Journal of Russian and East European Psychology, vol. 41, nos. 3/4,MayJune/JulyAugust, pp. 11528. 2003 M.E. Sharpe, Inc. All rights reserved.ISSN 10610405/2003 $9.50 + 0.00.

T.V. AKHUTINA

The Structure of the IndividualMental Lexicon from the Standpointof L.S. Vygotskys Ideas

Research on the structure of an individuals mental lexicon has shownthat, among the set of concepts belonging to a single semantic field,there is typically one particular concept that stands out. The pro-cesses of storage, identification, and retrieval of this concept arefaster and more successful than they are for the others. For example,in the vertical hierarchy of the concept fruitappleantonovka[McIntosh to give a familiar U.S. exampleTrans.], apple playsthis type of lead role and also occupies the most prominent positionin the horizontal series applepearplumcherriesorange.Concepts with such characteristics are referred to as basic, source,or primary concepts. What is the reason for their special status?

R. Brown (1958), who first raised this issue, associated the spe-cial status of one of the levels in a conceptual hierarchy with thefact that this level has maximum utility in the majority of con-texts. The child first learns words at the level of distinction that ismost useful to him. Thus, it is important for children to be able to

115

English translation 2003 M.E. Sharpe, Inc., from the Russian text 1994Moskovskii Gosudarstvennyi Universitet and Vestnik Moskovskogo Universiteta.Problema stroeniia individualnogo leksikona cheloveka v svete idei L.S.Vygotskogo. Psikhologiia, 1994, no. 4, pp. 4481. A publication of the Dept. ofPsychology, M.V. Lomonosov Moscow State University.

Translated by Lydia Razran Stone.

116 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

distinguish an apple from an orange, while the distinction betweena Golden Delicious and a Granny Smith [I have substituted Ameri-can for Russian apple varietiesTrans.] is not significant. Analo-gously the word tree is more useful than the word oak, and dog ismore useful than animal. The idea that the basic concepts werethose with maximum utility has been further developed in the worksof ethnographers (Berlin, 1972; Berlin et al., 1973), who showedthat the names of plants and animals in folk biology are dividedinto five (or six) levels: (1) unique names of the most primitive clas-sifier: plant, animal; (2) names of categories of living things: tree,bush; (3) generic names: pine, oak, birch; (4) species names: dwarfbirch; (5) varietal names: Karelian dwarf birch.

These ethnographers suppose that the basic level is occupied bygeneric names, which are at the center of the hierarchy. In thelanguages of the world this level contains the most terms and theterms that are the most useful. Thus, terms at this level providegreater discriminative ability than the hierarchical levels above itas well as simpler names (a simpler classification system) than thelevels below it.

Agreeing with this point of view, E. Rosch and her colleagues(Rosch et al., 1976) add that generic names are, on the one hand,relatively close to each other in semantics (as compared to thenames at the higher levels), and, on the other hand, are relativelyfar away from representatives of neighboring categories (as com-pared to the names at the lower levels). The use of generic namesis most convenient, expedient, and frequent, and, for this reason,children learn these terms earlier than the more general or specificterms at other levels.

However, investigations by German scholars (Khoffman, 1986;Hoffmann and Ziessler, 1982) have demonstrated that primary(basic) concepts may occupy different positions in the generic-specific hierarchy. For example, in the hierarchy, vehiclecarVolkswagen, the primary concept coincides with the generic name(car). However, in the hierarchy, treedeciduous treebirch, thebasic concept (tree) is located two levels above the generic name(birch).

MARCHAPRIL 2003 117

Is it possible to determine what place primary concepts occupyin generic-specific hierarchies? I. Hoffmann and M. Ziessler con-ducted an experiment in which subjects were presented with vari-ous concepts and asked to name their features. The authorsidentified two forms of (mental) representation. Concepts to whichsubjects attributed mainly pictorial features were called sensori-ally represented concepts. Concepts that were generally describedby abstract features and examples were called categorically repre-sented concepts (Hoffmann and Ziessler, 1982).1 These research-ers constructed a hierarchy whose lower levels were occupied bysensorially represented concepts, while the higher levels weredevoted to categorically represented ones. They then defined theprimary concepts as the most abstract concepts that were stillsensorially represented, that is, the highest level of the sensori-ally represented concepts. Such concepts can be identified morerapidly and have the greatest associative strength, as shown bythe fact that subjects generate the greatest number of features forthese concepts.

We reanalyzed the results cited by the above authors and foundthat it is not possible to accurately identify the primary conceptas being the highest hierarchical level among the sensorially rep-resented concepts in a given hierarchy. This is particularly clearin the musical instrument hierarchy. Subsidiary to this highestlevel concept, are two at the same level wind instruments andstring instruments. For the first of these categories abstract de-scriptors are predominant, while sensory features predominatefor the second. Consistent with this, the concepts at the next(lower) level are automatically treated differently: flute behaveslike a primary concept, while violin is a secondary one. How-ever, both flute and violin would be considered primary on thebasis of number of descriptors supplied insofar as the number offeatures subjects provided for them is higher than that for otherrepresentatives of this hierarchy. The situation is analogous forthe weapons hierarchy.

Thus, we can see that none of the approaches considered allowsus to accurately define the level of the basic concept in a generic-

118 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

specific hierarchy. U.S. investigators, using formal logical crite-ria, identify the central level of the generic-specific hierarchy asprimary. German psychologists combine formal logical and psy-chological approaches and provide an operational psychologicaldefinition of primary concepts, but at the same time they embedthese in rigid hierarchies (which fail to correspond to the naivelogic of language) on the basis of the erroneous assumption thatcategorically represented concepts must occupy the highest lev-els, while sensorially represented concepts lie at the lower ones.As we have already seen in the examples cited, in a number ofcases, this assumption is not justified. Furthermore, if we do notsimply divide concepts into two groups, but indicate their place inthe continuum of categoricallysensorially represented con-cepts, then we find that many higher concepts are closer to thesensory pole than lower concepts (e.g., treedeciduous tree;boatmotorboat).

Both of the models considered correspond to reality to someextent, but both allow too many exceptions. This suggests that thevery principle according to which basic concepts are identifiedmust be differentone that is consistently psychological, whichwe, like L.S. Vygotsky, understand to be historical genetic orgenetic-functional. If we are to describe primary concepts, we muststart by considering their origin and functioning.

On the basis of Vygotskys findings that the system of wordmeaning begins to form in a child while his thinking is prelogi-cal and based on complexes* (the same is undoubtedly true withrespect to philogeny) and that the child first learns words withconcrete, pictorial meanings, forming associations of the typewordgeneralized image, with the form of the generalizationchanging developmentally in accordance with various laws, wehypothesized that the meaning of the words first learned by achild will act as the initial (primary, basic) words for the con-struction of individual mental semantic fields (Akhutina, 1977;Akhutina, Petrenko, and Nistratov, 1978).

During our experimental verification of this hypothesis we were

MARCHAPRIL 2003 119

able to show that the primary meanings: (1) initially are imagemeaning complexes (sensorially represented; (2) because theyare images, they may, from a logical standpoint, be considerednonelementary; (3) are constructed according to the laws of think-ing in complexes; of these nonlogical structures, those that arereinforced by the linguistic context remain and those that are notso reinforced pass into a latent state or disappear; (4) are the refer-ence points for defining meanings acquired later (i.e., secondary)meanings in the same semantic field; (5) occupy a central positionin the semantic field, which makes them easy to retrieve, thusensuring that they will be used frequently.

In agreement with the ideas of R. Brown and E. Rosch, wehypothesize that primary concepts (which are the first words ofchildren) are maximally useful. It is significant for a young childto be able to distinguish between an apple and an orange or acarrot and a cucumber. Thus, for the categories of fruit and veg-etables it is the level of generic names that is the basic one andthat contains the primary meanings. At the same time, to a smallchild the difference between a birch and a linden or a bullfinchand a sparrow is not so significant and thus tree and bird (names ata higher level) form the primary concepts in the newly formingsemantic fields. The history of the semantic field determines itsstructure.

We have been speaking about the primacy of concepts in avertical series; however, members of horizontal series are also dis-tinguished by their functional characteristics.

When subjects were asked to list the features characteristic ofapples, pears, oranges, and other members of the fruit category,it was found that no single feature listed was common to all mem-bers of this category (Rosch and Mervis, 1975).

At the same time, when subjects are asked to classify membersof one or another category on the basis of how typical they are,they unanimously put the apple in the most typical category, andolives in the least typical one. If the subjects can evaluate thetypicality of category members, it may be hypothesized that they

120 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

have an idea of the most typical memberthe prototype. In Roschsopinion, this is just what they have in mind when they refer to thecategory. The prototypical category member, occupying the centralplace in the semantic category, has the greatest number of featuresin common with other members of this category and is furthestaway from representatives of other categories. All category mem-bers are grouped around the prototype according to the principleof family resemblance.

The degree of typicality determines how easy it is to operatewith the given concept. For example, if we ask subjects to evalu-ate the truth or falsity of the propositions: A crow is a bird, andA chick is a bird, then it will require less time to verify the firstproposition than the second, which names a nontypical categorymember.

The facts and proposed explanation advanced by E. Rosch andher colleagues accords fully with Vygotskys theory of conceptformation. Generalization of the thinking in complexes type andorganization of groups according to the principle of family resem-blance are two different descriptions of a single phenomenon.Despite their basic similarity, there is one difference between them.The U.S. investigators speak of the principle of typicality or fam-ily resemblance only with regard to the members of one category,one horizontal series. On the other hand, Vygotsky, who was mak-ing a more general assertion about the different types of verbalthinking, attributed a broader sphere of operation to complexes.

Despite the fact that Vygotsky is cited so often in semantic in-vestigations, his point of view is far from thoroughly or accuratelyunderstood (the review by the erudite and subtle scholar S. Carey[1982] is a case in point), and we will thus consider it in somewhatmore detail.

Experience with artificial concepts (using the Vygotsky-Sakharovprocedure) allowed Vygotsky to experimentally establish that themeaning of words develops during childhood and that the develop-ment goes through a series of stages (syncretism, complex,preconcept, concept), which involve different types of generaliza-

MARCHAPRIL 2003 121

tion.2 His conclusions here are well known. The following asser-tions are less well known. Every generalization structure is basedon a specific system of commonality and commonality relationshipsbetween general and particular concepts (Vygotsky, 1982, p.271). The commonality relationship of flowerrose differs inthe complex and concept structures. To a two-year-old child, whohas learned the word flower before learning rose, both conceptsstand alongside each other, and can replace each other, while toan eight-year-old child, one stands above the other and includesmore particulars.

Vygotsky believed that only at the highest developmental stagesof word meaning (and thus of relationships of commonality) doesa complete system of concepts arise that permits each concept tobe designated in an unlimited number of ways through use of otherconcepts to define its place in the system (Vygotsky, 1982, p. 273,and further). He also suggests that the new level of generalizationdoes not completely supersede the developmentally previous onein the linguistic consciousness of an adult; rather, old and newforms coexist in a relationship akin to geological stratification.Everyday thinking, which is based on everyday speaking, makesextensive use of the higher forms of complex-based thinking inthe form of pseudoconceptseveryday concepts, while scien-tific thinking relies on scientific concepts (ibid., pp. 168, 176).

The essential difference between everyday and scientific con-cepts is associated with the presence or absence of a system. Out-side a concept system, all that is possible are associationsestablished among objects, that is, empirical connections. When asystem of concepts exists, meta-empirical associations becomepossible (Vygotsky, 1982, p. 284) Thus, conceptual thinking dif-fers from thinking in complexes by virtue of the existence of asystem (cf. The different commonality relationships determinewhat types of operations are possible for a given level of thinking[ibid., p. 290]).

Let us now turn to two aspects of word meaningrelatednessto an object (= reference) and meaning per se. We will see that

122 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

the latter is part of a system and may be identified in terms of itsplace in the system, and that the laws of formal logic are appli-cable to this system. Reference, which is based on a pictorial asso-ciation between the word and the object, also entails associationbetween words, but these are complex-based associations of thefamily resemblance type. The meaning of a word that desig-nates a concrete object is thus stored in two coordinate systems:(1) in a formal logic-based hierarchical system utilizing a systemof meanings per se (categorical meanings), and (2) in a visualimage gallery of generalized images associated with the refer-ential meaning of the word.

These ideas of Vygotsky are supported by the most diverseempirical evidence. The natural division of meanings into sensoryand categorical representations, the characteristics of primary con-cepts, the typicality effects for everyday concepts and the size ofcategories for scientific conceptsall these facts can be explainedconsistently by Vygotskys ideas. For example, experiments onlexical recognition and naming of a word have shown that a con-crete word has two types of meaningperceptual and conceptualand that these are activated at differing rates (Flores dArcais,Schreuder, and Glazenborg, 1985).

The fact that Vygotskys ideas correspond to the findings aboutaphasia provides an important piece of evidence demonstratingthat they are correct. Special experiments targeted at analyzingthe retention of categorical meanings of words and referentialmeanings have shown that one of these can be impaired withoutimpairment of the other. Patients with a brain lesion localized inthe parietal-temporal-occipital areas (i.e., suffering from thesemantic aphasia syndromeLuria, 1947) have great difficultyoperating with categorical meanings. Categorical classification(either of words or of objects), retrieval of words with a givenmeaning (apple: fruit = dress:???) were the most difficult for thisgroup of patients (Akhutina and Malakhovskaia, 1985) When thefocus of the lesion is in the lower temporal-occipital cortex of thedominant cerebral hemisphere, patients have the most trouble oper-

MARCHAPRIL 2003 123

ating with the referential meanings. With my student N.V. Komolova,we developed an experimental methodology based on the idea thatunderlying such meaning is a generalized image-template, whichis used to segment the continuum of real phenomena. In these stud-ies, subjects were asked to view schematic depictions of animals,whose features changed smoothly along a single continuum andto divide them into groups.

In our experiment, in addition to perceptual classification, weused verbal classification in which words were to be placed ingroups not on the basis of generic-specific categories, but on thebasis of whether they referred to a single object (in particular, thewords: pussycat, kitty, and purr, had to be distinguished from vixen,sly, cub, yelps, or howl, pup). The difficulty of these tasks, whichwas intercorrelated for these tasks but not correlated with defects inperforming tasks involving categorical meanings, was greatest forpatients with damage located as described above. This suggests thatpatients with acoustic-mnestic and optical-mnestic (optical) apha-sia suffer primary impairment of referential meaning.

These data confirm Vygotskys ideas and suggest that the mean-ing of words denoting concrete objects is dually represented andmay be retrieved by two different routes: the categorical, withinthe verbal system of concepts (cf. the term concept network inLuria), and the imagistic, within the system of references. In thefirst case, the categorical (significative) meaning is activated,in the second the object denotation, to use Vygotskys term.Each meaning occupies a particular place in a continuum, at oneend of which are the most categorically represented and least im-agistically represented words and at the other end the most imag-istically and least categorically represented.

The fact that the two systems of word meaning generaliza-tions coexist in the linguistic consciousness of an adult is alsoclearly demonstrated by investigation of speech and thinking inpatients who have undergone a course of unilateral shock therapy(Balonov and Deglin, 1976; Deglin, Balonov, and Dolinina, 1983;Chernigovskaia and Deglin, 1986). The most striking data come

124 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

from word association experiments and syllogistic reasoning ex-ercises. Under conditions where the left hemisphere is depressed,in response to the stimuli words, there is an increase in responsesfrom the lexicosemantic area comprising the names of sets ofobjects and real phenomena. Particularly typical of this state areutterance-complexes (in Vygotskys term)naming of compo-nents of a concrete image that underlies the word (Deglin, Balonov,and Dolinina, 1983, p. 34). When the right hemisphere is depressed,the tendency to rubrification, to superimpose an abstract classifi-cation scheme on the external world is intensified. Responses basedon the principle of binary opposition, syntagmatic responses, re-sponses with semantically vacuous changes in words, and word for-mations all increase. The number of utterance-complexes reflectingindividual life experience, and concrete images decreases sharply(ibid., pp. 3637).

As noted above, in Vygotskys opinion, every generalizationstructure corresponds to a specific system of logical mental opera-tions that are possible with that structure (1982, p. 283). Study ofthe solution of logical problems after one hemisphere has beendepressed showed that the same individual may use either logical(when the right hemisphere is depressed) or empirical (when theleft hemisphere is depressed) methods of solution. Thus, after pre-senting the syllogism, Every artist can draw a rabbit. Drer is anartist. Can Drer draw a rabbit or not? a patient with a depressedright hemisphere answered, Drer can draw a portrait of his mothervery well, and he can also draw a rabbit, because he is an artist, Iknow of him (logical answer, reinforced by experience). The samepatient, after her left hemisphere had been depressed said: Drer?He probably can (with little confidence). I dont remember (thepatient does not use logical reasoning).3

Research on speech and thinking when one of the hemispheresis depressed poses the question of the lateralization of various formsof representation of meaning. While these investigations suggestthat the image glossary is stored in the right hemisphere and thecategorical in the left, the data from studies of aphasia suggest

MARCHAPRIL 2003 125

the possibility of both systems being impaired in cases of patho-logical lesions in the left hemisphere. As a working hypothesis,we take the viewpoint that object images are stored bilaterallyand categorical word meanings are stored in the left hemisphere.We understand, however, that this issue requires further carefulconsideration using data obtained from a variety of experimen-tal models.

Thus, we have considered the viewpoints in the literature on theorganization of the semantic field as components of an individualsmental lexicon. Studies performed on concrete lexicons have madeit possible to identify primary concepts, which occupy a favoredposition in the semantic field. These constitute the starting pointand center around which the entire field is structured according tothe principle of complexes (family resemblances). The imagisticnature of the generalization underlying the formation of a complexleads to a situation where a meaning logically subsumed under aparticular complex (cf. a chicken is a bird) may not be part ofthis complex but forms a separate semantic complex. The sys-tem of logical generic-specific hierarchies that forms later, un-der the influence of schooling, is superimposed on the first withoutreplacing it. This viewpoint, which follows from the experimentsand theoretical generalizations of L.S. Vygotsky, is confirmedby the data on aphasia.

Notes

1. Because this research used indirect criteria for judging the form of a conceptsrepresentationthat is, the verbal definitions provided by subjectsit is impor-tant to emphasize that there are other data in favor of this division, in particular,data on aphasia. Also testifying in favor of this distinction are analyses of problemswith naming in patients suffering from various forms of aphasia performed by A.R.Luria (1947), E.P. Kok (1967), E.D. Markova (1961), L.S. Tsvetkova (1972), N.G.Kalita (1974, 1979), T.V. Akhutina (1981a, 1981b, 1983), E.K. Warrington (1981a,1981b), A. Caramazza, R.S. Berndt, and H.H. Brownell (1982).

2. One can draw a parallel between the continuum complexpreconceptconcept proposed by Vygotsky and the continuum of E. Bates and B. MacWhinneygrouping on the basis of family resemblancegrouping on the basis of features

126 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

with different weightsgrouping according to a formal logic principle (Batesand MacWhinney, 1982).

3. For more detail on this see: Chernigovskaia and Deglin (1986). On theexistence of different types of thinking in members of modern cultures, see also,Riumina (1982); Tulviste (1982, 1988); Tulviste and Tulviste (1985); Frumkina(1984); Frumkina, Riumina, Mostovaia (1985).

References

Akhutina, T.V. 1977. [The Association Between Semantic Complexity, Time ofAcquisition and Word Frequency]. In [National Culture and Communica-tion], ed. Iu.A. Sorokina, pp. 1114. Moscow.

. 1981a. [Organization of the Human Lexicon Based on Data fromStudy of Aphasia]. In [Psychological Research in the Area of Vocabularyand Phonetics], ed. A.A. Zalevskaia, pp. 312. Kalinin.

. 1981b. [On Semantic Impairments in Aphasia Patients]. In [TheInvestigation of Speech. Abstracts from a Scientific Conference], pp. 67.Leningrad.

. 1983. [Impairments of Word Meaning in Aphasia Patients]. InProblems of Psychophysiology: Diagnosis of Impairments and Restorationof Human Psychological Functions. Abstracts from the Sixth Conference ofthe USSR Society of Psychologists, Part 2, pp. 33739. Moscow.

Akhutina, T.V., and Malakhovskaia, E.V. 1985. [The Significative WordMeaning in Aphasia]. Vestn. Mosk. Un-ta. Ser. 14, Psikhologiia, no. 1, pp.2938.

Akhutina, T.V.; Petrenko, V.F.; and Nistratov, A.A. 1978. [The SemanticComplexity of Words from the Standpoint of Linguistics and Psycho-linguistics]. Abstracts of the sixth All-union Symposium onPsycholinguistics, pp. 1012. Moscow.

Balonov, L.Ia., and Deglin, V.L. 1976. [Hearing and Speaking of the Dominantand Nondominant Hemispheres]. Leningrad: Nauka.

Bates, E., and MacWhinney, V. 1982. Functionalist Approaches to Grammar.In Language Acquisition. The State of the Art, ed. E. Wanner and L.Gleitman, pp. 173218. Cambridge.

Berlin, B. 1972. Speculations on the Growth of Ethno-Botanical Nomencla-ture. Language in Society, vol. 1, pp. 5186.

Berlin, B.; Breedlove, D.E.; and Raven, P.H. 1973. General Principles ofClassification and Nomenclature in Folk Biology. Amer. Anthropologist,vol. 75, pp. 29099.

Brown, R. 1958. How Shall a Thing Be Called? Psych. Rev., no. 65, pp.1421.

Caramazza, A.; Berndt, R.S.; and Brownell, H.H. 1982. The Semantic DeficitHypothesis: Perceptual Parsing and Object Classification by AphasicPatients. Brain and Language, vol. 15.

Carey, S. 1982. Semantic Development: The State of the Art. In Language

MARCHAPRIL 2003 127

Acquisition. The State of the Art, ed. E. Wanner and L. Gleitman, pp. 34789. Cambridge.

Chernigovskaia, T.V., and Deglin, V.L. 1986. [Metamorphic and SyllogisticThinking as a Symptom of Functional Asymmetry of the Brain]. In [Workson Sign Systems], vol. 19, pp. 6884. Tartu.

Deglin, V.L.; Balononov, L.Ia.; and Dolinina, I.B. 1982. [Language and theFunctional Asymmetry of the Brain]. In [Works on Sign Systems], vol. 16,pp. 3142. Tartu.

Flores dArcais, G.B; Schreuder, R.; and Glazenborg, G. 1985. SemanticActivation During Recognition of Referential Words. Psychol. Res., vol.47, no. 1, pp. 3350.

Frumkina, R.M. 1984. [Color, Meaning, and Similarity. Aspects ofPsycholinguistic Analysis]. Moscow: Nauka.

Frumkina, R.M.; Riumina, N.A.; and Mostovaia, A.D. 1985. [Typology ofClassification Decisions in Experiments on Free Classification (Based onSelf-Reports)]. In [The Linguistic and Psycholinguistic Structure ofSpeech], ed. R.M. Frumkina, pp. 2334. Moscow.

Hoffmann, J.; and Ziessler, M. 1982. Begriffe und ihre Merkmale. Z. frPsychologie, vol. 190, no. 1, pp. 4677.

Kalita. N.G. 1974. [On the Question of the Nature of Naming Impairments inAuditory Nominal Aphasia]. Psikhologicheskiie issledovaniia, no. 6, pp.3142. Moscow.

. 1979. [On the Issue of Impairment of Thinking in Patients withAcoustic-Amnestic Aphasia]. In [Problems of Aphasia and RehabilitativeTraining]. Moscow.

Khoffman, I. 1986. [Active Memory]. Moscow: Meditsina.Kok, E.P. 1967. [Visual Agnosias]. Moscow: Meditsina.Luria, A.R. 1947. [Traumatic Aphasias]. Moscow: Isdatelstvo AMN RSFSR.Markova, E.D. 1961. [Clinical and Pathophysiological Features of Amnestic

Aphasia]. In [Clinical and Pathophysiological Issues in Aphasia], ed. E.V.Shmidt and R.A. Tkachev. Moscow: Medgiz.

Riumina, N.A. 1982. [Free Classification as a Method of Studying the Semanticsof Concrete Words]. In [Experimental Studies in Psycholinguistics, ed. R.M.Frumkina and A.P. Vasilevich, pp. 2029. Moscow.

Rosch, E. 1975. Cognitive Representations of Semantic Categories. J. Experi-ment. Psychol. General, vol. 104, pp. 192233.

Rosch, E., and Mervis, C.B. 1975. Family Resemblances: Studies in theInternal Structure of Categories. Cognit. Psychol., vol. 7, pp. 573605.

Rosch, E.; Simpson, C.; and Miller, R.S. 1976. Structural Bases of TypicalityEffects. J. Experiment. Psychol: Human Perception and Performance, no.2, pp. 491502.

Tsvetkova, L.S. 1972. [The Process of Naming an Object and Its Impairment].Vopr. psikhol., no. 4, pp. 10717.

Tulviste, P. 1988. [The Cultural Historical Development of Verbal Thinking].Tallinn.

Tulviste, T., and Tulviste, P. 1985. [On the Correspondence Between the

128 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

Nature of Units and Operations in Verbal Thinking: An ExperimentalConfirmation of Vygotskys Hypothesis]. In [The Linguistic andPsycholinguistic Structure of Speech], ed. R.M. Frumkina, pp. 10915.Moscow.

Vygotsky, L.S. 1982. [Thought and Language]. In [Collected Works in SixVolumes], Vol. 2. Moscow: Pedagogika, pp. 5361.

Warrington, E.K. 1981a. Neuropsychological Evidence for Multiple MemorySystems. Acta neurol. scand., vol. 64, suppl. 89, pp. 1319.

. 1981b. Neuropsychological Studies of Verbal Semantic Systems.Philos Trans. Roy. Soc. London, vol. 295, pp. 41128.