the development of analogical reasoning in

8/12/2019 The Development of Analogical Reasoning In

1/23

The Development of Analogical Reasoning in

Deaf Children and Their Parents

Communication Mode

Marcin Bandurski Jakub Falkowski Institute of the Deaf, Warsaw

Tadeusz Ga 1kowskiFaculty of Psychology, University of Warsaw

The purpose of this article is to analyze the results of a studyof the development of analogical reasoning in deaf childrencoming from two different linguistic environments (deaf

children of deaf parentssign language, deaf children of hearing parentsspoken language) and in hearing children,as well as to compare two groups of deaf children to a groupof hearing children. In order to estimate the development of childrens analogical reasoning, especially the developmentof their understanding of different logical relations, twoage groups were singled out in each population of children:younger (9- and 10-year-olds) and older (12- and 13-year-olds). In this way it is possible to assess the inuence of earlyand consistent sign-language communication on the de-velopment of the conceptual system in deaf children andto establish whether early and consistent sign-languagecommunication with deaf children affects their mental

development to the same extent as early and consistentspoken-language communication with hearing children. Thechildren were given three series of analogy tasks based ondifferent logical relations: (a) a series of verbal analogy tasks(the relations of opposite, part-whole, and causality); (b)a series of numerical analogy tasks (the relations of classmembership, opposite, and part-whole); and (c) a series of gural-geometric analogy tasks (the relations of opposite andpart-whole). It was found that early and consistent sign-

language communication with deaf children plays an almostequivalent role in the development of verbal, numerical, andspatial reasoning by analogy as early and consistent spoken-

language communication with hearing children.

As Goswami (1991) notes, the essence of analogy isreasoning about relations, in particular about re-lational similarity, so that a correspondence is estab-lished between one set of relations and another (p. 1).The classical definition interprets analogy as anequality of proportions . . . involving at least 4 terms. . . when the second is related to the first as the fourthis to the third (cited in Goswami, 1991, p. 2). An

example of such analogy, called the classical analogy,is the following task of verbal analogy, ill : healthy :: poor : ____. The rst two elements make a pair witheach other, forming the opposite relation; an elementthat is selected as the solution to the third elementwould create a relation corresponding to the one estab-lishedin therstpair. The giventaskreects theequalityof proportions A:B::C:D, where A is to B as C is to D.

In Piagets opinion, the use of analogy as sophisti-cated reasoning strategy appears in early adolescence(Inhelder & Piaget, 1970). As Lunzer (1965) claims,

formal operational reasoning consists of operatingon the results of simple operations, called concreteoperations. Simple logical operation involves discover-ing the relations between objects ( leather : shoe),relations that are called relations of rst/lower order. Incontrast, analogies require an inference of the relationalsimilarity, that is, a recognition of the relations between

The research was supported by The Scientific Research Committee (theresearch project BST 773, carried out at Faculty of Psychology,University of Warsaw). We would like to thank the deaf and hearingchildren for their participation in the study, Angus Baird for editing ourmanuscript in English, Marcin Szewczyk for the technical assistance,Wojciech Kacprzak for preparing the geometric figures, Adam Tarnowskifor the statistical assistance, Karolina W 1odarczyk and Piotr Tomaszewskifor help in establishing a list of signs, Mariola B1aszczyk and AnnaKowalska for help in conveying the instructions to hearing children, andtwo anonymous reviewers for providing the helpful comments. Allcorrespondence should be sent to Marcin Bandurski, Jakub FalkowskiInstitute of the Deaf, Pl. Trzech Krzy _zzy 4/6, 00499 Warszawa, Poland(e-mail: [email protected]).

Journal of Deaf Studies and Deaf Education vol. 9 no. 2 Oxford University Press 2004; all rights reserved. DOI: 10.1093/deafed/enh018


2/23

relations (leather : shoe :: wool : cardigan ), relations thatare called relations of second/higher order. Hence,inferenceof relational similarityas typical formalopera-tional reasoning can emerge only in early adolescence.

A different viewpoint regarding analogical reason-ing development is represented by Jurkowski (1961,1967), who assumes that mental operations in earlyadolescence become independent of the content of analogy tasks. The content of the analogy tasks isunderstood as the category of logical relations on whichthese tasks are based. As studies show, analogicalreasoning and the rate of its development vary depend-ing on the category of relations contained in the pairsof particular analogies; this suggests that the extentto which mental operations depend on the categoryof logical relations on which these operations areperformed, decreases with age. Children of youngerschool age (910 years) reason correctly when thepremises concern the opposite relation ( east : west ::day : ____) or the class-membership relation ( carp : sh :: stork : ____), but they are unable to infer whenthepremises concern the causality relation ( knowledge :learn :: tiredness : ____). The difference that analogycontent makes decreases with age and hardly appears inchildren of older school age (1314 years).

The preschool childrens failures in the classicalanalogy tasks could be the result of applying relationsthat are not familiar to them in these tasks. As the study

by Goswami and Brown (1990b) shows,4-year-old children successfully solved tasks of pictorialanalogies that contained physical-causal relations, e.g.cake : cut cake :: apple : ____, choosing the appropriatepicture from the series of different pictures ( cut apple,cut bread , bruised apple, ball , banana ). When solvingcontrol-causal reasoning tasks that consisted of thepictures of causally transformed objects, such as cut cake, cut apple, and cut bread , these children often foundthe causal agent responsible for the transformation of the mentioned objects among the series of pictures of

possible agents, such as knife, water, and sun. Thesecontrol tasks examined childrens knowledge about therelations used in analogies. The results suggest thatchildren of preschool age are able to recognize thehigher-order relations as long as the analogies pre-sented to them are based on relations that are part of their knowledge.

Additional support for Goswamis (1991, 1996)claims about the importance of relational familiarity insolving the analogy tasks comes from studies byAlexander and her colleagues (1987, 1989). In theirstudies, 4-year-old children successfully solved geo-metric analogy tasks based on relations of color, shape,and size (large red circle : small red circle :: large blue square : ____), searching out the appropriate gure ina set. Goswami (1989) conducted a similar study withthe use of geometric analogy tasks that containedproportion relations, such as half circle : half rectangle ::quarter circle : ____. Children aged 4 years used thestrategy based on relational similarity, searching out theappropriate element of a gure in a set.

Sternberg and Nigro (1980) found that whensolving verbal analogy tasks, children aged 9 and 12years often use associations as a substitute for fullreasoning by analogy, because in the task narrow :wide :: question : (trial, statement, answer, ask ), the wordquestion is highly associated with the term answer. So, insome situations, the solutions associated with the Cterm can be correct. In contrast, 15-year-old childrenuse strategies based on relational similarity rather thanon associations with the C term. On this basis,researchers singled out two levels of verbal-analogysolutions that correspond to Piagetian stages of concrete and formal operations. Moreover, Goldmanand her colleagues (1982) conclude that 8-year-olds

solving verbal-analogy tasks have difculties in in-ferring the relation between the A and B terms. Thesedifculties can reect a general lack of attention to thementioned analogy terms. Hence, the children tend tochoose or to generate D terms that are stronglyassociated with the C term.

As a study by Goswami and Brown (1990a) shows,children aged 4 years, solving first the control thematictasks that were composed of the analogy C term ( dog : ____) and the alternative solutions ( doghouse, bone,cat , other dog ), chose the rst two answers as the

associations with the C term. Next these children,solving the pictorial analogy tasks based on thematicrelations (bird : nest :: dog : ____) and given the samealternative solutions, pointed mostly to solutions thatcopied the relation occurring between the A andB terms (doghouse). This implies that children areaware of the existence and necessity of the relational

154 Journal of Deaf Studies and Deaf Education 9:2 Spring 2004


3/23


4/23

nonverbal spheres of logical thinking such as abstractspatial reasoning (Braden, 1987; Conrad & Weiskrantz,1981) and understanding the principles of liquidconservation (Dolman, 1983). On the other hand,DCDPin contrast to DCHPhave the geneticallydetermined deafness that is not accompanied by otheradditional handicaps (e.g., partial deficits, a mildmental retardation); this indicates that the interactionof biological and environmental factors affects thegeneral cognitive development in these children(Akamatsu, Musselman, & Zwiebel, 2000; Kusche,Greenberg, & Garfield, 1983). Moreover, sign languageis treated as verbal language (Farris, 1994; Hoemann,1991; Klima & Bellugi, 1979; Marschark, 1993;Marschark, Siple, Lillo-Martin, Campbell, & Ever-hart, 1997) because this language is a full-fledgedlinguistic system that possesses complex syntactic,lexical, and morphological structures and allows thepossibility of operating with abstract expressions,logical connections, and well-organized sentences.Hence, it would be possible to examine the verbalspheres of logical thinking in deaf children whilemaintaining strict methodological procedures.

The development of analogical reasoning in deaf children was studied by Panasiuk (1990) and Sharpe(1985). The study by Panasiuk (1990) shows that deaf children aged 10, 12, and 14 years do not equal theirhearing peers in respect to understanding and using

different logical relations (opposite, class membership,and part-whole) in three spheres of analogical reason-ing (verbal, numerical, and spatial). Panasiuk claimsthat hearing impairment makes spoken-languagecommunication impossible and thus promotes thedevelopment of concretely imaged mental operationsin deaf children. In contrast, a study by Sharpe (1985)shows that deaf adolescents (entirely DCHP) aged 14 19 years do not understand higher-order relations intasks of verbal and geometric analogies as well as age-matched hearing adolescents do. In Sharpes opinion,

the primary communication mode, reflected throughthe use of spoken language or sign language, is aprerequisite to the development of higher complexcognitive processes but not facility in any languages andit accounts for deaf and hearing adolescents differentperformances on the mentioned tasks. Hearing child-rens oral-aural mode of communication provides the

sensory experience that facilitates the perception of contrast more effectively than does the gestural-visualmode of communication of deaf children. 2 Sucha perception of contrast involves an analysis of spokenutterances that contain a contrast at several levels, soit leads to complex cognition and hierarchic cognitivedevelopment, unlike the perception of contrast in-volving an analysis of signed utterances.

Assumptions and Hypotheses

As mentioned earlier, deaf adolescents who didnot haveearlyand consistentcontact withsign languageobtainedworse results in tasks of analogies than did hearingadolescents. According to Sharpe (1985), these out-comes canbe evidence of theineffectiveness of gestural-visual communication mode that is used bydeaf people.One can assume that overlooking deaf adolescents whohaveearlyand consistent contact withsign language hasled her to formulate the above-mentioned conclusion.Therefore one should reconsider the possibility thata gestural-visual communication mode and an oral-aural communication mode might facilitate to an equaldegree the perception of contrast necessary to complexcognition, if deaf children (sign language) as well ashearing children (spoken language) have consistentcontact with linguistic systems suitable for them frombirth. Yet one should emphasize that deaf childrens

gestural-visual communication mode provides themwith a somewhat other sensory experience than hearingchildrens oral-aural communication mode; this doesnot necessarily have to mean that a gestural-visualcommunication mode is inferior to an oral-auralcommunication mode. One also should assume thatsigns, which cannot be treated as substitutes for wordsbut as their equivalents, enable the information con-taining the contents of different levels of abstractnessto be conveyed, so they allow deaf children to enrichtheir cognitive experiences.

According to the above assumptions, the followinghypotheses that will be verified empirically wereformulated:

Hypothesis 1. The ability of deaf children tounderstand the analogical relations expressed verballyand nonverbally is comparable to the ability of



5/23

hearing children; early and consistent sign-languagecommunication determines the emergence of thisability in deaf children at the same time as in hearingchildren who have early and consistent exposure tospoken language.

The following detailed hypotheses were drawnfrom the general hypothesis above:

Hypothesis 2. Deaf children with early andconsistent exposure to sign language understand thedifferent relations expressed in sign language ina similar way as hearing children (with early andconsistent exposure to spoken language) who un-derstand the relations expressed in spoken language.

Hypothesis 3. The education of deaf children with

early and consistent exposure to sign language in anartificial sign system 3 or in spoken language contributesto the emergence of the ability to understand numericalrelations somewhat later than in hearing children withearly and consistent exposure to spoken language.

Hypothesis 4. Early and consistent exposure tothe grammatical structures of sign languageincontrast to early and consistent exposure to thegrammar of spoken languagedetermines the differ-ences between deaf and hearing children in the

recognition of spatial relations, to the disadvantage of hearing children.

To verify these hypotheses, an analysis of a specificindependent variablethe communication mode(sign language, spoken language)was performed.The dependent variable, which according to the aboveassumptions the independent variable can influence,was the level of analogical reasoning in deaf children, inparticular the level of their understanding of logicalrelations expressed verbally and nonverbally.

Method

Subjects

A total of 104 children took part in the study. Allchildren were of an intellectual norm. Children who

had an additional handicap from birth were eliminatedfrom the study. When selecting children for the study,the following factors influencing the level of cognitivedevelopment were taken into consideration:

AgeChildren aged 910 years (the youngerage group) and 1213 years (the older age group) were

tested. The age of each group reected the character-istic periods of thinking development described byPiaget: younger groupstage of concrete operations,older groupstage of formal operations.

Linguistic environmentThe following groupsof children were tested: (a) DCDP having early andconsistent contact with sign language; (b) DCHPhaving contact with spoken language; (c) HCHP havingearly and consistent contact with spoken language.

The information on the deaf subjects linguisticenvironment was provided by teachers and schoolpsychologists. Neither of the hearing parents of deaf children had contact with sign language until the childbegan to attend preschool (by age 6). The hearingparents did not communicate with their deaf childrenin sign language, and deaf parents used sign languagein interaction with their deaf children. Some DCHPhad deaf siblings (4 children in the younger group and2 children in the older group). The DCDP who did nothave early and consistent contact with sign languagewere eliminated from the study because the purpose of

study was to evaluate the influence of early andconsistent sign-language communication on the cog-nitive development of deaf children. These childrenwere unable to use sign language or they communicat-ed by means of some signs. This was confirmed byteachers who also mentioned that the childrens deaf parents did not communicate with them in signlanguage. All the deaf children attended schools whereteachers used either an artificial sign system or spokenlanguage. In the audiological respect, all children withhearing impairment had severe and profound pre-

lingual deafness (above 80 dB in the better ear).The information about the degree of hearing loss camefrom school records. The data concerning the groupsof children are presented in Table 1.

The study was conducted at elementary schoolsand secondary schools for deaf children in Warsaw,Cracow, Lodz, Wroclaw, Lublin, Wejherowo, and

Development of Analogical Reasoning 157


6/23

Otwock, and also at elementary schools and secondaryschools for hearing children in Warsaw.

Materials

Three series of analogy tasks based on verbal,numerical, and spatial relations were selected for thestudy. The particular series of tasks were given in thesame order for all the subjects: (a) a series of verbalanalogy tasks, (b) a series of numerical analogy tasks,and (c) a series of figural-geometric analogy tasks.

The time to solve the particular tasks was limitedto 90 sec, and the measurement of reaction time waseliminated, so none of the series of tasks required speedin solving them.

In order to eliminate the possibility of learning, allthe analogy tasks were given in an established orderaccording to the sequence principle. In each series, thetasks of analogies based on the same category of logicalrelations were presented separately.

The presentation of each series of analogy tasks waspreceded by instructions consistent with the content

of each series of tasks; the content of the instructionwas conveyed to deaf children in conventional signlanguage and to hearing children in spoken language.Additional explanations were given as needed.

Tasks of verbal analogies. The verbal analogies weredrawn from the series of verbal analogy tasks used by

Jurkowski (1967), Pietrulewicz (1983), and Panasiuk(1990, 1995). The tasks of verbal analogies were basedon three categories of relations that in Jurkowskis(1967) opinion have a privileged place in cognitivepsychology: opposite, part-whole, and causality. Thesecategories of relations were taken into considerationbecause the purpose of the study was to examinewhether deaf children of older school agethat is, theage at which the period of formal operations occurs, orin other words the period in which mental operationsbecome independent of the category of logical rela-

tionsare able to understand logical relations regard-less of type. An understanding of relations irrespectiveof type is evidence of attaining the level of formalthinking.

The verbal analogy tasks were inventive tasks, thatis, they required children to generate the missing termthat would complete the analogy.

The verbal analogy tasks were presented to the deaf children in sign language and to the hearing childrenfirst in spoken language, then in written language. Toestablish a list of verbal analogy tasks, words and signs

belonging to the same categories were selected: (a)grammatical categories (words and signs correspond-ing to them from groups of nouns, verbs, adjectives,or adverbs); (b) lexical frequency categories (words andsigns occurring in use equally frequently); (c) lexical-semantic categories (the meanings of words and signsalmost corresponding with each other). Such an action

Table 1 The data regarding the groups of children

Environmentaland agecategories Grade

Totalnumber of subjects

Numberof girls

Numberof boys

Age of youngestchild

Age of oldestchild

Meanage of subjects

Younger DCDP 3 ES* 8 4 4 8;8 10;8 9;104 ES 8 4 4

Older DCDP 6 ES 8 4 4 11;5 14;6 12;10

1 SS 8 4 4Younger DCHP 4 ES** 16 8 8 9;6 13;5 10;9Older DCHP 6 ES 8 4 4 12;1 14;0 13;1

1 SS 8 4 4Younger HCHP 3 ES 10 5 5 8;11 10;10 9;11

4 ES 10 5 5Older HCHP 6 ES 10 5 5 12;0 13;10 12;10

1 SS 10 5 5

*ESelementary school; SSsecondary school.

**Children from grade 3 (ES) did not participate in this experiment because they did not understand the instructions given in the conventional signlanguage; their number was 5, so the researchers decided not to examine these children.



7/23

is consistent with Jurkowskis (1990) recommendationthat the content of the native-language words and of the foreign-language words correspond. The wordsand signs matched in this way were the analogy terms.

To present the trial tasks of verbal analogies ( egg :hen :: milk : ____ and summer : rain :: winter : ____ ),the following instruction was given: Read attentivelythree successive words. The rst two words make a pairwith each other. Compare them to each other. Thenyou have a third word to which you will select thefourth word. You must complete it so as to form a pairsimilar to the pair that is formed by the rst twowords. To give the verbal analogy tasks to deaf chil-dren, wordsbeing the analogy termswere sub-stituted by signs, and the instruction was modied.

The series of verbal analogy tasks numbered 24trials; each logical relation was contained in 8 trials.The list of verbal analogy tasks is included in Appendix1. These tasks are grouped according to specific logicalrelation and are marked with successive numbers.

In order to estimate the solutions of verbal analogytasks, the principles of singling out the correct andincorrect solutions were applied. These principles wereformulated by Jurkowski (1967). In reference to thetasks concerning the part-whole relation, attention wasdirected to whether the whole given in solution wasapproximate to the whole in the first pair of analogy.With regard to the opposite relation, attention was

directed to the mixing of two relations that are notidentical in a logical respect: opposite and negation.With reference to the causality relation, attention wasdirected to whether the effect was of direct (notindirect) character. Furthermore, the answers had to becorrect in respect to form: they had to have theappropriate case, tense, number (singular or plural),and mood. According to Jurkowski (1967), whenestimating the solutions, researchers should omithearing childrens spelling mistakes, because thepurpose of the study is to examine the ability to reason

analogically, not the ability to write.

Tasks of numerical analogies. The series of numericalanalogy tasks was borrowed from the experimentconducted by Panasiuk (1990). The series of numericalanalogy tasks contained three categories of logicalrelations: opposite, class membership, and part-whole.

Taking these categories of relations into considerationaimed to examine whether older deaf children canunderstand the logical relations irrespective of theircategory.

The numerical analogy tasks were selective tasks,that is, they required children to mark the answer thatwould complete the analogy.

For first numerical analogy task, the followinginstruction was given: Look attentively at three rowsof numbers. The first two rows make a pair with eachother. Compare them to each other. Next you have thethird row of numbers, to which you must select sucha row of numbers from the rows given in parentheses soas to form a pair similar to the pair that is formed bythe first two rows of numbers.

The series of numerical analogy tasks contained 18trials; each logical relation numbered 6 trials. Thisseries presented the various mathematical functions.The set of numerical analogy tasks is included inAppendix 2. The tasks are marked with the successivenumbers and with the symbols related to a certainrelation (aopposite, bclass membership, cpart-whole).

Tasks of gural-geometric analogies. The series of figural-geometric analogy tasks was constructed byPanasiuk (1990, 1995), who used this series for thestudies of analogical reasoning in deaf and hearing

children. This series took only two relations intoconsideration: opposite and part-whole, because ac-cording to Panasiuk (1995) it was difficult to establisha graphic picture of logical connection for the classmembership relation (p. 133).

The figural-geometric analogy tasks were selectivetasks, that is, they required the children to choose onefigure from among four different figures in order tocomplete the second pair of the analogy according tothe particular logical relation.

For the first figural-geometric analogy task, the

following instruction was given: Look attentively atthree figures. The first two figures make a pair witheach other. Compare them to each other. Now look atthe third figure, to which you should select such a figurefrom four figures given in parentheses so as to forma pair similar to the pair which is formed by the firsttwo figures.



8/23

The series of figural-geometric analogy tasks had12 trials, 6 trials for each logical relation. The set of figural-geometric analogy tasks is included in Appen-dix 3. The tasks are marked with the successivenumbers and with the symbols related to a certainrelation (aopposite, bpart-whole).

Procedure

The study was conducted in school rooms in conditionsfavoring attention to the tasks. The study was of anindividual character, that is, only two persons (experi-menter and subject) took part in the experiment, whichconsisted of one session. The deaf and hearing childrenwere examined by a deaf researcher (the first author).This researcher is able to communicate with hearingpersons, reading the utterances from their lips. By age 3he had become deaf, so he can correctly pronounce thewords and sentences in Polish, in which he is fluent.Furthermore, the hearing educators helped him convey

the instructions concerning the analogy tasks whenhearing children did not understand these instructions.The session started by getting in touch with the

subject. In order to contact the subject, conversationswere initiated by asking questions concerning whathe/she was interested in, whether he/she liked tolearn, what subjects at school he/she liked, and soforth.Furthermore, DCDP were asked questions aboutwhether their parents used sign language, and it was

observed whether they could use this language fluently.This aimed to eliminate from the study those DCDPwho did not have early and consistent contact with signlanguage.

Each subject was informed that he/she wouldsolve interesting tasks. Next, after conveying theinstructions and additional explanations, three seriesof analogy tasks were given in the established order.The actions performed by the subject were observed;in the series of verbal analogy tasks, the actionsconsisted of generating the term that would completethe analogy; in the series of nonverbal analogy tasks,the actions consisted in marking the chosen answer.While solving the verbal analogy tasks, hearingchildren first gave the answers in spoken language,then wrote down these answers on the cardthemselves (their answers in both languages corre-sponded with each other); deaf children gave theanswers in sign language, and deaf childrens answerswere written down by the experimenter. After thesession, the childs outcomes were noted on the sheetof record of results. The session with youngerchildren lasted 45 to 60 min., and with olderchildren, 30 to 45 min.

Results

The Development of Verbal Analogical Reasoning

The present section is related to deaf and hearingchildrens understanding of verbal relations at dif-ferent levels of difficulty (opposite, part-whole, andcausality).

The opposite relation. The easiest relation for almostall the groups was the opposite relation, which wascharacterized by the highest mean scores.

Table 2 The means and standard deviations in tasks of verbal analogies based on the opposite relation

Age group

DCDP DCHP HCHP

M SD M SD M SD

Younger 5.2500 0.854 2.1250 1.994 5.9000 1.021Older 6.1250 0.881 5.8750 0.881 6.9500 1.097

Marithmetic mean; SDstandard deviation.

Table 3 The values of t test and the levels of signicance of differences between the mean scores of (a) particular

environmental groups at both levels of age, and (b) age groups (younger and older) in each populationA comparison of youngerenvironmental groups

A comparison of olderenvironmental groups

A comparison of age groups(younger and older) in population

DCDP-DCHP, t (20) 5.77, p , 0.05 DCDP-DCHP, t (30) 0.80, ns DCDP, t (30) 2.85, p , 0.05DCDP-HCHP, t (34) 2.04, p , 0.05 DCDP-HCHP, t (34) 2.44, p , 0.05 DCHP, t (21) 6.89, p , 0.05DCHP-HCHP, t (25) 6.90, p , 0.05 DCHP-HCHP, t (34) 3.18, p , 0.05 HCHP, t (38) 3.14, p , 0.05

t a value of function of students test; p a level of signicance.



9/23

With the exception of the comparison of oldergroups of DCDP and DCHP, the only signicantdifferences were found in the mean scores withregard to understanding the opposite relation. In thepopulation of DCHP, the increase in childrensunderstanding of the opposite relation is larger thanin the others (DCDP and HCHP). In eachpopulation, there is a signicant difference betweenage groups, younger and older, in terms of their

understanding of the opposite relation. A generalANOVA three-way model with repeated measures inone factor has been applied to explore the inter-actions between group membership, age, and task.Because the interaction of three factors had beensignicant, F (14, 686) 5 3.79, p , 0.0001, theanalysis was followed by separate ANOVAs forparticular tasks, with posthoc LSD tests. For theverbal opposite relation, the ANOVA showed a sig-nicant effect of age, F (1, 98) 5 65.82, p , 0.0001;a signicant effect of group, F (2, 98) 5 38.30, p ,

0.0001; and a signicant interaction between age andgroup, F (2, 98) 5 15.33, p , 0.0001. Posthoc LSDtests showed that in both DCHP groups the scoresare low. All younger groups had lower scores incomparison with older ones. The scores of theyounger DCDP group are not signicantly differentfrom the younger HCHP group, while in the olderDCDP group the scores are a little lower.

The qualitative analysis. When solving task 22 ( sell :buy :: rest : ____ ), hearing children completed theanalogy by giving the words work, make, run, and labor.In contrast, deaf children coming from two linguisticenvironments tended to complete this analogy bygiving one term (work). Hearing children are able to go

beyond the schematic connections and operate ona wider range of terms. A comparison of the wrongsolutions given by all the groups of children in manytasks allows one to conclude that there are noqualitative differences between them. When solvingtrial 22 ( sell : buy :: rest : ____ ), subjects gave terms of associative or situational character, such as lie, laze, andnot work. In contrast, the inappropriate solutions givenby younger DCHP constitute a separate category of errors. Almost all the tasks were completed by givingthe terms that have no content relationship with any

term of the analogy. In task 1 ( well : badly :: nicely : ____ ), the following terms appeared: mess, wise, clothes,and writing.

The part-whole relation. The part-whole relation isa somewhat more difficult relation from the previousconnection; the evidence is the lower mean score ineach group.

With regard to their understanding of the part-whole relation, significant differences occur betweenthe following groups: younger groups of DCDP and

HCHP; younger groups of DCHP and HCHP; oldergroups of DCDP and DCHP; and older groups of DCHP and HCHP. No significant differences arefound in the other comparisons. The developmentaltrend in the populations of DCDP and DCHP isgreater than in the population of HCHP. Yet, theincrease of results between age groups, younger andolder, attains a significant level in each population. For

Table 4 The means and standard deviations in tasks of verbal analogies based on the part-whole relation

Age group

DCDP DCHP HCHP

M SD M SD M SD

Younger 2.6250 0.957 1.9375 1.609 4.7000 1.894Older 6.3750 1.627 4.9375 1.178 6.3000 1.559


Table 5 The values of t test and the levels of signicance of differences between the mean scores of: a) particularenvironmental groups at both levels of age; b) age groups (younger and older) in each population

A comparison of youngerenvironmental groups


A comparison of age groups(younger and older)in population

DCDP-DCHP, t (24) 1.46, ns DCDP-D CHP, t (30) 2.86, p , 0.05 DCDP, t (24) 7.96, p , 0.05DCDP-HCHP, t (25) 4.27, p , 0.05 DCDP-HCHP, t (34) 0.14, ns DCHP, t (30) 6.03, p , 0.05DCHP-HCHP, t (34) 4.65, p , 0.05 DCHP-HCHP, t (34) 2.89, p , 0.05 HCHP, t (38) 2.91, p , 0.05




10/23

the verbal part-whole relation, the ANOVA showeda significant effect of age, F (1, 98)5 85.70, p , 0.0001;a signicant effect of group, F (2, 98) 5 16.30, p ,

0.0001; and a signicant interaction of factors,F (2, 98)5 4.65, p , 0.02. The results of LSD tests sug-gest that all differences are signicant, except the non-signicantdifference betweenyounger groups of DCDPand DCHP.

The qualitative analysis. When solving task 23(wool : sheep :: feathers : ____ ), younger DCDP tendedto complete the analogy by giving concrete conceptssuch as hen, duck, peacock, goose, and pigeon, whereasyounger HCHP tended to give the general conceptbird. These proportions (37.5% and 65% using theconcept bird ) are equalized in older children (68.75%and 65%). In reference to task 17 ( drawers : desk :: pockets : ____ ), one can observe a different tendency.DCDP from both age groups (18.75% and 37.5%)gave the general name clothes more frequently thanHCHP (5% and 20%) who in turn gave more names of certain kinds of clothes: jacket, trousers, coat , and blouse.Furthermore, in task 14 ( corridors : school :: streets : ____ ), one can notice associative terms, such as cars,courtyard, houses, shops, and street-lamps. In trial8 (minute : hour :: centimeter : ____ ), distance termssuch as kilometer, decimeter, millimeter were given bychildren. These errors occurred most frequently inyounger children and more frequently in the deaf thanin the hearing.

The causality relation. The most difficult relationexpressed verbally is the causality relation, which wascharacterized by the lowest mean scores.

With regard to their understanding of the causalityrelation, signicant differences are observed betweenyounger groups of DCDP and DCHP, younger groups

of DCHP and HCHP, older groups of DCDP andDCHP, and older groups of DCHP and HCHP. Theother comparisons did not show signicant differences.The increase in their understanding of the causalityrelation in the population of DCDP is somewhat lessthan in the other populations (DCHP and HCHP);the increase of scores is signicant in each population.For the verbal causality relation, the ANOVA showeda signicant effect of age, F (1, 98)5 61.12, p , 0.0001,and a signicant effect of group, F (2, 98) 5 17.63, p ,

0.0001. No signicant interaction, F (2, 98) 5 0.63, ns,was found. All younger groups had lower scores whencompared to older ones, and the DCHP groups hadlower scores in comparison to other groups.

The qualitative analysis. In task 6 ( frosts : ice ::rain : ____ ), which required children to recognizea direct causal relationship between atmosphericphenomena, DCDP gave the terms such as ood,mud , and puddles, and HCHP gave only this last label.Trial 12 (bomb : ruins :: thunderbolt : ____ ), which

concerns acoustic phenomena, was much easier forgroups of DCDP than for groups of HCHP. In DCDP,one can observe solutions that refer to visuallyperceived phenomena ( destructions, re, ame, andoverturned tree), and in HCHP one can observe only therst three words. Moreover, in task 6 ( frosts : ice :: rain: ____ ), one can notice associative responses such aswater, storm, wetly, and thunderbolt. When solving trial 9(inattention : accident :: lack of food : ____ ), instead of

Table 6 The means and standard deviations in tasks of verbal analogies based on the causality relation

Age group

DCDP DCHP HCHP

M SD M SD M SD

Younger 3.6250 1.543 0.9375 0.924 2.6000 1.666Older 5.5625 1.786 3.6875 1.578 5.2500 1.772


Table 7 The values of t test and the levels of signicance of differences between the mean scores of: a) particular

environmental groups at both levels of age; b) age groups (younger and older) in each populationA comparison of youngerenvironmental groups



DCDP-DCHP, t (24) 5.99, p , 0.05 DCDP-DCHP, t (30) 3.15, p , 0.05 DCDP, t (30) 3.28, p , 0.05DCDP-HCHP, t (34) 1.89, ns DCDP-HCHP, t (34) 0.52, ns DCHP, t (24) 6.03, p , 0.05DCHP-HCHP, t (26) 3.79, p , 0.05 DCHP-HCHP, t (34) 2.76, p , 0.05 HCHP, t (38) 4.88, p , 0.05




11/23

the required term hunger , terms such as death and illnessappeared. Here the effect is of an indirect character.These errors are characteristic for each group, yet mosterrors appear in younger children, more often inDCHP than in DCDP and HCHP.

The Development of Numerical AnalogicalReasoning

The next set of trials given to children was a series of numerical analogy tasks that represented three catego-ries of logical relations: class membership, opposite andpart-whole.

The class membership relation. The easiest connectionis the class membership relation, and the mean scoresin this relation have the highest values.

With regard to their understanding of the classmembership relation, a signicant difference appearssolely between younger groups of DCHP and HCHP;no signicant differences in the other comparisons of environmental groups at both levels of age are found. Alarger increase in terms of their understanding of theclass membership relation is observed in the popula-tion of DCHP, and in the others a smaller one. Incontrast, the increase of scores is signicant only in thepopulation of DCHP. For the numerical class member-ship relation, the ANOVA showed a signicant effect of

age, F (1, 98) 5 11.69, p , 0.001; a signicant effect of group, F (2, 98) 5 3.82, p , 0.05; and a nonsignicantinteraction of factors, F (2, 98) 5 2.70, p , 0.08. TheLSD tests show that only the younger DCHP grouphas signicantly lower scores in comparison to otheryounger ones.

The qualitative analysis. The mathematical functionsin all the numerical analogy tasks based on the classmembership relation are the same and take the form:(n n n n) : (4n) :: (m m m) : (3m). Thechildren had to infer the multiplication of someidentical elements by their number from the additionof these elements. The most frequent error, typical forall the groups, was a kind of repetition of the analogysB term. In task 2, (8 8 ) : (8 2) :: (2 2 2 2) :(____), the row of numbers ( 2 8 ) was mostly given.

The opposite relation. A somewhat more difficultrelation from the previous connection is the oppositerelation, which is characterized by lower mean scores.

The differences between scores of younger groupsof DCDP and DCHP and younger groups of DCHPand HCHP are signicant, and the differences betweenscores of the other groups do not show a signicantlevel. The increase in their understanding of the op-posite relationgreater in the environmental groupsof DCDP and DCHP, smaller in the population of HCHPis signicant only in the rst two groups. Forthe numerical opposite relation, the ANOVA showeda signicant effect of age, F (1, 98) 5 11.51, p , 0.001;a signicant effect of group, F (2, 98) 5 3.49, p , 0.05;and a nonsignicant interaction of factors, F (2, 98) 5

1.85, p , 0.2. The LSD tests show that only theyounger DCHP group has signicantly lower scores incomparison to other younger ones.

Table 8 The means and standard deviations in tasks of numerical analogies based on the class membership relation

Age group

DCDP DCHP HCHP

M SD M SD M SD

Younger 5.5000 0.816 4.6875 1.535 5.6000 0.681Older 5.8750 0.333 5.7500 0.443 5.8000 0.522


Table 9 The values of t test and the levels of signicance of differences between the mean scores of: a) particular

environmental groups at both levels of age; b) age groups (younger and older) in each populationA comparison of youngerenvironmental groups



DCDP-DCHP, t (23) 1.87, ns DCDP-DCHP, t (30) 0.90, ns DCDP, t (20) 1.71, nsDCDP-HCHP, t (34) 0.40, ns DCDP-HCHP, t (29) 0.53, ns DCHP, t (17) 2.65, p , 0.05DCHP-HCHP, t (23) 2.21, p , 0.05 DCHP-HCHP, t (34) 0.30, ns HCHP, t (38) 1.04, ns




12/23


13/23

identity and take a 180 rotation of figures intoconsideration. The most frequent error, which occursmostly in both groups of HCHP and younger group of DCHP, was the maintenance of the analogys C term.When solving task 3, the answer (number 2) was given,which does not take the 180 rotation into consider-ation. The chosen figure, a diamond with a particularelement on the left, is not rotated 180 in relation to theidentical figure in the C term.

The part-whole relation. A more difficult connectionis the part-whole relation, and mean indicators in thisrelation have lower values.

The differences in the understanding of the part-whole relation between the scores of particular environ-mental groups at both levels of age do not show asignicant level. The increase of scores in understand-ing the part-whole relation in the population of DCDPis larger than in the other populations (DCHP andHCHP); this increase is signicant only in the popula-

tions of DCDP and HCHP. For the spatial part-wholerelation, the ANOVA showed a signicant effect of age;F (1, 98) 5 20.47, p , 0.0001; a nonsignicant effect of group; F (2, 98)5 1.63, p , 0.2; and a nonsignicant in-teraction of factors, F (2, 98) 5 1.01, p , 0.4. The LSDtests show that the younger groups have lower scores.

The qualitative analysis. All the figural-geometricanalogy tasks based on the part-whole relation require

a precise analysis and synthesis of the appearance of elements of particular figures. The most frequenterror, typical for all the groups, was the transfer of opposite relation. This kind of error is observed in task9 (number 1). Here the figure chosen as the analogycompletion is rotated 180 in relation to the identicalfigure in the C term. The number of errors decreaseswith increasing age.

Discussion

The aim of the present study was to verify thehypotheses about the almost equivalent influence of early and consistent sign-language communicationon the development of verbal and nonverbal thoughtprocesses in deaf children. Hence, the empirical dataselected during the conducted study allow not only the

evaluation of the level of mental development of deaf children coming from two different linguistic environ-ments, but also the comparison of these children tohearing children.

The results of this study support the generalhypothesis that early and consistent sign-languagecommunication determines the emergence of similarabilities to understand the logical relations expressedverbally and nonverbally in deaf children at the sametime as in hearing children with early and consistentexposure to spoken language. The quantitative and

qualitative similarity in performance between DCDPand HCHP can mean that children from both pop-ulations possess a similar complex conceptual system,which is required to solve the analogy tasks that havea complicated logical structure. This system hasbeen shaped on the basis of a verbal and nonverbalsymbolic system. If DCDP, like HCHP, are able torecognize the analogy between similar logical relations,oneshouldexpectthat from early childhoodthe analogy

Table 12 The means and standard deviations in tasks of numerical analogies based on the part-whole relation

Age group

DCDP DCHP HCHP

M SD M SD M SD

Younger 2.6875 1.446 2.3750 1.257 2.7500 1.831Older 4.5625 1.894 3.2500 1.914 4.7500 1.369



environmental groups at both levels of age and (b) age groups (younger and older) in each populationA comparison of youngerenvironmental groups



DCDP-DCHP, t (30) 0.65, ns DCDP-DCHP, t (30) 1.95, ns DCDP, t (30) 3.15, p , 0.05DCDP-HCHP, t (34) 0.11, ns DCDP-HCHP, t (34) 0.34, ns DCHP, t (30) 1.53, nsDCHP-HCHP, t (34) 0.69, ns DCHP-HCHP, t (34) 2.74, p , 0.05 HCHP, t (38) 3.91, p , 0.05




14/23

playsa role in their cognitivedevelopment, on conditionthat these children have a consistent contact witha suitable linguistic system from birth. Sign-languagecommunication, like spoken-language communication,is an effective tool leading to the emergence of similarcognitive abilities in the appropriate time; this isconrmed by the results of studies carried out byBraden (1987), Brill (1974), Courtin (1997, 2000a,2000b; Courtin & Melot, 1998), Sisco and Anderson

(1980), and Zwiebel (1987). Yet the precise quantitativeand qualitative analysis of performances in the differenttasks hasrevealedsome differences between DCDP andHCHP, for which there are many possible explanations.

One of the three detailed hypotheses, whichsuggested that deaf children with early and consistentexposure to sign language understandthe different rela-tions expressed in sign language in a similar way ashearing children (with early and consistent exposureto spoken language) who understand the relationsexpressed in spoken language, is partially confirmed by

the results of this study, because significant differenceswere revealed in the series of verbal analogy tasks basedon the opposite relation to the disadvantage of both agegroups of DCDP. Yet, one should emphasize that withthe exception of one verbal analogy task (lie : truth ::love : ____ ), where deaf children had quite largedifculties in nding the appropriate label, 4 no quanti-tative and qualitative differences between DCDP andHCHP are found. It can point to a similarity be-

tween the cognitive structures that develop throughcommunication in the linguistic systems suitable forthem. An explanation for the failures of DCDP in theabove-mentioned task, which comprises the contentsseparated from reality, may be the lack of use of termswith abstract content by teachers during the lessons; asa result of this, these children solely use the concrete,familiar expressions from everyday life situations.Hence, they can easily recognize the relations betweenfamiliar terms in theother verbal analogy tasks based onthe opposite relation. As Furth (1966, 1971a) notices,the education of deaf children consists only in repeatingthe specic actions or expressions of a concretecharacter; it can lead to the impoverishment of theirlinguistic and nonlinguistic experiences. It seems thatnot taking into consideration the opposing terms of abstract character (signs, words) in the didactic processcanresult in theconcreteness of deaf childrens thinkingregardless of whether they have early and consistentcontact with sign language.

Yet, the series of verbal analogy tasks based onthe part-whole relation has revealed some interestinginformation. It refers especially to younger DCDPwhocompared to younger HCHPobtained worseresults in many verbal analogy tasks containing thepart-whole relation. The explanations for their failuresperhaps are such factors as the inappropriate acquisi-tion of the contents of some concepts or the difficulties

in the multisensory association of information comingfrom the setting. The fact that the poor performance of younger DCDP is due to the restricted transmission of information concerning the perceptual (color, shape,size, and element) and functional (application) prop-erties of perceived objects relates to the teachers useof ineffective communicative systems. In the Polisheducational system, the education of deaf childrentakes place mostly either in spoken language or in

Table 14 The means and standard deviations in tasks of gural-geometric analogies based on the opposite relation

Age group

DCDP DCHP HCHP

M SD M SD M SD

Younger 5.8125 0.539 5.4375 0.811 5.3000 1.081Older 5.9375 0.241 5.9375 0.241 5.5000 1.000



environmental groups at both levels of age, and (b) age groups (younger and older) in each populationA comparison of youngerenvironmental groups



DCDP-DCHP, t (30) 1.54, ns DCDP-DCHP, t (30) 0.00, ns DCDP, t (20) 0.86, nsDCDP-HCHP, t (25) 1.86, ns DCDP-HCHP, t (19) 1.89, ns DCHP, t (17) 2.35, p , 0.05DCHP-HCHP, t (34) 0.42, ns DCHP-HCHP, t (19) 1.89, ns HCHP, t (38) 0.60, ns




15/23

an artificial sign system (Tomaszewski, 1999); thisobstructs them in the appropriate acquisition of thecontents of concepts referring not only to the variousconcrete objects, but also to mathematics. Mathemat-ical concepts such as month, year, day, week, minute,hour, centimeter , and meter , contain many dimensions,so it can be necessary to possess the multidimensionalsemantic space in the mind in order to establisha degree of distance between these concepts (Biela,1981; Trzebin ski, 1981). As a result of being educatedin ineffective communicative systems, DCDP probablypossess the semantic space with a restricted number of dimensions in the early school education period; thiscan contribute to the poor formation of part-wholeconcepts in their mind.

Another possible source of the failure of youngerDCDP to understand the part-whole relation is theirdifficulty in the multisensory association of stimuli,which is mentioned by Ga 1kowski (1998) whendescribing Baron-Cohens experimental analysis of Eye-Direction Detector (EDD) and Shared-Attention Mechanism (SAM), important in nonlinguistic com-munication. The function of these regulative mecha-nisms is to form a common attention eld in thecommunication process, to which the necessity of theuse of both the visual and auditory modality channelsrelates. In the communication process that is of triadiccharacter (childparentobject), there is an interpreta-

tion phenomenon that is related to giving a meaning tothe observed objects. Then, during intensive concen-tration on the given object (e.g., a ower), the childsimultaneously associates the stimuli coming fromthe object with the symbolic stimuli coming fromthe parent. Multisensory association of information in-volves snifng, touching, and seeing the object andreceiving symbolic information about it, so it requiresthe simultaneous use of the different modalities of thesensesvision, hearing, smell, and touch. In DCDP,in contrast, owing to their deafness, the difculties inmultisensory association of stimuli emerge in the earlylife period. When encountering unfamiliar objects,these children can observe, yet they cannot simulta-neously receive information about their perceptual andfunctional properties. In this connection, as Tomas-zewski (2000a, 2001) notices, the process of shifting thechilds attention from the object to a person conveyingthe information about this object and back occurs. Thedelayed reception of information about objects in earlychildhood can lead to a restricted understanding of thepart-whole relation in the early school period, but onlyin some spheres. It refers to some verbal analogy tasks,such as juice : plant :: blood : ____. This task containsthe biological concepts that are certainly familiar toyounger DCDP, but which, linked according to thepart-whole relation, give them difculties in under-standing this relation.

In contrast, the series of verbal analogy tasksbased on the causality relation has not revealedsignificant differences between environmental groupsof DCDP and HCHP at both levels of age. Thiscould be evidence of the similarity between thecognitive structures that are responsible for theoperations on the most difficult relation, the causalityrelation. Yet, one should point out that in the case of both populations, the process of becoming mental

Table 16 The means and standard deviations in tasks of gural-geometric analogies based on the part-whole relation

Age group

DCDP DCHP HCHP

M SD M SD M SD

Younger 2.6250 1.359 2.5625 1.091 2.9500 1.699Older 4.5625 1.091 3.4375 1.929 4.2000 1.641



environmental groups at both levels of age and (b) age groups (younger and older) in each populationA comparison of youngerenvironmental groups



DCDP-DCHP, t (30) 0.14, ns DCDP-DCHP, t (24) 2.03, ns DCDP, t (30) 4.46, p , 0.05DCDP-HCHP, t (34) 0.62, ns DCDP-HCHP, t (34) 0.75, ns DCHP, t (24) 1.58, nsDCHP-HCHP, t (29) 0.82, ns DCHP-HCHP, t (34) 1.28, ns HCHP, t (38) 2.37, p , 0.05




16/23

operations independent of the categories of relationson which these operations are performed occurs inearly adolescence. Older DCDP and HCHP, incontrast to younger DCDP and HCHP, have shownan understanding of logical relations regardless of their category; this is evidence of attaining a formalthinking level. Furthermore, when solving the verbalanalogy tasks based on the causality relation, it isnecessary to coordinate the mental operations becausethese tasks require the recognition of the direct (notindirect) causal connection between the specificphenomena, and the knowledge about these phe-nomena itself is not sufficient to solve these taskscorrectly. The coordination of mental operations ispossible when these operations become independentof the category of logical relations (Jurkowski, 1967).This suggests that early and consistent sign-languagecommunication canto some extentinfluence theemergence of abilities to coordinate the mentaloperations, as does early and consistent spoken-language communication.

With regard to deaf children who do not haveearly and consistent contact with sign language (thatis, DCHP), they have shown a considerably lowerlevel of understanding of the different verbalrelations than do DCDP and HCHP. The lack of understanding by younger DCHP of some expres-sions in the analogy terms, and their difficulties in

understanding particular relations, can indicate thatthese children have a restricted conceptual systemthat does not allow them to rapidly grasp theconcepts of opposite, part-whole, and causality whilesolving the tasks. These data support the claim thatdeaf childrens early and consistent contact solelywith spoken language deprives them of the possibilityof knowing the world fully. This is because theinformation that reaches them by means of linguisticcode is not of visual character and often is omittedby them in cognitive and linguistic processing

(Tomlinson-Keasey & Kelly, 1974, 1978) if thisprocessing takes place in their mind. Moreover, theconsiderably lower level of understanding of thecausality relation found in older DCHP is a sign of lack of coordination of logical operations, and thustheir thinking does not attain a formal operationallevel. The cause of their delay may not only be the

lack of early and consistent sign-language communi-cation, but also an incomplete participation in thefamily life and events connected with it. Talkingabout some events connected with the family andsocial life in sign language is one factor that enrichesdeaf childrens linguistic and nonlinguistic experi-ence; it promotes the development of the cognitivestructures responsible for the coordination of mentaloperations. 5

The results of the conducted study in terms of theunderstanding of the numerical relations are incon-sistent with the second detailed hypothesis that theeducation of deaf children with early and consistentexposure to sign language in the artificial sign systemor spoken language contributes to the emergence of theabilities to understand the numerical relations in themsomewhat later than in hearing children with earlyand consistent exposure to spoken language. The lackof quantitative and qualitative differences betweenDCDP and HCHP in terms of their understanding of the different numerical relations may be evidence thata similar complex conceptual system does function intheir minds. This lack of significance could result fromthe fact that DCDPalready in preschool and theearly school periodprobably learn the variousarithmetic actions of reversible character, such asaddition, subtraction, multiplication, and division,from their parents. Such a situation is possible because

these parents use sign language in the interaction withtheir children and can convey basic mathematicalknowledge to them by means of this language. Earlylearning of arithmetic actions allows children tomemorize them and later leads to a widening of theknowledge about relations occurring between mathe-matical functions.

The somewhat poorer performances of both agegroups of DCHP in the numerical analogy tasks compared to performances of DCDP and HCHP should be attributed to a low level of cognitive

competence, which occurs in the early period of theirlife and which is a consequence of hearing parents useof spoken language in their interaction with thesechildren. This language does not fulfil the specificconditions of deaf childrens development becauseit requires the use of the auditory channel in thecommunication process. Deaf children cansolely in



17/23

a visual wayreceive the semantic, symbolic, andspatial information that in turn they process into newinformation also of visual character. Young DCHPcannot receive a variety of information referring to theperceptual and functional properties of perceivedobjects and phenomena, and for this reason theyprocess and produce less information; it makes themindividuals of a low thinking level. Hence, it is possiblethat these children may not show the readiness to learnnumerical concepts in school. Such an interpretationis consistent with Sisco and Andersons (1980) andZwiebels (1987) viewpoints. Yet, it is important to notethat older DCHP, in contrast to older DCDP andHCHP, correctly solved the numerical analogy tasksdepending on the category of logical relations that arecontained in them. This means that they are unableto coordinate the mental operations, showing thusa concrete thinking level. This also allows us to believethattheir cognitivestructures accountable for theopera-tions on the most difficult relationthe part-wholerelationdo not attain a higher level of functioningbecause of the delay in general mental development,which is found in the early period of their life. Thisdelay is due probably to a lack of communication insign language.

The results in terms of the childrens understand-ing of the spatial relations are partially consistent withthe last detailed hypothesis that early and consistent

exposure to the grammatical structures of sign lan-guage, in contrast to early and consistent exposureto the grammar of spoken language, determines thedifferences between deaf and hearing children in therecognition of spatial relations, to hearing childrensdisadvantage. Although no significant superiority of DCDP to HCHP is found in terms of their un-derstanding of the spatial relations, the differencesbetween them are so large that one can claim about aninfluence of grammatical structures of sign language onthe intensive development of visual-spatial abilities in

deaf children. In reference to the figural-geometricanalogy tasks based on the opposite relation, someimprovement in their understanding of this relation isobserved in DCDP from both age groups. These tasksrequire children to discover the principle of figureidentity when taking the 180 rotation of figures intoconsideration, 6 so the sophisticated improvement of

DCDP with regard to their understanding of theopposite relation can be connected with the reversalinterpretation, which is the skill revealed quite fre-quently in the processing of sign language grammar 7

(Emmorey, Kosslyn, & Bellugi, 1993). This means thatif the perceptual-spatial experience of DCDP corre-sponds with the content of abstract-spatial reasoningtasks, these children can show a higher level of under-standing of the logical spatial relations than HCHP.Furthermore, on account of the fact that the seriesof gural-geometric analogy tasks was not of speedcharacter, signicant differences could not occur be-tween the mentioned environmental groups. It isconrmed by the results of experiments conducted byAmerican psychologists (Braden, 1987; Emmorey,Kosslyn, & Bellugi, 1993), who state that DCDP canbe faster than HCHP in the visual-spatial informationprocessing.

That there is a quantitative and qualitativesimilarity between the particular environmentalgroupsDCDP, DCHP, and HCHPat both levelsof age in their understanding of the spatial relationsmay confirm Furths (1971b, 1991) thesis about thelack of influence of language on the development of some nonverbal spheres of logical thinking in deaf children. This is all the more so because he considersthe possibility that other factors play a role in thedevelopment of a conceptual system in deaf children,

especially DCHP, namely that DCHP construct theirown verbal and nonverbal symbols (gestures andimages), which can help them to develop logicalthinking. Furths reflections are confirmed by theresults of preliminary studies conducted by Tomas-zewski (2000b, 2001), who examined the influence of biological factors on the formation and developmentof grammatical structures of gestural language inDCHP having contact solely with spoken language inthe preschool environment. In such an environment,these children constructed three- and four-gesture

utterances that in the grammatical respect resembledsign language structures, so they were createdaccording to the rules of visual-spatial grammar.Moreover, these utterances consisted not only of gestures, but also of negations, with the gestures beingthe lexical means of a symbolic form, that is, they werenot idiosyncratic gestures. It seems that deaf child-



18/23

rens use of symbolic gestures can reflect their generalmental development and to some extent influence thedevelopment of logical thinking, as Furth (1971b,1991) suggests. In contrast, the understanding and useof negations by them could point to an understandingof specific social phenomena (e.g., a ban on makingsomething, denying the facts). So communication ingestural language, possessing some elements of visual-spatial grammar, can promote both the formation of concepts (e.g., spatial concepts) and the intensiveprocessing of visual-spatial information, and then leadto the understanding of spatial relations by DCHP inthe school period.

The understanding and use of logical relationsregardless of their category as a sign of formal thinking(Jurkowski, 1961, 1967) is an example of metacognitiveskill that is connected with the ability to consciouslyand purposefully use the higher-order relations whensolving the tasks irrespective of task content (Gos-wami, 1991). This metacognitive skill is found amongolder children only in the populations of DCDP andHCHP. When solving the successive analogy tasksbased on the most difficult relations (verbal causalityrelation, numerical part-whole relation, and spatialpart-whole relation), these children showed a high levelof conscious use of these relations; owing to this,most analogy tasks were correctly solved. In order tounderstand and use the mentioned logical relations, it

is necessary to possess the general knowledge as wellas select the mental operations appropriate for thetask content and combine them in a systematic way.Furthermore, the skill of solving the verbal andnonverbal analogy tasks that contain opposite, part-whole, causality, and class membership relations isthe ability to recognize the similarity between logicalrelations (not the similarity between the exteriorfeatures within these tasks). The fact that this skill isobserved entirely in DCDP and HCHP means thatthe concepts of opposite, part-whole, causality, and

class membership do function in their minds, so thesechildren do not rely on the perceptually visible featuresof tasks while solving them, but are using a complexconceptual system.

In contrast, DCHP from both age groups have lesslinguistic experience in the area of sign language. Thisexperience consists of both the familiarity of signs

as the equivalents of words and the knowledge aboutthe relations occurring between signs, being the sym-bolic labels of reality elements. The lack of some partof knowledge about the relations found in DCHPdefinitely distinguishes them from children from otherpopulations; it appears to be consistent with Goswa-mis (1991, 1996) thesis, according to which the skillof solving the analogy tasks is determined by thefamiliarity of relations on which these tasks are based.As regards their nonlinguistic experience, which ismost frequently connected with casual learning, it is sorich that it enables them to solve many nonverbal tasks,such as the figural-geometric analogy tasks and somenumerical analogy tasks. Attention should be paid tothe character of tasks presented to children. The verbaland numerical analogy tasks, in contrast to the figural-geometric analogy tasks, comprise linguistic andmathematical symbols that are socially shared symbols,and hence it is necessary to acquire them in order toachieve the possibility of functioning in the societynormally. Yet, because of a low level of cognitivecompetence resulting from the lack of exposure to signlanguage in early childhood, DCHP have difficulties inacquiring these symbols; owing to this, they showa delay in the development of understanding someverbal and numerical relations and thus do not attaina formal thinking level at the appropriate time. Itdoes not mean that the ability to coordinate mental

operations is not available to these children. Thisability can emerge in their development later than inthe case of HCHP (Furth, 1991; Furth & Youniss,1975). However, one should carefully interpret theoutcomes obtained by DCHP in the verbal analogytasks, because it could appear that poor linguisticabilitiesrather than delays in the analogical reasoningdevelopmentare accountable for the lower results ingroups of DCHP. The evidence for this is that theiroutcomes in the figural-geometric analogy tasks arecomparable to the results of children from the other

groups, indicating that these children are able to reasonby analogy.In summary, Sharpes (1985) claim that deaf

childrens gestural-visual communication mode is in-ferior to hearing childrens oral-aural communicationmode is not confirmed by the results of this study.One should state that not only early and consistent



19/23

sign-language communication, but also deaf chil-drens own activity, can affect the character of theirlinguistic and nonlinguistic experiences. Knowledge,thinking tools, and linguistic patterns are conveyedby adults by means of sign language to thesechildren, and their own activity is determined bythe biological endowment of organism (the quality of genes); these factorsenvironmental and genetic interact, leading through the successive stages of their development to the emergence of complexforms of cognition. On the other hand, it is worthemphasizing that the results of this study should betreated with caution. As mentioned earlier, the

hearing children were examined by a researcherwho is deaf but who also uses Polish languagefluently. It also should be pointed out that thehearing childrens results in this study are compara-ble to the (hearing) childrens outcomes in thestudies by Jurkowski (1961, 1967) and Pietrulewicz(1983). Therefore, it would be necessary to conducta similar study of the development of analogicalreasoning in deaf and hearing children in order toconfirm the results of the presented study. Thisstudy could be carried out by both the deaf andhearing researchers.

Appendix 1: A Set of Verbal Analogy Tasks

The series of verbal analogy tasks based on the opposite relation1. well : badly :: nicely : ____ 13. lie : truth :: love : ____ 4. light : heavy :: expensive : ____ 16. far : near :: long : ____ 7. sad : merry :: industrious : ____ 19. clearness : darkness :: white : ____

10. ill : healthy :: rich : ____ 22. sell : buy :: rest : ____

The series of verbal analogy tasks based on the part-whole relation

2. bath : bathroom :: blanket : ____ 14. corridors : school :: streets : ____ 5. shelves : cupboard :: storeys : ____ 17. drawers : desk :: pockets : ____ 8. minute : hour :: centimeter : ____ 20. juice : plant :: blood : ____

11. month : year :: day : ____ 23. wool : sheep :: feathers : ____

The series of verbal analogy tasks based on the causality relation

3. poison : death :: food : ____ 15. learn : knowledge :: work : ____ 6. frosts : ice :: rain : ____ 18. sadness : weeping :: joy : ____ 9. inattention : accident :: lack of food : ____ 21. downpour : wetly :: heat : ____

12. bomb : ruins :: thunderbolt : ____ 24. cloud : rain :: chill : ____

Appendix 2: A Set of Numerical Analogy Tasks (The Meanings of Symbols: aThe Opposite Relation,bThe Class Membership Relation, cThe Part-Whole Relation)

1a. (2 1) : (2 1) :: (7 3) : (____)(2 1) a(7 3) b(8 2) c(3 7) d

2b. (8 8) : (8 2) :: (2 2 2 2) : (____)(3 3) a(2 8) b(8 2) c(2 4) d



20/23

3c. 4 : 8 :: 3 : (____)(9) a(6) b(10) c(12) d

4b. (4 4 4) : (4 3) :: (5 5 5 5) : (____)(3 4) a(5 4) b(4 3) c(5 4) d

5a. (8 4) : (8 : 4) :: (6 2) : (____)(6 : 2) a(12 : 2) b(4 1) c(8 : 2) d

6c. 8 : 24 :: 5 : (____)(10) a(20) b(16) c(15) d

7b. (6 6 6 6) : (6 4) :: (2 2 2 2) : (____)(4 6) a(2 6) b(2 4) c(6 2) d

8a. (3, 6, 9) : (18, 15, 12) :: (6, 9, 12) : (____)(15, 18, 21) a(13, 14, 15) b(13, 10, 7) c(4, 7, 10) d

9c. 6 : 18 :: 4 : (____)(8) a(16) b(12) c

(20) d10b. (1 1 1) : (1 3) :: (4 4) : (____)

(4 2) a(4 2) b(1 4) c(3 1) d

11a. (2, 4, 6) : (6, 4, 2) :: (3, 6, 9) : (____)(2, 3, 4) a(3, 6, 10) b(9, 6, 3) c(4, 7, 10) d

12c. 5 : 10 :: 10 : (____)(15) a(20) b(25) c(30) d

13b. (2 2 2 2 2) : (2 5) :: (3 3 3) : (____)(3 3) a(2 5) b(3 2) c(5 2) d

14a. (2, 3, 4) : (4, 3, 2) :: (3, 4, 5) : (____)(6, 7, 8) a(5, 4, 3) b(4, 5, 6) c(3, 2, 4) d

15c. 4 : 16 :: 2 :(____)(8) a(6) b(12) c(4) d

16b. (3 3 3 3) : (3 4) :: (6 6 6) : (____)(6 6) a(4 3) b(6 3) c(3 3) d

17a. (2, 4, 6) : (8, 6, 4) :: (6, 8, 10) : (____)(12, 10, 8) a(3, 5, 7) b(4, 6, 8) c

(12, 14, 16) d18c. 9 : 27 :: 1 : (____)

(1) a(3) b(2) c(54) d



21/23

Appendix 3: A Set of Figural-Geometric Analogy Tasks (The Meanings of Symbols: aThe OppositeRelation, bThe Part-Whole Relation)

Notes

1. According to McNeils (1985) and Marscharks (1993)claims, gestures are verbal symbols, which can enable young deaf and hearing children to recognize the specic properties

underlying the meanings of their rst signs and words. So thesegestures perform a symbolic function that consists not only inrepresenting the reality elements, but also in enabling thesechildren to acquire all semantic features of signs and words.

2. Sharpe means the sign language used by deaf personswithin the deaf community, so this communication mode is notbased on the spoken language grammar.

3. An articial sign system is a gestural representation of spoken language, that is, the signs that accompany the words areproduced while formulating the sentences according to the rulesof spoken-language grammar. This is an entirely communica-tive system (not a linguistic system) that has been created foreducational goals. It is used mostly by hearing teachers ininteractions with deaf children. In this case, the articial signsystem is based on Polish.

4. As the analysis of results showed, in task 13 (lie : truth ::love : ____ ), DCDP from both age groups (0% and 6.25%)obtained considerably lower outcomes than HCHP from bothage groups (20% and 65%).

5. One should mention that four younger DCHP and twoolder DCHP had deaf siblings; it may suggest that these childrenpartially communicated with their siblings in sign language.

Although they can communicate with each other in everyday lifesituations by means of sign language, it does not indicate thatthey have consistent contact with this language. It is important totake into account whether their hearing parents, who have richexperiences, are inclined to accept sign language as an alternative

communication mode. As mentioned earlier, neither hearingparent of the deaf children used sign language in the interactionswith his/her deaf child. On the other hand, the deaf siblings maynot possess full linguistic competence in sign language, so it isdifcult to treat these children as having consistent contact withdeveloped sign language. In a family where all memberscommunicate with each other in sign language, the linguisticinteractions between them become more intensive. It could, inturn, enrich these childrens linguistic and nonlinguistic experi-ences.

6. The exception is the last gural-geometric analogy taskbased on the opposite relation. In this task, the gure in the Bterm is the mirror image of the gure in the A term, so this taskdoes not require the 180 rotation of the gure.

7. The reversal interpretation is a mental process whose usefacilitates the understanding of signed utterances formulated bythe signer. In this language, the syntax is organized spatially.This entails the necessity of forming the signed space whileconstructing the sentences. The signed space reects the realworld and is represented in the signers mind. This signer,creating his/her utterances in sign language, uses the spacefrom his or her own point of view, so the observer, being also



22/23

a signer, must mentally reverse the spatial relationships in orderto compute the correct spatial array (Lillo-Martin, 1997, p. 97)and understand the mentioned signed utterances.

References

Akamatsu, C. T., Musselman, C., & Zwiebel, A. (2000). Natureversus nurture in the development of cognition in deaf

people. In P. E. Spencer, C. J. Erting, & M. Marschark(Eds.), The deaf child in the family and at school (pp. 255 274). Mahwah, NJ: Erlbaum.

Alexander, P. A., Willson, V. L., White, C. S., & Fuqua, J. D.(1987). Analogical reasoning in young children. Journal of Educational Psychology, 79, 401408.

Alexander, P. A., Willson, V. L., White, C. S., Fuqua, J. D.,Clark, G. C., Wilson, A. F., et al. (1989). Development of analogical reasoning in 4- and 5-year-old children. CognitiveDevelopment, 4, 6588.

Biela, A. (1981). Psychologiczne podstawy wnioskowania przezanalogie [Psychological foundations of analogical inference].Warsaw, Poland: PWN.

Braden, J. P. (1987). An explanation of the superior performance

IQs of deaf children of deaf parents. American Annals of theDeaf, 132, 263266.

Braden, J. P. (1994). Deafness, deprivation, and IQ. New York:Plenum Press.

Brill, R. G. (1974). The superior IQs of deaf children of deaf parents. In P. J. Fine (Ed.), Deafness in infancy and earlychildhood (pp. 151161). New York: Medcom.

Conrad, R., & Weiskrantz, B. C. (1981). On the cognitive abilityof deaf children with deaf parents. American Annals of theDeaf, 126, 9951003.

Courtin, C. (1997). Does sign language provide deaf childrenwith an abstraction advantage? Evidence from a categoriza-tion task. Journal of Deaf Studies and Deaf Education, 2,161171.

Courtin, C. (2000a). Sign language and cognitive exibility.Current Psychology Letters: Behaviour, Brain and Cognition,3, 1930.

Courtin, C. (2000b). The impact of sign language on thecognitive development of deaf children: The case of theoriesof mind. Journal of Deaf Studies and Deaf Education, 5, 266 276.

Courtin, C., & Melot, A.-M. (1998). The development of theories of mind in deaf children. In M. Marschark & M. D.Clark (Eds.), Psychological perspectives on deafness (Vol. 2,pp. 79102). Mahwah, NJ: Erlbaum.

Dolman, D. (1983). A study of the relationship between syntacticdevelopment and concrete operations in deaf children. American Annals of the Deaf, 128, 813819.

Emmorey, K., Kosslyn, S. M., & Bellugi, U. (1993). Visualimagery and visual-spatial language: Enhanced imageryabilities in deaf and hearing ASL signers. Cognition, 46,139181.

Farris, M. A. (1994). Sign language research and Polish SignLanguage. Lingua Posnaniensis, 36, 1336.

Furth, H. G. (1964). Research with the deaf: Implications forlanguage and cognition. Psychological Bulletin, 62, 145164.

Furth, H. G. (1966). Thinking without language: Psychological implications of deafness. New York: Free Press.

Furth, H. G. (1971a). Education for thinking. Journal of Rehabilitation of the Deaf, 5, 717.

Furth, H. G. (1971b). Linguistic deciency and thinking:Research with deaf subjects 19641969. Psychological Bulletin, 76, 5872.

Furth, H. G. (1991). Thinking without language: A perspective

and review of research with deaf people. In D. P. Keating &H. Rosen (Eds.), Constructivist perspectives on developmental psychopathology and atypical development (pp. 203227).Hillsdale, NJ: Erlbaum.

Furth, H. G., & Youniss, J. (1975). Congenital deafness andthe development of thinking. In E. H. Lenneberg& E. Lenneberg (Eds.), Foundations of languagedevelopment: A multidisciplinary approach (Vol. 2, pp. 167 176). New York: Academic Press.

Ga1kowski, T. (1998). Przestrzenne i ruchowe komponentykomunikacji z dzieccmi g1uchymi [The spatial and motorcomponents of communication with deaf children]. Warsaw,Poland: Polish Committee of Audiophonology.

Goldman, S. R., Pellegrino, J. W., Parseghian, P., & Sallis, R.

(1982). Developmental and individual differences in verbalanalogical reasoning. Child Development, 53, 550559.

Goswami, U. (1989). Relational complexity and the develop-ment of analogical reasoning. Cognitive Development, 4,251268.

Goswami, U. (1991). Analogical reasoning: What develops? Areview of research and theory. Child Development, 62, 122.

Goswami, U. (1996). Analogical reasoning and cognitive de-velopment. In H. W. Reese (Ed.), Advances in child development and behavior (Vol. 26, pp. 91138). New York:Academic Press.

Goswami, U., & Brown, A. L. (1990a). Higher-order structureand relational reasoning: Contrasting analogical and the-matic relations. Cognition, 36, 207226.

Goswami, U., & Brown, A. L. (1990b). Melting chocolate andmelting snowmen: Analogical reasoning and causal relations.Cognition, 35, 6995.

Hoemann, H. W. (1991). Piagetian perspectives on researchwith deaf subjects. In D. P. Keating & H. Rosen (Eds.),Constructivist perspectives on developmental psychopathologyand atypical development (pp. 229245). Hillsdale, NJ:Erlbaum.

Inhelder, B., & Piaget, J. (1970). Od logiki dziecka do logiki m1 odzie _zzy [The growth of logical thinking from childhood toadolescence]. Warsaw, Poland: PWN.

Jurkowski, A. (1961). Rozwoj zdolnosci rozumowania przezanalogix w zale_zznosci od rodzaju stosunkow, na ktorychopiera six analogia [The development of analogical reasoningabilities depending on the kind of relations on which analogyis based]. Psychologia Wychowawcza, 4, 170176.

Jurkowski, A. (1967). Rozumowanie przez analogie u dzieci wwieku szkolnym [Analogical reasoning in children of schoolage]. Warsaw, Poland: PWN.

Jurkowski, A. (1990). Adaptacja testow s1ownikowych [Theadaptation of vocabulary tests]. In A. Ciechanowicz (Ed.),Kulturowa adaptacja testo w [Cultural adaptation of tests]



23/23

(pp. 7792). Warsaw, Poland: Polish PsychologicalAssociation.

Klima, E. S., & Bellugi, U. (1979). The signs of language.Cambridge, MA: Harvard University Press.

Kusche, C. A., Greenberg, M. T., & Gareld, T. S. (1983).Nonverbal intelligence and verbal achievement in deaf adolescents: An examination of heredity and environment. American Annals of the Deaf, 128, 458466.

Lillo-Martin, D. (1997). The modular effects of sign languageacquisition. In M. Marschark, P. Siple, D. Lillo-Martin,R. Campbell, & V. S. Everhart (Eds.), Relations of languageand thought: The view from sign language and deaf children(pp. 62109). New York: Oxford University Press.

Lunzer, E. A. (1965). Problems of formal reasoning in testsituations. Monographs of the Society for Research in Child Development, 30 (2, Serial no.100), 1946.

Marschark, M. (1993). Psychological development of deaf children.New York: Oxford University Press.

Marschark, M., Siple, P., Lillo-Martin, D., Campbell, R., &Everhart, V. S. (1997). Relations of language and thought: Theview from sign language and deaf children. New York: OxfordUniversity Press.

McNeill, D. (1985). So you think gestures are nonverbal?Psychological Review, 92, 350371.

Meadow, K. P. (1980). Deafness and child development. Berkeley:University of California Press.

Ottem, E. (1980). An analysis of cognitive studies with deaf subjects. American Annals of the Deaf, 125, 564575.

Panasiuk, M. (1990). Rozwo j mys lenia dzieci g 1 uchych [Thedevelopment of thinking in deaf children]. Warsaw, Poland:WSiP.

Panasiuk, M. (1995). Eksperymentalne badanie umiej x tnoscirozumowania przez analogix u dzieci z defektem s1uchu[Experimental study of analogical reasoning skills inchildren with hearing impairment]. In J. Stachyra (Ed.),

Wybrane metody diagnostyczne w surdopsychologii [Theselected diagnostic methods in deaf psychology] (pp. 121 150). Lublin, Poland: UMCS Press.

Pietrulewicz, B. (1983). Rozwo j rozumowania przez analogie udzieci niewidomych w wieku szkolnym [The developmentof analogical reasoning in blind children of school age].Wroc 1aw, Poland: Ossolineum Press.

Sharpe, S. L. (1985). The primary mode of human communi-cation and complex cognition. American Annals of the Deaf,130, 3946.

Sisco, F. H., & Anderson, R. J. (1980). Deaf childrensperformance on the WISC-R relative to hearing status of parents and child-rearing experiences. American Annals of the Deaf, 125, 923930.

Sternberg, R. J., & Nigro, G. (1980). Developmental patterns

in the solution of verbal analogies. Child Development, 51,2738.Tomaszewski, P. (1999). Depatologizacja g 1uchoty: Droga do

dwuj x zycznosci i dwukulturowos ci w wychowaniu dzieckag1uchego [The depatologization of deafness: A way tobilingual and bicultural education of deaf child]. InR. Ossowski (Ed.), Kszta 1 cenie specjalne i integracyjne [Thespecial and integrated education] (pp. 146155). Warsaw,Poland: MEN.

Tomaszewski, P. (2000a). Rozwoj jx zykowy dziecka g1uchego:wnioski dla edukacji szkolnej [The linguistic development of deaf child: Implications for school education]. Audiofonolo- gia, 16, 2157.

Tomaszewski, P. (2000b). Two rczos c jezykowa dzieci g 1 uchych w

ro _zznych sytuacjach zabawowych a koncepcje natywistyczne[The linguistic creativity of deaf children in different playsituations and the conceptions of nativism]. Paper presentedat the scientic session of Polish Committee of Audio-phonology, Warsaw, March.

Tomaszewski, P. (2001). Sign language development in youngdeaf children. Psychology of Language and Communication,5 (1), 6780.

Tomlinson-Keasey, C., & Kelly, R. R. (197

the development of analogical reasoning in

Documents