4-0.3-0.2-0.100.10.20.3 complex simple complex simple complex simple Arabic Hebrew English   c LVFRVF Figure 1.  Signal detection measures in Experiment 1: unilateral presen-tations.  c  criterion; LVF  left visual field; RVF  right visual field. 473 LANGUAGE STRUCTURE AND HEMISPHERIC ABILITY  results are presented in Table 1. The table highlights the similar-ities and differences between asymmetry patterns in the differentlanguages. Reflecting the sensitivity measures, all three groupsrevealed the expected RVF advantage in accuracy, with the En-glish language group showing this in RT as well. In addition, allthree groups showed an effect of lexicality, with words beingresponded to faster than nonwords, with the Hebrew and Arabicspeakers showing a word advantage in errors as well. Thus, over-all, performance in the lexical decision task reflected LH special-ization for the task and a word-superiority effect in the threelanguages. The higher order interactions are discussed separatelyfor each group.  English The cell means of the Lexicality  Morphology  Visual Fieldconditions are illustrated in the top panels of Figure 2. The resultsfor the English-language group were clear cut. It can be seen thatthe three-way interactions in both RT and errors were not signif-icant, because the patterns were similar for words and nonwords.In RT, the simple main effect of morphology was significant onlyin the RVF for both words,  F  (1, 19)  13.43,  p  .005,   p 2  .41,and nonwords,  F  (1, 19)    9.46,  p    .01,    p 2   .33. As listed inTable 1, in RT, both the Lexicality  Visual Field interaction andthe Morphology  Visual Field interaction were significant. Mor-phology had no main effect, but the main effects of both lexicalityand visual field were significant. Thus, the manipulation of lexi-cality and of morphology resulted in a processing dissociation,suggesting independent processing in the two hemispheres forwords and for nonwords.In errors, the Lexicality    Morphology interaction was signif-icant, because morphological complexity affected responses onlyto nonwords, to the same extent in both visual fields: LVF,  F  (1,19)    13.51,  p    .005,    p 2   .42; RVF,  F  (1, 19)    19.59,  p   .001,   p 2  .49.  Hebrew The cell means for the Hebrew speakers are illustrated in themiddle panels of Figure 2. The RT data show two effects: a maineffect of lexicality (see Table 1) and a simple main effect of morphology only for words in the RVF,  F  (1, 19)  6.74,  p  .05,   p 2   .26, with complex words being responded to more quicklythan simple words.In the error data, the three-way interaction between lexicality,morphology, and visual field approached significance. The maineffects of lexicality and of visual field are significant, as is a simplemain effect of morphology, only for nonwords and only in theLVF,  F  (1, 19)  8.87,  p  .01,   p 2  .34, with fewer errors beingmade on simple nonwords than on complex nonwords. These dataare similar to those for the English speakers, in that RT measuresrevealed sensitivity to morphological complexity of words only inthe RVF but differ from those for the English speakers, in that theerror measure revealed this sensitivity only for nonwords in theLVF.  Arabic The cell means of the Arabic speakers are illustrated in thebottom panels of Figure 2. As listed in Table 1, in RT the two-wayinteractions between lexicality and visual field and Lexicality   Morphology were significant, together with the main effect of lexicality. In RT, Arabic speakers showed an RVF advantage,  F  (1,19)  4.37,  p  .05,   p 2  .19, for words and showed no advantagefor nonwords. The cell means data showed that for words, thesimple main effect of morphology was significant only in the LVF,with complex words resulting in faster responses than simplewords,  F  (1, 19)    4.33,  p    .05,    p 2   .19; for nonwords, therewas a trend,  F  (1, 19)  3.33,  p  .08,   p 2  .15, for the oppositeeffect in the RVF, in which complex nonwords took longer toreject than simple nonwords.The error data showed very similar effects, with significantinteractions between lexicality and visual field and lexicality andmorphology, a significant main effect of lexicality, and a signifi-cant main effect of visual field. The advantage of simple overcomplex nonwords in the RVF was significant as well,  F  (1,19)    7.67,  p    .05,    p 2   .39. Again, the RVF advantage forwords was significant,  F  (1, 19)    18.94,  p    .01,    p 2   .50,whereas for nonwords it was not. Thus, like the Hebrew speakers,Table 1 Statistical Effects for Median Reaction Time and % Errors of Lexicality, Morphology, and Visual Field in Each Language Group EffectRT % errorEnglish Hebrew Arabic English Hebrew ArabicLex  Morph  VF  ns ns ns ns  3.25,  p  .087,   p 2  .15 ns Lex  VF 4.26,  p  .053,   p 2  .18 ns  4.26,  p  .053,   p 2  .18 ns ns  18.46,  p  .0001,   p 2  .49Lex  Morph  ns ns  7.07,  p  .05,   p 2  .275.30,  p  .05,   p 2  .22 ns  5.91,  p  .06,   p 2  .24Morph  VF 6.89,  p  .05,   p 2  .27 ns ns ns  3.0,  p  .099,   p 2  .14 ns Lex 28.79,  p  .001,   p 2  .6014.77,  p  .001,   p 2  .4415.45,  p  .001,   p 2  .74 ns  4.86,  p  .05,   p 2  .205.81,  p  .05,   p 2  .23Morph  ns ns ns  32.72,  p  .0001,   p 2  .52 ns ns VF 6.88,  p  .05,   p 2  .27 ns ns  21.35,  p  .0001,   p 2  .536.66,  p  .05,   p 2  .267.59,  p  .05,   p 2  .29Mean error (%) 22 22.5 36.6  Note. df   1, 19. RT  reaction time; Lex  lexicality; Morph  morphology; VF  visual field. 474  EVIATAR AND IBRAHIM
Search
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks
SAVE OUR EARTH

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!

x